Opinion: Birds of a Feather?

Taking into account the interaction of nuclear and mitochondrial genes in birds holds the promise of more objectively defining what constitutes a species.

By | March 10, 2017

Left: Blue-winged warbler | Right: Golden-winged warblerLeft: © istock.com/BirdImages; Right: © Jayne Gulbrand/Shutterstock.com

What defines a species? Because the boundaries between species can appear so fluid, pursuing such a question seems, at times, like academic esoterica—little different than discussing how many angels can dance on the head of a pin. But accurate species definitions lie at the heart of biological investigations and management of natural resources (e.g., the US Endangered Species Act). It is troublesome, therefore, that new information on the genetic structure of long-recognized species of birds could jeopardize their status as full species.

The problem, in a nutshell, is that the DNA of many familiar species of birds holds signatures of substantial exchange of nuclear genes with other bird species. Such gene exchange matters because, by decades-old definitions, it is the isolation of gene pools that defines species. Substantial genetic exchange raises questions about whether these populations truly constitute species.

A case-in-point concerns the blue- and golden-winged warblers, two beautiful and very distinctive little songbirds that have long been regarded as separate species. A recent study, however, showed that these two “species” share more than 99 percent of their nuclear genes—much more gene sharing that we would expect between full species. There is evidence that this shared nuclear genotype is a product of regular interbreeding between individual golden- and blue-winged warblers for millennia. In other words, there appears to be poor isolation of gene pools between these two songbird species. These data are leading some biologists to question whether blue- and golden-winged warblers really are separate species. The golden-winged warbler is currently a threatened species receiving conservation attention, so losing status as a separate species would change not only birdwatching field guides, but also federal and state funding to protect threatened avian populations.

Even as variation in nuclear genotypes indicate fuzzy boundaries between [bird] species, differences in mitochondrial genotypes form distinct and abrupt boundaries between species.

There are many other examples of iconic birds with uncertain species boundaries. The Baltimore oriole, namesake of a professional baseball team, was merged with other orioles to form the not-so-iconic “Northern oriole,” before a committee of the American Ornithologists’ Union voted the Baltimore oriole back to full-species status.

Most of the focus on the flow of genes between species concerns genes in the nucleus, which harbors the large majority of genes in all animals. But complex organisms also have a small set of genes in the mitochondrion that is distinct from nuclear genes. An interesting pattern is emerging repeatedly in genomic comparisons among closely related bird species: even as variation in nuclear genotypes indicates fuzzy boundaries between species, differences in mitochondrial genotypes form distinct and abrupt boundaries between species.

For instance, blue- and golden-winged warblers show a level of genetic differentiation in mitochondrial genotype that is typical of separate species. An abrupt transition in mitochondrial genotype exactly matches the abrupt transition in plumage color and song between these two species of warblers.

Time and again, the nuclear genetic boundaries between avian species are fuzzy while the mitochondrial genetic boundaries are clear. A possible implication is that mitochondrial genes play a special role in speciation.
In a paper published online on March?8 (The Auk, 134:393-409, 2017), I have proposed a new definition of species, which could explain these patterns of variation in nuclear and mitochondrial genes. I argue that what really makes each species distinct is a set of mitochondrial and nuclear genes—about 200 genes in total—that have evolved to work well together. I call this a “coadapted set of mitonuclear genes,” and present the idea that it is uniquely coadapted sets of mitochondrial and nuclear genes that define a species. Because most nuclear genes are not coadapted with particular mitochondrial genes, the large majority of nuclear genes can move across species’ boundaries during hybridization events without compromising species integrity, forming the fuzzy boundaries between species in nuclear genotype. Mitochondrial genes and the small number of nuclear genes that interact with them are prevented from flowing into other species because mismatches in these coadapted genes produce low-quality individuals that do not persist.

In my paper, I applied this mitonuclear compatibility species concept only to birds, but there is no reason to believe that speciation in birds is fundamentally different than speciation in other complex animals. Birds provide an ideal test case for species concepts because we have detailed knowledge of their distributions and patterns of interbreeding, and because the traits that birds use in sexual signaling—coloration and song—are conspicuous to human sensory systems. Also, because of their very high respiration rates (the body temperature of most birds is around 105 °F), birds are the animal group that is likely to suffer the highest fitness costs when mitonuclear genes are mismatched across species boundaries, as the loss of function is targeted to cellular respiration.

The good news about the mitonuclear compatibility species concept is that it makes specific predictions that can be tested with the burgeoning databases on both the nuclear and mitochondrial genotypes of birds. If the hypothesis is supported in birds, then it can be applied to other vertebrates and to other animal groups.

A species concept that can be objectively applied to define species would be a boon to taxonomists and conservation biologists alike. And it is somehow reassuring that the instincts of 19th century naturalists regarding the boundaries between species might be confirmed with 21st century genomic analysis. 

Geoffrey E. Hill is a professor of biological sciences and curator of birds at Auburn University in Alabama.

Add a Comment

Avatar of: You



Sign In with your LabX Media Group Passport to leave a comment

Not a member? Register Now!

LabX Media Group Passport Logo


Avatar of: eratosignis


Posts: 1

March 13, 2017

I think it most unlikely that this is a general "species concept", and the attempt to "objectively define" species in this way seems to me to be considerably less general than one based on nuclear-nuclear incompatibilities or "Dobzhansky-Muller incompatibilities". And personally, I don't think even that one works very well.

There are also many examples of considerable mitochondrial genome introgression among species which suggest a lack of strong interactions between the nuclear and mitochondrial DNA. There are also plenty of examples of pairs of sympatric species that show very little postzygotic isolation at all; they appear to do it mostly with mate choice and ecologically divergent adaptations -- the Darwin's finches are an obvious case in point.

The best definition of species I think is one based on sympatry, which I believe Dobzhansky, and particularly Mayr were working towards with their "biological species concept." Too bad that they put the accent on "reproductive isolation." Reproductive isolation is more like the mechanismof speciation, rather than the result of that process. Evolutionary biologists mean by separate species taxa that have the ability to overlap in sympatry and remain distinct in multiple phenotypic traits or loci. One can add the rider that they must be reproductively isolated enough to remain distinct in sympatry, but this is obvious from the simpler definition based on remaining distinct.

Reproductive isolation seems easy to test, and Dobzhansky in particular believed that this kind of postzygotic isolation was testable, but in fact it is not at all clear how hybrid sterility or inviability relates to an ability to overlap spatially without a pair of taxa collapsing to form a swarm. For this reason it's always going to be somewhat arbitrary how we decide what status allopatric forms have. How *much* reproductive isolation or mitonuclear incompatibility does one need?  It depends on a lot of other things like ecology and behaviour that are not to do with the incompatibility.


Avatar of: Memphis100


Posts: 2

March 13, 2017

A complex explanation may not be required.  Mitochondrial DNA is passed from the mother to all offspring.  If a female Golden-Wing mates with either a male Golden-Wing or Blue-Wing, both male and female offspring will have the maternal Golden-Wing mitochondrial DNA.  Similarly, all offspring of a female Blue-Wing will have maternal Blue-Wing mitochondrial DNA regardless of whether the parent male was Golden-Wing or Blue-Wing.  What is required is that the two populations were isolated from each other long enough for two distinct mitochondrial DNA patterns to arise.  Based on the extent of the difference in the two patterns it should be possible to estimate how long the two populations were separated from each other.  At a later time, the populations were no longer isolated and they began to interbreed, so that the nuclear DNA became very simliar, but the two distinct maternal mitochondrial DNA patterns persisted.  It seems that at one time the two were distinct species, and opinions can differ as to whether the ability to interbreed makes them no longer distinct.

Avatar of: Barry@DataSense


Posts: 18

March 14, 2017

We still have much to learn about population genetics. An individual or a species is much more that the cohort of genes they posess, and will depend on other factors like gene expression, epigenics, the environment and, I would suggest, the almost infinite possible outcomes from the interactions between these.

As an analogy, take a piano, which has a given number of notes (musical genes if you like). Most pianos would belong to the same 'species'. But once you add chords, or change the beat, or the tempo, you get symphonies, sonatas, lullabies, rock tunes etc. 

My gut feeling is that simply measuring the number of shared genes between individuals (or groups of individals ? = populations, or species, or sub-species) is FAR too simplistic to be a meaningful measure. I think we're losing sight of the value of 'traditional' taxonomy which has become 'unfashionable'.

March 15, 2017


Geoffrey Hill has made a valuable contribution to our understanding of mitonuclear interactions. However, for reasons relating to species classification using mitochondrial barcodes (as set out in a BioRxiv preprint), I do not share his views on the role of mitochondria in speciation:

Forsdyke DR (2017) Base composition, speciation, and why the mitochondrial barcode precisely classifies. (http://biorxiv.org/content/early/2017/03/14/116814). 

Furthermore, in his blog (March 10th) the author of a textbook on speciation is sceptical of a report on Hill's work as reported in the Christian Science Monitor: (https://whyevolutionistrue.wordpress.com/2017/03/10/defining-species-a-new-but-problematic-species-concept/ )

Popular Now

  1. A Coral to Outlast Climate Change
  2. Science Celebrities: Where Are the Women?
  3. First In Vivo Human Genome Editing to Be Tested in New Clinical Trial
  4. Understanding Body Ownership and Agency