Proposal (344) to South American Classification Committee
Merge Saltatricula into Saltator
Effect on SACC: This would merge a monotypic genus into an existing large one, Saltator.
Background: This proposal is triggered by the passage of Proposal to place Saltatricula next to Saltator. Now we need to decide whether the evidence is sufficient for merger of Saltatricula into Saltator. Without repeating all of Proposal 322, I note only the following critical sentence:
"The critical node (in their Fig. 1) for that placement [of Saltatricula in Saltator] has strong support (> 95% Bayesian), as does the node that supports a sister relationship to Saltator atricollis."
Therefore, not one but two strongly supported nodes place Saltatricula within our broadly defined genus Saltator. Six of six SACC voters who commented on 321 were already in favor of the merger. As Doug noted, the only question is whether to follow the Klicka et al. placement in the linear sequence (sister to atricollis) or put it at the end of the linear sequence. Given that the support for the sister relationship, I recommend not only the merger placing it to follow atricollis -- this represent the best estimate of its relationships given published data so far.
Van Remsen, May 2007 (in consultation with John Klicka and Kevin Burns)
Comments from Cadena: "YES for reasons stated in proposal 322."
Comments from Stiles: "YES. On both - merging of Saltatricula into Saltator and placing it next to S. atricollis - the genetic data are convincing; the nodes are well supported."
Comments from Pacheco: "YES. Vejo como uma boa solução acomodadora."
Comments from Jaramillo: "NO - I was a bit shocked at the results mainly because I was convinced that Saltator atricollis was in fact allied to Embernagra/Emberizoides. Upon second thought though it makes sense to me based on field experience with these birds that Saltatricula is close to S. atricollis. That I am comfortable with now, and now feel like I was misleading myself with Embernagra/S. atricollis. What spurred my interest in this creature was that the first times I saw S. atricollis what went through my head was "that ain't no Saltator." It was partly the shape, the song, the coloration … various bits of information you get when you see a bird in life. I think that it makes much more sense to keep Saltatricula in its own genus and add in S. atricollis there. Why have such a large and heterogeneous Saltator? I have not looked at the nomenclature and if Saltatricula is the valid option or if atricollis belonged to an earlier named genus, but the two should be sisters in their own genus, not in Saltator."
Comments from Santiago Claramunt and Luciano Naka: "We hope we can persuade the committee members to vote NO on this proposal. The statement that Saltatricula is embedded into Saltator contradicts our perception based on marked differences in size, habitat, and song between Saltatricula and typical Saltator. However, according to Klicka et al.'s tree, Saltatricula is not deeply embedded into Saltator; it is only the position of Saltator atricollis that makes Saltator paraphyletic. In the experience of one of us (L.N.), S. atricollis is an atypical Saltator, and its habitat (a Cerrado specialist) and particularly its song resemble more those of Saltatricula (a dry Chaco inhabitant) than any other Saltator. Another source of uncertainty is the fact that Bayesian probabilities associated with branches were not reported. A posterior probability of .99 or 1.00 indicates strong support, but a probability of 0.95 is marginal at best (it may be equivalent to a 65-75% bootstrap).
"Concerned about these ambiguities, we downloaded a subset of Klicka et al. (2007) data, and we discovered that the signal for the merging of Saltatricula into Saltator is not strong. First, the mitochondrial sequence signal is not strong for basal relationships mainly because Transitions at third codon positions reach levels of saturation even within the genus Saltator. Probably due to this saturation, basal relationships changed dramatically in our re-analyses when using different models of substitution or different tree building algorithms, and not one of the basal branches had a bootstrap support value over 50%. However, in all these re-analyses we recovered a sister relationship between Saltatricula and S. atricollis with relatively strong support (bootstrap 80-100%) but in most cases this clade was not sister to the other Saltator.
"In conclusion, the mitochondrial sequence data is suggesting that Saltator is not monophyletic not because Saltatricula is embedded in Saltator, but because atricollis may not be a Saltator. For the moment, the most conservative options are: to maintain the status quo and maintain Saltator paraphyletic until strong evidence is obtained); or transfer atricollis to Saltatricula (which although may sound a drastic measure, is not more drastic than merging Saltatricula into Saltator)."
Additional comments from Remsen: "Claramunt and Naka's re-analysis, along with Alvaro's comments, has convinced me to change my vote to NO. A proposal to treat S. atricollis in = Saltatricula should be the next step; I would regard that as the most conservative option in terms of maximizing chances for having monophyletic genera."
Additional comments from Stiles: "In view of these comments, I am also willing to vote NO, and agree that the best course might be to remove atricollis from Saltator and merge it into Saltatricula (an examination of synonymies in Bds. British Museum and Hellmayr indicates that atricollis has been placed in Tanagra, Loxia and Fringilla as well as Saltator and as none of these are appropriate now, so there appears to be no nomenclatural impediment to this move)- this would make Saltator monophyletic if I read things right, so perhaps a fresh proposal is in order?"
Additional comments from Cadena: "The points by Claramunt and Naka are very good ones, and I am happy to change my vote to NO based on their comments. I would definitively support transferring atricollis to Saltatricula as an alternative that seems entirely consistent with the strongly supported nodes in the phylogeny, and that preserves a good deal of phylogenetic information in the classification."