How Species Arise
Speciation. Jerry A. Coyne and H. Allen Orr. xiv + 545 pp.
Sinauer Associates, 2004. $89.95, casebound; $54.95, paper.
Charles Darwin was so convinced that species arose as an outcome of
adaptation driven by natural selection that he did not consider
alternatives. He also gave surprisingly little attention to species
formation, apparently considering it to be an inevitable consequence
of natural selection operating on diverse organisms in a
heterogeneous world. The 1930s and 1940s saw the publication of two
books focused on species: Theodosius Dobzhansky's Genetics and
the Origin of Species, and Ernst Mayr's Systematics and
the Origin of Species. Jerry Coyne and Allen Orr's new
book, Speciation, is clearly inspired by Dobzhansky's.
Coyne and Orr largely ignore the work of those who actually
identify, delimit and describe species. Nevertheless, their book,
which offers a critical analysis of the enormous literature dealing
with speciation, is an impressive achievement of great depth and
broad scope. It will be required reading for those studying species
and species formation.
The authors are major contributors to research on speciation. They
are Drosophila geneticists, and their work has taken
advantage of the experimental opportunities provided by those model
organisms and their relatives. In addition, Coyne and Orr have made
important contributions to many theoretical aspects of speciation.
Their experience is evident in the strategy developed for this book.
The 12 chapters have predictable titles and cover predictable
ground. But the book is novel in the depth of its scholarship, the
exceptionally thorough treatment of central questions and the superb
use of published literature. The references fill nearly 50 densely
packed pages, and most of the literature cited is recent, much of it
from the last 10 years, including a number of yet-unpublished
papers. This inclusion of recent material suggests that the field is
progressing so rapidly that the book will have a short life, but
what will surely last is the analytical and integrative treatment of
what we know now.
The authors' strategy is to select a topic and present a thorough
introduction that goes into considerable depth. Next they review
theory, treating some topics briefly and others in great detail.
They consider experimental evidence and, where appropriate, survey
evidence from nature. Topics that have an enormous
literature—for example, allopatric speciation—are broken
into subtopics (such as "concordance between present or past
geographic barriers and genetic discontinuities within
species"), which are then treated in detail. The logic of
organization and presentation is excellent, and the quality of the
writing is high.
Although the authors open with a vigorous argument for the reality
of species, they admit that "we lack the rigorous studies
needed to convince skeptics that nature is discontinuous."
Their discussion of species concepts is a spirited argument in favor
of biological species, viewed from many perspectives. They do
consider objections, often in depth. However, the second chapter
("Studying Speciation") makes clear that their research
program envisions the evolution of reproductive isolation as the
central, and nearly the only, issue. Coyne and Orr take the view
that complete cessation of gene flow between incipient species is
not required for speciation to occur; this gives them much latitude.
They emphasize "isolating barriers," arguing that although
some are likely to be more important than others, only comparative
analyses can determine which.
Allopatric speciation is the favored mode, and Coyne and Orr give it
the most attention, but they also consider parapatric speciation and
sympatric speciation. The sharpest critiques in the book are of
sympatric modes, which the authors consider to be possible but
probably of minor importance. Coyne and Orr note that
"parapatric and sympatric modes of speciation require that the
evolutionary forces causing populations to diverge must be stronger
than the gene flow causing them to fuse," whereas there are few
roadblocks to allopatric speciation. However, in lengthy chapters on
various kinds of isolation, the authors make clear that they have
high standards for evidence. The examples are especially rich,
although fruit flies, stickleback fishes, sunflowers and monkey
flowers are invoked repeatedly. The most fly-heavy chapter deals,
predictably, with the genetics of postzygotic isolation; here Coyne
and Orr focus attention on their research relating to hybrid
sterility and inviability. Their analysis of the number of genes
required for speciation concludes that "total
postzygotic isolation often involves many genes," although only
a few genes may be necessary to initiate such isolation.
Polyploid speciation (based on chromosomal dysfunction),
reinforcement (in which incipient isolation is enhanced by natural
selection when populations become sympatric), and natural selection
versus genetic drift are all considered in analytical detail. Coyne
and Orr argue their positions well. They maintain that polyploidy
has been overemphasized, even in plants. Although reinforcement has
been widely debated, they conclude that at least it is possible
("Put bluntly, theory said it could not happen until the data
said it probably did"). Founder-effect speciation (isolation
arises by dramatic genetic divergence in a newly established
population—as might occur, for example, when an island is
populated by a few individuals from the mainland) is unlikely, as
theory has shown, and Coyne and Orr argue that data provide little
evidence that various controversial models operate in nature. They
see selection as playing a far more important role in speciation
than does drift.
A final chapter on macroevolution is rather anticlimactic and very
different from the other chapters. As always, the analysis is deep,
and the standards for evidence are high, but in this area we know
less than we think we do (about such things as the foundation for
punctuated equilibrium, for example). The authors accept
species-level selection but question whether it is a central issue
in establishing evolutionary trends.
Although there is very much to admire in this book, as a systematist
I am left unsatisfied. I specialize in biodiversity issues,
including the discovery and naming of new species, criteria for
recognition of species, and the generation of robust phylogenetic
hypotheses about the relationships of those new species with known
taxa. Delimiting species and studying speciation are very different
activities, but they are related. How to interpret geographic
variation remains a perplexing problem, especially when there are
breaks in geographic ranges.
In a 25-page appendix, species concepts are catalogued and
critiqued. This is a fair-minded, but opinionated, discussion that
still leaves some major issues unresolved. I was dissatisfied with
the treatment here of evolutionary and phylogenetic species
concepts, and with the general lack of a true lineage perspective
throughout the book. Many phylogenists and systematists reject the
biological species concept, and although Coyne and Orr try to
explain the reasons for this deep division between evolutionary
biologists, they fail to satisfy me. The division exists because of
disagreements over the reality of species and the criteria for their
recognition. In order for Coyne and Orr to pursue their research
program, species must be real. Yet they ignore the important
theoretical contributions of workers such as Michael Ghiselin and
Kevin de Queiroz. Although they do cite some of de Queiroz's work,
they ignore two of his more recent, influential papers ("The
general lineage concept of species, species criteria, and the
process of speciation" in Endless Forms, edited by D.
J. Howard and S. H. Berlocher [Oxford University Press, 1998] and
"The general lineage concept of species and the defining
properties of the species category" in Species, edited
by R. A. Wilson [The MIT Press, 1999]). De Queiroz argues that
biologists implicitly share a common species concept: that of
general lineages that differentiate through time for various
reasons. A species under this view is a lineage concept. Other
so-called concepts (including even the biological species concept)
become criteria for deciding what to recognize as species. Coyne and
Orr do not bother with such philosophical issues. Furthermore, the
publications they cite for evolutionary and phylogenetic species
generally predate the integrative papers of de Queiroz. For example,
in rejecting the phylogenetic species concept, Coyne and Orr
contrast "thin" (gene trees) versus "fat"
(species trees) branches, and choose to focus on such important
matters as genetic coalescence. But they ignore the critical role of
extinction in delimiting species and in giving the impression of
species reality when what has happened is that continuous but
geographically variable lineages have been broken up.
I study a controversial "ring species" complex, and my job
would be far easier had a little extinction occurred. I suspect that
many species are "born" with geographic variation and are
never integrated by gene flow—as evidenced by the
phylogeographic structure and high FST values (indicating
greater population divergence) that are characteristic of widely
distributed species, especially those with low vagility. One of the
lessons of the many allozyme and phylogeographic studies is that
genetic variation within species is so great that species-wide gene
flow is rare. This finding raises doubts about
"speciation"—rather than fragmentation (with
differing patterns of adaptation in the fragments)—being the
primary cause of biological diversity. For me the central question
is not "What are species?" but rather "What do we
want them to be?" Coyne and Orr are explicit on this point, and
although many will agree with them, many others will disagree.
Setting aside my objections, some of which are admittedly
philosophical, I highly recommend Speciation. I advise
buying the hardback version; although the paperback is well bound,
it is heavy and slippery, and this is a book you will want to read
and use. Coyne and Orr are to be congratulated on their very
well-researched, well written and integrated, and highly stimulating book.