Proposals (220) to South American Classification Committee

Merge Nesomimus into Mimus

Background: The mockingbirds of the Galapagos Islands have traditionally been placed in their own genus, Nesomimus, based on plumage and morphology, although the rationale for recognition of Nesomimus has been questioned by some authors (see references in Arbogast et al. 2006).

New information: Arbogast et al. (2006) sequenced > 2650 bp of 4 mtDNA genes from Nesomimus populations from 11 islands as well as all members of the genus Mimus. Their molecular phylogeny demonstrates that Nesomimus is monophyletic. Their phylogeny suggests that it is embedded in Mimus, and that its closest relatives are, somewhat surprisingly in terms of plumage, probably the northern group (M. gundlachii, M. polyglottos, M. gilvus). [Of biogeographic interest is that several Galapagos taxa also have their closest relatives in the West Indies, and the northern Mimus clearly have the better track record as dispersers and colonists than the South American group.] [Also of interest is a problem in species limits within the Galapagos, but that's a separate proposal.]

Analysis: Although Arbogast et al. (2006) word their findings in a way that suggests that there is no doubt that Nesomimus is embedded in Mimus, a closer look shows that the support for the critical nodes is below the thresholds usually required for treating the findings as beyond a reasonable doubt. Specifically, the support level for the node that links Nesomimus with northern Mimus is only 49% (bootstrap probability). As my colleague Robb Brumfield noted, it would be interesting to do a Shimodaira-Hasegawa test to see if there's a significant difference between the trees they present and a tree in which Mimus is constrained to be monophyletic.

Nonetheless, I'm just quibbling. I suspect that additional sampling will elevate the support for the critical nodes. It seems not only sensible but also inevitable that Nesomimus is embedded in Mimus, i.e., derived from extant species or ancestral species within Mimus rather than from a shared common ancestor outside Mimus. In other words, it seems exceptionally unlikely that Nesomimus was derived from the shared ancestor of all extant Mimus, and even if it were, ranking it as a separate genus would boil down to a matter of taste. Even if Nesomimus is not embedded within extant Mimus, it is more closely related to Mimus than to the sister taxon to all mockingbirds, Oreoscoptes; that node IS strongly supported. Therefore, I would support the merger of Nesomimus into Mimus.

Recommendation: YES. Our current sequence does not reflect phylogenetic hypotheses. Although linear sequences are a poor way to portray relationships, for consistency we should make those sequences as consistent as possible with available data, and those data suggest that Nesomimus is embedded within or, at worst, the sister group to, all Mimus. To address the uncertainty for the support for the Nesomimus-northern Mimus sister relationship, I suggest we move the Galapagos species to an intermediate position between northern and southern Mimus, with a Note to explain our caution. Thus, our sequence would place the Galapagos species between M. gilvus and M. thenca in the current sequence.

Literature Cited (see SACC Biblio for the rest):

ARBOGAST, B., S. V. DROVETSKI, R. L. CURRY, P. T. BOAG, G. SEUTIN, P. R. GRANT, B. R. GRANT, AND D. J. ANDERSON. 2006. The origin and diversification of Galapagos mockingbirds. Evolution 60: 370-382.

V. Remsen, June 2006

Note added 8/22/07: See also Lovette & Rubenstein (2007), although as Daniel pointed out to me, they used mostly previously sequenced material: (p. 1035: "Some or all DNA sequences from 16 Mimidae taxa ... were derived from previous studies [Hunt et al. 2001, Barber et al. 2004, Arbogast et al. 2006])"

_______________________________________________________________________________________________

Comments from Jaramillo: "YES - Data are solid, and in life Nesomimus is not that different than let's say a Bahama Mockingbird, so this is not surprising. Van's point about how sensible it is that the Galapagos taxa would be expected to be imbedded in the genus, rather than an offshoot from a nearest relative of the mainland genus is a good one. This logic can be applied to many island taxa; the interesting situation is when the relationships do not fit this expectation (Caribbean trogons, todies etc.)."

Comments from Stiles: "YES. This arrangement might not be as surprising as it seems given the probable tectonic history of the Caribbean plate, which as I understand it moved from west to east through the gap between North and South America well before the final closure of this gap with the isthmus of Panama, producing and pushing island arcs ahead of it as it moved. Thus, I don't find it so terribly surprising that a Caribbean genus like Spindalis turns out to be sister to the rest of the extant tanagers (Burns?): it probably represents an old lineage present at the start of the differentiation of the tanagers that hopped aboard such an island arc as it passed through, remained isolated and suffered relatively limited differentiation in the Antilles thereafter, while the rest of the lineage either went extinct or differentiated much more explosively on the mainland(s) what with orogenies, sea level changes, glaciations and climatic changes, etc. not to mention the wealth of interspecific interactions (or alternatively, it evolved on an early island arc and hopped off onto the mainland as it passed by; subsequent differentiation on the mainland was explosive, on the islands very minor. The presence of related forms in the Galapagos and the Antilles could mean that representatives of some other groups (e. g., Darwin's finches got off the boat in the other direction at a relatively early date. Hence, the conclusion that "Nesomimus" are closer to northern-Antillean mockingbirds than to southern ones might not be wholly unique or unprecedented (unless I'm speculating overly wildly??)."

Comments from Robbins: "NO. The molecular data unequivocally demonstrate that Nesomimus is monophyletic; however, whether it is embedded within Mimus is open to question (note the poor nodal support) and even the sister taxon to Nesomimus is equivocal. Although including Nesomimus within Mimus likely will prove correct, the current data set does not establish that as fact. Before speculating about historical biogeographical scenarios, we should first solidly establish relationships and be consistent in our application of minimum criteria for evaluating those data."

Comments solicited from Peter Grant: "I do not have a strong opinion about the question of the appropriate genus. I tend to be conservative, and do not want to change a name unless the case is very clear. As Robbins indicates the critical node is not well supported, and if it were left to me I would not change the genus name until that node is resolved. I must admit that I expect Nesomimus to fall within Mimus. Another reason for being conservative is that I don’t like pinning a judgment on just one (mitochondrial) molecule, even when several genes have been assayed for greater statistical resolution. There are plenty of instances elsewhere of strong conflict between the signals of relatedness from nuclear and mitochondrial genes."

Comments from Pacheco: "NO. Em vista das opiniões sobretudo de Mark Robbins e Peter Grant, voto pela manutenção de Nesomimus, até que mais estudos complementem as sugestões ora levantadas e resolvam a questão."