Nick Atkinson, "Speciation's Defining Moment" (2005)

Speciation's Defining Moment
Genetics and genomics enliven an old controversy on reinforcement | By Nick Atkinson
Evolutionary biologists, both theoreticians and empiricists, have argued for decades about the relative merits of two speciation scenarios: allopatry and sympatry. The multifaceted debate bristles with any number of sharply contested points, but one that has provoked the greatest polarization is the concept of reinforcement. This is the crucial last step in the process, when behavioral mechanisms finally become established, driving a reproductive wedge between incipient species.

Broad consensus has been reached on what constitutes a "good" biological species. The late Ernst Mayr defined species as "groups of interbreeding natural populations that are reproductively isolated from other such groups." ¹ Nevertheless, at that critical point when one species becomes two, the distinction often becomes blurred. The whole theory of evolution rests on the notion that species can split and diverge, but the precise mechanics of that process have been difficult to discern. Reinforcement has long been suspected to be key, but empirical evidence has lagged behind the algebraic march.

"Selection can act no further on inviable or sterile hybrids," says Nick Barton, professor in population genetics at the University of Edinburgh's Institute of Cell, Animal, and Population Biology. But the genetic underpinnings of mate preferences are a different matter: Reinforcement therefore allows speciation to be adaptive, rather than arbitrary. Because of this, reinforcement has occupied a "special place amongst theories of speciation," says Barton. Now, modern genetic and genomic methods are increasingly being brought to bear on this lingering theoretical debate, unearthing surprising new findings along the way towards a resolution.

DIVIDING THE BRIDGE Allopatry and sympatry, the two pillars of speciation, are separated by matters of geography. Allopatric speciation, favored by Mayr, occurs when a species' range is split by some accident of nature, such as a volcanic eruption, an ice age, or a forest fire. The newly isolated populations, thus prevented from exchanging alleles, become free to adapt to their particular ecological conditions. In addition to any environmental differences that might exist between geographically isolated refugia, runaway sexual selection, in which female preferences drive the evolution of male sexual ornamentation and/or behavior, might in time lead to significant divergence in the characteristics of the local population.

Mutation and genetic drift can also play a major role in fixing alternative alleles. Ultimately, without the constant mixing of a single gene pool, differences between populations might accumulate. Should such diverging groups be united once again, the differences between them might be so great as to prevent conception (prezygotic isolation) or the production of fertile offspring (postzygotic isolation).

Speciation in sympatry is a subtler process. Lacking a geographic division, the parental population splits as the result of some other factor, such as ecological niche separation. Threespine sticklebacks within a single lake, for example, have been shown to exist in large- and small-mouthed forms, each of which is a specialized feeder. ² An intermediate-sized mouth endows a fish with nothing but the worst of both worlds, providing a clear selective advantage to individuals that mate only with their own kind. That is exactly what happens. Big-mouthed forms from different lakes mate more readily with each other than they do with small-mouthed, more closely related, forms from their own lakes.

"We still don't know the genetic underpinnings responsible for the increased discrimination observed in sympatric versus allopatric species," says Louisiana State University's Daniel Ortiz-Barrientos, coauthor on a recent paper that begins to address this exact problem. ³ The reason is partly owing to the theoretical difficulties that reinforcement has raised, many of which are now resolved, even though occasional echoes of former debate can still be heard.

A MISUNDERSTOOD STEP Roger Butlin, visiting professor at the University of Sheffield, UK, was one of the most visible critics of reinforcement during the phase that finally led to its acceptance among theoretical biologists. ⁴ He says that it has often been misunderstood. "The basic idea seems so intuitively appealing: If random mating results in a high risk of producing unfit offspring, then it's better to mate assortatively," he says. The problem, at least historically, is the beguiling logic of the idea, making it tempting to interpret evidence for increased levels of assortative mating (or divergence in mating signals) in terms of reinforcement mechanisms, without either considering problems with the process itself or excluding alternative explanations.

Barton says the problems date back to differences of opinion between Charles Darwin and the coarchitect of the theory of natural selection, Alfred Russell Wallace. "Darwin argued that the differences that keep species separate are an arbitrary by-product of their divergence," says Barton. "Wallace, on the other hand, saw them as being somehow adaptive, having been shaped by selection." This underlines the difficulties that researchers address as they try to discern whether what can be observed is functional adaptation or merely the consequence of some other evolutionary process. Teasing apart cause and effect, and adaptation and byproduct, is seldom straightforward in evolutionary studies, particularly those centered on the crucible of diversity: speciation.

The key to understanding reinforcement is the point in the mating game at which it operates. The reduced fitness of hybrid offspring does, however, create a selective pressure for the divergence of mating and preferences, although genetic factors and opposing selection pressures limit the scope of the response. Reinforcement acts at the behavioral level, though, and underlies the adaptive advantage enjoyed by individuals that choose compatible mates. Selecting a mate with whom viable offspring can be produced is obviously an advantage over random, disassortative mating.

HIGH-FIDELITY FRUIT FLIES Unfortunately, most studies of the genetics of speciation have centered on hybrid sterility and inviability. The reason, says Daven Presgraves, a speciation geneticist at the University of Rochester, NY, is that these represent easier aspects to study than behavior. Presgraves says that making the link between behavior, speciation, and genetics is now the major challenge in the field.

Presgraves notes that the "stand-out paper" by Ortiz-Barrientos and his LSU colleagues, Brian Counterman and Mohamed Noor, addresses this challenge by providing the first high-resolution genetic analysis of reinforcement. The study demonstrates that the genes underlying the baseline mating discrimination observed in allopatric populations differ from those responsible for the stronger preferences seen in regions where populations overlap, which, says Ortiz-Barrientos, is the "typical signature of reinforcement. ³" The fact that two different sets of genes are involved is interesting, he says, because of the light they throw on existing models of reinforcement.

Ortiz-Barrientos and colleagues took advantage of the recently published genome sequence of Drosophila pseudoobscura to map the genes responsible for its behavioral mating avoidance of Drosophila persimilis, a sister taxon that occurs sympatrically with it in parts of North America. By narrowing the search down to particular chromosomal regions, the researchers discovered that the most likely reinforcement mechanism was olfactory: D. pseudoobscura females avoid mating with sympatric D. persimilis males because the males don't have the correct smell.

THE GENETICS OF SPECIATION Ortiz-Barrientos talks of the data vacuum that has plagued empirical studies of reinforcement. "Most of the controversy in reinforcement research comes from periods in which compelling empirical evidence was unavailable, or from theoretical considerations suggesting that reinforcement was very unlikely to exist." The increasing number of published complete genome sequences should change all this as the genome sequences of a dozen Drosophila species become available over the next year or so. Researchers predict that the impact of these data will be enormous, yielding information on a range of questions, such as what genes cause reproductive isolation, what their "normal" functions are, and what evolutionary forces drove their divergence.

"We now have a novel way of thinking about reinforcement," says the University of North Carolina's Maria Servedio, whose own work centers on the process of speciation. Reinforcement is generally believed to strengthen existing species-recognition mechanisms, by fine-tuning them to allow recognition of incipient species. It is now clear, however, that reinforcement can involve new forms of recognition, as the olfactory mode of discrimination in sympatric D. pseudoobscura females shows. Servedio says these revelations should help guide the development of new theoretical models, but she cautions: "We need to keep an open mind about the mechanisms behind premating isolation."

Servedio says that harnessing the power of modern genetic studies will lend crucial support in the effort to understand the nature of speciation. Identifying the genes responsible for behavioral adaptations that lead to the reproductive isolation of diverging populations will provide the basis for several new research directions: rigorous empirical testing of evolutionary theories of speciation; understanding reinforcement's molecular mechanisms and the natural variation among reinforcement genes in nature; and identifying the forces that drive mating discrimination. Such questions date to the earliest analytical attempts to make sense of life's endless diversity.

These are exciting times for speciation research, says Ortiz-Barrientos. "Speciation genes are the doors to many unanswered questions in evolutionary biology."

References
1. E Mayr "What is a species, and what is not?" Philos Sci 1996, 63: 262-77.

2. HD Rundle, D Schluter "Reinforcement of stickleback mate preferences: sympatry breeds contempt," Evolution 1998, 52: 200-8.

3. D-Barrientos Ortiz et al, "The genetics of speciation by reinforcement," PLoS Biology 2004, 2: 2256-63. [Publisher Full Text]

4. R Butlin "Reinforcement of premating isolation," Speciation and its Consequences (Edited by: Otte D, Endler JA). Sunderland Mass.: Sinauer Associates 1989, 158-79.