Review TRENDS in Ecology & Vol.16 No.12 December 2001 693 thaliana and its wild relatives: a model system for ecology and evolution

Thomas Mitchell-Olds

The postgenomics era will bring many changes to ecology and evolution. diploids with eight pairs. Evolutionary Information about genomic sequence and function provides a new foundation and ecological studies have focused on A. lyrata ssp. for organismal . The crucifer and its wild relatives lyrata in , A. lyrata ssp. petraea and will play an important role in this synthesis of and ecology. We A. halleri in , and A. halleri ssp. gemmifera discuss the need for model systems in ecology, the biology and relationships in Japan. (Earlier literature refers to these taxa as of crucifers, and the molecular resources available for these experiments. The lyrata, Cardaminopsis petraea, C. halleri scientific potential of this model system is illustrated by several recent studies and Arabis gemmifera, respectively.) are in – interactions, developmental plasticity, comparative genomics interfertile within this group of closely related N = 8 and molecular evolution. taxa. Chromosome number is reduced to N = 5 in A. thaliana, so diploid mapping crosses with wild There is a rich history of field studies on many species relatives are impossible5. However, Arabidopsis in diverse environments in plant ecology and suecica is believed to be an allotetraploid derived from evolution. Molecular methods are sufficiently A. thaliana and A. arenosa. advanced that we can now study the function and Previous taxonomic treatments of Arabis do not evolution of that are important in ecology; for constitute a natural classification, because various example, those that control responses to temperature, distantly related species have been included in this drought, , and other environmental grouping2–4, and future taxonomic work is needed in challenges. However, studies in molecular ecology this area. Arabis refers to at least two distantly and evolution require large initial investments in related but natural groups: N = 7 species in North molecular technologies. To be cost effective, these America, and N = 8 species of Arabis sensu stricto in investments must be shared among multiple and North America. is the type laboratories and funding sources. Equally important, species of the Arabis, which is distantly related individual investigators can achieve more in-depth to Arabidopsis. This group of N = 8 Arabis spp. is biological understanding when they build upon found in Eurasia, East , northern California previous studies from a research community that and southern Oregon2,4. shares the tools and resources of model organisms. Approximately 50 named species of Arabis, such as Clearly, no single species or model system can address A. holboellii and A. drummondii, are found in North the diverse range of ecological and evolutionary America. This group has a basal number of N = 7 questions of current interest. Nevertheless, the , and might be reclassified as genus crucifer Arabidopsis thaliana and its wild relatives based on taxonomic considerations4. The provide a model system that has a vast array of group is closely related to Arabidopsis, and molecular tools1, genetic resources and biological reproduces by self-pollination, outcrossing or information that can be used to address fundamental APOMIXIS6,7 (see Glossary). Taxa from diverse habitats questions in ecology and evolution (Table 1). can be cross fertilized, so it is feasible to map quantitative trait loci (QTLs) that control local Systematics adaptation to biotic and abiotic factors. Arabidopsis thaliana belongs to the Cruciferae ( Agriculturally important crucifers include , Capparales). Over the past 40 million Brassica8 and Raphanus. Oilseed rape (canola) years, ~3000 crucifer species have radiated to many consists of varieties of B. napus and B. rapa. habitats worldwide2. Recent molecular and Horticultural crops include European , morphological studies have outlined systematic cauliflower and brussel sprouts (B. oleracea), Asian Thomas Mitchell-Olds Dept of and relationships across the Brassicaceae (Fig. 1), and vegetables such as Chinese cabbage and bok choy 2–4 Evolution, Max-Planck across the closest relatives of Arabidopsis . The (B. rapa), and radish (Raphanus spp.). Because Institute of Chemical genus Arabidopsis contains about ten species that are these weedy annuals are widely distributed in Ecology, Winzerlaer native to Eurasia, North Africa and North America. agricultural and other disturbed habitats, they have Strasse 10, 07745 Jena, Germany. The closest wild relatives of A. thaliana (Fig. 2) include featured in many ecological and evolutionary e-mail: [email protected] A. lyrata and A. halleri, which are self-incompatible studies. Rapid cycling genotypes are also useful for

http://tree.trends.com 0169-5347/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0169-5347(01)02291-1 694 Review TRENDS in Ecology & Evolution Vol.16 No.12 December 2001

Table 1. Examples of ecological and evolutionary questions that can be recombination is apparent in nucleotide addressed using model systems polymorphisms segregating among individuals. Questions Refs Species of Arabidopsis and Arabis have evolved a • What genes are responsible for ecologically important variation? 17 broad range of life-history, developmental and Why are they polymorphic? physiological traits that adapt them to a wide variety of Are quantitative trait locus alleles young, deleterious and spatially restricted, habitats. Combined with molecular tools and biological or are they old widespread polymorphisms contributing to local adaptation? information from A. thaliana, researchers can address • How often do disease resistance genes evolve via an arms race of sequential 18 many ecological questions in these species. Although SELECTIVE SWEEPS, versus ancient polymorphisms maintained by A. thaliana is common in agricultural fields and BALANCING SELECTION? • What ecological processes influence breeding system evolution and the 6 disturbed sites in the temperate zones, it is also found frequency of outcrossing, inbreeding, or asexual reproduction? in undisturbed rocky sites and forest openings • Is plasticity adaptive? 37 (T. Mitchell-Olds and M. Clauss, unpublished). Do developmental responses to environmental conditions increase fitness? Significant isolation by distance across its native range Have these responses been shaped by natural selection? suggests that populations of A. thaliana in Europe and • Can expression profiling identify candidate genes that influence insect 42,44 9 resistance, drought tolerance, flowering time and other ecologically predate human agricultural disturbance , hence important traits in natural populations? they have long ecological and evolutionary histories in Does natural selection promote adaptive evolution of these loci? association with insects, pathogens, competitors and • How do polyploids differ ecologically and physiologically from their 5 abiotic environmental factors. diploid relatives? Among closely related taxa, A. lyrata grows on How are these differences shaped by natural selection? sand dunes, riverbanks, cliffs, or forest understorey in • What is the genetic basis of ? 29,50 Does sexual selection of -stigma recognition genes influence sites with low densities of competing species (Fig. 3). ? By contrast, A. halleri thrives in perennial meadows What is the role of duplications in interspecific genomic incompatibility? under intense competition with grasses and forbs, as • What is the relative importance of changes in gene regulation versus 14,34 well as in heavy metal-contaminated sites. These function within and among species? short-lived perennial species inhabit primarily mesic Have these evolutionary changes resulted from neutral genetic drift or from natural selection? environments. Elsewhere, species of North American • Do neutral molecular markers predict patterns of ecologically important 51 Arabis inhabit and xeric grassland, whereas variation in rare ? other Arabis taxa grow in cold, wet, or high elevation What is the molecular basis of adaptively important polymorphisms in sites (Fig. 3). In some areas of western North America, endangered species? Arabis spp. can be studied in pristine environments that are largely free of human disturbance. laboratory research. Brassicas are the closest A broad geographical range is important for model agricultural relatives of Arabidopsis, so comparative systems in plant ecology, so that field populations will genomics should provide applied benefits. However, be readily accessible to many researchers. Arabidopsis because these species are ancient polyploids8, and its wild relatives fulfill this need3,12. Arabidopsis molecular studies can be far more complex than in thaliana is widely distributed in North Africa, Eurasia diploid Arabidopsis. and North America. is found in Canada, the American Midwest and East Coast, UK, Geography and environments Scandinavia, central Europe, and parts of Asia. A recent study of molecular markers in 142 accessions Arabidopsis halleri ranges from to Taiwan. sampled from the native range of A. thaliana found N = 7 Arabis (Boechera) spp. are found across North statistically significant isolation by distance9, America6, with vast diversity in the western USA. probably reflecting Pleistocene migrations across N = 8 species of Arabis sensu stricto occur in northern Europe and Asia. Although human disturbance California and across Eurasia. Finally, Brassica and clearly influences the biogeography of A. thaliana, related species are common in Eurasia and have been there is evidence for post-Pleistocene colonization of introduced to most temperate agricultural areas. Europe from glacial refugia near the Mediterranean Importantly, collections of many species and and in Asia9. Only a few of A. thaliana are genotypes are available from stock centers. available from central Asia, and these show tantalizing patterns of genetic diversity for molecular Genetic tools and biological resources and phenotypic traits. A model system for molecular ecology and evolution There is no evidence for different subspecies requires a diverse array of biological and molecular among the Arabidopsis ecotypes that have been resources that are freely available to the research examined9,10. Furthermore, studies of amplified community. Many microsatellite markers have been fragment length polymorphisms (AFLPs) and optimized for use in the wild relatives of A. thaliana13, nucleotide variation find evidence for modest, but and additional markers can be easily developed important historical recombination within and using genomic sequence information (M. Clauss, among genes, hence evolution does not unpublished). Bacterial artificial chromosome (BAC) proceed in a bifurcating, tree-like manner10,11. libraries are available from A. lyrata and A. drummondii Instead, evidence for historical outcrossing and (J. Nasrallah and T. Mitchell-Olds, unpublished),

http://tree.trends.com Review TRENDS in Ecology & Evolution Vol.16 No.12 December 2001 695

Fig. 1. Neighbor-joining distance tree based on matK and Chs 100 Ionopsidium prolongoi sequences2. Percentage 100 Ionopsidium abulense bootstrap values from Lepidieae 1000 replicates are shown Cochlearia danica above branches. Tribal Cochlearia pyrenaica assignments are given on Arabis alpina EUROPE the right. Approximate 100 divergence dates (million Arabis alpina AFRICA years before present) for nodes A–F are: A, 16–21; 95 E B, 13–19; C, 19–25; 100 Arabis pumila D, 10–14; E, 15–17; and Arabideae F,26–32. Several taxa 100 mentioned in the text are 96 73 not shown in this figure: Aubrieta deltoidea Arabidopsis arenosa and A. suecica are closely Aubrieta deltoidea F related to A. lyrata. Arabis turrita Brassica is near Raphanus and Sinapis. Finally, Microthlaspi perfoliatum Lepidieae Leavenworthia is related 100 to Barbarea32. 91 Reproduced, with Alliaria petiolata Sisymbrieae permission, from Ref. 2. 84 Sinapis alba 95 Brassiceae 100 Raphanus sativus Sisymbrium irio Sisymbrieae 89 Fourraea alpina 68 Barbarea vulgaris A 100 Rorippa palustris 100 Rorippa amphibia Arabideae Cardamine amara 100 Cardamine penziesii 100 Cardamine rivularis B Lepidium campestre Lepidieae 84 Arabidopsis lyrata ssp. petraea SWE 100 Arabidopsis lyrata ssp. petraea GER 100 92 Arabidopsis lyrata ssp. lyrata USA Sisymbrieae 100 88 Arabidopsis halleri Arabidopsis thaliana Turritis glabra 100 Arabideae 97 pumila C D 67 100 Olimarabidopsis cabulica Sisymbrieae Olimarabidopsis pumila rubella Lepidieae 87 Arabis lyallii Arabis parishii 100 100 Arabis drummondii Arabideae 100 Arabis drummondii Arabis divaricarpa Boechera Arabis lignifera 95 Halimolobus perplexa himalaica 100 Sisymbrieae Crucihimalaya wallichii Mathiola incana Hesperideae Aethionema grandiflora Lepidieae 0.05 Substitutions per site

TRENDS in Ecology & Evolution

http://tree.trends.com 696 Review TRENDS in Ecology & Evolution Vol.16 No.12 December 2001

Fig. 2. Wild relatives of Arabidopsis thaliana. (a) Arabidopsis lyrata ssp. lyrata. (b) Arabis holboellii. (c) Arabis fecunda. Figure 2c is reproduced, with permission, from Ref. 52.

providing large insert clones (>80 kb) for analyses of facilitate QTL mapping and progeny testing in gene families, molecular evolution and comparative advanced generation crosses, which can be hindered genomics. A linkage map comparing A. thaliana and by inbreeding depression in outcrossing species20. A. lyrata is approaching completion (O. Savolainen, However, many population genetic models assume unpublished). Levels of nucleotide identity are random mating, which is approximated more closely in generally high within the genus Arabidopsis populations of outcrossing species (e.g. A. lyrata and (>90% within coding regions)14–16, so carefully chosen A. halleri). Outcrossing species are predicted to PCR primers can be used to amplify most genes of maintain higher levels of genetic variation within interest. At greater evolutionary distances within populations21. This pattern has been observed among the Cruciferae, DNA or amino acid sequences can Leavenworthia spp. at the PGI (phosphoglucose be compared from other dicotyledenous families to isomerase) locus22, but is less clear when the ADH identify conserved regions suitable for primer design. () gene is compared between Genetic variation at known loci influences life A. lyrata and A. thaliana15. Data from additional loci histories17, pest resistance14,18, floral architecture19 are needed to resolve this discrepancy. Breeding system and other ecologically important traits. Consequently, also influences transposon abundance and mobility23, genes of interest are readily accessible from the close which, in turn, can affect genetic variation for wild relatives of A. thaliana. More information on quantitative traits. Consequently, breeding system can the wild relatives of Arabidopsis can be found at influence the genetic architecture of quantitative traits. http://www.ice.mpg.de/departments/Gen/wild.htm. Sequence variation at neutral genes can allow estimation of important population genetic µ Breeding system and genetic variation parameters, such as population growth rate, or Ne Breeding system influences experimental tractability and Nem, the products of effective population size and in A. thaliana and its relatives. High levels of self or migration24. Analysis of DNA sequence pollination (e.g. A. thaliana and A. drummondii) variation within and between species can compare observed patterns of variation with predictions of neutral equilibrium models. In Arabidopsis and related species, neutrality has been rejected by some studies (e.g. Refs 18,25,26) but not by others (e.g. Ref. 16). Several studies find patterns of genetic variation indicative of recent population expansion in A. thaliana16,27. A large study of nucleotide polymorphism at 400 loci (K. Schmid and T. Mitchell-Olds, unpublished) finds -wide evidence for population expansion, which corresponds to the large population size of A. thaliana resulting from human disturbance. Plants are ideal for molecular because they are sessile and are confronted by measurable environmental challenges, such as temperature, drought, pathogens and insects. Many known genes function in response to these environmental challenges, so these loci could be Fig. 3. Habitats of Arabidopsis and Arabis. (a) Populations of Arabidopsis lyrata grow adjacent to influenced by natural selection. Ecologically montane glaciers in the Pacific Northwest, USA. (b) Arabidopsis halleri grows under highly important polymorphisms have already been competitive conditions in perennial meadows, Thüringerwald, central Germany. (c) Arabis perennans grows in xeric prairie and environments, AZ, USA. (d) Arabis drummondii grows in meadow, identified at known loci in A. thaliana (e.g. Refs 17,18) forest and high-elevation habitats in the Rocky Mountains, USA. and corresponding genes can be isolated readily in

http://tree.trends.com Review TRENDS in Ecology & Evolution Vol.16 No.12 December 2001 697

Box 1. Expression profiling and the molecular ecology of plant defense

Expression profiling simultaneously quantifies issue can be approached using quantitative genetic expression (mRNA abundance) for hundreds or methods, it is desirable to understand the functional thousands of genes (Fig. I). Arabidopsis thaliana and molecular basis of plant defenses and putative ‘DNA chips’ with thousands of sequenced genes are costs. For example, manipulation of defensive available commercially and from public-sector physiology with plant hormones, mutants, or resource centers. Alternatively, expression of known transgenics can up- or downregulate the jasmonic- genes of special interest (e.g. drought or disease acid pathway. Several studies blots) can be assayed on nylon or glass at reasonable have shown that jasmonate-induced responses are costa,b with adequate replicationc. These tools can be costly, resulting in reduced growth or fitnesse–g. applied to many ecological questions, for example: However, jasmonates have extensive pleiotropic are so-called ‘inducible defenses’ actually expressed effects on growth and , and further under normal field conditions, even in the absence of research is required to determine whether fitness insects and pathogens? costs are attributable to either plant defense, or to Many researchers are examining possible costs of confounded pleiotropic effects of the jasmonate plant resistance to insect herbivoresd. Although this signal transduction pathwayd,g. Expression profiling (Fig. I) shows extensive crosstalk between the 3 jasmonate, salicylate, and ethylene response pathways, thus affecting many aspects of plant physiologyh. Consequently, manipulation of distal 2 defense mechanisms (downstream from signaling components responsible for crosstalk) is necessary 1 to determine whether costs are actually attributable to plant defensed. This implies that detailed 0 understanding of signal transduction pathways is required before costs of defense can be addressed by Ð1 either transgenic or physiological manipulations. References (response to jasmonate)

Changes in Ð2 a Reymond, P. et al. (2000) Differential gene expression in response to mechanical wounding and insect feeding in Arabidopsis. 12, 707–720 Ð3 b Sekia, M. et al. (2001) Monitoring the expression pattern Ð4 Ð2 0 2 4 of 1300 Arabidopsis genes under drought and cold stresses Changes in gene expression by using a full-length cDNA microarray. Plant Cell (response to ethylene) 13, 61–72 c Kerr, M.K. and Churchill, G.A. (2001) Statistical design and TRENDS in Ecology & Evolution the analysis of gene expression microarray data. Genet. Res. Camb. 77, 123–128 Fig. I. Changes in gene expression controlled by the methyl d Purrington, C.B. (2000) Costs of resistance. Curr. Opin. Plant jasmonate wound response pathway are positively correlated with effects of another plant hormone, ethylene. Schenk et al.h examined Biol. 3, 305–308 529 differentially regulated genes with repeatable expression levels e Baldwin, I.T. (1998) Jasmonate-induced responses are costly in control experiments. Application of ethylene or jasmonate but benefit plants under attack in native populations. Proc. hormones results in correlated changes in gene expression (r = 0.45; Natl. Acad. Sci. U. S. A. 95, 8113–8118 n = 529, P <0.001). Although these hormones influence gene f Agrawal, A.A. (1999) Induced responses to herbivory in wild expression in many aspects of , some of these radish: effects on several and plant fitness. Ecology responses have clear effects on plant resistance to pest speciesi. 80, 1713–1723 Data on expression levels of responsive genes was obtained from g Redman, A.M. et al. (2000) Fitness costs of jasmonic acid- http://www.pnas.org. Expression levels were log transformed, and induced defense in tomato, Lycopersicon esculentum. changes in response to hormone treatments (Z-scores) were Oecologia 126, 380–385 calculated as residuals from a third-order regression on expression h Schenk, P.M. et al. (2000) Coordinated plant defense levels in the control treatments. Of the genes, 529 had repeatable expression in replicated control experiments (−1.5

at first reproduction in A. thaliana17,19,36,37. -mediated light perception has received particular attention, because the red:far red light (R:FR) ratio provides an unambiguous indication of neighboring plants and community density37. As predicted, single locus mutants deficient in phytochrome activity display altered fitness responses to plant density36. Recently, a large manipulative study of A. thaliana genotypes from natural populations examined the adaptive significance of in response to environmental conditions. Dorn et al.37 measured plastic responses to light levels, R:FR, and plant density, and natural selection on phenology and growth form in these environments. Developmental changes in response to R:FR signals had beneficial effects on fitness, showing the adaptive importance of phenotypic plasticity to this environmental cue.

Insects and pathogens Arabidopsis and Arabis are attacked by a wide variety of pest species, offering potential for studies of , costs of plant defense and other topics in plant–insect and plant– interactions Fig. 4. Insects and pathogens of Arabidopsis and Arabis. (a) Pontia (Pieridae) consuming Arabis (Box 1). Many insect species complete their life holboellii in Montana. (b) Dipteran miner on Arabidopsis halleri. (c) Insect galls on on Arabidopsis relatives (Fig. 4), including the A. holboellii. (d) Puccinia infection of Arabis perennans causes elongation of pseudo emitting fungal spores. (e) Insect seed predators destroy many developing in Arabis populations. widely studied cabbage butterflies Pieris spp. QTL Fig. 4d reproduced, with permission, from J-Z. Lin. influencing insect resistance can be rapidly mapped in Arabidopsis38. Several leaf miners in the genus segmental duplications in the Arabidopsis attack A. thaliana and related species39. genome1,28,29. Extensive duplicated regions suggest Scaptomyza are closely related to virilis that a polyploidization event occurred ~100 million and have active P-element transposons40, offering years ago (Mya)28. Since then, progressive loss of potential for a plant–insect model system combining individual genes from duplicated segments has been a the genetic tools of Arabidopsis and Drosophila. fundamental feature of genome evolution. The half- Likewise, microbial pathogens such as Albugo, life of duplicated genes in Arabidopsis is estimated Peronospora, Pseudomonas and Puccinia are at approximately 3.2 My (Ref. 29). Continuing gene commonly found on wild and agricultural crucifers6,18. loss is opposed by individual gene duplications, Ecological models of plant–insect interactions have often in tandem gene families. During population suggested that perennial species might invest heavily differentiation, loss of gene function at alternative, in chemical defenses against insect herbivores, duplicated loci provides a plausible speciation whereas weedy and ephemeral species could lack mechanism consistent with existing models29. common mechanisms of plant defense, because they Several large genomic regions have been sequenced might escape detection by herbivores41. However, from relatives of Arabidopsis30,31. The genera Capsella recent molecular studies show that A. thaliana and Brassica diverged from Arabidopsis 10–14 and possesses typical mechanisms of plant defense that 16–21 Mya, respectively32 (Fig. 1). In comparison to have been documented in many plant families. For Arabidopsis, functional genes are highly conserved, example, ethylene, jasmonate and salicylate signal whereas intergenic regions display insertions, transduction pathways are conserved in A. thaliana, deletions and higher levels of nucleotide substitution. and influence plant resistance to insect and microbial Such sequence comparisons will be useful for genetic pests42,43. Genomic analyses have recently shown analysis of crop plants, and for annotation and that many familiar defense and signaling genes are regulatory studies of Arabidopsis genes33,34. induced by Pieris feeding on A. thaliana44. Likewise, Furthermore, evolutionary analysis of comparative are biologically active secondary sequence data can provide insights on the evolution of metabolites that are widely implicated in plant–insect gene families, the molecular basis of adaptation and and plant–pathogen interactions45–47. Several the rate of deleterious mutations35. biosynthetic and modification genes have recently been cloned in A. thaliana48,49. In Growth and phenology summary, A. thaliana employs pathways of plant Many ecological and evolutionary studies have defense that are widely conserved in many plant examined genetic variation for growth form and age families, hence Arabidopsis and Arabis provide

http://tree.trends.com Review TRENDS in Ecology & Evolution Vol.16 No.12 December 2001 699

representative models of physiology and evolution of Glossary plant–pest interactions in herbaceous dicots. Apomixis: asexual reproduction, producing seed progeny identical to the maternal genotype. Conclusion Balancing selection: natural selection that maintains a Arabidopsis thaliana and its wild relatives provide polymorphism at a stable equilibrium. Examples include Acknowledgements a model system with a diverse array of molecular heterozygote advantage and frequency dependent selection. I thank M. Clauss, tools, genetic resources and biological information. Ecotype: in evolutionary biology, ecotypes are genetically D. Kliebenstein, differentiated and locally adapted populations. However, in the J. Kroymann, K. Schmid, Crucifers are geographically widespread, and are Arabidopsis literature, ‘ecotype’ is synonymous with ‘accession’, J. Schultz and four adapted to a wide range of biotic and abiotic and does not necessarily imply local adaptation. anonymous reviewers for environments. Molecular tools can identify genes Selective sweep: when selection fixes a favorable mutation, comments. Figures 2c and that evolve in response to temperature, drought, linked, neutral polymorphisms can also become fixed via hitch 4d are generously hiking. Thus, nearby polymorphisms are swept away during provided by J-Z. Lin and herbivores, pathogens and other environmental selective fixation of advantageous . D. Siemens, respectively. factors. Such approaches can elucidate the ecological This work was supported and evolutionary forces influencing quantitative by the Max-Planck- genetic variation in the wild, and the history and genomic sequence can provide insights on the Gesellschaft, the US National Science geographical distribution of ecologically important evolution of gene families, the molecular basis of Foundation (Grant alleles. Recent studies of plastic responses to light adaptation and the rate of duplication and loss in DEB 9527725), the quality have shown the adaptive importance of gene families. The synthesis of organismal biology European Union, and the Bundesministerium für environmentally mediated changes in plant and genomics will change our understanding of Bildung und Forschung. development. Evolutionary analysis of comparative ecology and evolution.

References 13 van Treuren, R. et al. (1997) Evolution of balancing selection in the Leavenworthia PgiC 1 The Arabidopsis Genome Initiative (2000) microsatellites in Arabis petraea and Arabis locus. Genetics 153, 1423–1434 Analysis of the genome sequence of the flowering lyrata, outcrossing relatives of Arabidopsis 26 Kawabe, A. et al. (2000) DNA polymorphism at plant Arabidopsis thaliana. Nature 408, 796–815 thaliana. Mol. Biol. Evol. 14, 230–238 the cytosolic phosphoglucose isomerase (PgiC) 2 Koch, M. et al. (2001) Molecular systematics of the 14 Bishop, J.G. et al. (2000) Rapid evolution in plant locus of the wild plant Arabidopsis thaliana. Cruciferae: evidence from coding plastidic matK chitinases: molecular targets of selection in plant- Genetics 156, 1339–1347 and nuclear CHS sequences. Am. J. Bot. pathogen coevolution. Proc. Natl. Acad. Sci. 27 Innan, H. and Stephan, W. (2000) The coalescent 88, 534–544 U. S. A. 97, 5322–5327 in an exponentially growing metapopulation and 3 O’Kane, S.L. and Al-shehbaz, I.A. (1997) A 15 Savolainen, O. et al. (2000) Contrasting patterns its application to Arabidopsis thaliana. Genetics synopsis of Arabidopsis (Brassicaceae). Novon of nucleotide polymorphism at the alcohol 155, 2015–2019 7, 323–327 dehydrogenase locus in the outcrossing 28 Vision, T.J. et al. (2000) The origins of genomic 4 Galloway, G.L. et al. (1998) Phylogenetic utility of Arabidopsis lyrata and the selfing Arabidopsis duplications in Arabidopsis. Science 290, 2114–2117 the nuclear gene arginine decarboxylase: an thaliana. Mol. Biol. Evol. 17, 645–655 29 Lynch, M. and Conery, J.S. (2000) The example from Brassicaceae. Mol. Biol. Evol. 16 Aguade, M. (2001) Nucleotide sequence variation evolutionary fate and consequences of duplicate 15, 1312–1320 at two genes of the phenylpropanoid pathway, the genes. Science 290, 1151–1155 5 Nasrallah, M.E. et al. (2000) Arabidopsis species FAH1 and F3H genes, in Arabidopsis thaliana. 30 Quiros, C.F. et al. (2001) Arabidopsis and Brassica hybrids in the study of species differences and Mol. Biol. Evol. 18, 1–9 comparative genomics: sequence, structure and evolution of amphiploidy in plants. Plant Physiol. 17 Johanson, U. et al. (2000) Molecular analysis of gene content in the ABI1-Rps2-Ck1 chromosomal 124, 1605–1614 FRIGIDA, a major determinant of natural segment and related regions. Genetics 6 Roy, B.A. (1993) Patterns of rust infection as a variation in Arabidopsis flowering time. Science 157, 1321–1330 function of host genetic diversity and host density 290, 344–347 31 Rossberg, M. et al. (2001) Comparative sequence in natural populations of the apomictic crucifer, 18 Stahl, E.A. et al. (1999) Dynamics of disease analysis reveals extensive microcolinearity in the Arabis holboellii. Evolution 47, 111–124 resistance polymorphism at the Rpm1 locus of lateral suppressor regions of the tomato, 7 Sharbel, T.F. and Mitchell-Olds, T. (2001) Arabidopsis. Nature 400, 667–671 Arabidopsis, and Capsella genomes. Plant Cell Recurrent polyploid origins and 19 Juenger, T. et al. (2000) Quantitative trait loci for 13, 979–988 phylogeography in the Arabis holboellii complex floral morphology in Arabidopsis thaliana. 32 Koch, M.A. et al. (2000) Comparative evolutionary (Brassicaceae). Heredity 87, 59–68 Genetics 156, 1379–1392 analysis of the chalcone synthase and alcohol 8 Lagercrantz, U. (1998) Comparative mapping 20 Karkkainen, K. et al. (1999) Genetic basis of dehydrogenase loci among different lineages of between Arabidopsis thaliana and Brassica nigra inbreeding depression in Arabis petraea. Arabidopsis, Arabis and related genera indicates that Brassica genomes have evolved Evolution 53, 1354–1365 (Brassicaceae). Mol. Biol. Evol. 17, 1483–1498 through extensive genome replication 21 Charlesworth, D. and Charlesworth, B. (1995) 33 Wiehe, T. et al. (2000) SGP-1: Prediction and accompanied by chromosome fusions and frequent Quantitative genetics in plants: the effect of the validation of homologous genes based on sequence rearrangements. Genetics 150, 1217–1228 breeding system on genetic variability. Evolution alignments. Genome Res. 11, 1574–1583 9 Sharbel, T.F. et al. (2000) Genetic isolation by 49, 911–920 34 Koch, M.A. et al. (2001) Comparative genomics distance in Arabidopsis thaliana: biogeography 22 Liu, F. et al. (1999) The effect of mating system and regulatory evolution: conservation and and post-glacial colonization of Europe. Mol. Ecol. differences on nucleotide diversity at the function of the Chs and Apetala3 promoters. Mol. 9, 2109–2118 phosphoglucose isomerase l locus in the plant Biol. Evol. 18, 1882–1891 10 Miyashita, N.T. et al. (1999) DNA variation in the genus Leavenworthia. Genetics 151, 343–357 35 Charlesworth, D. et al. (2001) Genome sequences wild plant Arabidopsis thaliana revealed by 23 Charlesworth, D. and Charlesworth, B. (1995) and evolutionary biology, a two-way interaction. amplified fragment length polymorphism Transposable elements in inbreeding and Trends Ecol. Evol. 16, 235–242 analysis. Genetics 152, 1723–1731 outbreeding populations. Genetics 140, 415–417 36 Ballare, C.L. and Scopel, A.L. (1997) Phytochrome 11 Kuittinen, H. and Aguade, M. (2000) Nucleotide 24 Beerli, P. and Felsenstein, J. (2001) Maximum signalling in plant canopies: testing its variation at the chalcone isomerase locus in likelihood estimation of a migration matrix and population-level implications with photoreceptors Arabidopsis thaliana. Genetics 155, 863–872 effective population sizes in n subpopulations by mutants of Arabidopsis. Funct. Ecol. 11, 441–450 12 Alonso-Blanco, C. and Koornneef, M. (2000) using a coalescent approach. Proc. Natl. Acad. Sci. 37 Dorn, L. et al. (2000) Plasticity to light cues and Naturally occurring variation in Arabidopsis: an U. S. A. 98, 4563–4568 resources in Arabidopsis thaliana: testing for underexploited resource for . Trends 25 Filatov, D.A. and Charlesworth, D. (1999) DNA adaptive value and costs. Evolution Plant Sci. 5, 22–29 polymorphism, haplotype structure and 54, 1982–1994

http://tree.trends.com 700 Review TRENDS in Ecology & Evolution Vol.16 No.12 December 2001

38 Jander, G. et al. (2001) The TASTY locus on signaling reduces resistance of Arabidopsis thaliana L. catalyzes the conversion of chromosome 1 of Arabidopsis affects feeding of the against Egyptian cotton worm but not L-phenylalanine to phenylacetaldoxime in the insect Trichoplusia ni. Plant Physiol. diamondback moth. Plant Physiol. biosynthesis of benzylglucosinolate. J. Biol. 126, 890–898 124, 1007–1017 Chem. 275, 14659–14666 39 Hering, E.M. (1957) Blattminen von Europa, 44 Reymond, P. et al. (2000) Differential gene 49 Kliebenstein, D.J. et al. (2001) Gene Uitgeverij Dr W. Junk expression in response to mechanical wounding duplication and the diversification of secondary 40 Kwiatowski, J. and Ayala, F.J. (1999) Phylogeny of and insect feeding in Arabidopsis. Plant Cell metabolism: side chain modification of Drosophila and related genera: Conflict between 12, 707–720 glucosinolates in Arabidopsis thaliana. Plant molecular and anatomical analyses. Mol. Phylog. 45 Mauricio, R. (1998) Costs of resistance to natural Cell 13, 681–693 Evol. 13, 319–328 enemies in field populations of the 50 Kliman, R.M. et al. (2000) The population genetics 41 Feeny, P. (1976) Plant apparency and chemical Arabidopsis thaliana. Am. Nat. 151, 20–28 of the origin and divergence of the Drosophila defense. In Biochemical Interaction Between 46 Siemens, D. and Mitchell-Olds, T. (1998) simulans complex species. Genetics Plants and Insects (Wallace, I. et al., eds), Evolution of pest induced defenses in Brassica 156, 1913–1931 pp. 1–40, Plenum Press plants: tests of theory. Ecology 79, 632–646 51 McKay, J.K. et al. (2001) Local adaptation despite 42 Schenk, P.M. et al. (2000) Coordinated plant 47 Stowe, K.A. (1998) of absence of marker diversity in the rare Sapphire defense responses in Arabidopsis revealed by resistance in Brassica rapa: Correlated response Rockcress. Proc. R. Soc. B Biol. Sci. microarray analysis. Proc. Natl. Acad. Sci. of tolerance in lines selected for glucosinolate 268, 1715–1721 U. S. A. 97, 11655–11660 content. Evolution 52, 703–712 52 Settles, A.M. and Byrne, M. (1998) Opportunities 43 Stotz, H.U. et al. (2000) Induced plant defense 48 Wittstock, U. and Halkier, B.A. (2000) and challenges grow from Arabidopsis genome responses against chewing insects. Ethylene Cytochrome P450 CYP79A2 from Arabidopsis . Genome Res. 8, 83–85

Prospects for nuclear gene phylogeography

Matthew P.Hare

In phylogeography, an empirical focus on gene lineages enables the history of been the primary data sources making phylogeography population processes to be inferred from the simultaneous analysis of temporal such a productive empirical approach. Extending and and spatial patterns. Rapidly evolving cytoplasmic DNA has been the empirical generalizing phylogeography by using nuclear workhorse propelling the success of this nascent field. Now, as more HAPLOTYPE data has been a desirable goal in efforts to sophisticated historical models are being tested, there is a growing need for test ever more sophisticated historical models. phylogeography to expand from a largely marker-specific discipline to a more However, given the potential for technical and general analytical approach that can be applied across independent loci. Recent biological complications when analysing nuclear results using nuclear haplotypes to study phylogeography indicate that the haplotype data at the intraspecific level, it has been anticipated technical and biological hurdles can be overcome in many taxa to uncertain whether a phylogeographical study design achieve phylogeographical comparisons across unlinked loci. Although many could generally be extended to noncytoplasmic challenges remain, a more complete understanding of the historical, demographic markers1. An evaluation of achievements made to date and selective processes shaping phylogeographical patterns is emerging. in nuclear phylogeography arguably provides the most practical guide to its future potential. Over the past decade, phylogeography grew as a A gene tree for a single cytoplasmic or nuclear discipline because allelic phylogenies provided (nDNA) LOCUS provides a slim and sometimes explicitly historical tools for the study of geographical misleading representation of the population histories subdivision among populations1.Analysing the through which alleles were transmitted3,4. Because relative ages and historical relationships of alleles cytoplasmic loci are usually inherited uniparentally, in a geographical context could distinguish ongoing they will not have genealogical patterns that are processes such as GENE FLOW (see Glossary) from representative of the entire population history, previous events such as range expansion. The temporal especially when sex biases have affected fitness or resolution offered by genealogies was previously dispersal behavior5. Loci under selection can also unavailable through classical population genetic have genetic patterns that deviate from expectations analyses (e.g. hierarchical partitioning of variation based on population history and demography, yet 2 Matthew P.Hare using F-statistics) . In addition, using alleles rather the selected patterns often mimic alternative Dept of Biology, than populations as the basic unit in phylogenetic demographic histories. Testing alternative University of Maryland, clustering promoted integrated analyses of history hypotheses about evolutionary forces must ultimately College Park, MD 20742, above and below the species level without the need rely on the that selection acts locally within a USA. e-mail: matt.hare@ to make a priori taxonomic distinctions. genome, whereas population demography a umail.umd.edu mitochondrial (mt)DNA and chloroplast (cp)DNA have common signature across all neutral loci1,6.

http://tree.trends.com 0169-5347/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0169-5347(01)02326-6