Proposal (593) to South American Classification Committee

Revise classification of the genus Saltator

Effect on SACC: This proposal would (A) revise the linear sequence of species, and (B) merge the monotypic genus Saltatricula into Saltator.

Background and New information: Our current linear sequence is based largely on historical momentum and perceptions of relationships by various authors, including the study of phenotypic characters by Hellack and Schnell (1977). Chaves et al. (2013) produced a phylogeny of all species recognized in the genus using mtDNA (ND2, cyt-b; missing ND2 for S. maxillosus). A poor screen shot of their results is pasted in below, although the major tree is too large to have any resolution here; let me know if you need a pdf of the original.

These results largely support the traditional view of relationships with the following two exceptions: (a) S. nigriceps is not the sister to S. aurantiirostris + S. maxillosus but rather is the sister to those two plus S. grossus + S. fuliginosus; (b); S. striatipectus is part of the S. coerulescens group, as originally treated by Paynter and others; (c) and Saltatricula multicolor is the sister to Saltator atricollis. (They also found that inclusion of Middle American grandis in S. coerulescens makes that a paraphyletic species. I worry that this might be a gene tree/species tree problem, but Hilty [2003] already split grandis based on voice; regardless, this would requires a separate proposal.)

To incorporate these results into our classification requires a change in the linear sequence of species and transfer of monotypic Saltatricula into Saltator. I separate these into two sub-proposals.

A. Modify linear sequence of species.

Our current linear sequence is as follows:

Saltator grossus Slate-colored Grosbeak
Saltator fuliginosus Black-throated Grosbeak
Saltator maximus Buff-throated Saltator
Saltator atripennis Black-winged Saltator
Saltator coerulescens Grayish Saltator
Saltator similis Green-winged Saltator
Saltator maxillosus Thick-billed Saltator
Saltator orenocensis Orinocan Saltator
Saltator nigriceps Black-cowled Saltator
Saltator aurantiirostris Golden-billed Saltator
Saltator striatipectus Streaked Saltator
Saltator cinctus Masked Saltator
Saltator atricollis Black-throated Saltator
Saltatricula multicolor Many-colored Chaco Finch

As long as the sequence has to be modified to show the proper relationships of S. nigriceps and S. striatipectus, I think we should overhaul the sequence to reflect the tree produced by Chaves et al. using our conventions (least-diverse branch first, and so on; for sister species or allospecies in superspecies, northwestern-most listed first). The following is the proposed sequence – please check for better alternatives:

Saltator atricollis Black-throated Saltator
Saltator (Saltatricula) multicolor Many-colored Chaco Finch

Saltator maximus Buff-throated Saltator
Saltator atripennis Black-winged Saltator
Saltator orenocensis Orinocan Saltator
Saltator coerulescens Grayish Saltator
Saltator striatipectus Streaked Saltator
Saltator similis Green-winged Saltator
Saltator nigriceps Black-cowled Saltator
Saltator maxillosus Thick-billed Saltator
Saltator aurantiirostris Golden-billed Saltator
Saltator cinctus Masked Saltator
Saltator grossus Slate-colored Grosbeak
Saltator fuliginosus Black-throated Grosbeak

B. Merge Saltatricula into Saltator.

To maintain a monophyletic Saltator, there are two choices: (a) merge Saltatricula into it, or (b) transfer atricollis to Saltatricula. An earlier SACC proposal (344) to merge the two genera was voted down for various reasons, one of which was that atricollis and multicolor were very different from Saltator and another of which was that the genetic data for their sister relationship were substandard. With increased individual and taxon sampling, those genetic data appear to be stronger now, so I think this comes down to a matter of taste.

Analysis and Recommendation: The proposed new sequence reflects the findings of Chaves et al. (2013) and barring additional tweaks, removes previous misconceptions on relationships in the genus. As for the merger of Saltatricula, I lean slightly towards the merger at this point. Although I recognize that atricollis is an odd Saltator (see comments from Claramunt and Naka on prop. 344), I think these differences should be outlined in a short publication, with sonograms, or at least summarized in a proposal based on existing published information. Note also that the doubts concerning placement of atricollis and multicolor in Saltator raised by Naka and Claramunt in terms of the genetic data no longer apply. Also, Saltator is heterogeneous even if atricollis is extracted, so in the subjective realm of delimitation of boundaries of genera, I would like to see all this laid out in a full proposal. Finally, overturning of the Chavez et al. treatment of multicolor and atricollis would be better addressed in a separate proposal. If Part B is rejected, then that would force a follow-up proposal regardless.

Literature Cited:

CHAVES, J. C., J. R. HIDALGO, AND J. KLICKA. 2013. Biogeography and evolutionary history of the Neotropical genus Saltator (Aves: Thraupini). Journal of Biogeography 40: 2180–2190.

Van Remsen, October 2013

______________________________________________________________________________________________

Comments from Stiles: “I´m reserving my vote for the moment, in that the proposal (and the associated tree) bring up a point that may merit more discussion. On the one hand, the Chaves et al. study does clear up nicely the relationships in Saltator (although I don´t see S. rufiventris in their tree – has it been put somewhere else?). The sister-group status of multicolor and atricollis (NOT atriceps) seems reasonable, at least when eyeballing their portraits in Ridgely & Tudor’s songbird guide – atricollis looks rather like an oversized, washed-out multicolor – or if one prefers, the latter is a miniature, jazzed-up atricollis. However, the split between these two and the rest of Saltator is quite old, near the Miocene-Pliocene boundary – unless I´m very mistaken, most (?) generic splits among the nine-primaried oscines are considerably younger. If so, it might be at least as reasonable to consider the multicolor-atricollis clade as a separate genus (presumably Saltatricula, though I haven’t checked to see if any earlier names apply. This wouldn´t affect the linear arrangement of species in the proposal. Comments? Regarding the situation in S. coerulescens, I can attest to the vocal differences between at least all Colombian birds I have heard and the ones I knew well in Costa Rica – but I am bemused by the Venezuelan sample falling with the Middle American clade, not the South American clade – what´s going on here?

Comments from Zimmer: “I would vote YES on Sub-Proposal A, for accepting the revised sequence, with the modifications necessitated by a NO vote on Sub-Proposal B. I think we are better served by merging atricollis into Saltatricula, than by merging the latter into Saltator. I would agree with comments made by Claramunt and Naka on Proposal 344 regarding the ecological and vocal differences of atricollis from other saltators. As Van suggests, it really is a matter of taste, and my tastes run towards more narrowly defined, internally homogeneous genera.”

Comments from Pacheco: “ [YES on A and NO to B] Penso que há um avanço nas interrelações dos táxons tratados em Saltator a partir de Chaves et al., mas considero mais apropriado (a árvore permite) subordinar atricollis ao gênero Saltatricula que o contrário.”

Comments from Pérez-Emán: “YES to A. Updating linear sequence based on new and complete phylogenetic information (Chaves et al. (2013)) is necessary to have a classification that better reflects the current evolutionary knowledge of this group. However, many nodes in the so-called Clade C (coerulescens, striatipectus, similis, orenocensis) are not strongly supported, although it does not seem to affect the proposed linear sequence. Samples of coerulescens from Trinidad and Táchira (Venezuela) as well as samples of striatipectus from Zamora-Chinchipe (Ecuador) do not show clear relationships with others in the clade. As Gary indicated, it is perplexing and it would have been interesting to read some discussion in the article about the placement of these samples and which hypotheses are proposed to explain it. I would vote NO to B. This is another borderline proposal that could be voted either way. However, based on morphological, ecological and vocal similarities (as indicated by Claramunt and Naka) as well as the early split of the multicolor-atricollis clade, I would lean toward merging atricollis into Saltatricula.”

Comments from Cadena: “593A. YES. 593B. NO. I would favor placing atricollis in Saltatricula. The data have certainly improved in that there are more individuals sampled, but analyses are still based only on mtDNA, so there is no major change regarding the issues discussed under proposal 344 years ago. I voted to include multicolor in Saltator back then and later changed my mind based on the comments by Álvaro, Naka and Claramunt about atricollis being a very different "Saltator", which I still find valid. Considering we need to make a change anyway, I think our classification would be more informative of relationships if it were to convey that atricollis and multicolor are sister species by placing them in the same genus separate from Saltator.”

Comments from Jaramillo: “A. Revision of linear order as noted. YES.

B. NO. Given the genetic separation of multicolor and atricollis, from true Saltator I think it is equally valid to propose that atricollis be moved to Saltatricula. Saltator is a diverse looking genus, still multicolor is quite different not only in plumage but overall structure, and so is atricollis to a lesser extent. They also have an interesting biogeographic element that is worth considering. While true Saltator are birds of forest or more specifically forest edge; atricollis and multicolor are birds of the non-forest belt of South America, one in Chaco and Monte, the other typically in Cerrado habitats. I think that creating a two species Saltatricula is useful as it restricts Saltator to a more limited (although not nearly uniform) group, adding more information value to that genus. The time of division between Saltatricula and Saltator is distant, and that alone clarifies that such a broad Saltator is not the way to deal with this group.”