Proposal (879) to South American Classification Committee

Treat Saltator coerulescens as two or three species

Effect on SACC list: Saltator coerulescens would be split into three species.

Background: Paynter (1970) treated Saltator coerulescens (with 13 different subspecies) as a single species ranging from Mexico to Uruguay. In his Venezuela guide, Hilty (2003) included a taxonomic note indicating that the Middle American grandis subspecies group may be a separate species from the nominate South American Saltator coerulescens group (a return to the classification of W. Deppe 1830). Hilty (2003) also mentioned that vocal differences within South America may indicate additional species. Similar suggestions were also made elsewhere (Ridgely & Tudor 2009, del Hoyo et al. 2011). In the absence of any further study, all modern taxonomies however continued to treat Grayish Saltator as a single species (until recently).

New information:

Genetic data: Chaves et al. (2013) presented a phylogeny of the genus Saltator. All presently recognized species were found as monophyletic groups, with the exception of S. coerulescens.

The Amazonian group (coerulescens group) was found sister to Streaked Saltator S. striatipectus (making the present broadly defined S. coerulescens paraphyletic), whereas the Middle American grandis group was found sister to the Caribbean group of northern South-America (the olivascens group). Divergence times were estimated to be >3 million years in both cases.

Chaves et al. (2013) pointed out that taxonomically grouping the streaked vs non-streaked taxa is in fact contradictory to his genetic findings, and recommended additional research to better understand this apparent anomaly (while putting forward as possible hypothesis that a parallel evolution leading twice to a streaked plumage may be explained by a paedomorphic condition).

Vocal data: Boesman (2016) made a brief vocal analysis (without having learnt about the findings of Chaves et al.) and found three clear vocal groups:

· The northern (or Middle America) group (including vigorsii, plumbiceps, grandis, yucatanensis, hesperis and brevicaudus)

· The Caribbean group (including plumbeus, brewsteri and olivascens)

· The Southern (or Amazonian) group (including azarae, mutus, superciliaris and coerulescens)

The Caribbean group differs from both other groups by the lack of long, slurred notes, lack of a second song type, slower pace of stuttered song, etc. (The fact that this group lacks a whistled song is possibly an evolutionary adaptation to differ from the largely sympatric Streaked Saltator S. striatipectus). The Amazonian group differs from the Northern group based on a stuttered song with repeated notes (# of repeats) and a whistled song with fewer notes and upslurred ending.

These findings are very much congruent with Chaves’ findings of three groups (including also the fact that the Amazonian group seems vocally closer to S. striaticeps).

Morphological data: del Hoyo & Collar (2016) analyzed morphological differences, and concluded these were small. The northern group typically has a longer white eyebrow and more rufous-brown belly, and the Caribbean group has a more whitish central belly.

Discussion:

Although in the past two allopatric species have been suggested, it would seem that in fact rather three clear groups are involved (a split of Middle American vs. South American taxa would not amend paraphyly nor accommodate vocal differences, and thus despite earlier suggestions is not recommendable).

The Middle American group is allopatric, but the case of the two South American groups is more intriguing. These seem to meet both along the lower east slopes of the east Andes, and north of the Amazon delta. In both regions they are likely parapatric, but this requires further study (a situation identical to e.g. C. cyanoides vs. C. rothschildii along the Andes, where exact boundaries and possible interaction also still need to be uncovered). In both contact zones, there seems to be a clear-cut (and identical) change in voice, which eliminates the possibility of some type of ‘ring species’ based on voice.

Since the Boesman (2016) analysis, a few additional sound recordings have been deposited on-line from the contact zones, further confirming the sharp vocal transition along the Andes: ML59164651 is just north of the rio Meta (at a distance of c 30km from XC327455 !), is of the ‘Caribbean group’ and further indicates parapatry. (No new recordings from the eastern contact zone).

Learnt voice in oscine passerines calls for some caution, but it should be noted that in the genus Saltator, several other clear-cut cases are based on vocal differences between related species pairs (and confirmed by genetics) for which prior classifications based on morphology were not always in accordance (e.g. S. nigriceps vs S. aurantiirostris; Boesman 2016b).

Genetically, calculated time of divergence of the three groups was comparable to the widely accepted species pairs S. grossus vs S. fuliginosus, S. atripennis vs S atriceps or S. aurantiirostris vs. S. maxillosus (all pairwise sister species). A weakness is that the Caribbean group was only analyzed by two (admittedly widely separated) samples and that Bayesian PP<0.75.

Ideally, to make this case more robust, besides more extensive genetic sampling, play-back experiments could be added (although e.g. playing the whistled song of Amazonian group to Caribbean group is in fact about the same as playing song of the sympatric Streaked Saltator, with predictable result).

Furthermore, study of the situation in the contact zones of the 2 South-American groups would allow for a better assessment of interactions between both groups. The fact that such potential contact exists (twice) without any indication of clinal variation at the other hand is a strong argument absent when dealing with allopatric populations.

It would thus seem that the following viable taxonomic options exist:

· Retain the present treatment while awaiting more research, accepting paraphyly and highly divergent vocal groups within a single species

· Split the southern group, thus creating two monophyletic groups, but still having a (northern/Caribbean) species with two very distinct vocal groups

· Split into three species, all monophyletic and with distinct voice

Southern Grayish Saltator is the English common name given by Hilty (2003) presumably for all taxa in South America (but confusingly he described the range from Mexico to Uruguay), and by deduction the Middle American group may be called Northern Grayish Saltator. Del Hoyo & Collar (2016) recognized three species, and named them Northern Grey Saltator, Caribbean Grey Saltator, and Amazonian Grey Saltator. By keeping Grey (or Gray) in the name, the link with the former name Grayish is retained. ‘Caribbean’ is not very precise for a bird with a range well inland, but Caribbean Hornero, for example, has been used as well elsewhere, and the name ‘Northern’ is not an option here, with ‘Guianan’ not correct either etc. ‘Amazonian’ is also somewhat ‘stretched’ given the range of that group reaches the Chaco of Argentina. ‘Southern’ is an alternative.

Proposal:

A. Split S. coerulescens into two monophyletic species: S. grandis (including also vigorsii, plumbiceps, yucatanensis, hesperis, brevicaudus, plumbeus, brewsteri and olivascens) and Amazonian S. coerulescens (including also azarae, mutus and superciliaris)

B. If A is accepted, split S. grandis into two species: Middle American S. grandis (including also vigorsii, plumbiceps, yucatanensis, hesperis, brevicaudus) and Caribbean S. olivascens (including also plumbeus and brewsteri)

C. Give English common names respectively as Northern Gray Saltator, Caribbean Gray Saltator (if B is accepted) and Southern Gray Saltator (if NO, please provide alternative)

Literature:

Boesman, P. (2016). Notes on the vocalizations of Greyish Saltator (Saltator coerulescens). HBW Alive Ornithological Note 395. In: Handbook of the Birds of the World Alive. Lynx Edicions, Barcelona. https://doi.org/10.2173/bow-on.100395 https://static.birdsoftheworld.org/on395_greyish_saltator.pdf

Boesman, P. (2016b). Notes on the vocalizations of Black-cowled Saltator (Saltator nigriceps). HBW Alive Ornithological Note 440. In: Handbook of the Birds of the World Alive. Lynx Edicions, Barcelona. https://doi.org/10.2173/bow-on.100440 https://static.birdsoftheworld.org/on440_black-cowled_saltator.pdf

Chaves, J.A., Hidalgo, J.R. and Klicka, J. (2013). Biogeography and evolutionary history of the Neotropical genus Saltator (Aves: Thraupini). Journal of Biogeography. 40(11): 2180–2190.

Del Hoyo, J., Elliott, A. and Christie, D. (2011). Handbook of the Birds of the World Vol. 16. Lynx Edicions. Barcelona.

Del Hoyo, J. & Collar, N. (2016). Illustrated checklist of the Birds of the World. Lynx Edicions. Barcelona.

Hilty, S.L. (2003). Birds of Venezuela. Christopher Helm, London

Ridgely, R.S. and Tudor, G. (2009). Field Guide to the Songbirds of South America: The Passerines. University of Texas Press, Austin, Texas.

Paynter, R. A., JR. (1970). Subfamily Cardinalinae. Pp. 216-245 in "Check-list of birds of the World, Vol. 13" (R. A. Paynter Jr., ed.). Museum of Comparative Zoology, Cambridge, Massachusetts

Peter Boesman, August 2020

Comments from Areta:

“A. YES. The genetic differentiation, with coerulescens as sister to the vocally and morphologically different striatipectus support the split.

“B. YES. The depth of the split between olivascens and grandis is similar to that between coerulescens and striatipectus, and although they are less divergent in plumage, their vocalizations and amount of genetic differentiation argue in favor of this split as well.

“C. NO. Unfortunately, I am not allowed to vote. These compound names are awful and should be avoided, while the imply a relationship that does not exist. I would go with something along the lines of:

Cinnamon-bellied Saltator --- S. grandis

Drab-bellied Saltator --- S. olivascens

Grayish Saltator --- S. coerulescens

“I don´t see any need to change the name of the nominate taxon, which is also the most widespread.”

Comments from Stiles: “A: YES to splitting grandis from coerulescens: genetics, distributions and vocalizations provide strong support. B: YES to the further split of olivascens from coerulescens: genetics, vocalizations and considerable evidence for parapatry, albeit with only a small difference in plumage seem sufficient for this split, which also resolves the paraphyly of coerulescens. C. E-names: I like Nacho’s suggestion for these.”

Comments from Robbins: “Given the genetic and vocal data, YES, to both A & B for recognizing grandis and olivascens as species.”

Comments on English names from Josh Beck: “I strongly agree that proposed names are misleading with respect to relationships that don't exist. I like Nacho's novel name of Cinnamon-bellied for grandis, but I would like to suggest that Caribbean Saltator might be a better name than Drab-bellied for olivascens. The name Caribbean Gray(ish) Saltator is already in use in a few places (BirdLife, EcoRegistros, and as the group name for olivascens in Clements/EBird/BotW). Drab-bellied is not really all that informative/unique in a genus full of fairly drab bellied birds. But mostly, it seems a more useful name in that it ties the species to its distribution. To me it reminds me of the names "Rio Suno, Rio de Janeiro, and Yungas Antwrens" vs "Gray, Leaden, and Slaty." There's nothing really wrong with the colors for names, but the three geographically inspired names tell you a lot more right off the bat. One could follow that logic to suggest that Guianan Saltator might be an even better name based on the range of olivascens, but with precedent for Caribbean Gray Saltator, it seems Caribbean Saltator is a less disruptive choice than Guianan.”

Comments from Zimmer: “A. YES to splitting Middle American + Northern SA subspecies in this complex from “Amazonian” or southern coerulescens (including azarae, mutus and superciliaris, based upon genetics, voice, and morphology as outlined in the Proposal. (B) YES to further splitting Caribbean S. olivascens (including plumbeus and brewsteri) from the Middle American S. grandis group (including vigorsii, plumbiceps, yucatanensis, hesperis, and brevicaudus). (C) NO to the proposed compound English names of “Northern Grayish Saltator”, “Caribbean Grayish Saltator” and “Southern Grayish Saltator”. The compound names, although somewhat bland, are descriptive enough of the distributions of the three groups, but my objection lies in the genetics, which tell us that the southern coerulescens-group is sister to Streaked Saltator, making the broader coerulescens-group currently defining Grayish Saltator paraphyletic. Thus, it would seem inappropriate to use “Grayish Saltator” as a group name, when the implied group relationship doesn’t exist. Normally, I would want to steer clear of retaining the parental English name for one of the “daughters” in a 3-way split, but given that, in this case, the nominate taxon is the most widespread, and, technically, the other two species resulting from the split are not true daughters, perhaps retention of “Grayish Saltator” for post-split coerulescens is the best course, as Nacho suggests. I also rather like Nacho’s suggestion of “Cinnamon-bellied Saltator” for the grandis-group. For the olivascens-group, I would suggest either Olive-gray Saltator or Gray-olive Saltator, which is perhaps more descriptive than “Drab-bellied”, would square nicely with the species epithet, and would also provide a link to its former status as part of “Grayish” Saltator.”

Comments from Bonaccorso: “A. NO. From the Chaves et al. (2013) tree, it is not completely clear that the Caribbean clade is most closely related to the Middle American clade; they probably are, but there is no support for the node that unites those two clades. Also, their vocalizations are different enough to suggest they are both good biological species.

“B. YES. Given the phylogenetic uncertainty about the relationship between the Caribbean and the Middle American clade, it makes more sense to recognize three species of Saltator (S. grandis, S. olivascens, and S. coerulescens). All three species are easily differentiated by song, make sense geographically. The relationships between the Caribbean clade and other clades in Saltator will not change that.”

Comments from Jaramillo: “A. YES. B. YES. Vocal and genetic data suggests this is the way to go. I am impressed that the vocal change is clear at the contact zones, although I realize data is minimal still at the contact zones. C – NO.

“I would not keep Grayish Saltator as that will really cause confusion over time in this case. How about just shortening to Gray Saltator? Remember names do not need to be perfect. Cerulean is another option, but perhaps that is too much of a stretch. I like Cinnamon-bellied. However, I think it is a big stretch to use Caribbean for olivascens as someone who is novel to this situation will not be looking for it in Colombia and Venezuela, but in Cuba and Puerto Rico if you use Caribbean. Again, with leniency to names that are not perfect but at the same not entirely incorrect why not Olivaceous Saltator? Given the region, and Simon Bolivar’s wish for a “Gran Colombia” one could propose Gran Colombian Saltator, just throwing spaghetti at the wall here to see what sticks. Having some fun. Suggestions in the quest of good, memorable and simple names:

Cinnamon-bellied Saltator --- S. grandis

Olivaceous Saltator --- S. olivascens

Gray Saltator --- S. coerulescens

Comments from Lane: “Comments by Lane: A) YES. B) YES. C) NO to the proposed names by the proposal. I am inclined to use the names suggested by Alvaro, but perhaps modify his coerulescens name to "Blue-gray Saltator" to better distinguish it from "Grayish" and to better square with the blue implication of the scientific name.”

Comments from Pacheco: “YES, for A and B. Considering the genetic and vocal data, I vote to recognize both grandis and olivascens as full species.”

Comments from Claramunt: “NO. Given the morphological similarity, I think we should be more cautious. The relevant nodes in the ND2 tree are not strongly supported, so I think that the evidence for the paraphyly of coerulescens is not strong. Note also that there are no samples from nowhere near a putative contact zone in Colombia. The closest samples are in Ecuador and Venezuela. I would like to see some nuclear data and/or samples from Colombia, and/or a more rigorous analysis of the match between mtDNA and songs.”

Comments from Remsen: “A and B. NO. Declaring paraphyly based on an ND2 gene tree is just not acceptable, especially when the sampling is haphazard and not designed to detect gene flow near contact zones. Even so, as pointed out by Elisa and Santiago, support for the important nodes is not strong. This is a nice preliminary study that should set the stage for a more rigorous project that focuses on contact zones. As for the vocal data, they are indeed suggestive in that they map onto the genetic data fairly well. But again, I’d like to see a full-blown analysis with samples from near contact zones as well as playback trials. Peter has set the stage for a more formal analysis. I am tempted to vote YES just because I think the proposed species limits are probably a better match for the limited published data than treating them all as conspecific. But this group is complex and deserves a well-designed study with better genetic data, samples from the contact zones, and playback experiments, and comparisons including striatipectus and albicollis.

C. NO. The compound names (with or without hyphens) imply a monophyletic group, when in fact their potential non-monophyly is the basis for part of the proposal. As for some of the suggestions, “Grayish” is DOA. The nominate form may have a larger distribution, but even S. grandis sensu stricto has a large distribution that includes basically all Middle American countries (and likely therefore has more citations in literature). We need a separate proposal on English names if the taxonomic proposal passes.”

Comments from Oscar Johnson solicited by Remsen: “You're definitely right that this is just a gene tree, so it should clearly be interpreted with caution. Additional data from next-gen nuclear data will likely change some of the results, but looking at some recent studies that have included both mtDNA and next-gen nuclear data for the same samples, the mtDNA often gets quite close to the latter. So, I do think that mtDNA provide a reasonable estimate of relationships (with caveats) and are a good first pass at a phylogeny that can be useful in taxonomic decisions. MtDNA will give conflicting signals (vs next-gen) in cases of recent introgression (haploid, matrilineal, etc) or with incomplete lineage sorting (gene tree/species tree), but it does seem to do a good job of clustering populations and estimating relative divergence times. Having many fewer base pairs to work with does usually lead to lower support across the phylogeny, too. For examples of the similarities between mtDNA and next-gen phylogenies, check out the recent Aphelocoma papers from the McCormack lab, Ethan Linck's WEFL paper (buried in supplemental), or my Epinecrophylla paper. 

“For this Saltator paper, I would trust, for example, that there are two deeply diverged clades in coerulescens. The paraphyly of coerulescens could be due to gene tree/species tree conflict or insufficient signal in the mtDNA, but there is clearly a deep genetic separation across the Andes that is indicative of two species. I wouldn't be surprised if a study of nuclear data found coerulescens to be monophyletic with a deep split across the Andes. In that larger clade, I would say you've got four divergent genetic groups (each reasonably called species), but the relationships between them are poorly resolved (note low statistical support for many branches): two coerulescens with a split across the Andes, striatipectus, and albicollis/similis. Albicollis and similis are clearly very closely related (which is quite amazing, in my opinion), despite the geographic separation and clear morphological differences. 

“The situation with maximus is weird. there are clearly two deeply diverged clades suggestive of species, but the western Ecuador populations cluster with the Amazonian birds. I would be hesitant to split maximus, despite the deep divergence, as that result suggests that something interesting is going on in Colombia (unsampled), either with genetically intermediate populations that would "fill in" that deep divergence or with secondary contact somewhere between the closest samples in the study (western Ecuador and western Panama). By "filling in" the relationships for maximus, I mean that there could be additional deeply diverged clades in Colombia. It wouldn't change the fact that there is a deep divergence within the species. I think you would just need to get more information on where exactly those populations come back into contact.”

“For some of the other groups: 

“S. grossus and S. fuliginosus are clearly very closely related, or hybridized relatively recently. Maybe same species? More data probably needed.

“There are really not enough samples or statistical support to make any inferences about maxillosus and aurantiirostris, especially given reports of hybridization between these taxa. That's a situation that would need lots of nuclear data.”

“Regarding plumage/vocalizations, yes, hopefully those line up with the major genetic breaks, which would provide additional evidence for species status.”

Comments from Stiles: “Splitting up Saltator coerulescens. Three groups had both genetic and vocal support (again, thanks to Peter Boesman): call them M (Middle American grandis and sspp.), C (the ± Caribbean olivaceus and sspp.) and A (Amazonian coerulescens). Maintaining all in a single species is not an option because this leaves the species (Grayish Saltator) parapatric. Additional data: M is isolated geographically; probable parapatry exists between C and A). The proposal to split M from C+A passed 7-3; the proposal to split the two South American clades C and A passed 8-2. The problems derive from the derivation of E-names for the three groups. I’d suggest presenting the options here as two slates: geography, with the names Middle American (or Northern), Caribbean and Amazonian) vs. color differences: Cinnamon-bellied for grandis, Drab-bellied or Olive-gray (or Olivaceous?) for olivaceus, and Grayish (retaining the name for the original species), Blue-gray or Leaden (another option, perhaps more accurate and eliminating “gray” from the name) for coerulescens. First, vote for the slates: geography vs. color – if one of these gets a clear majority, then present the options for each of the species, perhaps numbering the options in order of preference for each (where two or more options have been suggested).”

Comments solicited from Don Roberson: “After reading the material, I quite like a combo of Alvaro's names plus Dan Lane's modification of the nominate, and off the cuff I think the following list is fine: short names, generally aimed at i.d. points, and a new name for the nominate that includes the concept of "gray/grey".

Cinnamon-bellied Saltator --- S. grandis
Olivaceous Saltator --- S. olivascens
Blue-gray Saltator --- S. coerulescens”