Portal de Periódicos da CAPES

Plant speciation – rise of the poor cousins

2003; Wiley; Volume: 161; Issue: 1 Linguagem: Inglês

10.1111/j.1469-8137.2004.00957.x

1469-8137

Loren H. Rieseberg, Jonathan F. Wendel,

Plant Reproductive Biology

In 1989, Coyne & Orr lamented that students of speciation were viewed 'as evolutionary biologists' poor cousins, doomed to eternal speculation about untestable theories' (Coyne & Orr, 1989). Fortunately, things have changed dramatically over the past decade. Advances in genetics, molecular biology, phylogeny reconstruction, and theory have led to a renaissance in speciation studies. The field has attracted numerous new empiricists and theorists, and major advances have been made along several fronts. Examples include elucidation of ecological character shifts associated with speciation, identification of genes that contribute to reproductive isolation, documentation of the porosity of reproductive barriers, new insights into genetic and genomic consequences of different modes of speciation, and the re-creation of naturally occurring species in the lab and glasshouse. Many of these advances have been particularly gratifying to botanists, because they confirm long-held hypotheses about the nature of plant species and speciation. Reviews in this Special Issue of New Phytologist highlight the progress made in plant speciation over the past 20 yr, and relate these discoveries to the work of early students of plant speciation, particularly Verne Grant (Box 1 Box 1 Verne Grant Through his more than 120 papers and 13 books, Verne Grant transformed our understanding of plant speciation. He started early. His first published paper, in 1949, provided the first credible model for sympatric speciation, in this case based on the constancy of pollinator behavior. A year later, he published a major review on the flower constancy of bees, establishing pollinator behavior as an important consideration in plants and heralding the exceptional work continuing in this area. Over the next decade, as a byproduct of his ongoing genetic, taxonomic, and pollinator studies in the Polemoniaceae, he authored influential treatments on the effects of chromosomal repatterning on adaptation, the nature of plant species, the role of hybridization in evolution, and the origins of agamospermous plants. These ideas culminated in his first major synthetic treatment, The origin of adaptations, which received the Phi Beta Kappa Award in Science in 1964. The 1960s also saw the publication of many, now classic empirical studies. These include characterization of mechanical isolating barriers in Salvia, artificial synthesis of new diploid and tetraploid hybrid species, and documentation of speciation by reinforcement. In 1971 he authored the first edition of what was perhaps his most influential book, Plant Speciation, and in 1975 he published the most important synthesis to date of evolutionary genetics in angiosperms, in a volume titled Genetics of flowering plants. These seminal monographs have been followed by even broader treatments of general evolutionary theory, including Organismic evolution (1977) and the Evolutionary process (1985). His research accomplishments have been honored by his election to the National Academy of Sciences in 1968 and to the American Academy of Arts & Sciences in 1975. He also received a Certificate of Merit from the Botanical Society of America in 1971 and served as President for the Society for the Study of Evolution in 1968. Although Verne Grant is now 87, his scholarship continues unabated. He continues to publish on the systematics and evolution of the Polemoniaceae, the theory and practice of taxonomy and cladistics, and the genealogy and history of the Edward Grant Family. ; see also the Commentary by Verne Grant on pp. 8–11 of this issue). Perhaps no other scientist has contributed more to our understanding of plant speciation or influenced more plant evolutionists, young and old – this issue honors his many empirical discoveries, conceptual advances, and masterful syntheses of the speciation literature. In many instances, the first step toward the formation of a new species is the differentiation of populations and races. Grant (1981) emphasized the importance of adaptation to environmental differences and the interaction between gene flow and local selection in accounting for the patterns of phenotypic variation we see in nature. Papers in this issue emphasize these same forces. For example, Nyberg Berglund et al. (2004) show that populations of Cerastium alpinum are differentiated through adaptation to serpentine soil and that serpentine adaptation has arisen in parallel during the postglacial colonization of Scandinavia. Latta (2004) takes a more theoretical approach, using computer simulations to demonstrate that diversifying selection generates greater divergence in quantitative traits than in the individual genes that underlie them. This is an important result, because it helps account for what is becoming a common empirical observation (greater divergence of traits than genes) in natural animal and plant populations. Among the many differences between now and 1981 has been the explosion of phylogeographic studies that reconstruct the evolutionary history of populations and races. The power of this approach is illustrated by Abbott & Comes (2004), who demonstrate the existence of refugia within Alaska that allowed the circumarctic plant, Saxifraga oppositifolia, to survive Pleistocene glaciations. Although Grant (1981) devoted a full section to the nature of species, this topic receives surprisingly little attention in this issue. Many of the papers discuss the evolution of reproductive barriers, however, suggesting that the biological species concept (or its close relatives) still provides the conceptual framework for most process-oriented studies of plant speciation. On the other hand, a quick scan of leading journals in plant systematics and taxonomy suggests that various forms of the phylogenetic species concept have attracted adherents among more pattern-oriented botanists. In this issue, Rieseberg et al. (2004) respond to the following, most frequent criticisms of the biological species concept: The ability to interbreed is symplesiomorphic. There is too little gene flow among populations of many species to hold them together. There is too much gene flow among other species to keep them apart. The authors note that the first criticism only makes sense if the phenotypic clusters recognized by naturalists represent monophyletic lineages, and this is demonstrably not the case for many of the clusters that have been analyzed. A solution to the second criticism may lie in the recognition that only very low migration rates are required for the rapid spread of advantageous alleles. Resolution of the third problem – too much gene flow – can be achieved through a conceptual shift, in which reproductive isolation is viewed in terms of individual genes or chromosomal rearrangements, rather than entire genomes. Most of the papers in this issue focus on the process by which new species diverge and become reproductively isolated. This may also be the area that has advanced the most since 1981. Thanks to the development of molecular technologies and quantitative trait locus (QTL) mapping methods, considerable progress has been made toward understanding the genetic basis of species' differences. Orr (2001) provides a recent survey of the literature, and the following tentative conclusions can be made. First, the number of genes that control species differences is highly variable. Second, there is compelling evidence that major genes sometimes play a role in the evolution of species differences (Bradshaw et al., 1998), but this is not always the case (Lin & Ritland, 1997). Third, strong epistasis is reasonably common in intraspecific comparisons, but rare in interspecific comparisons (Kim & Rieseberg, 2001). In this issue, Gottlieb (2004) extends our knowledge of the genetics of species differences by showing that differences between recently diverged progenitor–derivative species pairs are few and under simple genetic control. He argues that comparisons of recent derivative species with their progenitors are particularly valuable because divergence is minimal and the causes of speciation are therefore more apparent. Likewise, Rieseberg et al. (2004) reviews information about the direction of QTL effects to demonstrate that most of the phenotypic differences between species are caused by divergent natural selection. Considerably less is known about the fitness consequences of QTL differences in nature or the genes underlying them, but significant progress in this direction is presented by Lexer et al. (2004) on both fronts. They show that QTLs for salt adaptation increase survivorship in the habitat of the natural hybrid species, Helianthus paradoxus. Moreover, evidence is presented that implicates candidate genes that cosegregate with several of these QTLs. Another theme emerging in this issue is the attention given to divergent natural selection as a cause of divergence and the declining influence of models of speciation based on population bottlenecks and genetic drift. Indeed, two papers (Campbell, 2004; Levin, 2004) focus on the process of ecological speciation, in which reproductive isolation evolves as a byproduct of adaptation to different habitats. Levin argues that niche invasion is often facilitated by disturbance and that invasive species represent useful models for studying the early stages of ecological speciation. By contrast, Campbell shows how analyses of hybrid zones between species that have already diverged ecologically can be used to infer the selective forces that led to speciation in the first place – in this case a complex mixture of pollinator and habitat-mediated selection. Evidence of a primary role for selection also emerges from molecular phylogenetic studies. Pérez et al. (2004) demonstrate that two heterostylous species in Narcissus are independently derived. In Calochortus, another genus with dramatic floral differentiation, molecular phylogenetic evidence suggests that floral specialization is a secondary consequence of local diversification in habitat preference and associated pollinators (Patterson & Givnish, 2004). Igic et al. (2004) employ a completely different approach for partitioning the influence of selection vs. drift in evolutionary divergence. By analyzing sequence diversity and coalescence times at the self-incompatibility locus, they were only able to detect one instance of population restriction during the diversification of the Solanaceae, suggesting that population bottlenecks rarely facilitate speciation. Other evidence for the importance of diversifying selection in speciation derives from reports that most trait differences between species are maintained by selection and that divergence at genes causing reproductive isolation is driven by positive Darwinian selection (Rieseberg et al., 2004). The possible role of chromosomal rearrangements in speciation receives somewhat less attention in this issue than in Grant's Plant speciation. In part, this is a consequence of theory that has accumulated over the past two decades, which indicates that there is a paradox associated with chromosomal speciation. Chromosomal rearrangements that could contribute significantly to reproductive isolation (i.e. have strong negative effects on the fitness of individuals heterozygous for the rearrangement) are unlikely to become established in natural populations. Rearrangements with little effect on fitness can become established more readily, but will have little effect on reproductive isolation. Livingstone & Rieseberg (2004) offer a possible solution to this problem by suggesting that rearrangements act primarily as recombination modifiers, which aid in the accumulation of differences at genes within or linked to the rearrangement. This prediction is verified in a comparison of tomato and potato, in which the only rearrangement likely to have arisen in parapatry (chromosome 10) shows significantly more divergence at sequenced ESTs than do collinear chromosomes. Further evidence of an important role for chromosomal rearrangements in speciation comes from the Patterson & Givnish (2004) study of Calochortus, where it appears that chromosomal diversification has significantly enhanced species diversity by preventing gene flow between species from divergent clades, but with similar floral syndromes and ecological tolerances. Grant's most important conceptual contribution to plant speciation may have been his early recognition that the constancy of pollinators might frequently contribute to reproductive isolation and speciation. Thus, it is altogether fitting that many papers in this issue emphasize the role of pollinator isolation in speciation. For example, Kephart & Theiss (2004) employ manipulative experiments to demonstrate that pollinators contribute significantly to reproductive isolation among sympatric Asclepias species and that the pollinator isolation observed results from both mechanical and ethological differences. Hodges et al. (2004) discuss how differential pollinator visitation in Aquilegia formosa and A. pubescens results from divergence in floral morphology, concluding that floral isolation promotes reproductive isolation. Similarly, Campbell (2004) demonstrates that pollinator selection plays an important role in maintaining contemporary hybrid zones in Ipomopsis and may have initiated divergence between these species as well. By contrast, and as mentioned previously, Patterson & Givnish (2004) consider pollinator adaptation to be a secondary consequence of habitat diversification in Calochortus, rather than a primary cause of speciation. Grant (1981) devoted six chapters to natural hybridization and its products, indicating just how important he considered hybridization to be in plant evolution and speciation. However, there was surprisingly little hard evidence at the time for an important role for hybridization. This has changed dramatically in the past decade because of the widespread availability of molecular markers. Doyle et al. (2004) and Cronn & Wendel (2004) summarize a series of very detailed molecular phylogenetic studies of the soybean (Glycine) and cotton (Gossypium) genera, respectively. They show that the footprints of past hybridization are considerably more frequent than previously believed. In Glycine, for example, incongruence among the major genome groups in nuclear and chloroplast gene trees is best explained by ancient introgression events rather than phylogenetic sorting. The results are even more startling in Gossypium, where ancient introgressions are detected in at least 25% of Gossypium species. Although the widespread occurrence of ancient hybridization in Glycine and Gossypium implies an important evolutionary role, gene tree evidence alone is insufficient to prove that hybridization events were adaptive. Evaluating this possibility requires analysis of contemporary divergent populations that are experiencing secondary contact. In this respect, recent work on Iris is significant. Arnold and colleagues have demonstrated that the fitness of hybrid gene combinations varies widely, with some hybrid gene combinations having greater fitness than the parents in parental and/or novel habitats (Arnold et al., 2004). These findings are consistent with recent theoretical studies, which indicate that some hybrid genotypes are likely to be fitter than one or both parents, particularly when placed in novel habitats (Barton, 2001). Confirmation of the positive fitness consequences of some hybrid gene combinations is also illustrated by work in sunflower, where hybrid QTL combinations that result in increased fitness in a novel habitat have been identified (Lexer et al., 2004). Two other studies of hybridization illustrate its role as a window on the speciation process (Campbell, 2004), and a dispersal mechanism (Petit et al., 2004). As discussed earlier, Campbell (2004) used natural hybrid zones of Ipomopsis to infer the evolutionary forces that likely caused speciation. Petit et al. (2004) provide convincing evidence that asymmetric introgression between Quercus robur and Q. petraea is facilitating the colonization of the latter into regions already occupied by Q. robur. Our understanding of polyploidy has also seen major advances since 1981, as reviewed by Soltis et al. (2004). These include: The massive redundancy of flowering plant genomes, as shown by sequencing studies, much of which is the likely result of episodic, ancient rounds of genome doubling. Evidence that the recurrent formation of polyploids is common. Recognition that autopolyploidy is more frequent than previously believed and that it is associated with increased allelic diversity. That polyploid formation may stimulate shifts in reproductive biology. The extent and rapidity of genomic rearrangements and changes in gene expression in polyploids that potentially are visible to selection. The use of molecular markers to elucidate more precisely the parentage and reticulate histories of polyploid lineages. Point 6 is well-illustrated in the present volume by the reports of Ainouche et al. (2004) on the genus Spartina and Doyle et al. (2004) on Glycine. These and other studies cited in Soltis et al. (2004) herald a new era in our understanding of the genomic, genetic and ecological attributes of this important speciation mechanism in plants. This issue illustrates the spectrum of research being conducted in some of the leading research groups studying plant speciation and indicates that major shifts in thinking about speciation are under way. For example, it is clear that students of plant speciation have doubts about the importance of population bottlenecks and founder events in speciation, views that were widely held just a decade ago. Likewise, we have seen a shift in thinking about chromosomal speciation away from an emphasis on the fertility effects of the rearrangements and toward their role as recombination modifiers. There is a renewed emphasis on natural selection, particularly adaptation to different habitats, as the main force driving diversification, as well as a reduced emphasis on the necessity of geographic isolation in speciation. Fisher's infinitesimal model of adaptation, an important component of the neo-Darwinian synthesis, has been rejected by most students of plant speciation, but then it was never very widely held by botanists to begin with (Gottlieb, 1984). Evidence that hybridization and allopolyploidy are widespread and important components of species diversification is even stronger than it was 20 yr ago. Finally, the biological species concept continues to provide the framework for process-oriented studies of speciation, although this perspective is almost invariably enriched by phylogenetic underpinnings. Although this issue captures some of the important recent advances in the area of plant speciation, results from the next decade are likely to be even more dramatic. The primary reason for this is that many 'speciation' genes are likely to be identified over the next decade and their effects on phenotype, fitness and assortative mating will be determined. Patterns of molecular evolution at these genes should allow us confidently to assess the evolutionary processes responsible for the fixation of alleles with important effects on relevant aspects of the speciation process. Likewise, large genome sequencing projects will allow us to estimate rates and mechanisms of chromosomal evolution and estimate the role of gene duplication and gene silencing in speciation. Advances in phylogenetic methods and their ever-increasing application should make it possible to estimate more precisely the timing of speciation events and to estimate the frequency of different causes of speciation. Finally, we expect to see a true integration of molecular biology and ecology, which will connect genotypic and perhaps epigenetic changes with their functional effects in natural populations and ecological communities. Many of the papers highlighted here were inspired by the 11th New Phytologist Symposium, commemoratively titled 'Plant speciation' after Verne Grant's classic 1981 monograph and convened in June 2003 in Antigonish, Nova Scotia.