2021-A-1a: Revise the classification of the Euphoniinae (Fringillidae): transfer Euphonia musica and E. elegantissima to resurrected genus Cyanophonia or to Chlorophonia

YES. The transfer of these two species out of the genus Euphonia seems very well-supported. While I had originally supported placing them in their own genus, Cyanophonia, I think the other members bring up good points about the variability in Chlorophonia and similarities between the two groups. Plus, transferring these two species to Chlorophonia (Option 2) would also be in agreement with SACC.

YES. The exciting new study discussed in the proposal is comprehensive in both taxonomic coverage and number of loci used. The new, well-supported phylogenies that are presented show Euphonia as paraphyletic and thus make the current taxonomy untenable. In these situations, I prefer to make changes that cause the least disruption in current taxonomy, including using names that are currently in use as much as possible. Thus, the simplest thing to do would be to transfer elegantissima, cyanocephala, and musica to Chlorophonia (Option 2). Only three taxa would acquire a new name, no old generic name would need to be resurrected, and no new generic name would need to be invented. I appreciate the argument that a descriptive match might be lost when using Chlorophonia to apply to these species, but I would also argue that all species in a newly constituted Chlorophonia still have green females and most share light blue coloration on males and/or females. My second choice would be to apply Euphonia to all species. 

YES. This generic split of E. musica and E. elegantissima is well-supported by lots of data, and the proposal makes a good argument for placing these species in the resurrected genus Cyanophonia. This treatment is also consistent with the proposed treatment by the SACC.

YES. Remove E. musica and E. elegantissima from Euphonia. I’m on the fence as to which genus to place it in, but for now am persuaded to put them in Cyanophonia. Having said this after reading further comments from others a few days later, I’m now persuaded to change and put them with Chlorophonia (Option 2).

YES. I agree to move Euphonia musica, elegantissima and cyanocephala to the genus Cyanophonia, because the results show a clear close relationship with Chlorophonia and not with the genus Euphonia. And with regard to plumage, some characteristics such as the blue hood in Cyanophonia are related to Chlorophonia but there are other characteristics in the plumage that show a clear differentiation between these two genera.

YES. After reading other’s comments and some of the comments in the SACC, I have changed my vote to place the musica group in Chlorophonia (Option 2), instead of reviving the genus Cyanophonia. I agree that we should have some standards for groupings of higher taxa, and a time-calibrated standard makes sense. I have reconsidered my view that morphological divergence should override a tIme-based standard in this case. Given that some splits within Euphonia (ss) are deeper than that between the musica group and Chlorophonia, we should place the musica group in Chlorophonia and not use Cyanophonia

YES. Option 2. I had originally voted for Option 1 rather than Option 2, but I was under the misrepresentation that Option 1 had been approved by SACC. 

YES. Option 1, resurrect Cyanophonia and transfer musica and elegantissima to this genus. A revision of genera is mandated by the new phylogenetic information. The decision to place elegantissima, musica, and cyanocephala into either a resurrected Cyanophonia or a more broadly defined Chlorophonia hinges on whether one favors consistency in crown ages / node depth or phenotypic distinctiveness among genera. For me, plumage patterning and coloration is arguably the most distinctive feature among species in Euphoniinae that are otherwise quite similar in terms of diet, morphology, habitat preferences, and other aspects of their natural history. In particular, elegantissima, musica, and cyanocephala have very similar arrangement of plumage patches and variation in color appears modular, with differences restricted to the forecrown—they make sense to me as a cohesive unit. The species within the currently defined Chlorophonia are united by the prevalence and hue of green plumage among males and females. To me, Chlorophonia and the proposed Cyanophonia are quite distinct from one another, but are also both cohesive groupings in terms of plumage patterns and coloration, which is what the name Chlorophonia refers to. Genus-level classifications are ultimately arbitrary, but if we prioritize them as a way to facilitate communication about groups of closely related species, I think resurrecting Cyanophonia makes the most sense.

YES. While placing the musica group in its own genus would surely seem warranted on the basis of plumage characters, this is a slightly younger split than two of the major lineages of Euphonia, which do not obviously break out along plumage pattern lines and thus the decision depends on which measure, degree of phenotypic divergence or time of divergence and comparability between related genera, is given precedence where there is discord. Although the differences in plumage between the brilliant green Chlorophonia and the musica group seem profound and well-marked, the Chestnut-breasted Chlorophonia in particular really does sort of bridge the gap in some aspects of plumage. These factors lead me to marginally prefer the single genus Chlorophonia (Option 2) with two subgenera.

YES. First, it should be noted that SACC rejected the resurrection of Cyanophonia [here (b) Option 1] — this is not mentioned in the proposal. Standard principles of classification require the transfer (unless genus Euphonia is to be expanded to include the entire subfamily). Tangentially, I see a looming English name problem. Should the transferred taxa retain the soon-to-be-misleading name “Euphonia” or be called “Chlorophonias”, or even “Cyanophonias.” There is a rare opportunity here to maintain the monophyly of the English names and have them match the genus names – in fact, because the English group names are taken from the genus names, it makes no sense to me to call the blue-hooded group “Euphonias”.

Although the SACC link was provided, I suspect that some did not read the SACC comments, much less made aware that SACC went with Option 2; note especially the comments on phenotypic continuity with a broadly defined Chlorophonia. Note that the SACC proposal was voted down 3 to 6, and that on average, SACC members are more familiar with this group in terms of phenotype than are NACC members. The bottom-line for me is that if we marked all recent nodes in a phylogeny with genus names, then we could add hundreds of generic names. I also have to add that to resurrect Cyanophonia and to simultaneously vote down resurrecting Corthylio makes no sense to me. The former is a lineage that is only 5 my old, whereas the latter is 3 times as old (with caveat that UCE data set much better than the mtDNA data set). Gauging morphological similarities or differences is “eye-of-beholder” stuff that is unsatisfying to me, but the plumage differences between Chlorophonias and “Cyanophonias” is arguably less, especially if you restrict the comparison to females – see Zimmer’s comment above.

YES. Option 2 (Chlorophonia).

YES: Time-wise, a single genus solution (i.e., all Euphonia) would put this group more in line with e.g., Turdus (ca. 7.5 Ma; Voelker et al. 2008 Global Ecology & Biogeography 18: 44). However, I could live with two genera and, because these are judgment calls, retention of Chlorophonia gives more stability to supporting a two-genus solution. So YES to a) (move these two species out of Euphonia). These would be a perfect subgenus in Chlorophonia, and so YES to b2, put them there. And YES to c, revise linear sequence. Genera can be difficult things to delimit, but we will continue to divide things more finely, with less and less informative genera (e.g., we are undergoing an explosion of monotypic genera), if we don’t start doing more among-Aves contrasts of what it means to be an avian genus. Genus-level proposals are often failing to make contrasts with generic limits in other birds, being on average much too inward-looking. They also seem not to entertain subgeneric divisions enough. We need to start seeing more contrasts among genera across broader swaths of Aves so that an avian genus has more evenness or equivalency as a unit.

2021-A-1b: Revise the classification of the Euphoniinae (Fringillidae): revise the linear sequence of the subfamily

YES. 7 without comment.

YES. The proposed linear sequences matches the new phylogeny.

YES. Linear classifications should be updated as new phylogenies are published.

YES. Required book-keeping by our linear sequence conventions.

YES: Details above.

2021-A-2: Split Band-rumped Storm-Petrel Hydrobates castro

YES. Option B (2 species). This is a fantastic proposal that lays out all of the evidence for these splits. I was torn between option A and B, but currently think we should adopt the more conservative approach, and to wait until more taxa are included in playback experiments. Although the evidence from playback experiments for jabejabe isn’t as strong as it was for monteiroi, I think it is strong enough, especially given the genetic divergence involved. 

YES. Option A (3 species). Do not maintain the status quo because of the new data presented. Although more data is always desirable, at this point I think we have enough evidence for four biological species including the already split monteiroi (Option A). jabejabe is genetically very different from the other populations considered here. Splitting the remaining populations into two makes the most sense to me, given the data presented. castro and cryptoleucura are genetically very different from each other, and there is some playback evidence to support the split of cryptoleucura (at least study of Galapagos populations would indicate). I don’t think there is enough data to split helena and bangsia. I vote yes for option A, although option B (2 species) is my second choice. 

YES. Option B (2 species). Thanks to Oscar for a well-written, complete proposal on a complex situation. I could be talked into the recommended three-species treatment, but for now I favor the more conservative two-species approach pending further vocal data as noted in the proposal’s option B.

YES. Option B (2 species). A fascinating proposal and I have little doubt that multiple species are involved. Splitting monteiiroi was certainly mandated under the BSC concept. The genetic and especially the vocal (from H. castro) differences, well described and recorded by The Sound Approach (2008), persuade me to separate out the Cape Verde Islands population. The Sound approach (2008) describes the calls of Cape Verde Islands birds as elaborate chatter calls and purring, containing far more notes. For the breeders farther north in the Atlantic, excluding Monteiro’s, chatter calls and purring end with long-short flourishes. For the others we are dealing entirely with separate allopatric breeding populations so genetic divergence should not be surprising, and particularly between Atlantic and Pacific populations. So, yes makes sense perhaps for up to three more species (South Atlantic, Galapagos and Hawaii/Japan) on biogeographic grounds, but all of those issues need much more investigation, along the lines for the research for Monteiro’s and Cape Verde. 

YES. Option A (3 species). I support the option that divides this complex into three species, though I suspect that this is not the last word on this group. Kudos to the author of this proposal for doing a great job of synthesizing the complicated literature on this particularly challenging species complex (perhaps one of the most taxonomically challenging in all of ornithology?). The lack of reciprocal playback studies across all populations seems like an impossible bar to establish for this group, given that these birds breed on some of the most remote and far-flung islands in the world; I’m actually impressed that there is as much existing information on traits relevant to reproductive isolation as we have. I suspect that what we need to do here is what this committee has done in so many other situations (especially when dealing with allopatric populations): extrapolate from the known inferences about RI to situations where those comparisons are not possible.

To my mind the main ‘flaw’ in this situation is that our methods for reconstructing patterns of differentiation are advancing very quickly and the topology-based methods employed in the Taylor et al. study are already a bit out of date (not a criticism of that study or its authors: it’s just that our analytical toolbox is improving since that paper was submitted). This is particularly germane to situations like this that sit at the low-differentiation interface between what we used to refer to as ‘population genetics’ versus ‘phylogenetics,’ and someday soon I look forward to this complex being revisited with whole-genome data that allow the exploration of patterns of gene flow (or its absence) on a fine genomic scale.

I also note that this is a proposal where the number of possible options (most of them seem reasonable to me) will tend to divide us as voters and cause any change to fail. We might need to think about some version of ranked-choice voting for these circumstances…

YES. Option A (3 species). Taxonomy should be revised. At a minimum, jabejabe should be recognized as a separate species from castro. A thorough and well-written proposal. Very useful to make sense of a complicated situation. As detailed in the proposal, I feel the SNAPP phylogeny is likely closer to the true underlying phylogeny for the group. I think having four species (i.e., jabejabe, monetiroi, castro, cryptoleucura) within Hydrobates is the most sensible solution given the data on hand. With this option, each species is monophyletic. If we opt for Option B (my second preference), then the grouping of castro, helena, cryptoleucura, bangsi, and kumagai will be paraphyletic with respect to monetiroi. This is a fine option too as species can be paraphyletic under the BSC, and especially make sense in this case with monetiori as a geographically restricted species endemic to the Azores. There will likely be future proposals and changes regarding species statuses among helena, cryptoleucura, kumagai, and bangsi, but I don’t think we have sufficient data on reproductive isolation among these taxa to recognize them under the BSC.

YES. Option A (3 species), the recommended option to split castro into three species (castro, jabejabe, and cryptoleucura). I agree that more work is needed on the complex, but it seems to me that jabejabe is surely best considered a species, based especially on the nuclear data, and that playback experiments and the phylogenies strongly suggest that cryptoleucura is too, but less decisively. I would add that these splits are similar to those of several others we have recently made in Procellariiformes, and with the exception of S Atlantic helenae being conspecific with cryptoleucura, make biogeographic sense.

As for a 2-way split, my second choice is to vote YES to split jabejabe even if cryptoleucura isn’t split.

YES. Option A (3 species + H. monteiroi), as recommended in the proposal for the reasons elucidated therein. I think this is the best taxonomic interpretation of the data published so far. (By the way, Oscar wrote a terrific proposal on a complex system. Too bad that he wasn’t a reviewer for the paper!). If splits are adopted, then I think we should have a separate proposal on English names for all taxa, not just cryptoleucura, just to make them official.

NO. There is no strong evidence to separate Hydrobates castro into additional species.

NO. This is an amazingly detailed proposal and spells out well the current understanding in this complex complex of storm-petrels. The analyses have it all: recent divergences, incomplete lineage sorting, possible gene flow, little to no plumage evolution, a fragmentary knowledge of behavioral differences, different signals from different data sets, and some reproductive isolation. We have established the benchmark to split monteiroi, and can see how this plays into the various phylogenies. 

It seems we are relying almost completely on the phylogeny to make these splits, with very little data on vocalizations, plumage, or morphometrics. We are not using data that directly results from reproductively isolating mechanisms aside from geography. I have a hard time keeping the Atlantic taxon helenae lumped with the Pacific taxa, based entirely on the phylogeny, and maybe it should be split off by itself (pending results from the Gulf of Guinea populations). At the same time, it is difficult to see how helenae may be reproductively isolated from North Atlantic castro populations. The genetic distinctiveness of jabejabe is pronounced (at least in this clade) but it may not even be vocally distinct from castro. I do not mind that excluding monteiroi from castro (s.l.) renders the latter polyphyletic, as reproductive isolating mechanisms may have arisen within monteiroi as an autapomorphy. I think we need to know more before we split any of these taxa.

NO. No additional comment.

NO. Maintain castro as a polytypic biological species. Taylor et al. (2019) have done some great work on a complicated species complex, uncovering wonderful levels of cryptic diversity (evolutionarily significant units, ESUs). But are they biological species? They never actually measured gene flow; they do not consider a subspecies-level solution; and they adopt a different species concept (basically one favoring genotypic clusters). This makes interpretation of biological species limits rather difficult. Allochronic populations are no more species than allopatric ones are, and genotypic clusters are very common within biological species (including humans). And at these shallow levels of divergence, we expect phylogenetic discordance among datasets, including polyphyly and paraphyly. So I disagree with making species limits decisions by trying to impose monophyly, especially without reference to clear corroborative phenotypic traits that correspond with strong reproductive isolation (i.e., not subspecies-level differences). Good proposal. I appreciated the discussion of methodological issues and gene tree conflicts and also the observation that “Particularly relevant to taxonomic decisions would be studies that include estimates of rates of gene flow between taxa, detailed analyses of song structure, and vocal recognition experiments including all taxa.” I would only add that playback experiments are overrated, and that actual estimates of gene flow would be more definitive in these cases.

2021-A-3: Split Mew Gull Larus canus brachyrhynchus and rename as Short-billed Gull

NOTE: The vote on 2021-A-3(c) has been postponed.

(a) YES. The vocal differences are especially compelling to me for this split. (b, d-e) YES to all.

(a) YES. This is a close call for a couple reasons. Much of the data is in an unpublished dissertation. Does anyone know why that work remains unpublished? Secondly, the genetic data appears to be mtDNA alone. That said, it’s usually the case in gulls where introgression in mtDNA has led to similarity across taxa that are recognized species. This is the opposite case, with distinct difference for brachyrhynchus compared to other taxa. That, coupled with the vocal data, leads me to go ahead and vote a borderline Yes for a split. The morphological data to me are less compelling, given that we are using a biological species definition. (b, d-e) YES to all, based on arguments presented in the proposal. 

(a-b, d-e) YES. The proposal does a good job of presenting a case for this split, and for recognizing the English name Short-billed Gull for brachyrhynchus.

(a) YES. Part of this issue for me is what were the reasons it was lumped a century ago? As noted in the motion they (nominate canus from brachyrhynchus) were treated as different species through the 3rd edition of the Check-list (1910), but then lumped by the 4th edition of the Check-list (1931). Dan Gibson helped me track down what happened in between. Obersholser (1919) published a short article (Notes on North American Birds) doubting the validity of brachyrhynchus as a full species and relied on an analysis of two specimens, one from Slave River, Mackenzie (USNM 19351), the other from Bering Island, Commander Islands, Russian Far East (USNM 92894). He keys on what we call P8 and P9 (he counts the outermost as P1, then counts in), a key difference between these two taxa (nominate canus has more black). In particular he notes that the Bering Island specimen had a brachyrhynchus-like primary pattern. I’m not sure about the Mackenzie bird, but on the Commander Islands, both taxa occur as rare strays (Vaurie 1965). Vaurie (ibid) mentions three records of brachyrhynchus for the Commander Islands. Gibson and Byrd (2007) cite Hartert (1920) in citing two records of immatures for the Commanders. I’m wondering if Oberholser’s statement about the Bering Island specimen was actually one of those records of brachyrhynchus? One puzzlement to me is that Oberholser doesn’t even refer to kamtschatschensis even though it was named in 1857. Why not? Oberholser (ibid) concludes with the comment that based on his examination of the two specimens (cited above): “In view of the above facts it seems necessary to consider Larus brachyrhynchus a subspecies of Larus canus, and its name will therefore become Larus canus brachyrhyncus Richardson. Taxonomy was easy then, or so it seems. A year later (Oberholser 1920) he just lists this taxonomic treatment. Look at a couple of specimens and issue a decree, one that would be followed for a century. There is no consideration of anything other than definitive adult plumage. The immature plumages are much more distinctly different. Asian kamtschatschensis is closer to canus (and heinei) than brachyrhynchus. Slightly larger and darker mantled heinei (from canus) meets and intergrades with kamtschatschensis along the Lena River (Vaurie 1965), so any further splitting in this complex will need to await more studies there. East Asian kamtschatschensis is completely allopatric from brachyrhynchus. The latter is common in Nome on the Seward Peninsula. Both subspecies are very rare on St. Lawrence Island to the southwest. 

Rather than dwell on Oberholser, I wanted to cite Peter Grant’s (1986) comments on geographic variation: “Four subspecies are generally recognized, and are described below. Marked plumage differences between canus and brachyrhynchus are in themselves, apparently sufficient to warrant specific status for latter (thus Mew Gull L. brachyrhunchus), but the clinal nature of plumage and size differences between the four subspecies, especially intergradation between heinei and kamtschatschensis in Siberia, and the similarities of plumage (especially first-years) between kamtschatschensis and brachyrhynchus, are strong arguments against any split.” I don’t agree with Peter’s assessment about this entirely. The tail pattern of first cycle kamtschatschensis is like canus and heinei, not brachyrhuncus, and I would say the overall coloration is too. 

Literature cited: 

Gibson, D.D., and G. V. Byrd. 2007. Birds of the Aleutian Islands, Alaska. Series in Ornithology, No. 1. Nuttall Ornithological Club and the American Ornithologists’ Union. 

Hartert, E. 1920. The birds of the Commander Islands. Novitates. Zoologicae 27:128-158.

Grant, P.J. 1986. Gulls, a guide to identification, 2nd edition. T & A D Poyser. 

Oberholser, H.C. 1919. Notes on North American Birds VII. Auk 36:81-85.

Oberholser, H.C. 1920. Fifth annual list of proposed changes in AOU Checklist of North American Birds. Auk 37:276.

Vaurie, C. 1965. The Birds of the Palearctic Fauna. Non Passeriformes. H.F.& G. Witherby ltd. 

(b) YES. Retain kamtschatschensis in canus for the reasons listed above. Investigate the Lena. (d) YES to Short-billed Gull for the English name for the hopefully restored L. brachyrhynchus. (e) YES to retaining the thoroughly uninspiring name of Common Gull for the Old World taxa in this complex. You would have thought that the Europeans would have thought that the Herring Gull and all of its permutations would have been the “Common Gull.”

(a-b, d-e) YES: Yes to all components of this proposal. This boils down to an overall ‘weight of the evidence’ situation to me even though there are limitations to each of those components when viewed in isolation. The molecular data in particular are a bit old-school, and we certainly understand now the many ways that mtDNA results can be misleading (in both directions: populations of the same biological species can have high mtDNA divergence, and undisputedly valid biological species can have very low mtDNA divergence). But here, these are gulls, which tend to hybridize A LOT as a group, so you would expect mtDNA introgression if they were not functioning as species. Instead there is clean separation in the haplotype network across this proposed split; for me it is not the magnitude of that difference that is most telling, but rather the lack of evidence for mtDNA introgression facilitated by interbreeding. The detailed study of trait variation is impressive and highly consistent with what the available phylogenetic evidence also suggests. That combination seems enough to me.

I also have a different perspective from some on whether a thesis or dissertation counts as ‘published’ information. I strongly think that it does, because (1) they are archived and available as scientific products and (2) theses and dissertations are ‘peer reviewed’ by the examining committee.

(a-b, d-e) YES. Yes to all, for reasons stated in the proposal.

(a) YES on the species split, based primarily on the documentation of differences in display vocalizations. The other data are consistent with species treatment but not sufficient on their own for anything but subspecies rank for brachyrhynchus. By the way, this is a return to the species limits assigned by Ridgway (1919). (b) YES. Certainly that is what the genetic data, albeit unpublished, indicate. (d) YES. Nice match with species name, good description of a distinctive feature, and a name with historical precedent. The last time brachyrhynchus was considered a separate species, at least by major classifications in the Western Hemisphere, was by Ridgway (1919), who called it Short-billed Gull (and used Mew Gull for L. canus sensu stricto). In the W. Hemisphere, Mew Gull has thus been used strictly for the Palearctic (Ridgway) and for the combined species (e.g., AOU 57, 83, 98), so if that name disappears, that’s fine with me to prevent further confusion. AOU (1931) was the first treatment that I can find in the W. Hemisphere that considered them conspecific, and, as was the practice at the time, AOU (1931) applied Short-billed Gull to the subspecies L. c. brachyrhynchus, and this was followed by Hellmayr & Conover (1948); because none of the Palearctic canus group had been recorded in W. Hemisphere at that time, Short-billed referred only to the Nearctic taxon, L. c. brachyrhynchus. Therefore, from the historical perspective, use of Short-billed makes common sense. (e) YES, assuming this is the English name preferred by Palearctic people. 

(a-b, d-e) YES.

(a-b, d-e) YES. I do not find Sternkopf’s (2011) data convincing for species limits because it does not adequately assess gene flow between these forms (and they are gulls, after all), but coupled with the phenotypic evidence and additional yet-unpublished evidence I think the answer is pretty clear. While I might suggest waiting until the unpublished data are available, this is more efficient and I think reads existing evidence correctly. 

(a) NO. More data and analysis are needed to separate canus vs brachyrhynchus. The boundaries of the two species are not clear, even though they have been documented to have different songs. (b) YES. Retain the kamtschatschensis in canus. (d) NO. I disagree with the split. Retain English name of Mew Gull for L. canus brachyrhynchus. (e) NO. All taxa must have the name Mew Gull.

(a) NO. We have split some sister species of gulls (e.g., Olrog’s and Belcher’s Gulls), but lumped others (e.g., Thayer’s/Iceland Gull). Larus canus (ss) and brachyrhynchus are mostly diagnosable and seem to have some vocal differences, but I think the vocal analyses need to be expanded to be useful in drawing species boundaries at this time. Nothing is mentioned of any actual examples of parapatry, so I assume that there are none. Although the American brachyrhynchus is genetically quite divergent from the Old World forms, genetics in gulls is so messy, with what I feel is rampant gene flow and reticulation of phylogenies, that I have trouble inferring much usefulness in the results. In the end, I do not see solid evidence of reproductive isolation. (b) YES. (d) NO. I voted not to split. (e) NO. All taxa = Mew Gull.

(a) NO. One issue is that the genetic data have not been published in a peer-reviewed journal. Furthermore, % mtDNA divergence or differentiation in microsatellites is not sufficient to split. Many biological species have ≥2% mtDNA divergence among subspecies. Furthermore, there is substantial overlap in the NMDS plots of the morphological characters. Even if we accept that brachyrhynchus is morphologically diagnosable from kamtschatschensis, it is not clear that these slight differences in primary tip coloration translate to reproductive isolation between them under the BSC. Although the descriptions of vocal displays are suggestive of divergence and assist in identification, this current evidence seems purely qualitative. I would like to see a more rigorous analysis with larger sample sizes before accepting that brachyrhynchus is diagnosable in terms of its vocal displays. (b) YES. I do not think there is sufficient evidence to split kamtschatschensis at this time. It should be retained in canus. (d) NO. Given that I did not vote to split in part A, L. canus brachyrhnchus should retain the English name of Mew Gull. (e) NO. Given that I did not vote to split in part A, all taxa should retain the English name of Mew Gull.

2021-A-4: Revise generic limits in the Phalacrocoracidae (cormorants)

(a-b) YES. Changes are well-supported by the data.

(a-b) YES.

(a-b) YES.The data supported the changes.

(a-b) YES. Families with one genus comprising almost all species do not impart much taxonomic information. If splits are old (comparable to genera in related families), then it makes sense to split genera even if there is not a great a deal of morphological evolution (e.g., cormorants). It is unfortunate that Neotropic and Double-crested get saddled with “nano-winged” when they are fully flighted.

(a-b) YES. Two genera for all of Phalacrocoracidae (40 species) seems too few, and 35 species in Phalacrocorax is unwieldy—even if morphology and plumage is largely conserved among its constituent species. The seven-genus classification proposed by Kennedy and Spencer (2014) seems a very reasonable solution. The proposed genera make sense in various ways, including biogeography (Urile are northern Pacific Ocean cormorants; Leucocarbo share blue eyes, etc..) and are roughly similar in age. However, one strike against this change is that the genus Phalacrocorax is already monophyletic, and this proposal does not promote taxonomic stability. However, genera are most useful when they correspond to cohesive clades of a manageable size. I’m not sure that I agree in splitting Nannopterum from Leucocarbo, but 35 species in one genus is too many for my taste, and this classification is a good, stable remedy for this point forward. All this said, I feel we should consider this in conjunction with the results of the SACC votes and try to coordinate our decision to be consistent between our two groups, so maybe we can pool our votes together to see whether it passes 2/3 majority?

(a-b) YES. Reasons stated in the proposal.

(a-b) YES. As noted in the proposal, this proposal came within 1 vote of passing SACC in 2014, and that was before Kennedy et al. (2019), and I strongly support it for all the reasons in the proposal. Multiple genera that reflect the venerable ages of the several lineages also emphasizes a distinctive theme of cormorant evolution, namely morphological conservatism.

(a-b) YES. The concordance between the osteological and molecular phylogenies and the relative ages involved make these good genera, in my view.

(a) NO. I understand this is not an option in the proposal, but I do not agree that the cormorants should be split into so many different genera, although I do agree that additional genera are preferable, given the levels of divergence within the family. My preference would be to at least place brasilianus, auritus, and harrisi within Leucocarbo. These two clades are strongly supported as sister taxa, and the levels of divergence between them matches levels of divergence within even the narrowly defined Phalacrocorax (although generally speaking I’m also not necessarily a fan of defining genera/families by clade age), which would maintain consistency across the family and genera. (b) NO. This clade is more difficult to make a decision on, but I’m inclined to again vote NO, in preference of including penicillatus, pelagicus, and urile within a slightly broader (though still reduced) Phalacrocorax. These three taxa are strongly supported as sister to the Old World Phalacrocorax clade sensu stricto.

(a-b) NO. To maintain stability, I am not in favor of revising generic classification unless new evidence indicates lack of monophyly of existing genera. If there was some global consensus to use a new generic taxonomy for cormorants, then I might be more in favor of these changes. 

(a-b) NO. My rationale here will be unsatisfying but here it is: genera are arbitrary groupings (other than the criterion that they be monophyletic) and in this case there are many valid but alternative ways in which this family could be subdivided into more or fewer genera; there is not one outcome that clearly stands out as more reasonable than the others. Meanwhile this is a group with a worldwide distribution, with relatively few species/lineages occurring in the NACC area. Since there are so many valid outcomes here, I would not want to see NACC taking the lead on any particular one of them only to end up being inconsistent with other global taxonomic authorities. Thus is a great example of a group for which a global consensus approach really needs to be pursued.

2021-A-5(A): Adopt split of Long-billed Gnatwren Ramphocaenus melanurus

(a-c) YES. I vote for all the changes recommended in the proposal to bring us into agreement with SACC, including to use the name adopted by SACC.

(a-b) YES. To comply with changes adopted by SACC, including the name used by SACC. 

(c) NO. I’m not sure why the linear sequence needs to be changed, both Microbates and Ramphocaenus now have two species and they are sister taxa. 

(a-c) YES. I vote for all changes recommended in the proposal, for the reasons given and for consistency with other treatments.

(a-c) YES.

(a-c) YES. The data supported the suggested changes in all the three proposals.

(a-c) YES. The data are strong to split sticturus from melanurus, following the SACC. 

(a-c) YES. No additional comments.

(a-c) YES. Let’s follow SACC’s lead as this species is predominantly within their geographic jurisdiction. 

(a-c) YES. Reasons stated in the proposal.

(a) YES. Reasons indicated in SACC comments. (b) YES on English names already in use by SACC. (c) NO. I didn’t follow the explanation. Only two species of Microbates, so it should remain first – I must be missing something.

(a-c) YES.

(a) YES. Split out sticturus, following SACC and recognizing a narrower melanurus. The evidence is strong for this. (b) NO in the interest of stability, especially with the possibility of another change necessary if melanurus undergoes another split. With the type locality of nominate melanurus in Brazil, such a split would require yet another English name change to the currently (in our area) well recognized Long-billed Gnatwren. (c) YES. New sequence so Rhamphocaenus precedes Microbates.

2021-A-5(B): Retain the English name Long-billed Gnatwren for Ramphocaenus melanurus sensu stricto

YES. 12 without comment.

2021-A-6: Change the linear sequence of the gnatcatchers (Polioptila spp.)

YES. 6 without comment.

YES. This is well-supported and necessary based on new phylogenetic data.

YES. Proposed sequence matches the phylogeny. 

YES. The sequence appears to follow our sequencing criteria.

YES. Linear classifications should be updated when new phylogenetic data are published.

YES. Reasons stated in the proposal.

YES. Required book-keeping by our linear sequence conventions.

2021-A-7: Split Tropical Gnatcatcher Polioptila plumbea into two or more species

(a-b) YES. 2 without comment.

(a) YES. Given the polyphyly of the group based on the phylogeny of Smith et al. (2018), this split seems well-supported. (b) YES. Given its use already for this group, this is a good English name for this taxon. 

(a-b) YES. Reasons outlined in the proposal.

(a-b) YES. Reasons given in the proposal. English name of White-browed Gnatcatcher for the bilineata group seems good and is consistent with other authorities.

(a-b) YES. White-browed Gnatcatcher for English name. 

(a-b) YES. Enough support for the split, and the English name for bilineata group (White-browed Gnatcatcher) is good and I support it.

(a) YES. The rampant poloyphyly in Poilioptila requires plumbea to be split into multiple units. The trans-Andean taxon (bilineata) forms a nice clade composed of similar looking taxa. It may get a bit messy in Colombia, but that does not affect the distinctiveness of the bilineata clade from plumbea. (b) YES. English name of White-browed Gnatcatcher is mostly appropriate.

(a) YES. Given the polyphyly presented by Smith et al. (2018), it makes sense to recognize the plumbea group (i.e., plumbea, parvirostris, and atricapillus) as a separate species. The polyphyly has strong node support and is quite deep. If we continue to recognize Polioptila dumicola, the logical next step is to classify plumbea, parvirostris, and atricapillus as a separate species. (b) YES. Provided the name is already in use by Sibley and Moore (1993), this makes sense. White-browed Gnatcatcher is an appropriate name given that facial patterns are important field marks in this group.

(a-b) YES. Reasons stated in the proposal.

(a) YES. Essentially required by phylogenetic data. (b) YES, pending SACC vote on D.

(a-b) YES. This one makes me very leery, for several reasons. 1) It is based on mtDNA and we’ve found that to be wrong often enough (gene trees do not equal species trees) that we’ve largely abandoned using mtDNA alone to make especially species-level splits. But here polyphyly is doing the heaviest lifting as justification for a split. 2) Smith et al. (2018) themselves stated that “For most, determining whether or not they are biological species will require additional work in genetic sampling, study of song, morphology, behavior, and putative hybrid zones.” 3) mtDNA polyphyly and paraphyly are relatively common among biological species. The classic Funk & Omland (2003, Ann Rev Ecol Evol Syst) finding was “This survey detected species-level paraphyly or polyphyly in 23% of 2319 assayed species, demonstrating this phenomenon to be statistically supported, taxonomically widespread, and far more common than generally recognized.” All that said, plumbea is a mess. Did anyone else find it odd that Smith et al. (2018) did not give time estimates on key nodes? I went onto GenBank to get a quick estimate and found that these two clades in which plumbea occur are almost 14% divergent. With that (though I hate to use mtDNA divergence generally) and the phenotypic trait of the northern clade, “broad white eyebrow year-round” per above, I can vote YES on (a) (not primarily due to polyphyly), and YES on (b), and the rest for now is up to SACC.

2021-A-8: Split Fregata rothschildi from Magnificent Frigatebird F. magnificens

YES. The Galapagos population is very distinctive genetically. It would be nice to have vocal and other behavioral data, but I think the evidence is sufficient to support this split. With regard to the English name, I abstain for now. I’d like to consider other options, as I agree that retaining the English name Magnificent Frigatebird for the species that is not F. magnificens is confusing.

YES. The genetic differences (and associated lack of gene flow) are strong enough for me to consider these separate species, even without analyses of courtship, etc. As the proposal indicates, these kinds of differences are not seen across the isthmus of Panama, although this is a barrier for other seabirds. (b) NO. Let’s find something better than Magnificent, too confusing given species name change. (c) YES.

YES. Barely yes, though. The genetic data are convincing that the Galapagos population is functioning as an isolated entity, despite the high vagility of this taxon. But I also agree with the comments suggesting that comparisons of vocalizations and displays would be helpful, as would better sampling of mainland frigatebird populations.

NO. 1 without comment.

NO. While the genetic evidence is fairly compelling, I would want to see samples from additional populations, as well as vocal and display evidence, given the otherwise very minimal morphological differences. I think it is likely that these two taxa are distinct at the species level, however I don’t think the evidence is quite there yet to make the split.

NO. Let’s see more. For some, morphometric data wouldn’t even qualify for subspecies designation. 

NO. Although the genetic data supports the division, it seems to me that more data from other populations are needed to have the full history of this species.

NO. I would like to see evidence of differences in display or behavior that would lead to reproductive isolation. The morphometric differences are indicative of a subspecies level difference.

NO. Data are certainly suggestive of separate species status, but I think that additional sampling of coastal South American populations is needed before we can be sure. F. rothschildi populations from Ecuador are most likely to exchange gene flow with F. magnificens magnificens, and need to be considered alongside these data for a conclusive study.

NO. The existing data are suggestive, but much more is needed.

NO. SACC voted this one down unanimously because of incomplete sampling. Comparative data on displays, not just missing genetic samples from populations, would also be part of the incompleteness of that sampling,

NO. These are at best subspecies-level differences in phenotype (I don’t see clear phenotypic diagnosability, did I miss it?). The genetic data are pretty good in sample size, and there are clearly some barriers – but why are those barriers not isolation by environment (IBE) rather than actual reproductive isolating mechanisms? It looks to me like there is no way to answer that question given present evidence. In the absence of clear phenotypic divergence my bet is on IBE: weak phenotypic evidence but strong evidence of a difficult environmental barrier to overcome (i.e., “magnificent frigatebirds banded in Galapagos have been recovered as dead and/or emaciated vagrants in Central America”). If (a) should pass, I would vote YES on (b) and (c).

2021-A-9: Resurrect Corthylio for Ruby-crowned Kinglet

(a-b) YES. 1 without comment.

(a-b) YES. First, data of all types strongly support the distinctiveness of calendula relative to other species of Regulus, which are very similar among themselves. These data include plumage and other morphological characters, vocalizations and other behavior, protein electrophoresis, and mitochondrial DNA. All of these were cited in the proposal but, as Max has pointed out, UCE data relevant to this question but not cited in the proposal are available in Oliveros et al. (2019). The morphological data alone were considered sufficient to place calendula in the genus Corthylio in the 4th edition of the Check-list (although changed to a subgenus in the 5th edition), and the vocalizations (both songs and calls) are highly distinctive. Moreover, every type of molecular data that has been studied has shown the divergence between calendula and other species of Regulus to be typical of that of higher-level categories. The electrophoretic differences in Ingold et al. (1988), who sampled the North American species calendula and satrapa, were comparable to those between closely related families or highly divergent genera, and the mtDNA data of Päckert et al. (2009), who sampled all species, showed calendula to be as divergent from the rest of Regulus as most families (15 million years by their timescale) and twice as divergent as any other species of Regulus are from one another (7-8 million years). The UCE data of Oliveros et al. (2019) also showed a large divergence between calendula and the other species sampled (R. regulus), estimated at 12 million years. To me, this seems like more than ample data on which to base recognition of a new genus. We have phenotypic and mitochondrial data from all species, electrophoretic data that compare calendula with satrapa, and genomic data that compare calendula with regulus, and all point to the same result. We could wait for sampling of genomic data for all species (or nuclear data of some type for all species), but this seems to me to be a higher bar than we usually set, given the phenotypic and genetic data already available.

Second, the current generic arrangement, in which all species in the family Regulidae are placed in Regulus, provides no additional information other than the false impression that the six species are all very closely related because they are all placed in the same genus in addition to the same family. Monotypic genera do not provide information on relationships to other species in the same way that polytypic genera do, but in this case this information is provided at the family level: the relationship of calendula with the other five species of Regulus is clear because all belong to the same small family. The fact that they form a monophyletic group is obvious because they are all in the same family – a single genus is redundant (and misleading) in terms of information content. In this case, recognition of a monotypic genus would add taxonomic information without any tradeoff: placing calendula in Corthylio would produce a classification that conveys the distinctiveness of this species relative to the rest, a broad-based distinctiveness as detailed above, while simultaneously reflecting the monophyly of all species formerly in Regulus through their constituting the family Regulidae.

Third, our responsibility as a committee is to produce an up-to-date classification that reflects the best and most recent scientific data, and this includes making decisions about taxonomic categories that do not have precise definitions or criteria, but about which we can make informed judgments. Transferring calendula to Corthylio would be consistent with other recent changes to higher-level taxa (genus and above) based on criteria other than lack of monophyly. For example, we recognized Helopsaltes in 2020 based on deep genetic divergences within Locustella and we recognized Leiothlypis in 2019 based on deeply divergent genetic clades within Oreothlypis consistent with long-recognized phenotypic differences. We’ve also recently recognized other higher-level taxa, such as families, subfamilies, and orders, based on similar rationale (i.e., very divergent and deserving of separate recognition based on comparisons with other taxa of the same rank). Virtually every supplement contains a change of this type, and some contain many.

(a-b) YES. The long branch leading to R. calendula, and its phenotypic and vocal distinctiveness from other Regulus, warrant recognizing the monotypic genus Corthylio

(a-b) YES. This is one of the commonest passerines in much of the United States. I agree with comments from others that commented how different this species is from other Regulus. The song and call notes bear no resemblance to the other kinglets. Habitat is different too as is behavior. The definition of a genus is pretty vague with no clear standards, but I sure see them as completely different. Then again, I said that about Wilson’s Phalarope. I’m not sure what to think about the category of a subgenus?

(a-b) YES. First a single genus in a family doesn’t convey any taxonomic information. Calendula is far the outlier. Also, not a taxonomic argument as much as a biogeographical one: with the two species in our area and both in the same genus, one would think that they may be sister taxa. But satrapa is far more closely related to the four other Regulus not in our area.

(a) YES. I had initially voted against this outcome but was persuaded to change my vote by the lines of reasoning advanced in other comments. (b) YES. Warranted by the phylogenetic hypothesis.

(a) YES. I originally voted NO as I was uncertain how NACC considers changes in taxonomic ranks above species level for reasons other than lack of monophyly. These changes occur regularly, and I think that the degree of phenotypic ifferentiation—especially the node depth of 15 mya—merits transfer into a resurrected Corthylio. In this case, I think a monotypic genus does make sense, but we should be careful to employ monotypic genera sparingly in my opinion. (b) YES. R. calendula should come first based on new phylogenetic data. 

(a-b) YES. Reasons stated in the proposal.

(a-b) YES. A 15 mya split almost demands a separate genus. Of note is that their voices, both song and calls, differ to a remarkable degree given that they are both tiny passerines of roughly similar acoustic habitats (as documented by papers cited in the proposal).

As for the argument that because Regulus as currently delimited is monophyletic, then there is no reason for a change, that same argument could be applied more broadly ad absurdum to the point that all birds are in one genus. I don’t “like” or “dislike” monotypic genera. I don’t get it. If a lineage has only one species and genetic data reveal it is highly divergent, possibly evolving separately since the Miocene, then placing it in a monotypic genus injects information into the classification.

Concerning the point that a change would disrupt stability, my response is … this stability needs to be disrupted! Continuing to think of them as congeners when we now know they have been evolving separately for a longer time than some taxa we rank as families (e.g. in Thraupidae) masks this important feature of evolution in Regulidae. The purpose of a classification is to reflect change revealed by new research, not preserve status quo over all other considerations. Stability in this case means that users will continue to think of the kinglets as a bunch of closely related species when in fact one of them, calendula, is distantly related to the others; thus, “stability” hides a degree of genetic divergence that is greater than that between tanager taxa we vote to recognize at the family level not long ago.

The only argument for retaining them as congeners would be if mtDNA can overestimate time-since-divergence by more than an order of magnitude. I made a quick attempt to find a time-calibrated phylogeny in the Sylvioidea and could not find anything useful. If someone can convince me that the mtDNA divergence estimate could be so far off that it could be, say, under 5 mya, then I would change my vote.

(a-b) YES. While I do not like monotypic genera, Corthylio has both the morphological and time attributes of a good bird genus. I resurrect some tidbits from a prior set of comments (in which I favored hummingbird subgenera): we have Turdus as an example genus (ca. 7.5 Ma; Voelker et al. 2008 Global Ecology & Biogeography 18: 44), and our recently (2017) revised view of Anas and the genera that used to be part of it, with clade ages of 9.4 – ~11.2 Ma.

(a) NO. I don’t think the morphological differences are strong enough to warrant moving Ruby-crowned Kinglet to a new genus, and I agree with the authors of the proposal, Brandon and Natalia, that it disrupts stability unnecessarily. (b) YES. 

(a) NO. At this time, I don’t think we should change generic names, unless changes in phylogenetic topology require it. I would prefer to keep Ruby-crowned in Regulus for stability. While I appreciate the argument for ranking taxa based on age, I think a temporal banding approach should wait some years until phylogeny is more stabilized. I would also note the next node in the kinglet tree is also pretty old, should we not also provide names for these clades as well? One argument against temporal banding is the idea that classifications are primarily about trying to organize diversity. For example, even though Thraupidae is a young family, does it make sense to put all those species in one genus? By doing so, you might learn that the speciation rate is high in the family, but it makes it harder to talk about groups within that family. (b) YES. Ruby-crowned is sister to the remaining taxa, so needs to come first per convention. 

(a) NO. I think more data is needed to consider this. (b) YES.

2021-A-10: Change the linear sequence of species in Chaetura

YES. 5 without comment.

YES. These changes fit very well with the recent genetic evidence, and are very clear-cut.

YES. Proposed sequence matches phylogeny and rules for drafting such linear sequences. 

YES. Reasons are given in the proposal.

YES. Except should we not have vauxi before pelagica (vauxi is more NW)?

YES. Linear sequences should be updated when new phylogenetic information arises.

YES. Reasons stated in the proposal.

YES. Proposal has now passed at SACC, unanimously so far.

2021-A-11: Reconsider the case to split Saxicola maurus and S. stejnegeri from Common Stonechat S. torquatus


2021-A-12: Resurrect Philodice as a separate genus from Calliphlox

YES. 3 without comment.

YES. Given the paraphyly of the genus (sensu lato), and the fact that merging bryantae and mitchelli into another, existing genus would require many more changes, resurrecting Philodice is the simplest solution that maintains the most stability of the group.

YES. Change required to keep genera monophyletic. 

YES. Change is required and well-supported by the data.

YES. The proposal is supported by the data.

YES.  The polyphyly in what we have left of Calliphlox requires the change.

YES. This is a sensible way to resolve the polyphyly of Calliphlox revealed from recent phylogenetic studies.

YES. Granted, there needs to be a full revision of the woodstars that would probably reduce the number of genera, perhaps dramatically. However, it is uncertain when that might happen, and it is highly untenable to keep these two taxa in Calliphlox

YES.  Required in a phylogenetic classification, and YES to retaining Woodstar. 

NO. This clade is barely out of the Pleistocene on average (minimum date occurs in the early Pleistocene; see Licona-Vera & Ornelas 2017: fig. 2, where they have misdrawn the beginning of the Pleistocene at 2.0 Ma rather than the correct 2.6 Ma). This is a perfect subgenus. In fact, going by Licona-Vera & Ornelas (2017: fig. 2) the entire clade of woodstars would be a single genus (Calliphlox) on a par with many other bird genera (e.g., genera formerly part of Anas), including the proposed regulid genus Corthylio elsewhere in this proposal set. I could live with a two-genus solution: that would span ~5 Myr of evolution and be a monotypic Calliphlox and the rest in whatever genus name has priority among the – yes – SEVEN other genera (this is getting insane, especially the number of monotypic genera). Yes, morphological diversity is high in hummingbirds, causing them to be long recognized as oversplit at the generic level relative to other groups of birds. When genera were based exclusively on phenotype, this was more excusable, but with increasingly strong molecular dating we can begin to move avian genera toward a more meaningful average. It’s a difficult category, especially given our longstanding commitment to using morphological together with genomic evidence to determine their limits. But that’s why they pay us the big bucks.

2021-A-13: Treat McKay’s Bunting Plectrophenax hyperboreus as conspecific with Snow Bunting P. nivalis

YES. Species recognized under the BSC should have little to no gene flow with other species. The UCE data from Winker et al. (2018) indicate substantial gene flow (i.e., ≥3 per generation in both directions) between hyperboreus and nivalis to the degree that hyperboreus is better treated as a subspecies of nivalis rather than as separate species. The lack of any fixed differences among thousands of UCE loci is also quite telling that there is likely extensive ongoing gene flow. While the two may be phenotypically diagnosable, it also seems like they approach a phenotypic continuum and are very similar in their life history and ecology. I think these are a single species under the BSC.

YES. Reluctantly, hyperboreus is clearly better treated as a subspecies.

NO. 1 without comment.

NO. This is a very, very challenging case for me, and I am very much on the fence. However, I ultimately vote NO to lump hyperboreus with nivalis. A couple of thoughts: 1) the proposal discusses ‘extralimital pairings’ between the two taxa, and in my mind, extralimital pairings do not tell us much about reproductive isolation. Hybridization has been noted in many cases when one taxa is much rarer than the other, as in the case of an individual pairing outside of its normal distribution (e.g., extralimital Glossy Ibis pairing with White-faced Ibis, Oswald et al. 2019); 2) I think the fact that the two species breed/migrate at different times, such that nivalis does not occur with hyperboreus when they are breeding is at least evidence for some prezygotic isolation; 3) the proposal notes that the extreme plumages of each (hyperboreus and nivalis) approach each other; however, the darkest male hyperboreus is still much lighter than the lightest nivalis; and 4) the lack of fixed differences in fewer than 4,000 loci is not terribly surprising to me in such recently diverged taxa, and to me is not necessarily sufficient to lump them. That being said, taken together, evidence for species status for hyperboreus is not strong. For me to overturn such a long-standing taxon though, more evidence is necessary one way or the other. 

NO. This one is difficult. When I look at the figure of plumage presented in the proposal, the two taxa look pretty distinct to me. I don’t see intergradation between the two, but maybe I’m missing something? However, with little genetic difference, this must be either a very recent split or lots of ongoing gene flow. Given the overall evidence, it’s likely a recent split. In thinking this over, I keep coming back to the plumage differences and temporal differences in breeding. I recommend continuing to recognize these as separate species, but further evidence or discussion could convince me.

NO. This is a questionable case, and I originally voted YES. However, I’ve read through other committee member comments and there is some merit to the case for keeping them as separate species for now (although the proposal also makes good arguments for making them conspecific). Ultimately, my vote is based on the fact that there remain unanswered questions about vocal differences and whether that would affect mate choice, and about whether the differences in migration timing may act as isolating mechanisms. If these are separate species, clearly it’s a recent event. Further study is needed for making a definitive case on lumping these taxa.

NO. Primarily for the reasons outlined by dissenting comments within the Committee, one member having been to St. Matthew for an extended period!. The earlier arrival on St. Matthew and Hall Islands (from Snow Bunting) of McKay’s and their potentially defending territories from later arriving or passing through Snow Buntings, thus preventing much interbreeding, fits my definition of a species. I recall that was one of the suggestions for why the redpolls didn’t interbreed. at least in parts of its range. Hoary Redpolls were already entrenched on breeding territories before Common Redpolls arrived (I recall this was on Baffin Island). Regardless it seems that one would see more intergrades if these were just subspecies. Perhaps something is being missed. The species isn’t that well known. Another NACC member is right about how widespread Snow Bunting is in the arctic north. It is certainly a very common species on St. Lawrence Island, a few hundred miles to the northwest, where a few Mckay’s are seen, mainly as migrants, occasionally in winter. I need to look at Jack’s paper again, but issues of SY year vs. AHY might have accounted for the increased numbers of perceived hybrids. Still, wayward McKay’s summering at places like St. Lawrence Island or the Pribilof Islands could easily pair with Snow Buntings given the inability to assortatively mate. It would be helpful to have a biologist stationed on St. Matthew during spring, one to see the arrival of breeding McKay’s, then to study interactions with Snow Buntings that arrive later. I wonder if there are territorial battles, or perhaps the Snow Buntings know they are in the transient lounge and will be moving on to everywhere else in the high arctic around the globe. 

As for the Saw-whet Owls, Lance Benner believes that some recordings of brooksi Northern Saw-whet Owls are distinctly higher pitched than the nominate subspecies. My view was that I heard NO DIFFERENCE in the calls of Unspotted Saw-whet Owl from Northern Saw-whet, and I listened to one call at close range for over an hour. This included the primary song (toots) and call notes. I see our treatment in this complex as inconsistent. And some Northern Saw-whets get to Haida Gwaii, at least in winter. None apparently stay and try to breed. A small group of us, including Lance, would like to visit Haida Gwaii in a year or so and get additional recordings for this partly marine owl. 

NO. What strikes me most is that hyperboreus has maintained a distinctive phenotype despite being surrounded by nivalis, two highly migratory taxa that live in a highly variable environment that should facilitate lots of exchange. These taxa have maintained their distinctiveness even though the environment around where they now breed has completely changed in the past 15K years. If reproductive isolation between the two taxa did not exist, then I would expect that hyperboreus would long ago have been swamped by nivalis.

NO. We treat as separate species many island forms with similar or lower levels of phenotypic divergence. Although the extremes of these forms can be similar, the ‘average’ birds of each species are quite distinct (especially males). I also note that there is a logical disconnect between the evidence and some of the other comments here: if gene flow is high, then hyperboreus and nivalis should not be phenotypically different. The opportunity for interbreeding over the climate cycles of the Pleistocene must have been large, yet they are currently functioning as isolated taxa.

NO. Although the mechanism is uncertain, these two must be reproductively isolated by BSC definition; otherwise, Snow Buntings would have swamped McKay’s long ago. How could this tiny population on just one of perhaps hundreds of Arctic islands on which Snow Buntings breed maintain its distinctive phenotype if there were no barriers to free gene flow? Speaking of which, the proposal reads as if the two taxa overlap in plumage, but the graphic presented, specifically to include extremes in plumage, shows no overlap as far as I can see, in both sexes, thus contradicting that claim. That there are occasional hybridization events is, to me, unconvincing and totally expected if the species ratios are highly skewed when they co-occur. That McKay’s has maintained its phenotypic integrity (based on the extremes presented in the proposal) in the face of that occasional hybridization could actually be used to support species rank. That they do not differ genetically at loci sampled so far could reflect that the genes controlling the plumage differences and the differences in timing of migration and winter range have not been sampled, and that the isolation is so recent that minimum differentiation at most loci is expected. As noted by the proposal, vocalizations have not been properly analyzed. As noted by Maley and Winker (2010), reproductive isolation could be maintained by differences in arrival dates; also, McKay’s winters virtually entirely on the shore of the Bering Sea in western Alaska (one of the harshest winter habitats of any passerine), where there are few Snow Buntings in winter, and thus have an East-West (actually SW to NE) migration route. Snow Buntings are longer-distance migrants that winter primarily south of the breeding range, and so on average must have a N-S migration route. These are small differences indeed, but perhaps enough to prevent free interbreeding. If these were sedentary taxa, then their allopatric distributions would be a hindrance to the evaluation of species limits. But these two taxa are highly migratory, and Snow Bunting has a circumpolar distribution that includes many islands that are as isolated as St. Matthew; in fact, I can’t tell for sure for some of the high Arctic Russian Islands but otherwise I can’t find any Arctic or subarctic islands not inhabited by breeding Snow Buntings, including every other Bering Sea Island except of course St. Matthew and Hall. That McKay’s can maintain itself as a distinct phenotype in this vast sea of Snow Bunting populations to me strongly suggests reproductive isolation. Snow Bunting is an exceptionally vagile species, as reflected in its minimal phenotypic differentiation across its vast circumpolar range. There are only four subspecies: nominate (W. Palearctic and Nearctic), vlasowae (E. Palearctic), townsendi (W. Aleutians and Commander islands; unlikely to be diagnosable because described on basis of average differences in measurements), and insulae (Iceland)

What worries me about my NO vote is that I also voted NO to recognize two species in the reasonably similar Saw-whet Owl proposal, so I wonder if I am being consistent. I think my reluctance on the Saw-whet split is because species limits in nocturnal birds are dictated by voie, and because plumage does not seem to make any difference in gene flow in owls in general, e.g., the color morphs in Megascops asio. If playback experiments show that there is discrimination, then I would change my vote immediately.

ABSTAIN. 2 without comment.