Speciation – a rebirth

Plant speciation – the 11th New Phytologist Symposium, St. Francis Xavier University, Nova Scotia, Canada, June 2003

Contemporary studies of speciation span several levels of biological organization and many subdisciplines within biology. For instance, they include identification of single gene changes that are responsible for hybrid inviability (Presgraves et al., 2003), analysis of physiological characters associated with habitat differences between sister taxa (Lexer et al., 2003; Rajakaruna et al., 2003b), phylogenetic reconstruction of character evolution and diversification rates (Hodges, 1997; Magallón & Sanderson, 2001), and study of genomic additivity in polyploids (Liu et al., 2001). The diversity of empirical approaches reflects the complexity of how species arise and are maintained through time. However, while species and an understanding of speciation play a central role in biology, a comprehensive description of the speciation process remains elusive. The 11th New Phytologist Symposium (June 2003) brought together scientists to discuss recent, substantial progress in plant speciation research and to honor Verne Grant for his long-lasting and profound conceptual and empirical contributions – by utilizing a diversity of methods and modern tools, significant progress is now being made and there is also a wealth of novel insights.

Genetics of speciation

Speciation is the evolution of reproductive and genetic isolation between populations. We have numerous conceptual models of the circumstances that might lead to this process occurring (Turelli et al., 2001) – what we lack is an understanding of the number and types of genetic changes that are required for populations to set off on independent evolutionary trajectories. Similarly, we are only beginning to describe the genetic basis of trait evolution and adaptation through natural selection (Via, 2002). In fact, the genetics of reproductive isolation and of adaptation are overlapping fields of inquiry (Orr, 2001). In many cases reproductive isolation is associated with adaptive phenotypic differences between species (e.g. niche differences), and comparisons of closely related species are often particularly interesting and tractable (e.g. outcrossing vs selfing species, and species with different pollinator syndromes).

Aided by genetic maps, researchers are estimating the genomic distribution, number, and effects of quantitative trait loci (i.e. the genetic architecture) contributing to reproductive isolation in plants (Bradshaw et al., 1998; Kim & Rieseberg, 1999; Fishman et al., 2002; Hodges et al., 2002; Lexer et al., 2003). It is probably premature to make general statements about the genetic architecture of species’ boundaries based on these empirical studies, except to note that a small number of quantitative trait loci explain a majority of the phenotypic variation between some species (Bradshaw et al., 1998). This observation is consistent with loci of large effect underlying phenotypic differences between domesticated and artificially selected species, but contrary to the classical view of adaptation proceeding by many genes of small effect (Orr, 1998a).

Further study of the genetic architecture of reproductive isolation and of species’ differences in general will yield a greater understanding of the roles of adaptation and genetic drift in shaping long-term evolutionary trajectories. For instance, the accumulation of data on quantitative trait loci has spurred the development of new theories to assess the contribution of selection and genetic drift to macroevolutionary patterns (Orr, 1998b). Using this theory and associated statistical tests, a recent review of empirical studies of plants and animals found that divergent selection is very frequently responsible for phenotypic differences between species (Rieseberg et al., 2002).

Localization of quantitative trait loci is also leading to studies of candidate genes and molecular mechanisms involved in reproductive isolation. Experiments could involve transgenes to validate the role of specific loci in functions that contribute to isolation, as has been performed in animals (Presgraves et al., 2003). Analysis at the level of single genes and molecular mechanisms will more precisely and conclusively describe the genetic architectures of phenotypic traits.

Compared to estimates of the number and size of effect of loci, less attention has been given to the genomic distribution of loci contributing to reproductive isolation. The distribution of loci is of particular relevance to naturally hybridizing taxa, for which the genomic location of factors subject to selection in hybrids will dictate what regions of the genome will be protected from or will experience introgression (i.e. maintenance of linkage disequilibria). Relevant data from sympatric European oak species were presented during the symposium (Caroline Saintagne, INRA, France). Loci responsible for quantitative differences in leaf morphology between pedunculate and sessile oak (Quercus robur and Q. petraea) lie within genomic regions that experience the lowest levels of interspecific gene flow at neutral genetic markers.

Beyond their importance to the genetics of speciation, these various approaches to studying genetic architectures complement efforts to describe the genetics of domestication and studies of gene exchange between domesticated or non-native plants and their wild, native relatives (Abbott et al., 2003).

Ecology and natural selection

Participants in the symposium repeatedly drew attention to the importance of ecological traits in speciation and the isolation that ecological divergence may bring about (Schemske, 2000; Schluter, 2001; Via, 2002). While this intuitive perspective is not new, the emphasis is required in part because disproportionate attention has been given to intrinsic, postzygotic isolating mechanisms and, in particular, to the genetic basis of hybrid male sterility in Drosophila. There are good reasons to believe that ecological and spatial barriers constitute the bulk of reproductive isolation for many species, with intrinsic postzygotic isolation playing a secondary role (Schemske, 2000; Turelli et al., 2001).

Two studies of the salt tolerance of plants in extreme edaphic conditions are good examples of current research that incorporates ecology into studies of speciation. The history and genetics of ecological differentiation are being studied in Helianthus paradoxus, a homoploid hybrid species that may have escaped gene flow from parental species by colonizing an extreme habitat, salt marshes (Lexer et al., 2003). Recombinant hybrid plants (BC2) from an initial cross of the parental species (H. annuus and H. petiolaris) were transplanted into a salt marsh and quantitative trait loci for ion uptake traits and survivorship were identified. Selection intensities on the survivorship loci were estimated to be of a magnitude sufficient to contribute to the isolation of a hybrid neospecies. Similarly, tolerance of extreme ionic stress plays a role in differentiation between two races of plants that have arisen in parallel within two closely related species of Lasthenia (Rajakaruna et al., 2003a). One race is associated with habitats with high soil ion concentrations and the second is restricted to more benign habitats. Differences in habitat preferences are matched by differences in ion uptake rates and tolerances under laboratory conditions, suggesting that races have diverged genetically for critical ecophysiological traits (Rajakaruna et al., 2003b). Because salt tolerant races have apparently arisen in parallel within two very closely related species, this system presents an opportunity to examine whether repeated evolutionary transitions are achieved in a consistent and possibly predictable manner.

Phylogenetics, hybridization and parallel evolution

Robust phylogenies form a basis for the study of speciation. Along with biogeographic information, they may provide tests for hypotheses about the roles of geography, ecology, and hybridization in the origin of species (Schemske, 2000). Phylogenies also provide a means to estimate diversification rates among different clades (Magallón & Sanderson, 2001) and test hypotheses about the rate of speciation within clades (Hodges, 1997; Wendel & Cronn, 2003).

Some recent phylogenies are challenging our understanding of species boundaries and limits to gene transfer. Allopolyploid cottons are thought to have arisen within the last million years as a result of polyploidization and intercontinental genomic transfer between Gossypium species (Old World to New World, Wendel & Cronn, 2003). While this finding is puzzling, it is perhaps less so than recent evidence of transfer of genes between monocot and dicot species (Bergthorsson et al., 2003).

For studies of speciation, one of the particularly significant outcomes of phylogenetic analysis is the discovery of taxa that have originated repeatedly and independently (Soltis & Soltis, 1993; Schwarzbach & Rieseberg, 2002; Rajakaruna et al., 2003a). As noted above, the parallel evolution of independent lineages provides evolutionary biologists with replicated natural experiments for tests of hypotheses about the origin of species (Schluter, 2001).


Given the amount of attention given to species concepts in the literature, some might be surprised to learn that none of the symposium presentations dealt with species concepts, except perhaps to lament their proliferation. Various reasons might exist for this omission, but it is clear that plant speciation will continue to progress significantly without a universally accepted species concept, and that students of plant speciation are content with the ‘biological species concept’ for many applications.

In recent years novel applications of various research methodologies and tools have produced long-sought empirical data to test hypotheses about the origin of species. For instance, the increasing resolution and broad applicability of genetic tools allow the construction of comparative genomic maps, and studies of genome and molecular evolution within the context of speciation; we can now search for molecular evidence that a species had a single or multiple origins; and we can use genetic data to study the basis of species-specific ecological traits. We will continue to accumulate information about speciation within individual taxonomic groups and certain ecological settings, but generalizations and a comprehensive understanding of speciation will require complementary studies of a wide diversity of plant taxa. Current empirical studies of the genetics, ecology, physiology and evolutionary history of plants hold great promise for the future.


Two years spent as a postdoctoral student in Loren Rieseberg's lab served as a form of special education that helped mitigate the author's zoological training and potential consequent retardation with respect to speciation and hybridization. Thanks to Loren Rieseberg, Jonathan Wendel, Holly Slater and David Garbary for a great meeting.

C. Alex Buerkle Department of Biology, University of Wisconsin-Eau Claire, Eau Claire, WI 54702, USA (tel +1 715 836 3383; fax +1 715 836 5089; email buerkla@uwec.edu)