Translate to: Español | Français | Português

AOU Committee on Classification and Nomenclature (North & Middle America)

Votes on Proposals 2009-B

2009-B-1: Split the Elepaio (Chasiempis sandwichensis) into three species
2009-B-2: Change the English name of Puffinus (=Ardenna) gravis
2009-B-3: Recognize the parulid genus Leiothlypis
2009-B-4: Recognize the parulid genus Oreothlypis
2009-B-5: Recognize the parulid genus Parkesia
2009-B-6: Recognize a new scientific name for the Blue-winged Warbler
2009-B-7: Split Icterus dominicensis into four species
2009-B-8: Separate Melanitta americana from M. nigra
2009-B-9: Transfer Oceanodroma monorhis from the Appendix to the Main List

Split the Elepaio (Chasiempis sandwichensis) into three species

YES - 4 without comment.

YES. This is a nicely done proposal that presents a good case for splitting the three island populations into different species (congruence of differences in phenotype, mtDNA, and song). Because the populations are allopatric and song recognition is not 100% (e.g., Hawaii Elapaio responded to songs from Kaua'i), we can only only infer that they would be reproductively isolated if they came into contact. Nonetheless, I think the combined evidence is sufficient to make this split.

YES (weakly). Not sure if there is anything different here than what we voted before. Molecular divergence is hardly surprising given the isolation and don't have the BNA reference at hand to check on song differences. I wonder if the calls differ too? But splitting seems consistent in what we have done with other different island groups (e.g. Amakihis), including from the Hawaiian Islands. I can hardly agree with the comment that molecular evidence does not suport subspecific recognition with the Elepaios on the big island of Hawaii - since when was that a prerequisite for subspecies recognition?! If the plumage differences are consistent and this reflects the different ranges on the island then what is the problem?

YES. The three populations appear to be diagnosable in morphology and mtDNA, and show some differences in song recognition. Not mentioned in the proposal is the lack of reciprocal monophyly in the sole nuclear marker examined in the genetic study — but the observed sharing of common LDH haplotypes across the three populations is not at all surprising, nor inconsistent with the observation that the faster-evolving and more rapidly coalescing mtDNA marker has diverged. The proposal itself is particularly nicely crafted; I hope that we get an increasing number of such robust proposals from submitters who are not on the committee.

YES, but barely. The playback responses are not symmetrical – although the Hawaiian taxon responds more strongly (N=9) to itself than to Kauai songs, the difference cannot be significantly different -- the SE’s are widely overlapping and the means barely differ. Reciprocal monophyly, in my opinion, is interesting but of no direct relevance to species limits, at least in allopatric, especially insular, populations.

NO. Although I suspect the suggestion is probably correct, the proposal relies heavily on three areas of evidence that are each individually fairly weak for determining species limits: reciprocal monophyly, genetic distance, and playback experiments in an oscine passerine. On these grounds alone I do not accept the proposal; why they are not subspecies is not adequately addressed (especially given the apparent desire to synonymize bryani and ridgwayi in sandwichensis). It seems rather de-emphasized in the proposal that the fairly well-marked subspecies on the island of Hawaii enable a comparison among subspecies. In classic comparative biology (in which we infer what might occur in allopatric forms by examining closely related sympatric forms), the fact that these well-marked subspecies interbreed enables inference that the other two allopatric forms would interbreed also if the opportunity occurred. Also, the goal of the proposal is a revised taxonomy “to make it congruent with phylogenetic units and biologically distinct populations,” both of which can and do apply to subspecies as well as species. The proposal glosses over these inconvenient facts that are integral to taxonomy under the biological species concept. In short, the full dimensions of the taxonomic problem are not addressed.

My objections to the strongest evidence presented here are that cultural and neutral genetic evolution in allopatric forms predict exactly what has been found in the genetic (mtDNA) and song data. If these forms were identical in plumage, we would not be discussing a taxonomic revision in this oscine passerine. Thus, too little is made of the plumage variation. I do not find the morphological evidence and conclusion of Vanderwerf et al. 2009 that there should be no recognized subspecies on Hawaii convincing. The abstract inference that subspecies within Hawaii were examined morphologically is not borne out in the text. The evolution of sexual dichromatism (likely just once) and the subsequent plumage diversification in different areas on Hawaii are major phenomena in this group that need to be better integrated into the genetic and song data so that we may better infer how allopatric forms would behave if in sympatry (Figure 2c in Vanderwerf 2007 and interbreeding between well-marked forms on Hawaii provide evidence against the proposal). Why aren’t there just two biological species?

The statement that “song could serve as a behavioral isolating mechanism,” when only males are tested (female responses to potential mates were not) and its divergence is not assuredly producing such a result even in males (Figure 2c in Vanderwerf 2007), is weak. Subspecies can also be “phylogenetic units and biologically distinct populations;” that descriptor is not exclusive to species. In sum, this problem needs to be addressed in the context of biological species with the appropriate evidence (especially how sympatric forms of this and closely related species behave) to confidently infer the degree of difference present in allopatric forms. The little evidence of that type in this complex is inconvenient with respect to a desire to split it into three species, and its importance is overly diminished.

(Aside: if this split is accepted, I recommend a re-ordering of the taxonomic sequence so that the Kauai form (sclateri) is first, given that it is the basal form.)

YES. This change has long been needed, and the proposal nicely puts together the case. These are surely equally good or better candidates for species status as a great many other allopatric taxa generally accepted as such. We are unlikely to get higher standards or more convincing evidence either way, at least not soon (before one or more taxa go extinct?), and for me the burden of proof is now on showing that they should continue to be treated as one extremely polytypic species. I agree that the lack of genetic differentiation within Hawaii should not be taken as proof that no subspecies should be recognized. Surely a careful morphological reanalysis would be necessary to establish this.

YES. I my opinion, Pacific Monarchs are horribly undersplit, even though there are two cases where fairly close relatives occur sympatrically (two spp of Mayrornis in Fiji, two spp. of Clytorhynchus in Fiji).


Change the English name of Puffinus (=Ardenna) gravis

YES - 2 without comment.


NO. This is really inconsequential. It has been Greater from the first edition of the Check-list, when it was P. major

YES - for the reasons given in the proposal.

ABSTAIN. I don't have strong feelings one way or another. On the one hand, I can understand the reluctance to change a name that has been in the AOU Check-list since the first edition. On the other hand, it makes sense to match the most commonly used name globally since this is a wide-ranging species.

YES. Howell makes a reasonable case here and in the spirit of being occasionally cooperative with our Old World counterparts, I favor this change.

YES. I agree with the points made in the proposal.

YES. A minor, low-cost change that improves world-wide standardization. I appreciate Andy’s point that the AOU has been using this name forever, and I’m generally opposed to name-meddling, but the cost in this case is minor (vs. changing, for example, the good name Red-backed Sandpiper, also in use from the first checklist, to the dumb name Dunlin).

NO. Admittedly, English usage in the Check-List is mixed here -- so stay with what is currently in use here.

NO. It’s Greater Shearwater in the 1st ed. of the AOU Check-List (1886), why change now?

YES - unless there are further considerations that aren’t summarized in the proposal.


Recognize the parulid genus Leiothlypis

YES - 1 without comment.

NO - 1 without comment.

NO. It is marginally better to have one genus than two.

NO. I vote yes on splitting this group from Vermivora but follow the suggestion of lumping them with Oreothlypis (proposal 2009-B-4).

NO, since the species that would be included therein can with at least equal justification be included within Oreothlypis.

YES. The phylogenetic results require at least one new name, and I agree with this admittedly arbitrary decision – in my subjective tastes on where to draw the line, I see little in common between the sedentary, strongly patterned, tropical latitude Oreothlypis and the dull, migratory, mostly temperate latitude Leiothlypis group.

YES. I find it hard to think of these dull, crowned ex-Vermivora as being very much like the two ex-Parula even if they are sister in the molecular studies. I think that considering this group as a distinct genus rather than lumping it with the rather different superciliosus and gutturalis makes the most sense.

NO (weakly). YES to the idea that this group should be split from Vermivora, but a weak NO vote to the name Leiothlypis for reasons related to proposal 2009-B-4. The evidence for separating Leiothlypis from Vermivora is sound, but there is a pure judgement call to be made relating to the name for the new group. One reasonable possibility is to recognize Leiothlypis, as proposed here. There is good evidence that these species form a monophlyletic group. The alternative possibility is to recognize a slightly more inclusive monophyletic group comprised of these “Leiothlypis” taxa plus their sister lineage, which comprises the taxa we currently know as Parula superciliosa (Crescent-chested Warbler) and P. gutturalis (Flame-throated W.); this is also a well supported clade. Under this second scenario, the genus name with precedence is Oreothlypis, as described in proposal 2009-B-04. There is really no right or wrong here, just a judgement call on whether genera should be more or less inclusive, and on whether the morphological distinctiveness of these erstwhile Parula are enough to separate them from these erstwhile Vermivora. I lean, but only slightly, toward lumping them together in Oreothlypis.

NO. I am really agnostic, but I would prefer fewer, most inclusive taxa, if they work with phylogeny.

NO. For reasons given in 2009-B-4.


Recognize the parulid genus Oreothlypis

YES - 3 without comment.

YES - with the provision that the species from proposal 2009-B-3 be included.

YES - but including the taxa proposed under Leiothlypis (proposal 2009-B-3).

YES - with the inclusion of the species in “Leiothlypis.” An expanded genus is less likely to be overturned at some point in the future.

YES - to the recognition of Oreothlypis under either the proposed scenario or the alternative where Oreothlypis also includes the “Leiothlypis” species currently placed in Vermivora.

YES - including Leiothlypis spp.

YES. I am really agnostic, but I would prefer fewer, most inclusive taxa, if they work with phylogeny.

YES - but to include those species in 2009-B-3, providing a more inclusive genus-level group in this morphologically rather similar group of parulids.


Recognize the parulid genus Parkesia

YES - 6 without comment.

YES - for reasons given in the proposal.

YES. Parkes would be happy as I believe he was the first to state how different the juvenal plumage of the Ovenbird was with the two waterthrushes.

YES - of course assuming that no preexisting name is available.

YES. With same comments as for Leiothlypis.

YES - although I wish another name had been chosen. I spoke with Ken about this approximately 20 years ago, and he agreed that a new name was needed given molecular and morphological evidence (the latter not dealt with fully by Sangster 2008).


Recognize a new scientific name for the Blue-winged Warbler

YES - 7 without comment.

YES - for reasons given in the proposal.

YES. And by chance it removes a real confusing misnomer. The species has little to do with pines over most of the range.

YES. I trust the thoroughness in making sure that all this is correct. [A trivial positive consequence is the removal of a misnomer pinus for this species and the removal of potential confusion with D. pinus.]

YES. Let's be sure that we have this right, however.


Split Icterus dominicensis into four species

YES - 2 without comment.

YES - finally.

YES (weakly). I am still not totally convinced (i.e., no "smoking gun") but I think the evidence is sufficient and will follow the recommendation.

YES. As pointed out, little has changed in terms of evidence, but I favored splitting before. Too bad there is not more detail about vocal difference, and none about call notes, which might be more significant than songs (less learned). I know I. northropi best and it certainly gave some call notes routinely that I did not hear from its closest relative (I. melanopsis), but my time with the latter subspecies was more limited. Still, the plumage differences between the two are striking, indeed the most striking within the Greater Antilles as a whole, including the Bahamas.

As for English name, I favor Northrop's Oriole for the Bahamas bird (indeed only on Andros now, no good records since the 1970's on Great Abaco). But it would be inconsistent to call it Bahamas Oriole in view of the English names we use for the other Bahamian endemics: Bahama Woodstar, Bahama Swallow, and Bahama Yellowthroat. So, first choice is Northrop's Oriole, 2nd choice is Bahama Oriole.

YES. I found Garrido’s et al.’s song descriptions rather not of much help with regard to the species question, but given the genetic, morphological, and some ecological distinctiveness of each group, I think it is best to treat them as species.

YES. The weight of the evidence now supports this split. I noticed that northropi (Bahamas) and melanopsis (Cuba) show very little mtDNA differentiation, yet these forms are quite distinct in plumage. The cleanest approach is simply to split them all.

YES - but with some reservations. The genetic data (mtDNA only) indicate that Icterus dominicensis is polyphyletic with respect to the Lesser Antillean group, although, despite the explicit statement in Sturge et al. that this is a case of polyphyly caused by incorrect taxonomy, not incomplete lineage-sorting, I still worry that the results represent just that.

The voice descriptions and sonograms are consistent with species rank, but are definitely not the sort of vocal data that would stand on their own in terms of taxon-ranking. I would disagree with dismissal of vocal data in oscines, however; yes, they are learned, but there is a strong genetic component to what they learn - -all experiments that I know of so far indicate that they inherit a template that strongly favors learning of their own species’ song. In fact, common sense suggests that it has to be the case – otherwise, oscines would sing all kinds of unrelated species’ songs learned while in their sensitive learning period.

I strongly agree on the lack of importance of the ecological differences – nowhere are these more flexible depending on local conditions than on islands.

Concerning the Price & Hayes paper, I don’t think it adds much. Has anyone questioned whether the taxa involved aren’t 100% diagnosable? If they had found that they were only, say, 98% diagnosable, would that change anything? How many specimens are necessary for determining whether two taxa are “100% diagnosable” in an analysis in which N plays such a pivotal role? Not only is morphological diagnosability not a sufficient criterion for species rank, but it is also not a necessary one – otherwise Willow and Alder flycatchers and the two wood pewees would not be ranked as species. That said, those with access to collections should check the three of the proposed species-level taxa: melanopsis and portoricensis are very similar to nominate dominicensis, and this is not a case of overly enthusiastic lumping. They are so similar that the Raffaele et al. field guide only illustrates one of the three; they differ only in slightly greater extent of yellow on lower back and undertail areas. Thus, it is indeed a relief to me that Price & Hayes demonstrated 100% diagnosability.

YES - but with separate comments on each of the two proposals.

2009-B-7. NO. The first proposal effectively follows Price and Hayes (2009), which recommends this split based on phylogenetic and evolutionary species concepts. There appears to be little effective addressing of the problem under the biological species concept; the proposal equates diagnosability with full species status (the statement “diagnosability equivalent to sympatric continental species” is misleading – while true, I am genotypically as diagnosable as a chimpanzee [100%] from my neighbor; that does not make us separate species; within the text of that paper they do better, however). Because diagnosability is also a characteristic of subspecies and even populations, I am not convinced that a change is required. (I am also troubled when conservation drives taxonomy, as appears to be occurring here.) The second proposal is much better than the first because it wrangles with these difficult problems (notes below). But I ask: Why are they not 2 or 3 biological species? Melanopsis andnorthropi appear to be young allopatric taxa and in most ways identical to many allopatric subspecies of birds.

2009-B-7. YES. The preponderance of evidence and the outlining of the difficulties of its interpretation are here well considered. Differences in juvenal plumage between melanopsis and northropi are very interesting, and, while I agree that there is no smoking gun, I will follow the recommendation and vote in favor. This vote trumps my vote on 2009-B-7a and demonstrates that how a proposal is written and evidence interpreted can make a big difference.

YES. I am not completely satisfied with the data here, but the arguments in the Rising version of this proposal convince me that splitting these up is the best option.

YES. I voted for this the last time around!

Regarding the taxonomic utility of oscine vocalizations: SOME species learn PARTS of their repertoires, to be sure, but I would maintain there is often considerable taxonomic information in oscine vocalizations. Even canary breeds pass along genetically based songs!


Separate Melanitta americana from M. nigra

YES - 3 without comment.

NO - 1 without comment.

YES - for the reasons given in the proposal.

NO. I think that the acoustic differences could be important, but I would like to see some data on the region where their breeding ranges approach.

NO. I think we should continue to refrain from splitting these taxa pending molecular evidence on their distinctiveness and/or a better idea of what might be happening (in terms of bill morphology, call variation, assortative mating, etc.) in and near their potential zone of contact in Russia. Also, how robust is the geographic sampling in the morphological and acoustic studies, especially near the contact zone?

YES. The evidence published by Sangster in the Wilson Bulletin is compelling. And, Black Scoters really are talkative on the winter grounds, at least from the one area where Black Scoters are fairly numerous in California on the north side of Tomales Bay. And I routinely hear them on lakes in Alaska too. The calls I hear indeed do sound long. The fact that there are no known hybrids between the two races is also significant given that the ranges around the Lena River in Russian Far East aren't that far apart.

The ranges within nominate fusca and stejnegeri White-winged Scoters closely approach, but farther west in Russia, and there are no known hybrids, as is the case with stejnegeri and North American deglandi in North America where the Bering Sea separates them, but they do overlap in migration. But from my experience Black Scoter is the only real talkative scoter. I haven't heard much from Surf Scoters or deglandi White-wings, even on boreal lakes where they likely breed.

Given the wide separation of North American breeding populations within M. n. americana, I wonder if there is any genetic difference?

YES. This seems analogous to the Anas crecca/carolinensis split (which we did not accept), except that in that case the vocalizations are essentially if not entirely identical. Assuming that, as stated, the vocalizations are homologous and used in courtship, burden of proof shifts to demonstrating that they are not separate species. Note that, according to the proposal, there is no known contact zone.

YES. The differences in vocalizations appear to be stable across their ranges, and samples of americana from the eastern Russia and Japan are included. Although it would be better to zoom in a bit closer to the area of closest approach, there does not appear to be any signs that characters change in either species in this area. Courtship vocalizations and soft part colors and pattern appear to have great importance for species recognition in ducks.

YES - but with strong reservations. Although the vocal data are suggestive, I worry about N and the geography of sampling. And given the generally high tendency to hybridize in waterfowl, I wonder if a difference in bill color combined with a slight difference in voice is at the level of difference associated with assortative mating in other diving ducks. [Someone should do a similar analysis of Nearctic vs. Palearctic Somateria mollissima, in which the bill color and patterns differ more strongly than they do in these scoters.] What would genetic data show in a case of allopatry like this? We already know that they differ genetically assuming that the bill color and nostril shape differences are under genetic control. Would we be looking for reciprocal monophyly or evidence of gene flow in the areas where they approach parapatry? To me, it wouldn’t make any difference if they were a worse-case-scenario, e.g., total mishmash of haplotypes or indistinguishable, if the current level of phenotypic differences was at the level associated with lack of gene flow in parapatric duck taxa because the genetic data would simply indicate to me a recent split and the phenotypic data a split that was going to stay that way, i.e., independent gene pools henceforth.

NO. These seem like subspecies-level characteristics to me given that these are ducks. I agree that until we know what is occurring in the zone of closest approach/contact (e.g., assortative mating, lack of gene flow) these are better off treated as subspecies.


Transfer Oceanodroma monorhis from the Appendix to the Main List

YES - 7 without comment.

YES - as indicated from the motion.

YES - now incontrovertible.

YES. If Jon, Steve Howell, and others are confident that the photos adequately clinch the identification, then so am I.

YES. This is a very well written and supported proposal.