American Birding Podcast



Open Mic: Never Mind the Bullock’s – Taking Molecular Studies with a Grain of Salt

At the Mic: Steve N.G. Howell

Steve is a senior international bird tour leader for WINGS and has written several books including A Guide to the Birds of Mexico and Northern Central AmericaGulls of the Americas (with Jon Dunn), and most recently, Petrels, Shearwaters, and Albatrosses of North America.  He lives near Point Reyes, California.


There’s no denying it: Molecular studies are helping us unravel the mysteries of avian relationships, and quickly. But until we fully understand the finer principles of genetics, the history of genes, and what that all entails, we will continue to make mistakes in our interpretations of the results of molecular studies. Is there anything we can do about this? Perhaps we can start with a grain of salt, and if that doesn’t work then it’s
time to break out the lime and tequila.

A Tale of Two Orioles 

As an example of the tribulations involved in molecular studies, let’s review some recent developments with a time-honored favorite in the North American taxonomic arena – the Northern Oriole complex. The taxonomic tale of Baltimore Oriole Icterus galbula and Bullock’s Oriole I. bullocki is fairly well known – two species, then one, and now three (what, three?). Less well known to most ABA members, perhaps, is the Mexican cousin of Baltimore and Bullock’s – Abeille’s Oriole I. abeillei (unhelpfully named Black-backed Oriole by those seemingly unaware that 12 of 16 oriole species in Mexico have black backs, including both Baltimore and Bullock’s).

In 1999, a molecular phylogeny of New World orioles was produced, in which, supposedly, Baltimore Oriole and Abeille’s Oriole were closely related, while Bullock’s Oriole was more closely related to Streak-backed Oriole I. pustulatus (Omland et al. 1999, Molecular Phylogenetics and Evolution 12:224-239).

The authors realized how odd this looked, so a fuller genetic analysis was undertaken, which confirmed the earlier result (Kondo et al. 2004, Condor 106:674-680); it was then postulated that this was an amazing example of adult male plumage patterns diverging very quickly in an evolutionary time scale (between Baltimore and Abeille’s). A follow-up paper by the same team (Kondo et al. 2008, Evolution
62-5:1182-1191) went on to show that Abeille’s Oriole was derived from Baltimore Oriole (Bullock’s Oriole was ignored). This was thus a remarkable case of a tropical resident species evolving from a migratory ancestor, contrary to the  paradigm of resident tropical avifaunas tending to be the reservoir from which migration flows. Another paper used the new and “robust phylogeny” to explore the evolutionary history of plumage patterns in orioles (Omland & Lanyon 2000, Evolution 54:2119-2133). Another paper using the new phylogeny postulated that orioles colonized mainland South America from Caribbean islands (Sturge et al. 2009, Condor 111:575-579). And so on.

Ciudad Guzman, Jal, Mexico (11 of 37)
The adult male of Abeille’s (or Black-backed) Oriole (shown here) is a striking bird, although the female looks very similar to Bullock’s Oriole and the two species can be difficult to separate in female/immature plumage. Ciudad Guzmán, Jalisco, Mexico, March 2012. © Steve N. G. Howell.

Amazing stuff. But this is where the trail of genetic clues led, so how could it not be true? Hence, the AOU (2000, Auk 117:847-858) moved Bullock’s Oriole to follow Streak-backed Oriole, and thus separated it from Baltimore Oriole by five other species. Whiskey Tango Foxtrot was the response of many field ornithologists, but what did they know of genetics?

However, as knowledge of genetics increased, further molecular analyses of these orioles showed that the earlier grouping of Baltimore and Abeille’s was in error (Jacobsen et al. 2010, Molecular Phylogenetics and Evolution 56:419-427; Jacobsen & Omland 2010, Ecology & Evolution 2:2413-2429). Today it is believed that Abeille’s is more closely related to Bullock’s Oriole, which fits with older conclusions based on morphology, vocalizations, and biogeography. These two orioles are a good example of how messy the progress has been in applying new genetic tools to avian taxonomy, for only a few years ago these two were “among the most closely related of avian species” (Kondo et al. 2008:1183).

Along with a clearer understanding of “Northern Oriole” relationships, the claims of rapid plumage evolution and of a resident tropical species deriving from a northern migrant are also now debunked. And given that the baseline phylogeny was flawed to an unknown degree, is it still true (as claimed by Omland & Lanyon 2000:2119) that “plumage patterns and colors are highly labile between species of orioles?”

Still, after a litany of at least five publications it would appear that we have established something more firmly than before, and hopefully we have a better understanding of the evolutionary history of these three species. But these three species of orioles live in North America, where many people were familiar enough with the birds to notice that the genetic results were incongruous with intuition. For the other species, less familiar to most people, the original results of Omland et al. (1999) appear not to have been questioned. My opinion is that most of their results look reasonable, yet it is entirely possible that some of the other relationships “proven” by genetics and statistics are in error.

For example, Orange Oriole I. auratus and Streak-backed Oriole might be closer than reported. Either way, the unfortunate error involving Baltimore Oriole and friends does not instill confidence. Moreover, some well-marked, potentially species-level taxa were omitted from the analysis, despite “the importance of dense taxon sampling” in the title of Omland et al. (1999); an example would be Dickey’s Oriole I.  [graduacauda] dickeyae of west Mexico.

These molecular studies on relatively well-sampled orioles exemplify how little scientists really understand/understood about genetics. And these flaws went on to color or discolor a number of ecological and environmental studies and theories. Some might cry: “But that was then, this is now; today we really do understand things.” Certainly one would like to think that our knowledge of genetics has improved in the past few years, but how much?

What Do We Not Know?

Although some may claim otherwise, avian geneticists are still groping around in a recently tilled but dimly lit field, learning how to use their new tools. Thus it seems only common sense to treat their early harvests as working hypotheses, especially at the taxonomic levels of genus and  species. This is particularly true given that ornithologists can’t even agree on what constitutes a species. In this regard, check out the eloquent but sobering essay entitled “A species is whatever I say it is” by Nigel Collar in the March 2013 issue of British Birds (vol. 106:130-142), which highlights numerous other examples of the unresolved issues with genetic studies.

Although problems with genetic analyses may be fewer than those associated with studies based on non-genetic data, they are still problems. Despite this, taxonomic committees appear to have fallen in love with “glamorous” DNA studies, seduced by the promise of ultimate truth. Hence it often seems they are running amok on freshly plowed soil of uncertain viscosity, undervaluing or ignoring non-genetic tools that  might help them avoid getting needlessly stuck. One day, who knows, the genetic tools may be so refined that we won’t need any other lines of evidence, but there will be years or decades of messiness before we reach that point – if we ever do.

In the meantime, species and genera are constantly being shifted around hither and thither. Bullock’s Oriole is only the tip of the iceberg. Of course, birders only “suffer” from all this if they try to keep up with lists that are changing every week, or with using field guides compelled to adopt the latest and greatest changes – regardless of whether these changes are helpful for somebody trying to identify a bird. I can only think that poor Roger Tory Peterson must be rolling in his grave, now that the new Peterson field guides to birds try to follow taxonomic sequence – which is contrary to the simple brilliance of Peterson’s original system.

One cannot help but wonder how many more taxonomic decisions fueled or driven solely by genetics are simply errors in analyses. The merger of skuas into the genus Stercorarius, because Pomarine Jaeger is supposedly more closely related to Great Skua than to the other  jaegers? Putting Willet in between Lesser Yellowlegs and Greater Yellowlegs? Are all the recent wood-warbler genus reshuffles truly accurate, or might some be changed again? And how about them sparrows and towhees? And so on… How many things do we not know that we don’t know?

Thanks to Ned Brinkley, Burr Heneman, Alvaro Jaramillo, Peter Pyle, and Brian Sullivan for comments on an earlier version of this essay.

P.S. As of March 2013, the AOU check-list still separates Bullock’s Oriole from Baltimore Oriole by 5 species of orioles: Orange, Jamaican I. leucopteryx, Spot-breasted I. pectoralis, Altamira I. gularis, and Audubon’s I. graduacauda. Yet Yellow-backed Oriole
I. chrysater is 9 species distant from Audubon’s, despite the phylogeny of Omland et al. (1999), which placed Yellow-backed and Audubon’s as each other’s closest relatives (a view that is supported by morphology, voice, and biogeography, pending analysis of the unsampled taxon dickeyae, mentioned earlier). Time to pass the tequila…