Speciation. Jerry A. Coyne and H. Allen Orr. Sinauer, Sunderland, MA, 2004. 545 pp., illus. $89.95 (ISBN 0878930914 cloth).
“Why are there are so many species and how did they form?” is a fundamental question in biology. Jerry Coyne and Allen Orr's book is the first major summary of the data and ideas on speciation since Animal Species and Evolution, Ernst Mayr's highly influential book published in 1963. Although several books have appeared in the interim, none has been as broad as Coyne and Orr's.
The first two chapters of Speciation, which provide the foundations for subsequent arguments, present an admirable discussion of the various concepts of species that have proliferated in the past 20 years. Coyne and Orr opt for a slightly modified version of Mayr's biological species concept: “Species are groups of interbreeding natural populations that are reproductively isolated from other such groups” (p. 30). Species are maintained by isolating barriers, or “those biological features of organisms that impede the exchange of genes with members of other populations” (p. 29).
The real strength of the book lies in these elaborations: “Groups of interbreeding natural populations are distinct species if either (1) their genetic differences preclude them from living in the same area or (2) they inhabit the same area but their genetic differences make them unable to produce fertile hybrids”; and “Distinct species are characterized by substantial but not necessarily complete reproductive isolation” (p. 30). Maintaining the distinction between definition and recognition of species in this way removes many of the problems that plagued earlier approaches. Equating speciation with the formation of isolating mechanisms defines evolutionarily independent entities, which are presumably the units of evolutionary phenomena. This approach will be unsatisfactory to systematists, who are using a different definition to ask entirely different questions, but it is probably the best approach to take in evolutionary biology. However, it ignores processes that lead to anagenesis (which might explain some of the divergence leading to isolation, because anagenesis can occur differently in various populations within species).
Isolating barriers include the premating and postmating isolating mechanisms discussed by others, but Coyne and Orr make a valuable distinction between postmating prezygotic barriers and postmating postzygotic barriers. This allows the inclusion of all of the recent work on cryptic mate choice and sexual selection, which can occur after mating, while keeping them distinct from classic postmating postzygotic barriers. Another refreshing improvement is the explicit inclusion of ecological factors in the list of isolation barriers. Ecological factors appear in both the premating and the postmating postzygotic sections (the first including habitat isolation, temporal isolation, and pollinator isolation, and the second ecological inviability, when hybrids cannot find an appropriate niche or mates). This is an important point because the evolutionary dynamics can be quite different between premating and postmating postzygotic barriers; for example, only postmating postzygotic ecological isolation could lead to reinforcement of isolation.
Ecologically minded readers will be amused by what Coyne and Orr call “ecological”—they actually mean factors other than conventional genetic ones—but this is understandable, given that they both work on ecologically intractable organisms (Drosophila). Readers should ignore this quirk and attend to the useful new insights that arise from this approach; ecological and behavioral factors are considered later in the book.
The authors do not define species as “reproductively isolated entities having sufficient divergence to permit their coexistence” (p. 35), because “coexistence of nearly identical species can be maintained by spatial and temporal fluctuation in resources, or by subtle and virtually undetectable differences in ecology” (p. 35). This fact, Coyne and Orr maintain, would make the “sufficient divergence” species concept untestable in practice. A related point is that ecological isolation can arise either directly, by divergent adaptation to the local environments, or incidentally, as a result of competitive divergence unrelated to the habitat differences, again making distinctions untestable. The authors' strongly empirical approach, which permeates the book, sometimes leads to inconsistencies. For example, they point out the important distinction between definition and recognition of species, but later they reject a valid definition because it does not lead to an ability to distinguish species in all cases. They do believe niche differences are necessary to aid the persistence of species, and point out that there is no necessary correlation between reproductive isolation and ecological differentiation.
Coyne and Orr advocate and then criticize various modes of speciation, presenting an excellent discussion of the recent literature and evidence. However, they made no effort to compare the conditions for each mode and say which predictions or properties are unique to each mode. For example, they list six critical conditions for, and properties of, allopatric speciation, but almost all of these apply also to parapatric speciation, and some apply to sympatric speciation. A table and chapter contrasting the modes and their conditions would have been very helpful. The authors eventually opt for allopatric speciation as the most common mode and as the null hypothesis, but mostly on the basis of testability and simplicity (although genetic drift, which they rightly discard on examination of the empirical evidence, is even simpler). Considerations of the commonness of speciation modes should, however, also consider the commonness of their necessary conditions; isolation by distance and geographic variation in ecological conditions is more common than absolute isolation, so on that basis parapatric speciation might be thought more common than allopatric speciation. But the strong overlap in properties makes them difficult to distinguish in practice, and Coyne and Orr decide arbitrarily to assume that allopatric speciation should be assumed as a null hypothesis.
After discussing the modes of speciation, the authors present several excellent chapters on various isolating mechanisms: ecological isolation, behavioral and nonecological isolation, postzygotic isolation and its genetics, polyploidy and hybridization, reinforcement, and genetic drift. This is a major advance over previous treatments because each subject is discussed in detail, excellent summaries of the empirical data are given, and we have a chance to consider the strengths and weaknesses of the evidence. It was particularly refreshing to find a chapter on ecological isolation, a subject that has been almost entirely neglected.
These chapters are useful also because they point out glaring holes in our knowledge. For example, the summary of adaptation to new environments revealed that no study has showed that such adaptation yielded behavioral preferences for the new environments, which would be critical for speciation. Another example: Although selection may increase prezygotic isolation as a way to decrease postzygotic isolation effects (reinforcement), selection could also decrease postzygotic isolation directly by purging the deleterious alleles involved. One strong, general, and important new conclusion emerges from recent work on the population genetics of speciation: Reproductive isolation in the form of postzygotic isolation can easily result from natural selection within species (chapter 4), and genetic drift is probably unimportant in most cases of speciation (chapter 11).
The only notable weakness of the book is that only two implications of sexual selection are mentioned: It can cause rapid divergence of mate choice and traits, and hence isolation, and it increases the likelihood of reinforcement (choice is driven by both the need to avoid postmating costs and by indirect selection on preferences from sexual selection itself). But there is more to sexual selection than this. In choosing mates, females try to get the best mate that they can find. This would result in (premating) isolation as an incidental by-product of sexual selection, because a mate of the wrong species (with postmating isolation) would not be as good a mate as one of the same species. This in turn implies that in concentrating on isolation and the biological species concept, we may have the wrong end of the stick, and that Coyne and Orr's dismissal of the mate recognition concept may be premature. We might come to different conclusions about speciation and species if we viewed it as a problem of mate choice.
Consideration of sexual selection also reminds us that sexual selection may run in different directions in different populations of the same species. This means that there can be geographic variation in mating preferences, as has been shown in many species; what is a biological species in one location may not be one, using the same criteria, in other locations. This could also be true even with a conventional model of speciation if newly evolved isolating mechanisms have not yet spread throughout the species range. An examination of the factors causing and effecting the evolution of mate choice and mating traits should clarify this problem, but it may also undermine our conception of species if they are more amorphous than we currently think.
The last chapter presents some interesting ideas and speculations about evolution above the species level. The discussion of species selection was particularly objective and welcome. I would, however, have enjoyed a final chapter with general conclusions and questions.
The book is excellent but frustrating. It is an excellent summary of the ideas and data, but the chapters are not well connected to one another and the work does not yield a clear conclusion about the process of speciation or what conditions favor it. In a way this is a good thing, because the subject was held back for 40 years by overly strong opinions. Nonetheless, it would have been useful to set up a series of general questions and predictions.
Readers will come away thinking that speciation research is a mess, but at least—thanks to Coyne and Orr—it is now an organized mess, and we can get on with figuring out how speciation works and when we should and should not see it.