Proposal (351) to South American Classification Committee


Transfer Saltator and Saltatricula from Incertae Sedis to Thraupidae


Effect on SACC: This would transfer two genera from Incertae Sedis to Thraupidae.


Background We recently moved Saltator from Cardinalidae (Proposal 321) and Saltatricula from Emberizidae (Proposal 322), placing them in Incertae Sedis in adjacent positions. Passage of that proposal triggers a follow-up to move them now into Thraupidae. There is a concurrent proposal (344) to merge Saltatricula into Saltator, but that outcome does not affect this proposal.


Klicka et al.'s (2007) analysis included 102 genera of tanagers, emberizines, and cardinalines. The genetic sampling consisted of 2281 bp of two mitochondrial genes, ND2 and cyt-b ... a nice sample. Not only is Saltator not in the Cardinalidae, but there is support for placement within the Thraupidae. The critical node (#2 in their Fig. 1) for that placement has strong support (> 95% Bayesian). That node places Saltator plus Saltatricula as sister to the rest of the tanagers.  Saltator rufiventris is not a Saltator but is deeply embedded within the Thraupidae (to be addressed in another proposal).


Analysis: MtDNA is widely considered a reliable predictor of phylogeny at these levels of taxonomy, and certainly these data sets represent the first truly scientific estimates of the phylogeny and classification of this group.


The only question is whether to place them within Thraupidae or leave them as Incertae Sedis. Given that their placement in Thraupidae is based on one node and one study (and no nuclear DNA analyzed so far), and given that Burns, Klicka et al. will undoubtedly be publishing more on Thraupidae and Saltator, I previously suggested a cautious approach by placing them temporarily as Incertae Sedis, including even S. rufiventris, with all appropriate footnotes indicating their likely relationship to Thraupidae. However, SACC comments received on #321 were strongly in favor of a direct transfer to Thraupidae, thus catalyzing this proposal.


However, here's John Klicka's response:


"If Saltator is to be placed within one of the existing nine-primaried oscine families, it will almost certainly have to be placed among the Thraupidae. I think that it will end up as a basal lineage within that group, but I am a bit hesitant to move it formally based only on the data presented in our recent paper. It probably wouldn't hurt to wait and see how the nuclear data affects this placement; work that Kevin should have done fairly soon. "


Therefore, I recommend a continued cautious approach and wait for the forthcoming results by voting NO.




KLICKA, J., K. BURNS, AND G. M. SPELLMAN. 2007. Defining a monophyletic Cardinalini: A molecular perspective. Molecular Phylogenetics and Evolution 45: 1014-1032.

[See SACC Literature Cited for others]


Van Remsen, May 2008





Comments from Cadena: "YES for reasons stated in proposal 321. Even after John's comment, I am still comfortable with a YES vote. Sure, nuclear data would be nice, but the mtDNA results are solid (see my comments on proposal 321), and I think we should go with the best available information."


Comments from Stiles: "YES. I had already indicated that I favored this move as the evidence (even though from only one gene, one study) seemed exceptionally clean-cut. If the nuclear data continue to indicate a very deep split between Saltator and the tanagers "proper", I would recommend that Kevin formally propose subfamily rank for the saltators. For the moment, I agree with Daniel that the best available evidence is to place the saltators in Thraupidae."


Comments from Stotz: "YES. Once we took Saltator and Saltatricula out of Cardinalidae, moving them to Thraupidae is the obvious placement for them."


Comments from Santiago Claramunt: "I agree with John in that it is better to wait for stronger evidence. The alternative hypothesis of a sister relationship between Saltator and other Cardinalidae could not be rejected (P = 0.245), and the branch uniting Saltator and thraupids is very short, and we don't know the exact posterior probability associated. According to some simulations, a posterior probability of 0.95 may be equivalent to a bootstrap value as low as 60%.


"Superficially, Saltator fits naturally into Cardinalidae based on external morphology, voice, and habits, although these "characters" may represent convergence to a forest large-seed-eater syndrome. Genetic similarity based on protein electrophoresis data (Tamplin et al. 1993) as well as overall morphological similarity (Hellack & Schnell 1977) does not show Saltator to be distinct from other cardinalids. Tordoff (1954a) mentioned that some authors considered Saltator "a thick-billed tanager". He noted that Saltator's palatal structure is closer to Piranga than to Cardinalis. Given that Piranga is a cardinalid, this is not evidence of a closer relationship to true thraupids. I need to check those other authors to se if there is any evidence of such relationship at all.


"In conclusion, I think the committee has to be conservative until stronger evidence is presented. I think it was premature to remove Saltator from Cardinalidae (proposal #321)."


Comments from Jaramillo: "YES - This is where the data is pointing. If additional data causes a further change in placement, I would prefer to adjust at that point rather than leave these genera in Incertae Sedis for a longer period of time. I guess we are getting to this question in an upcoming proposal, but in general I would rather avoid use I.S."


Comments from Robbins: "NO. Given John Klicka's comments, I vote "no" for now, and look forward to seeing their nuclear data set."


Comments from Zimmer: "NO. I don't think it will hurt to wait for the nuclear data set, especially in light of Santiago's comments."


Comments from Schulenberg: "Yes. I'm with Daniel, Gary, Doug, and Alvaro on this. The evidence points (strongly) in this direction. And while there may be times when the use of "incertae sedis" can not be avoided, when we really don't know anything other than that "taxon x isn't at all what we thought it was," I also don't like to see "incertae sedis" relied on when we have at least some idea what the relationships of "taxon x" may be."


Additional comments from Stiles: "The comments by Santiago are not wholly convincing because the evidence he cites may not be solid by modern criteria. For instance, palate types (Tordoff): the seed-cracking habit of typical cardinalines would make for strong selection on the form of their palates. Neither Piranga nor Saltator are specialized seed-crackers - both are mainly frugivorous to omnivorous (both take many insects, Saltator also eats buds and young leaves); hence a closer resemblance to Piranga than to Cardinalis in their palates is hardly surprising. The key question, not addressed by Tordoff, is how easily it would have been for a Saltator-type palate to evolve from a typical tanager-type palate in association with taking larger or tougher-husked fruits. Protein electrophoresis has a very mixed record in elucidating taxonomic relationships, as Sibley was forced to conclude after investing many years on this method: he stated that interpreting electrophoretic peaks was "like reading tea leaves". The record of DNA-DNA hybridization is also pretty mixed, with results of some studies being well supported by DNA sequence data and some not. Overall morphological similarity may simply reflect the fact that both saltators and typical cardinalines are both rather large and robust compared to most tanagers."


Comments from Pacheco: "NO. A partir dos comentários, sobretudo, do John e do Santiago, considero mais apropriado votar "Não" e aguardar por novos dados."