Split or Get Off the Pot: Genetics and the AOU vs. Reality?

“Surely one of the core virtues of birding is its independent, even anarchic spirit” (Gordon, 2015).

The North American Classification Committee (NACC) of the American Ornithologists’ Union (AOU) comprises appointed members who pass judgment and make determinations on species limits, English names, and related issues for the birds of North America. To this end, the NACC proclaims the AOU Checklist as “the official source on the taxonomy of birds found in North and Middle America” (http://checklist.aou.org, accessed May 2015). The American Birding Association (ABA) Checklist follows the AOU, for the most part, but will that union endure? 

I often hear puzzled birders bemoan AOU decisions, perceiving them as seemingly inconsistent and idiosyncratic: Baltimore and Bullock’s orioles, split, lumped, split; Myrtle and Audubon’s warblers, split, lumped, in limbo; Xantus’s Murrelet split; Leach’s Storm-Petrel not split, and so on. There’s a very simple reason for this perception—the decisions are inconsistent and idiosyncratic! They are simply opinions, some with more baggage than others.

n a pelagic trip off California in the 1980s I heard somebody say of this species (a Scripps’s Murrelet, then a Xantus’s Murrelet): “Hey folks, this is an imminent split...” After all, it looks different, sounds different, and there’s no viable evidence of interbreeding with other taxa. Straightforward? Not so much. Not until a genetic study ‘confirmed’ the obvious in 2012 did the AOU split Xantus’s Murrelet into two species. Cordell Bank, Marin County, California, 5 August 2012. © Steve N. G. Howell

On a pelagic trip off California in the 1980s I heard somebody say of this species (a Scripps’s Murrelet, then a Xantus’s Murrelet): “Hey folks, this is an imminent split…” After all, it looks different, sounds different, and there’s no viable evidence of interbreeding with other taxa. Straightforward? Not so much. Not until a genetic study ‘confirmed’ the obvious in 2012 did the AOU split Xantus’s Murrelet into two species. Cordell Bank, Marin County, California, 5 August 2012. © Steve N. G. Howell.

Inconsistency? You don’t have to look far to find it. One of my favorite examples was the 41st supplement to the AOU Checklist (Auk 114:542-552, 1997). A lot of decisions made in that supplement appear reasonable, although several might not pass the present committee’s criteria. In that supplement, Scarlet-rumped Tanager was split into Passerini’s and Cherrie’s tanagers, based on a genetic study published in 1996. That ‘new-school’ molecular study of the 1990s would surely be viewed today as simplistic, even flawed, and reinterpretation of the data might even lead to rejecting the split. But who wants to go back and re-do a genetics study? Or even re-evaluate it?

In the very same supplement, however, Yellow-throated and White-naped brushfinches (which AOU calls Brush-Finches, the capital F wrongly implying they are ‘true’ finches, but that’s another subject…) were lumped. This was based on an old-school study published in 1978, based on comparison of museum specimens and a dated philosophy of species concepts—hardly consistent with the concurrent tanager split. A comparable genetic study of the brushfinches would surely have revealed at least two ‘species’—witness, for example, the recent description of a ‘new species’ of brushfinch from southern Mexico, apparently indistinguishable except by genetics (Navarro et al. 2013)! That’s right: no plumage, no morphology, no voice, nothing but a genetic barcode. Based on that approach, every isolated population could be treated as a species, and who’s to say that’s wrong? It’s just a different opinion.

The reason for such idiosyncrasy and inconsistency is very simple: decisions are made by that most imperfect of mammals—the human species, us. Egos, personal philosophies, nepotism, and numerous other factors ebb and flow as committee members come and go. Think of your state or provincial bird records committee, or any committee, of the personalities and politics involved; the AOU is no different. They’re only human and, as in any human endeavor, trends also come and go.

ccording to the AOU, this adult White-breasted Hawk Accipiter chionogaster (carrying a Black-headed Siskin!) is just a Sharp-shinned Hawk. Yet Bicolored Hawk is not lumped by the AOU with Cooper’s Hawk, nor Rufous-tailed Hawk Buteo ventralis (of South America) with Red-tailed Hawk, which would surely be the case if consistency were an issue. Near Tegucigalpa, Honduras, 22 March 2015. © Steve N. G. Howell.

According to the AOU, this adult White-breasted Hawk Accipiter chionogaster (carrying a Black-headed Siskin!) is just a Sharp-shinned Hawk. Yet Bicolored Hawk is not lumped by the AOU with Cooper’s Hawk, nor Rufous-tailed Hawk Buteo ventralis (of South America) with Red-tailed Hawk, which would surely be the case if consistency were an issue. Near Tegucigalpa, Honduras, 22 March 2015. © Steve N. G. Howell.

The trend that spurred this commentary is the strong impression held by me and others that today’s AOU seems to use genetic data as a crutch. They are reluctant to take a step without it, and in the process appear crippled, or at least seriously handicapped. Not surprisingly, then, the AOU lags well behind most of the world’s other taxonomic bodies on the road to reality. Potentially relevant data—such as voice, plumage, morphology, ecology—are usually relegated to “that’s interesting, but…” or are thrown into a genetics-based proposal as lip service to make it appear well rounded.

An example? Anyone with normal hearing and a basic grasp of biogeography could tell many years ago that the whip-poor-wills were two different species, but not until 2010 were they split by AOU because… you’ve guessed it, somebody finally did a genetic study. Of course, when the decision was published, AOU cited differences in voice, morphology, and even egg pigmentation, in publications dating back to 1964, but not till the genetics were published did the crippled body actually move forward with a long overdue split.

I agree in principle that genetics may hold the ‘ultimate answer,’ but I remain unconvinced that humans can yet conduct let alone interpret meaningful molecular studies, at least to divine species limits in birds. We still have some genetic distance to go (see Howell 2013, Never Mind the Bullock’s).

At the other end of the spectrum, the recent world checklist embarked upon by Birdlife International and the Handbook of the Birds of the World (HBW) team (del Hoyo & Collar 2014) explicitly avoids including genetic data in its scoring system to evaluate species limits. Genetic studies are potentially just as flawed as any other type of study, rarely if ever do two papers use the same methods such that they can be compared directly, and ‘bad’ genetic studies still get published.

he dramatic vocal differences between Northern Potoo and Common Potoo were published in 1978. Yet not till 1995 did AOU split these species by finally acknowledging expert opinions from field ornithologists, a trend that shortly thereafter started to dry up with the advent of ‘omniscient’ genetic data. Northern Potoo, San Blas, Nayarit, Mexico, 25 January 2013. © Steve N. G. Howell.

The dramatic vocal differences between Northern Potoo and Common Potoo were published in 1978. Yet not till 1995 did AOU split these species by finally acknowledging expert opinions from field ornithologists, a trend that shortly thereafter started to dry up with the advent of ‘omniscient’ genetic data. Northern Potoo, San Blas, Nayarit, Mexico, 25 January 2013. © Steve N. G. Howell.

For example, a recent genetics paper argued that Newell’s Shearwater should be retained as a subspecies of Townsend’s Shearwater (Martinez et al. 2015). ‘Retained’ is perhaps a disingenuous word—the rest of the world splits Newell’s and Townsend’s as species, based on differences in plumage, morphology, ecology, voice, etc. But now a genetic study argues the reverse, which could have implications for conservation, not just listing. In the view of some, peer review failed miserably for that paper, but these days anything can be published—even this essay.

Sadly, however, while it is closer to reality than the AOU checklist, the HBW approach to taxonomy is somewhat naive and inconsistent, and has been rightly criticized (e.g., Remsen 2015). Ideally, all data, including genetics, should be evaluated critically, and an objective decision made. If genetic data are unavailable, decisions can still be made using the best available information (a scientific maxim of which the present AOU appears largely unaware), rather than wait years, decades even, for someone to ‘do the genetics.’ Meanwhile, striking differences in owl vocalizations, for example, might be dismissed because of small sample size or some other excuse. Yet how can we argue against genetic distances from a small sample, from only one or two individual birds, even if we have no idea what these distances actually mean?

Although in theory anyone can submit a proposal to the AOU, summarizing data and arguing for a split, or lump, the present committee as a whole is—wittingly or not—sending a pretty clear message: if you don’t have genetic data, don’t bother us. (In fairness, I should point out that the AOU does contain people who are field-oriented, aware of reality, and vote against some genetics-based proposals, but they are usually in a minority.)

Of 10 proposed species splits among extant species submitted this year to the AOU (http://checklist.aou.org/nacc/proposals/current_proposals.html, 14 May 2015), 7 primarily revolved around genetic data, and likely would not had been suggested without it. (Some of these species have been recognized for years by non-AOU authors, however, without the need for genetic crutches.) The remaining 3 proposals involved trying not to be so woefully behind the times when it comes to seabirds, which is encouraging to see. So how did these proposals shake out?

Big bromeliad, small owl! The best available data today argue that Guatemalan Pygmy-Owl Glaucidium cobanense (shown here) is a species, but the split was rejected recently by AOU. Yet the 1997 AOU supplement split Least Pygmy-Owl into five species, based on the same type of non-genetic data dismissed by today’s committee. Other world checklists recognize Guatemalan Pygmy-Owl as a species, and the AOU stands alone in its entrenched conservatism. Chanal Road, Chiapas, Mexico, 3 March 2010. © Steve N. G. Howell.

Big bromeliad, small owl! The best available data today argue that Guatemalan Pygmy-Owl Glaucidium cobanense (shown here) is a species, but the split was rejected recently by AOU. Yet the 1997 AOU supplement split Least Pygmy-Owl into five species, based on the same type of non-genetic data dismissed by today’s committee. Other world checklists recognize Guatemalan Pygmy-Owl as a species, and the AOU stands alone in its entrenched conservatism. Chanal Road, Chiapas, Mexico, 3 March 2010. © Steve N. G. Howell.

Well, in the 56th supplement to the AOU Checklist (published in the July 2015 issue of Auk, the AOU’s magazine), 4 of the 10 proposals, including 2 that relied largely on genetic data, were rejected. The other 2 rejected splits may well be valid, however, and one of them (Hen Harrier) is widely split elsewhere.

Of the 6 accepted splits, 3 were long overdue and involved seabirds, although in one case the proposal was a triumph of sophistry over science—but I suppose the end justifies the means. Remarkably, Townsend’s Shearwater was split in spite of the bad genetics paper mentioned above (which was published after the original proposal was submitted), suggesting that the NACC retains vestiges of common sense. Of the final 3 splits, all were split or suggested as splits years ago (1989, 1995, 2002) by field observers, but they were not ‘officially’ split until, you’ve guessed it, a genetics study came along to painfully elaborate the obvious.

On the one hand, AOU blindness to real-life field data is great news. It means as birders we can spend more time watching and enjoying birds, and not worrying that our observations on striking vocal, plumage, or ecological differences might have any import. Cryptic species can remain cryptic, some may even go extinct without formal recognition, but who cares? On the other hand, some people might wonder if the highly conservative approach of the AOU is doing a disservice to ornithology, and to conservation? But again, who cares?

In the meantime, birders might discard any misapprehensions they have that avian taxonomy is a science. Which is fine, as most birders aren’t scientists and all they really want is a list to tell them what can be counted. Traditionally the ABA has used the AOU Checklist as its model, although I don’t believe ABA is legally bound to this course. It’s a convention, a choice. Many thinking birders these days choose instead to use the International Ornithological Committee (IOC) list, which is more realistic than the AOU checklist, and other options exist.

exican Hermit, originally described as a species, lumped years later for no good reason, and ‘retained’ as a species by Howell & Webb (1995). Now, 20 years later, a genetics study comes along and this species is ‘officially’ split again. Amazing. Near San Blas, Nayarit, Mexico, 10 March 2013. © Steve N. G. Howell.

Mexican Hermit, originally described as a species, lumped years later for no good reason, and ‘retained’ as a species by Howell & Webb (1995). Now, 20 years later, a genetics study comes along and this species is ‘officially’ split again. Amazing. Near San Blas, Nayarit, Mexico, 10 March 2013. © Steve N. G. Howell.

If the AOU continues on a heading beguiled by genetics, I suspect that sooner or later most birders, including the ABA, may leave the taxonomically constipated NACC to play alone in their ivory tower. So, if you want to start splitting the Warbling Vireos, the Marsh Wrens, the Leach’s Storm-Petrels, the Fox Sparrows, the Northern Pygmy-Owls, and so on, go ahead, the data are there. To paraphrase Democritus: Nothing exists except atoms and empty space; everything else is opinion.

 


References

Del Hoyo, J., & N. J. Collar. 2014. Illustrated Checklist of the Birds of the World Volume 1: Non-passerines. Lynx Ediciones, Barcelona.

Gordon, J. A. (2015, June). Birding Together. Birding 47(3): 8-10.

Howell, S. N. G., & S. Webb. 1995. A Guide to the Birds of Mexico and Northern Central America. Oxford University Press, Oxford.

Martinez, J. et al. 2015. Phylogenetic placement of the critically endangered Townsend’s Shearwater (Puffinus auricularis auricularis): evidence for its conspecific status with Newell’s Shearwater (Puffinus a. newelli) and a mismatch between genetic and phenotypic differentiation. Journal of Ornithology DOI 10.1007/s10336-015-1189-2.

Navarro, A. G. et al. 2013. A new species of Brush-Finch (Arremon; Emerizidae) from western Mexico. Wilson Journal of Ornithology 125(3):443-453.

Remsen, J. V. 2015. Review of HBW/Birdlife International Illustrated Checklist of the Birds of the World Volume 1: Non-passerines. Journal of Field Ornithology 86(2):182-187.