Linked paper: The mitonuclear compatibility species concept by G.E. Hill, The Auk: Ornithological Advances 134:2, April 2017.
A couple of years ago I was at the Bentsen State Park hawk watch tower with a dozen other birders when the topic of species boundaries came up. No one knew me or that I was a professional ornithologist—I was deep undercover as a mediocre birder—so I just listened to the conversation. Someone suggested that the “experts” changed species boundaries periodically so that new field guides would have to be printed and bought. The not-so-subtle suggestion was that the checklist committee was in the pockets of powerful publishers. Then it was suggested that researcher elites manipulated the system to get government money. Everyone quickly reached consensus that both of these views were correct. From that conversation, it would seem that scientists have not been very effective at conveying to the public the rationale for changes to the checklist. But maybe a skeptical public is not surprising given that some professional ornithologists are also not enamored with how species boundaries of birds are demarcated. This uncertainty regarding what is or is not a species could simply reflect the reality that at the tips of the branches of the tree of life, species boundaries are nebulous. But I don’t think that is correct. I agree with Ernst Mayr’s conclusions that species are real and fundamental biological entities. I’m convinced that the problem with the current approach to assessing species boundaries lies in a poor understanding of the process of speciation leading to an incomplete and confused species concept.
I lived the first fifty years of my life without offering any opinion about the process of speciation or species boundaries, and I had no plans to ever wade into the morass. It was therefore quite unexpected when I fell head first into the deep end of the speciation controversy by proposing a new definition for how to define an avian species, published as an opinion piece in The Auk. This new speciation concept arose not from phylogeographic studies—the genesis of all previous species concepts—but from the study of mating patterns in relation to colorful feathers (see Hill and Johnson 2012). Sexual selection and speciation have always been closely associated topics in evolutionary biology, but what I realized about five years ago is that both of these fundamental evolutionary processes emerge as a necessary outcome of the critical need for coadaptation between mitochondrial and nuclear genes and that mitonuclear coadaptation is the foundation of any comprehensive species concept. Before I explain what I mean by “mitonuclear coadaptation,” however, I need explain current dogma for recognizing species.
Whether it is explicitly stated or not, decisions regarding species boundaries are based on a species concept, and the Biological Species Concept currently being followed by the American Ornithological Society checklist committee was proposed by Ernst Mayr about 75 years ago, long before the age of modern genomics. As a matter of fact, the current species concept was written before biologists even knew that mitochondria have DNA that is inherited independently from the nuclear DNA.
It turns out the knowing about mitochondrial DNA is very important because mitochondrial DNA codes for components of the electron transport system, which is the biochemical machine that generates most of the energy for birds, mammals, and other complex life. But mitochondria only carry a few genes—not nearly enough to code for the entire electron-transport system—so much of the system is encoded by nuclear DNA, with the products moving over to the mitochondria to co-function with the products of mitochondrial genes. Having two distinct genomes coding for components that have to work together in a fully coordinated manner means that these genomes have to be coadapted; put another way, the genes in the mitochondria and nucleus have to coevolve so that their products are matched in physical shape and complementary of function. When populations become isolated, their mitochondrial and nuclear genes can coevolve to be different than the coevolved mitochondrial and nuclear genes of any other population such that they cannot be mixed without incompatibilities and fitness loss. What I propose in my opinion paper is that it is uniquely coadapted sets of mitochondrial and nuclear genes that define a species and that make hybrid offspring resulting from pairings between species less viable, thereby maintaining species boundaries.
Currently, it is not possible to assess the mitonuclear compatibility of any two populations of birds, although by simple extrapolation of advancing technology we will soon have such capability. What we currently have is information on the mitochondrial genotype of many species of birds, including many populations that are potentially distinct species. I propose that uniquely co-adapted mitonuclear genotypes necessarily involve unique mitochondrial genotypes. Thus, mitochondrial genotype becomes an excellent and available proxy for species boundaries.
Many biologists already look to mitochondrial genotypes when assessing species status. My theory simply provides stronger justification for placing more emphasis on mitochondrial genotypes and less emphasis on nuclear genotypes. Indeed, advocates of DNA barcoding assert that most species can be unambiguously diagnosed by mitochondrial gene sequence, and my theory potentially explains why.
As with any new idea, this hypothesis for the process of speciation and nature of species boundaries must be tested. Whether the mitochondrial compatibility species concept proves to be accurate or an overstatement, ornithology will benefit by considering mitonuclear interactions when they ponder species boundaries.