Plant Diversity and Evolution Genotypic and Phenotypic Variation in Higher Plants

Plant Diversity and Evolution Genotypic and Phenotypic Variation in Higher Plants Edited by Robert J. Henry Centre for Plant Conservation Genetics Southern Cross University Lismore, Australia CABI Publishing

CABI Publishing is a division of CAB International CABI Publishing CABI Publishing CAB International 875 Massachusetts Avenue Wallingford Oxfordshire OX10 8DE 7th Floor UK Cambridge, MA 02139 USA Tel: +44 (0)1491 832111 Tel: +1 617 395 4056 Fax: +44 (0)1491 833508 Fax: +1 617 354 6875 E-mail: [email protected] E-mail: [email protected] Website: www.cabi-publishing.org © CAB International 2005. All rights reserved. No part of this publication may be reproduced in any form or by any means, electronically, mechanically, by photocopying, recording or otherwise, without the prior permission of the copyright owners. A catalogue record for this book is available from the British Library, London, UK. Library of Congress Cataloging-in-Publication Data Henry, Robert J. Plant diversity and evolution : genotypic and phenotypic variation in higher plants / Robert J Henry. p. cm. Includes bibliographical references (p. ). ISBN 0-85199-904-2 (alk. paper) 1. Plant diversity. 2. Plants--Evolution. I. Title. QK46.5.D58H46 2005 581.7--dc22 2004008213 ISBN 0 85199 904 2 Typeset in 9/11pt Baskerville by Columns Design Ltd, Reading. Printed and bound in the UK by Cromwell Press, Trowbridge.

Contents Contributors vii 1 Importance of plant diversity 1 Robert J. Henry 2 Relationships between the families of flowering plants 7 Mark Chase 3 Diversity and evolution of gymnosperms 25 Ken Hill 4 Chloroplast genomes of plants 45 Linda A. Raubeson and Robert K. Jansen 5 The mitochondrial genome of higher plants: a target for natural adaptation 69 Sally A. Mackenzie 6 Reticulate evolution in higher plants 81 Gay McKinnon 7 Polyploidy and evolution in plants 97 Jonathan Wendel and Jeff Doyle 8 Crucifer evolution in the post-genomic era 119 Thomas Mitchell-Olds, Ihsan A. Al-Shehbaz, Marcus A. Koch and Tim F. Sharbel 9 Genetic variation in plant populations: assessing cause and pattern 139 David J. Coates and Margaret Byrne 10 Evolution of the flower 165 Douglas E. Soltis, Victor A. Albert, Sangtae Kim, Mi-Jeong Yoo, Pamela S. Soltis, Michael W. Frohlich, James Leebens-Mack, Hongzhi Kong, Kerr Wall, Claude dePamphilis and Hong Ma 11 Diversity in plant cell walls 201 Philip J. Harris 12 Diversity in secondary metabolism in plants 229 Peter G. Waterman v

vi Contents 249 287 13 Ecological importance of species diversity 317 Carl Beierkuhnlein and Anke Jentsch 327 14 Genomic diversity in nature and domestication Eviatar Nevo 15 Conserving genetic diversity in plants of environmental, social or economic importance Robert J. Henry Index

Contributors Victor A. Albert, The Natural History Museums and Botanical Garden, University of Oslo, NO-0318 Oslo, Norway Ihsan A. Al-Shehbaz, Missouri Botanical Gardens, PO Box 299, St Louis, MO 63166-0299, USA, Email: [email protected] Carl Beierkuhnlein, University Bayreuth, Lehrstuhl fur Biogeografie, D-95440 Bayreuth, Germany, Email: [email protected] Margaret Byrne, Science Division, Department of Conservation and Land Management, Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia, Email: [email protected] Mark Chase, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UK, Email: [email protected] David J. Coates, Science Division, Department of Conservation and Land Management, Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia, Email: [email protected] Claude dePamphilis, Department of Biology, The Huck Institutes of the Life Sciences and Institute of Molecular Evolutionary Genetics, The Pennsylvania State University, University Park, PA 16802, USA Jeff Doyle, Department of Plant Biology, 228 Plant Science Building, Cornell University, Ithaca, NY 14853–4301, USA Michael W. Frohlich, Department of Botany, Natural History Museum, London SW7 5BD, UK Philip J. Harris, School of Biological Sciences, The University of Auckland, Private Bag 92019, Auckland, New Zealand, Email: [email protected] Robert J. Henry, Centre for Plant Conservation Genetics, Southern Cross University, PO Box 157, Lismore, NSW 2480, Australia, Email: [email protected] Ken Hill, Royal Botanic Gardens, Mrs Macquaries Road, Sydney NSW 2000, Australia, Email: [email protected] Robert K. Jansen, Integrative Biology, University of Texas, Austin, TX 78712-0253, USA, Email: [email protected] Anke Jentsch, UFZ Centre for Environmental Research Leipzig, Conservation Biology and Ecological Modelling, Permoserstr. 15, D-04318 Leipzig, Germany Sangtae Kim, Department of Botany and the Genetics Institute, University of Florida, Gainesville, FL 32611, USA Marcus A. Koch, Heidelberg Institute of Plant Sciences, Biodiversity and Plant Systematics, Im Neuenheimer Feld 345, D69129, Heidelberg, Germany, Email: [email protected] heidelberg.de Hongzhi Kong, Laboratory of Systematic and Evolutionary Botany, Institute of Botany, The Chinese Academy of Sciences, Beijing 100093, China and Department of Biology, The Huck Institutes of the Life Sciences and Institute of Molecular Evolutionary Genetics, The Pennsylvania State University, University Park, PA 16802, USA vii

viii Contributors James Leebens-Mack, Department of Biology, The Huck Institutes of the Life Sciences and Institute of Molecular Evolutionary Genetics, The Pennsylvania State University, University Park, PA 16802, USA Hong Ma, The Huck Institutes of the Life Sciences and Institute of Molecular Evolutionary Genetics, The Pennsylvania State University, University Park, PA 16802, USA Sally A. Mackenzie, Plant Science Initiative, N305 Beadle Center for Genetics Research, University of Nebraska, Lincoln, NE 68588-0660, USA, Email: [email protected] Gay McKinnon, School of Plant Science, University of Tasmania, Private Bag 55, Hobart, TAS 7001, Australia, Email: [email protected] Thomas Mitchell-Olds, Department of Genetics and Evolution, Max Planck Institute of Chemical Ecology, Hans-Knoll Strasse 8, 07745, Jena, Germany, Email: [email protected] Eviatar Nevo, Institute of Evolution, University of Haifa, Mt Carmel, Haifa, Israel, Email: [email protected] Linda A. Raubeson, Department of Biological Sciences, Central Washington University, Ellensburg, WA 98926-7537, Email: [email protected] Tim F. Sharbel, Laboratoire IFREMER de Genetique et Pathologie, 17390 La Tremblade, France, Email: [email protected] Douglas E. Soltis, Department of Botany and the Genetics Institute, University of Florida, Gainesville, FL 32611, USA, Email: [email protected]fl.edu Pamela S. Soltis, Florida Museum of Natural History and the Genetics Institute, University of Florida, Gainesville, FL 32611, USA Kerr Wall, Department of Biology, The Huck Institutes of the Life Sciences and Institute of Molecular Evolutionary Genetics, The Pennsylvania State University, University Park, PA 16802, USA Peter G. Waterman, Centre for Phytochemistry, Southern Cross University, Lismore, NSW 2480, Australia, Email: [email protected], [email protected] Jonathan Wendel, Department of Ecology, Evolution and Organismal Biology, Iowa State University, Ames, IA 50011, USA, Email: [email protected] Mi-Jeong Yoo, Department of Botany and the Genetics Institute, University of Florida, Gainesville, FL 32611, USA

1 Importance of plant diversity Robert J. Henry Centre for Plant Conservation Genetics, Southern Cross University, PO Box 157, Lismore, NSW 2480, Australia Introduction (DNA). Analysis of expressed genes (tran- scriptome), proteins (proteome), metabolites Plants are fundamental to life, providing the (metabolome) and ultimately phenotypes basic and immediate needs of humans for (phenome) provides a range of related lay- food and shelter and acting as an essential ers for investigation of plant diversity. component of the biosphere maintaining life on the planet. Higher plant species occupy a Diversity of Plant Species wide variety of habitats over most of the land surface except for the most extreme More than a quarter of a million higher plant environments and extend to fresh water and species have been described. Continual analy- marine habitats. Plant diversity is important sis identifies new, previously undescribed for the environment in the most general species and may group more than one sense and is an essential economic and social species together (lumping) or divide species resource. The seed plants (including the into more than one (splitting). The use of flowering plants) are the major focus of this DNA-based analysis has begun to provide book and are related to the ferns and other more objective evidence for such reclassifica- plant groups as shown in Fig. 1.1. tions. Evolutionary relationships may be deduced using these approaches. The analy- Types of Plant Diversity sis of plant diversity at higher taxonomic lev- els allows identification of genetic Plant diversity can be considered at many relationships between different groups of different levels and using many different cri- plants. The family is the most useful and teria. Phenotypic variation is important in important of these classification levels. A the role of plants in the environment and in knowledge of evolutionary relationships is practical use. Analysis of genotypic variation important in ensuring that management of provides a basis for understanding the plant populations is conducted to allow con- genetic basis of this variation. Modern bio- tinuation of effective plant evolution, allowing logical research allows consideration of vari- longer-term plant diversity and survival to be ation at all levels from the DNA to the plant maintained. The use of DNA analysis has characteristic (Table 1.1). Genomics studies greatly improved the reliability and likely sta- the organism at the level of the genome bility of such classifications. Chase presents an © CAB International 2005. Plant Diversity and Evolution: Genotypic and 1 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

2 R.J. Henry Gymnosperms Angiosperms (flowering plants) Lycophytes Ferns and (clubmosses) horsetails Bryophytes (liverworts, hornworts and mosses) Fig. 1.1. Phylogenetic relationships between higher plants (based upon Pryer et al., 2001). Table 1.1. Levels of analysis of diversity in plants. Level Whole system Study of whole system DNA Genome Genomics RNA Transcriptome Transcriptomics Protein Proteome Proteomics Metabolite Metabolome Metabolomics Phenotype Phenome Phenomics updated review of the relationships between Diversity within Plant Species the major groups of flowering plants in Chapter 2. This analysis draws together Diversity within a population of plants of the recent evidence from plant DNA sequence same species may be considered a primary analysis. The rate of evolution of new species level of variation. Coates and Byrne present varies widely in different plant groups (Klak an analysis of the causes and patterns of et al., 2004). The factors determining these variation within plant species in Chapter 9. differences are likely to be important deter- Principles of population genetics can be minants of evolutionary processes. used to analyse and understand the varia- tion within and between populations of a Evolutionary relationships are important species. Reproductive mechanisms are a key in plant conservation and also in plant determinant of plant diversity. Plants may improvement. Plant breeders increasingly reproduce by either sexual or asexual look to source genes from wild relatives for means. Clonal or vegetative propagation use in the introduction of novel traits or the usually results in relatively little genetic vari- development of durable pest and disease ation except that arising from somatic muta- resistance (Godwin, 2003).

Importance of plant diversity 3 tions. Sexual reproduction can involve many The diversity of different plant communi- different reproductive mechanisms that pro- ties that make up the wider ecosystem is duce different levels of variation within the another level to be considered. Plant com- population. Outbreeding species are gener- munities may extend over very wide geo- ally much more variable than inbreeding or graphic ranges while in others a complex self-pollinating species. Some species use mosaic of different plant communities can more than one of these methods of repro- exist in close proximity. This is usually duction. Examples include a mix of vegeta- determined by the uniformity of the envi- tive variants, mixed outcrossing and ronment, which, in turn, is determined by mechanisms such as apomixis. differences in substrate or microenviron- Morphological and other phenotypic varia- ment. This is an important level of analysis tion within species can be extreme. Variation of plant diversity for use in the conservation in one or a small number of genes can result of plant and more general biodiversity. in very large morphological differences in the plant. Maize was domesticated from Plant Diversity Enriching and Sustaining teosinte, a very different plant in appear- Life ance. However, a mutation in a single gene has been shown to explain the major mor- Plants and plant diversity contribute directly phological differences (Wang et al., 1999). and indirectly to the enrichment of life This emphasizes the importance of DNA experiences for humans. A world in which analysis in determination of plant diversity. few other life forms existed would in a nar- row sense limit opportunities for eco- Factors determining diversity within tourism, but this is a much wider issue. A species are also being better defined by the key driver for support for nature conserva- use of DNA analysis methods. The influence tion is the human perception that diversity of environmental factors in driving adaptive of life forms has a value beyond that associ- selection relative to other factors of evolution- ated with the importance, however great ary history in determining genetic structure that might be, of diversity for environmental of plant populations can now be examined sustainability and economic reasons. experimentally. Nevo explores these issues in Chapter 14. Habitat fragmentation may limit Human food is sourced directly or indi- gene flow in wild plant populations (Rossetto, rectly from plants. The role of plants in the 2004). This has become an important issue in food chain is dominant for all animal life. managing the impact of human activities on This provides immediate and everyday plant diversity and evolutionary capacity. examples of the importance of plant diver- sity in contributing to a diversity of foods. A Plant Diversity at the Community and small number of plant species account for a Ecosystem Level relatively large proportion of the calories and protein in human diets. Most human diets Diversity can also be considered at the plant include smaller amounts of a larger number community level. Indeed this is probably what of plant species. Many more plant species are most people think of when they consider regionally important as human food. plant diversity. This diversity of species within Chapter 15 (Henry) expands on these issues. any given plant community is often termed the species richness. The number of species is Environmental Importance of Plant one measure of this diversity but the fre- Diversity quency of different species in the population is another. Populations may contain only one Plant diversity is a key contributor to envi- or a few dominant species and very small ronmental sustainability on a global scale. numbers of individuals from a large number Studies of species richness demonstrate the of species or they may be composed of much greater productivity of more diverse plant more equal numbers of different species.

4 R.J. Henry communities. The mechanisms that promote and relationships in the flowering plants is the co-existence of large numbers of species provided for angiosperms in Chapter 2 and may include the ability of competitors to the gymnosperms in Chapter 3. Diversity in thrive at different times and places (Clark non-nuclear genomes is analysed for the and McLauchlan, 2003). More research is chloroplast in Chapter 4 and the mitochon- needed in this area because of the scale of dria in Chapter 5. The complication of retic- the potential environmental importance of ulate evolution in the interpretation of plant this issue. This topic is reviewed by relationships is evaluated by McKinnon in Beierkuhnlein and Jentsch in Chapter 13. Chapter 6. The evolution and role of poly- ploidy in plants is reviewed by Wendel and Social and Economic Importance of Doyle in Chapter 7. In Chapter 8, Mitchell- Plant Diversity Olds et al. provide an analysis of a plant fam- ily, the Brassicaceae, which includes Social uses of plants may include ceremonial Arabidopsis, the first plant for which a com- and other specific social applications. plete genome sequence was determined. However, the greatest social use of plants Patterns of variation in plant populations probably relates to their use as ornamentals. and their basis are explored by Coates and Ornamental plants often reflect social status Byrne in Chapter 9. The evolution of the or identity. Foods from some plants have a key organ, the flower, is reviewed by Soltis et social value extending beyond that con- al. in Chapter 10. Two key features of plants tributed by their nutritional value. – the cell wall and diverse secondary metab- olism – are described in an evolutionary Agriculture and forestry are primary context by Harris and Waterman in industries of great economic importance. Chapters 11 and 12, respectively. The plant The food industry as an extension of agri- cell is characterized by the presence of a cell culture can be considered to depend upon wall essential to the structure of plants. The plant diversity. Ornamental plants are also cell wall is not only of biological significance. of considerable economic importance. Fibre The chemistry of cell walls is the basis of crops (such as cotton and hemp) provide a wood and paper chemistry. The secondary major source of materials for clothing. metabolites in plants play a major role in the Forest species are key sources of building defence of the plant. These compounds are materials for shelter for many human popu- also of use to humans in many applications, lations. Plants remain the source of many including use as drugs or drug precursors in medicinal compounds. All of these uses have medicine. The ecological significance of social and economic importance. plant diversity is the subject of Chapter 13. Nevo explores the impact of domestication Overview of Plant Diversity and on plant diversity in Chapter 14 and Henry Evolution describes conservation of diversity in plants of environmental, social and economic This book brings together a wide range of importance in Chapter 15. issues and perspectives on plant diversity and evolution. Diversity at the genome This compilation brings together infor- (gene) and phenome (trait) level is consid- mation on plant diversity and evolution in a ered. A contemporary analysis of diversity general sense and provides essential back- ground for an understanding of plant biol- ogy and plant use in industry. References Clark, J.S. and McLauchlan, J.S. (2003) Stability of forest biodiversity. Nature 423, 635–638. Godwin, I. (2003) Plant germplasm collections as sources of useful genes. In: Newbury, H.J. (ed.) Plant Molecular Breeding. Blackwell, Oxford, pp. 134–151.

Importance of plant diversity 5 Klak, C., Reeves, G. and Hedderson, T. (2004) Unmatched tempo of evolution in Southern African semi- desert ice plants. Nature 427, 63–65. Pryer, K.M., Schnelder, H., Smith, A.R., Cranfill, R., Wolf, P.G., Hunt, J.S. and Sipes, S.D. (2001) Horsetails and ferns are the monophyletic group and the closest living relatives of seed plants. Nature 409, 618–622. Rossetto, M. (2004) Impact of habitat fragmentation on plant populations. In: Henry, R.J. (ed.) Plant Conservation. Haworth Press, New York. Wang, R.L., Stec, A., Hey, J., Lukens, L. and Dooebley, J. (1999) The limits of selection during maize domes- tication. Nature 398, 236–239.

2 Relationships between the families of flowering plants Mark W. Chase Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UK Introduction It is clear that an improved understand- ing of all phenotypic patterns is important, In the past 10 years, enormous improve- but it is equally clear that assessments of phy- ments have been made to our ideas of logenetic patterns should involve as few angiosperm classification, which have interpretations and as many data points as is involved new sources of information as well possible. Other forms of DNA data (e.g. gene as new approaches for handling of system- order and restriction endonuclease data) suf- atic data. The former is the topic of this fer from limitations similar to those of mor- chapter, but a few comments on the latter phology, and thus also should be abandoned are appropriate. Before the Angiosperm as appropriate data for phylogenetic analy- Phylogeny Group classification (APG, 1998), ses. Prior to the APG effort (1998), there was the process of assessing relationships was no single, widely accepted phylogenetic clas- mired in the use of gross morphology and a sification of the angiosperms, regardless of largely intuitive understanding of which the data type upon which a classification was characters should be emphasized (effectively based. Instead, classifications were estab- a method of character weighting). lished largely on the authority of the author; Morphological features and other non-mole- choice of which of the many in simultaneous cular traits (such as development, biosyn- existence should be used depended to a thetic pathways and physiology) are worthy large degree on geography, such that in the of study, but their use in phylogenetic analy- USA the system of Cronquist (1981) was pre- ses is limited by the prior information pos- dominant, whereas in Europe those of sessed by the researcher through which the Dahlgren (1980) or Takhtajan (1997) were acquisition of new data is filtered and the more likely to be used. To a large degree, inherently complex and largely unknown these competing systems agreed on most genetic basis of nearly all traits. It has issues, but in the end they disagreed on become increasingly clear that morphology many points, including the relationships of and other phenotypic data are not appropri- some of the largest families, such as ate for phylogenetic studies (Chase et al., Asteraceae, Fabaceae, Orchidaceae and 2000a), but instead should be interpreted in Poaceae. When trying to establish why these the light of phylogenetic trees produced by differences existed, it soon becomes evident analysis of DNA data, preferably DNA that the authors of these classifications were sequences. using the same data but interpreting them © CAB International 2005. Plant Diversity and Evolution: Genotypic and 7 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

8 M. Chase differently, usually in line with their intuitive intuitive basis meant that they were not sub- assessments of which suites of characters ject to improvement through evaluation of were most informative. emerging new data. The only way changes could be incorporated was by the original The issue of ranks and authority author changing his or her mind. This intu- itive basis made researchers in other fields of Other differences between morphologically science view classification as more akin to based classifications (e.g. Cronquist, 1981; philosophy than science. Thus, in spite of Thorne, 1992; Takhtajan, 1997) have to do many years of careful study and syntheses of with the hierarchical ranks given to the same many data, plant taxonomy came to be groups of lower taxa. For example, viewed as an outmoded field of research. It Platanaceae (one genus, Platanus) were placed was clear that all of the different ideas of in the order Hamamelidales by Cronquist relationships for a given family, Fabaceae for (1981), the order Platanales by Thorne example, observed in competing modern (1992), and the subclass Platanidae by classifications could not be simultaneously Takhtajan (1997), but only in the first case was correct, and if selection of one over the oth- it associated with any other families. In APG ers was based on an assessment of which (1998, 2003), Platanaceae were included in author was the most authoritative, then per- Proteales along with Nelumbonaceae and haps framing a research programme around Proteaceae and were listed as an optional syn- a classification was unwise. It would perhaps onym of Proteaceae (APG, 2003). Higher cate- be better to think that predictivity should not gories composed of single taxa are a be an attribute of classification and to ignore redundancy in classification and make them the evolutionary implications for research in less informative than systems with many taxa other fields. Although it is immediately clear in each higher category. All clades in a clado- to researchers in other areas of science that gram should not be named, and lumping to classifications should be subject to modifica- an extreme degree can also make the system tion on the basis of being demonstrated to less informative, but monogeneric families, such put together unrelated taxa, this did not as Platanaceae, should not then be the sole com- appear to matter to many taxonomists. ponent of yet higher taxa unless such a taxon is sister to a larger clade composed of many The APG classification is not the work of higher taxa. Thus recognition of Zygophyllales a single author, and the data are analysed composed of only Zygophyllaceae was included phylogenetically, that is, without any influ- in APG (2003) for exactly this reason, but had ence of preconceived ideas of which charac- Zygophyllaceae been shown to be sister to any ters are more reliable or informative, other single order, they would have been included than that DNA sequences from all three there so that redundancy of the classification genetic compartments that agree about pat- could have been reduced. terns of relationships (Soltis et al., 2000) are likely to produce a predictive classification. Regardless of these considerations, all If new data emerge that demonstrate that classifications prior to APG (1998) could only any component of the APG system places be revised or improved by the originating together unrelated taxa, then the system will author; if an author made changes (usually be modified to take these data into account. viewed as ‘improvements’) to the classifica- There is no longer a need for competing tion of another author, then what resulted classifications, and over time the APG system was viewed as the second author’s classifica- should be improved by more study and the tion, not merely as a revision of the first. The addition of more data. long succession of major classifications of the angiosperms was the result of the fact that Monophyly and classification these were not composed of sets of falsifiable hypotheses. They were indisputably The concept of monophyly has had a long hypotheses of relationships, but their highly and problematic history, and some have

Relationships between flowering plant families 9 claimed that the phylogeneticists have genetic basis, and what may appear to be a twisted its original meaning. It is not worth- ‘major trait’ could in fact be a genetically while to include these arguments here, but it simple change. Therefore, recognition of is appropriate to mention that the APG sys- paraphyletic taxa does not involve an appre- tem follows the priorities for making deci- ciation of how ‘major’ underlying genetic sions about which families to recognize that change might be and assumes that the tax- were proposed by Backlund and Bremer onomist can determine this simply by (1998), which means that the first priority is appearances, which we know to be incorrect. that all taxa are monophyletic in the phylo- The use of paraphyly in classification there- genetic sense of the word, i.e. that all mem- fore decreases predictivity of the system and bers of a taxon must be more closely related on this basis should also be avoided. to all other members of that taxon than they are to the members of any other taxon. This What follows in this chapter is compatible is in contrast to what an evolutionary taxon- with the use of monophyly in what has come omist would propose; in such an evolution- to be known as ‘Hennigian monophyly’, after ary system, if some of the members of a the German taxonomist, Hennig, whose group had developed one or more major ideas formed the basis for phylogenetic novel traits then that group could be segre- (cladistic) classification. It is of no importance gated into a separate family, leaving behind that an earlier definition of ‘monophyly’ may in another family the closest relatives of the or may not have existed. The term as used in removed group (the phylogenetic taxono- this sense has been widely accepted as of mist would term the remnant group as prime importance in the construction of a being paraphyletic to the removed group, predictive system of classification, and classi- which is not permitted in a phylogenetic fication should be as practical as possible and classification). Aside from the philosophical as devoid of historical and philosophical con- considerations, which have been debated cepts as possible because this makes classifica- extensively, there is a practical reason for tion subject to change simply because new eliminating paraphyletic groups: it is impos- generations develop new philosophies, which sible to get two evolutionary taxonomists to inevitably means that classification must agree on when to split a monophyletic change. Change of classification is undesir- group in this manner. Is one major novel able on this basis, and therefore the tenets trait enough or should there be two or under which a classification is formulated more? How do we define a ‘major trait’ such should be as far removed from historical and that everyone understands when to split a philosophical frameworks as possible because monophyletic group? This problem is simi- if a classification is to be used by scientists lar to that of falsifying hypotheses that are in other fields, it should change as little as based on someone else’s intuition. If given possible. the same set of taxa, how likely is it that two evolutionary taxonomists would split them Angiosperm Relationships in the same manner and how would either be able to prove the other wrong? The overall framework of extant Therefore, the practical solution is to avoid angiosperm relationships (Fig. 2.1) has the use of paraphyly, which is what the APG become clear only since the use of DNA system did. It is simply impractical to sequences to elucidate phylogenetic pat- include paraphyletic taxon in a system, terns, beginning with Chase et al. (1993). because to do so forces the process of classi- Analyses using up to 15 genes from all three fication back into the hands of authority and genomic compartments of plant cells incorporates intuition in the process, which (nucleus, mitochondrion and plastid) have is not only undesirable but also unscientific. yielded consistent and well-supported esti- mates of relationships (Qiu et al., 2000; Zanis From the standpoint of the genetics, use et al., 2002). Studies of genes have placed of paraphyly is also unwise. This is because the previously poorly known monogeneric there are few traits for which we know the

10 M. Chase Amborellaceae Nymphaeaceae angiosperms Austrobaileyales Chloranthaceae Magnoliales Laurales Canellales magnoliids Piperales Acorales Alismatales Pandanales monocots Dioscoreales Liliales Asparagales Arecales Poales Dasypogonaceae commelinids Commelinales Zingiberales eudicots Ceratophyllales Ranunculales Sabiaceae Proteales Buxaceae Trochodendraceae Gunnerales Aextoxicaceae core eudicots Berberidopsidaceae rosids Dilleniaceae Caryophyllales Santalales Saxifragales Crossosomatales Geraniales Myrtales Celastrales Malpighiales Oxalidales Rosales Fabales eurosid I Fagales Cucurbitales Brassicales Sapindales eurosid II Malvales asterids Cornales euasterid I Ericales euasterid II Garryales Lamiales Solanales Gentianales Aquifoliales Apiales Asterales Dipsacales Fig. 2.1. The APG classification displayed in cladogram format. The patterns of relationships shown are those that were well supported in Soltis et al. (2000) or other studies; the data analysed in these studies included at least plastid rbcL and atpB and nuclear 18S rDNA sequences. Rosid and asterid families not yet placed in one of the established orders are not shown (modified from APG, 2003). family Amborellaceae as sister to the rest of subject of a great number of other studies the angiosperms. Amborella, restricted to and has been shown to have a number of New Caledonia, has, since the three-gene not particularly primitive traits, such as sep- analysis of Soltis et al. (1999, 2000), been the arately sexed plants. One study (Barkman et

Relationships between flowering plant families 11 al., 2000) used a technique to ‘reduce’ noise The remainder of the angiosperms fall in DNA sequences, which resulted in Amborella into two large groups, the monocots and being placed sister to Nymphaeaceae (the eudicots (dicots with triaperturate pollen), waterlilies). It is not clear how the subject of and a number of smaller clades: Canellales, noise in DNA sequences should be identi- Laurales, Magnoliales, Piperales (these four fied, but several other techniques were orders collectively known as the ‘eumagnoli- employed by Zanis et al. (2002), and they ids’ or simply ‘magnoliids’), Ceratophyllaceae found that the rooting at the node with (monogeneric) and Chloranthaceae (four Amborella alone could not be rejected by any genera). These smaller groups were in pre- partition of the data (e.g. codons, transi- vious systems typically included with the tions/tranversions, synonymous/non-synony- eudicots in the ‘dicots’ because, like the mous). Thus it seems reasonable to conclude eudicots, they have two cotyledons. None that the rooting issue was resolved in favour the less, they share with the monocots unia- of that of Amborella, but more study is perturate pollen, and it would appear that required. Following Amborella, the next node the magnoliids are collectively sister to the splits Nymphaeaceae from the rest, followed monocots (Duvall et al., 2005). The relation- by a clade composed of Austrobaileyaceae, ships of Ceratophyllum and Chloranthaceae Schisandraceae and Trimeniaceae. This have been difficult to establish, but it would arrangement of families (the ANITA grade appear that the former are related to the of Qiu et al., 1999) results in each being monocots and the latter perhaps sister to the given ordinal status: Amborellales, monocots plus magnoliids. More study is Nymphaeales and Austrobaileyales. None of required before these issues can be settled. these families is large (Nymphaeaceae is the largest with eight genera and 64 species), As stated above, the monocots were con- and were it not for their phylogenetic place- sidered one of the two groups of ment, they would probably receive little angiosperms, but they share with the primi- attention. They are critical in terms of tive dicots pollen with a single germination understanding patterns of morphological pore. In this respect, they are not an obvi- and genomic change within the ous group on their own, but they deviate angiosperms, and thus no study purporting substantially from the primitive dicots in to present a comprehensive overview can having scattered vascular bundles in their ignore them. They have thus been studied stems (as opposed to having them arranged extensively but are problematic none the in a ring) and leaves generally with parallel less because it is clear that they are the last venation (as opposed to a net-like reticu- remnants of their lineages. As such they are lum). Their flowers are generally composed unlikely to represent adequately the traits of of whorls of three parts, typically two whorls these lineages, so their use in the study of each of perianth parts and stamens and a how morphological characters have changed single whorl of carpels, but there are numer- must be qualified by an appreciation of the ous exceptions to this format. instability caused by having so few represen- tatives of these earliest lineages to diverge Within the monocots, the relationships of from the rest of the angiosperms. It could nearly all families are well established as well well be that the traits ancestral for the as the general branching order of the orders angiosperms are not to be found in the fam- sensu APG (1998, 2003). Monogeneric ilies of the ANITA grade, but rather in the Acoraceae (Acorales) are sister to the rest descendants of the other line, the bulk of the (Chase et al., 1993, 2000b; Duvall et al., families of angiosperms. ‘Basal’ families in a 1993a,b); the sole genus, Acorus, in most sys- phylogenetic sense are not necessarily primi- tems of classification was included in Araceae tive (the concept of heterobathmy applies (the aroids), but most morphologists had con- here: most plants are mixtures of advanced cluded that it did not belong there (Grayum, and primitive traits, for example dioecy and 1987). The issue of what is the most primitive vesselless wood, respectively, in Amborella). monocot family was not settled by the posi- tion of Acorus because most of the characters judged to be primitive in the monocots are

12 M. Chase found in Alismatales (Dahlgren et al., 1985). Asparagales (14 families) is the largest Alismatales (13 families), which include order of the monocots and contains the Araceae, Tofieldiaceae and the largest family, Orchidaceae (the orchids, 750 alismatid families (Alismataceae, Apono- genera, 20,000 species; one of the two getonaceae, Butomaceae, Cymodoceaceae, largest families of the angiosperms, the Hydrocharitaceae, Juncaginaceae, Limno- other being Asteraceae). The onion and daf- charitaceae, Posidoniaceae, Potamo- fodil family (Alliaceae) and the asparagus getonaceae, Ruppiaceae and Zosteraceae), are and hyacinth family (Asparagaceae) are the then the next successive sister to the rest of enlarged optional concepts of these families the monocots. The alismatid families were proposed by APG (2003). Up to 30 smaller previously the only components of families have sometimes been recognized in Alismatales, but analyses of DNA data have Asparagales, but this large number of mostly indicated a close relationship of these to small families makes learning the families of Araceae and Tofieldiaceae, the former being the order difficult and trivializes the concept considered either an isolated family or of family. Therefore, I favour the optional related to Areceae (the palms) and the latter a fewer/larger families recommended by APG part of Melanthiaceae, all of which have now (2003). For example, APG II proposed to been proven to be erroneous placements. lump the following in Asparagaceae: Agavaceae (already including Anemarrh- Alismatales include a large number of neaceae, Anthericaceae, Behniaceae and aquatic taxa, both freshwater and marine. Hostaceae), Aphyllanthaceae, Hyacinthaceae, The flowering rush family (Butomaceae) and Laxmanniaceae, Ruscaceae (already includ- water plantain family (Alismataceae) include ing Convallariaceae, Dracaenaceae, mostly emergent species, whereas others, such Eriospermaceae and Nolinaceae) and as the pondweed family (Potamogetonaceae) Themidaceae. Hesperocallidaceae have and frog’s bit family (Hydrocharitaceae), have recently been shown to be embedded in species that are submerged, with perhaps only Agavaceae, thus further reducing the num- their flowers reaching the surface. Yet others, ber of families in Asparagales. Asparagales such as Najadaceae, have underwater pollina- include a number of genera that can pro- tion. The eel grass family (Zosteraceae) and duce a form of secondary growth, which per- the sea grass families (Cymodoceaceae and mits them to become tree-like; these include Posidoniaceae) are all marine and ecologically the Joshua tree (Yucca), aloes (Aloe) and the important; they are also among the relatively grass trees of Australia (Xanthorrhoea). small number of angiosperms that have con- quered marine habitats. Orchidaceae are famous for their extrav- agant flowers and bizarre pollination biol- The next several orders have typically ogy, but only one, the vanilla orchid been considered the ‘lilioid’ monocots because (Vanilla), is of agricultural value. Many are they were by and large included in the hetero- important in the cut flower and pot plant geneous broad concept of Liliaceae by most trade worldwide. Other well-known mem- authors (Hutchinson, 1934, 1967; Cronquist, bers of Asparagales include Iris, Crocus and 1981). Liliaceae in this expansive circumscrip- Gladiolus (Iridaceae), Aloe, Phormium and tion included all monocots with six showy Hemerocallis (Xanthorrhoeaceae), Allium tepals (in which the sepals looked like petals), (onion), Narcissus (daffodils), Hippeastrum six stamens and three fused carpels. If the (amaryllis) and Galanthus (snowdrops; all plants were either arborescent (e.g. Agave, Alliaceae), Asparagus, Hyacinthus (hyacinth), Dracaena) or had broad leaves with net-like Agave (century plant), Hosta and Yucca, venation (e.g. Dioscorea, Trillium), they were Convallaria (lily of the valley), Dracaena, placed in segregate families, but we now know Cordyline and Triteleia (all Asparagaceae). that these distinctions are not reliable for the There are many of these that are of minor purposes of family delimitation. Instead of horticultural importance. Asparagus, onion one large family, we now have five orders, and agave (fibre and tequila) are the only Asparagales, Dioscoreales, Liliales, Pandanales agriculturally exploited species. and Petrosaviales (Chase et al., 2000b).

Relationships between flowering plant families 13 Dioscoreales are composed of three fami- Commelinales include the bloodroots lies, but only Dioscoreaceae, which are large (Haemodoraceae), pickerelweed and water forest understorey plants or vines, are large hyacinths (Pontederiaceae) and the large and well known. Species of Dioscorea (yams) spiderwort family, Commelinaceae. The gin- are a source of starch in some parts of the gers, Zingiberaceae, and bananas, Musaceae, world, as well as of medicines (e.g. birth con- are members of Zingiberales, whereas the trol compounds). A few species are grown as largest commelinid order, Poales, contains ornamentals (e.g. bat flower, Tacca). the wind-pollinated grasses, Poaceae Burmanniaceae are all peculiar mycopara- (Graminae), and sedges, Cyperaceae, which sitic herbs, some of which are without dominate regions where woody plants can- chlorophyll, but these are not common and not grow, as well as the Spanish mosses, have no commercial uses. Bromeliaceae, which like the orchids (Orchidaceae; Asparagales) are epiphytes. In Liliales have 11 families, including the well- addition to being ecologically important, known Liliaceae (in the narrow sense) and the grasses are the foundation of agriculture cat-briars, Smilacaceae (another group of vines worldwide and include maize (Zea), rice with a nearly worldwide distribution). Like a (Oryza) and wheat (Triticum), as well as a number of genera in Asparagales (e.g. number of minor grains, such as barley Narcissus, Allium), many members of Liliaceae (Hordeum) and oats (Avena). have bulbs; Lilium and Tulipa (tulips) are horti- culturally important. Colchicaceae also have Eudicots many species with bulbs, but unlike Liliaceae, which has a north temperate distribution, Eudicots are composed of three major Colchicaceae are primarily found in the south- groups: caryophyllids (a single order, ern hemisphere, although the autumn crocus Caryophyllales), rosids (13 orders) and aster- (Colchicum) is found in Europe and is the ids (nine orders). In addition to these (the source of colchicine, an alkaloid that interferes core eudicots), there are a number of smaller with meiosis and causes chromosome doubling families and orders that form a grade with (polyploidy). Alstroemeriaceae, Peruvian lily, is respect to the core eudicots. The largest of also used in horticulture. these are Ranunculales, which include the buttercups (Ranunculaceae) and poppies Pandanales are a tropical order contain- (Papaveraceae), and Proteales, which include ing the screw pines, Pandanaceae, and the the plane tree (Platanaceae), lotus Panama hat family, Cyclanthaceae. Screw (Nelumbonaceae) and protea (Proteaceae) pines, Pandanus, are immense herbs without families. The last is an important family in secondary growth; the leaves are used as South Africa and Australia where they are thatch, and the fruits are eaten. one of the dominant groups of plants. The Cyclanthaceae are straggling vines that look placement of the lotus (Nelumbo) in this order similar to palms (but they are distantly was one of the most controversial aspects of related); they are local sources of fibre and the early phylogenetic studies based on DNA of course are used for Panama hats. sequences, but subsequent studies have demonstrated that this is a robust result. The The remaining monocots were recog- lotus is a ‘waterlily’ (an herbaceous plant with nized as a group, the commelinids, before rhizome and round leaves attached to the the advent of DNA phylogenetics because of stem in their middle), but its similarities to their shared possession of silica bodies and the true waterlilies are due to convergence. UV-fluorescent compounds in their epider- mal cells. They are otherwise a diverse The so-called ‘basal’ eudicots (i.e. group of plants and include small herbs, a Ranunculales and Proteales) have flowers few vines and tree-like herbs such as the that lack the organization typical for the palms and bananas. Arecales include only larger group. The strict breakdown into the palms, Arecaceae (or the more tradition- sepals, petals, stamens and carpels is not ally used Palmae), which are important throughout the tropics as sources of food, beverage and building materials.

14 M. Chase obvious in many of these taxa. Some have and often small trees or shrubs, but the way what appears to be a regular organization, in which they make wood does not follow but upon closer inspection this breaks down. the typical pattern for angiosperms, which is For example, some Ranunculaceae have probably an indication that these plants are whorls of typical appearance, but the sepals derived from herbs that lost the ability to are instead bracts and the petals are most make woody growth. None the less, some of likely derived from either sepals or stamens. these groups do make wood that appears to Numbers of parts are also not regular, and be typical, so it is not yet clear whether or fusion within whorls or between whorls is not Caryophyllales are ancestrally herba- rare, whereas in the core eudicots flowers ceous. Good examples of this anomalous take on a characteristic ‘synorganization’ in woodiness are the cacti (Cactaceae). Well- which numbers are regular and whorls of known examples of core Caryophyllales adjacent parts are often fused or otherwise families include Amaranthaceae (which interdependent. This is not to say that there include spinach and beets), Caryophyllaceae are not complicated flowers in these basal (carnations), Cactaceae and Portulacaceae lineages because there are some rather (pusley and spring beauty). Cactaceae and extraordinary ones: for example, in several other families adapted to arid zones Ranunculaceae, there are Delphinium species are known to be closely related to various with highly zygomorphic flowers in which members of Portulacaceae, but a formal the parts are highly organized. None the transfer of these families to the last has not less, synorganization is typically the hall- yet been proposed (although it will almost mark of the core eudicots. certainly be treated this way in a future update of the APG system). Caryophyllids In the DNA studies, Centrospermae (core The flowers of Caryophyllales (29 families; caryophyllids) were found to have a number APG, 2003) often look like those of other core of previously undetected relatives. Many of eudicot families, and thus some of the mem- these have chemical and pollen similarities to bers of this order were previously thought to the core group, and some have anomalous be rosids (e.g. the sundews, Droseraceae, secondary growth as well. The core set of which were thought to be related to families are well known for their abilities to Saxifragaceae) or asterids (e.g. the leadworts, adapt to harsh environments, particularly Plumbaginaceae, which many authors deserts and salty sites, and their newly dis- thought were related to Primulaceae because covered relatives are similarly adapted. of their similar pollen and breeding systems For example, the tamarisks (Tamaricaceae) with stamens of different lengths). The core and frankenias (Frankeniaceae) have salt- Caryophyllales have a long history of recogni- secreting glands, and jojoba (Simondsiaceae) tion, and in the past they have been called grows in the arid zones of western North the Centrospermae because of their capsules America along with cacti. The leadworts with seeds arranged on centrally located pla- (Plumbaginaceae) and jewelweeds centa. This group was clearly identified in the (Polygonaceae) also include a number of first DNA studies (Chase et al., 1993), so pre- plants adapted to dry and salty conditions. vious workers were correct in recognizing The ecological diversity displayed by these this group, but the DNA analyses placed a plants was increased by the recognition that number of additional families with the core several families of carnivorous plants are Caryophyllales. In addition to their fruit members of Caryophyllales. These are the characters, Centrospermae also have betalain sundews and Venus fly trap (Droseraceae) floral pigments that have replaced the antho- and the Asian pitcher plants (Nepenthaceae). cyanins typically found in angiosperms. Another common characteristic is anomalous Carnivory evolved several times in the secondary growth; such plants are woody angiosperms, and there are members in each of the major groups: Brochinnia (Bromeliaceae) in the monocots, the Australian pitcher plants (Cephalotus,

Relationships between flowering plant families 15 Cephalotaceae) in the rosids and the blad- these). Asterids are similar except that there derworts (Lentibulariaceae) and New World is a single whorl of stamens. Santalales have pitcher plants (Sarraceniaceae), each related typically many whorls of some parts, partic- to different groups of the asterids. Botanists ularly stamens (up to as many as 16 in some had debated the affinities of each of these cases), so they clearly deviate from the main groups of carnivorous plants for many years, themes of the core eudicots. It is likely that and most had proposed multiple origins. Santalales evolved before the number of However, there was little agreement about whorls became fixed or that they have sim- which of the carnivorous plants might be ply retained a degree of developmental flex- closely related and with which other families ibility that was lost in the other major they shared a common history. DNA data groups. were crucial to establish patterns of relation- ships (Albert et al., 1992) because the highly Saxifragales modified morphology of these plants as well as the diversity of floral types made assess- Unlike Santalales, Saxifragales (12 families) ments of their relationships largely a matter is a novel order in the APG system (1998, of intuitive weighting of the reliability of 2003). The name has been used previously these characters. by some authors, but the circumscription of the order is different. Some of the families Santalales are woody and wind-pollinated, for example the witch hazel family (Hamamelidaceae, Before turning to the rosids, I would like to although some genera are pollinated by mention briefly two APG orders of core insects) and the sweet gum family eudicots that have not been placed in the (Altingiaceae), and these were previously three major groups because they have yet to considered to be related to the other wind- obtain a clear position in the results of the pollinated families (see Hamamelidae DNA studies. The first of these are below). Others are woody and insect-polli- Santalales (six families), which include a nated, for example the gooseberry and cur- large number of parasitic plants, all of which rant family (Grossulariaceae), and yet others are photosynthetic but none the less obligate are herbaceous and insect-pollinated, for parasites. Some, like the sandalwood family example the stonecrops (Crassulaceae), (Santalaceae), attach to their hosts via peonies (Paeoniaceae) and saxifrages underground haustoria, whereas others, like (Saxifragaceae). The order has many species the mistletoes (Loranthaceae), grow directly with a particular type of vein endings in on the branches of their woody host plants. their leaves, but in general they are diverse Although most are parasites on woody in most traits. If not thought to be related to species, some, such as the Western Hamamelidae, then they were thought to be Australian Christmas tree (Nuytsia), attack related to the rosids in Rosales and clustered herbaceous plants (they are one of the few near Saxifragaceae. New results have shown trees in the areas where they grow). that a small tropical family, Peridiscaceae, Santalales have a long history of recognition are also related (Davis and Chase, 2004). as a group, and nearly all proposed classifi- cations have included them, more or less Dilleniaceae with the same circumscription as in APG (1998, 2003). Like other core eudicots, This small family is only mentioned here species in Santalales have organized flowers, because, although it is an unplaced-to-order but they have unusual numbers of whorls. core eudicot, it is the namesake of subclass Rosids and caryophyllids generally have one Dilleniidae, which figured importantly in whorl each of calyx (sepals), corolla (petals) many previous systems of angiosperm classi- and carpels, whereas there are two whorls of fication (e.g. Cronquist, 1981). They occupy stamens (sometimes with an amplification of

16 M. Chase a potentially critical position within the core on long filaments so that they can dangle in eudicots as sister to one of the other major the wind), determination of other relation- groups (i.e. asterids, caryophyllids or rosids) ships was made difficult, leading most work- or perhaps to a pair or all three, so, when ers to place them together. DNA studies they are placed, an understanding of their have been of major significance in sorting floral organization might be key to under- out the diverse patterns of relationships; standing floral evolution of the eudicots in some families are now placed among the general. In the three-gene analysis of Soltis non-core eudicots (e.g. Platanaceae in et al. (2000), they were sister to Proteales; Trochodendraceae, unplaced to Caryophyllales but this was not a clear order), Saxifragales (e.g. Daphniphyllaceae result. If additional gene data also place and Hamamelidaceae), rosids (most of the them in this position, they will be included ‘higher’ Hamamelidae such as Betulaceae in Caryophyllales. and Fagaceae in Fagales, see below) or even asterids (e.g. Eucommiaceae in Garryales). Rosids At least in the case of Hamamelidae, Like Carophyllales, rosids and asterids have botanists had the characters associated with a long history of recognition, and similarly wind pollination as the basis for placing the the DNA sequence studies have considerably families in one taxonomic category, but the enlarged the number of groups associated basis for Dilleniidae was always much with them (see below). In contrast to the weaker and less consistent among the Caryophyllales and the asterids, many authors who recognized the group. Basically groups of plants long thought to be rosids (and explaining their characters in APG ter- have been demonstrated to have relation- minology), they were core eudicots that ships to the first two groups, and thus the tended to have many petals and stamens, rosids have somewhat fewer families than in with the latter maturing centrifugally. In all many systems of classification. The addi- other respects, they were diverse and diffi- tional families have come mostly from the cult to place. With respect to the APG system group called by many previous authors the (1998, 2003), families of this subclass are dilleniids (e.g. in Cronquist, 1981, subclass now placed in either the rosids (e.g. Dilleniidae) and hamamelids (subclass Brassicaceae, Clusiaceae, Cucurbitaceae, Hamamelidae, sensu Cronquist). Before dis- Malvaceae and Passifloraceae) or asterids cussing the rosids, it is appropriate to first (Ericaceae, Primulaceae and Theaceae). The discuss these two groups that are not pre- only exceptions to this are Paeoniaceae and sent in the APG system. Dilleniaceae, which are Saxifragales and unplaced in the core eudicots thus far, Hamamelidae (Cronquist, 1981) con- respectively. Thus with respect to all previ- tained nearly all of the families of wind-pol- ous systems of angiosperm classification, that linated trees, including such well-known of APG (1998, 2003) does not contain in any families as the beeches and oaks (Fagaceae), form two of the previously recognized major birches (Betulaceae) and plane tree taxa, which have been shown by DNA stud- (Platanaceae). They were often split into ies to be polyphyletic (Chase et al., 1993; ‘lower’ and ‘higher’ Hamamelidae, in recog- Savolainen et al., 2000; Soltis et al., 2000). nition of their degree of advancement. The syndrome of wing pollination is highly con- Within the rosids, there are still several straining of floral morphology on a mechan- orders not yet placed to either of the two ical basis, and convergence in distantly larger groups, eurosid I and II: related families was always suspected. Crossosomatales, Geraniales and Myrtales. Nevertheless, since the syndrome is one Crossosomatales are a small order, with three associated with either great modification or families, none of which is well known. It is loss of many floral organs (e.g. petals are another of the APG orders that no one had nearly always absent and stamens are held predicted. Geraniales have four families, of which only Geraniaceae are well known (the temperate genera Geranium and largely South

Relationships between flowering plant families 17 African Pelargonium, the ‘geranium’ of com- nodules. This trait is important because merce, which are both important horticultur- these plants can thereby grow on poorer soil ally). On the other hand, Myrtales (13 families) and enrich it (e.g. farmers alternate crops so have several important families, including the that in some years they plant legumes, one combretum family (Combretaceae), the melas- of the major nitrogen-fixing families). It has tome family (Melastomataceae), the myrtle and been hypothesized (Soltis et al., 1995) that guava family (Myrtaceae) and the fuchsia and this trait evolved in the common ancestor of evening primrose family (Onagraceae). this clade and then was lost in many of the Melastomataceae and Myrtaceae are both large genera, although the reasons why such a and ecologically important in the tropics, valuable trait would be lost is not clear. The whereas Onagraceae are horticulturally alternative hypothesis, and perhaps the important. Onagraceae have been studied for more likely one, is that there are some pre- many years by several American botanists and conditions that are required for the trait to have become a minor model family. evolve and these were present in the com- mon ancestor; possession of the precondi- The remainder of the rosids are split into tions then made it more likely that the trait two major clades, which have been referred would evolve. If nitrogen fixation can be to as eurosid I and eurosid II. Alternative engineered in plants that currently are not names, fabids and malvids, have also been capable of this, then it is more likely that this suggested for these two clades, respectively. will be possible in non-fixing species in this Celastrales (three families) are another clade than those in other clades. order unique to the APG classification (in the sense of their circumscription). The rea- Cucurbitales (seven families) contain the sonably large spindle family, Celastraceae, is familiar cucumber and melon family the only one of any particular note in this (Cucurbitaceae) as well as the begonia family order, which is sister to one of the larger (Begoniceae), which is common in our gar- orders, Malpighiales (28 families). dens. They are sister to Fagales (seven fami- Malpighiales and Celastrales share a particu- lies), which are important (mostly) north lar seed type with a fibrous middle layer. temperate forest trees. These include the Seed characters appear to be significant tax- birch family (Betulaceae), the she-oak family onomic characters in the angiosperms as a (Casuarinaceae, one of the tropical members whole, but unfortunately they are relatively of this order), the beech and oak family poorly studied. Within Malpighiales, the (Fagaceae, with some tropical genera), the most important families are the mangosteen walnut, pecan and hickory nut family family (Clusiaceae or Guttiferae), a large (Juglandaceae) and the southern beeches tropical family with several species impor- (Nothofagaceae). These families are well tant for their fruit or timber, the spurge fam- known for their timbers as well as their fruits ily (Euphorbiaceae), the passionfruit family (nuts), and they are dominant members of (Passifloraceae) and the violet family many temperate and tropical ecosystems. (Violaceae). Also related to these two orders are Oxalidales (six families), in which the Fabales (four families) are important oxalis (Oxalidaceae) and elaeocarp because they include the legume family, (Elaeocarpaceae) families are placed. Both of which, as mentioned above, are capable of these are sources of ornamentals, and some fixing nitrogen and thus enriching many of species of oxalis are important weeds. The the soils in which they grow. They are also southern hemisphere cunon family important as food-producing plants and are (Cunoniaceae) includes some important tree grown as crops throughout the world. Beans species. and pulses are a good source of protein; soy- bean is widely grown and soya is a widely The rest of the families make up a clade used meat substitute. Many legumes, both that has been termed the ‘nitrogen-fixing herbs and woody species, are also common clade’ (Soltis et al., 1995) because at least ornamentals, and some of the tropical gen- some members of each order are known to era are timber species. Most are ecologically harbour nitrogen-fixing bacteria in root important throughout the world. The other

18 M. Chase large family of Fabales is the milkwort fam- them in a single order (Dahlgren, 1980). ily, Polygalaceae. Both of these families have This circumscription was so highly criticized highly characteristic zygomorphic flowers of by other taxonomists that in this next classifi- similar general construction, although no cation, he split them again into several unre- previous author had suggested that they lated orders. The basis for including them in were closely related. DNA studies were the a single order was simply due to the pres- first to place these families in one clade ence of mustards oils, which involve a com- (Chase et al., 1993). Milkworts curiously are plicated biosynthetic pathway for their unable to fix nitrogen. synthesis; chemists interested in plant nat- ural chemistry had long believed that it was Rosales (nine families) in the APG circum- highly unlikely that such a process could scription (2003) are radically different from have evolved so many times in distantly those of most previous systems (e.g. related groups (up to six times if you con- Cronquist, 1981). Among the important sider the placement of these families in the families are the rose family (Rosaceae), system of Cronquist, 1981). Thus DNA data which include many ornamentals as well as figured importantly in the recognition of this fruit-bearing species, such as apples, cher- circumscription of the order. The largest ries, peaches, plums, raspberries and straw- family in the order is the mustard family, berries. A few are also important timbers Brassicaceae (Cruciferae), which include the (e.g. cherry and white beam). Both Rosaceae well-known broccoli, Brussels sprouts, cab- and Rhamnaceae include a number of nitro- bage and cauliflower, all of which are gen-fixing genera, and the latter include a selected forms of the same species. In APG number of timber species as well as some (2003), the circumscription of Brassicaceae minor fruit-bearing genera (e.g. jujube). included the caper family (Capparaceae), but Circumscription of the last set of families in recent studies have shown that by segregat- Rosales is in flux, but these have long been ing a third family, Cleomaceae, it would then recognized as a natural group. Relative to be appropriate to reinstate Capparaceae as a their limits as used in APG (2003), the mari- recognized family. Other commonly encoun- juana and hops family (Cannabaceae) tered families of Brassicales are the papaya should now include the hop-hornbeam fam- (pawpaw) family (Caricaceae) and the nastur- ily (Celtidaceae), which has been split from tium family (Tropaeolaceae). Ulmaceae. The nettle family (Urticaceae) have a number of temperate herbs of minor Malvales (nine families) are well known importance and a larger number of tropical for their production (in various parts of the trees that are timber species; many are plants) of mucilaginous compounds (e.g. the sources of fibres. The fig and mulberry fam- original source of marshmallow is the marsh ily (Moraceae) are a mostly tropical group, mallow, a species in Malvaceae; sugar mixed which are important ecologically and as a with these polysaccharides is what was origi- source of fruits. Figs are well known for nally used to make the candy, but it is now their symbiotic relationships with their polli- artificially synthesized). Nearly all of the nators, fig-wasps, each species of which gen- nine families produce at least some of these erally has a one-to-one relationship with a compounds. The best-known family of the species of fig. This relationship is one of the order is the mallow and hibiscus family longest enduring known; it probably dates (Malvaceae), which before the application of back to 90 million years ago, when the first DNA data was typically split into four fami- fig-wasp fossils are known. lies, Bombacaceae, Malvaceae, Sterculiaceae and Tiliaceae. Chocolate is also a commer- In the second major clade of the rosids, cial product from a species in the family, and there are only three orders, Brassicales, okra is an edible fruit of a species of hibis- Malvales and Sapindales. Brassicales (15 fam- cus. A number of ornamentals are found in ilies, most of them small) include all of the the next largest family in Malvales, the families that produce mustard oils, but their thymelea family (Thymelaeaceae), which morphological traits were so diverse that include the daphne, whereas the next only one author ever previously included

Relationships between flowering plant families 19 largest family is the dipterocarps Caricaceae), but this is rarely encountered; (Dipterocarpaceae), which is the most rosids of most orders have two whorls of sta- important family of the Old World tropical mens that are rarely attached to the petals forests and produces timbers. (Celastraceae and Rhamnaceae are two exceptions, but they have lost different The last order of the eurosid II clade is whorls). The caryophyllids are more similar Sapindales (nine families), which are nearly to the asterids in some ways (seed and pollen all woody species, whereas Brassicales and to characters), but to rosids in others (e.g. lack a lesser degree Malvales have many herba- of fused petals). ceous species. The largest family of the order is that of the maple and litchi (Sapindaceae), Cornales (six families) were previously which is a largely tropical group; the well- associated with the rosids because of their known north temperate maples and horse unfused petals. The dogwood family chestnuts (buckeyes) are two exceptions to (Cornaceae) is the best known of the order this distribution. Another important family and is largely north temperate. The of tropical forest trees is the mahogany fam- hydrangea family (Hydrangeaceae) is well ily (Meliaceae), from which also comes an known for its ornamental species; it had been important insecticide, neem. The citrus or previously associated with Saxifragaceae by rue family (Rutaceae) is also an important nearly all authors. The loasa family woody group, but there are some herba- (Loasaceae) had been frequently placed near ceous species, such as rue itself, which is a the passionflower family (Passifloraceae); this temperate genus. Grapefruit, lemons, limes family includes a number of plants with sting- and oranges, as well as a number of minor ing hairs (such as the nettles, Urticaceae). fruits, are important commercially. The poi- son ivy and cashew family (Anacardiaceae) is Ericales (23 families) have previously another largely tropical group; the family is been split into as many as seven orders by well known for its highly allergenic oils, some authors (e.g. Cronquist, 1981; which cause severe and sometimes fatal reac- Diapensiales, Ebenales, Lecythidales, tions in many people. Cashews, mangoes and Polemoniales, Primulales, Theales and pistachios are important commercial mem- Sarraceniales), but DNA data do not dis- bers of the family. criminate among these clearly so the order has been broadly defined in APG (1998). Asterids Well-known families among Ericales include the heath and rhododendron family The second major group of eudicots is the (Ericaceae), ebony family (Ebenaceae), phlox asterids, which are subdivided into three family (Polemoniaceae), primula family major subgroups, only the last two of which (Primulaceae), North American pitcher have typically been considered to be mem- plants (Sarraceniaceae), zapote family bers of formally recognized asterid taxa. (Sapotaceae) and tea family (Theaceae). Asterids differ in a number of technical and Commercially important timber families chemical characters from the rosids, but include the ebonies and zapotes, and their flowers differ in having fused petals to Ericaceae include a number of ornamentals which a single whorl of stamens is typically (azaleas, ericas and rhododendrons). attached. This sympetalous corolla fused to the stamens is sometimes modified late in flo- In the first of the two core or euasterid ral development, such that when these flow- clades (four orders), which some authors ers open they appear to have free petals, but have termed the lamiids (Bremer et al., in terms of their development they are none 2001), there are four families unplaced to the less derived from a fused condition (this order, the most important of which is the situation has been termed ‘early sympetaly’ borage family (Boraginaceae) and in which a by Erbar and Leins (1996)). Some rosids can number of ornamentals are included (for- also be sympetalous (e.g. the papaya family, get-me-not, etc.). Garryales is a small order with two small families: Garryaceae include the ornamentals Garrya and Aucuba. Eucommia (Eucommoniaceae) is a wind-

20 M. Chase pollinated genus formerly placed in subclass lavender (Lavendula), rosemary (Rosmarinus), Hamamelidae (Cronquist, 1981; see above). marjoram and oregano (Oreganum) and sage Gentianales (five families) include the milk- (Salvia), the last of which also has a number weeds (Apocynaceae), gentians (Gentianaceae) of ornamentals. and the fifth largest family in the angiosperms, the madders (Rubiaceae). The last of the lamiid orders is Solanales Milkweeds are a largely tropical family of (five families), which include the morning glory vines and trees, some of which are locally family (Convolvulaceae) and the potato and important timbers; the succulent milkweeds tomato family (Solanaceae). Convolvulaceae of South Africa are common in cultivation also contain the sweet potato (Ipomoea), which and include the carrion flowers that attract is of major importance as a staple (starch) flies to pollinate them and which deceive the crop in some tropical regions (e.g. New female flies so well that they lay eggs on Guinea). In addition to potato and tomato what they think is a rotting animal carcass. (Solanum), Solanaceae also include aubergine Gentians are common herbs, some orna- (also Solanum), sweet and hot peppers mental, in temperate zones but include as (Capsicum) and tomatillo (Physalis), as well as well some tropical trees. Rubiaceae are many ornamentals, such as petunia (Petunia), largely tropical woody plants, but in the poor man’s orchid (Schizanthus) and devil’s temperate zones there are some herbs; trumpet (Brugmannsia). Solanaceae are also many are important timber species, such as well known for their drug plants, including teak (Tectonia), as well as medicinal plants belladonna (Atropa) and tobacco (Nicotiana), and coffee (Coffea species). the most widely used drug plant of all. The largest order of the lamiids is The last clade of euasterids is the lobeli- Lamiales (21 families), which include the ids, which includes four orders. There are acanths (Acanthaceae), Catalpa and bignon still a number of small families that are not family (Bignoniaceae), African violet family yet placed in one of these orders (e.g. the (Gesneriaceae), mints (Lamiaceae or escallonia family, Escalloniaceae, and brunia Labiatae), olive and lilac family (Oleaceae), family, Bruniaceae). Two of the orders were veronica family (Plantaginaceae), snapdragon previously not considered asterids at all by family (Scrophulariaceae), broomrape family most previous authors (e.g. Cronquist, 1981; (Orobanchaceae) and verbena and teak fam- Thorne, 1992). Apiales and Aquifoliales ily (Verbenaceae). Scrophulariaceae have were usually allied to rosid families, been much studied and remain problematic although authors such as Cronquist (1981) in their circumscription. Orobanchaceae admitted that at least the former was transi- include the obligate, non-photosynthetic gen- tional between his subclasses Rosidae and era that most previous authors assigned Asteridae. Apiales (ten families) include the there, but along with these the former ‘hemi- carrot family (Apiaceae or Umbelliferae), ivy parasitic’ genera, such as the Indian paint family (Araliaceae) and pittosporum family brush (Castileja) and lousewort (Pedicularis), (Pittosporaceae). Apiaceae include mostly which had been included in herbaceous plants, and in addition to carrots Scrophulariaceae, have been transferred to (Daucus), they produce parsnips (Pastinaca) Orobanchaceae. The genera related to and fennel (Foeniculum, which is both a veg- Veronica, such as foxglove (Digitalis, the source etable and a herb). Other genera provide of the heart medicine, digitalin), are now con- herbs, such as dill (Anethum), parsley sidered to be Plataginaceae, which had for- (Petroselinum) and chervil (Anthriscus). merly been a monogeneric family. A number Araliaceae include the common English ivy of other segregates from Scrophulariaceae (Hedera; other types of ivy, such as Virginia have recently been proposed as well, such as creeper and poison ivy, are included in the pocketbook plant (Calceolariaceae). Vitaceae and Anacardiaceae, respectively). Further changes are likely as more studies are Other well-known members of Araliaceae completed. The mints (Lamiaceae) are the include aralia (Aralia) and ginseng (Panax), sources of many herbs, such as basil (Ocimum), the latter of which is considered an impor- tant tonic in the Far East.

Relationships between flowering plant families 21 Aquifoliales (five families, none of them Dipsacales (two families) is the last order of asterids. Caprifoliaceae and Adoxaceae large) include the holly family are the only included families, the former often treated as five more narrowly circum- (Aquifoliaceae) and the stemonura family scribed families. The former includes elder (Sambucus, used as a fruit; the flowers as a (Stemonuraceae, which is largely tropical). drink) and snowball bush (Viburnum). Caprifoliaceae include a number of orna- Two small families, Helwingiaceae and mentals, such as honeysuckle (Lonicera), abelia (Abelia), morina (Morina) and scabious Phyllonomaceae, include shrubs with flow- (Scabiosa). Dipsacus, teasel, has in the past been used to card wool, but is now an intro- ers borne in the middle of their leaves. duced weed in many parts of the world. Holly (Ilex) was important in the religious The Future rituals of the pre-Christians in Europe and One of the major questions regarding the APG system of classification is its stability. As later became identified with Christmas is obvious when comparing the original APG classification with APG II, only a few changes because its leaves persisted through the win- have been made. The orders described in the original have been uniformly retained, ter. One species of Ilex is commonly used as and even family circumscription has changed relatively little. The evidence produced by 10 a tea in southern South America, principally years of phylogenetic studies on the angiosperms has been remarkably consistent, Argentina, and others are frequently used as and thus a system built upon such a base is likely to be stable. Changes anticipated are ornamentals. continued refinement of familial circum- scriptions and likely recognition of some Asterales (11 families) have the greatest small orders for families that consistently fall outside the major clades, such as number of species in the asterids because Berberidopsidales (composed of just Berberidosidaceae and Aextoxicaceae). they contain the daisy and sunflower family There are also some small families and (Asteraceae or Compositae), which is one of genera that should be placed once suitable material becomes available for DNA work, the two largest families of flowering plants but in the general scheme of things these are trivial matters. From time to time a (the other is the orchids, Orchidaceae). genus misplaced within a family is also dis- covered. For example, Aphanopetalum was Asteraceae are economically and ecologically considered a member of Cunoniaceae by most authors, but it does not fall into either important and contain herbaceous plants as Cunoniaceae or even Oxalidales, and instead is related to Saxifragales, in which well as woody genera. Helianthus is the sun- APG II placed it. None the less, these sorts of change have little effect on the overall sys- flower, which is cultivated for its seeds that tem and do not complicate matters greatly (most people did not know Aphanopetalum so are rich in proteins, as well as the Jerusalem such changes have little effect on users of the APG classification). artichoke (Jerusalem in this case is a corrup- tion of hira sol, sunflower in Spanish); Cynara is the true artichoke, which is the large flower head that is harvested before it opens; Lactuca is lettuce, and Chicorium is chicory. Other species are important weeds, such as dandelion (Taraxacum), sticktight (Bidens), English daisy (Bellis) and ragweed (ironically named Ambrosia). Many cultivated ornamen- tals are also members of Asteraceae, includ- ing marigold (Calendula), African marigold (Tagetes, which is native to Mexico, in spite of its common name), dahlia (Dahlia), cosmos (Cosmos), batchelor’s button (Centaurea), daisy and chrysanthemum (Chrysanthemum) and aster (Aster and several segregate genera). Other important families in Asterales include the bluebell family (Campanulaceae), the goodenia family (Goodeniaceae) and the bog-bean family (Menyanthaceae). Campanulaceae include several ornamentals, such as lobelia and cardinal flower (both Lobelia), Canterbury bells (Campanula) and bellflowers (Platycodon). Goodeniaceae is an Australasian family that has produced some ornamentals, such as Scaevola.

22 M. Chase The major improvement that is needed in major groups of eudicots (e.g. asterids, the APG system is for there to be greater caryophyllids and rosids) and the basal confidence in the higher-level relationships clades of angiosperms (Chloranthaceae, mag- (above orders) so that a formal nomenclature noliids, monocots, eudicots and probably can be adopted for superorders or sub- Ceratophyllaceae) are not clear. classes, but to achieve this will require addi- Collaborative efforts are under way to tional data collected in an organized manner. address these uncertainties, and in the future At present, the relationships among the we can expect clarification of these issues. References Albert, V.A., Williams, S.E. and Chase, M.W. (1992) Carnivorous plants: phylogeny and structural evolution. Science 257, 1491–1495. APG (Angiosperm Phylogeny Group) (1998) An ordinal classification of the families of flowering plants. Annals of the Missouri Botanical Garden 85, 531–553. APG (Angiosperm Phylogeny Group) (2003) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnaean Society 141, 399–436. Backlund, A. and Bremer, K. (1998) To be or not to be: principles of classification and monotypic plant fami- lies. Taxon 47, 391–400. Barkman, T.J., Chenery, G., McNeal, J.R., Lyons-Weiler, J., Ellisens, W.J., Moore, M., Wolfe, A.D. and dePamphilis, C.W. (2000) Independent and combined analyses of sequences from all three genome compartments converge on the root of flowering plant phylogeny. Proceedings of the National Academy of Sciences USA 97, 13166–13171. Bremer, K., Backlund, A., Sennblad, B., Swenson, U., Andreasen, K., Hjertson, M., Lundberg, J., Backlund, M. and Bremer, B. (2001) A phylogenetic analysis of 100+ genera and 50+ families of euasterids based on morphological and molecular data with notes on possible higher level morphological synapomor- phies. Plant Systematics and Evolution 229, 137–169. Chase, M.W., Soltis, D.E., Olmstead, R.G., Morgan, D., Les, D.H., Mishler, B.D., Duvall, M.R., Price, R.A., Hills, H.G., Qiu, Y.-L., Kron, K.A., Rettig, J.H., Conti, E., Palmer, J.D., Manhart, J.R., Sytsma, K.J., Michael, H.J., Kress, W.J., Karol, K.G., Clark, W.D., Hedrén, M., Gaut, B.S., Jansen, R.K., Kim, K.J., Wimpee, C.F., Smith, J.F., Furnier, G.R., Strauss, S.H., Xiang, Q.Y., Plunkett, G.M., Soltis, P.S., Swensen, S.M., Williams, S.E., Gadek, P.A., Quinn, C.J., Eguiarte, L.E., Golenberg, E., Learn, G.H. Jr, Graham, S.W., Barrett, S.C.H., Dayanandan, S. and Albert, V.A. (1993) Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80, 528–580. Chase, M.W., Fay, M.F. and Savolainen, V. (2000a) Higher-level classification in the angiosperms: new insights from the perspective of DNA sequence data. Taxon 49, 685–704. Chase, M.W., Soltis, D.E., Soltis, P.S., Rudall, P.J., Fay, M.F., Hahn, W.H., Sullivan, S., Joseph, J., Givnish, T., Sytsma, K.J. and Pires, J.C. (2000b) Higher-level systematics of the monocotyledons: an assessment of current knowledge and a new classification. In: Wilson, K.L. and Morrison, D.A. (eds) Monocots: Systematics and Evolution. CSIRO, Melbourne, pp. 3–16. Cronquist, A. (1981) An Integrated System of Classification of Flowering Plants. Columbia University Press, New York. Dahlgren, R.M.T. (1980) A revised system of classification of the angiosperms. Botanical Journal of the Linnaean Society 80, 91–124. Dahlgren, R.M.T., Clifford, H.T. and Yeo, P.F. (1985) The Families of the Monocotyledons: Structure, Evolution and Taxonomy. Springer, Berlin. Davis, C.C. and Chase, M.W. (2004) Elatinaceae are sister to Malpighiaceae, and Peridiscaceae are mem- bers of Saxifragales. American Journal of Botany 91, 149–157. Duvall, M.R., Clegg, M.T., Chase, M.W., Lark, W.D., Kress, W.J., Hills, H.G., Eguiarte, L.E., Smith, J.F., Gaut, B.S., Zimmer, E.A. and Learn, G.H. Jr (1993a) Phylogenetic hypotheses for the monocotyledons con- structed from rbcL sequences. Annals of Missouri Botanical Garden 80, 607–619.

Relationships between flowering plant families 23 Duvall, M.R., Learn, G.H. Jr, Eguiarte, L.E. and Clegg, M.T. (1993b) Phylogenetic analysis of rbcL sequences identifies Acorus calamus as the primal extant monocotyledon. Proceedings of the National Academy of Sciences USA 90, 4611–4644. Duvall, M., Mathews, S., Mohammad, N. and Russell, T. (2005) Placing the monocots: conflicting signal from trigenomic analyses. In: Columbus, J.T. (ed.) Proceedings of the Third International Conference on Monocots. Aliso Press, Los Angeles, California. Erbar, C. and Leins, P. (1996) Distribution of the character state ‘early sympetaly’ and ‘late sympetaly’ within the ‘sympetalae tetracyclicae’ and presumably allied groups. Botanica Acta 109, 427–440. Grayum, M.H. (1987) A summary of evidence and arguments supporting the removal of Acorus from the Araceae. Taxon 36, 723–729. Hutchinson, J. (1934) The Families of Flowering Plants. Oxford University Press, Oxford. Hutchinson, R. (1967) The Genera of Flowering Plants. Clarendon Press, Oxford. Qiu, Y.-L., Lee, J., Bernasconi-Quadroni, F., Soltis, D.E., Soltis, P.S., Zanis, M., Chen, Z., Savolainen, V. and Chase, M.W. (1999) The earliest angiosperms: evidence from mitochondrial, plastid and nuclear genomes. Nature 402, 404–407. Qiu, Y.-L., Lee, J., Bernasconi-Quadroni, F., Soltis, D.E., Soltis, P.S., Zanis, M., Chen, Z., Savolainen, V. and Chase, M.W. (2000) Phylogeny of basal angiosperms: analysis of five genes from three genomes. International Journal of Plant Sciences 161, S3–S27. Savolainen, V., Chase, M.W., Hoot, S.B., Morton, C.M., Soltis, D.E., Bayer, C., Fay, M.F., de Bruijn, A.Y., Sullivan, S. and Qiu, Y.-L. (2000) Phylogenetics of flowering plants based upon a combined analysis of plastid atpB and rbcL gene sequences. Systematic Biology 49, 306–362. Soltis, D.E., Soltis, P.S., Morgan, D.R., Swensen, S.M., Mullin, B.C., Dowd, J.M. and Martin, P.G. (1995) Chloroplast gene sequence data suggest a single origin of the predisposition for symbiotic nitrogen fix- ation in angiosperms. Proceedings of the National Academy of Sciences USA 92, 2647–2651. Soltis, D.E., Soltis, P.S., Chase, M.W., Mort, M.E., Albach, D.C., Zanis, M., Savolainen, V., Hahn, W.H., Hoot, S.B., Fay, M.F., Axtell, M., Swensen, S.M., Nixon, K.C. and Farris, J.S. (2000) Angiosperm phy- logeny inferred from a combined data set of 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnaean Society 133, 381–461. Soltis, P.S., Soltis, D.E. and Chase, M.W. (1999) Angiosperm phylogeny inferred from multiple genes as a tool for comparative biology. Nature 402, 402–404. Takhtajan, A. (1997) Diversity and Classification of Flowering Plants. Columbia University Press, New York. Thorne, R.F. (1992) An updated phylogenetic classification of the flowering plants. Aliso 13, 365–389. Zanis, M.J., Soltis, D.E., Soltis, P.S., Mathews, S. and Donoghue, M.J. (2002) The root of the angiosperms revisited. Proceedings of the National Academy of Sciences USA 99, 6848–6853.

3 Diversity and evolution of gymnosperms Ken Hill Royal Botanic Gardens, Mrs Macquaries Road, Sydney, NSW 2000, Australia Introduction of orders, classes, subclasses, divisions, subdi- visions or phyla, giving rise to the different Lindley (1830) introduced the class spellings often seen in the literature (e.g Gymnospermae for that group of seed plants Cycadales, Cycadae, Cycadinae, Cycadophyta, possessing exposed or uncovered ovules as Cycadophytina). All subgroups including the one of four classes of seed plants. Two were flowering plants or angiosperms are treated the monocots and the dicots, with the here as divisions with the termination ‘-phyta’. Gymnospermae placed between these, and a fourth group (the Rhizanths) was added for A Monophyletic Group? a number of highly modified hemiparasites such as Rafflesia and Balanophora. Within the The extant seed plants (the Spermatophyta) class Gymnospermae, Lindley recognized have been shown to be a monophyletic five ‘natural orders’, Gnetaceae, Cycadaceae, group; that is, the entire group arose from a Coniferae, Taxaceae and Equisetaceae. single common ancestor, with initial radiation Equisetaceae have since been shown not to in the Late Palaeozoic (Stewart and Rothwell, be seed plants, Coniferae and Taxaceae have 1993). The five lineages recognized within been combined and an additional group has the seed plants have been shown to be mono- been introduced for the then unknown phyletic by most studies (e.g. Crane, 1988; ginkgo. This gives us the four divisions of Loconte and Stevenson, 1990; Qiu and gymnosperms recognized today, Cycadophyta, Palmer, 1999), although the status of the Ginkgophyta, Pinophyta and Gnetophyta, Gnetophyta and Pinophyta has been ques- with all of the flowering plants treated as the tioned by some recent molecular studies fifth division of seed plants, the Magnoliophyta (Bowe et al., 2000; Chaw et al., 2000; Rydin et (Judd et al., 2002). al., 2002; Soltis et al., 2002). However, exact relationships among these lineages and the Nomenclature pattern and chronology of divergence remain unclear. A number of morphological and The gymnosperms have been variously molecular cladistic studies published over the placed in a class Gymnospermae or a division past 10 years on all or part of the Gymnospermophyta. Within this group, the Spermatophyta differ in details of divergence, subgroups have been recognized at the rank and no consensus is yet available (Fig. 3.1). © CAB International 2005. Plant Diversity and Evolution: Genotypic and 25 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

26 K. Hill Fig. 3.1. Differing hypotheses published in recent years on the phylogenetic relationships of angiosperms and the four clades constituting the gymnosperms. (a) Parenti, 1980; (b) Hill and Crane, 1982; (c) Crane, 1985a; Doyle, 1998a; (d) Doyle and Donoghue, 1986; (e) Loconte and Stevenson, 1990; (f) Qiu et al., 1999; (g) Soltis et al., 2002; (h) Goremykin et al., 1996; (i) Chaw et al., 1997; Bowe et al., 2000; (j) Rydin et al., 2002. Although regarded as a natural group for placing the Gnetophyta on the stem lineage many years, more recent morphological stud- of the Magnoliophyta (Crane, 1985a,b; Doyle ies have suggested that the Gymnospermae and Donoghue, 1986, 1992; Friedman, 1992; may be paraphyletic (Parenti, 1980; Hill and Donoghue, 1994; Doyle, 1996; Frohlich and Crane, 1982; Crane, 1985a,b; Doyle and Meyerwitz, 1997; Nickrent et al., 2000). Still Donoghue, 1986; Bremer et al., 1987; more recently, molecular phylogenetic studies Loconte and Stevenson, 1990; Nixon et al., have failed to corroborate the Anthophyte 1994; Rothwell and Serbet, 1994; Doyle, clade and in many cases have supported a 1996, 1998a,b). Central to these conclusions monophyletic Gymnospermae (Hasebe et al., was the recognition of the ‘Anthophyte’ clade, 1992; Goremykin et al., 1996; Chaw et al.,

Diversity and evolution of gymnosperms 27 1997, 2000; Stefanovic et al., 1998; Bowe et tive spaces’ (vacant or underutilized ecologi- al., 2000; Pryer et al., 2001; Soltis et al., cal niches) where they are suffering little 2002). Other recent studies support a close competition (Lewin, 1988). This allows phylogenetic relationship between ginkgo many, sometimes impractical, forms to and cycads and between Gnetales and develop and coexist. Later, when members conifers (Raubeson, 1998). However, sup- of the new lineage begin to compete, many port for the deeper phylogenetic structure of the early morphologies are removed by in all of these studies has been low and selection. In this case the adaptive space is results have been inconsistent, and the the dry land made accessible by the acquisi- monophyly of the Gymnospermae must still tion of ‘key adaptations’ allowing the plants be regarded as an open question (Rydin et to resist desiccation. The ‘key adaptations’ al., 2002). The four divisions will be dis- allowing this movement were the protection cussed separately below. of the fragile gametophyte stage of the life cycle by the evolution of pollen (which also Origins of the Gymnosperms eliminates dependency on water for fertil- ization) and the evolution of seeds (which Primitive plants (bryophytes and pterido- also enable the transport and protection of phytes and some of their even more primi- plant embryos). These features also elimi- tive algal progenitors) disperse by means of nate the fragile free-living gametophyte haploid spores, which establish a free-living stage from the life cycle. haploid gametophyte generation. These reproduce using motile flagellated sperm, This ‘experimental’ period lasted less which must swim through free water to find than 40 million years in the case of the seed and then fertilize ova (Raven et al., 1992). plants, and was largely over by the middle of This limits habitat to sites with free water. the Carboniferous. A few basic designs Their gametophyte generations are free-liv- became established and common. At this ing and also lack conductive vascular tis- point, lineages assigned to modern-day sues, are not differentiated into true organs Cycadophyta, Ginkgophyta and Pinophyta such as leaves and roots, have fixed stom- were in existence, and seed plants became ates that cannot close and have poorly more species rich. Although three of the five developed cuticles. They are consequently extant lineages were in existence by the end sensitive to environmental conditions, and of the Carboniferous, these progenitors dif- in particular cannot withstand desiccation. fered in many ways from their living descen- Evolution of seed plants represents a major dents (Florin, 1939; Miller, 1982). step in surviving different and varying envi- ronmental conditions. The advent of pollen Seed plants have many anatomical and eliminates dependency on water for fertil- reproductive features in common (Foster ization, and the seed allows wider and more and Gifford, 1989). The primary vascular successful dispersal. structure is a eustele (vascular bundles are organized into bundles of xylem internally The earliest known seed plants have been flanked by bundles of phloem on the out- reported from the Late Devonian side). All have secondary growth occurring (Famennian) of West Virginia (Rothwell et from a bifacial vascular cambium (produc- al., 1989). A number of other seed structures ing cells on both sides; in seed plants have been reported from the latest phloem is produced on the outside and Devonian and Early Carboniferous. Many xylem on the inside). Gametophytes are have unusual morphologies and show no wholly developed within the spore mother similarities to extant seed plants. cell walls with the exception of the sperm cell transfer by haustorial growth of the It has been suggested that extensive microgametophyte as the pollen tube. morphological variability often seen early in There is thus no free-living gametophyte the history of lineages occurs because the generation in seed plants; the entire game- new organisms are moving into new ‘adap- tophyte generation is sustained by or para- sitic on the sporophyte generation.

28 K. Hill Gymnosperm Characteristics lineage has been identified beyond the com- plex of Mesozoic seed plants clearly recog- Gymnosperms and angiosperms are differ- nizable as flowering plants from which all of entiated by several features (also see Foster the major lineages branched. The first and Gifford, 1989; Raven et al., 1992). unequivocally true angiosperms appear in the fossil record as both pollen and macro- 1. Gametophytes. Angiosperms are character- fossils in the Early Cretaceous (Beck, 1976; ized by simplified megagametophytes Friis et al., 1987). It has been suggested that reduced to only embryos in seeds, with the earlier angiosperms lived in upland, endosperm not derived from megagameto- possibly arid, regions where they were phyte tissue. However, gymnosperm gameto- unlikely to enter the fossil record (Cleal, phytes vary in structure between the divisions, 1989). The lack of any trace (including and may represent a gradational reduction in pollen, which can be widely transported) of complexity to the angiosperm condition. these pre-Cretaceous angiosperms through 2. Integuments. Most but not all the Carboniferous to Cretaceous time gap angiosperms have ovules completely sur- makes this hypothesis unlikely. If the gym- rounded by two integuments (bitegmic). nosperms are indeed monophyletic, their The exceptions apparently represent sec- sister group the angiosperms must date ondary loss of one integument. Most gym- from the same period, the Carboniferous. nosperms are unitegmic except in the This leaves a gap of over 150 million years gnetophytes, where the second integument with no fossil record of angiosperms – a may not be homologous with that of the period longer than their entire known fossil angiosperms. history. This could be either because the 3. Pollen wall morphology. The pollen of gymnosperms are not a natural group, or angiosperms is different from that of all because the stem lineage of the other seed plants in having a tectate– angiosperms lacked distinguishing columellate structure, in which the outer angiosperm synapomorphies. layer of the pollen wall (exine) is differenti- ated into two layers separated by columns. Reproductive Features Common to All In all other seed plants, pollen has a two- Gymnosperms layer structure (exine and intine) but within these layers structure is homogeneous. A feature of all gymnosperms is the aggrega- 4. Vessels. Xylem vessels characterize most tion of reproductive structures into separate angiosperms. However, several basal male and female cones or strobili angiosperm families lack vessels, while some (Chamberlain, 1935). ferns, Selaginella, Equisetum and Gnetum, all have vessel-like cells. Developmental mor- The male cone or microstrobilus in phology can differentiate vessels as probable cycads, conifers and ginkgo consists of a independent developments in these groups. cone axis with spirally arranged modified 5. Sieve elements with companion cells in leaves (microsporophylls), each bearing two phloem. Companion cells also occur in gne- or several pollen sacs (microsporangia) tophytes, but apparently again through a abaxially (on the underside). Gnetophyta different developmental pathway. have a more complex microstrobilus struc- 6. Carpels. Angiosperms have ovules ture that varies from family to family (Foster enclosed in carpels, whereas gymnosperms and Gifford, 1989). have exposed ovules (hence the name). However, the nature of the carpel varies Diploid cells inside the microsporangium widely and is not completely enclosing in (microsporocytes) undergo meiosis to pro- some cases. duce four haploid cells (microspores). Each microspore divides mitotically to produce a The origin and ancestry of the flowering microgametophyte, which becomes a pollen plants remains a mystery. No clear ancestral grain. The development of the microgameto- phyte occurs inside the microspore wall, and

Diversity and evolution of gymnosperms 29 this all occurs inside the microsporangium. The seed is made up of the embryo (2n), The outer wall of the microspore forms the the endosperm or food source that comes pollen wall. The microgametophyte thus from the megagametophyte (1n), and the consists of four nuclei: two prothallial nuclei, seed coat that is derived from the integu- one tube nucleus and one generative ment (from the 2n parent sporophyte). nucleus. The tube nucleus forms a pollen There is no double fertilization as in tube that digests its way thorough the mega- angiosperms, although gnetophytes show a sporangium. The generative nucleus divides double fertilization of a different kind (see mitotically to produce two sperm cells. below). The female cone or megastrobilus differs Cycadophyta in different gymnosperm groups (Foster and Gifford, 1989). Cycads have a simple struc- The cycads are a distinct monophyletic ture consisting of a cone axis, modified group, defined by the presence of cycasin, leaves (sporophylls) and two or several girdling leaf traces, simple megasporophylls, ovules on the underside of the sporophylls. the absence of axillary buds and the primary thickening meristem, which gives rise to the In conifers the female cone consists of a pachycaul habit (Stevenson, 1981, 1990). cone axis and cone scales (modified branches because they are subtended by a Present-day occurrence bract (a type of a leaf)). On the surface of a cone scale, there are two or several ovules. The modern cycads comprise two families Gnetophyta also have a compound cone with ten genera and about 300 species dis- scale structure, but in both male and female tributed across the warm, subtropical envi- cones. Female reproductive structures in ronments of the Americas, Africa, eastern ginkgo are highly reduced, and homologies Asia and Australasia. Most individual gen- with either the cycad megasporophyll or era, however, have more limited geographi- conifer cone scale have been disputed cal ranges. Many extant cycads show (Florin, 1951; Meyen, 1981). relictual distributions, although other groups are clearly actively evolving (Gregory An ovule consists of an integument, a and Chemnick, 2004). megasporangium, and a diploid megaspore mother cell (megasporocyte). The megas- Vegetative morphology porocyte divides meiotically to produce four haploid megaspores. Three of these degen- All living cycads are dioecious, long-lived, erate leaving one megaspore, which divides slow-growing woody perennials (also see mitotically many times to produce a Foster and Gifford, 1989; Norstog and megagametophyte. Specialized regions of Nichols, 1997). the megagametophyte will differentiate into two archegonia, and each of these archego- Stems are pachycaul (short and thick) nia will produce a single egg cell. with a broad pith and cortex and manoxylic Development of the megagametophyte wood (a wood type that contains abundant occurs inside the megaspore (all inside the parenchyma, typical of cycads), and may be megasporangium and the integument). subterranean or aerial. Internal stem struc- ture is characterized by a eustele with The pollen is carried by wind or insects endarch protoxylem, where a small amount to a mucilaginous droplet, which exudes of manoxylic wood is produced from a bifa- from the micropyles of the ovules. The cial vascular cambium. Leaf vasculature drop retracts (or evaporates) bringing the traces in the stem are girdling; that is, traces pollen into the pollen chamber where a haustorial pollen tube forms and the final stages of male gametophyte development take place. At fertilization, one of the sperm cells unites with the egg cell to produce a 2n zygote that will divide mitotically to pro- duce an embryo.

30 K. Hill arise from the stele at a point opposite the Female sporophylls are simple and entire point of leaf attachment and dichotomously (dissected in Cycadaceae), and carry naked, branch, with the two branches girdling the unitegmic ovules. Seeds are large, with a stele as they transverse the cortex and meet- two-layered testa: a fleshy and distinctly ing again before entering the petiole. This coloured outer layer, and a woody inner condition is known in no other extant group layer. The embryo is straight, with two of seed plants. Axillary buds are absent, and cotyledons, which are usually united at the vegetative branching is either dichotomous tips; germination is cryptocotular. or adventitious. Although widely accepted in the past to Cycad roots are heteromorphic, with con- be wind pollinated (Chamberlain, 1935), tractile and coralloid roots in addition to recent studies in several regions indicate normally functioning roots. Contractile tis- that cycads are mostly insect pollinated, sue is present in roots and, to a lesser often by closely commensal beetles (Norstog extent, stems, especially in juvenile plants et al., 1986; Tang, 1987; Donaldson et al., (Stevenson, 1980). Coralloid roots are highly 1995; Stevenson et al., 1998). This contrasts modified roots, with apogeotropic growth with both Ginkgo and the conifers, all of and extensive dichotomous branching, with which are wind pollinated (Page, 1990). the branches shortened, thickened and Chemistry of the pollinator-attractants in modified to internally accommodate symbi- cycads is markedly different from that of any otic cyanobacteria (Nathanielsz and Staff, flowering plants (Pellmyr et al., 1991), sug- 1975). gesting an independent origin for this polli- nation syndrome. Leaves are large, spirally arranged, pin- nate, bipinnate or bipinnatifid, exstipulate Male gametophytes produce large, multi- or stipulate or with a stipular hood, loosely flagellate and motile sperm cells, sharing pubescent at least when young, and usually some similarities with those of Ginkgo but oth- arranged in crowns on the stem-apex. The erwise unlike those of any other seed plants. leaves are often scleromorphic, owing to the strong fibres, thick cuticle and thick hypo- Cycad seeds are large, with a fleshy outer dermis. Leaf-bases may be persistent or coat (sarcotesta) over a hard, stony layer abscisent, depending on species. Leaves are (sclerotesta), and copious haploid, mater- interspersed with scale-leaves (cataphylls), nally derived endosperm. The fertilized except in Stangeria and Bowenia. embryo develops slowly but continuously until germination, with short-term chemical The pachycaul habit of modern-day inhibition of germination by the sarcotesta cycads is thought by some to be a Tertiary but no real dormancy (Dehgan and Yuen, development, many Mesozoic cycads having 1983). This makes seeds relatively short- dense wood and a leptocaul habit lived and subject to damage by desiccation. (Delevoryas, 1993). Reproductive morphology Dispersal Sporophylls of both sexes are simple and The fleshy sarcotesta attracts animals, spirally arranged in determinate strobilate mainly birds, rodents, small marsupials and structures (except in Cycadaceae) carried on fruit-eating bats, which serve as dispersal stem apices. The strobilate structure is lack- agents (Burbidge and Whelan, 1982; Tang, ing in Cycadaceae, with flushes of sporo- 1987). In most cases, the fleshy coat is eaten phylls developing at the stem apex in the off the seed and the entire seed is not con- same manner as flushes of leaves. The abax- sumed. Dispersal is consequently limited to ial surfaces of male sporophylls carry the usually short distance that the animals numerous sporangia in two ‘patches’ that can carry the seed. open by slits. Pollen is cymbiform, monosul- cate and bilaterally symmetrical. Cycas subsection Rumphiae has seeds with a spongy endocarp not seen elsewhere among the cycads (Guppy, 1906; Dehgan

Diversity and evolution of gymnosperms 31 and Yuen, 1983; Hill, 1994), which gives a Zamiaceae potential for oceanic dispersal, and it has been demonstrated that seeds maintain via- Zamiaceae shows a distinct break into bility after prolonged immersion in sea Laurasian and Gondwanan elements, possi- water (Dehgan and Yuen, 1983). Subsection bly from an ancestral disjunction resulting Rumphiae is the only subgroup of the genus from the breakup of Pangaea. Fossil evidence to occur on oceanic islands, and is widely places the extant genera in Australia at least distributed through the Indian and western back into the Eocene (Cookson, 1953; Hill, Pacific oceans, as well as all non-mainland 1978, 1980; Carpenter, 1991). Macrozamia parts of South-east Asia (Hill, 1994). has also speciated widely, with 38 species rec- ognized in Australia. Many species are com- Distribution and ecology ponents of complexes with narrow geographic replacement patterns, suggesting Cycad plants are long-lived and slow grow- that speciation is active and ongoing. ing, with slow recruitment and population turnover. The fleshy and starch-rich stems Bowenia and Stangeria were placed in are highly susceptible to fungal attack, and Stangeriaceae, but more recent studies have almost all species grow in well-drained soils. failed to corroborate their sister relation- Habitats range from closed tropical forests to ship, and have indicated that both genera semideserts, the majority in tropical or sub- may be best included in Zamiaceae. Both are tropical climates in regions of predominantly Gondwanan, with one genus in Australia summer rainfall. Cycads often occur on or and another in southern Africa. Bowenia are restricted to specialized and/or localized occurs as understorey shrubs in moist euca- sites, such as nutritionally deficient sites, lypt woodlands or forests, or in closed meso- limestone or serpentinite outcrops, beach phyll forests. Fossil evidence places the dune deposits or precipitously steep sites. genus Bowenia in southern Australia in the Early Tertiary (Hill, 1978). Contractile roots are present in all cycads (above), particularly in juvenile plants. Evolution and fossil record These draw the sensitive growing apex of seedlings below the soil surface, affording While the extant cycads have been clearly protection against drought and the fires shown to be a monophyletic group by both that are a frequent feature of many cycad morphological and molecular studies habitats. (Stevenson, 1990; Chase et al., 1993), ances- try and relationships of the group remain Coralloid roots host symbiotic cyanobac- unclear. The group is acknowledged as teria, which fix atmospheric nitrogen and extremely ancient, with a fossil record contribute to the nutrient needs of the extending back to the Early Permian (Gao plant. This provides an advantage in the and Thomas, 1989). The Palaeozoic and nutritionally deficient soils occurring in Mesozoic cycads were, however, very differ- many cycad habitats. ent from those of the present day, and fossil evidence of the extant genera is known only Cycadaceae from the Tertiary. The cycads have been proposed as the sister group to all other liv- The monogeneric Cycadaceae is apparently ing seed plants (Nixon et al., 1994), although Laurasian in origin, and relatively recently other studies have suggested different rela- dispersed into the Australasian region. This tionships (Pryer et al., 2001). is supported by the fossil record, with Cycas fossils known only from the Eocene of China Relationships among the cycad genera and Japan (Yokoyama, 1911; Liu et al., are still not well understood. A detailed clas- 1991). Australia stands out as a major centre sification of taxa at and above generic level of speciation for Cycas, with some 27 of the c. based on morphological and molecular data 100 species. has been presented by Stevenson (1992), although more recent molecular studies

32 K. Hill indicate somewhat different relationships Ginkgophyta (Hill et al., 2003; Rai et al., 2003). Present-day occurrence Cycads first appeared in the Late Carboniferous, and are thought to have The Ginkgophyta is represented today by a arisen from the Medullosan seed ferns single species, Ginkgo biloba, which is (Taylor and Taylor, 1993). Fossils from this restricted to central China in the wild, but period are somewhat problematical and extensively propagated as an ornamental unlike later or modern cycads. Spermopteris is tree. There have been suggestions that it has characterized by two rows of ovules attached been extinct in the wild for centuries and to the abaxial surface of Taeniopteris, a com- maintained only in cultivation in temples. mon simple leaf type from the Upper However, there have also been reports of Paleozoic. This plant has been placed with wild occurrences in the Tianmu Mountains the cycads on the basis of foliar features, in Zhejiang province (Fu et al., 1999). mainly the haplocheilic stomata, in which the guard mother cell gives rise to only two Vegetative morphology guard cells. Venation patterns and a non- bifurcating leaf-base also represent derived Ginkgophytes are large, long-lived, decidu- characters shared between Taeniopteris and ous, dioecious trees. Roots are fibrous to cycads. However, haplocheilic stomata also woody and undifferentiated. Stems are dif- occur in several other types of seed plants, ferentiated into long and short shoots, most such as conifers, ginkgoes, Ephedra, glos- probably in a parallel development to that sopterids and Cordaites. Fossil foliage present in some conifers and angiosperms. assigned to cycads is commonly pinnately In short shoots, internodes are very short, in compound (e.g. Nilssonia), but includes some contrast to long-internode long shoots. simple, entire leaves (e.g. some Nilssonia and Secondary xylem in long shoots is pycnoxylic Taeniopteris). Cycad-like pinnate fossil foliage like that of conifers and Cordaites; wood in of the common Mesozoic group, short shoots is rich in parenchyma, Bennetitales, is separated by possession of approaching the manoxylic condition of syndetocheilic stomata, in which guard cycads. Leaves are exstipulate with multiple mother cells produce four guard cells, a fea- dichotomous venation and double vascular ture shared with angiosperms. traces and develop on both long shoots and short shoots although the shape of leaves During the Permian, fossil genera with varies somewhat between long and short greater resemblance to living cycads begin shoots. Leaves are simple although at times to appear, for example, Crossozamia. Among quite deeply lobed and triangular in shape these are taxa that produce megasporo- with veins that bifurcate to fill space as the phylls in a helical arrangement similar to the wedge of the leaf widens. Ginkgo has a lepto- simple ovulate cones of extant Zamia, and caul habit with well-developed, dense wood, taxa with foliaceous megasporophylls closely and also has well-developed axillary buds resembling extant Cycas. and branching. Resin canals are absent. The cycads reached their peak in species Reproductive morphology richness and ecological importance during the Mesozoic, and have been declining since Ginkgo biloba is dioecious, a character in com- that time (Harris, 1976). Many cycad genera mon between Ginkgo and the living cycads. from the Mesozoic have been reconstructed Pollen is produced in paired microsporangia from fairly complete fossils, for example (also termed sporangiophores) on simple Leptocycas. Based on these reconstructions, stalked microsporophylls, which are spirally some workers think that the Mesozoic cycads in general had relatively slender trunks with widely spaced leaves that abscised, and that modern cycads with the short, thick stems did not evolve until the Tertiary (Delevoryas, 1993).

Diversity and evolution of gymnosperms 33 aggregated into simple catkin-like strobili ancient Ginkgo leaves closely resemble mod- (lacking bracts), which are also carried on ern Ginkgo. Others are highly dissected and short shoots. The pollen grains are monosul- resemble the typical Ginkgo leaf but without cate, spherical and wind-dispersed (see also tissue between the veins (Tralau, 1968). In Foster and Gifford, 1989). living Ginkgo, leaf shape is highly variable, suggesting that more taxa may have been Ovules are erect and borne in pairs sub- described on the basis of leaves only than tended by a collar of uncertain origin on might have actually existed. axillary stalks on short shoots. The ovule has a single integument that becomes three-lay- Pinophyta (The Conifers) ered and develops in the seed into a fleshy outer sarcotesta, a stony inner sclerotesta The conifers are uniquely defined by the and a thin endotesta. This is superficially reduced (non-megaphyllous) leaves, the similar to integument differentiation in the presence of resin canals, the compound cycads and Medullosa, and is an adaptation female sporophylls, and the undifferentiated to animal dispersal. shoot apex of fertile axillary shoots. Molecular studies corroborate the mono- Development of the male gametophyte is phyletic nature of the Pinophyta most dis- very similar to that in the cycads except that tinctly in that all members show loss of the there are two prothallial cells instead of one. inverted repeat unit in the chloroplast The mature spermatozoid is similar to the genome (Raubeson and Jansen, 1992). cycad sperm, but it is smaller and has only 2.5 turns of the spiral, compared with 5 or 6 Present-day occurrence in the cycads. Numerous flagella (10,000–12,000 in cycads, uncounted in Ginkgo) are attached along the spiral. Evolution and fossil record Globally, there are about 650 species of conifers. These are placed in 68 genera in The ginkgophytes first appeared in the fossil seven families (Farjon and Page, 2001). record in the Permian and became impor- tant components of Mesozoic ecosystems Conifers occur on all continents except worldwide, apparently reaching maximum Antarctica, but their abundance is unevenly diversity in the Jurassic (Thomas and Spicer, distributed in terms of both individuals and 1986). Today this group is represented by a taxa (Table 3.1). Where the vast boreal single species, Ginkgo biloba. conifer forests stretch across continents and contain billions of trees, they sustain no A range of Permian and Mesozoic more than a handful of species. In contrast, ginkgophytes has been described based on more southerly latitudes in the northern leaves, wood and some reproductive struc- hemisphere and all of the southern hemi- tures (Thomas and Spicer, 1986). Some sphere have either scattered conifer forests, Table 3.1. The extant conifer families: diversity, distribution and earliest appearance in the fossil record. Family Time range Genera/species Present distribution Araucariaceae Cretaceous to recent 3/37 Southern hemisphere Cephalotaxaceae Jurassic to recent 1/5 Northern hemisphere Cupressaceae Triassic to recent 30/157 Both hemispheres Pinaceae Cretaceous to recent 10/250 Northern hemisphere Podocarpaceae Triassic to recent 18/180 Southern hemisphere Sciadopityaceae Jurassic to recent 1/1 Northern hemisphere Taxaceae Jurassic to recent 5/17 Northern hemisphere Note: distributions are given for extant taxa only.

34 K. Hill or mixed conifer/hardwood forests in which thick, and stomata sunken. Both epidermal conifers occur in low densities, dispersed and hypodermal cells are frequently ligni- among other trees or shrubs. Many species fied. Venation appears parallel and leaf vas- occupy very small areas, often as relict pop- culature does not show more than one ulations of once greater abundance. Some discrete order, although highly variable, and areas have a high diversity of species, but venation is often characteristic at the family hardly any of these species are abundant or genus level. When numerous vascular enough to form forests of any appreciable bundles are present, a clear midrib is absent size. A good example is New Caledonia in and vasculature is often dichotomously the south-west Pacific, an island with 43 branched. Leaf bases may or may not be species of conifers, all endemic, in an area persistent, reflecting presence or absence of about the size of Wales. Mexico has 42 an abscission layer. species of pines (Pinus), compared with eight species in all of Canada and Alaska. About Wood anatomy is relatively uniform across 200 species of conifers are restricted to the the conifers. Xylem is composed entirely of southern hemisphere, where vast conifer tracheids and wood is generally parenchyma forests are unknown. It is this scattered poor and thus pycnoxylic (dense wood that diversity that is most threatened with extinc- contains little parenchyma). Tracheids of tion. Families and genera are unevenly dis- many conifers have a characteristic circular tributed and show a number of relictual bordered pitting on element walls that has biogeographic patterns. been used to identify fossil conifer wood to the family level. Both xylem and phloem tis- Despite the relatively low numbers in sues are characteristic at the family level, with comparison with the flowering plants, the slight differences between the families conifers are an economically important (Hardin et al., 2001). group of plants. Vegetative morphology Reproductive morphology All conifers are woody perennials with aerial Extant conifers are monoecious, with stems. Roots are fibrous to woody and undif- strongly dimorphic, unisexual, strobilate ferentiated. Shoots are similar in all conifer reproductive structures. families, with a leptocaul habit and pyc- noxylic secondary xylem. An architecture Pollen is borne in microsporangia carried with strongly differentiated plagiotropic and on simple microsporophylls lacking sub- orthotropic shoots is present in a number of tending bracts, which are arranged in deter- genera in different families, and several gen- minate strobili that may be solitary or era in the Pinaceae show a differentiation clustered, and axillary or terminal. into long and short shoots. Most conifers are Sporophylls carry two or more abaxial evergreen but a few (Larix and Metasequoia) microsporangia, which open by slits. are deciduous. Pollen grains are monosulcate, with Conifer leaves are quite diverse in shape somewhat different morphologies according and form, although basic structure is rather to family. Araucariaceae has spherical pollen uniform and no conifer has megaphyllous with little ornament. Cupressaceae, or compound leaves. Phyllotaxis may be Podocarpaceae and Pinaceae have saccate opposite, verticillate or spiral, and all conifer pollen with two distinct sacs or bladders that leaves are simple and without stipules. A sin- are thought to add buoyancy that assists gle vascular trace enters the leaf from the with wind pollination. stele. Axillary buds are present, although frequently vestigial or reduced to small pads Fertilization in all conifers is by wind. A of undifferentiated meristematic tissue pollen-drop mechanism is employed to cap- (Burrows, 1987). Cuticle is characteristically ture pollen, and pollen tubes then grow haustorially to the archegonia in the apical section of the embryo sac. The male gameto- phyte has non-motile sperm cells.

Diversity and evolution of gymnosperms 35 Megasporangiate structures are some- Distribution and ecology what more complex, with each fertile scale representing a modified short shoot with its Many extant conifer taxa show relictual dis- subtending bract (Florin, 1951). In all extant tributions and, at least at the familial and conifers, the modified shoot is reduced to a generic level, were far more diverse, wide- single fertile scale-leaf, with one, two or spread and abundant in the past than they more adaxial ovules (depending on family are today (Hill, 1995). Distribution and ecol- and genus). The fertile scale may be free ogy differ in the different families, and will from, or wholly or partially fused with, the be discussed below. subtending bract. Bract–scale complexes are borne in usually large, woody strobili, The conifer families although these may be highly reduced, in some cases to single bract–scale structures Pinaceae that show very little resemblance to the orig- inal strobilate structure. Not all conifers pro- Leaves are linear, acicular, spirally duce woody strobili and, in some taxa, arranged, or dimorphic, bract-like and aci- ovule-bearing structures are reduced to cular, on specialized short shoots. Male fleshy, berry-like forms (e.g. Juniper, Taxus, cones are lateral, comprising numerous Podocarpus). These structures are autapo- spirally arranged fertile scales, each scale morphies derived in relation to seed dis- with two abaxial microsporangia; pollen is persers. Ovules are unitegmic, and may be winged, with two air sacs (except in Larix erect or inverted. and Pseudotsuga). Female cones are termi- nal on specialized lateral shoots, compris- Endosperm is haploid, derived from the ing numerous spirally arranged imbricate female gametophyte tissue. Seeds are medium woody or dry fertile scales; each scale has a to large, dry or fleshy, and winged or wingless. free bract and two inverted adaxial ovules. Dispersal is by wind (dry seeds) or by birds or Seeds have a single terminal wing. small mammals (fleshy seeds). The fleshy Germination is phanerocotular. Cotyledons component that functions as an attractant is 3–18. variously derived, developing from an aril, epimatium or receptacle in different genera. A northern hemisphere family of ten The embryo is straight, with two or more genera; several genera include important cotyledons, and germination is cryptocotular forest community dominants, many of which or phanerocotular. are important economic timber sources. Many species are widely cultivated as orna- Modes of perennation range from mental plants. resprouting to obligate reseeding. Population dynamics of individual species Taxaceae are generally reflected by the dynamics of the host communities: species occurring in Pollen cones are axillary and solitary or clus- rainforests where fires are infrequent are tered, with sporophylls bearing 2–16 reseeders and show regular recruitment, microsporangia (pollen sacs); pollen is more whereas those in sclerophyllous habitats are or less spherical and not winged. Seed cones either resprouters or reseeders with episodic are reduced to one to two ovules subtended recruitment patterns responsive to fire by inconspicuous, decussate bracts. Each regimes. Mast seeding behaviour has also ‘cone’ has one erect ovule, which produces been reported in some subalpine taxa one unwinged seed with a hard seed coat par- (Gibson et al., 1995). Some episodic reseed- tially or wholly surrounded by a juicy, fleshy ers are dependent on fire for seed release or leathery aril. Embryos have two cotyledons. (some Callitris species; Bowman and Harris, 1995), although most are not. Most display Five genera and 17 species, mainly north- continuous recruitment in undisturbed ern hemisphere. Usually secondary compo- habitats and episodic recruitment in distur- nents of the vegetation. bance-prone habitats.

36 K. Hill Cephalotaxaceae in the extinct Lebachiaceae have been sug- gested as possible ancestral Cupressaceae Pollen cones are each subtended by one (Miller, 1982). The basal lineages formerly bract and are aggregated into axillary capit- known as Taxodiaceae are well represented ula of six to eight cones, each with 4–16 in the fossil record of the later Mesozoic and microsporophylls. Each microsporophyll has Tertiary; the lineages previously placed in two to four (usually three) pollen sacs bear- the Cupressaceae in the narrower sense are ing subspherical, non-saccate pollen. Seed much less so (Ohsawa, 1994). A number of cones are borne from axils of terminal bud extant genera of the latter have been scales, with one to six or occasionally up to recorded as fossils from the later Mesozoic eight long pedunculate cones per bud. Each and Tertiary, but these identifications are cone axis has several pairs of decussate based on similarites rather than synapomor- bracts, each bearing two erect, axillary phies, and must be regarded as doubtful ovules. Seeds ripen in their second year and (Thomas and Spicer, 1986). Many fossils at are drupelike and completely enclosed by a first placed in the narrow Cupressaceae have succulent aril. Embryos have two cotyledons, since been proven to belong to the extinct and germination is epigeal. conifer family Cheirolepidaceae (Miller, 1988). Cephalotaxaceae differs from Taxaceae in Male cones are small and comprise oppo- its seed cones, which have several two-ovu- site, whorled or spirally arranged scales with late bracts, instead of a single fertile, one- two to nine microsporangia on the abaxial ovulate bract. surface. Pollen is not winged or saccate. Female cones comprise one or more (c. 20) One genus and five species, northern fertile scales, each with a fully fused bract. hemisphere. Also usually secondary compo- Cone scales are alternate, opposite or nents of the vegetation. whorled in the same phyllotaxis as the foliage leaves. Fertile scales are imbricate or Sciadopityaceae valvate, persistent and usually woody at maturity (secondarily fleshy in Juniperus and Shoots are dimorphic (long or short) with Arceuthos). Each scale has 1–12 erect ovules leaves of two types, scale leaves on the stem, on the adaxial surface. Seeds are winged or and photosynthetic leaves at the apex of not, and embryos usually have two but both long and short shoots. Photosynthetic rarely up to nine cotyledons. leaves variously interpreted as a pair of true leaves fused together, or as highly modified A family of 30 genera (many monotypic) shoots (cladodes). Pollen cones are borne in and about 155 species. Higher taxonomy of dense terminal clusters. Seed cones are frag- this group has been unstable, with the wide ile, breaking up soon after seed release. recognition of two families, Taxodiaceae and Each cone has 15–40 thin fertile scales, each Cupressaceae. Recent studies indicate that with five to nine flattened, narrowly winged these do not represent natural groups, and seeds. Embryos have two cotyledons. that the former concept of the Cupressaceae represents one of several lineages of com- The family was formerly included as a mon descent. The remaining lines were genus within Taxodiaceae (now included in aggregated into the Taxodiaceae, creating a Cupressaceae), but recent genetic studies paraphyletic assemblage. have shown that it is clearly not allied with that group (Brunsfeld et al., 1994). Podocarpaceae A single genus with a single species, Taxonomy of the Podocarpaceae has received a Sciadopitys verticillata, endemic in Japan. great deal of attention in recent years, and the formerly large and diverse genera Podocarpus Cupressaceae and Dacrydium have been extensively split into smaller segregate genera (Quinn, 1970; de The Cupressaceae first appear in the fossil Laubenfels, 1978, 1987; Page, 1988; Molloy, record in the Late Triassic (Bock, 1969), although a number of earlier forms placed

Diversity and evolution of gymnosperms 37 1995). Phyllocladus has at times been removed back as the Middle Cretaceous. Agathis, on to its own family (Keng, 1978), but more the other hand, is only known as fossils recently has been shown to be nested within from the Tertiary, and from sites geographi- Podocarpaceae (Conran et al., 2000). cally close to present-day occurrences. Wollemia is today relictual, with an Podocarpaceae is the first of the extant extremely reduced living population. Fossil families to appear in the fossil record, with pollen comparable to Wollemia indicates an Rissikia from the Early Triassic (Miller, appearance in the Middle Cretaceous, 1977). Many of the extant genera are known becoming more widespread and abundant as fossils, but almost exclusively from the in the Late Cretaceous and Early Tertiary, Tertiary (Hill, 1995). Strong evidence exists and persisting in some areas until the for a progressive decline in abundance and Pleistocene (MacPhail et al., 1995). diversity of this family throughout the Tertiary, continuing until as recently as the Male cones are made up of numerous Late Pleistocene (Hill, 1995). spirally arranged fertile scales, each with 4–9 pendulous abaxial microsporangia. Male cones are made up of numerous, Pollen is unwinged. Female cones are made spirally inserted sporophylls, each with two up of numerous imbricate, spirally abaxial microsporangia. Pollen grains are arranged, fertile scales, each with a fully or saccate. Female cones consist of one to many mostly adnate bract and a central, adaxial, fleshy or dry but not woody fertile scales inverted ovule. Seeds may be winged or not. with fully adnate bracts, each with one (or Embryos have two or, less commonly, four two) erect or inverted ovules. Cone scales cotyledons. are persistent or deciduous, and scales and axes are fleshy or dry, not woody at matu- A family of three genera and c. 38 rity. Germination is phanerocotular and species, mostly Malaysian and western Pacific embryos have two cotyledons. in distribution. A family of c. 18 genera and 180 species, Evolution and fossil record largely southern hemisphere in its distribu- tion, extending through much of Africa, True conifers first appeared in the from Japan through Malaysia to Australia Carboniferous and increased their abun- and New Zealand, and through Central and dance, dominance and taxonomic richness South America. during the Permian and into the Mesozoic. Palaeozoic conifers show little resemblance Araucariaceae to modern or Mesozoic conifers (Florin, 1950; Rothwell, 1982), and modern families Much of the present-day distribution of the appeared only in the Mesozoic (Table 3.1). family is relictual, although Araucaria and Fossil taxa from the Palaeozoic such as Agathis have radiated widely in New Lebachia and Walchia are regarded, however, Caledonia. The family today is essentially as true conifers (Miller, 1982). Australasian in distribution, with a minor presence in South-east Asia and the western Conifers were ecosystem dominants Pacific. Two species of Araucaria occur in through most of the Mesozoic and reached South America, illustrating another their pinnacle in species diversity during that Gondwanan link. time, declining with the diversification of the angiosperms in the Cretaceous and more so This family appears in the fossil record in the Tertiary. Many of the Mesozoic in the Early Cretaceous, and is abundant conifers in fact represent now-extinct fami- worldwide in deposits from the Late lies, for example, Cheirolepidaceae. Many Mesozoic (Stockey, 1982). The family extant taxa are relictual, although recent became restricted to the southern hemi- evolutionary radiations are evident in some sphere (Gondwana) from the end of the genera such as Pinus, Cupressus, Callitris and Mesozoic. Araucariaceae is remarkable in Podocarpus. Many conifers are well repre- that extant sections of the genus Araucaria can be recognized in the fossil record as far

38 K. Hill sented as fossils and their distribution can Gnetophyta clearly document the diminution of biogeo- graphic range through the Tertiary. The gnetophytes are highly diverse in gross morphology but are clearly shown to be a Florin (1951, 1954) suggested that the monophyletic group by numerous molecu- ovuliferous cone scale with its subtending lar studies (Chaw et al., 2000; Donoghue and bract of the modern conifer cone was Doyle, 2000; Nickrent et al., 2000; Pryer et homologous to the reproductive short shoot al., 2001). and its subtending leaf in Cordaites. With this interpretation, the entire axis of ovulif- Present-day occurrence erous Cordaianthus can be homologized with the conifer seed cone. Florin argued from The group comprises three families, this that Cordaites may be the ancestor to Gnetaceae, Ephedraceae and Welwitschiaceae, the conifers, although the two groups coex- each with only a single extant genus isted for a long period in the upper (Gnetum, Ephedra and Welwitschia) and 30, 60 Carboniferous and Permian. and 1 extant species, respectively. Cordaites (Cordaitales) was abundant in Today, Gnetum is a tropical moist-forest the Late Palaeozoic, and reconstructions plant of both the Old and New Worlds. In suggest it was one of the largest trees of the contrast, both Ephedra and Welwitschia are Carboniferous. The leaves were long (up to dry climate or desert plants. Ephedra con- 1 m) and strap-like, up to 12–15 cm wide, tains about 30 species distributed on all con- similar in appearance to some cycad leaflets. tinents except Australia and Antarctica, The reproductive organs were cones. while Welwitschia is represented by a single Modified lateral shoots produced needle- species that is restricted to the Namib Desert like bracts subtending a short axis bearing in south-west Africa. overlapping, sterile bracts at the base, and a spiral of sporophylls above. Each Vegetative morphology microsporophyll usually had six terminal microsporangia exerted just beyond the Ephedra species are shrubs or occasionally apex of the cone at maturity. Ovules were clambering vines with jointed whorled or produced in similar structures in place of fascicled branches. Leaves are simple, scale- microsporophylls, with one or sometimes like, opposite and decussate or whorled, more terminal ovules extending beyond the connate at the base forming a sheath, gener- sterile bracts. ally ephemeral and mostly not photosyn- thetic. Leaves are vascularized by a pair of In conifer pollen cones, microsporangia traces that exit from the eustele and enter are borne on a single bract without a trace of the leaf. This contrasts with conifers, where a second subtending structure. This is dif- leaves are vascularized by a single trace. ferent from Cordaites in which the microsporangia were borne on a branch Gnetum species are mostly woody subtended by a bract on a branch system climbers, rarely shrubs or trees. Leaves are that was subtended by a second order of opposite, simple, elliptic, petiolate, without bract. Thus, the homologies between pollen- stipules, pinnately veined with reticulate bearing structures in Cordaites and the secondary venation and entire margins. conifers are not clear. Both Ephedra and Gnetum possess a lepto- caul habit with axillary branching and pyc- Saccate pollen is common among the noxylic secondary xylem similar to the conifers as well as many other Mesozoic seed conifers and Ginkgo. plants. Cordaites, glossopterids, Caytonia, Callistophyton and corystosperms all have this Welwitschia is a bizarre plant unlike any- type of pollen. Cycads, Ginkgo, peltasperms, thing else in the plant kingdom, with a gnetophytes and angiosperms lack saccate pollen. The saccate condition has been inter- preted as primitive or plesiomorphic within the conifers.

Diversity and evolution of gymnosperms 39 short, woody, unbranched stem producing pretations regard the inner layer as the true only two or rarely three strap-shaped leaves integument and the outer layer as a reduced that grow from a basal meristem throughout bracteole similar to that surrounding the the life of the plant. The leaves contain microsporangia. The two layers develop into numerous subparallel veins that may anasto- a fleshy, leathery or corky outer layer and a mose or terminate blindly in the mesophyll woody inner layer as the seed matures, (this character is unique in Welwitschia implying animal dispersal. among the gymnosperms). The woody stem widens with age to become a concave disc up Strobili in Gnetum are arranged on axes to a metre across. The branched reproduc- with a conspicuous node–internode organi- tive shoots arise from near the leaf bases at zation. Microsporangiate strobili have two the outside edge of the disc. fused bracts that form a cupule that sur- rounds fertile shoots. Each fertile shoot is The gnetophytes show a number of vege- composed of two fused bracteoles, which tative features that were thought to ally surround the microsporophyll. In the them closely with the angiosperms. The megasporangiate strobilus, the cupule sub- shoot apex meristem is differentated into tends a whorl of ovules, which are wrapped tunica and corpus. The leaves of Gnetum in two layers of tissue outside of the integu- have reticulate venation that bears a striking ment. These external ‘envelopes’ of tissue resemblance to that of some dicot may be sclerified and fused to the integu- angiosperms, although this has not been ment to form a seed-coat-like structure, shown to be a stricly homologous state. Also which makes the Gnetum seed appear very like the flowering plants, the gnetophytes angiosperm-like. The fleshy coat implies ani- possess vessels in their wood. However, gne- mal dispersal. tophyte vessels have a different develop- mental origin and thus may not be strictly Pollen cones of Welwitschia are red and homologous with those of angiosperms. resemble those of Ephedra, appearing in Similarly, the sieve tube–companion cell groups of two to three terminally on each associations are now thought to be parallel branch. Ovulate cones are also red and arise developments (Thomas and Spicer, 1986). from branched reproductive shoots; each cone consists of a single unitegmic ovule and Reproductive morphology another layer derived from two confluent primordia (sometimes called a ‘perianth’) All three genera are dioecious. Pollen and with two ‘bracts’. Normally, only one seed seeds are borne in complex strobilate struc- develops within each cone; it is dispersed by tures, and cone, pollen and seed morphol- wind with the ‘perianth’ as a wing. ogy varies among genera. All appear wind pollinated, although there are some sugges- Ephedra and Gnetum are also similar to tions that Gnetum may be insect pollinated. angiosperms in displaying a form of double Spermatozoioids are non-motile. Ovules are fertilization. In most angiosperms, two erect and unitegmic or apparently bitegmic. sperm cells are involved in fertilization. One unites with the egg cell to form a diploid In Ephedra, microsporangiate strobili con- zygote; the other fuses with the two polar sist of several pairs of bracts. A shoot, bear- nuclei to form a triploid endosperm. The ing ‘bracteoles’ (a second level of bracts), endosperm serves as nutrition for the devel- arises in the axis of each larger bract. Each oping embryo. In Ephedra, two egg cells are bracteole surrounds a stalked microsporo- produced, one of which is fertilized to phyll bearing two or more pollen sacs become the embryo while the other begins (microsporangia). The megasporangiate but does not complete development into an cone is similarly arranged with pairs of embryo (Friedman, 1992). In Ephedra, this bracts, the top two of which subtend an second embryo does not provide nutrition ovule. The ovule appears to be surrounded to the developing embryo as does the by two integuments, although most inter- endosperm of angiosperms; however, the degenerate embryo is believed to be an intermediary step between the primitive

40 K. Hill seed plant condition (where the embryo is Donoghue and Doyle, 2000), although a nourished by a larger gametophyte) and number of recent molecular studies dispute that of angiosperms (megagametophyte so this relationship (Winter et al., 1999; Pryer et reduced as to not be effective for nourishing al., 2001; Rydin et al., 2002; Schmidt and the embryo). Another way to think of this is Schneider-Poetsch, 2002; Soltis et al., 2002). as a reduction in the polyembryony These studies do not, however, agree on the observed in many non-angiosperm seed position of the Gnetophyta in the evolution of plants. Polyembryonic cycad seeds have the seed plants, and this remains at present been reported, with several embryos actually an unresolved question. The fossil record is capable of germination and growing to meagre, with records of pollen grains that maturity. In some conifers, several embryos resemble Ephedra and Welwitschia from the regularly form from fertilizations by several Triassic and Cretaceous. Welwitschia-like fossil pollen grains, but only one embryo per seed cones are known from the Late Triassic survives. The multi-embryo condition (Cornet, 1996). Fossil sporophyll structures appears to be plesiomorphic for the seed associated with ephedroid pollen are known plants and has been reduced in both the from the Jurassic (van Konijnenburg-van gnetophyte and angiosperm lineages. Cittert, 1992), and may represent gneto- phyte progenitors. Leaf fossils with Gnetum- Evolution and fossil record like venation patterns are known from the Early Cretaceous (Crane and Upchurch, The gnetophytes have long been believed to 1987). This age is compatible with the place- be the closest living relatives of the flowering ment of the Gnetophyta as sister to the plants (Doyle and Donoghue, 1986, 1992; angiosperms rather than in a monophyletic Gymnospermae. References Beck, C.B. (ed.) (1976) Origin and Early Evolution of Angiosperms. Cambridge University Press, Cambridge. Bock, W. (1969) The American Triassic flora and global distribution. Geological Center Research Series, Volumes 3 and 4. Geological Center, North Wales, Pennsylvania. Bowe, L.M., Coat, G. and dePamphilis, C.W. (2000) Phylogeny of seed plants based on all three genomic compartments: extant gymnosperms are monophyletic and Gnetales’ closest relatives are conifers. Proceedings of the National Academy of Sciences USA 97, 4092–4097. Bowman, D.M.J.S. and Harris, S. (1995) Conifers of Australia’s dry forests and open woodlands. In: Enright, N.J. and Hill, R.S. (eds) Ecology of the Southern Conifers. Melbourne University Press, Melbourne, pp. 252–270. Bremer, K., Humphries, C.J., Mishler, B.D. and Churchill, S.P. (1987) On cladistic relationships in green plants. Taxon 36, 339–349. Brunsfeld, S.J., Soltis, P.E., Soltis, D.E., Gadek, P.A., Quinn, C.J., Strenge, D.D. and Ranker, T.A. (1994) Phylogenetic relationships among the genera of Taxodiaceae and Cupressaceae: evidence from rbcL sequences. Systematic Botany 19, 253–262. Burbidge, A.H. and Whelan, R.J. (1982) Seed dispersal in a cycad, Macrozamia riedlei. Australian Journal of Ecology 7, 63–67. Burrows, G.E. (1987) Leaf axil anatomy in the Araucariaceae. Australian Journal of Botany 35, 631–640. Carpenter, R.J. (1991) Macrozamia from the early Tertiary of Tasmania and a study of the cuticles of extant species. Australian Systematic Botany 4, 433–444. Chamberlain, C.J. (1935) Gymnosperms. Structure and Evolution. University of Chicago Press, Chicago, Illinois. Chase, M.W., Soltis, D.E., Olmstead, R.G., Morgan, D., Les, D.H., Mishler, B.D., Duvall, M.R., Price, R.A., Hills, H.G., Qiu, Y.-L., Kron, K.A., Rettig, J.H., Conti, E., Palmer, J.D., Manhart, J.R., Sytsma, K.J., Michaels, H.J., Kress, W.J., Karol, K.G., Clark, W.D., Hendren, M., Gaut, B.S., Jansen, K.R., Kim, K.-J., Wimpee, C.F., Smith, J.F., Furnier, G.R., Strauss, S.H., Xiang, Q.-Y., Plunkett, G.M., Soltis, P.S., Swensen, S.M., Williams, S.E., Gadek, P.A., Quinn, C.J., Eguiarte, L.E., Golenburg, E., Learn, G.H., Graham, S.W., Barrett, S.C.H., Dayanandan, S. and Albert, V.A. (1993) Phylogenetics of seed plants: an

Diversity and evolution of gymnosperms 41 analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80, 528–580. Chaw, S.-M., Zharkikh, A., Sung, H.-M., Lau, T.-C. and Li, W.-H. (1997) Molecular phylogeny of extant gym- nosperms and seed plant evolution: analysis of nuclear 18S rRNA sequences. Molecular Biology and Evolution 4, 56–68. Chaw, S.-M., Parkinson, C.L., Cheng, Y., Vincent, T.M. and Palmer, J.D. (2000) Seed plant phylogeny inferred from all three plant genomes: monophyly of extant gymnosperms and origin of Gnetales from conifers. Proceedings of the National Academy of Sciences USA 97, 4086–4086. Cleal, C. (1989) Angiosperm phylogeny: evolution in hidden forests. Nature 339, 16. Conran, J.G., Wood, G.A., Martin, P.G., Dowd, J.M., Quinn, C.J., Gadek, P.A. and Price, R.A. (2000) Generic relationships within and between the gymnosperm families Podocarpaceae and Phyllocladaceae based on an analysis of the CP gene rbcL. Australian Systematic Botany 48(6), 715–724. Cookson, I.C. (1953) On Macrozamia hopeites: an early Tertiary cycad from eastern Australia. Phytomorphology 3, 306–312. Cornet, B. (1996) A new gnetophyte from the late Carnian (Late Triassic) of Texas and its bearing on the ori- gin of the angiosperm carpel and stamen. In: Taylor, D.W. and Hickey, L.J. (eds) Angiosperm Origin, Evolution and Phylogeny. Chapman & Hall, New York, pp. 32–67. Crane, P.R. (1985a) Phylogenetic relationships in seed plants. Cladistics 1, 329–385. Crane, P.R. (1985b) Phylogenetic analysis of seed plants and the origin of angiosperms. Annals of the Missouri Botanical Garden 72, 716–793. Crane, P.R. (1988) Major clades and relationships in the higher gymnosperms. In: Beck, C.B. (ed.) Origin and Evolution of the Gymnosperms. Columbia University Press, New York, pp. 218–272. Crane, P.R. and Upchurch, G.R. Jr (1987) Drewria potomacensis gen. et sp. nov., an Early Cretaceous mem- ber of Gnetales from the Potomac Group of Virginia. American Journal of Botany 74, 1723–1738. de Laubenfels, D.J. (1978) The genus Prumnopitys (Podocarpaceae) in Malesia. Blumea 24, 189–190. de Laubenfels, D.J. (1987) Revision of the genus Nageia (Podocarpaceae). Blumea 32, 209–211. Dehgan, B. and Yuen, C.K.K.H. (1983) Seed morphology in relation to dispersal, evolution and propagation of Cycas L. Botanical Gazette 144, 412–418. Delevoryas, T. (1993) Origin, evolution and growth patterns in cycads. In: Stevenson, D.W. (ed.) Proceedings of CYCAD 90, the Second International Conference on Cycad Biology. Palm and Cycad Societies of Australia Limited, Brisbane, pp. 236–245. Donaldson, J.S., Nanni, I. and de Wet Bosenberg, J. (1995) Beetle pollination of Encephalartos cycadifolius. In: Vorster, P. (ed.) Proceedings of the Third International Conference on Cycad Biology. The Cycad Society of South Africa, Stellenbosch, pp. 423–434. Donoghue, M.J. (1994) Progress and prospects in reconstructing plant phylogeny. Annals of the Missouri Botanical Garden 81, 405–418. Donoghue, M.J. and Doyle, J.A. (2000) Seed plant phylogeny: demise of the anthophyte hypothesis? Current Biology 10, R106–R109. Doyle, J.A. (1996) Seed plant phylogeny and the relationships of Gnetales. International Journal of Plant Science 157(6), Supplement S3–S39. Doyle, J.A. (1998a) Molecules, morphology, fossils and the relationship of angiosperms and Gnetales. Molecular Phylogenetics and Evolution 9, 448–462. Doyle, J.A. (1998b) Phylogeny of vascular plants. Annual Review of Ecology and Systematics 29, 567–599. Doyle, J.A. and Donoghue, J.M. (1986) Seed plant phylogeny and the origin of the angiosperms: an experi- mental cladistic approach. Botanical Review 52, 331–429. Doyle, J.A. and Donoghue, J.M. (1992) Fossils and plant phylogeny reanalyzed. Brittonia 44, 89–106. Farjon, A. and Page, C.N. (2001) Conifers Status Survey and Conservation Action Plan. IUCN, Gland, Switzerland. Florin, R. (1939) The morphology of the female fructifications in Cordaites and conifers of Palaeozoic age. Botaniska Notiser 36, 547–565. Florin, R. (1950) Upper Carboniferous and Lower Permian conifers. Botanical Review 16, 258–282. Florin, R. (1951) Evolution in Cordaites and Conifers. Acta Horti Bergiani 15, 285–388. Florin, R. (1954) The female reproductive organs of conifers and taxads. Biological Reviews 29, 367–389. Foster, A.S. and Gifford, E.M. Jr (1989) Comparative Morphology of Vascular Plants. W.H. Freeman and Company, San Francisco, California. Friedman, W.E. (1992) Evidence of a pre-angiosperm origin of endosperm: implications for the evolution of flowering plants. Science 255, 336–339.

42 K. Hill Friis, E.M., Chaloner, W.G. and Crane, P.R. (eds) (1987) The Origins of Angiosperms and their Biological Consequences. Cambridge University Press, Cambridge, UK. Frohlich, M.W. and Meyerwitz, E.M. (1997) The search for flower homeotic gene homologs in basal angiosperms and Gnetales: a potential new source of data on the evolutionary origin of flowers. International Journal of Plant Science 158, S131–S142. Fu, L.-G., Nan, L. and Mill, R.R. (1999) Ginkgoaceae. In: Wu, Z.-Y. and Raven, P.H. (eds) Flora of China, Vol. 4. Science Press, Beijing; Missouri Botanical Garden, St Louis, Missouri, p. 8. Gao, Z.-F. and Thomas, B.A. (1989) A review of fossil cycad megasporophylls, with new evidence of Crossozamia Pomel and its associated leaves from the lower Permian of Taiyuan, China. Review of Palaeobotany and Palynology 60, 205–223. Gibson, N., Barker, P.C.J., Cullen, P.J. and Shapcott, A. (1995) Conifers of southern Australia. In: Enright, N.J. and Hill, R.S. (eds) Ecology of the Southern Conifers. Melbourne University Press, Melbourne, pp. 252–270. Goremykin, V.V., Bobrova, V.K., Pahnke, J., Troitsky, A.V., Antonov, A.S. and Martin, W. (1996) Noncoding sequences from the slowly evolving chloroplast inverted repeat in addition to rbcL data do not support gnetalean affinities of Angiosperms. Molecular Biology and Evolution 13, 383–396. Gregory, T.J. and Chemnick, J. (2004) Hypotheses on the relationship between biogeography and speciation in Dioon (Zamiaceae). In: Walters, T. and Osborne, R. (eds) Cycad Classification Concepts and Recommendations. CAB International, Wallingford, UK, pp. 137–148. Guppy, H.B. (1906) Observations of a Naturalist in the Pacific. Macmillan, London. Hardin, J.W., Harding, J.W. and White, F.M. (2001) Textbook of Dendrology. McGraw-Hill, New York. Harris, T.M. (1976) The Mesozoic gymnosperms. Review of Palaeobotany and Palynology 21, 119–134. Hasebe, M., Kofuji, R., Ito, M., Kato, K., Iwatsuki, M. and Ueda, K. (1992) Phylogeny of gymnosperms inferred from rbcL gene sequences. Botanical Magazine (Tokyo) 105, 673–679. Hill, C.R. and Crane, P.R. (1982) Evolutionary cladistics and the origin of angiosperms. In: Joysey, K.A. and Friday, E.A. (eds) Problems of Phylogenetic Reconstruction. Academic Press, London, pp. 269–361. Hill, K.D. (1994) The Cycas rumphii complex (Cycadaceae) in New Guinea and the western Pacific. Australian Systematic Botany 7, 543–567. Hill, K.D., Chase, M.W., Stevenson, D.W., Hills, H.G. and Schutzman, B. (2003) The families and genera of cycads: a molecular phylogenetic analysis of cycadophyta based on nuclear and plastid DNA sequences. International Journal of Plant Science 164(6), 933–948. Hill, R.S. (1978) Two new species of Bowenia Hook. ex Hook. F. from the Eocene of eastern Australia. Australian Journal of Botany 26, 837–846. Hill, R.S. (1980) Three new Eocene cycads from eastern Australia. Australian Journal of Botany 28, 105–122. Hill, R.S. (1995) Conifer origin, evolution and diversification in the Southern Hemisphere. In: Enright, N.J. and Hill, R.S. (eds) Ecology of the Southern Conifers. Melbourne University Press, Melbourne, pp. 10–29. Judd, W.S., Campbell, C.S., Kellogg, E.A., Stevens, P.F. and Donoghue, M.J. (2002) Plant Systematics: a Phylogenetic Approach, 2nd edn. Sinauer Associates, Sunderland, Massachusetts. Keng, H. (1978) The genus Phyllocladus (Phyllocladaceae). Journal of the Arnold Arboretum 59, 249–273. Lewin, R. (1988) A lopsided look at evolution. Science 241, 291. Lindley, J. (1830) An Introduction to the Natural System of Botany. Longman & Co., London. Liu, Y.S., Zhou, Z.Y. and Li, H.M. (1991) First discovery of Cycas fossil leaf in northeast China. Science Bulletin 22, 1758–1759. Loconte, H. and Stevenson, D.W. (1990) Cladistics of the Spermatophyta. Brittonia 42, 197–211. MacPhail, M., Hill, K., Partridge, A., Truswell, E. and Foster, C. (1995) ‘Wollemi Pine’ – old pollen records for a newly discovered genus of gymnosperm. Geology Today 11(2), 48–50. Meyen, S. (1981) Ginkgo: a living pteridosperm. IOP Newsletter 15, 6–11. Miller, C.N. (1977) Mesozoic conifers. Botanical Review 43, 218–280. Miller, C.N. (1982) Current status of Palaeozoic and Mesozoic conifers. Review of Palaeobotany and Palynology 37, 99–114. Miller, C.N. (1988) The origin of modern conifer families. In: Beck, C.B. (ed.) Origin and Evolution of Gymnosperms. Columbia University Press, New York, pp. 448–486. Molloy, B.J.P. (1995) Manoao (Podocarpaceae), a new monotypic conifer genus endemic to New Zealand. New Zealand Journal of Botany 33, 183–201. Nathanielsz, C.P. and Staff, I.A. (1975) On the occurrence of intra cellular blue-green algae in cortical cells of apogeotropic roots of Macrozamia communis. Annals of Botany 39, 363–368.

Diversity and evolution of gymnosperms 43 Nickrent, D.L., Parkinson, C.L., Palmer, J.D. and Duff, R.J. (2000) Multigene phylogeny of land plants with special reference to bryophytes and the earliest land plants. Molecular Biology and Evolution 17, 1885–1895. Nixon, K.C., Crepet, W.L., Stevenson, D.W. and Friis, E.M. (1994) A reevaluation of seed plant phylogeny. Annals of the Missouri Botanical Garden 81, 484–533. Norstog, K.J. and Nichols, T.J. (1997) The Biology of the Cycads. Cornell University Press, Ithaca, New York. Norstog, K.J., Stevenson, D.W. and Niklas, K.J. (1986) The role of beetles in the pollination of Zamia fur- furacea L. fil. (Zamiaceae). Biotropica 18, 300–306. Ohsawa, T. (1994) Anatomy and relationships of petrified seed cones of the Cupressaceae, Taxodiaceae and Sciadopityaceae. Journal of Plant Research 107, 503–512. Page, C.N. (1988) New and maintained genera in the conifer families Podocarpaceae and Pinaceae. Notes from the Royal Botanic Gardens, Edinburgh 45, 377–395. Page, C.N. (1990) The families and genera of vascular plants, Vol. 1: Pteridophytes and gymnosperms. In: Kramer, K.U. and Green, P.S. (series eds) The Families and Genera of Vascular Plants. Springer-Verlag, Berlin. Parenti, L.R. (1980) A phylogenetic analysis of land plants. Journal of the Linnaean Society, Biology 13, 225–242. Pellmyr, O., Tang, W., Groth, I., Bergstrom, G. and Thien, L.B. (1991) Cycad cone and angiosperm floral volatiles: inferences for the evolution of insect pollination. Biochemical Systematics and Ecology 19, 623–627. Pryer, K.M., Schneider, H., Smith, A.R., Cranfill, R., Wolf, P.G., Hunt, J.S. and Sipes, S.D. (2001) Horsetails and ferns are a monophyletic group and the closest living relatives to seed plants. Nature 409, 618–622. Qiu, Y.-L. and Palmer, J.D. (1999) Phylogeny of early land plants: insights from genes and genomes. Trends in Plant Science 4, 26–30. Qiu, Y.-L., Lee, J., Bernasconi-Quadroni, F., Soltis, D.E., Soltis, P.S., Zanis, M., Zimmer, E.A., Chen, Z., Savolainen, V. and Chase, M.W. (1999) The earliest angiosperms: evidence from mitochondrial, plastid and nuclear genomes. Nature 402, 404–407. Quinn, C.J. (1970) Generic boundaries in the Podocarpaceae. Proceedings of the Linnaean Society of New South Wales 94, 166–172. Rai, H.S., O’Brien, H.E., Reeves, P.A., Olmstead, R.G. and Graham, S.W. (2003) Inference of higher-order relationships in the cycads from a large chloroplast data set. Molecular Phylogenetics and Evolution 29(2), 350–359. Raubeson, L.A. (1998) Chloroplast DNA structural similarities between conifers and gnetales: coincidence of common ancestry. American Journal of Botany (suppl.) 85, 149 [abstract]. Raubeson, L.A. and Jansen, R.K. (1992) A rare chloroplast-DNA mutation shared by all conifers. Biochemical Systematics and Ecology 20, 17–24. Raven, P.H., Evert, R.F. and Eichhorn, S.E. (1992) Biology of Plants, 5th edn. Worth Publishers, New York. Rothwell, G.W. (1982) New interpretations of the earliest conifers. Review of Palaeobotany and Palynology 37, 7–28. Rothwell, G.W. and Serbet, R. (1994) Lignophyte phylogeny and the evolution of spermatophytes: a numeri- cal cladistic analysis. Systematic Botany 19, 443–482. Rothwell, G.W., Scheckler, S.E. and Gillespie, W.H. (1989) Elkinsia gen. nov., a late Devonian gymnosperm with cupulate ovules. Botanical Gazette 150, 170–189. Rydin, C., Källersjö, M. and Friis, E.M. (2002) Seed plant relationships and the systematic position of gne- tales based on nuclear and chloroplast DNA: conflicting data, rooting problems, and the monophyly of conifers. International Journal of Plant Science 163, 197–214. Schmidt, M. and Schneider-Poetsch, H.A.W. (2002) The evolution of gymnosperms redrawn by phytochrome genes: the Gnetatae appear at the base of the gymnosperms. Journal of Molecular Evolution 54, 715–724. Soltis, D.E., Soltis, P.A. and Zanis, M.J. (2002) Phylogeny of seed plants based on evidence from eight genes. Journal of Botany 89, 1670–1681. Stefanovic, S., Jager, M., Deutsch, J., Broutin, J. and Masselot, M. (1998) Phylogenetic relationships of conifers inferred from partial 28S rRNA gene sequences. American Journal of Botany 85, 688–697. Stevenson, D.W. (1980) Observations on root and stem contraction in cycads (Cycadales) with special refer- ence to Zamia pumila L. Botanical Journal of the Linnaean Society (London) 81, 275–281. Stevenson, D.W. (1981) Observations on ptyxis, phenology, and trichomes in the Cycadales and their sys- tematic implications. Journal of Botany 68, 1104–1114.

44 K. Hill Stevenson, D.W. (1990) Morphology and systematics of the Cycadales. Memoirs of the New York Botanical Garden 57, 8–55. Stevenson, D.W. (1992) A formal classification of the extant cycads. Brittonia 44, 220–223. Stevenson, D.W., Norstog, K.J. and Fawcet, P.K.S. (1998) Pollination biology of cycads. In: Owens, S.J. and Rudall, P.J. (eds) Reproductive Biology. Royal Botanic Gardens, Kew, UK, pp. 277–294. Stewart, W.N. and Rothwell, G.W. (1993) Paleobotany and the Evolution of Plants. Cambridge University Press, New York. Stockey, R.A. (1982) The Araucariaceae: an evolutionary perspective. Review of Palaeobotany and Palynology 37, 133–154. Tang, W. (1987) Insect pollination in the cycad Zamia pumila (Zamiaceae). American Journal of Botany 74, 90–99. Taylor, T.N. and Taylor, E.L. (1993) The Biology and Evolution of Fossil Plants. Prentice Hall, New Jersey. Thomas, B. and Spicer, R.A. (1986) Evolution & Palaeobiology of Land Plants. Croom Helm, London. Tralau, H. (1968) Evolutionary trends in the genus Ginkgo. Lethaia 1, 63–101. van Konijnenburg-van Cittert, J.H.A. (1992) An enigmatic Liassic microsporophyll, yielding Ephedripites pollen. Review of Palaeobotany and Palynology 71, 239–254. Winter, K.-U., Becker, A., Munster, T., Kim, J.T., Saedler, H. and Thiessen, G. (1999) The MADS-box genes reveal that gnetophytes are more closely related to conifers than to flowering plants. Proceedings of the National Academy of Sciences USA 96, 7342–7347. Yokoyama, M. (1911) Some tertiary fossils from the Mippe coalfields. Journal of the College of Sciences Imperial University of Tokyo 20, 1–16.

4 Chloroplast genomes of plants Linda A. Raubeson1 and Robert K. Jansen2 1Department of Biological Sciences, Central Washington University, Ellensburg, WA 98926-7537, USA; 2Integrative Biology, University of Texas, Austin, TX 78712-0253, USA Introduction to occur only once in evolutionary history. Three major lineages of primary endosym- The chloroplast is descended from a for- bionts are extant: red, green and a small merly free-living bacterium. Thus the ‘primitive’ group, the glaucocystophytes. chloroplast genome is eubacterial and as For information on algal chloroplast such is circular, attached to the inner genomes the reader is referred to Palmer organellar membrane, unassociated with and Delwiche (1998) and Simpson and proteins, and uses (at least in part) gene Stern (2002) as well as the references above. regulatory and replication machinery simi- The focus of this chapter will be the chloro- lar to that characterized in the model plast genome of the land plants, a derived organism, Escherichia coli. Chloroplasts are group within the green lineage. thought to have descended from one pri- mary endosymbiotic event, where the free- Sugiura (2003) has recently published a living bacterium was engulfed and enslaved concise history of work on the chloroplast by a host eukaryotic cell (Douglas, 1998; genome. To summarize here: through the McFadden, 2001; Moreira and Phillipe, 1950s and early 1960s scientists demon- 2001; although see Palmer, 2003, and strated that the chloroplast contained its Delwiche and Palmer, 1997, for a considera- own unique genome. Then attention turned tion of controversies concerning the mono- to characterizing the chloroplast DNA phyly of plastids; and Stiller et al., 2003, for (cpDNA) in various plants using restriction an opposing viewpoint). Over time a com- site mapping, electron microscopy, and plex genetic symbiosis developed that other techniques of the times. The first com- involved the loss of genes from the chloro- pletely sequenced chloroplast genomes plast genome, with any essential genes (Table 4.1) were tobacco (Nicotiana, Fig. 4.1; being transferred to the nucleus. In order Shinozaki et al., 1986) and liverwort for this transfer to result in a protein that is (Marchantia; Ohyama et al., 1986). Since that functional in the organelle, the nuclear copy time, many additional genomes have been first must be expressed with the product completely sequenced (Table 4.1), tech- then targeted back to the chloroplast. niques have advanced, and molecular bio- Apparently the original transition from an logical research has progressed to functional independent bacterium to an organelle genomics. involved such a complex series of events as As molecular techniques have progressed and basic knowledge of the chloroplast © CAB International 2005. Plant Diversity and Evolution: Genotypic and 45 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

Table 4.1. Completely sequenced land plant chloroplast genomes (as of 1 November Higher Subclass Family Species classification (for angiosperms) (common name) (informal) Bryophytes Marchantiaceae Marchantia polymorpha Anthocerotaceae (liverwort) Vascular plants Funariaceae Anthoceros formosae Pteridophytes Psilotaceae (hornwort) Fern ally Physcomitrella patens Adiantaceae (moss) Fern Pinaceae Psilotum nudum Seed plants Pinaceae (whisk fern) Gymnosperms Adiantum capillus-veneris Conifers Calycanthaceae (maiden-hair fern) Amborellaceae Angiosperms Magnoliidae Chenopodiaceae Pinus thunbergii Dicots Magnoliidae Brassicaceae (black pine) Caryophyllidae Onagraceae Pinus koraiensis Fabaceae (Korean pine) Dilleniidae Fabaceae Rosidae Calycanthus fertilis Amborella trichopoda Rosidae Spinacia oleracea (spinach) Rosidae Arabidopsis thaliana Oenothera elata (evening primrose) Lotus corniculatus (lotus) Medicago truncatula (lucerne)

r 2003). Publication 46 L.A. Raubeson and R.K. Jansen Ohyama et al. (1986) Accession number Kugita et al. (2003a) (GenBank) Sugiura et al. (2003) X04465 AB086179 Wakasugi et al., unpublished AP005672 Wolf et al. (2003) AP004638 Wakasugi et al. (1994) s AY178864 Noh et al., unpublished D17510 Goremykin et al. (2003b) NC_004677 Goremykin et al. (2003a) Schmitz-Linneweber et al. (2001) AJ428413 Sato et al. (1999) AJ506156 Hupfer et al. (2000) AJ400848 Kato et al. (2000) AP000423 Lin et al., unpublished AJ271079 AP002983 AC093544

Monocots Asteridae Scrophulariaceae Epifagus virginiana Asteridae Solanaceae (beech drops) Asteridae Solanaceae Nicotiana tabacum Commelinidae Poaceae (tobacco) Commelinidae Poaceae Atropa belladonna Commelinidae Poaceae (belladonna) Triticum aestivum (wheat) Zea mays (maize) Oryza sativa (rice) Updated list of completely sequenced chloroplast genomes (for land plants and algae other/cp_list.html

M81884 Wolfe et al. (1992) Z00044 Shinozaki et al. (1986) AJ316582 Schmitz-Linneweber et al. (2002) AB042240 Ogihara et al. (2002) X86563 Maier et al. (1995) X15901 Hiratsuka et al. (1989) e) can be found at http://megasun.bch.umontreal.ca/ogmp/projects/ Chloroplast genomes of plants 47

48 L.A. Raubeson and R.K. Jansen Fig. 4.1. Gene map of tobacco chloroplast genome (adapted from Wakasugi et al., 1998). The inner circle shows the four major regions of the genome – the two copies of the inverted repeat (IRA and IRB) and the large and small single-copy regions (LSC and SSC). The outer circle represents the tobacco genome with the transcribed regions shown as boxes proportional to gene size. Genes shown on the inside of the circle are transcribed in a clockwise direction, whereas genes on the outside of the circle are transcribed anti- clockwise. The IR extent is also shown by the increased width of the circle representing the tobacco genome. Genes with introns are marked with asterisks (*). Arrows between the gene boxes and gene names show those operons known to occur in tobacco cpDNA. Other operons could be present. Genes coding for products that function in protein synthesis are darker grey; genes coding for products that function in photosynthesis are stippled; genes coding for products with various other functions are lighter grey. genome has expanded, evolutionary biolo- genome has been utilized more than any gists have been able to extend the use of other plant genome as a marker for investi- cpDNA in comparative studies. Such studies gating plant evolution and diversity due to have contributed to the understanding of its many advantages. Because of the mutational processes operating in chloro- genome’s small size (generally 120–160 kilo- plast genomes as well as providing data for base pairs, kbp) and high copy number (as phylogenetic purposes. The chloroplast many as 1000 per cell), it is relatively

Chloroplast genomes of plants 49 straightforward to isolate and characterize For the remainder of this chapter, we will cpDNA. Plus, the conservative nature of the focus on two aspects of the plant chloroplast genome allows for the use of DNA probes genome: (i) its organization and evolution; from even distantly related species and the and (ii) the phylogenetic utility of different design of ‘universal’ primers. In addition, approaches to cpDNA characterization. the genome is a good phylogenetic marker because rates of nucleotide change (while Organization and Evolution of Land overall being slower than in the nuclear Plant Chloroplast Genomes genome) show a range of rates making dif- ferent parts of the genome appropriate for Chloroplast genome organization is highly different levels of comparison, rare changes conserved within land plants. Most land in gene order can be informative even at plant genomes have a quadripartite struc- deep phylogenetic levels, and the usual pat- ture with two copies of a large inverted tern of uniparental inheritance and lack of repeat (IR) separating two single copy recombination simplify analysis. regions (refer to the inner circle of Fig. 4.1). As the two regions of unique genes are of The use of cpDNA in phylogenetic stud- unequal size, these regions are referred to as ies dates back to the early 1980s when a few the large and small single copy regions (LSC plant biologists used the genome to address and SSC, respectively). Land plant cpDNAs species relationships in several groups of usually contain 110–130 different genes. crop plants by comparing fragment patterns The majority of these genes (about 80; see of purified cpDNA digested with restriction Table 4.1) code for proteins, mostly involved enzymes (Palmer and Zamir, 1982; Bowman in photosynthesis or gene expression; the et al., 1983; Clegg et al., 1984; Hosaka et al., remainder are transfer RNA (about 30) or 1984). Later studies compared restriction ribosomal RNA (4) genes. Most chloroplast site changes via filter hybridization at higher genes are part of polycistronic transcription taxonomic levels (e.g. Sytsma and Gottlieb, units (Fig. 4.1; Palmer, 1991; Mullet, 1993); 1986; Jansen and Palmer, 1988). Some stud- that is, they occur in operons where the ies also mapped gene order and used genes within each operon are under the rearrangements to address evolutionary control of the same promoter. Often oper- relationships (e.g. Jansen and Palmer, ons contain multiple promoters that allow 1987a; Raubeson and Jansen, 1992a,b). transcription of a subset of genes within the More recently the vast majority of phyloge- operon (e.g. Miyagi et al., 1998; Kuroda and netic and systematic studies have employed Maliga, 2002). Some polycistronic tran- sequence data for cpDNA-based phyloge- scripts are subject to cis- (within the same netic comparisons. The first sequencing transcript) or trans- (between different tran- studies (Doebley et al., 1990; Soltis et al., scripts) splicing or both. For example 1990) utilized the large subunit of ribulose (Hubschmann et al., 1996), the rps12 gene 1,5-bisphosphate carboxylase/oxygenase exists as three exons. The 5Ј exon occurs as (rbcL). This has been the most widely part of the clpP operon. The remaining two sequenced chloroplast gene and the empha- exons occur (in a quite distant location) sis on this gene culminated in a multi- together in an operon with ndhB. In the authored study involving 499 species of seed construction of the mature rps12 mRNA, the plants (Chase et al., 1993). Many other indi- intron between exons 2 and 3 is removed vidual chloroplast genes and intergenic and the exons joined (cis-splicing) and exon regions have now been utilized (reviewed in 1 (from the separate operon) is joined to Soltis and Soltis, 1998). Several recent stud- exons 2 and 3 (trans-splicing). Both group I ies have used ten or more protein-coding and group II types of self-splicing introns genes from partially or completely are found in land plant cpDNAs; the major- sequenced chloroplast genomes to estimate ity are group II (Palmer, 1991). Although phylogenetic relationships of plants (e.g. intron content is quite variable in algal Graham and Olmstead, 2000; Lemieux et al., 2000; Martin et al., 2002).

50 L.A. Raubeson and R.K. Jansen genomes, it is conserved in land plants. 78 of 81 protein-coding genes found in the Nineteen of the 20 introns found in the tobacco genome also occur in the genome of tobacco genome (Wakasugi et al., 1998) liverwort (this and additional comparisons occur also in the hornwort, Anthoceros, are shown in Table 4.2). genome (Kugita et al., 2003a); the only intron not shared between the two genomes Loss events can occur if genes are not occurs in rps16, a gene that is absent from essential to organismal function or if a func- the hornwort genome (and other non-seed tional copy of the gene can be transferred to plant cpDNAs). the nucleus. The entire gene could be lost simply by deletion in a single mutational Although the gene content and organiza- event but more likely point mutations make tion of the chloroplast genome is evolution- the gene non-functional and then the arily conservative, changes do occur. In the pseudogene gradually decays until it is no remainder of this section we will consider in longer recognizable in the genome. For more detail three classes of gene order instance, in the conifer Pinus thunbergii changes in the land plant chloroplast chloroplast (Wakasugi et al., 1994), the 11 genome: gene and intron loss; inverted ndh genes are lost (four) or present only as repeat changes; and inversions. These types pseudogenes (seven). The ndh genes are of mutations are often referred to as struc- homologues of the mitochondrial NADH tural changes or rearrangements in this lit- dehydrogenase genes. In organisms with erature. functional chloroplast copies, the gene prod- ucts are active within the chloroplast; for Gene and intron loss example, expression increases when plants are under oxidative stress (Casano et al., As mentioned above, the loss of genetic 2001). In a second example of the loss of information from the chloroplast genome multiple related genes, none of the genes has been a general pattern over evolution- involved in photosynthesis is present in the ary time since the original free-living highly reduced chloroplast genome of the prokaryote was first engulfed. Most genes non-photosynthetic plant Epifagus (Wolfe et were lost early in the process of endosym- al., 1992). In a third example, angiosperms biosis (Martin et al., 1998); however, some Welwitschia and Psilotum have all lost (pre- loss events have continued to occur during sumably independently) copies of chlL, chlN land plant diversification. Most losses within and chlB from their chloroplasts (Burke et land plants are restricted to individual lin- al., 1993). These three genes encode the eages; thus, gene content is mostly shared three subunits of a protein that allows among chloroplast genomes. For example, chlorophyll to mature in the absence of light. A separate nuclear-encoded protein Table 4.2. Distribution of protein-coding genes in a bryophyte (Marchantia), a conifer (Pinus), a dicot (Nicotiana) and a monocot (Zea). Calculated from Table 5 of Martin et al. (2002). Bold values on the diagonal indicate the number of genes in each genome. Above the diagonal, the number of genes shared between genomes is shown. Below the diagonal (and following the slash on the diagonal) are the number of unique genes found in a single genome or found in only one pair of genomes. Marchantia Pinus Nicotiana Zea Marchantia 86/4 72 78 75 Pinus 5 73/0 67 64 Nicotiana 00 81/0 78 Zea 01 3 79/0

Chloroplast genomes of plants 51 matures chlorophyll but only in the pres- porated into the IR. Within land plants, the ence of light. Because angiosperms are length of the IR has ‘grown’ from 10 kbp in unable to green (i.e. mature chlorophyll) in liverwort to 25 kbp in tobacco (Palmer and the dark we can infer a simple loss of the chl Stein, 1986). Most, but not all, examples of genes from the chloroplast without a func- gene duplication within the chloroplast tional transfer to the nucleus. In a final genome occur through IR expansion. example, an exceptional case involving a single gene, infA (coding for a translation The underlying mechanism of IR expan- initiation factor), has been lost multiple sion is not well understood. Gene conver- times (about 24) within angiosperms with an sion is thought to be involved since the estimated four independent functional existing DNA sequence is used as the tem- transfers of the gene to the nucleus (Millen plate for forming the new copy. Evidence et al., 2001). supports gene conversion or copy correction acting on the chloroplast genome. Rates of Introns are also lost from within genes in nucleotide substitution are reduced in the the chloroplast. Recombination of the IR relative to single-copy regions (Palmer, processed mRNA with the genomic DNA is 1991) and in the same genes when in the IR probably the mechanism responsible versus when single copy (Perry and Wolfe, (Palmer, 1991). As long as the intron is pre- 2002). Where examined the two copies of cisely removed this mutation would be selec- the repeat are identical (Palmer, 1991). tively neutral, unless regulatory or other Presumably, these patterns (of rate and functional elements are contained within the identity) occur because large amounts of intron sequence. Some instances of intron homologous recombination (and copy cor- loss have occurred repeatedly. For example, rection) take place between the two copies of the rpoC1 intron has been lost indepen- the repeat. So much recombination occurs dently in grasses (Katayama and Ogihara, in fact that two different versions of the 1996) and in a subfamily of Cactaceae genome are present (with opposite orienta- (Wallace and Cota, 1996), as well as a mini- tions of the small and large single copy mum of four additional times within dicots regions relative to one another) in equamo- (Downie et al., 1996). These gene and intron lar quantities (Palmer, 1983; Stein et al., loss events, which can occur multiple times 1986). Minor changes, of about 100 base independently, might be locally useful char- pairs (bp) or less, in the endpoints of the IR acters for phylogenetic inference; however, are probably relatively common and gene they must be treated cautiously in broader conversion alone is an adequate explanation comparisons. (Goulding et al., 1996). However, this mech- anism does not account for the fact that, in Inverted repeat some cases, no existing material is lost. Thus major changes (those incorporating one or As mentioned above, the vast majority of multiple genes into the IR) must be land plant chloroplast genomes contain a explained by additional mechanisms such as large duplicated region, the inverted repeat, double reciprocal recombination (Palmer et where the two copies are reverse comple- al., 1985; Yamada, 1991) or double-stranded ments of each other. The genes that form break repair (Goulding et al., 1996) com- the core of the repeat encode the ribosomal bined with gene conversion. RNAs (23S, 16S, 5S and 4.5S). This rDNA- containing IR appears to be an ancestral An increase in the length of the IR is genomic feature as it is found in charo- much more common than its decrease, phytes, basal green algae and some red although decrease is easier to explain mech- algae (Turmel et al., 1999). Gene content anistically by deletion. Only two accounts of other than the rDNA varies. Genes, formerly significant decrease in IR gene content are single copy at the boundaries of the single published: in the Apiaceae where a series of copy regions, can be duplicated and incor- sequential deletion forms has been charac- terized (Plunkett and Downie, 2000) and in Cuscuta, where a probable 7 kbp contraction

52 L.A. Raubeson and R.K. Jansen has occurred (Bömmer et al., 1993). Increase, whereas an almost-complete loss is known in contrast, has been exemplified on many from conifers (Tsudzuki et al., 1992). Thus, occasions: the growth into the LSC within on a minimum of five independent occasions land plants mentioned previously (Palmer the IR has been lost from land plant chloro- and Stein, 1986; Raubeson, 1991); a phyloge- plast genomes. At times the legume loss and netically informative addition of about 11.5 the conifer loss have been equated and thus kbp of the LSC within Berberidaceae (Kim seen as an example of homoplasy. However, and Jansen, 1998); incorporation of SSC the two cases differ in the extent of the loss, genes into the IR shared among families in the gene content of the IR prior to loss, within the Campanulales (Knox and Palmer, and in the copy of the IR that is lost in the 1999); and a spectacular expansion of the IR event (Fig. 4.2). The loss of the IR defines six to 75 kbp in Pelargonium (Geraniaceae; tribes of legumes (Fabaceae; Palmer et al., Palmer et al., 1987b) among others. Where 1987b; Lavin et al., 1990), whereas the conifer the IR has grown extensively beyond the loss event supports conifer monophyly boundaries seen in tobacco, such as in (Raubeson and Jansen, 1992a). In both of Pelargonium (Price et al., 1990) and also in the these instances, other gene order changes Campanulaceae (Cosner et al., 1997), the usually co-occur with the loss of the IR. growth is associated with multiple additional changes in genome organization. Inversions It has been suggested that the presence So far we have discussed deletion of genetic of the IR promotes stability (i.e. reduces information in the context of gene loss, gene order changes) in the remainder of the intron loss and IR loss or contraction as well molecule (Palmer et al., 1987a). Reasons why as the addition of information in the context the IR may facilitate gene order conserva- of IR expansion. To conclude this section on tion include: (i) that enzymes mediating genome organization we will discuss changes recombination are active at the IR, leaving in gene order and orientation within the few copies of these enzymes available to genome. The most common mechanism modify other parts of the molecule; or (ii) leading to gene order change in the chloro- that the interactions between the two IR plast genome is inversion, where a section of copies physically hold the SSC and LSC in a the genome is reversed in order and orien- more open orientation, diminishing the like- tation relative to the remainder of the lihood that portions interact and recombine. genome (Palmer, 1991). Inversions can (For a more detailed discussion and addi- occur via homologous recombination tional possible reasons, see Palmer, 1991.) between small inverted repeats or through The correlation is not perfect, but most double-stranded break repair (Palmer, genomes without the IR or with a greatly 1991). In discussing the nature and utility of enlarged IR have unusually high numbers inversion characters, it is important to dis- of changes in gene order. Perhaps the same tinguish scale. Inversions commonly occur mechanism that promotes gene order within non-coding regions of cpDNA changes also promotes changes in extent of sequence where small repeats associated the IR, or perhaps there is some stability with hairpin or stem–loop structures pro- provided to the molecule by the ‘normal’ IR. vide foci for inversions (Kelchner, 2000). These small-scale (c. 2–200 bp) inversions In a few lineages the IR has been lost; one may be very prone to homoplasy and com- copy has been eliminated leaving each gene plicate interpretation of non-coding only in the retained copy. A complete loss of sequence in phylogenetic studies (Kelchner, the IR is known or suspected from the 2000). However, large-scale changes where chloroplast genomes of some members of the inversions reverse the order and orientation legume family, two members of the of multiple genes have different characteris- Geraniaceae (Price et al., 1990), Conophilis (a tics that will be discussed below. non-photosynthetic plant in the Orobanchaceae; Downie and Palmer, 1992) and Striga (Scrophulariaceae; Palmer, 1991),

Chloroplast genomes of plants 53 Fig. 4.2. Inverted repeat (IR) loss in conifer and legume chloroplast genomes. Genes in the IR prior to the loss events are shown stippled in the ‘before’ circles. Genes in the remaining copy are still stippled in the ‘after’ circles and the site of the lost copy is shown as a stippled triangle. Only selected genes are shown and distance between the genes is not to scale. Extent of the IR is shown as a bar along the genome circle. Note that the extent of the IR differs between ‘conifer before’ and ‘legume before’; trnH, trnI and 3Ј psbA are duplicated in the conifer but not the legume, whereas in the legume a portion of the S10 operon is in the IR but is single copy in the conifer. A comparison of the two ‘after’ circles will reveal a difference in the extent of the loss; the conifer loss is partial (trnI and 3Ј psbA remain as a small remnant IR), whereas in the legume the loss is complete. The majority of chloroplast genomes that dicot) cpDNAs have an identical gene order. have been characterized and compared lack However, in some lineages, changes do any changes in gene order (Palmer, 1991; occur (last reviewed in Downie and Palmer, Downie and Palmer, 1992). It appears that, 1992). Most inversions occur with the end- in most lineages, gene order remains points in non-coding regions so that no unchanged over vast periods of evolutionary genes are disrupted, and only on rare occa- time. For example, Amborella (in several phy- sions are operons split (Palmer, 1991). logenetic studies, the most basal extant Additionally, both endpoints of inversions angiosperm) and tobacco (a derived asterid usually occur within the LSC, perhaps sim-

54 L.A. Raubeson and R.K. Jansen ply because this region is largest and there- incongruent with the most parsimonious fore has the most regions that could serve as trees based on the restriction site characters. endpoints that would not disrupt genes or Hoot and Palmer suggested that the conflict- operons. In a few instances, inversions have ing inversions occurred in parallel. It is pos- occurred that have one endpoint in the LSC sible for identical inversions to occur and the other in the IR. Most such inver- independently if they form due to homolo- sions are probably lethal because they result gous recombination across the same repeat in direct repeats that would promote dele- structure. However, repetitive sequences are tion of genes within the inversion. The few uncommon in land plant chloroplast junction-spanning inversions that occur (e.g. genomes (Palmer, 1991). Also, it is unclear in the leptosporangiate ferns (Stein et al., whether, in general, inversions occur because 1992; Raubeson and Stein, 1995), buck- of repeats or whether repeats occur because wheat (Ali et al., 1997) and adzuki bean of inversions. For example, during double- (Perry et al., 2002)) are associated with stranded break repair it is common for small expansion of the IR. Incorporation of the duplicated segments of DNA (filler DNA) to inversion into the IR would eliminate the be inserted at the site of the break potentially disruptive direct repeats. (Gorbunova and Levy, 1999). This may explain why repeats, including the duplica- Even a small number of inversions, tion of transfer RNA genes or portions of depending on their distribution, can serve larger genes, are associated with the end- as powerful phylogenetic markers. For points of inversions (or other rearrange- example, a single inversion identified the ments). Even where repeats occur in basal members of the Asteraceae (as will be genomes and many inversions have discussed in more detail later in this chap- occurred, independent occurrence of identi- ter). In a second case, a 30-kbp inversion cal inversions is unusual. In a study of the (first recognized as a difference between liv- Campanulaceae, over 40 inversions occur in erwort and tobacco) was found to occur in a data set of 18 taxa with very little homo- all vascular plant cpDNAs except those of plasy evident in the data (Cosner, 1993). lycopsids (Fig. 4.3), marking lycopsids as basal lineage of vascular plants (Raubeson Thus, as rare and complex genomic and Jansen, 1992b). Although somewhat changes, inversions are especially useful controversial at the time, sequence-based phylogenetic markers (Rokas and Holland, studies since (e.g. Nickrent et al., 2000; 2000). Of course no characters are perfect Pryer et al., 2001) have supported this basal and these, as any other, should be carefully position of the lycopsids. Additionally, two investigated and interpreted in the light of inversions and an expansion of the IR clar- all other evidence. In the remainder of the ify basal nodes in leptosporangiate ferns chapter we will more explicity compare the (Stein et al., 1992; Raubeson and Stein, utility of three types of cpDNA data in phy- 1995) and informative inversions have been logenetic studies: restriction site polymor- characterized in legumes (Lavin et al., 1990). phisms, rearrangement characters and Of three inversions shared throughout the nucleotide sequence data. Poaceae, one is restricted to the family, one is shared with Joinvilleaceae and one is Phylogenetic Utility of cpDNA Data shared with Joinvilleaceae and Restoniaceae, thus clarifying the sister groups of the Chloroplast genomes have been characterized grasses (Doyle et al., 1992). for studies of plant diversity using three dif- ferent approaches: restriction fragment/site Few cases have been published comparing comparisons, structural rearrangements, and genomes of taxa among which numerous sequencing of genes or non-coding regions. inversions co-occur. In one of the earliest The utility of these various approaches has such studies, Hoot and Palmer (1994) gener- been reviewed in detail in several papers ated restriction site and mapping data for (Palmer et al., 1988; Downie and Palmer, members of the Ranunculaceae. The distrib- ution of two of the inversion characters was

Chloroplast genomes of plants 55 Fig. 4.3. Inversion distribution in land plants. Bryophytes (e.g. liverwort) and lycopsids have the ancestral gene order for land plants, whereas horsetails and the fern Osmunda differ only in the orientation of the 30-kbp region, shown stippled. Angiosperm (represented by tobacco, Amborella) cpDNAs have the 30-kbp inversion plus the further modification of additional genes incorporated into the IR (shown in lighter grey). Only selected genes are shown as landmarks. Distance between the genes is not to scale. Extent of the IR is shown as a bar along the genome circle. 1992; Doyle, 1993; Olmstead and Palmer, Restriction fragment/site comparisons 1994; Jansen et al., 1998; Soltis and Soltis, 1998; Graham et al., 2000; Rokas and Until about 1998, restriction enzyme Holland, 2000). Below we briefly review each approaches were the most widely used tech- of the three primary methods for using niques for estimating phylogenetic relation- cpDNA and provide examples of their use. ships among plants (see Jansen et al., 1998). We end by discussing the relative utility of The earliest applications of this approach these three approaches for reconstructing the used highly purified cpDNA (e.g. Palmer and phylogenetic history of plants. Zamir, 1982). By the late 1980s, researchers

56 L.A. Raubeson and R.K. Jansen utilized total genomic DNA from which the To illustrate the power of this approach, cpDNA could be visualized by Southern we will describe an early study of hybridization (reviewed in Palmer et al., Heterogaura (Sytsma and Gottlieb, 1986) 1988). The latter approach overcame the dif- because of its historical importance and the ficulties of isolating sufficient quantities of surprising results that it uncovered. pure cpDNA, allowed for a better assessment Heterogaura is a monotypic genus in the of the homology of restriction fragments and evening primrose family (Onagraceae) that has been used in numerous studies from a has been considered distinct since 1866. It wide diversity of plants at a wide range of tax- is closely related to Clarkia, a genus that onomic levels (reviewed in Jansen et al., has served as a model system for studies of 1998). At lower taxonomic levels where speciation in plants, but differs from cpDNA variation is generally quite low (< Clarkia in several features, especially floral 1%) it has been possible to estimate fragment morphology. Sytsma and Gottlieb (1986) homology with a great deal of confidence by mapped sites for 29 restriction enzymes for simple inspection of fragment patterns. At eight species of Heterogaura and Clarkia higher levels of sequence divergence, the using the Southern hybridization more labour-intensive mapping of restriction approach. They surveyed 605 restriction sites was essential to accurately assess charac- sites and found 119 variable sites, 55 of ter homology. Two advantages (Givnish and which were shared by two or more taxa Sytsma, 1997; Jansen et al., 1998) of the (i.e. were parsimony informative). restriction site mapping approach over DNA Phylogenetic analyses of these data gener- sequencing of individual genes are that: (i) by ated a single most parsimonious tree with a using a large number of enzymes that recog- consistency index of 0.95 (Fig. 4.4). nize sequences scattered throughout the Surprisingly, the genus Heterogaura was entire chloroplast genome it is possible to nested within Clarkia, sister to Clarkia dud- gather a very large number of phylogeneti- leyana. These results clearly indicated that cally informative characters; and (ii) compar- the morphologically distinct Heterogaura isons of cpDNA sequences of individual genes should be merged with Clarkia and that and whole chloroplast genome restriction site previous morphological comparisons were studies suggest that the restriction site data misleading with regard to the relationships exhibit less homoplasy than DNA sequences. in this group. 15 C. epilobioides 3 C. rostrata C. lewisii 13 C. cylindrica 4 C. dudleyana 5 Heterogaura 2 C. modesta 14 C. lingulata C. biloba Fig. 4.4. Phylogenetic tree of Clarkia section Peripetasma based on cpDNA restriction site data. Numbers above nodes indicate number of restriction site changes. Adapted from Systma and Gottieb (1986).

Chloroplast genomes of plants 57 The results of the Heterogaura study are survey for the distribution of structural particularly noteworthy for several reasons. rearrangements that were initially identified First, the Onagraceae was viewed as one of by gene mapping. This included the the best-studied angiosperm families, so it hybridization of cloned cpDNA fragments was shocking to learn that a genus that was that spanned the rearrangement endpoints recognized as distinct for 120 years was (e.g. Jansen and Palmer, 1987a) or poly- nested within the well-studied genus Clarkia. merase chain reaction (PCR) using primers Second, Clarkia was considered a model sys- that closely flank rearrangement endpoints tem for studying speciation in plants so it (e.g. Doyle et al., 1996). Hundreds of chloro- was surprising that such a novel result could plast genomes were mapped using the have gone undetected by previous workers. Southern hybridization approach (see Table Third, this was one of the earliest studies 1 in Downie and Palmer, 1992) and, in a that used this approach in plant systematics, number of the groups investigated, struc- and, in combination with several other early tural changes of various types were detected. studies (e.g. Sytsma and Schaal, 1985; In most cases, only one or a few structural Coates and Cullis, 1987; Jansen and Palmer, rearrangements occurred (e.g. in the 1988), it set the stage for a rapid surge in the angiosperm families Asteraceae, Fabaceae, use of cpDNA restriction site comparisons Poaceae; see Table 2 in Downie and Palmer, for examining plant diversity. 1992). However, there were several plant groups in which the chloroplast genomes Structural rearrangements were highly rearranged (i.e. conifers and the angiosperm families Campanulaceae, The second approach for using chloroplast Geraniaceae and Lobeliaceae). genomes for reconstructing phylogenies of plants involves major structural rearrange- One of the most notable examples, ments. As stated earlier, the overall structure demonstrating the powerful utility of of the chloroplast genome is highly con- cpDNA rearrangements for phylogenetic served among land plants and major struc- studies in plants, comes from the tural changes, including inversions, deletions angiosperm family Asteraceae. We describe of genes and introns, expansion/contraction this example for three reasons: (i) it was the of the inverted repeat, and loss of the first, extensive study to demonstrate the inverted repeat, are relatively uncommon power of cpDNA rearrangements for assess- events. In most cases, cpDNA structural ing relationships among deep nodes; (ii) it rearrangements have little or no homoplasy resolved a long-standing controversy making them excellent characters for phylo- regarding the identification of the basal lin- genetic analysis (Palmer et al., 1988). Some eage of this large, extensively studied family; types of changes, such as gene and intron and (iii) the surprising result obtained from losses and expansion and contraction of the the cpDNA rearrangement generated con- inverted repeat, have occurred multiple siderable controversy among angiosperm times (as discussed above) but others, espe- systematists but was later confirmed by mul- cially inversions, have virtually no homoplasy tiple lines of evidence. (Soltis and Soltis, 1998). Here we will focus on inversions, which, because of their rare The Asteraceae (composite or daisy fam- occurrence and low levels of homoplasy, ily) is one of the largest flowering-plant fami- make especially robust phylogenetic indica- lies, with approximately 1535 genera and tors, especially for deep nodes in the phy- 25,000 species (Bremer, 1994). Although the logeny of plants. Early approaches for family has been the focus of numerous stud- examining cpDNA structure involved the ies, considerable controversy existed about very labour-intensive method of constructing the identity of the basal lineage. Five of the restriction site and gene maps. In some stud- 16 recognized tribes of Asteraceae had been ies, other faster methods were developed to suggested as being ancestral based on mor- phological, biogeographical and chemical evidence. Comparative restriction site and gene mapping studies by Jansen and Palmer

58 L.A. Raubeson and R.K. Jansen (1987b) identified a 22-kbp inversion (Fig. were basal in the Asteraceae and that the 4.5a) in the large single-copy region of the cpDNA inversion marks an ancient evolu- chloroplast genome of lettuce (Lactuca sativa, tionary split in the family. The relationships tribe Lactuceae), although the inversion was inferred from the inversion distribution were absent from the chloroplast genome of later confirmed by morphological (Bremer, Barnadesia caryophylla (tribe Mutisieae). Eighty 1987) and DNA sequence data (Kim et al., species representing all tribes of Asteraceae 1992; Kim and Jansen, 1995; Jansen and and ten related families were examined for Kim, 1996). The implications of this finding the distribution of this inversion using were very significant in altering the classifica- cloned cpDNA fragments that spanned the tion of the Asteraceae (Barnadesiinae were inversion endpoints (Jansen and Palmer, elevated to subfamilial status) and improving 1987a). The results showed that all related our understanding of the biogeography and families and members of the subtribe character evolution in the family. Barnadesiinae of the tribe Mutisieae lack the inversion, whereas all other members of the More recently, complete chloroplast Asteraceae have this structural change (Fig. genome sequences have been generated for 4.5b). This suggested that the Barnadesiinae 31 taxa, including 19 land plants (Table 4.1). These data are facilitating the exploration of (a) 2136SS 23S rpl23 rps16 psaBpsaA rbcL petD petD Lactuca 16S23S rpl23 rps16 rpl16 rpl16 Barnadesia rpl23 psbA rpoB rpl23 atpA 2136SS psaBpsaA rbcL 16S psbA atpA rpoB (b) Other } Inversion seed plants absent All other angiosperms Barnadesioideae All other } Inversion Asteraceae present Fig. 4.5. (a) Comparison of chloroplast genome organization between two members of the Asteraceae (Barnadesia and Lactuca) that differ by a 22-kbp inversion. Grey bars indicate the extent of the inverted repeats. Arrows show the region in which inversion occurs. (b) Phylogenetic tree showing the taxonomic distribution of the 22-kbp inversion. Adapted from Jansen and Palmer (1987b).

Chloroplast genomes of plants 59 structural rearrangements in phylogeny they equal rates of mitochondrial single- reconstruction at the deepest nodes of the copy genes (Gaut, 1998). Whereas some plant evolutionary tree. Ongoing genomic studies have found the rates of nucleotide sequencing projects by our group and others substitution in chloroplast-coding regions will greatly expand this sampling during the more conservative than those in non-coding next several years, with a special emphasis regions (see references in Kelchner, 2000), on sequencing genomes from all of the other studies have suggested that some non- major lineages of green plants. The accumu- coding regions may evolve no faster than lation of more chloroplast genome sequences coding regions in the chloroplast genome from a wider diversity of taxa will make it (e.g. Manen and Natali, 1995). Lineage necessary to develop better computational effects in nucleotide substitution rates have methods to deal with gene order characters been detected; for example, grasses have an for phylogeny reconstruction. Considerable elevated rate relative to tobacco or pine work has already been done in this area (Muse and Gaut, 1997). Locus-dependent (Cosner et al., 2000; Moret et al., 2001; Wang differences in rates of non-synonymous sub- et al., 2002; Bourque and Pevzner, 2002), but stitution also occur in some genes (Gaut et additional research is needed in order to al., 1997; Muse and Gaut, 1997; Matsuoka et analyse more highly rearranged genomes al., 2002). RNA editing has been detected in than are sequenced currently. land plant cpDNAs (Miyamoto et al., 2002; Sabater et al., 2002; Kugita et al., 2003b) and DNA sequencing rate of DNA change is usually accelerated in genes with editing of transcripts (Shields The third approach for comparing chloro- and Wolfe, 1997; Bock, 2000). plast genomes in studies of plant phyloge- nies is DNA sequencing. Early When comparing DNA sequence in a sequence-based phylogenetic studies were phylogenetic study, too little variation results hampered by several factors, including the in trees that are highly unresolved and there need to clone genes being sequenced, the are numerous examples of this in the litera- lack of universal primers and the labour- ture. Too much variation can result in exces- intensive nature and high expense of man- sively high levels of homoplasy leading to ual DNA sequencing. Thus, many early suspect relationships, a problem encoun- studies suffered from limited taxon sam- tered (as just one example) in a phyloge- pling and the use of inappropriate genes netic analysis of rbcL among all land plants (i.e. ones with inadequate levels of varia- and their green algal relatives (Manhart, tion). The advent of PCR technology and 1994). Too much variation can also lead to automated DNA sequencing has made it difficulties in alignment, caused by high possible to sequence many more taxa and to rates of nucleotide substitution or the pres- explore the utility of additional chloroplast ence of many deletions and insertions genes. Once these new methods increased (indels). In general, coding regions tend to the capacity for comparative chloroplast be more easily aligned because indels must DNA sequencing projects, genes could be be in multiples of three to maintain func- selected based on criteria of appropriateness tionality of the genes. The increased use of rather than simple logistics. The gene many more variable intergenic regions and sequenced should exhibit the appropriate introns has caused considerable difficulty in amount of variation for the taxonomic level alignment of these chloroplast sequences and group being studied. (see Kelchner, 2000, for a review). In general, nucleotide substitution rates There are numerous examples of the in cpDNA are slower than those of the application of cpDNA sequences for estimat- nuclear genome and faster than those of the ing phylogenetic studies of plants. One mitochondrial DNA (Wolfe et al., 1987). notable example is the analysis of 499 rbcL Substitution rates are lower in the IR where sequences from seed plants (Chase et al., 1993). At the time of its publication this study represented the largest data set of DNA

60 L.A. Raubeson and R.K. Jansen sequences for any group of organisms. Forty- The 499-taxon rbcL data set generated a two scientists contributed to the analysis, well-resolved tree with some very important making the study one of the most amazing implications for resolving the major clades examples of cooperation among the systemat- of seed plants (Fig. 4.6; readers should refer ics community. The resulting phylogenies to Chase et al. (1993) for the details of rela- defined many major clades of flowering tionships within these major clades of seed plants, which formed a set of hypotheses of plants). Some notable results included the relationships that could be tested with other placement of the Gnetales as the sister data sets (Fig. 4.6). And finally, the computa- group of the flowering plants, the position tional phylogenetics community has utilized of the aquatic genus Ceratophyllum in a basal this data set extensively for evaluating many position in angiosperms, the division of issues, including the effects of taxon sampling angiosperms into two major groups corre- on the accuracy of phylogeny reconstruction sponding to those taxa with uniaperturate (Hillis, 1998) and the development of faster and triaperturate pollen, the occurrence of parsimony methods to handle such large data the Magnoliidae as a polyphyletic group sets (Rice et al., 1997; Nixon, 1999). at the base of the angiosperms, and the Asterid I Asterid II Asterid III Asterid IV Asterid V Rosid I Rosid II Rosid III Rosid IV Hamamelid II Hamamelid I Ranunculids Paleoherbs Monocots Laurales Magnoliales Ceratophyllum Gnetales Pinaceae Other conifers Cycads Fig. 4.6. Phylogenetic tree of angiosperms based on rbcL sequences. This tree illustrates the major clades that were part of the analysis including 499 species. Adapted from Chase et al. (1993).

Chloroplast genomes of plants 61 presence of several large clades that corre- Comparative utility of these approaches spond with the major recognized subclasses of angiosperms. Although more recent mol- All three of the approaches for using chloro- ecular phylogenies based on other genes plast DNA data for studies of plant diversifi- and/or different phylogenetic methods have cation have been shown to be very valuable suggested alternative relationships for some for resolving phylogenetic relationships. of these groups, many of the major clades, The first approach, restriction site/fragment especially within angiosperms, have been comparisons, is not widely used anymore, confirmed (Soltis et al., 2000). primarily because it is now much easier to generate DNA sequence data than it was in The number of chloroplast genes available the past. The primary advantage of this for sequence-based studies has grown rapidly approach was the capacity to examine many with the widespread use of PCR and auto- restriction sites (scattered sequences of 4–8 mated sequencing (Table 4.3). The most widely bp) from throughout the genome. However, sequenced coding regions include rbcL, atpB, it is now possible to easily sequence numer- matK and ndhF. Large data sets have been ous genes (or regions) or even to undertake examined for all angiosperms, or for large relatively automated sequencing of entire clades within angiosperms, for each of the chloroplast genomes. The other two types of genes or combinations of these genes. These characters, structural rearrangements and analyses have provided many new insights into DNA sequences will continue to be widely phylogenetic relationships in a wide diversity of used in the future. Structural characters are plants from the earliest land plants to the most extremely valuable for assessing relation- derived clades of angiosperms. There are too ships at the deepest nodes as they exhibit many results from phylogenetic studies of less homoplasy than DNA sequence data. plants using these genes to summarize here so The rapid increase in the availability of com- we refer the reader to some of the recent plete genome sequences will provide more papers using these four genes at various taxo- of these types of characters in the future. nomic levels (Olmstead and Palmer, 1994; The only limitation of structural changes is Soltis and Soltis, 1998; Graham and Olmstead, that there are many fewer characters avail- 2000). Two of these genes, atpB and rbcL, are able; however, this should not discourage highly conserved in nucleotide sequence and plant systematists from utilizing these char- have been most useful for assessing relation- acters where they are present. Chloroplast ships among the major lineages of plants. The DNA sequence data certainly will continue other two, matK and ndhF, provide two to four to be widely used for reconstructing the times more phylogenetically informative char- phylogeny of plants. The systematics com- acters and, therefore, have been used to exam- munity is moving away from relying on one ine relationships among more recently or a few chloroplast genes or regions as evi- diverged taxa in the families Acanthaceae, denced by the increase in the number of Asteraceae, Brassicaceae, Orchidaceae, papers using multiple sets of genes or whole Poaceae, Polemoniaceae, Saxifragaceae, genomes (Turmel et al., 1999; Graham and Scrophulariaceae and Solanaceae (reviewed in Olmstead, 2000; Lemieux et al., 2000; Soltis and Soltis, 1998). At lower taxonomic Martin et al., 2002; Maul et al., 2002; Rai et levels many systematists have utilized al., 2003). sequences from introns and intergenic regions to reconstruct phylogenies (Table 4.3). Summary Although this approach has been successful in many instances, there are a number of prob- We have reviewed aspects of chloroplast lems associated with using these markers. A genome diversity and evolution in land number of molecular mechanisms (indels, sec- plants, especially with regard to their phy- ondary structure, slipped-strand mispairing logenetic utility. In general, the molecule is and localized intramolecular recombination) evolutionarily conservative in both struc- can make it difficult to align these sequences (see Kelchner, 2000, for a detailed review).

62 L.A. Raubeson and R.K. Jansen Table 4.3. Chloroplast genes and regions used for phylogenetic studies of plants arranged by the number of sequences in GenBank (>100) as of 28 July 2003. LSC = large single copy region; SSC = small single copy region; IR = inverted repeat (gene position, if variable, given for tobacco); NA = not applicable. Gene and intergenic region length is in base pairs using the tobacco genome as a reference. Gene or region Location Length Number in Protein product GenBank rbcL LSC 1,434 17,656 Rubisco – large subunit trnL-trnF spacer LSC 358 10,938 NA trnL intron LSC 503 NA matK LSC 9,951 Maturase within trnK intron atpB LSC 1,530 8,572 ATP synthase beta subunit ndhF SSC 1,497 5,475 NADH dehydrogenase ND5 rps16 intron LSC 2,133 4,422 NA atpB-rbcL spacer LSC 2,324 NA psbA LSC 860 2,302 Photosystem II 32 kDa protein rpoB LSC 711 1,515 RNA polymerase beta subunit trnT-trnL spacer LSC 1,062 1,494 NA atpA LSC 3,213 1,441 ATP synthase alpha subunit rpl16 intron LSC 711 1,290 NA rpoC1 LSC 1,524 1,269 RNA polymerase betaЈ subunit psbA-trnH spacer LSC 1,020 NA psaA LSC 2,046 870 Photosystem I P700 apoprotein A1 rpoA LSC 454 768 RNA polymerase alpha subunit psaB LSC 2,253 520 Photosystem I P700 apoprotein A2 rps7 IR 1,014 447 Ribosomal protein S7 psbB LSC 2,205 399 Photosystem II 47 kDa protein rpl2 IR 468 387 Ribosomal protein L2 petB LSC 1,527 379 Cytochrome b6/f apoprotein ndhI SSC 1,491 318 NADH dehydrogenase subunit I psbC LSC 1,401 303 Photosystem II 44 kDa protein psbE LSC 504 302 Photosystem II 8 kDa subunit petD LSC 1,386 283 Cytochrome b/f complex subunit IV rpoC2 LSC 252 283 RNA polymerase betaЈ subunit psbT LSC 1,225 266 Photosystem II T-protein psbF LSC 4,179 260 Photosystem II 4 kDa protein psbL LSC 105 255 Photosystem II L-protein psbJ LSC 120 248 Photosystem II J-protein psbN LSC 117 237 Photosystem II N-protein psbH LSC 123 231 Photosystem II 10 kDa phosphoprotein ndhD SSC 132 212 NADH dehydrogenase subunit D ndhG SSC 222 205 NADH dehydrogenase subunit G psbD LSC 1,503 198 Photosystem II D2 protein ndhB IR 531 177 NADH dehydrogenase subunit B 1,062 162 2,212 160 ture (gene order and organization) and and suggest types and patterns of muta- rates of nucleotide substitution. However, tional processes affecting gene organization structural rearrangements (e.g. gene loss, in the chloroplast genome. In addition, IR loss or expansion, inversions) do occur changes at the nucleotide level can also be in some lineages. These changes can be used to reconstruct phylogenies of plant used to infer relationships of plant lineages groups. The vast majority of phylogenetic

Chloroplast genomes of plants 63 studies utilize cpDNA characteristics as Acknowledgements markers. Most evolutionary studies com- pare nucleotide sequence data, although We thank Steve Wagner and our students historically comparing restriction digest (Rhiannon Peery, Nichole Fine, Melissa patterns was important. Where available, Phillips, Tim Chumley, Andy Alverson, structural changes have provided some Stacia Wyman, Josh McDill, Aneke Padolina, important insights into the evolution of Ruth Timme, Elizabeth Ruck, Mike Moore land plants. Nuclear and mitochondrial and Mary Guisinger) for reading and com- markers are likely to be better utilized by menting on an earlier version of our manu- the plant systematics community in the script. NSF has provided continued funding future; however, the chloroplast genome for our work on chloroplast genomes (cur- will continue to be important due to its rently, RUI/DEB0075700 to LAR and logistical advantages and its evolutionary DEB0120709 to RKJ and LAR). Figures 4.1, characteristics. 4.2 and 4.3 were prepared by Gwen Gage. References Ali, J., Kishima, Y., Mikami, T. and Adachi, T. (1997) Expansion of the IR in the chloroplast genomes of buck- wheat species is due to incorporation of an SSC sequence that could be mediated by an inversion. Current Genetics 31, 276–279. Bock, R. (2000) Sense from nonsense: how the genetic information of chloroplasts is altered by RNA editing. Biochimie 82, 549–557. Bömmer, D., Haberhausen, G. and Zetsche, K. (1993) A large deletion in the plastid DNA of the holopara- sitic flowering plant Cuscuta relexa concerning two ribosomal proteins (rpl2, rpl23), on transfer RNA (trnI) and an ORF2280 homologue. Current Genetics 24, 171–176. Bourque, G. and Pevzner, P.A. (2002) Genome-scale evolution: reconstructing gene order in the ancestral species. Genome Research 12, 26–36. Bowman, C.M., Bonnard, G. and Dyer, K. (1983) Chloroplast DNA variation between species of Triticum and Aegilops. Location of variation on the chloroplast genome and its relevance to the inheritance and classification of the cytoplasm. Theoretical and Applied Genetics 65, 247–262. Bremer, K. (1987) Tribal interrelationships of the Asteraceae. Cladistics 3, 210–253. Bremer, K. (1994) Asteraceae: Cladistics and Classification. Timber Press, Portland, Oregon. Burke, D.H., Raubeson, L.A., Alberti, M., Hearst, J.E., Jordan, E.T., Kirch, S.A., Valinski, A.E.C., Conant, D.S. and Stein, D.B. (1993) The chlL (frxC) gene: phylogenetic distribution in vascular plants and DNA sequence from Polystichum acrostichoides (Pteridophyta) and Synechococcus sp. 7002 (Cyanobacteria). Plant Systematics and Evolution 187, 89–102. Casano, L.M., Martin, M. and Sabater, B. (2001) Hydrogen peroxide mediates the induction of chloroplastic ndh complex under photooxidative stress in barley. Plant Physiology 125, 1450–1458. Chase, M., Soltis, D., Olmstead, R., Morgan, D., Les, D., Mishler, B., Duvall, M., Price, R., Hills, H., Qui, Y.- L., Kron, K., Rettig, J., Conti, E., Palmer, J., Manhart, J., Sytsma, K., Michaels, H., Kress, J., Karol, K., Clark, D., Hedren, M., Gaut, B., Jansen, R., Kim, K.-J., Wimpee, C., Smith, J., Furnier, G., Straus, S., Xiang, Q.-Y., Plunkett, G., Soltis, P., Swensen, S., Williams, S., Gadek, P., Quinn, C., Equiarte, L., Golenberg, E., Learn, G. Jr, Graham, S., Barrett, S., Dayanandan, S. and Albert, V. (1993) Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80, 528–580. Clegg, M.T., Brown, A.D.H. and Whitfield, P.R. (1984) Chloroplast DNA diversity in wild and cultivated bar- ley: implications for genetic conservation. Genetic Research 43, 339–343. Coates, D. and Cullis, C.A. (1987) Chloroplast DNA variability among Linum species. American Journal of Botany 74, 260–268. Cosner, M.E. (1993) Phylogenetic and molecular evolutionary studies of chloroplast DNA variation in the Campanulaceae. PhD dissertation, Ohio State University, Columbus, Ohio. Cosner, M.E., Jansen, R.K., Palmer, J.D. and Downie, S.R. (1997) The highly rearranged chloroplast genome of Trachelium caeruleum (Campanulaceae): multiple inversions, inverted repeat expansion and con- traction, transposition, insertions/deletions, and several repeat families. Current Genetics 31, 419–429.

64 L.A. Raubeson and R.K. Jansen Cosner, M.E., Jansen, R.K., Moret, B.M.E., Raubeson, L.A., Wang, L.-S., Warnow, T. and Wyman, S. (2000) A new fast heuristic for computing the breakpoint phylogeny and experimental analyses of real and syn- thetic data. In: Bourne, P., Gribskov, M., Altman, R., Jensen, N., Hope, D., Lengauer, T., Mitchell, J., Scheeff, E., Smith, C., Strande, S. and Weissig, H. (eds) Proceedings of the 8th International Conference on Intelligent Systems for Molecular Biology. La Jolla, California, pp. 104–115. Delwiche, C.F. and Palmer, J.D. (1997) The origin of plastids and their spread via secondary symbiosis. Plant Systematics and Evolution 11 (suppl.), 53–86. Doebley, J.F., Durbin, M., Golenberg, E.M., Clegg, M.T. and Ma, D.P. (1990) Evolutionary analysis of the large subunit of carboxylase (rbcL) nucleotide sequence among the grasses (Gramineae). Evolution 44, 1097–1108. Douglas, S.E. (1998) Plastid evolution: origins, diversity, trends. Current Opinions in Genetics and Development 8, 655–661. Downie, S.R. and Palmer, J.D. (1992) Use of chloroplast DNA rearrangements in reconstructing plant phy- logeny. In: Soltis, P.S., Soltis, D.E. and Doyle, J.J. (eds) Molecular Systematics of Plants. Chapman & Hall, New York, pp. 14–35. Downie, S.R., Llanas, E. and Katz-Downie, D.S. (1996) Multiple independent losses of the rpoC1 intron in angiosperm chloroplast DNAs. Sytematic Botany 21, 135–151. Doyle, J.J. (1993) DNA, phylogeny, and the flowering of systematics. Bioscience 43, 380–389. Doyle, J.J., Davis, J.I., Soreng, R.J., Garvin, D. and Anderson, M.J. (1992) Chloroplast DNA inversions and the origin of the grass family (Poaceae). Proceedings of the National Academy of Sciences USA 89, 7723–7726. Doyle, J.J., Doyle, J.L. and Palmer, J.D. (1996) The distribution and phylogenetic significance of a 50-kbp chloroplast DNA inversion in the flowering plant family Leguminosae. Molecular Phylogenetics and Evolution 5, 429–438. Gaut, B.S. (1998) Molecular clocks and nucleotide substitution rates in higher plants. In: Hecht, M.K., MacIntyre, R.J. and Clegg, M.T. (eds) Evolutionary Biology, Vol. 30. Plenum Press, New York, pp. 93–120. Gaut, B.S., Clark, L.G., Wendel, J.F. and Muse, S.V. (1997) Comparisons of the molecular evolutionary process at rbcL and ndhF in the grass family (Poaceae). Molecular Biology and Evolution 14, 769–777. Givnish, T.J and Sytsma, K.J. (1997) Homoplasy in molecular vs. morphological data: the likelihood of cor- rect phylogenetic inference. In: Givnish, T.J. and Sytsma, K.J. (eds) Molecular Evolution and Adaptive Radiation. Cambridge University Press, Cambridge, pp. 55–101. Gorbunova, V. and Levy, A.A. (1999) How plants make ends meet: DNA double-strand break repair. Trends in Plant Science 4, 263–269. Goremykin, V.V., Hirsch-Ernst, K.I., Wolfl, S. and Hellwig, F.H. (2003a) Analysis of the Amborella trichopoda chloroplast genome sequence suggests that Amborella is not a basal angiosperm. Molecular Biology and Evolution 20, 1499–1505. Goremykin, V.V., Hirsch-Ernst, K.I., Wolfl, S. and Hellwig, F.H. (2003b) The chloroplast genome of the ‘basal’ angiosperm Calycanthus fertilis – structural and phylogenetic analyses. Plant Systematics and Evolution 242, 119–135. Goulding, S.E., Olmstead, R.G., Morden, C.W. and Wolfe, K.H. (1996) Ebb and flow of the chloroplast inverted repeat. Molecular General Genetics 252, 195–206. Graham, S.W. and Olmstead, R.W. (2000) Utility of 17 chloroplast genes from inferring the phylogeny of the basal angiosperms. American Journal of Botany 87, 1712–1730. Graham, S.W., Reeves, P.A., Burns, A.C.E. and Olmstead, R.G. (2000) Microstructural changes in noncoding chloroplast DNA: interpretation, evolution, and utility of indels and inversions in basal angiosperm phylogenetic inference. International Journal of Plant Science 161, S83–S96. Hillis, D.M. (1998) Taxon sampling, phylogenetic accuracy, and investigator bias. Systematic Biology 47, 3–8. Hiratsuka, J., Shimada, H., Whittier, R., Ishibashi, T., Sakamoto, M., Mori, M., Kondo, C., Honji, Y., Sun, C.R., Meng, B.Y., Li, Y.Q., Kanno, A., Nishizawa, Y., Hirai, A., Shinozaki, K. and Sugiura, M. (1989) The complete sequence of the rice (Oryza sativa) chloroplast genome: intermolecular recombination between distinct tRNA genes accounts for a major plastid DNA inversion during the evolution of the cereals. Molecular and General Genetics 217, 185–194. Hoot, S.B. and Palmer, J.D. (1994) Structural rearrangements, including parallel inversions, within the chloroplast genome of Anemone and related genera. Journal of Molecular Evolution 38, 274. Hosaka, K., Ogihara, Y., Matsubayaski, M. and Tsunewaki, K. (1984) Phylogenetic relationships between the tuberous Solanum species as revealed by restriction endonuclease analysis of chloroplast DNA. Japanese Journal of Genetics 59, 349–369.

Chloroplast genomes of plants 65 Hubschmann, T., Hess, W.R. and Borner, T. (1996) Impaired splicing of the rps12 transcript in ribosome-defi- cient plastids. Plant Molecular Biology 30, 109–123. Hupfer, H., Swiatek, M., Hornung, S., Herrmann, R.G., Maier, R.M., Chiu, W.L. and Sears, B. (2000) Complete nucleotide sequence of the Oenothera elata plastid chromosome, representing plastome I of the five distinguishable euoenothera plastomes. Molecular and General Genetics 263, 581–585. Jansen, R.K. and Kim, K.-J. (1996) Implications of chloroplast DNA data for the classification and phylogeny of the Asteraceae. In: Hind, N., Beentje, H. and Pope, G. (eds) Proceedings of Compositae Symposium. Royal Botanic Gardens, Kew, UK, pp. 317–339. Jansen, R.K. and Palmer, J.D. (1987a) A chloroplast DNA inversion marks an ancient evolutionary split in the sunflower family (Asteraceae). Proceedings of the National Academy of Sciences USA 84, 5818–5822. Jansen, R.K. and Palmer, J.D. (1987b) Chloroplast DNA from lettuce and Barnadesia (Asteraceae): structure, gene localization, and characterization of a large inversion. Current Genetics 11, 553–564. Jansen, R.K. and Palmer, J.D. (1988) Phylogenetic implications of chloroplast DNA restriction site variation in the Mutisieae (Asteraceae). American Journal of Botany 75, 751–764. Jansen, R.K., Wee, J.L. and Millie, D. (1998) Comparative utility of restriction site and DNA sequence data for phylogenetic studies in plants. In: Soltis, D.E., Soltis, P.S. and Doyle, J.J. (eds) Molecular Systematics of Plants II: DNA Sequencing. Chapman & Hall, New York, pp. 87–100. Katayama, H. and Ogihara, Y. (1996) Phylogenetic affinities of the grasses to other monocots as revealed by molecular analysis of chloroplast DNA. Current Genetics 29, 572–581. Kato, T., Kaneko, T., Sato, S., Nakamura, Y. and Tabata, S. (2000) Complete structure of the chloroplast genome of a legume, Lotus japonicus. DNA Research 7, 323–330. Kelchner, S.A. (2000) The evolution of non-coding chloroplast DNA and its application in plant systematics. Annals of the Missouri Botanical Garden 87, 482–498. Kim, K.-J. and Jansen, R.K. (1995) ndhF sequence evolution and the major clades in the sunflower family. Proceedings of the National Academy of Sciences USA 92, 10379–10383. Kim, Y.-D. and Jansen, R.K. (1998) Chloroplast DNA restriction site variation and phylogeny of the Berberidaceae. American Journal of Botany 85, 1766–1778. Kim, K.-J., Jansen, R.K., Wallace, R.S., Michaels, H.J. and Palmer, J.D. (1992) Phylogenetic implications of rbcL sequence variation in the Asteraceae. Annals of the Missouri Botanical Garden 79, 428–445. Knox, E.B. and Palmer, J.D. (1999) The chloroplast genome arrangement of Lobelia thuliana (Lobeliaceae): expansion of the inverted repeat in an ancestor or the Campanulales. Plant Systematics and Evolution 214, 49–64. Kugita, M., Kaneko, A., Yamamoto, Y., Takeya, Y., Matsumoto, T. and Yoshinaga, K. (2003a) The complete nucleotide sequence of the hornwort (Anthoceros formosae) chloroplast genome: insight into the earli- est land plants. Nucleic Acids Research 31, 716–721. Kugita, M., Yamamoto, Y., Fukikawa, T. and Matsumoto, T. (2003b) RNA editing in hornwort chloroplasts makes more than half the genes functional. Nucleic Acids Research 31, 2417–2423. Kuroda, H. and Maliga, P. (2002) Overexpression of the clpP 5’-untranslated region in a chimeric context causes a mutant phenotype, suggesting competition for a clpP-specific RNA maturation factor in tobacco chloroplasts. Plant Physiology 129, 1600–1606. Lavin, M., Doyle, J.J. and Palmer, J.D. (1990) Systematic and evolutionary significance of the loss of the large chloroplast DNA inverted repeat in the family Leguminosae. Evolution 44, 390–402. Lemieux, C., Otis, C. and Turmel, M. (2000) Ancestral chloroplast genome in Mesostigma viride reveals an early branch of green plant evolution. Nature 403, 649–652. Maier, R.M., Neckermann, K., Igloi, G.L. and Kossel, H. (1995) Complete sequence of the maize chloroplast genome: gene content, hotspots of divergence and fine tuning of genetic information by transcript edit- ing. Journal of Molecular Biology 251, 614–628. Manen, J-F. and Natali, A. (1995) Comparison of the evolution of ribulose-1,5-bisphosphate carboxylase (rbcL) and atpB-rbcL non-coding spacer sequences in a recent plant group in the tribe Rubieae (Rubiaceae). Journal of Molecular Evolution 41, 920–927. Manhart, J.R. (1994) Phylogenetic analysis of green plant rbcL sequences. Molecular Phylogenetics and Evolution 3, 114–127. Martin, M., Rujan, T., Richly, T., Hansen, A., Cornelsen, S., Lins, T., Leister, D., Stoebe, B., Hasegawa, M. and Penny, D. (2002) Evolutionary analysis of Arabidopsis, cyanobacterial, and chloroplast genomes reveals plastid phylogeny and thousands of cyanobacterial genes in the nucleus. Proceedings of the National Academy of Sciences USA 99, 12246–12251. Martin, W., Stoebe, B., Goremykin, V., Hansmann, S., Hasegawa, M. and Kowallik, K.V. (1998) Gene trans- fer to the nucleus and the evolution of chloroplasts. Nature 393, 162–165.

66 L.A. Raubeson and R.K. Jansen Matsuoka, Y., Yamazaki, Y., Ogihara, Y. and Tsunewaki, K. (2002) Whole chloroplast genome comparison of rice, maize, and wheat: implications for chloroplast gene diversification and phylogeny of cereals. Molecular Biology and Evolution 19, 2084–2091. Maul, J.E., Lilly, J.W., Cui, L., dePamphilis, C.W., Miller, W., Harris, E.H. and Stern, D.B. (2002) The Chlamydomonas reinhardtii plastid chromosome: islands of genes in a sea of repeats. The Plant Cell 14, 1–22. McFadden, G.I. (2001) Primary and secondary endosymbiosis and the origin of plastids. Journal of Phycology 37, 951–959. Millen, R.S., Olmstead, R.G., Adams, K.L., Palmer, J.D., Lao, N.T., Heggie, L., Kavanagh, T.A., Hibberd, J.M., Gray, J.C., Morden, C.W., Calie, P.J., Jermlin, L.S. and Wolfe, K.H. (2001) Many parallel losses of infA from chloroplast DNA during angiosperm evolution with multiple independent transfers to the nucleus. The Plant Cell 13, 645–658. Miyagi, T., Kapoor, S., Sugita, M. and Sugiura, M. (1998) Transcript analysis of the tobacco plastid operon rps2/atpI/H/F/A reveals the existence of a non-consensus type (NCII) promoter upstream of the atpI coding sequence. Molecular and General Genetics 257, 299–307. Miyamoto, T., Obokata, J. and Sugiura, M. (2002) Recognition of RNA editing sites is directed by unique proteins in chloroplasts: biochemical identification of cis-acting elements and trans-acting factors involved in RNA editing in tobacco and pea chloroplasts. Molecular and Cellular Biology 22, 6726–6734. Moreira, D. and Philippe, H. (2001) Sure facts and open questions about the origin and evolution of photo- synthetic plastids. Research in Microbiology 152, 771–780. Moret, B., Wang, L., Warnow, T. and Wyman, S. (2001) New approaches for reconstructing phylogenies from gene order data. In: Proceedings of the International Conference on Intelligent Systems in Molecular Biology (ISMB 2001), 21–25 July 2001, Copenhagen, Denmark, pp. 165–173. Mullet, J.E. (1993) Dynamic regulation of chloroplast transcription. Plant Physiology 103, 309–313. Muse, S.V. and Gaut, B.S. (1997) Comparing patterns of nucleotide substitution rates among chloroplast loci using relative ratio test. Genetics 146, 393–399. Nickrent, D.L., Parkinson, C.L., Palmer, J.D. and Duff, R.J. (2000) Multigene phylogeny of land plants with spe- cial reference to bryophytes and the earliest land plants. Molecular Biology and Evolution 17, 1885–1895. Nixon, K. (1999) The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15, 407–414. Ogihara, Y., Isono, K., Kojima, T., Endo, A., Hanaoka, M., Shiina, T., Terachi, T., Utsugi, S., Murata, M., Mori, N., Takumi, S., Ikeo, K., Gojobori, T., Murai, R., Murai, K., Matsuoka, Y., Ohnishi, Y., Tajiri, H. and Tsunewaki, K. (2002) Structural features of a wheat plastome as revealed by complete sequencing of chloroplast DNA. Molecular Genetics and Genomics 266, 740–746. Ohyama, K., Fukuzawa, H., Kohchi, T., Shirai, H., Sano, T., Sano, S., Umesono, K., Shiki, Y., Takeuchi, M., Chang, Z., Aota, S., Inokuchi, H. and Ozeki, H. (1986) Chloroplast gene organization deduced from complete sequence of liverwort Marchantia polymorpha chloroplast DNA. Nature 322, 572–574. Olmstead, R.G. and Palmer, J.D. (1994) Chloroplast DNA systematics: a review of methods and data analy- sis. American Journal of Botany 81, 1205–1224. Palmer, J.D. (1983) Chloroplast DNA exists in two orientations. Nature 301, 92–93. Palmer, J.D. (1991) Plastid chromosomes: structure and evolution. In: Hermann, R.G. (ed.) The Molecular Biology of Plastids. Cell Culture and Somatic Cell Genetics of Plants, Vol. 7A. Springer-Verlag, Vienna, pp. 5–53. Palmer, J.D. (2003) The symbiotic birth and spread of plastids: how many times and whodunit? Journal of Phycology 39, 4–11. Palmer, J.D. and Delwiche, C.F. (1998) Origin and evolution of plastids and their genomes. In: Soltis, D.E., Soltis, P.S. and Doyle, J.J. (eds) Molecular Systematics of Plants II: DNA Sequencing. Chapman & Hall, New York, pp. 375–409. Palmer, J.D. and Stein, D.B. (1986) Conservation of chloroplast genome structure among vascular plants. Current Genetics 10, 823–833. Palmer, J.D. and Zamir, D. (1982) Chloroplast DNA evolution and phylogenetic relationships in Lycopersicon. Proceedings of the National Academy of Sciences USA 81, 8014–8018. Palmer, J.D., Boynton, J.E., Gillham, N.W. and Harris, E.H. (1985) Evolution and recombination of the large inverted repeat in Chlamydomonas chloroplast DNA. In: Steinbeck, K.E., Bonitz, S., Arntzen, C.J. and Bogorad, L. (eds) Molecular Biology of the Photosynthetic Apparatus. Cold Spring Harbor Laboratory Press, New York, pp. 269–278.

Chloroplast genomes of plants 67 Palmer, J.D., Aldrich, J. and Thompson, W.F. (1987a) Chloroplast DNA evolution among legumes: loss of a large inverted repeat occurred prior to other sequence rearrangements. Current Genetics 11, 275–286. Palmer, J.D., Nugent, J.M. and Herbon, L.A. (1987b) Unusual structure of geranium chloroplast DNA: a triple-sized inverted repeat, extensive gene duplications, multiple inversions and two repeat families. Proceedings of the National Academy of Sciences USA 84, 769–773. Palmer, J.D., Jansen, R.K., Michaels, H., Manhart, J. and Chase, M. (1988) Chloroplast DNA variation and plant phylogeny. Annals of the Missouri Botanical Garden 75, 1180–1206. Perry, A.S. and Wolfe, K.H. (2002) Nucleotide substitution rates in legume chloroplast DNA depend on the presence of the inverted repeat. Journal of Molecular Evolution 55, 501–508. Perry, A.S., Brennan, S., Murphy, D.J., Kavanagh, T.A. and Wolfe, K.H. (2002) Evolutionary re-organisation of a large operon in adzuki bean chloroplast DNA caused by inverted repeat movement. DNA Research 9, 157–162. Plunkett, G.M. and Downie, S.R. (2000) Expansion and contraction of the chloroplast inverted repeat in Apiaceae subfamily Apiodeae. Systematic Botany 25, 648–667. Price, R.A., Caile, P.J., Downie, S.R., Logsdon, J.M. Jr and Palmer, J.D. (1990) Chloroplast DNA variation in the Geraniaceae: a preliminary report. In: Vorster, P. (ed.) Proceedings of the International Geraniaceae Symposium, 24–26 September 1990, Stellenbosch, South Africa, pp. 1–8. Pryer, K.M., Schneider, H., Smith, A.R., Cranfill, R., Wolf, P.G., Hunt, J.S. and Sipes, S.D. (2001) Horsetails and ferns are a monophyletic group and the closest living relatives to seed plants. Nature 409, 618–621. Rai, H.S., O’Brien, H.E., Reeves, P.A., Olmstead, R.G. and Graham, S.W. (2003) Inference of higher-order relationships in the cycads from a large chloroplast data set. Molecular Phylogenetics and Evolution 29, 350–359. Raubeson, L.A. (1991) Structural variation in the chloroplast genome of vascular land plants. PhD disserta- tion, Yale University, New Haven, Connecticut. Raubeson, L.A. and Jansen, R.K. (1992a) A rare chloroplast-DNA structural mutation is shared by all conifers. Biochemical Systematics and Ecology 20, 17–24. Raubeson, L.A. and Jansen, R.K. (1992b) Chloroplast DNA evidence on the ancient evolutionary split in vas- cular land plants. Science 255, 1697–1699. Raubeson, L.A. and Stein, D.B. (1995) Insights into fern evolution from mapping chloroplast genomes. American Fern Journal 85, 193–204. Rice, K.A., Donoghue, M.J. and Olmstead, R.G. (1997) Analyzing large data sets: rbcL 500 revisited. Systematic Biology 46, 554–563. Rokas, A. and Holland, P.W.H. (2000) Rare genomic changes as tools for phylogenetics. Trends in Ecology and Evolution 15, 454–459. Sabater, B., Martin, M., Schmitz-Linneweber, C. and Maier, R.M. (2002) Is clustering of plastid RNA editing sites a consequence of transitory loss of gene function? Implications for past environmental and evolu- tionary events in plants. Perspectives in Plant Ecology, Evolution and Systematics 5, 81–89. Sato, S., Nakamura, Y., Kaneko, T., Asamizu, E. and Tabata, S. (1999) Complete structure of the chloroplast genome of Arabidopsis thaliana. DNA Research 6, 283–290. Schmitz-Linneweber, C., Maier, R.M., Alcaraz, J.P., Cottet, A., Herrmann, R.G. and Mache, R. (2001) The plastid chromosome of spinach (Spinacia oleracea): complete nucleotide sequence and gene organiza- tion. Plant Molecular Biology 45, 307–315. Schmitz-Linneweber, C., Regel, R., Du, T.G., Hupfer, H., Herrmann, R.G. and Maier, R.M. (2002) The plastid chromosome of Atropa belladonna and its comparison with that of Nicotiana tabacum: the role of RNA editing in generating divergence in the process of plant speciation. Molecular Biology and Evolution 19, 1602–1612. Shields, D.C. and Wolfe, K.H. (1997) Accelerated rates of evolution of sites undergoing RNA editing in plant mitochondria and chloroplasts. Molecular Biology and Evolution 14, 344–349. Shinozaki, K., Ohme, M., Tanaka, M., Wakasugi, T., Hayashida, N., Matsubayashi, T., Zaita, N., Chunwongse, J., Obokata, J., Yamaguchi-Shinozaki, K., Ohto, C., Torazawa, K., Meng, B.Y., Sugita, M., Deno, H., Kamogashira, T., Yamada, K., Kusuda, J., Takaiwa, F., Kato, A., Tohdoh, N., Shimada, H. and Sugiura, M. (1986) The complete nucleotide sequence of the tobacco chloroplast genome: its gene organization and expression. EMBO Journal 5, 2043–2049. Simpson, C.L. and Stern, D.B. (2002) The treasure trove of algal chloroplast genomes. Surprises in architec- ture and gene content, and their functional implications. Plant Physiology 129, 957–966. Soltis, D.E. and Soltis, P.S. (1998) Choosing an approach and an appropriate gene for phylogenetic analysis. In: Soltis, D.E., Soltis, P.S. and Doyle, J.J. (eds) Molecular Systematics of Plants II: DNA Sequencing. Chapman & Hall, New York, pp. 1–42.

68 L.A. Raubeson and R.K. Jansen Soltis, D.E., Soltis, P.S., Clegg, M.T. and Durbin, M. (1990) rbcL sequence divergence and phylogenetic rela- tionships among Saxifragaceae sensu lato. Proceedings of the National Academy of Sciences USA 87, 4640–4644. Soltis, D.E., Soltis, P.S., Chase, M.W., Mort, M.W., Albach, D.C., Zanis, M., Savolainen, V., Hahn, W.H., Hoot, S.B., Fay, M.F., Axtell, M., Swensen, S.M., Prince, L.M., Kress, W.J., Nixon, K.J. and Farris, J.S. (2000) Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnaean Society 133, 381–461. Stein, D.B., Palmer, J.D. and Thompson, W.F. (1986) Structural evolution and flip-flop recombination of chloroplast DNA in the fern genus Osmunda. Current Genetics 10, 835–841. Stein, D.B., Conant, D.S., Ahearn, M.E., Jordan, E.T., Kirch, S.A., Hasebe, M., Iwatsuki, K., Tan, M.K. and Thomson, J.A. (1992) Structural rearrangements of the chloroplast genome provide an important phylo- genetic link in ferns. Proceedings of the National Academy of Sciences USA 89, 1856–1860. Stiller, J.W., Reel, D.C. and Johnson, J.C. (2003) A single origin of plastids revisited: convergent evolution in organellar gene content. Journal of Phycology 39, 95–105. Sugiura, C., Kobayashi, Y., Aoki, S., Sugita, C. and Sugita, M. (2003) Complete chloroplast DNA sequence of the moss Physcomitrella patens: evidence for the loss and relocation of rpoA from the chloroplast to the nucleus. Nucleic Acids Research 31, 5324–5331. Sugiura, M. (2003) History of chloroplast genomics. Photosynthesis Research 76, 371–377. Sytsma, K.J. and Gottlieb, L.D. (1986) Chloroplast DNA evidence for the origin of the genus Heterogaura from a species of Clarkia (Onagraceae). Proceedings of the National Academy of Sciences USA 83, 5554–5557. Sytsma, K.J. and Schaal, B.A. (1985) Phylogenetics of the Lisianthius skinneri (Gentianaceae) species com- plex in Panama utilizing DNA restriction fragment analysis. Evolution 39, 594–608. Tsudzuki, J., Nakashima, K., Tsudzuki, T., Hiratsuka, J., Shibata, M., Wakasugi, T. and Sugiura, M. (1992) Chloroplast DNA of black pine retains a residual inverted repeat lacking rRNA genes: nucleotide sequences of trnQ, trnK, psbA, trnI and trnH and the absence of rps16. Molecular and General Genetics 232, 206–214. Turmel, M., Otis, C. and Lemieux, C. (1999) The complete chloroplast DNA sequence of the green alga Nephroselmis olivacea: insights into architecture of ancestral chloroplast genomes. Proceedings of the National Academy of Sciences USA 96, 10248–10253. Wakasugi, T., Tsudzuki, J., Ito, S., Nakashima, K., Tsudzuki, T. and Sugiura, M. (1994) Loss of all ndh genes as determined by sequencing the entire chloroplast genome of the black pine Pinus thunbergii. Proceedings of the National Academy of Sciences USA 91, 9794–9798. Wakasugi, T., Sugita, M., Tsudzuki, T. and Sugiura, M. (1998) Updated gene map of tobacco chloroplast DNA. Plant Molecular Biology Reporter 16, 231–241. Wallace, R.S. and Cota, J.H. (1996) An intron loss in the chloroplast gene rpoC1 supports a monophyletic origin for the subfamily Cactoideae of the Cactaceae. Current Genetics 29, 275–281. Wang, L.-S., Jansen, R.K., Moret, B.M.E., Raubeson, L.A. and Warnow, T. (2002) Fast phylogenetic methods for analysis of genome rearrangement data: an empirical study. In: Proceedings of the 7th Pacific Symposium Biocomputing PSB 2002, 3–7 January 2002, Lihue, Hawaii, pp. 524–535. Wolf, P.G., Rowe, C.A., Sinclair, R.B. and Hasebe, M. (2003) Complete nucleotide sequence of the chloro- plast genome from a leptosporangiate fern, Adiantum capillus-veneris L. DNA Research 10, 59–65. Wolfe, K.H., Li, W.-H. and Sharp, P.M. (1987) Rates of nucleotide substitution vary greatly among plant mitochondria, chloroplast, and nuclear DNAs. Proceedings of the National Academy of Sciences USA 84, 9054–9058. Wolfe, K.H., Morden, C.W. and Palmer, J.D. (1992) Function and evolution of a minimal plastid genome from a nonphotosynthetic parasitic plant. Proceedings of the National Academy of Sciences USA 89, 10648–10652. Yamada, T. (1991) Repetitive sequence-mediated rearrangements in Chlorella ellipsoidea chloroplast DNA: completion of nucleotide sequence of the large inverted repeat. Current Genetics 19, 139–147.

5 The mitochondrial genome of higher plants: a target for natural adaptation Sally A. Mackenzie Plant Science Initiative, N305 Beadle Center for Genetics Research, University of Nebraska, Lincoln, NE 68588-0660, USA Introduction terns might be tolerated more in plant sys- tems than in animal, and the mechanisms Some of the most intriguing examples of underlying these selective organelle trans- adaptation in eukaryotes arise within the mission patterns, are not yet well under- plant kingdom, many in response to a plant’s stood. Mechanisms for selective exclusion of immotility and consequent inability to escape paternal organelles vary. Whereas some sys- environmental stresses (Hawkesford and tems appear to target paternally derived Buchner, 2001). These unique attributes organellar DNA for selective destruction or occur in various forms to produce wonders suppressed replication (Nagata et al., 1999; of plant architecture, specialized physiology Sodmergen et al., 2002; Moriyama and and reproductive strategies. At a cellular Kawano, 2003), some animal systems are level, several unique features of plant metab- postulated to exclude or destroy the pater- olism and organellar genome maintenance nal organelles themselves (Sutovsky, 2003). are evident in plants (Mackenzie and McIntosh, 1999). Some of these cellular When considering organelle inheritance attributes are thought to be the outcome of and segregation processes, one must keep in the endosymbiotic process that has led to the mind the distinct dynamics of organelle present-day plastid and consequent mito- behaviour. In contrast to nuclear genetic chondrial–plastid co-evolution (Allen, 1993; information, which undergoes replication at Adams et al., 2002; Elo et al., 2003). a precise point within a tightly regulated cell cycle, segregating to daughter cells in equal As in the case of most animal systems, and unchanging copy number each cellular organellar genomes generally show strict generation, organellar genomes obey very maternal inheritance in plants. However, different rules. Organellar DNA replication there are exceptions to this pattern. In some does not maintain tight synchrony with the cases, paternal inheritance is observed, cell cycle (Birky, 1983), and the numbers of though varying degrees of biparental inheri- genomes per organelle and organelles per tance are also seen (Reboud and Zeyl, 1994; cell vary dramatically depending on tissue Zhang et al., 2003). In nearly all such excep- type. In general, mitochondrial biogenesis is tions, the plastid has been more likely to highest in meristematic and reproductive show variation from strict maternal inheri- tissues, where numbers of genomes per tance than the mitochondrion. Why the mitochondrion and mitochondria per cell relaxation of strict maternal inheritance pat- generally range in the hundreds, while © CAB International 2005. Plant Diversity and Evolution: Genotypic and 69 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

70 S.A. Mackenzie mitochondrial numbers decrease markedly The endosymbiotic processes believed to in vegetative tissues (Lamppa and Bendich, have given rise to present day organelles are 1984; Fujie et al., 1993; Robertson et al., generally thought to have involved the 1995). The ‘segregation’ of organelles, exist- transfer of large amounts of genetic infor- ing as multi-genomic populations within mation from mitochondrial and plastid cells, occurs by a process of cytoplasmic sort- progenitor genomes to the nucleus. Very ing throughout development. Although early on, these transfers might have cytoplasmic sorting is generally considered a involved large genomic segments that stochastic process, nuclear gene influence on encompassed many genes simultaneously. cytoplasmic segregation is evident. More recent gene transfers, following the advent of RNA editing processes, have prob- This review will describe the unusual ably occurred as singular gene events that nature of plant mitochondrial genomes, con- involve an RNA intermediate in the process trasting their features and behaviour with (Nugent and Palmer, 1991; Covello and what is known of mammalian and fungal sys- Gray, 1992). With the massive nuclear tems. One anticipates that the considerable genomic rearrangements that have occurred divergence observed in plants derives, at in plants subsequent to the endosymbiotic least in part, from the unusual plant cellular events (Blanc et al., 2000), it is difficult to context of mitochondrial–chloroplast co-evo- envisage how genetic linkage has been main- lution. With availability of complete plant tained for large numbers of transferred genome sequence information, considerable genes without selection. One possibility is evidence has accumulated recently in sup- that maintenance of related genes in a link- port of this assumption. However, it is also age might facilitate coordinate gene regula- important to note that the vast majority of tion during key points in development genes essential for mitochondrial processes (Boutanaev et al., 2002). For example, at the are nuclear encoded; the mitochondrial point immediately following pollination, a genome, though essential, encodes less than maternally derived cytoplasm must immedi- 5% of the information required for its varied ately establish compatibility with the newly functions. Therefore, it is impossible to con- introduced paternal nuclear contribution. sider plant mitochondrial genome evolution Possibly epigenetic regulation would be cru- and adaptation without addressing the criti- cial for re-establishing necessary interge- cal role of the nucleus in these ongoing nomic compatibility. processes. A second intriguing observation regard- Nuclear Regulation of Mitochondrial ing the nuclear genes that participate in DNA and RNA Metabolism organellar genome maintenance is the large number predicted to encode proteins func- The availability of complete genome tional in both mitochondria and plastids sequences for Arabidopsis and rice has allowed (Hedtke et al., 1997; Beardslee et al., 2002; the identification of several candidate genes Elo et al., 2003). This assumption is sup- predicted to function in organellar DNA and ported by individual genes that encode RNA metabolism functions. Two striking fea- dual-targeted proteins, as well as genes that tures of nuclear genes that appear to direct have undergone duplication, with duplicate organellar genome maintenance are evident. members targeting distinct organelle types. The first surprising property of these genes The prevalence of proteins that apparently is their organization within the nuclear function in both mitochondria and plastids genome. Nuclear genes encoding organellar suggests that a substantial component of the DNA and RNA metabolism loci appear to be DNA and RNA metabolism apparatus is largely clustered in a few regions of the overlapping in the two organelle types plant genome (Elo et al., 2003). Moreover, (Small et al., 1998; Peeters and Small, 2001). this genomic arrangement may not be lim- At one time this would have seemed an ited to plants alone (Lefai et al., 2000). incongruous idea, given the numerous dif- ferences that exist in mitochondrial and

The mitochondrial genome of higher plants 71 plastid functions, genome structure and frequency DNA exchange at these sites pro- gene organization. These distinctions are duces a complex assemblage of large inver- likely to be remnant features of their pro- sions and subgenomic DNA molecules, each genitors, as the number of reports of shared, containing a portion of the genome. probably acquired, features increases. Whether this recombination activity is con- tinuous throughout plant development, or The Mitochondrial Genome of Higher restricted to a particular cell type, is not Plants clear. The technical difficulties that have complicated these studies arise from the Plant mitochondrial genomes are distin- entwined nature of replicative and recombi- guished by their extreme variation in size, national processes. ranging from about 200 to 2400 kbp, in com- parison to the 16-kbp genome of human In addition to high-frequency homolo- mitochondria and intermediate but less vari- gous recombination, plant mitochondrial ably sized genomes of fungi. With the recent genomes commonly undergo low-frequency availability of complete mitochondrial recombinations at non-homologous, often genome sequences for at least four plant intragenic, sites. This ectopic recombination species, we now find that the dramatic differ- activity gives rise to chimeric gene configu- ences in genome size are not accounted for rations, often expressive, within a wide by vast discrepancies in coding capacity. In array of plant species. In many cases, these fact, plant mitochondrial genomes encode unusual gene chimeras are discovered by somewhere between 55 and 70 genes (Oda et their causative association with cytoplasmic al., 1992; Unseld et al., 1997; Notsu et al., male sterility (Schnable and Wise, 1998). 2002; Handa, 2003), less than twice the However, not all ectopic recombinations number of genes found in the human mito- necessarily produce a detectable phenotype chondrion. Considerable sequence redun- (Marienfeld et al., 1997). dancy, integration of non-mitochondrial DNA, and ectopic recombination have con- Cytoplasmic male sterility (CMS) is a con- tributed to the observed variation. dition in which the plant is unable to pro- duce or shed viable pollen as a consequence The mitochondrial genome of plants con- of mitochondrial mutation. In nearly all sists of a heterogeneous population of both cases investigated, the associated mitochon- circular and linear DNA molecules, many drial mutations are dominant, stemming existing in highly branched configurations from expression of unique sequence (Backert et al., 1996; Bendich, 1996; chimeras that form aberrant open reading Oldenburg and Bendich, 1996; Backert and frames. To date, all cases of CMS appear to Borner, 2000). To date, an origin of replica- be associated with ectopic recombination tion has not been defined in plants, and evi- within the genome, implying an adaptive dence suggests that replication may occur, at advantage to this activity. No two CMS least in part, by a rolling circle mechanism. mutations have been identical, although In fact, it has been suggested that replication striking similarities have been observed in may initiate by a strand invasion process some cases (Tang et al., 1996). Interestingly, perhaps resembling that of T4 phage the frequency of non-homologous recombi- (Backert and Borner, 2000). nation, or the relative copy number of the derived recombinant forms, appears to be In contrast to mammalian mitochondria, controlled by nuclear genes. plant mitochondrial genomes are unusually dynamic in their structure, in part a conse- Nuclear Regulation of Mitochondrial quence of prolific intra- and intergenomic Genome Structure recombination activity. Within most plant mitochondrial genomes are dispersed sev- Whereas evolutionary pressures appear to eral repeated sequences, present in both have selected for a highly conservative, sta- inverted and direct orientations. High- ble and compact mitochondrial genome

72 S.A. Mackenzie configuration within animals, adaptive plant is male fertile. Interestingly, the condi- selection within the plant kingdom has pro- tion of male fertility is not reversed by the duced a system that appears to benefit segregation of Fr, suggesting that the prod- from a high degree of variability uct of Fr acts unidirectionally or in a limited (Marienfeld et al., 1999). A fascinating context, and the reversal of Fr action might example of the dynamic nature of the plant require additional nuclear components. mitochondrial genome is the widespread phenomenon termed stoichiometric shift- In Arabidopsis, mutation at the nuclear ing (Small et al., 1987). locus CHM results in the copy number amplification of a mitochondrial chimeric Certain subgenomic mitochondrial DNA DNA configuration and the appearance of configurations change dramatically in rela- green–white variegation (Martinez-Zapater tive copy number during the development et al., 1992; Sakamoto et al., 1996). Several of the plant. When present substoichiomet- mutant alleles of CHM are available in rically, these components of the mitochon- Arabidopsis, presenting an opportunity to drial genome have been estimated at one clone the gene. copy per every 100–200 cells of the plant (Arrieta-Montiel et al., 2001), representing a Recently, the product of the CHM locus heteroplasmic (heterogeneous cytoplasmic) was shown to resemble the MUTS protein of condition. However, stoichiometric shifting Escherichia coli, and the gene has now been can result in preferential amplification of designated MSH1 (MutS Homologue 1; these molecules to levels equimolar with the Abdelnoor et al., 2003). MutS is a component principal mitochondrial genome. The selec- of the DNA mismatch repair apparatus, and tive amplification or suppression of particu- several of its homologues have been identi- lar portions of the mitochondrial genome is fied to function within the nuclear genome influenced by nuclear genotype. of higher eukaryotes. One other nuclear gene encoding a mitochondrial MutS homo- Stoichiometric shifting has been reported logue was reported several years ago in in a wide array of plant species (Mackenzie yeast (Reenan and Kolodner, 1992). In that and McIntosh, 1999), and shifting events case, the mitochondrial protein was sug- are apparently induced under conditions of gested to function in mismatch repair (Chi cell culture or cybridization (Kanazawa et al., and Kolodner, 1994). 1994; Gutierres et al., 1997; Bellaoui et al., 1998), alloplasmy (Kaul, 1988), spontaneous Mismatch repair components appear to CMS reversion to pollen fertility (Mackenzie serve two important functions within the et al., 1988; Smith and Chowdhury, 1991), eukaryotic genome: to bind and repair and the introduction or mutation of specific nucleotide mismatches and to suppress non- nuclear genes. homologous recombination activity (Modrich and Lahue, 1996; Harfe and In the CMS system of common bean, the Jinks-Robertson, 2000). Enhanced ectopic mitochondrial sequence associated with recombination appears to be the effect pollen sterility, designated pvs-orf239, resides within the plant mitochondrial genome in on a mitochondrial molecule that is shifted response to MSH1 mutation (Abdelnoor et to substoichiometric levels in response to the al., 2003). introduction of the dominant nuclear gene Fr (Mackenzie and Chase, 1990). This is an Interestingly, allelic variation for MutS in appealing system for study because intro- microbial populations appears to provide duction of Fr, by standard crossing, results certain adaptive advantage under severe in Mendelian segregation for a particular, selection conditions by enhancing mutation reproducible mitochondrial rearrangement. frequency (LeClerc et al., 1996; LeClerc and The pvs-orf239 sequence, when present in Cebula, 1997; Bjedov et al., 2003; Chopra et high copy number, is expressed and the al., 2003). This ‘mutator’ phenomenon, aris- plant is male sterile. When Fr is introduced, ing from MutS variation, has been reported and the mitochondrial pvs-orf239 sequence in a number of organisms including is reduced to substoichiometric levels, the humans, where such variation contributes to cancer incidence (Li, 1999). A phenomenon

The mitochondrial genome of higher plants 73 resembling mitochondrial stoichiometric Interestingly, molecular studies of fertility shifting has also been described in restoration mechanisms in several CMS sys- Drosophila, although the nuclear effector is tems reveal a role of most restorer genes in not yet identified (Le Goff et al., 2002). It RNA processing (Schnable and Wise, 1998). now appears likely that evolutionary advan- Over the past few years, five nuclear genes tage might have been realized by permitting that restore pollen fertility to CMS mutants a degree of genetic variation within the have been cloned. Of these, four have been mitochondrial MSH1 gene of higher plants. shown to encode PPR proteins. These include A possible adaptive role of MSH1 variation the restorers of fertility in petunia (Bentolila et in plant populations will be discussed in a al., 2002), Kosena radish (Koizuka et al., later section. 2003), Ogura radish (Brown et al., 2003; Desloire et al., 2003) and rice (Kazama and Nuclear Regulation of Mitochondrial Toriyama, 2003). It would not be surprising to Transcript Processing and Fertility find PPR proteins implicated in the fertility Restoration restoration of several other CMS plant species in the near future. The RNA specificity that is A striking example of a derived plant fea- postulated by PPR protein:RNA binding, ture shared by both mitochondria and plas- combined with the intragenic recombination tids is their dependence on RNA processing, origins of most CMS mutations, appears an editing and stabilizing functions for organel- ideal system for nuclear control of CMS-asso- lar gene expression (Hoffmann et al., 2001; ciated aberrant gene expression. Binder and Brennicke, 2003). RNA process- ing, which involves the cleavage of RNA at Evolutionary Implications of precise sites as part of RNA maturation, and Mitochondrial Genome Dynamics in RNA editing, which generally involves spe- cific C to U conversions within a transcript, Higher Plants are both found to occur in a wide array of plastid and mitochondrial RNAs. The The mitochondrial mutations that confer expansion of these processes has apparently cytoplasmic male sterility have been of great been accompanied, or driven by, a concomi- interest for their value to the hybrid seed tant expansion in the number of nuclear industry (Fig. 5.1). In a broad range of plant genes associated with these functions. species, the phenomenon of heterosis, or hybrid vigour, is well documented (Tsaftaris The pentatricopeptide repeat (PPR) fam- and Kafka, 1998; Rieseberg et al., 2000). The ily of proteins in Arabidopsis numbers over genomic condition produced by hybridiza- 500 members, with over two-thirds encod- tion, probably associated with higher levels ing proteins predicted to target mitochon- of gene heterozygosity and perhaps epige- dria or plastids (Small and Peeters, 2000). netic processes, provides markedly The PPR proteins share almost no enhanced reproductive capacity and plant detectable sequence homology. Rather, they vigour. Although of obvious agricultural are linked by their unusual structural simi- benefit, the heterotic state is also clearly larities. Although highly divergent at their advantageous in natural populations amino termini, each PPR protein contains a (Rieseberg et al., 2000). series of 35-amino-acid repeat structures, present in variable numbers. These repeats Most domesticated crop species are cate- are predicted to confer a helical structure to gorized as predominantly outcrossing or the protein that is postulated to interact self-pollinating. However, in natural popu- directly with RNA or proteins. It has been lations exist the more heterogeneous and suggested that this family of nuclear pro- dynamic conditions of gynodioecy. teins may provide the RNA recognition Gynodioecious populations are composed specificity necessary for RNA processing of both female and hermaphroditic individ- activities. uals, permitting population expansion within geographically isolated environments

74 S.A. Mackenzie Fig. 5.1. Patterns of alteration distinguishing the plant mitochondrial genome. Homologous intra- and intermolecular recombination occurs at repeated sequences (boxes) that, when in direct orientation, can produce an equilibrium of non-recombinant and recombinant subgenomic molecules. Intragenic (␣–␤) ectopic recombination can occur to produce sequence chimeras. What is shown is the simplest scenario; often these chimeric sequences are derived from multiple recombination and/or insertion events (Schnable and Wise, 1998). Note that the final population of molecules does not necessarily include all parental and recombinant forms. Stoichiometric shifting represents a nuclear-directed process that can modulate the relative copy number of particular recombinant molecules within the genome, often reducing them to one copy per 100–200 cells of the plant. The shifting mechanism is not yet understood. Mitochondrial DNA molecules are shown in a circular form, as they map, for convenience. In vivo structures are predominantly linear and often multimeric. together with enhanced genomic recombi- gynodioecy involves the interaction of one to nation. Gynodioecy comprises self- and multiple mitochondrial mutations that con- cross-pollination activity in dynamic equilib- dition CMS together with maintainer (non- rium (Couvet et al., 1990). Whereas many of fertility restoring) and fertility restorer our domesticated crop species are predomi- nuclear genotypes. The CMS cytoplasm, in nantly self-pollinating, it is possible that sev- combination with a fertility-restoring eral of these species originated from nucleus, constitutes a hermaphrodite capa- gynodioecious natural populations. ble of self-pollination, while a CMS cyto- plasm combined with a maintainer genotype Most detailed studies of gynodioecy have constitutes a female, outcrossing form. been conducted in non-crop species such as Extensive cross-pollination activity eventu- Silene vulgaris (Olson and McCauley, 2002), ally introduces the restorer genotype to the Thymus vulgaris (Manicacci et al., 1997) and female, shifting the localized frequency of Plantago lanceolata (Van Damme, 1983). In hermaphrodites. the various natural populations studied,

The mitochondrial genome of higher plants 75 Several research groups have investi- 1998). Surveys of core germplasm collec- gated the environmental factors that influ- tions, encompassing all available genetic ence frequencies of fertility restoring, diversity from natural populations, reveal non-restoring and sterility-determining the presence of the CMS-associated pvs- components of this process within naturally orf239 mitochondrial sequence, highly con- occurring and widely dispersed populations served, in 100% of the lines, although the (Bailey et al., 2003; Jacobs and Wade, 2003). sequence is substoichiometric in over 90% However, these models generally do not of them (Arrieta-Montiel et al., 2001). The account for an additional genetic compo- pvs-orf239 sequence is also present in nent that is probably essential to under- Phaseolus coccineus, Phaseolus polyanthus and standing the maintenance of females in the Phaseolus acutafolius (Hervieu et al., 1993), population. This missing factor is the spon- implying that its evolution predates taneous inter-convertibility of the male- Phaseolus speciation. sterile and male-fertile condition. The nuclear–mitochondrial interactions that In bean, hermaphrodites arise by two dis- effect mitochondrial stoichiometric shifting tinct mechanisms: suppression of the male permit the spontaneous inter-conversion of sterility phenotype can be conditioned by females and hermaphrodites. Female plants the prevalent Fr2 locus (Mackenzie, 1991), located in isolation are dependent on the while substoichiometric shifting of pvs-orf239 process of low frequency spontaneous rever- is conditioned by the more rare Fr locus sion to fertility to facilitate reproduction late (Mackenzie and Chase, 1990). Spontaneous in the plant’s cycle. Spontaneous reversion reversion to fertility ranges in frequency of a CMS plant to male fertility has been depending on nuclear background described in several plant systems (Mackenzie et al., 1988), but even the most (Laughnan and Gabay-Laughnan, 1983; effective maintainer genotypes often result Smith and Chowdhury, 1991; Bellaoui et al., in a small number of seed pods produced 1998; Janska et al., 1998; Andersson, 1999), just prior to plant senescence if the plant has where the frequency of stoichiometric shift- not been artificially pollinated (Mackenzie, ing depends on nuclear background. unpublished data). Although not yet documented in nature, These observations suggest the existence it appears likely that genetic variability at the of a naturally occurring, highly refined MSH1 locus accounts for some portion of the genetic system to facilitate cross-pollination inter-conversion of females and hermaphro- in a species that, upon domestication, has dites. Whether MSH1 displays higher-than- become largely inbreeding. This fascinating average natural mutability in plants has not genetic system apparently integrates yet been investigated, but its unusually com- nuclear suppressors of mitochondrial gene plex gene structure, comprising 21 introns expression with nuclear mechanisms that in at least five plant species studied to date influence heteroplasmic sorting to (Abdelnoor et al., 2003; Abdelnoor and reversibly modulate female:hermaphrodite Mackenzie, unpublished), raises the possibil- ratios. A system integrated in this manner ity of alternative splicing (Black, 2003). would permit a more dynamic response to environmental changes than is generally Recent literature suggests that several of predicted in most models. our present-day crop plants probably derive from gynodioecious natural populations. Evidence suggests that nuclear–mito- These include beet (Beta vulgaris; Cuguen et chondrial genetic interactions, and the al., 1994), pearl millet (Pennisetum ameri- dynamic features of the plant mitochondrial canum; Delorme et al., 1997), sunflower genome, are highly amenable to adaptation (Helianthus spp.; Jan, 2000) and common for reproductive fitness in natural popula- bean (Phaseolus vulgaris; Mackenzie, 1991). tions. Thus, it is not surprising that the most In the common bean, domestication common problems arising in the deploy- occurred from multiple centres of origin ment of cytoplasmic male sterility for com- extending from Mexico to Ecuador (Gepts, mercial hybrid seed production are cytoplasmic instability and incomplete fertil-

76 S.A. Mackenzie ity restoration. These features, undesirable to have served plant populations well for within the highly controlled environmental reproductive success in their ever-changing conditions of today’s agriculture, are likely natural environments. References Abdelnoor, R.V., Yule, R., Elo, A., Christensen, A., Meyer-Gauen, G. and Mackenzie, S. (2003) Substoichiometric shifting in the plant mitochondrial genome is influenced by a gene homologous to MutS. Proceedings of the National Academy of Sciences USA 100, 5968–5973. Adams, K.L., Dalen, D.O., Whelan, J. and Palmer, J.D. (2002) Genes for two mitochondrial ribosomal pro- teins in flowering plants are derived from their chloroplast or cytosolic counterparts. Plant Cell 14, 931–943. Allen, J.F. (1993) Control of gene expression by redox potential and the requirement of chloroplast and mitochondrial genomes. Journal of Theoretical Biology 165, 609–631. Andersson, H. (1999) Female and hermaphrodite flowers on a chimeric gynomonoecious Silene vulgaris plant produce offspring with different genders: a case of heteroplasmic sex determination. Journal of Heredity 90, 563–565. Arrieta-Montiel, M., Lyznik, A., Woloszynska, M., Janska, H., Tohme, J. and Mackenzie, S. (2001) Tracing evolutionary and developmental implications of mitochondrial genomic shifting in the common bean. Genetics 158, 851–864. Backert, S. and Borner, T. (2000) Phage T4-like intermediates of DNA replication and recombination in the mitochondria of the higher plant Chenopodium album (L.). Current Genetics 37, 304–314. Backert, S., Lurz, R. and Borner, T. (1996) Electron microscopic investigation of mitochondrial DNA from Chenopodium album (L.). Current Genetics 29, 427–436. Bailey, M.F., Delph, L.F. and Lively, C.A. (2003) Modeling gynodioecy: novel scenarios for maintaining poly- morphism. American Naturalist 161, 762–776. Beardslee, T.A., Roy-Chowdhury, S., Jaiswal, P., Buhot, L., Lerbs-Mache, S., Stern, D.B. and Allison, L.A. (2002) A nucleus-encoded maize protein with sigma factor activity accumulates in mitochondria and chloroplasts. Plant Journal 31, 199–209. Bellaoui, M., Martin-Canadell, A., Pelletier, G. and Budar, F. (1998) Low-copy-number molecules are pro- duced by recombination, actively maintained and can be amplified in the mitochondrial genome of Brassicaceae: relationship to reversion of the male sterile phenotype in some cybrids. Molecular and General Genetics 257, 177–185. Bendich, A.J. (1996) Structural analysis of mitochondrial DNA molecules from fungi and plants using mov- ing pictures and pulsed-field gel electrophoresis. Journal of Molecular Biology 255, 564–588. Bentolila, S., Alfonso, A.A. and Hanson, M.R. (2002) A pentatricopeptide repeat-containing gene restores fertility to cytoplasmic male-sterile plants. Proceedings of the National Academy of Sciences USA 99, 10887–10892. Binder, S. and Brennicke, A. (2003) Gene expression in plant mitochondria: transcriptional and post-tran- scriptional control. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 358, 181–188. Birky, C.W. Jr (1983) Relaxed cellular controls and organelle heredity. Science 222, 468–475. Bjedov, I., Tenaillon, O., Gerard, B., Souza, V., Denamur, E., Radman, M., Taddei, F. and Matic, I. (2003) Stress-induced mutagenesis in bacteria. Science 300, 1404–1409. Black, D.L. (2003) Mechanisms of alternative pre-messenger RNA splicing. Annual Review of Biochemistry 72, 291–336. Blanc, G., Barakat, A., Guyot, R., Cooke, R. and Delseny, M. (2000) Extensive duplication and reshuffling in the Arabidopsis genome. Plant Cell 12, 1093–1101. Boutanaev, A.M., Kalmykova, A.I., Sheveloyov, Y.Y. and Nuriminsky, D.I. (2002) Large clusters of co- expressed genes in the Drosophila genome. Nature 420, 666–669. Brown, G.G., Formanova, N., Jin, H., Wargachuk, R., Dendy, C., Patil, P., Laforest, M., Zhang, J.F., Cheung, W.Y. and Landry, B.S. (2003) The radish Rfo restorer gene of Ogura cytoplasmic male sterility encodes a protein with multiple pentatricopeptide repeats. Plant Journal 35, 262–272. Chi, N.W. and Kolodner, R.D. (1994) Purification and characterization of MSH1, a yeast mitochondrial pro- tein that binds to DNA mismatches. Journal of Biological Chemistry 269, 29984–29992.

The mitochondrial genome of higher plants 77 Chopra, I., O’Neill, A.J. and Miller, K. (2003) The role of mutators in the emergence of antibiotic-resistant bacteria. Drug Resistance Updates 6, 137–145. Couvet, D., Atlan, A., Belhassen, E., Gliddon, C.J., Gouyon, P.H. and Kjellberg, F. (1990) Co-evolution between two symbionts: the case of cytoplasmic male-sterility in higher plants. In: Futuyama, D. and Antonovics, J. (eds) Oxford Survey in Evolutionary Biology, Vol. 7. Oxford University Press, Oxford, pp. 225–249. Covello, P.S. and Gray, M.W. (1992) Silent mitochondrial and active nuclear genes for subunit 2 of cytochrome c oxidase (Cox2) in soybean: evidence for RNA-mediated gene transfer. EMBO Journal 11, 3815–3820. Cuguen, J., Wattier, R., Saumitou-Laprade, P., Forcioli, D., Morchen, M., Van Dijk, H. and Vernet, P. (1994) Gynodioecy and mitochondrial DNA polymorphism in natural populations of Beta vulgaris ssp. mar- itima. Genetics Selection Evolution 26, 87s–101s. Delorme, V., Keen, C.L., Rai, K.N. and Leaver, C.J. (1997) Cytoplasmic-nuclear male sterility in pearl millet: comparative RFLP and transcript analyses of isonuclear male-sterile lines. Theoretical and Applied Genetics 95, 961–968. Desloire, S., Gherbi, H., Laloui, W., Marhadour, S., Clouet, V., Cattolico, L., Falentin, C., Giancola, S., Renard, M., Budar, F., Small, I., Caboche, M., Delourme, R. and Bendahmane, A. (2003) Identification of the fertility restoration locus, Rfo, in radish, as a member of the pentatricopeptide-repeat protein family. EMBO Reports 4, 588–594. Elo, A., Lyznik, A., Gonzalez, D.O., Kachman, S.D. and Mackenzie, S. (2003) Nuclear genes encoding mito- chondrial proteins for DNA and RNA metabolism are clustered in the Arabidopsis genome. Plant Cell 15, 1619–1631. Fujie, M., Kuroiwa, H., Kawano, S. and Kuroiwa, T. (1993) Studies on the behavior of organelles and their nucleoids in the root apical meristem of Arabidopsis thaliana (L.) Col. Planta 189, 443–452. Gepts, P. (1998) Origin and evolution of common bean: past events and recent trends. HortScience 33, 1124–1130. Gutierres, S., Sabar, M., Lelandais, G., Chetrit, P., Diolez, P., Degand, H., Boutry, M., Vedel, F., de Kouchkovsky, Y. and DePaepe, R. (1997) Lack of mitochondrial and nuclear-encoded subunits of com- plex I and alteration of the respiratory chain in Nicotiana sylvestris mitochondrial deletion mutants. Proceedings of the National Academy of Sciences USA 94, 3436–3441. Handa, H. (2003) The complete nucleotide sequence and RNA editing content of the mitochondrial genome of rapeseed (Brassica napus L.): comparative analysis of the mitochondrial genomes of rapeseed and Arabidopsis thaliana. Nucleic Acids Research 31, 5907–5916. Harfe, B.D. and Jinks-Robertson, S. (2000) DNA mismatch repair and genetic instability. Annual Review of Genetics 34, 359–399. Hawkesford, M.J. and Buchner, P. (eds) (2001) Molecular Analysis of Plant Adaptation to the Environment, 1st edn. Kluwer Academic Publishers, Dordrecht, The Netherlands. Hedtke, B., Borner, T. and Weihe, A. (1997) Mitochondrial and chloroplast phage-type RNA polymerases in Arabidopsis. Science 277, 809–811. Hervieu, F., Charbonnier, L., Bannerot, H. and Pelletier, G. (1993) The cytoplasmic male sterility (CMS) determinant of common bean is widespread in Phaseolus coccineus L. and Phaseolus vulgaris. L. Current Genetics 24, 149–155. Hoffmann, M., Kuhn, J., Daschner, K. and Binder, S. (2001) The RNA world of plant mitochondria. Nucleic Acid Research & Molecular Biology 70, 119–154. Jacobs, M.S. and Wade, M.J. (2003) A synthetic review of the theory of gynodioecy. American Naturalist 161, 837–851. Jan, C.C. (2000) Cytoplasmic male sterility in two wild Helianthus annuus L. accessions and their fertility restoration. Crop Science 40, 1535–1538. Janska, H., Sarria, R., Woloszynska, M., Arrieta-Montiel, M. and Mackenzie, S. (1998) Stoichiometric shifts in the common bean mitochondrial genome leading to male sterility and spontaneous reversion to fer- tility. Plant Cell 10, 1163–1180. Kanazawa, A., Tsutsumi, N. and Hirai, A. (1994) Reversible changes in the composition of the population of mtDNAs during dedifferentiation and regeneration in tobacco. Genetics 138, 865–870. Kaul, M.L.H. (1988) Male Sterility in Higher Plants. Springer, Berlin. Kazama, T. and Toriyama, K. (2003) A pentatricopeptide repeat-containing gene that promotes the process- ing of aberrant atp6 RNA of cytoplasmic male-sterile rice. FEBS Letters 544, 99–102.

78 S.A. Mackenzie Koizuka, N., Imai, R., Fujimoto, H., Hayakawa, T., Kumura, Y., Kohno-Murase, J., Sakai, T., Kawasaki, S. and Imamura, J. (2003) Genetic characterization of a pentatricopeptide repeat protein gene, orf687, that restores fertility in the cytoplasmic male-sterile Kosena radish. Plant Journal 34, 407–415. Lamppa, G.K. and Bendich, A.J. (1984) Changes in mitochondrial DNA levels during development of pea (Pisum sativum L.). Planta 162, 463–468. Laughnan, J.R. and Gabay-Laughnan, S. (1983) Cytoplasmic male sterility in maize. Annual Review of Genetics 17, 27–48. LeClerc, J.E. and Cebula, T.A. (1997) Hypermutability and homeologous recombination: ingredients for rapid evolution. Bulletin Institute Pasteur 95, 97–106. LeClerc, J.E., Li, B., Payne, W.L. and Cebula, T.A. (1996) High mutation frequencies among Escherichia coli and Salmonella pathogens. Science 274, 1208–1211. Lefai, E., Fernandez-Moreno, M.A., Kaguni, L.S. and Garesse, R. (2000) The highly compact structure of the mitochondrial DNA polymerase genomic region of Drosophila melanogaster: functional and evolu- tionary implications. Insect Molecular Biology 9, 315–322. Le Goff, S., Lachaume, P., Touraille, S. and Alziari, S. (2002) The nuclear genome of a Drosophila mutant strain increases the frequency of rearranged mitochondrial DNA molecules. Current Genetics 40, 345–354. Li, G.M. (1999) The role of mismatch repair in DNA damage-induced apoptosis. Oncology Research 11, 393–400. Mackenzie, S. (1991) Identification of a sterility-inducing cytoplasm in a fertile accession line of Phaseolus vulgaris L. Genetics 127, 411–416. Mackenzie, S. and Chase, C. (1990) Fertility restoration is associated with loss of a portion of the mitochon- drial genome in cytoplasmic male sterile common bean. Plant Cell 2, 905–912. Mackenzie, S. and McIntosh, L. (1999) Higher plant mitochondria. Plant Cell 11, 571–585. Mackenzie, S., Pring, D., Bassett, M. and Chase, C. (1988) Mitochondrial DNA rearrangement associated with fertility restoration and reversion to fertility in cytoplasmic male sterile Phaseolus vulgaris L. Proceedings of the National Academy of Sciences USA 85, 2714–2717. Manicacci, D., Atlan, A. and Couvet, D. (1997) Spatial structure of nuclear factors involved in sex determi- nation in the gynodioecious Thymus vulgaris L. Journal of Evolutionary Biology 10, 889–907. Marienfeld, J.R., Unseld, M., Brandt, P. and Brennicke, A. (1997) Mosaic open reading frames in the Arabidopsis thaliana mitochondrial genome. Biological Chemistry 378, 859–862. Marienfeld, J., Unseld, M. and Brennicke, A. (1999) The mitochondrial genome of Arabidopsis is composed of both native and immigrant information. Trends in Plant Science 4, 495–502. Martinez-Zapater, J., Gil, P., Capel, J. and Somerville, C. (1992) Mutations at the Arabidopsis CHM3 locus promote rearrangements of the mitochondrial genome. Plant Cell 4, 889–899. Modrich, P. and Lahue, R. (1996) Mismatch repair in replication fidelity, genetic recombination, and cancer biology. Annual Review of Biochemistry 65, 101–133. Moriyama, Y. and Kawano, S. (2003) Rapid, selective digestion of mitochondrial DNA in accordance with the matA hierarchy of multiallelic mating types in the mitochondrial inheritance of Physarum poly- cephalum. Genetics 164, 963–975. Nagata, N., Saito, C., Sakai, A., Kuroiwa, H. and Kuroiwa, T. (1999) The selective increase or decrease of organellar DNA in generative cells just after pollen mitosis one controls cytoplasmic inheritance. Planta 209, 53–65. Notsu, Y., Masood, S., Nishikawa, T., Kubo, N., Akiduki, G., Nakazono, M., Hirai, A. and Kadowaki, K. (2002) The complete sequence of the rice (Oryza sativa L.) mitochondrial genome: frequent DNA sequence acquisition and loss during the evolution of flowering plants. Molecular Genetics and Genomics 268, 434–445. Nugent, J.M. and Palmer, J.D. (1991) RNA-mediated transfer of the gene coxII from the mitochondrion to the nucleus during flowering plant evolution. Cell 66, 473–481. Oda, K., Yamato, K., Ohta, E., Nakamura, Y., Takemura, M., Nozato, N., Akashi, K., Kanegae, T., Ogura, Y., Kohchi, T. and Ohyama, K. (1992) Gene organization deduced from the complete sequence of liver- wort Marchantia polymorpha mitochondrial DNA: a primitive form of plant mitochondrial genome. Journal of Molecular Biology 223, 1–7. Oldenburg, D.J. and Bendich, A.J. (1996) Size and structure of replicating mitochondrial DNA in cultured tobacco cells. Plant Cell 8, 447–461. Olson, M.S. and McCauley, D.E. (2002) Mitochondrial DNA diversity, population structure, and gender association in the gynodioecious plant Silene vulgaris. Evolution 56, 253–262.

The mitochondrial genome of higher plants 79 Peeters, N. and Small, I. (2001) Dual targeting to mitochondria and chloroplasts. Biochimica et Biophysica Acta 1541, 54–63. Reboud, X. and Zeyl, C. (1994) Organelle inheritance in plants. Heredity 72, 132–140. Reenan, R.A.G. and Kolodner, R.D. (1992) Isolation and characterization of two Saccharomyces cerevisiae genes encoding homologs of the bacterial HexA and MutS mismatch repair proteins. Genetics 132, 963–973. Rieseberg, L.H., Baird, S.J.E. and Gardner, K.A. (2000) Hybridization, introgression, and linkage evolution. Plant Molecular Biology 42, 205–224. Robertson, E.J., Williams, M., Harwood, J.L., Lindsay, J.G., Leaver, C.J. and Leech, R.M. (1995) Mitochondria increase three-fold and mitochondrial proteins and lipid change dramatically in postmeristematic cells in young wheat leaves grown in elevated CO2. Plant Physiology 108, 469–474. Sakamoto, W., Kondo, H., Murata, M. and Motoyoshi, F. (1996) Altered mitochondrial genome expression in a maternal distorted leaf mutant of Arabidopsis induced by chloroplast mutator. Plant Cell 8, 1377–1390. Schnable, P.S. and Wise, R.P. (1998) The molecular basis of cytoplasmic male sterility and fertility restora- tion. Trends in Plant Science 3, 175–180. Small, I. and Peeters, N. (2000) The PPR motif, a TPR-related motif prevalent in plant organellar proteins. Trends in Biological Sciences 25, 46–47. Small, I.D., Isaac, P.G. and Leaver, C.J. (1987) Stoichiometric differences in DNA molecules containing the atpA gene suggest mechanisms for the generation of mitochondrial genome diversity in maize. EMBO Journal 6, 865–869. Small, I., Wintz, H., Akashi, K. and Mireau, H. (1998) Two birds with one stone: genes that encode products targeted to two or more compartments. Plant Molecular Biology 38, 265–277. Smith, R.L. and Chowdhury, M.K.U. (1991) Characterization of pearl millet mitochondrial DNA fragments rearranged by reversion from cytoplasmic male sterility to fertility. Theoretical and Applied Genetics 81, 793–799. Sodmergen, S., Zhang, Q.A., Zhang, Y.T., Sakamoto, W. and Kuroiwa, T. (2002) Reduction in amounts of mitochondrial DNA in the sperm cells as a mechanism for maternal inheritance in Hordeum vulgare. Planta 216, 235–244. Sutovsky, P. (2003) Ubiquitin-dependent proteolysis in mammalian spermatogenesis, fertilization, and sperm quality control: killing three birds with one stone. Microscopy Research and Technique 61, 88–102. Tang, H.V., Pring, D.R., Shaw, L.C., Salazar, R.A., Muza, F.R., Yan, B. and Schertz, K.F. (1996) Transcript pro- cessing internal to the mitochondrial open reading frame is correlated with fertility restoration in male- sterile sorghum. Plant Journal 10, 123–133. Tsaftaris, A.S. and Kafka, M. (1998) Mechanisms of heterosis in crop plants. Journal of Crop Production 1, 95–111. Unseld, M., Marienfeld, J.R., Brandt, P. and Brennicke, A. (1997) The mitochondrial genome of Arabidopsis thaliana contains 57 genes in 366,924 nucleotides. Nature Genetics 15, 57–61. Van Damme, J.M.M. (1983) Gynodioecy in Plantago lanceolata L. II. Inheritance of three male sterility types. Heredity 50, 253–273. Zhang, Q., Liu, Y. and Sodmergen, S. (2003) Examination of the cytoplasmic DNA in male reproductive cells to determine the potential for cytoplasmic inheritance in 295 angiosperm species. Plant and Cell Physiology 44, 941–951.

6 Reticulate evolution in higher plants Gay McKinnon School of Plant Science, University of Tasmania, Private Bag 55, Hobart, Tasmania 7001, Australia Introduction of a species’ gene pool. In others, reticula- tion may be destructive of diversity, causing Many evolutionary analyses focus exclu- the merging or assimilation of species. This sively on divergence. Evolution is seen as a chapter reviews evidence for reticulation in branching tree, from which new lineages higher plants and discusses its significance are continually splitting or budding. Most as an evolutionary mechanism. commonly used methods for calculating the evolutionary history of a group are Natural Hybridization in Plants designed to estimate the order in which dif- ferent genera or species diverged. Yet diver- In the plant kingdom, periods of genetic gence is only part of the evolutionary divergence followed by recombination take pattern. Particularly in the plant kingdom, place at various taxonomic levels, including populations and species may undergo the population and subspecies levels. repeated episodes of divergence, followed However, the term ‘reticulate evolution’ is by episodes of recombination. Through normally restricted to the description of hybridization between taxa, new genetic gene flow between taxonomically distinct combinations are formed. This process, species. If two species are able to cross suc- whereby branches of the evolutionary tree cessfully to form a first-generation (F1) exchange genes or are grafted together, is hybrid (Fig. 6.1a), a range of different con- called reticulate (net-like) evolution. sequences may ensue: The study of reticulate evolution is a 1. The F1 hybrid fails to reproduce, and challenging and dynamic area of research. forms an evolutionary dead-end. Methods for reconstructing reticulate evolu- 2. The F1 hybrid self-pollinates, or mates tionary pathways are only now being devel- with other F1 hybrids, to give second-gener- oped (e.g. Lapointe, 2000; Xu, 2000). These ation (F2) hybrid progeny. High genotypic methods should find wide application, as and phenotypic variability is expected in the new molecular studies confirm that reticula- F2 generation, because genes from the tion is both widespread and important to parental species can segregate into numer- evolution and diversity in higher plants. In ous different combinations (Fig. 6.1b). Some some cases reticulation is creative, leading to F2 progeny may closely resemble the F1 gen- enhanced diversity through the establish- eration or the original parental species. ment of new hybrid species or the expansion © CAB International 2005. Plant Diversity and Evolution: Genotypic and 81 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

82 G. McKinnon (a) Species 1 x Species 2 AA BB F1 hybrid aa bb Gene A Gene B Aa Bb Gene A Gene B (b) F1 hybrid x F1 hybrid Aa Bb F2 hybrid Aa Bb AA BB AA bb AA Bb aa BB aa bb aa Bb Aa BB Aa bb Aa Bb (c) Species 1 x F1 hybrid AA BB Aa Bb B1 back-cross AA BB AA Bb Aa BB Aa Bb Fig. 6.1. Possible genetic recombinations at two segregating nuclear loci (A and B) in the diploid progeny of two hybridizing diploid species (1 and 2). (a) First-generation hybrids; (b) second-generation hybrids; (c) back-crosses to species 1. 3. The F1 hybrid mates with members of lished, containing a mixture of F1 progeny, either or both parental species to produce advanced generation hybrids, back-crosses back-crossed (B1) progeny. Given that mem- and pure parental species. In some cases, bers of the parental species are often both such hybrid zones are both local and more numerous in the population and more ephemeral, and may have little evolutionary fertile than F1 hybrids, this is a likely scenario. significance. In others, hybridization can Back-crossed progeny also have a range of have far-reaching consequences including different genotypes and phenotypes, which the formation of new varieties, subspecies, may resemble the F1 generation or the origi- species or polyploid complexes. nal parental species (Fig. 6.1c). In some cases back-crossing occurs with one parental The frequency of natural hybridization species preferentially, biasing the genomic composition of subsequent generations. Natural hybridization is quite common in higher plants, but its frequency is not If numerous mating combinations are possi- ble, a complex hybrid zone may be estab-

Reticulate evolution in higher plants 83 evenly distributed among taxa. Certain transferred between species, successful fer- genera and families, as well as certain tilization requires pollen germination, pen- regions, show high hybridization frequen- etration of the stigma, growth of the pollen cies relative to others. Collated evidence on tube through the style, discharge of gametes the distribution of spontaneous plant with fertilization of the ovule, and the pro- hybrids from five floras (the British Isles, duction of viable seed. Incompatibility Scandinavia, the Hawaiian Islands, and the between species can affect any of these Great Plains and Intermountain West of stages, resulting in pollen tube arrest or North America) shows that within each embryo abortion. For example, hybridiza- flora, only 16 to 34% of families have any tion between Rhododendron species with dif- reported hybrids (Ellstrand et al., 1996). fering flower structures is restricted by Most of the hybrids are concentrated mismatches between style length and pollen within a few genera, which share certain tube length. Pollen tubes from species with characteristics that may promote hybrid much shorter or longer styles than the formation and persistence: outcrossing, recipient species either fail to penetrate the perennial habit and mechanisms for clonal ovary, or grow too far and fail to release reproduction (e.g. vegetative spread). their gametes (Williams et al., 1986; Families that show high hybrid frequencies Williams and Rouse, 1988). A complex com- in more than one flora are the Poaceae, bination of barriers to fertilization can co- Cyperaceae, Scrophulariaceae, Salicaceae, exist in a given species pair. In Penstemon, Rosaceae and Asteraceae. Hybridization is hybridization between the naturally sym- also common in the Betulaceae, Onagraceae, patric species Penstemon spectabilis and Orchidaceae and Pinaceae (Stace, 1975). Penstemon centranthifolius is limited by reduced pollen tube growth and seed set The failure to observe hybridization when P. spectabilis is the ovule parent, and within some plant families today does not by poor pollen grain germination and fruit mean that the evolutionary history of those set when P. centranthifolius is the ovule par- families has been free of reticulation. ent (Chari and Wilson, 2001). In addition, According to one estimate, about 70% of partial pollinator specificity helps to main- angiosperms have polyploidy in their history tain isolation between these species (Chari (Masterson, 1994). Since many polyploids and Wilson, 2001). have apparently arisen following hybridiza- tion, a major evolutionary role for Even when two species are reproduc- hybridization can be inferred for higher tively compatible, heterospecific (other- plants. species) pollen must generally overcome competition with conspecific (same-species) Barriers to hybridization pollen to fertilize the ovules. Such competi- tion is usually shown to favour conspecific What factors control the extent of hybridiza- pollen as the seed sire, for example in tion among plant taxa? Successful hybridiza- Hibiscus (Klips, 1999), Piriqueta (Wang and tion requires: (i) transfer of pollen between Cruzan, 1998) and Helianthus (Rieseberg et species; (ii) successful ovule fertilization and al., 1995a). The success of pollen competi- seed maturation; and (iii) ability of hybrid tion may be dependent on the direction of progeny to survive and attain reproductive the cross, leading to asymmetric hybridiza- maturity. tion. In crosses between Mimulus nasutus and Mimulus guttatus, pollen tubes from M. nasu- Barriers to one or more of these require- tus grow much more slowly than those of M. ments commonly operate to prevent gene guttatus in styles of M. guttatus, whereas flow between plant species. Pre-pollination pollen tubes from either species grow barriers include differences in flowering equally fast in styles of M. nasutus. As pre- time, adaptation to different pollinators, dicted, mixed pollen loads produce far more and ecological differences that prevent hybrid seed in M. nasutus than in M. guttatus species growing in sympatry. If pollen is (Diaz and Macnair, 1999).

84 G. McKinnon Hybrid fitness shows that hybrids are particularly common in disrupted habitats. For example, scarlet The fitness of hybrids, once formed, is vari- oak (Quercus coccinea) and black oak (Quercus velutina), both of which occur naturally in able. The new genetic combinations gener- the eastern USA, are normally reproduc- tively isolated by their preferences for moist ated by hybridization range from low areas and dry well-drained areas, respectively. Hybrids between the two advantageous to deleterious. First-generation species are common only when the natural environment has been disturbed by human hybrids often show superior growth or size, activity such as cutting and burning. In Rorippa, different types of habitat distur- termed hybrid vigour or heterosis, com- bance promote different patterns of hybridization (Bleeker and Hurka, 2001). In pared with parental species (particularly the River Elbe, which shows a natural dynamic of erosion leading to periodic habi- where the parents are inbred). However, tat disturbance, hybridization with bi-direc- tional introgression occurs between Rorippa their fertility may be low. This can result amphibia and Rorippa sylvestris; in artificial drainage ditches, hybridization occurs from chromosomal differences between the instead between R. amphibia and Rorippa palustris, with unidirectional introgression of parents that prevent successful pairing dur- genetic markers into R. amphibia. ing meiosis, or from unfavourable genic There are several ways in which habitat disturbance is likely to promote interspe- interactions, including cytonuclear interac- cific hybridization. The first is by altering species distributions, thereby creating new tions (reviewed in Burke and Arnold, 2001). mating opportunities. Upheavals caused by floods, fires, landslides, farming and road If hybrids succeed in reproducing, fitness building create corridors for dispersal that increase contact between species. The sec- may be reduced in later generations, ond is by the provision of novel or open environments in which hybrids are able to through the loss of favourably interacting establish themselves. While new hybrids are unlikely to be as well adapted as their par- gene complexes or the generation of lethal ents to the habitats in which those parents evolved, they may be successful in habitats gene combinations. This is referred to as that have been cleared of competitors and/or that contain new ecological niches. hybrid breakdown. Anderson (1948, 1949) attributed the pres- ence of numerous phenotypically variable Nevertheless, some hybrid combinations Iris hybrids in farmland on the Mississippi Delta to the rich variety of novel habitats equal or excel their parent species in fecun- created by land use. Later molecular analy- ses of Iris hybrid zones showed that certain dity or tolerance to environmental stresses. hybrid genotypes were indeed associated with novel habitats (Cruzan and Arnold, New, invasive lineages can evolve following 1993). A third, less commonly reported consequence of habitat disturbance is a hybridization, as documented in 12 differ- change in flowering synchrony between species. Lamont et al. (2003) showed that ent plant families by Ellstrand and hybrids between Banksia hookeriana and Schierenbeck (2000). Their success may be due to fixed heterosis and/or to evolution- ary novelty. Later-generation hybrids fre- quently demonstrate transgressive segregation: the production of phenotypes that are novel and extreme, rather than intermediate between their parents. Quantitative genetic studies show that this is chiefly due to the action of complementary genes (Rieseberg et al., 1999a). These novel phenotypes may contribute to evolutionary success. The invasive cordgrass, Spartina anglica, a hybrid derivative of Spartina mar- itima and Spartina alterniflora, is cited as a possible example of success through both fixed heterosis and evolutionary novelty (Ellstrand and Schierenbeck, 2000). The role of environment Environmental conditions are important in determining the frequency and results of hybridization. A large body of evidence

Reticulate evolution in higher plants 85 Banksia prionotes occurred in disturbed veg- thought that hybridization between pedun- etation as a result of a shift in flowering culate oak (Quercus robur) and sessile oak phenology. Affected populations showed (Quercus petraea) probably occurred at a time increased fecundity, earlier flowering of B. of low population size when the two species hookeriana and prolonged flowering of B. were confined to a glacial refuge (Ferris et prionotes, so that interspecific pollination al., 1993). Later post-glacial expansion became possible. Undisturbed interspersed could then have led to the two species shar- populations rarely coflowered and showed ing cpDNA haplotypes in recolonized ter- no hybridization. rain. However, a fine-scale analysis (Petit et al., 1997) showed that hybridization could Climate change and hybridization have played an important role in the recolo- nization process itself. Q. robur may have Climate change is a form of natural distur- acted as the pioneer species recolonizing bance that causes both species migrations new territories through seed dispersal, while and habitat transformations. The Q. petraea followed by pollinating established Quaternary Ice Ages led to cyclic expansion populations of Q. robur and ultimately and contraction of many species’ ranges in replacing it through back-crossing. response to changes in aridity and tempera- ture (Hewitt, 1999, 2000). Alpine plants Forms of Reticulate Evolution descended and ascended mountains, while lowland species underwent major geograph- The term ‘reticulate evolution’ embraces a ical redistributions. Worldwide, the lowering range of evolutionary outcomes which may of sea levels linked landmasses periodically, follow hybridization between species. These allowing contact between different floras. In depend on many factors such as the degree some cases, these changes apparently led to of genetic compatibility between the hybridization. For instance, chloroplast (cp) parental species, ecological preferences of DNA diversity in northern populations of the parents and hybrids, and the fitness and Packera pseudaurea in Alberta suggests former fertility of hybrids. Although intuitively it hybridization with other Packera species would seem that low levels of hybrid forma- which migrated southwards during periods tion are unlikely to attain evolutionary sig- of glaciation (Yates et al., 1999; Golden and nificance, this is not always the case. Many Bain, 2000). In south-east Spain, species of species which have quite low interfertility Armeria, which now occur at different alti- nevertheless form stable and persistent tudes, may have hybridized when in tempo- hybrid zones, or give rise to hybrid lineages. rary sympatry (Larena et al., 2002). On the For example, the formation of F1 hybrid island of Tasmania, cpDNA patterns in seed between the North American milk- Eucalyptus are consistent with hybridization weeds Asclepias exaltata and Asclepias syriaca is between local endemics and species which a rare event, yet natural hybrids occur migrated into Tasmania from mainland throughout areas of sympatry, and act as Australia (McKinnon et al., 2004). bridges for interspecific gene flow through back-crossing (Broyles, 2002). Only 5.6% of One of the best-studied examples of a pollen is viable in F1 hybrids between species complex that was redistributed dur- Helianthus annuus and Helianthus petiolaris, ing the Ice Ages is that of the oaks. Pollen yet over 90% fertility is regained after only evidence and molecular markers show that four generations of sib-mating or back-cross- oaks have recolonized Europe from refugia ing (Ungerer et al., 1998). In nature, these in Iberia, Italy and the Balkans since the two species commonly form hybrid zones Last Glacial Maximum. A systematic sharing and are believed to be the progenitors of of local cpDNA markers, revealing wide- three different hybrid species: Helianthus spread hybridization, has been found in anomalus, Helianthus deserticola and Helianthus seven species throughout Europe (Dumolin- paradoxus (Rieseberg, 1991). Lapègue et al., 1997). Initially, it was

86 G. McKinnon The three major forms of reticulate evo- Ecological divergence between the hybrid lution, which will be considered in more species and its parents is also likely to rein- detail below, are: force genetic isolation of the hybrid. 1. The formation of new species, termed Studying hybrid speciation hybrid speciation. 2. The transfer of genes between species, Studies of hybrid speciation fall into two termed introgression. broad categories. The first is the investiga- 3. Merging of species or extinction by tion of naturally occurring hybrid species. assimilation. Modern molecular techniques make it possi- ble to confirm or disprove hypotheses of Hybrid speciation hybrid speciation, since hybrid species are expected to carry a combination of genetic There is little doubt that hybrid speciation is markers from their putative parents. In of major evolutionary importance in the some cases, hybrid speciation is discovered plant kingdom. The potential role of during phylogenetic analysis of a species hybridization in plant speciation was recog- complex. For instance, the sunflower nized by early botanists such as Linnaeus Helianthus anomalus was found to combine and Mendel, and present-day understand- the chloroplast and nuclear ribosomal mark- ing of hybrid speciation is based on over two ers of two species, H. annuus and H. centuries’ worth of investigation (reviewed petiolaris, suggesting a hybrid origin in Rieseberg, 1997). The current view is that (Rieseberg, 1991). Matching evidence from some recognized ‘species’ have arisen more multiple, unlinked genetic markers provides than once through separate hybridization the strongest support for a hypothesis of events. Furthermore, in some genera, hybrid speciation. Many studies now employ hybridization between the same parental combined evidence from cytological analysis, taxa has given rise to different hybrid nuclear markers (e.g. allozymes, nuclear species. Certain hybrid taxa are derived ribosomal DNA, low-copy nuclear genes, from more than two parental species; for random amplified polymorphic DNA) example, Iris nelsonii is apparently a deriva- and/or chloroplast markers. tive of three species: Iris hexagona, Iris fulva and Iris breviligulata (Arnold, 1993). These The second category is the manipulation findings illustrate the complexity of species of experimental hybrid lineages to deter- relationships in plants and the inadequacy mine the mechanisms governing hybrid spe- of simple bifurcating phylogenies to describe ciation. Mapped molecular markers have such relationships. been used to study gene segregation in syn- thetic hybrids, with remarkable results. The formation of a new species through Rieseberg et al. (1996) used 197 markers hybridization requires the development of a covering the sunflower genome to investi- reproductive barrier between the newly gate three experimentally created hybrid formed hybrid lineage and its parents. lineages between the sunflowers H. annuus Without such a barrier, the new lineage will and H. petiolaris. If chance governed the seg- be swamped by gene flow with one or both regation of markers in later generation parental species, particularly during the hybrids, independent hybrid lineages would early stages when the parents and hybrids have widely differing genotypes. are sympatric. For this reason, hybrid speci- Astonishingly, however, all three lineages ation often involves a change in ploidy, or were very similar in genomic composition to some other form of chromosomal or genic each other and to the naturally occurring incompatibility between a hybrid and its hybrid species H. anomalus. The fact that parents. The ability to reproduce clonally, almost identical lineages can be reproduced for instance through apomixis, may help to by independent hybridization events shows stabilize a newly formed hybrid lineage. that selection, rather than chance, governs the genomic composition of hybrid species.

Reticulate evolution in higher plants 87 This research lends support to the con- times, respectively, in 70 years. Polyploid tention that many hybrid species have arisen species of Draba and Saxifraga may also have repeatedly. had multiple origins from diploid progeni- tors (Brochmann and Elven, 1992; Hybrid speciation with an increase in ploidy Brochmann et al., 1998). Polyploidy is a common mode of speciation in Since most parental species will exhibit plants, and in many cases arises following some genetic variation across their natural hybridization between genetically differenti- geographic ranges, polyploid species arising ated species whose chromosomes are too dis- from separate hybridization events in differ- similar to pair correctly during meiosis. ent localities will constitute a series of geneti- Meiotic failure in the F1 hybrid leads to the cally differentiated populations. Gene flow production of unreduced (2n) gametes. The may then occur between the different poly- union of two unreduced 2n gametes gives rise ploid populations, creating even greater to a 4n (tetraploid) zygote, termed an allote- genetic variability. This variability will be traploid. Alternatively, union of a 2n gamete further enhanced by different chromosomal with a normal haploid gamete gives rise to a rearrangements arising in different popula- triploid (3n) offspring, which may produce tions following polyploidy. Recent evidence triploid gametes; these can then unite with shows that allopolyploids undergo extensive haploid gametes to produce tetraploid prog- and rapid genomic reshuffling after forma- eny. The latter mechanism was demonstrated tion (Soltis and Soltis, 1999). In Brassica, by Müntzing in crosses of mint (Galeopsis extensive genetic and phenotypic diversity pubescens ϫ Galeopsis speciosa) as long ago as developed after only a few generations in 1930. Higher level allopolyploids, which com- experimentally created allopolyploids (Song bine three or more genomes, may also arise et al., 1995). Thus, polyploid hybrid specia- following hybridization. tion represents a particularly dynamic form of reticulate evolution. Speciation by polyploidy is a form of instantaneous, sympatric speciation. The Homoploid hybrid speciation new allopolyploid is often fully fertile, and fully or partially reproductively isolated Hybridization can also give rise to new from its nearest relatives. Polyploid specia- species with the same ploidy as the parental tion may give rise to polyploid series (with species. Molecular studies have confirmed multiples of the basic chromosome number, the natural occurrence of homoploid hybrid as for instance in Chrysanthemum) or com- speciation in Stephanomeria (Gallez and plexes in which species with different basic Gottlieb, 1982), Helianthus (Rieseberg, 1991), chromosome numbers have hybridized and Iris (Arnold, 1993), Pinus (Wang and Szmidt, become polyploid (for instance Clarkia; 1994; Wang et al., 2001), Paeonia (Sang et al., Lewis and Lewis, 1955). Current research 1995) and Penstemon (Wolfe et al., 1998). indicates that many polyploid species have Evidence from Paeonia (Sang et al., 1995) arisen recurrently, contradicting the princi- suggests that homoploid hybrid species in ple that biological species have a unique, some cases may go on to found speciose lin- monophyletic origin (Soltis and Soltis, eages. Homoploid hybrid speciation is often 1999). In fact, multiple origins for polyploid reported for diploids, but may also occur species may be the rule rather than the naturally between allotetraploids without an exception. Molecular studies on the intermediate stage of genome diploidization Tragopogon tetraploids, Tragopogon miscellus or a further increase in ploidy (Ferguson and Tragopogon mirus, indicate that spread of and Sang, 2001). each species is occurring not through dis- persal from a single origin but through How does a hybrid develop reproductive repeated instances of recreation (Soltis et al., isolation from its parental species and found 1995; Cook et al., 1998). These two species a new lineage without a change in ploidy? may have formed as often as 20 and 12 Stebbins (1950) proposed a mechanism, later termed ‘recombinational speciation’

88 G. McKinnon (Grant, 1971), involving parental species dif- H. anomalus (Rieseberg, 1991), but occur in fering by two or more separable chromoso- different habitats, suggesting a role for eco- mal rearrangements. In theory, two such logical selection. species could found a new hybrid lineage through the following sequence of events: Introgression 1. The F1 hybrid between the two parental Another common consequence of hybridiza- species is formed but has low fertility, since tion between species is introgression, the most of its gametes carry unbalanced chro- infiltration of genes from one species into mosome complements. the gene pool of the other. F1 hybrids, once 2. Some offspring of the F1 hybrid, through formed, act as a bridge to gene flow through chance segregation and recombination, are back-crossing to either or both of the homozygous for balanced chromosome com- parental species. Introgression may be uni- plements. directional, with genes flowing into only one 3. These offspring are fertile, but reproduc- of the species involved, or bidirectional. Its tively isolated from other such homozygotes nature depends on a complex combination and from the parental species, enabling the of factors, including mating patterns and establishment of a new lineage. chromosomal and genic incompatibilities. A distinction is drawn between localized intro- The mechanism of recombinational specia- gression, which refers to the exchange of tion has been verified experimentally in a genetic markers within an obvious hybrid number of genera, including Nicotiana zone, and dispersed introgression, which (Smith and Daly, 1959) and Gilia (Grant, refers to the flow of genes from one species 1966). More general models for homoploid into another at a distance from the hybrid hybrid speciation have since been pro- zone. Dispersed introgression may be due posed (e.g. Templeton, 1981). The modern to: (i) flow of introgressed genes across a view is that a variety of mechanisms population through pollen dispersal; (ii) including both chromosomal and genic seed dispersal of progeny carrying intro- incompatibility, and ecological divergence gressed genes; or (iii) the movement or dis- and selection for hybrids, can promote this appearance over time of a hybrid zone, form of speciation. leaving behind introgressed individuals. Recent studies have shown that, like Introgressive hybridization has been pro- allopolyploids, diploid hybrid species may posed as an important mechanism leading arise recurrently. Schwarzbach and to race formation in plant groups including Rieseberg (2002) deduced from chloroplast Pinus, Abies, Quercus, Purshia, Cistus, DNA and crossability data that the diploid Coprosma, Dracophyllum, Helianthus, Gilia and hybrid sunflower species H. anomalus proba- Tradescantia (Stebbins, 1950; Grant, 1971). bly arose on three occasions independently Molecular evidence confirms that certain from crosses between its parental species, H. intraspecific taxa such as the sunflower sub- annuus and H. petiolaris. Recurrent diploid species, H. annuus ssp. texanus, and the speciation has also been suggested for Pinus groundsel variant, Senecio vulgaris var. hiber- densata (Wang et al., 2001) and nicus, have arisen by introgression (reviewed Argyranthemum sundingii (Brochmann et al., in Abbott, 1992). In theory, introgression 2000). In addition, different diploid hybrid should increase the genetic diversity of a species can arise naturally from the same species and allow it to occupy new habitats cross. For example, Argyranthemum lemsiii through the capture or development of use- and A. sundingii share the parental species ful adaptations. Such increased genetic Argyranthemum frutescens and Argyranthemum diversity has been demonstrated in species broussonetii, but are derived by crosses in of Cypripedium (Klier et al., 1991) and opposite directions and show different chro- Aesculus (dePamphilis and Wyatt, 1990). mosomal rearrangements (Borgen et al., However, the role of introgression as a 2003). The sunflowers H. deserticola and H. paradoxus have the same parental species as

Reticulate evolution in higher plants 89 means for transferring or creating beneficial combined with the cytoplasmic markers of adaptations remains difficult to prove B. Repeated generations of back-crossing in through direct evidence. A possible example the same direction will dilute the nuclear is that of Rhododendron ponticum in the markers of B until they are difficult to British Isles. Combined molecular and bio- detect, but cannot remove the cytoplasmic geographic evidence suggest that this markers of B. Another possible reason for species may have acquired enhanced cold higher levels of cytoplasmic marker intro- tolerance through introgressive hybridiza- gression is selection against alien nuclear tion with Rhododendron catawbiense (Milne genes, but not alien cytoplasmic genes. and Abbott, 2000). The most exciting recent research on Studying introgression introgression uses molecular markers which have been mapped to different chromoso- Introgressive hybridization is sometimes dif- mal regions. These mapped markers allow ficult to distinguish from other processes tracking of the movement of chromosomal which give rise to similar patterns of mor- segments between species. A detailed study phological variation. Individuals that are by Martinsen et al. (2001) of contemporary morphologically intermediate between two recognized species might have arisen by nuclear nuclear hybridization. Alternatively, they might be genome A genome B remnants of an ancestral population from which the two species arose, or members of cpA mtA cpB mtB different species converging in morphology under natural selection. In addition, AB advanced generation hybrids sometimes resemble their parental species so strongly cpB mtB that their hybrid nature goes undetected. For this reason molecular markers are AB widely applied to the study of introgression. Evidence from cytoplasmic (chloroplast, cp, cpB mtB or mitochondrial, mt) DNA markers is com- monly used in combination with evidence A from nuclear genomic markers. The latter include allozymes, random amplified poly- cpB mtB morphic DNA (RAPDs), ribosomal DNA (rDNA) sequences, microsatellites and Fig. 6.2. Unidirectional introgression of maternally restriction fragment length polymorphisms inherited cp and mtDNA following hybridization (RFLPs). and back-crossing between two species, A and B. Species A acts as the pollen parent in all crosses. Recent studies have shown that introgres- First-generation hybrids carry the combined nuclear sion can be remarkably selective. Typically, genes of A and B, with the cp and mtDNA of B. cytoplasmic markers such as cpDNA are Successive back-crosses to pollen parent A dilute exchanged far more readily than nuclear the nuclear genomic contribution of B, but can markers (Rieseberg and Soltis, 1991). This is never erase the cp and mtDNA of B. due partly to their uniparental mode of inheritance (Fig. 6.2). In most (although not all) flowering plants, both mitochondria and chloroplasts are inherited from the maternal parent. An F1 hybrid between two species, A (male) and B (female) therefore inherits mtDNA and cpDNA only from B. If the hybrid is pollinated by A, its progeny will carry nuclear markers characteristic of A,

90 G. McKinnon unidirectional introgression between A follow-up study (Rieseberg et al., Fremont cottonwood (Populus fremontii) and 1999b) used three sympatric populations of the higher-altitude species, narrowleaf cot- H. petiolaris and H. annuus to investigate tonwood (Populus angustifolia), along the introgression in natural hybrid zones. Weber River in Utah, used 35 genetically Patterns of introgression were similar to mapped RFLP markers. These markers those seen in experimental hybrids, showed that the majority of the nuclear although with greater recombination and genome was not exchanged between introgression across rearranged parts of the species. However, a small percentage (21%) genome. They were also remarkably consis- of nuclear markers from Fremont cotton- tent across the three different hybrid zones, wood was able to introgress into ‘pure’ nar- suggesting that chromosomal segments were rowleaf cottonwood. Different markers under similar selective regimes in different showed different levels of introgression. populations. Many of the chromosomal seg- Some were found in narrowleaf cottonwood ments that failed to introgress were associ- only short distances from the 13 km long ated with reduced pollen fertility, which hybrid zone, but others showed dispersed would create a selective disadvantage in introgression up to 100 km from the hybrid hybrids. These studies show that, like hybrid zone. This pattern may have arisen when a speciation, introgression is a non-random hybrid zone between the two species moved process that can produce similar patterns of downhill gradually in response to climate genetic variation in separate locations. change. The authors suggested that hybrids could act as evolutionary filters that allow Introgression and phylogenetic incongruence introgression of beneficial genes between species, while preventing the transfer of The ability of some species to capture genes deleterious genes. from others by introgressive hybridization has important consequences for molecular phylo- In Helianthus, the genetic architecture of genetic analysis in plants. Until quite recently, barriers to introgression of nuclear genes is evolutionary relationships among plant now being uncovered. The species H. annuus species were often estimated by phylogenetic and H. petiolaris, described above as the analysis of cpDNA sequences, using one or a progenitors of three different hybrid few individuals to represent each species. species, also demonstrate localized intro- More extensive sampling has now shown that gression. Both species are diploid (n = 17) many plant species carry multiple cpDNA lin- but only seven of their chromosomes are co- eages, and that these lineages are not always linear; the remaining ten differ by a number species-specific. In some cases, this is due to a of translocations and inversions. Rieseberg et phenomenon called incomplete lineage sort- al. (1995b) used mapped RAPD markers to ing (Fig. 6.3a). Under lineage sorting, the study the introgression of nuclear genomic ancestor to a group of species carries multiple segments from H. petiolaris into H. annuus in divergent cpDNA lineages (A, B, C, D), all or experimental hybrids. They found that some of which are passed on to each descen- chromosomal rearrangements acted as a dant species. These lineages are subject to strong barrier to introgression. Only 2.4% of random drift in the daughter species, so that the rearranged portion of the genome intro- by chance some species will retain only A, gressed, whereas 40% of the co-linear por- some retain A and D, others retain B and C, tion was able to introgress. For both and so on. As a result, the cpDNA phylogeny portions of the genome, marker introgres- may not accurately reflect the species relation- sion was significantly less than would be pre- ships. Introgression of cpDNA following dicted by chance, although a few markers hybridization between species (Fig. 6.3b) can introgressed at higher than expected fre- also create a pattern of shared lineages that quencies. Thus, both chromosomal obscures the real species relationships. The rearrangements and selection against cer- same principles apply to nuclear gene phylo- tain H. petiolaris genes appeared to be limit- genies. As a result, phylogenies generated ing introgression.

Reticulate evolution in higher plants 91 Species 123 4 Time of sampling (a) Lineage sorting Species tree Ancestral population Species 1234 (b) Introgression cpDNA lineage A cpDNA lineage B Fig. 6.3. Two situations in which a gene phylogeny does not accurately reflect the species phylogeny. Both situations may occur together. (a) Lineage sorting. The ancestral population to species 1–4 carries two different cpDNA lineages, A and B. By chance, B is eliminated from species 1 and 4, and A is eliminated from species 2. From the cpDNA relationships, species 1 appears more closely related to species 4 than to species 2. (b) Introgression. Species 2 has acquired lineage B through hybridization with species 3. At the time of sampling, species 2 appears more closely related to species 3 than to species 1. using cp and nuclear DNA sequences may be 1. Under lineage sorting, enough time in conflict with one another, or with phyloge- may have elapsed since speciation for fur- nies based on morphological characters. ther sequence divergence within lineages A to D. Species which inherited B from their How can phylogenetic incongruence common ancestor will therefore carry caused by introgression be distinguished somewhat divergent copies of B. By con- from incongruence due to lineage sorting? A trast, recent introgression will result in dif- number of criteria are helpful in separating ferent species carrying identical copies of the two processes:

92 G. McKinnon B. The more ancient the introgression, the Merging of species and genetic assimilation more difficult it will be to distinguish it from lineage sorting by this criterion. An extreme consequence of introgressive Separation of the two phenomena is com- hybridization is the merging of species at plicated by the fact that hybridization one or more sympatric populations. In within some genera is a continuous process. plants with short generation times, this can According to fossil analysis, hybridization happen rapidly. Carney et al. (2000a) stud- between species of Populus has been occur- ied the change in genetic and morphological ring for at least 12 million years composition of a hybridizing population of (Eckenwalder, 1984). This is likely to gener- the sunflowers Helianthus bolanderi and H. ate a complex pattern resulting from both annuus after 50 years. They found that few recent and ancient gene flow between genetically pure parental plants remained in species. the population. The average phenotype had 2. Assuming that the marker in question is shifted in bias from H. bolanderi to H. annuus not under selection, lineage sorting is in this time. The trend was towards assimila- unlikely to give a matching geographical tion of H. bolanderi in this population, and pattern of genetic variation across species. potentially others throughout its range. A pattern of shared markers between two species only in regions of sympatry is Genetic assimilation is presently receiving therefore more likely to result from intro- attention because of its implications for the gression. This is particularly the case if conservation of rare species. Hybrids formed multiple markers are shared in the region between a rare species and a more abundant of sympatry. congener may contribute to the demise of the rare species by replacing its conspecific Phylogenetic incongruence has led to the (pure) progeny with increasing numbers of discovery of unsuspected, historical intro- hybrid and back-crossed progeny in each gression in higher plant genera. Wendel et generation. One well-known example is that al. (1995) ‘stumbled across’ a case of of the Catalina Island mahogany, Cercocarpus ancient, cryptic introgression in Gossypium traskiae, whose population size has shrunk to during phylogenetic analysis of the nuclear about six ‘pure’ individuals (Rieseberg and rDNA. Sequence data for the rDNA of Gerber, 1995). This species has declined in American Gossypium gossypioides placed it in number over the last century through over- the same clade as African species of grazing, but is also apparently under threat Gossypium, conflicting with evidence from through hybridization with its more abun- fertility relationships, cytogenetics, mor- dant congener Cercocarpus betuloides var. phology, allozymes and cpDNA. They con- blancheae. RAPD data show that in addition cluded that an ancient hybridization event to the six pure Cercocarpus traskiae individu- must have taken place between species that als, five adult hybrid trees and at least five now occupy different hemispheres. Comes seedlings of hybrid origin are present in the and Abbott (1999) found that, for two surviving population. species of Senecio, both rDNA and cpDNA evidence conflicted with morphological Island plants appear to be particularly classification. The most likely explanation susceptible to genetic assimilation through was historical capture of both rDNA and hybridization. This is due to their small pop- cpDNA, following introgressive hybridiza- ulation size, the likelihood of invasion of tion among species. In one of the two their habitat by related congeners from species, this former capture was quite other landmasses, and sometimes a lack of undetectable by RAPD profile or morphol- strong reproductive barriers between ogy. An increasing number of such discov- species because of unspecialized pollinators eries shows that extensive sampling of and/or relatively recent divergence (for species and markers is wise when analysing example, following adaptive radiations). In evolutionary histories. the Canary Islands, the rare endemic Argyranthemum coronopifolium is undergoing assimilation by hybridization with a wide-

Reticulate evolution in higher plants 93 spread weed, A. frutescens (Levin et al., ing genes that raise its fitness. Hybrid popu- 1996). Following an encounter between the lations could act as a genetic reservoir for two species, a population of A. coronopifolium reconstituting the parental genotypes under was gradually swamped by hybrids and A. favourable conditions (Anderson, 1949). frutescens over a period of 30 years. Numerous other examples of hybridization Summary between rare species and abundant con- geners have been documented in the British 1. Reticulation (hybridization between Isles (e.g. Saxifraga, Salix, Sorbus) and the divergent taxa) contributes to biodiversity in Hawaiian Islands (e.g. Cyrtandra, Dubautia; higher plants through the creation of new reviewed in Levin et al., 1996, and Carney et hybrid lineages and the transfer of genes al., 2000b). between species. The frequency of hybridization varies greatly between plant Factors such as absolute and relative pop- genera, and some large plant families have ulation sizes, geographical proximity, rates of few reported hybrids. Nevertheless, it is esti- hybridization and the fitness of hybrids are mated that most flowering plants have all likely to influence the rate of genetic episodes of reticulation in their evolutionary assimilation. The typical scenario is that of a history. rare species undergoing assimilation by a 2. Natural hybridization between species is more abundant invader. However, studies of controlled by numerous barriers to gene Spartina (Anttila et al., 1998) show that even flow. The frequency and success of hybridiza- an abundant species may be threatened by tion depend on the compatibility of the par- serial hybridization with a small population ent species and on environmental conditions. of an invader that produces large quantities Natural or anthropogenic disturbances of of superior pollen. The evolutionary conse- the environment can promote hybridization quences for the assimilating species are by bringing allopatric species together, creat- rarely considered, but must include the ing new habitats that favour hybrids, and/or acquisition of new genetic variability. In cases changing flowering synchrony. where a species has been completely assimi- 3. Even low levels of hybrid formation can lated, this variability may appear to have have a significant evolutionary outcome. arisen within the assimilating species, when Some species with low inter-fertility never- in fact it has been acquired through reticula- theless form persistent hybrid zones, and are tion. Harlan and deWet (1963) proposed the known to have founded new hybrid lineages. term ‘compilospecies’ to describe an aggres- 4. Hybrid speciation is of major importance sive species that plunders the gene pools of in the plant kingdom and can occur with or congeners, thereby increasing its own ecolog- without a change in ploidy. Recent research ical tolerance and geographic range. shows that many hybrid species have multi- ple origins. The genomic composition of Paradoxically, it has been suggested that hybrid species is not governed by chance hybridization might be one way to conserve segregation, but by selection for certain gene genes from extremely rare or threatened combinations. species. When only a few individuals of a 5. Introgressive hybridization contributes to species remain, inbreeding is likely to genetic variability, and probably adaptability, become catastrophic. Hybridization with a within established species. The exchange of congener may produce healthy progeny that genes between species by introgression is can propagate the genes of the endangered remarkably selective. Cytoplasmic genes are species. This method has been used to pre- exchanged more readily than nuclear genes, serve the genes of the St Helena redwood and some nuclear genes are exchanged (Trochetiopsis erythroxylon) and the St Helena more readily than others. Like hybrid speci- ebony (Trochetiopsis ebenus). The two species ation, introgression is apparently directed are almost extinct, but have been crossed to by selection. produce vigorous hybrids (Cronk, 1995). In theory, natural hybridization could therefore enrich an endangered species by contribut-

94 G. McKinnon 6. Hybridization can lead to extinction by a simple, branching evolutionary pathway. assimilation in rare species. Introgression, hybrid speciation and multi- 7. The study of reticulate evolution is ple origins for hybrid species do not fit demanding, and requires data of high within this assumption. The importance of quality and quantity. Reticulation is prob- reticulation is becoming increasingly lematic for phylogenetic reconstruction. apparent, and future phylogenetic analyses Most methods for phylogenetic analysis are likely to place more emphasis on this assume that species have evolved through important feature of plant evolution. References Abbott, R.J. (1992) Plant invasions, interspecific hybridization and the evolution of new plant taxa. Trends in Ecology and Evolution 7, 401–405. Anderson, E. (1948) Hybridization of the habitat. Evolution 2, 1–9. Anderson, E. (1949) Introgressive Hybridization. John Wiley & Sons, New York. Anttila, C.K., Daehler, C.C., Rank, N.E. and Strong, D.R. (1998) Greater male fitness of a rare invader (Spartina alterniflora, Poaceae) threatens a common native (Spartina foliosa) with hybridization. American Journal of Botany 85, 1597–1601. Arnold, M.L. (1993) Iris nelsonii (Iridaceae); origin and genetic composition of a homoploid hybrid species. American Journal of Botany 80, 577–583. Bleeker, W. and Hurka, H. (2001) Introgressive hybridization in Rorippa (Brassicaceae): gene flow and its consequences in natural and anthropogenic habitats. Molecular Ecology 10, 2013–2022. Borgen, L., Leitch, I. and Santos-Guerra, A.T.I. (2003) Genome organization in diploid hybrid species of Argyranthemum (Asteraceae) in the Canary Islands. Botanical Journal of the Linnaean Society 141, 491–501. Brochmann, C. and Elven, R. (1992) Ecological and genetic consequences of polyploidy in arctic Draba (Brassicaceae). Evolutionary Trends in Plants 6, 111–124. Brochmann, C., Xiang, Q.Y., Brunsfeld, S.J., Soltis, D.E. and Soltis, P.S. (1998) Molecular evidence for poly- ploid origins in Saxifraga (Saxifragaceae): the narrow arctic endemic S. svalbardensis and its wide- spread allies. American Journal of Botany 85, 135–143. Brochmann, C., Borgen, L. and Stabbetorp, O.E. (2000) Multiple diploid hybrid speciation of the Canary Island endemic Argyranthemum sundingii (Asteraceae). Plant Systematics and Evolution 220, 77–92. Broyles, S.B. (2002) Hybrid bridges to gene flow: a case study in milkweeds (Asclepias). Evolution 56, 1943–1953. Burke, J.M. and Arnold, M.L. (2001) Genetics and the fitness of hybrids. Annual Review of Genetics 35, 31–52. Carney, S.E., Gardner, K.A. and Rieseberg, L.H. (2000a) Evolutionary changes over the fifty-year history of a hybrid population of sunflowers (Helianthus). Evolution 54, 462–474. Carney, S.E., Wolf, D.E. and Rieseberg, L.H. (2000b) Hybridization and forest conservation. In: Young, A., Boshier, D. and Boyle, T. (eds) Forest Conservation Genetics: Principles and Practice. CSIRO Publishing, Collingwood, Australia, pp.167–182. Chari, J. and Wilson, P. (2001) Factors limiting hybridization between Penstemon spectabilis and Penstemon centranthifolius. Canadian Journal of Botany – Revue Cananadienne de Botanique 79, 1439–1448. Comes, H.P. and Abbott, R.J. (1999) Reticulate evolution in the Mediterranean species complex of Senecio sect. Senecio: uniting phylogenetic and population-level approaches. In: Hollingsworth, P.M., Bateman, R.M. and Gornall, R.J. (eds) Molecular Systematics and Plant Evolution. Taylor & Francis, London, pp. 171–198. Cook, L.M., Soltis, P.S., Brunsfeld, S.J. and Soltis, D.E. (1998) Multiple independent formations of Tragopogon tetraploids (Asteraceae): evidence from RAPD markers. Molecular Ecology 7, 1293–1302. Cronk, Q.C.B. (1995) A new species and hybrid in the St. Helena endemic genus Trochetiopsis. Edinburgh Journal of Botany 52, 205–213. Cruzan, M.B. and Arnold, M.L. (1993) Ecological and genetic associations in an Iris hybrid zone. Evolution 47, 1432–1445. dePamphilis, C.W. and Wyatt, R. (1990) Electrophoretic confirmation of interspecific hybridization in Aesculus (Hippocastanaceae) and the genetic structure of a broad hybrid zone. Evolution 44, 1295–1317.

Reticulate evolution in higher plants 95 Diaz, A. and Macnair, M.R. (1999) Pollen tube competition as a mechanism of prezygotic reproductive isola- tion between Mimulus nasutus and its presumed progenitor M. guttatus. New Phytologist 144, 471–478. Dumolin-Lapègue, S., Demesure, B., Fineschi, S., Le Corre, V. and Petit, R.J. (1997) Phylogeographic struc- ture of white oaks throughout the European continent. Genetics 146, 1475–1487. Eckenwalder, J.E. (1984) Natural intersectional hybridization between North American species of Populus (Salicaceae) in sections Aiegeiros and Tacamahaca. III. Paleobotany and evolution. Canadian Journal of Botany 62, 336–342. Ellstrand, N.C. and Schierenbeck, K.A. (2000) Hybridization as a stimulus for the evolution of invasiveness in plants? Proceedings of the National Academy of Sciences USA 97, 7043–7050. Ellstrand, N.C., Whitkus, R. and Rieseberg, L.H. (1996) Distribution of spontaneous plant hybrids. Proceedings of the National Academy of Sciences USA 93, 5090–5093. Ferguson, D. and Sang, T.T.I. (2001) Speciation through homoploid hybridization between allotetraploids in peonies (Paeonia). Proceedings of the National Academy of Sciences USA 98, 3915–3919. Ferris, C., Oliver, R.P., Davy, A.J. and Hewitt, G.M. (1993) Native oak chloroplasts reveal an ancient divide across Europe. Molecular Ecology 2, 337–344. Gallez, G.P. and Gottlieb, L.D. (1982) Genetic evidence for the hybrid origin of the diploid plant Stephanomeria diegensis. Evolution 36, 1158–1167. Golden, J.L. and Bain, J.F. (2000) Phylogeographic patterns and high levels of chloroplast DNA diversity in four Packera (Asteraceae) species in southwestern Alberta. Evolution 54, 1566–1579. Grant, V. (1966) The origin of a new species of Gilia in a hybridization experiment. Genetics 54, 1189–1199. Grant, V. (1971) Plant Speciation. Columbia University Press, New York. Harlan, J.R. and deWet, J.M.J. (1963) The compilospecies concept. Evolution 17, 497–501. Hewitt, G.M. (1999) Post-glacial recolonization of European biota. Biological Journal of the Linnaean Society 68, 87–112. Hewitt, G. (2000) The genetic legacy of the Quaternary ice ages. Nature 405, 907–913. Klier, K., Leoschke, M.J. and Wendel, J.F. (1991) Hybridization and introgression in white and yellow lady- slipper orchids (Cypripedium candidum and C. pubescens). Journal of Heredity 82, 305–319. Klips, R.A. (1999) Pollen competition as a reproductive isolating mechanism between two sympatric Hibiscus species (Malvaceae). American Journal of Botany 86, 269–272. Lamont, B.B., He, T., Enright, N.J., Krauss, S.L. and Miller, B.P. (2003) Anthropogenic disturbance promotes hybridization between Banksia species by altering their biology. Journal of Evolutionary Biology 16, 551–557. Lapointe, F.-J. (2000) How to account for reticulation events in phylogenetic analysis: a comparison of dis- tance-based methods. Journal of Classification 17, 175–184. Larena, B.G., Aguilar, J.F. and Feliner, G.N. (2002) Glacial-induced altitudinal migrations in Armeria (Plumbaginaceae) inferred from patterns of chloroplast DNA haplotype sharing. Molecular Ecology 11, 1965–1974. Levin, D.A., Francisco-Ortega, J.K. and Jansen, R.K. (1996) Hybridization and the extinction of rare plant species. Conservation Biology 10, 10–16. Lewis, H. and Lewis, M. (1955) The genus Clarkia. University of California Publications in Botany 20, 241–392. Martinsen, G.D., Whitham, T.G., Turek, R.J. and Keim, P. (2001) Hybrid populations selectively filter gene introgression between species. Evolution 55, 1325–1335. Masterson, J. (1994) Stomatal size in fossil plants: evidence for polyploidy in majority of angiosperms. Science 264, 421–423. McKinnon, G.E., Jordan, G.J., Vaillancourt, R.E., Steane, D.A. and Potts, B.M. (2004) Glacial refugia and reticulate evolution: the case of the Tasmanian eucalypts. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 359, 275–284. Milne, R.I. and Abbott, R.J. (2000) Origin and evolution of invasive naturalized material of Rhododendron ponticum L. in the British Isles. Molecular Ecology 9, 541–556. Petit, R.J., Pineau, E., Demesure, B., Bacilieri, R., Ducousso, A. and Kremer, A. (1997) Chloroplast DNA footprints of postglacial recolonization by oaks. Proceedings of the National Academy of Sciences USA 94, 9996–10001. Rieseberg, L.H. (1991) Homoploid reticulate evolution in Helianthus (Asteraceae): evidence from ribosomal genes. American Journal of Botany 78, 1218–1237. Rieseberg, L.H. (1997) Hybrid origins of plant species. Annual Review of Ecology and Systematics 28, 359–389.

96 G. McKinnon Rieseberg, L.H. and Gerber, D. (1995) Hybridization in the Catalina Island Mountain Mahogany (Cercocarpus traskiae): RAPD evidence. Conservation Biology 9, 199–203. Rieseberg, L.H. and Soltis, D.E. (1991) Phylogenetic consequences of cytoplasmic gene flow in plants. Evolutionary Trends in Plants 5, 65–84. Rieseberg, L.H., DesRochers, A.M. and Youn, S.J. (1995a) Interspecific pollen competition as a reproductive barrier between sympatric species of Helianthus (Asteraceae). American Journal of Botany 82, 515–519. Rieseberg, L.H., Linder, C.R. and Seiler, G.J. (1995b) Chromosomal and genic barriers to introgression in Helianthus. Genetics 141, 1163–1171. Rieseberg, L.H., Sinervo, B., Linder, C.R., Ungerer, M.C. and Arias, D.M. (1996) Role of gene interactions in hybrid speciation: evidence from ancient and experimental hybrids. Science 272, 741–745. Rieseberg, L.H., Archer, M.A. and Wayne, R.K. (1999a) Transgressive segregation, adaptation and speciation. Heredity 83, 363–372. Rieseberg, L.H., Whitton, J. and Gardner, K. (1999b) Hybrid zones and the genetic architecture of a barrier to gene flow between two sunflower species. Genetics 152, 713–727. Sang, T., Crawford, D.J. and Stuessy, T.F. (1995) Documentation of reticulate evolution in peonies (Paeonia) using ITS sequences of nrDNA: implications for biogeography and concerted evolution. Proceedings of the National Academy of Sciences USA 92, 6813–6817. Schwarzbach, A.E. and Rieseberg, L.H. (2002) Likely multiple origins of a diploid hybrid sunflower species. Molecular Ecology 11, 1703–1715. Smith, H.H. and Daly, K. (1959) Discrete populations derived by interspecific hybridization and selection. Evolution 13, 476–487. Soltis, D.E. and Soltis, P.S. (1999) Polyploidy: recurrent formation and genome evolution. Trends in Ecology and Evolution 14, 348–352. Soltis, P.S., Plunkett, G.M., Novak, S.J. and Soltis, D.E. (1995) Genetic variation in Tragopogon species: additional origins of the allotetraploids T. mirus and T. miscellus (Compositae). American Journal of Botany 82, 1329–1341. Song, K., Lu, P., Tang, K. and Osborn, T.C. (1995) Rapid genome change in synthetic polyploids of Brassica and its implications for polyploid evolution. Proceedings of the National Academy of Sciences USA 92, 7719–7723. Stace, C.A. (1975) Introduction. In: Stace, C.A. (ed.) Hybridization and the Flora of the British Isles. Academic Press, London, pp. 1–90. Stebbins, G.L. Jr (1950) Variation and Evolution in Plants. Columbia University Press, New York. Templeton, A.R. (1981) Mechanisms of speciation: a population genetic approach. Annual Review of Ecology and Systematics 12, 23–48. Ungerer, M.C., Baird, S.J.E., Pan, J. and Rieseberg, L.H. (1998) Rapid hybrid speciation in wild sunflowers. Proceedings of the National Academy of Sciences USA 95, 11757–11762. Wang, J.Q. and Cruzan, M.B. (1998) Interspecific mating in the Piriqueta caroliniana (Turneracea) complex: effects of pollen load size and competition. American Journal of Botany 85, 1172–1179. Wang, X.R. and Szmidt, A.E. (1994) Hybridization and chloroplast DNA variation in a Pinus complex from Asia. Evolution 48, 1020–1031. Wang, X.R., Szmidt, A.E. and Savolainen, O.T.I. (2001) Genetic composition and diploid hybrid speciation of a high mountain pine, Pinus densata, native to the Tibetan plateau. Genetics 159, 337–346. Wendel, J.F., Schnabel, A. and Seelanan, T. (1995) An unusual ribosomal DNA sequence from Gossypium gossypioides reveals ancient, cryptic, intergenomic introgression. Molecular Phylogenetics and Evolution 4, 298–313. Williams, E.G. and Rouse, J.L. (1988) Disparate style lengths contribute to isolation of species in Rhododendron. Australian Journal of Botany 36, 183–191. Williams, E.G., Kaul, V., Rouse, J.L. and Palser, B.F. (1986) Overgrowth of pollen tubes in embryo sacs of Rhododendron following interspecific pollination. Australian Journal of Botany 34, 413–423. Wolfe, A.D., Xiange, Q.Y. and Kephart, S.R. (1998) Diploid hybrid speciation in Penstemon (Scrophulariaceae). Proceedings of the National Academy of Sciences USA 95, 5112–5115. Xu, S. (2000) Phylogenetic analysis under reticulate evolution. Molecular Biology and Evolution 17, 897–907. Yates, J.S., Golden, J.L. and Bain, J.F. (1999) A preliminary phytogeographical analysis of inter- and intra- populational chloroplast DNA variation in Packera pseudaurea (Asteraceae: Senecioneae) from south- western Alberta and adjacent Montana. Canadian Journal of Botany 77, 305–311.

7 Polyploidy and evolution in plants Jonathan Wendel1 and Jeff Doyle2 1Department of Ecology, Evolution and Organismal Biology, Iowa State University, Ames, IA 50011, USA; 2Department of Plant Biology, 228 Plant Science Building, Cornell University, Ithaca, NY 14853-4301, USA One of the central goals of evolutionary Prevalence of Polyploidy in Plants biology is to understand the origin of new lineages and species. Accordingly, there is Although genome sequencing and compara- an abiding interest in the processes by tive mapping studies have demonstrated that which biodiversity arises, and in elucidating all eukaryotes have experienced one or more the full spectrum of intrinsic mechanisms rounds of genome doubling at some point in and extrinsic forces that shape the specia- their evolutionary history (Wolfe and tion process. In plants, one of the more Shields, 1997; Pébusque et al., 1998; Hughes prominent mechanisms of speciation et al., 2000; Gu et al., 2002), the phenomenon involves genome doubling, or polyploidy. appears to have been especially prevalent in This phenomenon exemplifies the complex higher plants. In fact, it is difficult to over- interplay between ‘intrinsic mechanisms’ state the importance of polyploidy in the and ‘extrinsic forces’, as it entails a suite of evolutionary history of flowering plants. internal genetic, genomic and physiological Based on the distribution of chromosome processes as well as external population- numbers among extant species (Stebbins, level and ecological forces. Thus, a chapter 1950, 1971; Lewis, 1980b; Grant, 1981), or devoted to the subject of polyploidy is by comparisons of stomatal size in living and interesting not only because of its impor- fossil taxa (Masterson, 1994), it has been esti- tance to plant speciation, but also because mated that perhaps three-quarters of of how much the subject enriches our angiosperms have experienced one or more understanding of the evolutionary process. episodes of ancient chromosome doubling. In this chapter we summarize some of the Although polyploidy is uncommon in gym- salient features of polyploidy in plants, nosperms and liverworts, it is common in including a brief description of its preva- mosses (Kuta and Przywara, 1997) and lence and modes of formation. We also exceptionally so in ferns: perhaps 95–100% introduce several model systems for the of pteridophytes have experienced at least study of polyploids and provide example one round of polyploidization in their past case studies, hoping to illuminate more (Masterson, 1994; Leitch and Bennett, 1997; richly how the ‘internal’ and ‘external’ Otto and Whitton, 2000). Thus, the notion processes associated with polyploidy con- that ‘polyploidy has contributed little to pro- tribute to evolutionary success and to the gressive evolution’ (Stebbins, 1971) has been generation of biodiversity. replaced by a consensus view that polyploidy © CAB International 2005. Plant Diversity and Evolution: Genotypic and 97 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

98 J. Wendel and J. Doyle is a prominent force in plant evolution so much attention. In plants, though, it is (Leitch and Bennett, 1997; Otto and an active, ongoing process in many lin- Whitton, 2000; Soltis and Soltis, 2000; eages (Stebbins, 1950; Lewis, 1980b; Wendel, 2000; Soltis et al., 2004). Grant, 1981). Many plant genera contain a high percentage of polyploid species as It is helpful to provide some additional well as diploids, showing that polyploid perspective on the use of the term ‘poly- formation occurred repeatedly in each ploidy’. Because genome doubling has been genus since its origin. Also, polyploid continuing since angiosperms first appeared series are commonly observed within in the Cretaceous and because this remains angiosperm and pteridophyte genera; an active, ongoing process (Otto and these comprise species that differ in multi- Whitton, 2000), many angiosperm genomes ples of a single base chromosome number, have experienced several cycles of poly- for example, 7, 14, 28… (for many exam- ploidization at various times in the past. ples, see Stebbins, 1950; Grant, 1981). Thus, most angiosperms are appropriately Thus, genome multiplication occurs considered to have ‘paleopolyploid’ genomes beyond the single round giving rise to a resulting from one or more rounds of tetraploid, generating higher ploidy levels. genome doubling. The more ancient of these Extraordinary examples abound of poly- past events may be difficult to discern because ploid series that attain very high ploidy of potentially rapid evolutionary restoration levels, including Potentilla (up to 16-ploid), of diploid-like chromosomal behaviour Chrysanthemum (up to 22-ploid) and Poa (up and/or other evolutionary changes following to 38-ploid). According to Grant (1981), polyploidization. Moreover, relatively recent Kalanchoe ranks among the leaders in this episodes of genome doubling may become category, with chromosome multiples superimposed on earlier cycles of poly- approaching 60-ploid, whereas even ploidization. Consequently, the polyploid higher ploidy levels have been suggested nature of many plant genomes was not evi- for Sedum and Ophioglossum (see Otto and dent until the advent of comparative Whitton, 2000). These widespread obser- genomics and genome sequencing projects, vations provide cytogenetic evidence that which commonly reveal duplicate (or higher polyploidy occurs repeatedly on the evolu- multiples) genomic regions or chromosomes tionary timescale of individual genera and that are most readily explained by polyploidy that it is widely dispersed among (Gaut and Doebley, 1997; Wilson et al., 1999; angiosperms and other plant groups. Devos and Gale, 2000; Paterson et al., 2000; Additional evidence on the frequency of Vision et al., 2000; Wendel, 2000; Simillion et polyploid speciation events has come from al., 2002; Blanc et al., 2003; Paterson et al., studies of the distribution of haploid chro- 2003). Prominent examples include many of mosome numbers (Otto and Whitton, our most important crop species (Paterson et 2000). Because genome doubling will al., 2003; Arnold et al., 2004), as well as the immediately create even haploid numbers, model plant Arabidopsis (Vision et al., 2000; there should be an excess of even as Simillion et al., 2002; Blanc et al., 2003), opposed to odd haploid numbers if poly- which was once considered a quintessential ploidy is common. Otto and Whitton diploid because of its small genome and low (2000) demonstrated that this is the case chromosome number. Given these and many for both ferns and angiosperms, and pro- other examples, it is probably safe to state vided minimal estimates of the percentage that no higher plant has escaped the influ- of speciation events that result from poly- ence of polyploidy. ploidization events. The conclusion sug- gested by these and other studies is that Frequency of Polyploidy polyploidy represents the most common mode of sympatric speciation in plants, If polyploid formation was only a rare and and hence is extraordinarily important to ancient phenomenon, it would not deserve a discussion of plant diversity.

Polyploidy and evolution in plants 99 Types of Polyploids and Modes of holed into the terms ‘autopolyploidy’ and Formation ‘allopolyploidy.’ In actuality, these two modes of formation represent endpoints in So far we have not addressed the manner in a taxonomic–genetic continuum. In general, which genome doubling takes place or the individuals within species will have genomes modes by which polyploids form. that are less diverged from one another than Traditionally, polyploidy has been thought will individuals from different species, and, to result from either duplication of a single to the extent that pairing of homoeologous genome (autopolyploidy) or from the combi- chromosomes becomes more difficult as nation of two or more differentiated genomes diverge, taxonomic allopolyploids genomes (allopolyploidy) (Stebbins, 1947, will more often be genetically allopolyploid 1950; Grant, 1981). However, polyploids than will taxonomic autopolyploids. form in many different ways, running the full gamut from a single genetically uniform An important consequence of the mode diploid plant doubling its chromosome com- of formation is that the two endpoints of plement to hybridization between individu- autopolyploidy and allopolyploidy often als from highly divergent species. For make different predictions with respect to systematists, the primary distinction is taxo- chromosome behaviour and genetic segre- nomic: an autopolyploid arises within a gation. These data, in turn, may offer species, and an allopolyploid involves insight into the mode of formation of any hybridization between two or more species particular polyploid. In strict autopoly- (Lewis, 1980b; Grant, 1981; Ramsey and ploids, there are four homologous chromo- Schemske, 1998; Soltis et al., 2004). somes capable of associating with one Although these definitions work well in another at meiosis, resulting in random many taxa, the terms would be more useful bivalents and, often, in multivalents. In con- if species themselves were both objectively trast, in a genetic allopolyploid the homolo- more definable and taxonomically more sta- gous chromosomes contributed by the two ble. Taxonomic definitions are hampered by diploid progenitor species are by definition the realities of natural variation: some unable to pair at meiosis, and as a conse- named species harbour tremendous quence only bivalents are formed at meiosis amounts of genetic and chromosomal varia- in the allotetraploid. Because of these differ- tion, and may in fact comprise several cryp- ences in chromosome behaviour, autopoly- tic species, whereas other species may be ploids and allopolyploids display different genetically nearly identical to their closest patterns of genetic segregation. The case relatives. Geneticists are more interested in with allopolyploids is straightforward; the behaviour of genes and chromosomes although each gene is doubled, the two following genome doubling than in taxo- homoeologous genomes behave indepen- nomic definitions; here the important dis- dently and genetic segregation at each is tinction is whether or not chromosomes disomic, as in its diploid progenitors. That is, from the different (homoeologous) comple- the genetic behaviour of any single locus in ments are capable of pairing with one an allopolyploid is expected to be very much another at meiosis. If so, the plant is like that of a diploid (interesting and myste- autopolyploid, if not, and chromosome com- rious exceptions to this expectation will be plements instead are maintained as two sep- discussed below). Random pairing and mul- arate sets that generally do not interact, the tivalent formation in autopolyploids, how- plant is an allopolyploid. ever, leads to quasi-random patterns of segregation among the multiple chromo- Clearly, there is expected to be broad some copies, and hence patterns of genetic overlap between the taxonomic and genetic segregation that differ from simple, disomic, definitions of polyploids, and in fact much Mendelian inheritance. In an autotetra- of the richness of the complexity of poly- ploid, for example, tetrasomic ratios are ploid formation is lost when it is pigeon- observed, either chromosomal or chromati- dal, depending on whether or not the locus

100 J. Wendel and J. Doyle in question resides close to the centromere Autopolyploids in nature are known to arise (for a lucid description, see pp. 240ff. in either from such mechanisms as spontaneous Grant, 1975). In practice, the segregation somatic doubling, merger of two unreduced ratio observed in any situation depends on gametes, or by a triploid ‘bridge’, the latter many variables, including the mode of poly- having arisen as described above. Triploid ploid formation, amount of pairing among individuals, although usually highly sterile, doubled chromosomes, time since polyploid typically produce an appreciable percentage formation and chromosomal location of the of viable gametes, thereby facilitating autote- gene in question. traploid formation either by self-pollination or crossing with diploid individuals that also pro- The genetic behaviour of duplicated loci duce a low percentage of unreduced gametes. has important consequences. Genetic and evo- Allotetraploids also arise by several means, lutionary advantages accrue to both classes of including ‘one-step’ and ‘two-step’ pathways polyploids, but in very different ways: (Harlan and DeWet, 1975; Ramsey and allopolyploids offer the presumed advantage Schemske, 1998), the former from the merger of fixed heterozygosity at all loci, whereas for of two unreduced gametes from two different autopolyploids the smaller proportion of species, and the latter either via a triploid homozygotes due to tetrasomic inheritance bridge or from spontaneous somatic doubling buffers them against the loss of genetic varia- of a sterile, interspecific diploid. Unreduced tion. The coalescent times of loci under di- gamete formation is nearly 50-fold greater in somic and tetrasomic inheritance can differ hybrids than in non-hybrids (Ramsey and dramatically. Disomically inherited loci typical Schemske, 1998), increasing the likelihood of of allopolyploids trace their origin to the allopolyploid formation by this route, after an diploid progenitors, and thus should coalesce interspecific hybrid has been formed. more deeply than tetrasomically inherited loci, where segregational loss of alleles contin- Allopolyploidy is probably more common ues after polyploid formation. The difference than autopolyploidy in nature (Ramsey and in coalescent times has been used to infer a Schemske, 1998; Soltis et al., 2004), although complex history for the maize genome; Gaut the latter is far more prevalent than was once and Doebley (1997) found a bimodal distribu- thought (Lewis, 1980a). Autopolyploidy has tion of coalescent times among low copy historically been considered maladaptive or nuclear genes, and suggested that maize is a at least uncommon relative to allopolyploidy, segmental allopolyploid, a class of polyploid in part because of fertility reductions associ- intermediate between strict auto- and ated with multivalent formation (and the allopolyploidy. Presumably many polyploids attendant production of gametes with an show some features of both auto- and unbalanced chromosome complement), but allopolyploids, and the full spectrum of evolu- also because of the fitness advantages pre- tionary possibilities is not adequately captured sumed to accompany the ‘fixed heterozygos- with only two or three terms. ity’ of allopolyploids (Stebbins, 1950, 1971; Grant, 1981). More recent empirical studies Both autopolyploids and allopolyploids are have drawn attention to many examples of known to be formed by several different successful autopolyploidy in plants (Soltis mechanisms (Harlan and DeWet, 1975; and Soltis, 1993, 1999; Soltis et al., 2004). In Ramsey and Schemske, 1998). One key fea- addition, Ramsey and Schemske (1998) point ture of these various mechanisms is that meio- out that the relative frequencies of autopoly- sis is an imperfect process. Specifically, failure ploidy and allopolyploidy are strongly in chromosome segregation may lead to the dependent on the frequency of interspecific formation of ‘unreduced’ or ‘2n’ gametes, F1 hybrid formation as well as mating system, which have a somatic complement of chromo- because of the aforementioned boost in somes. A union of two unreduced gametes unreduced gamete formation in hybrid rela- may subsequently lead to the formation of a tive to non-hybrid individuals. Although tetraploid embryo, for example, whereas the quantitative data on the frequency of inter- union of a normal, reduced gamete (1n) with specific hybrid formation in plants are lack- an unreduced gamete will generate a triploid.

Polyploidy and evolution in plants 101 ing, hybridization is known to be common tions (Figs 7.1–7.6). These include: (i) the (Grant, 1981; Rieseberg and Wendel, 1993; surprising and relatively recent realization Rieseberg, 1995; Arnold, 1997). Ramsey and that many polyploid taxa have experienced Schemske, however, conclude that interspe- multiple origins from genotypically similar cific hybridization rates are probably too low antecedents, which is expected not only to in many situations to generate an abundance increase genetic diversity within and among of allopolyploids, and hence that the rate of polyploid populations but also to provide an formation of autopolyploids may often opportunity for (ii) novel recombinational exceed that of allopolyploids. More recently, products at the polyploid level. The possibil- Ramsey and Schemske (2002) have chal- ities created by these twin processes lead lenged the primary assumption that directly to (iii) speciation mechanisms not autopolyploidy is associated with high fertil- available at the diploid level. Each of these is ity cost caused by chromosomally unbal- discussed briefly in turn below. To illustrate anced gametes, because, as mentioned these and related points more clearly, exam- above, natural selection acts quickly to ples are provided from some of the more restore fertility. Taken together, the evidence thoroughly studied angiosperm genera. suggests that both allopolyploidy and autopolyploidy are common evolutionary Multiple origins and genetic variation in outcomes; much remains to be learned about polyploids overall frequencies, relative rates in specific genera, or the long-term evolutionary fates Of all the accomplishments of molecular of these alternative products. phylogenetics in the area of polyploidy, per- haps none has received as much attention as Processes in Polyploids that Contribute the rigorous documentation that polyploids to Biological Diversity can form recurrently. This is not entirely a The pervasiveness of polyploidy in the plant 8x AAAAAABB kingdom, as discussed above, offers the most obvious measure of the importance of the 6x AABBDD phenomenon with respect to the genesis of biodiversity. Yet the full significance of poly- 4x AAAA AABB ploidy requires an understanding of the eco- logical context in which polyploids form and 2x AA BB CC DD the full suite of interactions with their animal AB CD pollinators and herbivores. Recent work has underscored this important point, showing Fig. 7.1. Four diploid (2x) species (A–D), along with how sympatric diploid and autotetraploid their phylogenetic relationships. Two types of Heuchera grossulariifolia plants differ with tetraploids (4x) are formed from these species: respect to the pollinators that they attract AAAA is an autotetraploid, and AABB is an and in their phytophagous insects (Nuismer allopolyploid. Additional levels of complexity occur and Thompson, 2001; Thompson et al., when the AABB allopolyploid hybridizes with 2004). Thompson et al. (2004) review the rel- species D (DD diploid) to form a hexaploid (6x) evance of polyploidy to biodiversity in ani- AABBDD, and when the two tetraploid species form mals, and conclude that polyploidy in plants the autoallooctoploid (8x) AAAAAABB. represents a significant diversifying force in animals by virtue of the many ecological interactions with herbivores and pollinators. Embedded within the biology of poly- ploids are other processes that play subtle roles in fostering genetic and phenotypic variation within and among plant popula-

102 J. Wendel and J. Doyle BBCC AABB BBCxC 1 12 AB BBC 2 AxB BBxC 3 AA BB CC (a) 4 (b) Fig. 7.2. Modes of origin of two allopolyploids are 3 shown. Diploid species AA and BB each produce haploid gametes (A and B) that unite (AxB) to form Fig. 7.3. Allele networks for a pair of homoeologous the diploid hybrid AB, which will usually be sterile single-copy nuclear gene loci in allopolyploid AABB because of poor chromosome pairing. If this plant and its diploid progenitors (AA and BB). Alleles doubles its chromosome number, a fixed hybrid sampled from diploids are shown as squares, alleles allotetraploid (AABB) is formed, in which case sampled from tetraploid individuals are shown as homologous pairing can occur and fertility is diamonds; lines connecting alleles represent one restored. A comparable allopolyploid (BBCC) is mutational step; small circles represent unsampled formed by a ‘triploid bridge’ (BBC). Diploid CC alleles. (a) Alleles from diploid AA and the A- produces a reduced gamete (C), but diploid BB homoeologue alleles from the polyploid. In this produces an unreduced, diploid gamete (BB); these network, two alleles from the polyploid are identical unite to form triploid BBC. Meiotic aberrations in to alleles from the diploid progenitor; each is the triploid increase the frequency of unreduced inferred to be derived from a separate origin of the gamete production; in the example shown, a polyploid. Alleles 3 and 4 from the polyploid are triploid gamete (BBC) unites with a haploid gamete differentiated from any diploid alleles. They could from the CC diploid to form the BBCC tetraploid. represent either additional independent origins or could simply reflect a lack of sampling of the diploid or extinction of allele lineages in the diploid. Allele network (b) represents the BB diploid and the B-homoeologues of the polyploid. The network illustrates recombination among alleles from the diploid (the loop in the network). new concept, and in fact was implicit if not 1 C directly stated early in the biosystematics era (Soltis et al., 2004). Nevertheless, the ability A2 to assess variation at the level of DNA sequences, and, in particular, to identify 1 D with great precision particular nuclear gene alleles and chloroplast haplotypes shared by 2B a polyploid and its diploid progenitors elim- inated any notion that most polyploids are Fig. 7.4. Chloroplast haplotype network for species products of single events (Soltis and Soltis, A–D. Polyploid AABB has haplotypes from both the 1999; Wendel, 2000). Although in some A and B diploid species, illustrating bidirectional cases the number of origins may be overesti- origins if chloroplast DNA is maternally inherited. mated, when, for example, the possibility of heterozygous unreduced gametes is ignored The phenomenon of multiple, indepen- (Watanabe et al., 1991; Vogel et al., 1999), dent origins of polyploids is central to our more commonly the number of indepen- understanding of polyploid evolution and dent origins will be underestimated, because the generation of biodiversity. Most obvious of the joint processes of lineage extinction in diploids and polyploids and allelic diver- gence over time.

Polyploidy and evolution in plants 103 Plant Locus over a long enough time to permit sampling A B cp Genotype of progenitors which themselves are experi- 1a encing allelic divergence from their 1b 1 3 C1 P antecedents. In such cases, the polyploid 1c 1 3 C1 P gene pool would receive periodic infusions 1d 4 2 B2 Q of new allelic variation, the importance of 4 2 B2 Q which is not fully understood but which clearly could provide fodder for evolution- 2a 1 3 C1 P ary diversification. There is as yet no docu- 2b 4 2 B2 Q mentation of the phenomenon of temporally 2c 4 3 B2 R recurrent polyploid formation, and it may 2d 1 2 C1 S be experimentally challenging to gather the quality of evidence necessary to rule out Fig. 7.5. Lineage recombination. The genotypes of alternative explanations for an observed pat- four AABB polyploid plants (a–d) are shown, based tern of diversity within a given polyploid lin- on the networks shown in Figs 7.3 and 7.4. Two eage. For example, one phylogenetic cases are shown. In the first (1a–1d), there are two signature of this scenario might be that a classes of plants, each with the same alleles at the polyploid would share some alleles with a two homoeologous loci and having the same diploid progenitor and others that are chloroplast haplotype; these are classified as unique to the polyploid. This pattern, how- genotypes P and Q. In the second example (plants ever, may also be produced by incomplete 2a–2d), these same two genotypes are also found, sampling of, or lineage extinction within, but lineage recombination has occurred to produce the diploid. Notwithstanding the absence of two new genotypes, R and S. In Case 1 there is compelling empirical examples, there is no evidence for two origins of the AABB polyploid. In reason to suspect that polyploids do not Case 2 there is also evidence of more than one form recurrently over time, at least in those origin, but the number of different genotypes can be cases where long-term sympatry among explained either by four separate origins or by progenitors may be expected and where lineage recombination following gene exchange barriers to hybridization are relatively weak. between plants from a smaller number of origins. Multiple origins are well illustrated by is the inference that polyploids have the Tragopogon, a genus of Asteraceae that pro- potential to sample extensively from the vides the best-studied example of recent pool of genetic variation found in their allopolyploid origin. Early work by Ownbey progenitors. This is true spatially, through (1950) showed that two polyploid species, the contemporaneous formation of poly- Tragopogon miscellus and Tragopogon mirus, ploids from different diploid populations originated through allopolyploidy in the with different genotypes. It is also true tem- western United States around the turn of the porally, if polyploid formation continues Locus Locus Plant A B cp 1a x 1c A B 1a 1 (3) C1 F1-1 1 (3) 1b 1 (3) C1 F1-2 12 1c (4) 2 B2 F1-3 (4) 2 1d (4) 2 B2 F1-4 (4) (3) Fig. 7.6. Reciprocal silencing. Plants from the two separate origins of the AABB polyploid are shown, as in Fig. 7.5. Brackets around alleles from the two homoeologous loci indicate an allele that is constitutively silenced and may be a pseudogene. All plants are viable, because in each there is one functional homoeologous locus. However, when plants from the two different origins are crossed, some F1 individuals (represented by F1–4) have non-functional alleles at both homoeologous loci; such plants are inviable. When this occurs at many unlinked loci, the result will be a barrier to gene flow that can lead to speciation.

104 J. Wendel and J. Doyle 20th century. This novel creation of entirely gene exchange between two or more recur- new species turned out to be a consequence rently formed entities. This, in fact, may be of the introduction to the USA of three critical to the survival and diversification of Eurasion diploid species followed by range polyploid lineages because, in the absence of expansion and subsequent sympatry among recurrent formation, many nascent poly- pairs of these three species. Occasionally, ploids are expected to be relatively depau- these sympatric species pairs underwent perate genetically. With gene flow between hybridization followed by allopolyploid for- polyploids (‘lineage recombination’, sensu mation. This now classic story is referred to Doyle et al., 1999), a genetically diverse and as the Tragopogon triangle. coherent polyploid species is formed, which in principle may be enriched by periodic The evolutionary history of these nascent and ongoing infusions of genetic variation allopolyploids has been extensively studied from diploid progenitors. by Douglas and Pamela Soltis and their col- leagues (Soltis et al., 2004). Among the more The best-documented example of this is interesting aspects of this complex is its in the soybean genus (Glycine), which dynamic nature over a relatively limited includes a large allopolyploid complex range, with new local populations having involving at least eight Australian diploid originated multiple times, and with poly- species and eight allopolyploids that com- ploids replacing diploids as prevalent weeds. bine diploid genomes in various ways (Doyle Not only have the polyploids originated et al., 2004a,b). In some of these polyploids, multiple times, but in some cases in both all alleles are identical or nearly identical to directions; namely, T. miscellus has formed alleles in the diploids, suggesting an origin by crosses involving each of its diploid prog- within the last 50,000 years, coincident with enitors as maternal parent, resulting in reci- climatic changes in Australia that may have procal morphological differences between T. been anthropogenic (Doyle et al., 1999, miscellus individuals depending on maternal 2004a,b). parentage (Soltis et al., 2004). Another inter- esting observation is that the Tragopogon The history of the complex genetic and allopolyploids have formed only in North taxonomic system comprising Glycine America, and not in the native European diploids and allopolyploids has been investi- ranges of the diploid species. gated in detail using molecular markers. This work has demonstrated that different, Among the more enduring lessons to be closely related polyploids vary in their man- learned from Tragopogon is that new poly- ner and frequency of formation, ecology, life ploid species may originate in a telescoped history, geographical range and patterns of timeframe that is apparent within individual molecular evolution (Doyle et al., 2004a,b). human lifetimes. This lesson has been re- Nearly all polyploids have arisen more than taught on several other occasions in other once, but some show exclusively unidirec- plant genera, most notably in the fascinating tional maternal origins whereas in others stories of Spartina anglica (Ainouche et al., both diploid progenitors have contributed 2004) and Senecio cambrensis (Abbott and chloroplast genomes. Some show evidence Comes, 2004). The latter species has arisen of extensive interbreeding among polyploid more than once, whereas Spartina anglica populations with separate origins, resulting appears to have only a single origin. in extensive lineage recombination, whereas in others multilocus genotypes formed by Recombination among allopolyploid lineages different combinations of the same diploid genomes have persisted intact. Several poly- In addition to the potential evolutionary sig- ploids have extensive ranges outside of nificance of infusion of new variation, recur- Australia, and appear to be far better colo- rent polyploid formation may be important nizers than their diploid progenitors, in facilitating novel recombinant genotypes whereas other polyploids have highly to be produced at the polyploid level by restricted ranges. This diversity in life histo- ries and population histories is paralleled by

Polyploidy and evolution in plants 105 the range of molecular evolutionary the epigenetic control of flowering time vari- responses; some polyploids have retained ation in synthetic Brassica allopolyploids nearly equal amounts of homoeologous (Schranz and Osborn, 2000); because flow- nuclear ribosomal DNA repeats, whereas in ering phenology is so obviously important to others one homoeologue predominates reproduction, one can imagine that epige- (Rauscher et al., 2004), apparently as a con- netic divergence can lead to reproductive sequence of repeat loss and concerted evolu- isolation, even in the face of little or no tion (Joly et al., 2004). genetic differentiation. Novel mechanisms of speciation Following polyploid formation, gene silencing is one of the several possible fates In general, polyploids and their diploid of the genome-wide duplication of all genes progenitors are assumed to be reproduc- (reviewed in Wendel, 2000; Lawton-Rauh, tively isolated from one another due to cyto- 2003). This process occurs with the onset of logical incompatibilities and triploid sterility. polyploid formation but increases with time, This makes polyploidy a pervasive mode of such that one of the two duplicate genes is sympatric speciation. However, the issue of expected to be silenced at many different isolation between cytotypes has received rel- loci. Werth and Windham (1991) hypothe- atively little study (but see e.g. Trinti and sized that what they termed ‘reciprocal Scali, 1996; Husband and Schemske, 2000; silencing’ – the loss of expression of differ- Husband et al., 2002). More generally, the ent duplicated gene copies in allopatric role of the cytoplasm in polyploid evolution polyploid populations – could lead to repro- is not well understood (Wendel, 2000; ductive isolation among polyploids, and thus Levin, 2003). One aspect of particular rele- to speciation. The basic idea is that hybrids vance to speciation at the polyploid level is between individuals bearing reciprocally the possibility of cytonuclear interactions silenced duplicated genes would segregate that create barriers to gene flow between progeny that are silenced at both gene polyploid lineages that may share nuclear copies for one to many loci. These double genomes but have different cytoplasms. As mutants would presumably be inviable or discussed above, bidirectional formation of deleterious due to their negative phenotypic allopolyploids has been demonstrated in at or physiological effects. This idea has least some groups. In principle, reproduc- recently been elaborated under the term tive isolation may arise following either mul- ‘divergent resolution’ by Lynch and Force tiple origins or recombination among (2000), and is suggested as a possible factor different allopolyploid lineages, thereby in the evolution of teleost fishes (Taylor et promoting diversification. al., 2001). In addition to speciation promoted by Reasons for the Evolutionary Success of cytonuclear differentiation, other mecha- Polyploids nisms may drive polyploid diversification. Dramatic and potentially rapid molecular The abundance of polyploid plant species in evolution of polyploids (discussed below) nature suggests that polyploidy confers a might lead to speciation. For example, selective advantage over diploidy in some Barrier et al. (2001) have suggested that settings. Traditional explanations for the rapid evolution of floral transcription factors success of polyploids have included a diver- is in part responsible for the dramatic radia- sity of proposals that, broadly speaking, may tion of polyploid Hawaiian silverswords. As be divided into ‘ecological’ and ‘genetic’. discussed in more detail below, polyploidy The former include fitness advantages may be associated with a high level of latent inferred from the greater ecological breadth epigenetic variation, some of which may and amplitude, or colonization of new habi- have phenotypic effects that are potentially tats, which is observed more frequently in visible to selection. An interesting example is polyploids than in their diploid antecedents

106 J. Wendel and J. Doyle (Stebbins, 1950; Grant, 1981). Genetic ‘orthologs’; see Koonin et al., 2004) as being explanations typically invoke a form of het- those that originated by speciation, in con- erosis (hybrid vigour) in polyploids, particu- trast to paralogous genes, which are formed larly in allopolyploids, resulting from the by gene duplication. Discriminating paralo- merger into a single nucleus of complemen- gous genes is theoretically a simple task: if tary alleles. In actuality, ecological and two non-allelic homologues (= homologs) genetic perspectives are not exclusive but are present in a plant genome, then by defi- instead may be complementary, the latter nition they must have arisen by duplication, representing an explanation for the former. and not by speciation, and they are par- For example, the early molecular study of alogues (= paralogs). Orthology is more dif- Tragopogon by Roose and Gottlieb (1976) ficult to establish, requiring a phylogenetic documented ‘fixed hybridity’ at isozyme loci test in which the gene tree must be identical in polyploids (as did many subsequent stud- to a known species phylogeny. ies), from which they speculated that this genetic diversity enabled the polyploids to Where do homoeologues (= homoe- become successful weeds and extend their ologs) fit in? In the genome of an allopoly- ranges. Other genetic explanations for the ploid, loci that were orthologous in the two adaptive superiority of polyploids invoke diploid progenitors become homoeologous. genetic ‘buffering’ conferred by heterozy- The term ‘homoeologous’ predates the gosity (Grant, 1981) or the perceived orthology/paralogy terminology by nearly favourable consequences of gene redun- 40 years and is defined in the Glossary of dancy and its attendant release from func- Genetics and Cytogenetics (Rieger et al., 1976) tional constraint for one gene copy, and/or as ‘the residual homology of originally com- functional divergence (Harland, 1936; pletely homologous chromosomes’. Ohno, 1970; Force et al., 1999; Lynch and ‘Homology’ is used here in the cytogenetic Conery, 2000; Wendel, 2000; Lynch, 2002; sense of the term: ‘chromosomes or chro- Lawton-Rauh, 2003). More recent genetic mosome segments … identical with respect and genomic proposals include adaptive to their constituent genetic loci (the same genome-wide allelic and/or non-allelic inter- loci in the same order) and their visible actions (Pikaard, 2002), and altered regula- structure’; definition 1 in Rieger et al. tory interactions that generate novel (1976). Definition 2 is the more familiar variation that is visible to selection (Osborn evolutionary definition of homology: simi- et al., 2003; Riddle and Birchler, 2003). larity due to common origin. Otholog(ue)s, Paralog(ue)s and So, are homoeologues paralogues or Hom(o)eolog(ue)s: Some Essential orthologues? The answer is yes! They have characteristics of both, but they are more Terminology like paralogues than like orthologues. Because homologous chromosomes become With the discovery that most plant genes homoeologous due to divergence following belong to gene families, molecular biologists speciation, the orthologous genes of two and systematists alike have come to embrace diploid sister species could be said to be the terminology for duplicate genes devel- homoeologous. However, once these homoe- oped many years ago by Fitch (1970). This ologues are united in the compound issue is more than simply a nomenclatural genome of an allopolyploid, they meet the exercise, in that its understanding is central criterion for paralogues: they are formed by to an appreciation of a diversity of biological gene duplication, like any other paralogue, phenomena, such as functional diversifica- the only difference being that they are tion of duplicated genes, as well as the formed by whole genome duplication. appropriate use of gene sequence data for Perhaps more importantly from a practical phylogeny reconstruction. Fitch defined perspective, today’s obvious homoeologues orthologous genes (‘orthologues’ or are tomorrow’s paralogues, and their origin as orthologues may be obscured as evidence for polyploidy is lost owing to diploidization.

Polyploidy and evolution in plants 107 As discussed above, the episodic process of duplicated genes and the acquisition of new polyploidization that we now know is typical function (Ohno, 1970; Force et al., 1999; of plant genomes generates repeated cycles Lynch and Conery, 2000; Lynch and Force, of homoeology followed by gene divergence 2000; Lynch et al., 2001; Altschmied et al., which, over vast amounts of evolutionary 2002; Lynch, 2002; Lawton-Rauh, 2003). time, become multigene families consisting This process is widely perceived to provide of paralogues with varying degrees of the raw material for adaptive diversification. sequence similarity. An alternative and common outcome of gene doubling is that one member of the Gene and Genome Evolution in duplicated gene pair will become silenced Polyploids and ultimately degenerate as a pseudogene (reviewed in Wendel, 2000; Lawton-Rauh, Among the many aspects of polyploidy that 2003). Several other possibilities also exist, have been studied recently, one of the more as modelled in Fig. 7.7, and these have only intriguing has been the question of how two recently come to light as a consequence of divergent genomes coordinately adjust and molecular genetic investigations conducted evolve to guide growth and development in several plant systems, including Brassica once they become united in a common (Song et al., 1995), wheat (Feldman et al., nucleus. The pervasiveness of polyploidy 1997; Liu et al., 1997, 1998a,b; Ozkan et al., constitutes prima facie evidence that such 2001; Shaked et al., 2001; Kashkush et al., adjustments occur and that some fraction of 2002), Arabidopsis (Comai et al., 2000; Lee them have positive fitness consequences. and Chen, 2001; Madlung et al., 2002) and Hence it is of interest to ask about the cotton (Wendel et al., 1995; Hanson et al., genetic, genomic and adaptive consequences 1998, 1999; Jiang et al., 1998; Zhao et al., of genome doubling. 1998). A useful device for conceptualizing gene Thus, there has been a growing aware- and genome evolution in polyploids is ness of a diversity of phenomena associated offered by Fig. 7.7. The most immediate and with polyploidy that were previously important genomic consequence of poly- unknown or unsuspected. The notion that ploidization is the simultaneous duplication has emerged is that polyploid genomes are of all nuclear genes, a phenomenon long ‘dynamic’ (Soltis and Soltis, 1995) at the thought to be central to the evolutionary molecular level, generating an array of success of polyploids (Stebbins, 1950; novel genomic instabilities or changes dur- Stephens, 1951a,b; Ohno, 1970; Lewis, ing the initial stages of polyploid formation 1980b; Levin, 1983). As explained above, or over longer time spans. Some of these genes duplicated by polyploidy are termed alterations are not readily explained by homoeologues. As modelled in Fig. 7.7, at Mendelian principles, but may none the less the time of polyploid formation each gene in have contributed to the evolutionary success the genome will become duplicated, such of polyploids (Soltis and Soltis, 1995; that two homoeologues (‘At’ and ‘Bt’) will Wendel, 2000; Finnegan, 2001a; Rieseberg, exist for each locus, with each homoeologue 2001; Liu and Wendel, 2002; Pikaard, 2002; being phylogenetically sister to its counter- Osborn et al., 2003). Examples of recent part in the progenitor diploid (A with At, B insights into the genetic and genomic behav- with Bt). One possible evolutionary outcome iour of polyploids include: (i) rapid genomic is the long-term preservation of both changes and intergenomic interactions that homoeologues, as well as retention of ances- become possible as a consequence of the tral functions by both copies. This scenario merger of two genomes into a single provides a useful null model against which nucleus; and (ii) epigenetic alterations that other possibilities may be evaluated. One may accompany new polyploids. We summa- long-recognized possibility is relaxation of rize briefly some of the important observa- selection, allowing divergence between the tions and explore the possible biological significance of the various phenomena.

108 J. Wendel and J. Doyle Fig. 7.7. Gene duplication and evolution in polyploids. Shown on the left is a hypothetical set of organisms comprising two diploids and their allopolyploid derivative. With the onset of allopolyploid formation all genes become duplicated. The two homoeologues (At, Bt) are not expected to be phylogenetically sister to each other, but are instead expected to be sister to their counterparts from each respective diploid (A, B). In addition, all else being equal, evolutionary rates are expected to be equivalent. These expectations provide convenient null hypotheses (middle), which may be falsified by a number of processes, including gene conversion (top centre), unequal rates (bottom centre), gene silencing (top right), the evolution of new function (centre right) and intergenomic transfer (bottom right). Recent work has demonstrated all of these possibilities in one or more natural and/or synthetic allopolyploids. Genome change and intergenomic important early paper by Song et al. (1995) interactions resulting from polyploidy demonstrated the novel appearance as well as disappearance of different restriction frag- Recent explorations of polyploidy have led to ments in synthetic Brassica polyploids and the discovery of a number of somewhat mys- their progeny. This work was soon followed terious phenomena associated with polyploid by similar observations in tetraploid and formation. Though we still know relatively lit- hexaploid wheats, with the added twist that tle about most details of genomic merger, the some of the changes observed in newly syn- initial stages evidently are moulded by an thesized wheat allopolyploids mimicked those array of molecular genetic mechanisms and observed in natural wheats with the same processes that collectively lead to polyploid genomic composition (Feldman et al., 1997; stabilization (Song et al., 1995; Soltis and Liu et al., 1998a,b; Shaked et al., 2001). The Soltis, 1999; Comai et al., 2000; Wendel, 2000; latter observation implied that the non- Lee and Chen, 2001; Kashkush et al., 2002; Mendelian response to allopolyploid forma- Liu and Wendel, 2002; Soltis et al., 2004). An tion was to a certain extent ‘directed’ by the

Polyploidy and evolution in plants 109 specificities of the genomes involved. These dormant transposable elements. The extent and other studies showed that polyploid and tempo at which these events will occur genomes are neither static nor need be strictly undoubtedly varies among plant species and additive with respect to the genomes of prog- genome combinations. In general, however, enitor diploids. Instead, the merger of two the inherently higher level of tolerance to different genomes in a common nucleus may insertions makes it more likely that transpos- be accompanied by genomic ‘reorganization’ able elements have played a role in genome of unknown aetiology (Wendel, 2000; Liu and evolution in polyploid than in diploid species. Wendel, 2002). Proposed mechanisms to Evidence in support of this supposition is cir- account for this non-Mendelian behaviour cumstantial, consisting primarily of the obser- include homologous and non-homologous vation that transposable elements and other recombination, methylation alterations and repeated sequences have spread among other epigenetic changes (see below) and per- genomes in tetraploid cotton (Hanson et al., haps deletional processes that are not well 1998, 1999) and wheat (Belyayev et al., 2000). understood. In Nicotiana, there apparently has been a massive proliferation of pararetroviruses fol- A different but related phenomenology lowing polyploid formation (Matzke et al., associated with polyploidy concerns trans- 2004). In all cases, the adaptively relevant posable elements. Genome merger in an effects, if any, of these inter-genomic, intra- allopolyploid creates the potential for the nuclear colonizations are not known. spread of transposable elements between two formerly isolated genomes. In addition to the spread of transposable Transposable elements are ubiquitous in elements among genomes, polyploidy cre- plant genomes (Kumar and Bennetzen, ates the opportunity for various types of 1999; Bennetzen, 2000), where they con- interactions between homoeologous genes tribute to genome evolution and genetic or repeated sequences. Tandemly repeated diversity by transposition and the atten- sequences, such as ribosomal genes, have dant effects on gene expression (Kidwell been demonstrated to experience interlocus and Lisch, 2001; Wessler, 2001). Most homogenization or concerted evolution, transposable elements are inactive under whereby sequences from one genome over- normal conditions, but they may become write the homoeologous sequences from the activated under stress (McClintock, 1984; other genome. First convincingly demon- Hirochika et al., 1996; Wessler, 1996; strated in allopolyploid Gossypium (Wendel et Grandbastien, 1998; Beguiristain et al., al., 1995), the phenomenon has now been 2001). In diploid hybrids, enhanced trans- described in a number of genera (Wendel, posable element activity is likely to be mal- 2000; Joly et al., 2004). New twists on this adaptive because insertions may disrupt phenomenon were recently reported in syn- essential gene functions. In polyploids, thetic Nicotiana allopolyploids (Skalicka et al., however, the harmful effects of transpos- 2003; Kovarik et al., 2004), where the rapid able element activity may be buffered by evolution of rDNA types was observed genomic redundancy, and hence insertions within a few generations, as was the appar- would be more likely to be tolerated ent birth of a new rDNA locus. This latter (Matzke et al., 1999; Wendel, 2000). observation may well represent the real-time Hence, it is noteworthy that in newly gen- capture of the well-known but unexplained erated allotetraploid wheat plants (Shaked phenomenon of birth and death of riboso- et al., 2001) and in Orzya ϫ Zizania hybrids mal arrays (Dubcovsky and Dvorák, 1995). (Liu and Wendel, 2000), transposable ele- ments have been shown to be activated. Epigenetics and polyploid evolution Although few natural plant hybrids and Epigenetics refers to heritable alterations in allopolyploids have been experimentally eval- gene expression that do not entail changes uated for transposable element activity, stud- in nucleotide sequence, but which neverthe- ies to date suggest that wide hybridization and allopolyploidy may trigger activation of

110 J. Wendel and J. Doyle less may have phenotypic and hence evolu- approach, Shaked et al. (2001) showed that, tionary consequences. Epigenetic effects can in wheat, cytosine methylation alterations be accomplished by several interrelated are genomically widespread but may signifi- covalent modifications of DNA and/or chro- cantly differ in frequency between the two mosomal proteins, such as DNA methylation constituent genomes of an allopolyploid; in and histone modifications (Nathan et al., first generation allotetraploid wheat, ten of 2003), and by chromatin remodelling, such the 11 bands that showed heritable methyla- as repositioning of nucleosomes. These heri- tion changes were from one of the two table modifications are collectively termed parental genomes. Genome-wide and non- ‘epigenetic codes’ (reviewed in Richards and random changes in DNA methylation pat- Elgin, 2002). Programmed epigenetic con- terns are also observed in synthetic trol of gene expression is essential during allotetraploid Arabidopsis and Arabidopsis normal growth and development (Wolffe arenosa (Madlung et al., 2002). Given the and Matzke, 1999; Finnegan, 2001b) and, importance of DNA methylation to gene because of this, the epigenetic arena is a expression, the foregoing examples indicate vibrant field in current biological research. that polyploid formation could have Given this fundamental significance, it is of genome-wide epigenetic consequences of interest to discuss the possible connections relevance to gene expression and hence between epigenetics and polyploidy. polyploid evolution. An integral component of the develop- This suggestion of epigenetic effects on mental control of gene expression is pro- gene expression may be related to the gen- grammed cytosine methylation (Richards, eral observation that polyploids are often 1997; Finnegan et al., 2000). Hyper- associated with variation and instability in methylation is usually a hallmark of hete- phenotypes that cannot be accounted for by rochromatin and is characteristic of conventional Mendelian transmission genet- euchromatic gene silencing, whereas ics or chromosomal aberrations (Comai, hypomethylation is often associated with 2000; Comai et al., 2000). The affected traits active gene expression (Martienssen and are diverse, including timing of flowering, Colot, 2001; Grewal and Moazed, 2003). In overall plant habit, leaf morphology and plants, cytosine methylation patterns are usu- homeotic transformations in floral morphol- ally stably maintained through meiosis and ogy. These allopolyploidy-associated changes over generations. Experimental disruption of in phenotypes may arise from altered gene cytosine methylation patterns may lead to expression due to variation in dosage-regu- aberrant plant morphology (Finnegan et al., lated gene expression, altered regulatory 1998, 2000; Finnegan, 2001b). As a potential interactions, and rapid genetic and epige- genome defence system (Yoder et al., 1997), netic changes (reviewed in Osborn et al., the cytosine methylation machinery may 2003). respond to environmental or genomic chal- lenges by causing alterations in methylation One of the surprising recent findings that are thought to mediate physiologically concerning polyploids has been the degree meaningful responses. Polyploidy, by uniting to which gene expression may be altered by divergent genomes into one nucleus, may genome doubling (Liu and Wendel, 2002; constitute such a challenge, or ‘genomic Adams et al., 2003; Comai et al., 2003; shock’ (McClintock, 1984; Comai et al., 2003). Osborn et al., 2003; Riddle and Birchler, 2003). Studies of gene expression in natural This suggestion has garnered recent and synthetic polyploids have shown that experimental support. In synthetic Brassica genes may be silenced immediately upon or (Song et al., 1995) and wheat allopolyploids shortly following polyploidization. For (Liu et al., 1998a; Shaked et al., 2001), DNA example, ribosomal RNA arrays from one methylation changes, including both hypo- parent may be silenced in some organs of and hypermethylations, were shown to Brassica napus, although both parental rRNA occur at anonymous genomic loci and in sets are expressed in floral organs (Chen cDNAs. Using a genome-wide fingerprinting and Pikaard, 1997). This is the well-known

Polyploidy and evolution in plants 111 phenomenon of nucleolar dominance, In cotton, natural and artificially gener- where in hybrids or allopolyploids nucleoli ated allotetraploids were studied by Adams et form in association with ribosomal RNA al. (2003) using an electrophoretic approach genes on chromosomes inherited from only to separate the transcripts of 40 different one of the two parents (Pikaard, 1999, duplicate gene pairs. A remarkably high 2000a,b). These demonstrations for riboso- level of transcription bias was observed, with mal genes suggested the possibility that respect to the duplicated copies, in that hybridization and polyploidization might about 25% of the genes studied exhibited similarly induce epigenetic modifications of altered expression in one or more plant protein-coding genes. This suspicion has organs. The most surprising result was the now been confirmed in several model plant observation of organ-specific gene silencing systems (Comai, 2000; Comai et al., 2000; that in some cases was reciprocal, meaning Lee and Chen, 2001; Kashkush et al., 2002; that one duplicate was expressed in one Madlung et al., 2002), including Arabidopsis, organ (e.g. stamens), while its counterpart wheat and cotton polyploids, where silenc- was expressed in a different organ (e.g. ing of numerous protein-coding genes has carpels). Moreover, this organ-specific parti- been demonstrated (Comai et al., 2000; Lee tioning of duplicate expression was also evi- and Chen, 2001; Kashkush et al., 2002; He et dent in newly generated allotetraploids. al., 2003). Remarkably, the silencing patterns observed in natural cotton allopolyploids, estimated to These studies indicate that allopolyploid be approximately 1.5 million years old formation in Arabidopsis is accompanied by (Senchina et al., 2003; Wendel and Cronn, epigenetic gene silencing, and that this 2003), were rather similar in some cases to silencing may affect a variety of genes with those observed in the newly generated diverse biological functions. The silencing tetraploids. This observation implicates events may occur rapidly (F2 generation of either long-term evolutionary maintenance synthetic allopolyploid) or over a longer of epigenetically induced expression states, evolutionary time span, but their reversibil- or subsequent fixation of expression states in ity may be retained in natural allopolyploid the natural tetraploids through normal species for thousands to perhaps millions of mutational processes during the 1–2 million years. Of particular significance is the years since polyploid formation in the genus. remarkable similarity or concordance in the silencing patterns between synthetic and Collectively, recent studies in several model natural allopolyploids, which suggests that plant systems have revealed that polyploid allopolyploidy not only induces epigenetic formation may be accompanied by epigenetic changes but also that the changes may be alterations in gene expression throughout the visible to natural selection and, judging genome. These epigenetic changes may occur from their persistence, adaptive. with the onset of polyploidy or accrue more slowly on an evolutionary timeframe. In at The scale of the phenomenon, and hence least some cases, rapid epigenetic modifica- its potential level of evolutionary impor- tions that arise with the onset of allopoly- tance, is illustrated by recent work involving ploidy may be preserved on an evolutionary synthetic and natural allopolyploid wheat timescale through multiple speciation events. (Kashkush et al., 2002) and cotton A more intriguing suggestion is that genome (Gossypium) (Adams et al., 2003). In newly doubling or merger creates a massive and generated tetraploid wheat, an appreciable sudden pulse of novel epigenetic variation, frequency (1% to 5%) of the genes surveyed which may be released and become visible to experienced silencing within the first gener- natural selection over periods of time ranging ation, and novel transcripts were occasion- up to millions of years. ally observed. Interestingly, the novel transcripts activated by polyploidy that The studies discussed in this section illus- could be assigned a function are retrotrans- trate a number of important phenomena posons, suggesting release from epigenetic that may bear directly on the evolution of control or suppression (see below). polyploids. This includes non-random

112 J. Wendel and J. Doyle genomic changes observed in natural reservoir of epigenetic/genetic combinations allopolyploid species and their synthetic for later release and evaluation by natural counterparts, gene silencing, novel gene selection, perhaps after millions of years expression, the possibility of organ-specific (Adams et al., 2003). This intriguing possibil- partitioning of duplicate gene function, and ity requires further study, but it may be transposable element activation. How each important to our understanding of the evo- of these processes translates into phenotypic lution of polyploids. or physiological variation that may be visible to natural selection is not yet known in most Conclusion cases, but the scale and scope of epigenetic alterations accompanying polyploidy and As summarized in this review, genome dou- the importance of epigenetics to growth and bling has been a pervasive phenomenon in development suggest that these connections plant evolution and remains a prominent have significance to our understanding of process by which biodiversity is generated polyploid evolution. In this respect a partic- today. We have illustrated some of the intrin- ularly relevant example may be epigeneti- sic features of polyploids, generated by cally controlled flowering time variation in genome doubling and/or merger, which pro- synthetic Brassica polyploids (Schranz and vide novel opportunities for creating pheno- Osborn, 2000), because flowering phenology typic variation, and have highlighted some of is so obviously important to reproduction. A the extrinsic factors that guide polyploid for- second example of an epigenetic modifica- mation and subsequent evolution. Much tion that is evolutionarily consequential is remains to be learned regarding the func- the natural flower symmetry mutant (from tional consequences of gene and genome dou- wild-type bilateral to radial) in Linaria vul- bling and the array of molecular genetic garis, originally described by Linnaeus more mechanisms that they both engender and are than 250 years ago. The molecular basis of subject to, as well as the interplay between this mutation is hypermethylation and these internal forces and external ecological silencing of a gene controlling flower form and population-level phenomena. Many (Cubas et al., 1999). When one extrapolates insights are likely in the near future, however, these examples of epigenetic regulation of as molecular genetic and genomic approaches single genes to the entire genome, it are increasingly brought to bear on natural becomes evident that allopolyploid lineages and artificial model polyploid systems. may harbour a nearly infinite and latent References Abbott, R.J. and Comes, H.P. (2004) Evolution in the Arctic: a phylogeographic analysis of the circumarctic plant, Saxifraga oppositifolia (Purple saxifrage). New Phytologist 161, 211–224. Adams, K.L., Cronn, R., Percifield, R. and Wendel, J.F. (2003) Genes duplicated by polyploidy show unequal contributions to the transcriptome and organ-specific reciprocal silencing. Proceedings of the National Academy of Sciences USA 100, 4649–4654. Ainouche, M.L., Baumel, A., Salmon, A. and Yannic, G. (2004) Hybridization, polyploidy and speciation in Spartina (Poaceae). New Phytologist 161, 165–172. Altschmied, J., Delfgaauw, J., Wilde, B., Duschl, J., Bouneau, L., Volff, J.-N. and Schartl, M. (2002) Subfunctionalization of duplicate mitf genes associated with differential degeneration of alternative exons in fish. Genetics 161, 259–267. Arnold, M.L. (1997) Natural Hybridization and Evolution. Oxford University Press, New York. Arnold, M.L., Bouck, A.C. and Scott Cornman, R. (2004) Verne Grant and Louisiana Irises: is there anything new under the sun? New Phytologist 161, 143–149. Barrier, M., Robichaux, R.H. and Purugganan, M.D. (2001) Accelerated regulatory gene evolution in an adaptive radiation. Proceedings of the National Academy of Sciences USA 98, 10208–10213.

Polyploidy and evolution in plants 113 Beguiristain, T., Grandbastien, M.-A., Puigdomenech, P. and Casacuberta, J.M. (2001) Three Tnt1 subfamilies show different stress-associated patterns of expression in tobacco. Consequences for retrotransposon control and evolution in plants. Plant Physiology 127, 212–221. Belyayev, A., Raskina, O., Korol, A. and Nevo, E. (2000) Coevolution of A and B genomes in allotetraploid Triticum dicoccoides. Genome 43, 1021–1026. Bennetzen, J.L. (2000) Transposable element contributions to plant gene and genome evolution. Plant Molecular Biology 42, 251–269. Blanc, G., Hokamp, K. and Wolfe, K.H. (2003) A recent polyploidy superimposed on older large-scale duplications in the Arabidopsis genome. Genome Research 13, 137–144. Chen, Z.J. and Pikaard, C.S. (1997) Transcriptional analysis of nucleolar dominance in polyploid plants: biased expression/silencing of progenitor rRNA genes is developmentally regulated in Brassica. Proceedings of the National Academy of Sciences USA 94, 3442–3447. Comai, L. (2000) Genetic and epigenetic interactions in allopolyploid plants. Plant Molecular Biology 43, 387–399. Comai, L., Tyagi, A.P., Winter, K., Holmes-Davis, R., Reynolds, S.H., Stevens, Y. and Byers, B. (2000) Phenotypic instability and rapid gene silencing in newly formed Arabidopsis allotetraploids. The Plant Cell 12, 1551–1567. Comai, L., Madlung, A., Joseffson, C. and Tyagi, A. (2003) Do the different parental ‘heteromes’ cause genomic shock in newly formed allopolyploids? Philosophical Transactions of the Royal Society of London B 358, 1149–1155. Cubas, P., Vincent, C. and Coen, E. (1999) An epigenetic mutation responsible for natural variation in floral symmetry. Nature 401, 157–161. Devos, K.M. and Gale, M.D. (2000) Genome relationships: the grass model in current research. The Plant Cell 12, 637–646. Doyle, J.J., Doyle, J.L. and Brown, A.H.D. (1999) Origins, colonization, and lineage recombination in a widespread perennial soybean polyploid complex. Proceedings of the National Academy of Sciences USA 96, 10741–10745. Doyle, J.J., Doyle, J.L., Rauscher, J.T. and Brown, A.H.D. (2004a) Diploid and polyploid reticulate evolution throughout the history of the perennial soybeans (Glycine subgenus Glycine). New Phytologist 161, 121–132. Doyle, J.J., Doyle, J.L., Rauscher, J.T. and Brown, A.H.D. (2004b) Evolution of the perennial soybean poly- ploid complex (Glycine subgenus Glycine): a study of contrasts. Biological Journal of the Linnean Society 82, 583–597. Dubcovsky, J. and Dvorák, J. (1995) Ribosomal RNA multigene loci: nomads of the Triticeae genomes. Genetics 140, 1367–1377. Feldman, M., Liu, B., Segal, G., Abbo, S., Levy, A.A. and Vega, J.M. (1997) Rapid elimination of low-copy DNA sequences in polyploid wheat: a possible mechanism for differentiation of homoeologous chro- mosomes. Genetics 147, 1381–1387. Finnegan, E.J. (2001a) Epialleles: a source of random variation in times of stress. Current Opinion in Plant Biology 5, 101–106. Finnegan, E.J. (2001b) Is plant gene expression regulated globally. Trends in Genetics 17, 361–365. Finnegan, E.J., Genger, R.K., Peacock, W.J. and Dennis, E.S. (1998) DNA methylation in plants. Annual Review of Plant Physiology and Plant Molecular Biology 49, 223–247. Finnegan, E.J., Peacock, W.J. and Dennis, E.S. (2000) DNA methylation, a key regulator of plant develop- ment and other processes. Current Opinion in Genetics and Development 10, 217–223. Fitch, W.M. (1970) Distinguishing homologous from analogous proteins. Systematic Zoology 19, 99–113. Force, A., Lynch, M., Pickett, F.B., Amores, A., Yan, Y.-L. and Postlethwait, J. (1999) Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151, 1531–1545. Gaut, B.S. and Doebley, J.F. (1997) DNA sequence evidence for the segmental allotetraploid origin of maize. Proceedings of the National Academy of Sciences USA 94, 6808–6814. Grandbastien, M.-A. (1998) Activation of plant retrotransposons under stress conditions. Trends in Plant Science 3, 181–187. Grant, V. (1975) Genetics of Flowering Plants. Columbia University Press, New York. Grant, V. (1981) Plant Speciation. Columbia University Press, New York. Grewal, S.I.S. and Moazed, D. (2003) Heterochromatin and epigenetic control of gene expression. Science 301, 798–802. Gu, X., Wang, Y. and Gu, J. (2002) Age distribution of human gene families shows significant roles of both large- and small-scale duplications in vertebrate evolution. Nature Genetics 31, 205–209.

114 J. Wendel and J. Doyle Hanson, R.E., Zhao, X.-P., Islam-Faridi, M.N., Paterson, A.H., Zwick, M.S., Crane, C.F., McKnight, T.D., Stelly, D.M. and Price, H.J. (1998) Evolution of interspersed repetitive elements in Gossypium (Malvaceae). American Journal of Botany 85, 1364–1368. Hanson, R.E., Islam-Faridi, M.N., Crane, C.F., Zwick, M.S., Czeschin, D.G., Wendel, J.F., McKnight, T.D., Price, H.J. and Stelly, D.M. (1999) Ty1-copia-retrotransposon behavior in a polyploid cotton. Chromosome Research 8, 73–76. Harlan, J.R. and DeWet, J.M.J. (1975) On Ö. Winge and a prayer: the origins of polyploidy. Botanical Review 41, 361–390. Harland, S.C. (1936) The genetical conception of the species. Cambridge Philosophical Society Biological Reviews 11, 83–112. He, P., Friebe, B., Gill, B. and Zhou, J.-M. (2003) Allopolyploidy alters gene expression in the highly stable hexaploid wheat. Plant Molecular Biology 52, 401–414. Hirochika, H., Sugimoto, K., Otsuki, Y. and Kanda, M. (1996) Retrotransposons of rice involved in mutations induced by tissue culture. Proceedings of the National Academy of Sciences USA 93, 7783–7788. Hughes, A.L., Green, J.A., Garbayo, J.M. and Roberts, R.M. (2000) Adaptive diversification within a large family of recently duplicated, placentally expressed genes. Proceedings of the National Academy of Sciences USA 97, 3319–3323. Husband, B.C. and Schemske, D.W. (2000) Ecological mechanisms of reproductive isolation between diploid and tetraploid Chamerion angustifolium. Journal of Ecology 88, 689–701. Husband, B.C., Schemske, D.W., Burton, T.L. and Goodwillie, C. (2002) Pollen competition as a unilateral reproductive barrier between sympatric Chamerion angustifolium. Proceedings of the Royal Society of London Series B: Biological Sciences 269, 2565–2571. Jiang, C., Wright, R., El-Zik, K. and Paterson, A. (1998) Polyploid formation created unique avenues for response to selection in Gossypium (cotton). Proceedings of the National Academy of Sciences USA 95, 4419–4424. Joly, S., Rauscher, J.T., Sherman-Broyles, S.L., Brown, A.H.D. and Doyle, J.J. (2004) Evolution of the 18S- 5.8S-26S nuclear ribosomal gene family and its expression in natural and artificial Glycine allopoly- loids. Molecular Biology and Evolution 21, 1409–1421. Kashkush, K., Feldman, M. and Levy, A.A. (2002) Gene loss, silencing, and activation in a newly synthe- sized wheat allotetraploid. Genetics 160, 1651–1659. Kidwell, M.G. and Lisch, D.R. (2001) Perspective: transposable elements, parasitic DNA, and genome evo- lution. Evolution 55, 1–24. Koonin, E., Fedorova, N., Jackson, J., Jacobs, A., Krylov, D., Makarova, K., Mazumder, R., Mekhedov, S., Nikolskaya, A., Rao, B., Rogozin, I., Smirnov, S., Sorokin, A., Sverdlov, A., Vasudevan, S., Wolf, Y., Yin, J. and Natale, D. (2004) A comprehensive evolutionary classification of proteins encoded in complete eukaryotic genomes. Genome Biology 5, R7. Kovarik, A., Matyasek, R., Lim, K.Y., Skalicka, K., Koukalova, B., Knapp, S., Chase, M.W. and Leitch, A.R. (2004) Concerted evolution of 18-5.8-26S rDNA repeats in Nicotiana allotetraploids. Biological Journal of the Linnean Society 82, 615–625. Kumar, A. and Bennetzen, J.L. (1999) Plant retrotransposons. Annual Review of Genetics 33, 479–532. Kuta, E. and Przywara, L. (1997) Polyploidy in mosses. Acta Biologica Cracoviensia 39, 17–26. Lawton-Rauh, A. (2003) Evolutionary dynamics of duplicated genes in plants. Molecular Phylogenetics and Evolution 29, 396–409. Lee, H.-S. and Chen, Z.J. (2001) Protein-coding genes are epigenetically regulated in Arabidopsis poly- ploids. Proceedings of the National Academy of Sciences USA 98, 6753–6758. Leitch, I.J. and Bennett, M.D. (1997) Polyploidy in angiosperms. Trends in Plant Science 2, 470–476. Levin, D.A. (1983) Polyploidy and novelty in flowering plants. American Naturalist 122, 1–25. Levin, D.A. (2003) The cytoplasmic factor in plant speciation. Systematic Botany 28, 5–11. Lewis, W.H. (1980a) Polyploidy in species populations. In: Lewis, W.H. (ed.) Polyploidy: Biological Relevance. Plenum, New York, pp. 103–144. Lewis, W.H. (1980b) Polyploidy: Biological Relevance. Plenum, New York. Liu, B. and Wendel, J.F. (2000) Retroelement activation followed by rapid repression in interspecific hybrid plants. Genome 43, 874–880. Liu, B. and Wendel, J.F. (2002) Non-Mendelian phenomena in allopolyploid genome evolution. Current Genomics 3, 489–506. Liu, B., Segal, G., Vega, J.M., Feldman, M. and Abbo, S. (1997) Isolation and characterization of chromo- some-specific sequences from a chromosome arm genomic library of common wheat. Plant Journal 11, 959–965.

Polyploidy and evolution in plants 115 Liu, B., Vega, J.M. and Feldman, M. (1998a) Rapid genomic changes in newly synthesized amphiploids of Triticum and Aegilops. II. Changes in low-copy coding DNA sequences. Genome 41, 535–542. Liu, B., Vega, J.M., Segal, G., Abbo, S., Rodova, M. and Feldman, M. (1998b) Rapid genomic changes in newly synthesized amphiploids of Triticum and Aegilops. I. Changes in low-copy non-coding DNA sequences. Genome 41, 272–277. Lynch, M. (2002) Genomics: gene duplication and evolution. Science 297, 945–947. Lynch, M. and Conery, J.S. (2000) The evolutionary fate and consequences of duplicate genes. Science 290, 1151–1155. Lynch, M. and Force, A. (2000) The probability of duplicate gene preservation by subfunctionalization. Genetics 154, 459–473. Lynch, M., O’Hely, M., Walsh, B. and Force, A. (2001) The probability of preservation of a newly arisen gene duplicate. Genetics 159, 1789–1804. Madlung, A., Masuelli, R.W., Watson, B., Reynolds, S.H., Davison, J. and Comai, L. (2002) Remodeling of DNA methylation and phenotypic and transcriptional changes in synthetic Arabidopsis allotetraploids. Plant Physiology 129, 733–746. Martienssen, R.A. and Colot, V. (2001) DNA methylation and epigenetic inheritance in plants and filamen- tous fungi. Science 293, 1070–1074. Masterson, J. (1994) Stomatal size in fossil plants: evidence for polyploidy in majority of angiosperms. Science 264, 421–424. Matzke, M.A., Scheid, O.M. and Matzke, A.J.M. (1999) Rapid structural and epigenetic changes in poly- ploid and aneuploid genomes. Bioessays 21, 761–767. Matzke, M.A., Gregor, W., Mette, M.F., Aufsatz, W., Kanno, T., Jakowitsch, J. and Matzke, J.M. (2004) Endogenous pararetroviruses of polyploid tobacco (Nicotiana tabacum) and its diploid progenitors, N. sylvestris and N. tomentosiformis. Biological Journal of the Linnean Society 82, 627–638. McClintock, B. (1984) The significance of responses of the genome to challenge. Science 226, 792–801. Nathan, D., Sterner, D.E. and Berger, S.L. (2003) Histone modifications: now summoning sumoylation. Proceedings of the National Academy of Sciences USA 100, 13118–13120. Nuismer, S.L. and Thompson, J.N. (2001) Plant polyploidy and non-uniform effects on insect herbivores. Proceedings of the Royal Society of London Series B: Biological Sciences 268, 1937–1940. Ohno, S. (1970) Evolution by Gene Duplication. Springer-Verlag, New York. Osborn, T.C., Pires, J.C., Birchler, J.A., Auger, D.L., Chen, Z.J., Lee, H.-S., Comai, L., Madlung, A., Doerge, R.W., Colot, V. and Martienssen, R.A. (2003) Understanding mechanisms of novel gene expression in polyploids. Trends in Genetics 19, 141–147. Otto, S.P. and Whitton, J. (2000) Polyploid incidence and evolution. Annual Review of Genetics 34, 401–437. Ownbey, M. (1950) Natural hybridization and amphiploidy in the genus Tragopogon. American Journal of Botany 37, 487–499. Ozkan, H., Levy, A.A. and Feldman, M. (2001) Allopolyploidy-induced rapid genome evolution in the wheat (Aegilops-Triticum) group. The Plant Cell 13, 1735–1747. Paterson, A.H., Bowers, J.E., Burow, M.D., Draye, X., Elsik, C.G., Jiang, C.-X., Katsar, C.S., Lan, Y.-R., Ming, R. and Wright, R.J. (2000) Comparative genomics of plant chromosomes. The Plant Cell 12, 1523–1539. Paterson, A.H., Bowers, J.E., Peterson, D.G., Estill, J.C. and Chapman, B.A. (2003) Structure and evolution of cereal genomes. Current Opinion in Genetics and Development 13, 644–650. Pébusque, M.-J., Coulier, F., Birnbaum, D. and Pontarotti, P. (1998) Ancient large-scale genome duplica- tions: phylogenetic and linkage analyses shed light on chordate genome evolution. Molecular Biology and Evolution 15, 1145–1159. Pikaard, C.S. (1999) Nucleolar dominance and silencing of transcription. Trends in Plant Science 4, 478–483. Pikaard, C.S. (2000a) The epigenetics of nucleolar dominance. Trends in Genetics 16, 495–500. Pikaard, C.S. (2000b) Nucleolar dominance: uniparental gene silencing on a multi-megabase scale in genetic hybrids. Plant Molecular Biology 43, 163–177. Pikaard, C.S. (2002) Genomic change and gene silencing in polyploids. Trends in Genetics 17, 675–677. Ramsey, J. and Schemske, D.W. (1998) Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29, 467–501. Ramsey, J. and Schemske, D.W. (2002) Neopolyploidy in flowering plants. Annual Review of Ecology and Systematics 33, 589–639. Rauscher, J.T., Doyle, J.J. and Brown, A.H.D. (2004) Multiple origins and nrDNA ITS homoeologue evolu- tion in the Glycine tomentella (Leguminosae) allopolyploid complex. Genetics 166, 987–998.

116 J. Wendel and J. Doyle Richards, E. (1997) DNA methylation and plant development. Trends in Genetics 13, 319–323. Richards, E.J. and Elgin, S.C.R. (2002) Epigenetic codes for heterochromatin formation and silencing: round- ing up the usual suspects. Cell 108, 489–500. Riddle, N.C. and Birchler, J.A. (2003) Effects of reunited diverged regulatory hierarchies in allopolyploids and species hybrids. Trends in Genetics 19, 597–600. Rieger, R., Michaelis, A. and Green, M.M. (1976) Glossary of Genetics and Cytogenetics. Springer-Verlag, New York. Rieseberg, L.H. (1995) The role of hybridization in evolution: old wine in new skins. American Journal of Botany 82, 944–953. Rieseberg, L.H. (2001) Polyploid evolution: keeping the peace at genomic reunions. Current Biology 13, R925–R928. Rieseberg, L.H. and Wendel, J.F. (1993) Introgression and its consequences in plants. In: Harrison, R. (ed.) Hybrid Zones and the Evolutionary Process. Oxford University Press, New York, pp. 70–109. Roose, M.L. and Gottlieb, L.D. (1976) Genetic and biochemical consequences of polyploidy in Tragopogon. Evolution 30, 818–830. Schranz, M.E. and Osborn, T.C. (2000) Novel flowering time variation in the resynthesized polyploid Brassica napus. Journal of Heredity 91, 242–246. Senchina, D., Alvarez, I., Cronn, R., Liu, B., Rong, J., Noyes, R., Paterson, A.H., Wing, R.A., Wilkins, T.A. and Wendel, J.F. (2003) Rate variation among nuclear genes and the age of polyploidy in Gossypium. Molecular Biology and Evolution 20, 633–643. Shaked, H., Kashkush, K., Ozkan, H., Feldman, M. and Levy, A.A. (2001) Sequence elimination and cyto- sine methylation are rapid and reproducible responses of the genome to wide hybridization and allopolyploidy in wheat. The Plant Cell 13, 1749–1759. Simillion, C., Vandepoele, K., Van Montagu, M.C.E., Zabeau, M. and Van de Peer, Y. (2002) The hidden duplication past of Arabidopsis thaliana. Proceedings of the National Academy of Sciences USA 99, 13627–13632. Skalicka, K., Lim, K.Y., Matyasek, R., Koukalova, B., Leitch, A.R. and Kovarik, A. (2003) Rapid evolution of parental rDNA in a synthetic tobacco allotetraploid line. American Journal of Botany 90, 988–996. Soltis, D.E. and Soltis, P.S. (1993) Molecular data and the dynamic nature of polyploidy. Critical Reviews in Plant Science 12, 243–273. Soltis, D.E. and Soltis, P.S. (1995) The dynamic nature of polyploid genomes. Proceedings of the National Academy of Sciences USA, 92, 8089–8091. Soltis, D.E. and Soltis, P.S. (1999) Polyploidy: origins of species and genome evolution. Trends in Ecology and Evolution 9, 348–352. Soltis, P.S. and Soltis, D.E. (2000) The role of genetic and genomic attributes in the success of polyploids. Proceedings of the National Academy of Sciences USA 97, 7051–7057. Soltis, D.E., Soltis, P.S. and Tate, J.A. (2004) Advances in the study of polyploidy since plant speciation. New Phytologist 161, 173–191. Song, K., Lu, P., Tang, K. and Osborn, T.C. (1995) Rapid genome change in synthetic polyploids of Brassica and its implications for polyploid evolution. Proceedings of the National Academy of Sciences USA 92, 7719–7723. Stebbins, G.L. (1947) Types of polyploids: their classification and significance. Advances in Genetics 1, 403–429. Stebbins, G.L. (1950) Variation and Evolution in Plants. Columbia University Press, New York. Stebbins, G.L. (1971) Chromosomal Evolution in Higher Plants. Edward Arnold, London. Stephens, S.G. (1951a) Evolution of the gene: ‘homologous’ genetic loci in Gossypium. Cold Spring Harbor Symposium on Quantitative Biology 16, 131–140. Stephens, S.G. (1951b) Possible significance of duplication in evolution. Advances in Genetics 4, 247–265. Taylor, J.S., Van de Peer, Y., Braasch, I. and Meyer, A. (2001) Comparative genomics provides evidence for an ancient genome duplication event in fish. Proceedings of the Royal Society of London Series B: Biological Sciences 356, 1661–1679. Thompson, J.N., Nuismer, S.L. and Merg, K. (2004) Plant polyploidy and the evolutionary ecology of plant/animal interactions. Biological Journal of the Linnean Society 82, 511–519. Trinti, F. and Scali, V. (1996) Androgenetics and triploids from an interacting parthenogenetic hybrid and its ancestors in stick insects. Evolution 50, 1251–1258. Vision, T.J., Brown, D.G. and Tanksley, S.D. (2000) The origins of genomic duplications in Arabidopsis. Science 290, 2114–2117.

Polyploidy and evolution in plants 117 Vogel, J.C., Barrett, J.A., Rumsey, F.J. and Gibby, M. (1999) Identifying multiple origins in polyploid homo- sporous pteridophytes. In: Hollingsworth, P.M., Bateman, R.M. and Gornall, R.J. (eds) Molecular Systematics and Plant Evolution. Taylor & Francis, London, pp. 101–117. Watanabe, K., Peloquin, S.J. and Endo, M. (1991) Genetic significance of mode of polyploidization: somatic doubling or 2n gametes. Genome 34, 28–34. Wendel, J.F. (2000) Genome evolution in polyploids. Plant Molecular Biology 42, 225–249. Wendel, J.F. and Cronn, R.C. (2003) Polyploidy and the evolutionary history of cotton. Advances in Agronomy 78, 139–186. Wendel, J.F., Schnabel, A. and Seelanan, T. (1995) Bidirectional interlocus concerted evolution following allopolyploid speciation in cotton (Gossypium). Proceedings of the National Academy of Sciences USA 92, 280–284. Werth, C.R. and Windham, M.D. (1991) A model for divergent, allopatric speciation of polyploid pterido- phytes resulting from silencing of duplicate-gene expression. American Naturalist 137, 515–526. Wessler, S.R. (1996) Plant retrotransposons: turned on by stress. Current Biology 6, 959–961. Wessler, S.R. (2001) Plant transposable elements: a hard act to follow. Plant Physiology 125, 149–151. Wilson, W.A., Harrington, S.E., Woodman, W.L., Lee, M., Sorrells, M.E. and McCouch, S.R. (1999) Inferences on the genome structure of progenitor maize through comparative analysis of rice, maize and the domesticated Panicoids. Genetics 153, 453–473. Wolfe, K.H. and Shields, D.C. (1997) Molecular evidence for an ancient duplication of the entire yeast genome. Nature 387, 708–713. Wolffe, A.P. and Matzke, M.A. (1999) Epigenetics: regulation through repression. Science 286, 481–486. Yoder, J.A., Walsh, C.P. and Bestor, T.H. (1997) Cytosine methylation and the ecology of intragenomic para- sites. Trends in Genetics 13, 335–340. Zhao, X.-P., Si, Y., Hanson, R.E., Crane, C.F., Price, H.J., Stelly, D.M., Wendel, J.F. and Paterson, A.H. (1998) Dispersed repetitive DNA has spread to new genomes since polyploid formation in cotton. Genome Research 8, 479–492.

8 Crucifer evolution in the post-genomic era Thomas Mitchell-Olds,1 Ihsan A. Al-Shehbaz,2 Marcus A. Koch3 and Tim F. Sharbel4 1Department of Genetics and Evolution, Max Planck Institute of Chemical Ecology, Hans-Knoll Strasse 8, 07745, Jena, Germany; 2Missouri Botanical Gardens, PO Box 299, St Louis, MO 63166-0299, USA; 3Heidelberg Institute of Plant Sciences, Biodiversity and Plant Systematics, Im Neuenheimer Feld 345, D-69129 Heidelberg, Germany; 4Laboratoire IFREMER de Genetique et Pathologie, 17390 La Tremblade, France Introduction regions of the northern hemisphere (Al- Shehbaz, 1984; Appel and Al-Shehbaz, The worldwide genomics initiative in 2003). It includes important crop plants cul- Arabidopsis thaliana has facilitated a renais- tivated worldwide as vegetables and orna- sance in evolutionary studies of the mentals and as sources of cooking and Brassicaceae. More than 20,000 published industrial oils, condiments and forage (Koch papers examine aspects of Arabidopsis biol- et al., 2003a). One species, Arabidopsis ogy, providing detailed understanding of thaliana (thale cress), is the model flowering molecular biology, genetics, biochemistry, plant in nearly every field of experimental physiology and development of this model biology, and its entire genome has recently plant. Increasingly, comparative analyses been sequenced (The Arabidopsis Genome (Hall et al., 2002a; Mitchell-Olds and Clauss, Initiative, 2000). The Brassicaceae are easily 2002) build upon Arabidopsis genomics to recognized by having flowers with four elucidate biology of crucifer species, and the petals forming a cross (sometimes reduced evolutionary processes which influence or lacking), six stamens (the outer two being adaptation and diversification in the shorter than the inner four, although some- Brassicaceae. Here we review aspects of sys- times only two or four stamens are present), tematics, speciation and functional variation and often a two-valved capsule with a sep- in this important plant family. tum dividing it into two chambers. Systematics and Taxonomy Family limits and relatives Family characteristics and importance The Brassicaceae were previously thought to have evolved in the New World, either through The mustard family (Brassicaceae or the putatively basal tribe Thelypodieae Cruciferae) includes some 340 genera and (Stanleyeae) from the Capparaceae subfamily about 3350 species distributed worldwide, Cleomoideae (Janchen, 1942; Al-Shehbaz, especially in the temperate and alpine 1973; Takhtajan, 1997), or to share a com- © CAB International 2005. Plant Diversity and Evolution: Genotypic and 119 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

120 T. Mitchell-Olds et al. mon ancestry with that subfamily (Al- (flattened at a right angle to the septum), a Shehbaz, 1985). Dvorák (1973) suggested an feature not found among the Old World alternative origin in the Old World and con- Cleomaceae. Although published molecular sidered the tribe Hesperideae as the link data are based on two, highly variable between the Brassicaceae and Cleomoideae. species, Aethionema grandiflorum and Hauser and Crovello (1982) tested these Aethionema saxatile, phylogenetic studies in hypotheses and favoured a New World origin progress (Menke, personal communication) from the Cleomoideae. should resolve its monophyly, determine its nearest relatives and shed light on what Based on a small sample (two genera of makes it basal in the Brassicaceae. Brassicaceae and four of Capparaceae) and a set of only 16 characters, Judd et al. (1994) ITS and other molecular markers at the tribal concluded in a cladistic morphological study level that the Brassicaceae is nested within the Capparaceae and that the two families should All major classification systems (Prantl, 1891; be united in one, Brassicaceae. Although this Hayek, 1911; Schulz, 1936; Janchen, 1942), merger was followed (Angiosperm Phylogeny which divide the family into 4–19 tribes, are Group, 1998) or recommended (Appel and based on a limited number of morphological Al-Shehbaz, 2003), thorough molecular and characters and do not recognize convergence morphological data (Rodman et al., 1996, as a factor in the evolution of Brassicaceae. 1998) suggested a closer relationship of the Molecular studies (Price et al., 1994; Zunk et Brassicaceae to Cleome than to the rest of al., 1996, 1999; Koch et al., 1999a, 2000, Capparaceae. Hall et al. (2002b), who con- 2001a, 2003a; Bailey et al., 2002; Koch, 2003; ducted detailed molecular studies using the O’Kane and Al-Shehbaz, 2003) have demon- chloroplast regions trnL-trnF and ndhF, advo- strated the polyphyly and artificiality of cated that three, well-supported mono- almost all tribes recognized in these systems. phyletic families (Brassicaceae, Cleomaceae, For example, Capsella and Arabidopsis, treated Capparaceae) should be recognized, with the by Schulz (1936) in unrelated tribes, have Brassicaceae and Cleomaceae as sister fami- been shown (Koch et al., 2001a; O’Kane and lies sharing a common ancestor. Al-Shehbaz, 2003) to be very closely related. Numerous other examples can be cited, and The remarkable morphological similari- the interested reader should consult Koch et ties between Thelypodieae and Cleomaceae al. (2003a). (exserted stamens equal in length, sessile stigmas, linear fruits, dense racemes, long Extensive molecular data, summarized by gynophores, linear anthers coiled at dehis- Warwick and Black (1997a,b), show that the cence, to name some) all appear to be the Brassiceae, characterized by segmented result of convergence rather than synapo- fruits and/or conduplicate cotyledons (Appel morphy. Using internal transcribed spacer of and Al-Shehbaz, 2003), is the only mono- nuclear ribosomal DNA (ITS) data, Warwick phyletic group among Schulz’s (1936) 19 et al. (2002) have clearly demonstrated that tribes. However, generic limits, as tradition- the Thelypodieae and many New World gen- ally recognized (Gómez-Campo, 1999), era form an unresolved, rather advanced, remain problematic, and a major revision of terminal polytomy. These findings agree the boundaries of most genera is needed in with those of Galloway et al. (1998). light of molecular data (Koch et al., 2003a). All broad-based molecular studies of Based on ITS sequence data (Kropf et al., Brassicaceae (Zunk et al., 1996, 1999; 2002; Koch et al., 2003a; O’Kane and Al- Galloway et al., 1998; Koch et al., 2001a, Shehbaz, 2003) and other markers 2003a) demonstrated that Aethionema occu- (Galloway et al., 1998; Koch et al., 2000, pies the most basal position in the family. 2001a, 2003a), several monophyletic clades Aethionema, a highly polymorphic genus of are readily recognized (Fig. 8.1). One such 50–60 species, is distributed primarily in group, the Brassica alliance (c. 550 species), Turkey and the Middle East (Appel and Al- Shehbaz, 2003). It has angustiseptate fruits

Crucifer evolution in the post-genomic era 121 100 Ionopsidium prolongoi Lepidieae 100 Ionopsidium abulense Arabideae Cochlearia danica Lepidieae Cochlearia pyrenaica Sisymbrieae Brassiceae 100 Arabis alpina EUROPE Sisymbrieae Arabis alpina AFRICA Arabideae 95 E 100 Arabis procurrens Arabis pumila Lepidieae Sisymbrieae Arabis blepharophylla Arabideae 96 100 73 Arabis hirsuta Sisymbrieae Lepidieae Aubrieta deltoidea Arabideae Aubrieta deltoidea Sisymbrieae F Hesperideae Lepidieae Arabis turrita Microthlaspi perfoliatum 91 100 Thlaspi arvense Alliaria petiolata 95 84 Sinapis alba 100 Raphanus sativus Sisymbrium irio 89 Fourraea alpina 68 A 100 Barbarea vulgaris Rorippa palustris 100 Rorippa amphibia Cardamine amara 100 Cardamine penziesii B 100 Cardamine rivularis 84 Lepidium campestre 88 100 Arabidopsis lyrata ssp . petraea SWE 100 Arabidopsis lyrata ssp . petraea GER 100 92 Arabidopsis lyrata ssp . lyrata USA Arabidopsis halleri Arabidopsis thaliana 100 Turritis glabra C D 97 67 Olimarabidopsis pumila 100 Olimarabidopsis cabulica Olimarabidopsis pumila Capsella rubella 87 Arabis lyallii 100 Arabis parishii Boechera 100 Arabis drummondii Arabis drummondii 100 Arabis divaricarpa Arabis lignifera 95 Halimolobus perplexa 100 Crucihimalaya himalaica Crucihimalaya wallichii Mathiola incana Aethionema grandiflora 0.05 substitutions/site Fig. 8.1. Neighbour-joining distance tree based on matK and Chs sequences (Koch et al., 2001a). Percentage bootstrap values from 1000 replicates are shown on each branch. Of the six monophyletic clades discussed in the text, four are represented here: Brassica alliance (Sinapis, Raphanus, Alliaria, Thlaspi, Microthlaspi), Arabidopsis alliance (Arabidopsis, Boechera, Halimolobos, Capsella, Olimarabidopsis, Crucihimalaya, Turritis), Arabis alliance (Arabis and Aubrieta) and Cardaminine alliance (Cardamine, Barbarea, Rorippa). Tribal assignments are given on the right. Approximate divergence dates (million years before present) for nodes A–F are: A, 16–21; B, 13–19; C, 19–25; D, 10–14; E, 15–17; and F, 26–32. Several taxa mentioned in the text are not shown in this figure: Brassica is near Raphanus and Sinapis, and Leavenworthia is related to Barbarea. Reproduced, with permission, from Koch et al. (2001a).

122 T. Mitchell-Olds et al. includes all members of the tribe Brassiceae, (Mummenhoff, 1995; Bruggemann, 2000; Mummenhoff et al., 2001) and morphologi- the expanded New World Thelypodieae and cal (Al-Shehbaz et al., 2002) evidence. The genera Acanthocardamum, Delpinophytum, Sisymbrium sensu Warwick et al. (2002), and Winklera, Stubendorffia, Megacarpaea and Biscutella should be studied in connection Thlaspi and its segregates (Mummenhoff et with this clade, and it is likely that some, if not all, are allied to Lepidium. The clade has al., 1997a,b; Koch and Mummenhoff, 2001). angustiseptate fruits, two ovules per ovary and simple or no trichomes. The vast majority of genera in this clade Another clade is the Cardaminine have species either glabrous or with simple alliance (c. 350 species), which includes Cardamine (including Dentaria and Iti), trichomes, though branched trichomes Armoracia, Barbarea, Iodanthus, Leavenworthia, Nasturtium, Planodes, Rorippa, Selenia, and apparently evolved independently a few perhaps several Himalayan genera. Members of this clade have accumbent times, especially in some South American cotyledons, latiseptate fruits, dissected or compound leaves, and simple or no tri- Thelypodieae and southern African chomes, and they frequently occupy aquatic, wet or mesic habitats (Franzke et al., 1998; Sisymbrium. ITS sequence data provide little Mitchell and Heenan, 2000; Sweeney and Price, 2000). resolution within the Thelypodieae. This Another alliance much in need of further suggests a relatively recent evolution of the studies includes Alyssum and related genera (c. 220 species). The remaining 1200 species group and insufficient time for ITS of Brassicaceae perhaps fall within these six major clades. sequences to diverge, in agreement with the Molecular data and generic delimitation absence of morphological differentiation Differences in fruit morphology and embryo and the difficulty in recognizing individual position have been used extensively in the delimitation of genera (Al-Shehbaz, 1984; genera in the group (Warwick et al., 2002). Rollins, 1993; Appel and Al-Shehbaz, 2003). However, such differences are often overem- The second clade (c. 300 species), desig- phasized and vegetative and floral characters are largely neglected. Sequence comparisons nated herein as the Arabidopsis alliance, of relatively rapidly evolving regions such as ITS and ndhF suggest that fruit morphology includes Boechera sensu Al-Shehbaz (2003b), and embryo type are subject to frequent con- vergence. The high degree of sequence simi- the halimolobine clade (Halimolobos, Mancoa, larity among taxa with different fruits (Fig. 8.2) emphasizes the rapid rate at which Pennellia, Sphaerocardamum) sensu Bailey et al. major changes in fruits and embryo mor- phology can occur. It is plausible that the (2002), Arabidopsis and its recent segregates number of genes responsible for changes in fruit shape and embryo position may be rela- (Al-Shehbaz et al., 1999; O’Kane and Al- tively small, thus allowing rapid bursts of evolution uncoupled from other aspects of Shehbaz, 2003), Camelina, Capsella, morphology or molecular markers. Cusickiella, Neslia, the polycolpate clade (Dimorphocarpa, Dithyrea, Lyrocarpa, Nerisyrenia, Paysonia, Physaria including Lesquerella, Synthlipsis) sensu O’Kane and Al- Shehbaz (2003), Pachycladon sensu Heenan et al. (2002), Erysimum and Transberingia (Beringia sensu Price et al., 2001). This clade is characterized by the preponderance of forked and/or stellate trichomes. A third clade, the Arabis alliance (c. 450 species), involves Arabis (but excludes Boechera, Turritis and Fourraea), Draba (including Erophila and Drabopsis) and Aubrieta. This group has branched tri- chomes, accumbent cotyledons, and often latiseptate fruits (flattened parallel to the septum). Although the same combination of characters is found in Boechera, the similari- ties are superficial and result from conver- gence rather than synapomorphy (Koch et al., 1999a; Koch and Al-Shehbaz, 2002). A fourth monophyletic clade (c. 250 species) contains Lepidium including Cardaria, Coronopus and Stroganowia. This clade is well supported by molecular

Crucifer evolution in the post-genomic era 123 Fig. 8.2. Some examples of fruit morphologies ITS and ndhF sequence data, and they differ among relatives of Arabidopsis. Three of these only in one base substitution for each species are members of the Arabidopsis alliance marker (Price, personal communication). (Halimolobos perplexa, Capsella rubella and Arabis They are now recognized as one genus, drummondii [syn. Boechera stricta]). Reproduced, Tropidocarpum (Al-Shehbaz and Price, 2001; with permission, from Koch et al. (1999a). Al-Shehbaz, 2003a). More than 225 genera (60%) of The other extreme involves taxa that are Brassicaceae include four species or less very difficult to distinguish based on fruit (Appel and Al-Shehbaz, 2003). It is likely that morphology but their ITS and other this number will be reduced significantly sequence data show enormous divergence with detailed molecular studies, as illustrated well supported by high bootstrap values. by two examples. First, based on chloroplast Three classic examples are given. First, ITS DNA restriction site variation (Warwick and (Koch et al., 1999a; O’Kane and Al-Shehbaz, Black, 1994) and ITS sequence data (Crespo 2003) and chalcone synthase and alcohol et al., 2000), Euzumodendron and Boleum are dehydrogenase data (Koch et al., 2000) clearly well nested within and hardly differ from demonstrate that the North American Arabis Vella. Although these genera have very dif- sensu Rollins (1993) is polyphyletic and that ferent fruit morphology, all include shrubs all except 15 of the 80 species should be with x=17 and united paired stamens, a assigned to Boechera (Al-Shehbaz, 2003b). character combination not found elsewhere Although the fruits of Boechera and Arabis are in the tribe Brassiceae. Molecular data, cou- quite similar, significant morphological differ- pled with unique morphology and cytology, ences have been found, and the two genera led Warwick and Al-Shehbaz (1998) to unite are clearly unrelated (Al-Shehbaz, 2003b). these genera in one, Vella, a position now Second, ITS data (Koch et al., 1999a; O’Kane widely accepted. and Al-Shehbaz, 2003) strongly support split- ting the c. 60 species of Arabidopsis into several Another classic example involves the genera differing significantly in characters Chilean Agallis and Californian Twisselmannia other than fruit morphology (Al-Shehbaz et and Tropidocarpum. These genera also have al., 1999; Al-Shehbaz and O’Kane, 2002). dramatically different fruit morphology but Finally, molecular data on Thlaspi sensu lato are indistinguishable in floral and vegetative (Mummenhoff and Koch, 1994; characters. All three are basically identical in Mummenhoff et al., 1997a,b; Koch and Mummenhoff, 2001) provide ample support for most of Meyer’s (1973, 1979) segregates of the genus based on seed-coat anatomy. In conclusion, extreme care should be taken in evaluating generic limits based solely on fruit and/or embryo morphology because both structures are highly homo- plastic. Monotypic genera are always sus- pect, and they should not be erected without thorough molecular studies using both nuclear and chloroplast markers. Speciation and Differentiation Hybridization and polyploidization Since at least Stebbins’s (1940) work it has been clear that hybridization and poly- ploidization have played an important role in

124 T. Mitchell-Olds et al. the evolution of angiosperms (Ehrendorfer, ferent base chromosome numbers, such as 1980; Soltis and Soltis, 1999). This is also true among Brassica species (U, 1935; Erickson et for the Brassicaceae. The majority of taxa are al., 1983; Palmer et al., 1983) or Arabidopsis recent or ancient polyploids with largely (Mummenhoff and Hurka, 1994, 1995; duplicated genomes (Appel and Al-Shehbaz, O’Kane and Al-Shehbaz, 1997; O’Kane et al., 2003; Koch et al., 2003a), and chromosome 1997; Sall et al., 2003). In the latter case arti- numbers varying from n = 4 in some species ficial hybrids have been obtained between A. of the North American Physaria and thaliana (n = 5) and other Arabidopsis species Australian Stenopetalum to n = 128 in with n = 8 (Comai et al., 2000, 2003; Cardamine concatenata (as Cardamine laciniata, Nasrallah et al., 2000), which can be see Al-Shehbaz, 1984). Polyploid complexes analysed genetically and physiologically. with reticulate evolutionary patterns are found frequently and date back to different Species divergence time periods. A late Pleistocene history of spe- ciation has been characterized in genera such Evolutionary studies of species differences as Cochlearia (Koch et al., 1996, 1999b; Koch, employ two main approaches: (i) genetic 2002) and Central European Cardamine mapping of trait differences (Schemske and (Franzke and Hurka, 2000; Marhold et al., Bradshaw, 1999; Rieseberg et al., 2003); 2002) or closely related Nasturtium and and (ii) divergence population genetics Rorippa (Bleeker et al., 1999; Bleeker, 2003). (Kliman et al., 2000). These approaches Other genera such as Yinshania (Koch and Al- provide complementary insights into the Shehbaz, 2000) and Draba (Koch and Al- evolutionary processes and functional Shehbaz, 2002) have been identified as changes during speciation. Several out- polyploid complexes with speciation standing studies of quantitative genetic dif- processes dating back to the Tertiary (mostly ferences between sister species have been Pliocene). Pleistocene differentiation, reported (e.g. Bradshaw and Schemske, reported for North American Boechera (Dobes 2003; Rieseberg et al., 2003). Many studies et al., 2004), was also greatly affected by vari- have examined quantitative trait loci (QTL) ous glacial and interglacial cycles. among cultivars or subspecies in Brassica (e.g. Lan and Paterson, 2000; Schranz et Within the tribe Brassiceae (an appar- al., 2002). To date, QTL mapping from ently monophyletic lineage; see above) the undisturbed natural populations of cru- genome underwent extensive duplications cifers has not yet been reported, although a of large genomic regions (Kowalski et al., linkage map of molecular markers is now 1994; Lagercrantz and Lydiate, 1996; available for Arabidopsis lyrata (Kuittinen et Lagercrantz, 1998), which led to the conclu- al., 2004). sion that ‘diploid’ species such as Brassica oleracea and Brassica rapa (n = 9 and n = 10, Divergence population genetics compares respectively) are ancestral hexaploids. The allele genealogies for multiple loci among impact of hybridization on several traits has several species (Kliman et al., 2000). Early in been demonstrated within Lepidium for speciation, alleles will be shared between sis- flower morphology (Bowman et al., 1999; ter species. In contrast, species-specific poly- Lee et al., 2002) or life history traits within morphisms predominate later in speciation. Capsella (Hurka and Neuffer, 1997). Detailed Between these two extremes, polymor- studies show hybridization between different phisms are especially informative at an parental taxa with identical chromosome intermediate stage of speciation, when some numbers accompanied by subsequent poly- genes have shared alleles, while other loci ploidization (e.g. Capsella: Mummenhoff and have private alleles. These differences Hurka, 1990; Draba: Brochmann et al., among loci may be attributable to natural 1992; Cardamine: Urbanska et al., 1997; selection on functional differences, random Microthlaspi: Koch and Hurka, 1999). variation across the genome, or occasional However, there are also detailed descrip- gene flow between species. tions of hybridization between taxa with dif-

Crucifer evolution in the post-genomic era 125 Ramos-Onsins et al. (2004) examined torical and population processes has nucleotide polymorphism at eight unlinked emerged as a central focus. For these pur- loci in species-wide samples of four taxa in poses molecular markers have been utilized Arabidopsis, comparing the highly-inbreed- to elaborate geographic distribution pat- ing A. thaliana with the closely related out- terns of presumably neutral loci. Maternally crossing Arabidopsis halleri, A. lyrata ssp. inherited DNA markers (e.g. the plastome of petraea and A. lyrata ssp. lyrata. Average levels most angiosperms, see Harris and Ingram, of nucleotide polymorphism were highest in 1991; Reboud and Zeyl, 1994) can be used ssp. petraea and lowest in ssp. lyrata, presum- to trace maternal lineages. A variety of ably reflecting differences in effective popu- nuclear markers are available (Sunnucks, lation size between subspecies. This 2000), including co-dominant isozymes and relatively low nucleotide polymorphism in microsatellites, dominant markers such as ssp. lyrata may reflect a population bottle- RAPDs (random amplified polymorphic neck during Pleistocene colonization of DNAs) or AFLPs (amplified fragment length North America (Wright et al., 2003; Ramos- polymorphisms), nuclear DNA sequences Onsins et al., 2004). Population genetic such as ITS1 and ITS2, single-copy nuclear analysis suggests that introgression has genes (Savolainen et al., 2000; Ramos-Onsins occurred between A. halleri and A. lyrata ssp. et al., 2004), and SNPs (single nucleotide petraea subsequent to speciation (Ramos- polymorphisms, e.g. Brumfield et al., 2003; Onsins et al., 2004). Schmid et al., 2003). Gamete recognition genes play an impor- Among crucifers, phylogeographic stud- tant role in speciation (Howard, 1999), and ies are available from all regions of their dis- they experience rapid adaptive evolution in tribution and on very different geographic a number of animal systems (Swanson and scales. Worldwide phylogeographies were Vacquier, 2002a,b). Similar rapidly evolving conducted on Lepidium (Mummenhoff et al., mate recognition genes have also been sug- 2001) and Arabidopsis (Ramos-Onsins et al., gested in plants. Binding of pollen to A. 2004), and central European phylogeogra- thaliana stigmas is controlled by an oleosin- phies were focused on Biscutella domain protein belonging to a small gene (Tremetsberger et al., 2002), Arabidopsis family (Mayfield et al., 2001). Genomic com- (Sharbel et al., 2000), Cochlearia (Koch, 2002; parisons from A. thaliana, Boechera stricta Koch et al., 2003b), Microthlaspi (Koch et al., (formerly Arabis drummondii) and Brassica 1998; Koch and Bernhardt, 2004), oleracea show rapid evolution due to gene Hornungia (as Pritzelago) (Kropf et al., 2003) duplication and deletion, accelerated amino and alpine Draba (Widmer and Baltisberger, acid substitution, and insertions and dele- 1999). These studies elucidated colonization tions within the coding region (Mayfield et routes from refugial areas into formerly al., 2001; Schein et al., 2004). These results glaciated areas of north and central Europe, are consistent with a hypothesized function and they also identified Pleistocene refugial in species recognition. Further functional areas in the Iberian Peninsula, northern and population genetic analyses are Italy, or the Balkans (Comes and Kadereit, required to elucidate causes of rapid evolu- 1998). tion in this gene family. A complex speciation and migration sce- Phylogeography nario in Draba was elaborated for North, Central, and South America (Koch and Al- Since its development by Avise et al. (1987), Shehbaz, 2002). This genus also shows strong phylogeography has become an increasingly affinities to high alpine regions (e.g. in the important field of research within biogeog- Alps, Scandinavia, the Himalayas, Rocky raphy. Originally the aim was to describe the Mountains and the Andes) and therefore its distribution of genetic variation in space and evolution was influenced by glacial and inter- time. More recently, understanding of his- glacial periods throughout the Pleistocene. Phylogeographic disjunction between North and South America was demonstrated in

126 T. Mitchell-Olds et al. Noccaea (Koch and Al-Shehbaz, 2004) and However, depending on the type of seed Halimolobus/Sphaerocardamon (Bailey et al., bank and reproductive biology of the 2002). Studies on the Chinese Yinshania species, the seed bank of a particular popu- (Koch and Al-Shehbaz, 2000) and several lation may be essential for recruitment and Australian and New Zealand mustards establishment of new cohorts. Depending on (Bleeker et al., 2002; Heenan and Mitchell, the spatial genetic structure of the subpopu- 2003) demonstrated rapid range expansion lations (surface and aboveground popula- and evolutionary radiation. Detailed molecu- tions stored as seeds or other diaspores), lar studies in Boechera demonstrated the major changes in the genetic constitution of impact of glacial and interglacial cycles on plant populations may occur during their reticulate evolution and radiation, as well as history (Levin, 1990). Of less than ten stud- migration and extinction (Sharbel and ies focusing on seed-bank genetics, two are Mitchell-Olds, 2001; Koch et al., 2003a; on members of the Brassicaceae (Cabin, Dobes et al., 2004). This genus is greatly 1996; Koch et al., 2003c). affected by asexual reproduction (see below), which has important impacts on patterns of Comparative physiology and development genetic and phenotypic diversity. Differentiation and wild populations Arabidopsis genomics has revolutionized our understanding of plant biology and enabled Many studies have focused on speciation functional analyses of many scientifically and processes and phylogenetic relationships at economically important traits. Nevertheless, or below the species level. Several phenotypic some mechanistic and evolutionary hypothe- traits and characters have been investigated, ses are better addressed using wild relatives either to learn more about their evolutionary of A. thaliana, which are not confined to significance or to elucidate plasticity and eco- well-watered, temperate, ephemeral envi- logical relevance. These traits include local ronments that follow agricultural distur- adaptation across climatic gradients in Arabis bance. In particular, understanding of fecunda (McKay et al., 2001) and Capsella resistance to drought, heavy metals and a (Neuffer and Hurka, 1999), survivorship in broad range of microbial pathogens can Boechera (as Arabis) laevigata (Bloom et al., benefit from comparative genomics using 2001), glucosinolate accumulation during the wild relatives of A. thaliana. plant/insect interaction in A. thaliana (Kliebenstein et al., 2001), herbivore resis- Water availability is fundamental to tance (Agrawal et al., 2002; Kroymann et al., almost all aspects of plant physiology (Bray, 2003; Weinig et al., 2003b), pollination 1997), and plant distribution and abun- (Strauss et al., 1999), as well as leaf morphol- dance in agricultural and natural ecosystems ogy, flowering and maternal effects in Capsella are largely determined by water availability. (Neuffer, 1989, 1990; Neuffer and Koch, Although Arabidopsis genomics has enabled 1996). Detailed analyses of host–pathogen much progress in understanding responses interaction have been presented in Boechera to drought (Abe et al., 2003; Boyce et al., (Arabis) (Roy, 2001) and A. thaliana (Stahl et 2003; Cheong et al., 2003; McKay et al., al., 1999; Tian et al., 2002, 2003). Several 2003; Oono et al., 2003), A. thaliana is con- studies have focused on the evolutionary sig- fined to mesic habitats, and therefore pro- nificance of phenotypic plasticity and reaction vides an incomplete view of adaptive norms in A. thaliana (Pigliucci and Byrd, changes in water relations. In contrast, other 1998; Pigliucci et al., 1999; Pigliucci and crucifers are adapted to desert, mesic and Marlow, 2001; Pollard et al., 2001; Weinig et aquatic habitats (Rollins, 1993; Bressan et al., al., 2002, 2003a; Ungerer et al., 2003). 2001; Mitchell-Olds, 2001; Xiong and Zhu, 2002) and display a broad range of adaptive Most population studies neglected the differences in water use efficiency (McKay et genetic diversity stored in the soil seed bank. al., 2001) and salt tolerance, which can be elucidated by comparative genomics.

Crucifer evolution in the post-genomic era 127 Resistance to deleterious effects of heavy genomic sequences between related species metals occurs among several mustards, (Cooper and Sidow, 2003). Phylogenetic including Thlaspi (Persans et al., 2001; footprinting considers a small number of Lombi et al., 2002; Pineros and Kochian, distant evolutionary comparisons, whereas 2003), A. halleri (Sarret et al., 2002; Bert et phylogenetic shadowing examines a set of al., 2003) and Alyssum (Kerkeb and Kramer, closely related species (Boffelli et al., 2003). 2003). Both natural variation (Sarret et al., Koch et al. (2001b) applied this approach to 2002; Bert et al., 2003) and genomic meth- promoters of chalcone synthase (Chs) in 22 ods (Persans et al., 2001; Sarret et al., 2002; crucifer species at increasing evolutionary Pineros and Kochian, 2003) have been distances from Arabidopsis. They identified employed to elucidate the physiological and conserved regions of the Chs promoter and molecular basis of heavy metal tolerance. verified their functional importance by expressing promoter fragments from six Crucifers are attacked by a wide variety of crucifer species in Arabidopsis protoplasts. pests, offering potential for functional and Hong et al. (2003) examined cis-regulatory evolutionary genomic studies of biotic interac- sequences of the Agamous locus in 29 tions. In addition to the extensive literature Brassicaceae species. Although they identified on microbial pathogens of A. thaliana (e.g. motifs conserved among taxa, some previ- Kunkel and Brooks, 2002; Nurnberger and ously identified, functionally important LFY Brunner, 2002; Farmer et al., 2003; Shah, and WUS binding sites were not highly con- 2003), comparative genomics in the served. Functional significance of several Brassicaceae will allow studies of pathogens conserved motifs was verified by reporter whose host range does not include Arabidopsis, gene analyses. Phylogenetic footprinting can including Puccinia (Basidiomycetes; Roy, identify important regulatory regions in 1993), Pyrenopeziza brassicae (Discomycetes; many species, and is now being applied to Singh et al., 1999) and Leptosphaeria maculans large-scale analyses in grasses (Inada et al., (Loculoascomycetes; Mitchell-Olds et al., 2003) and many animals (Cooper and 1995). Furthermore, although earlier studies Sidow, 2003). concluded that arbuscular mycorrhizal fungi do not colonize Brassicaceae, recent research Breeding systems has found intraradical hyphae, vesicles, coils and arbuscules formed by mycorrhizal fungi Self-incompatibility is ancestral in the in the roots of several Thlaspi species (Regvar Brassicaceae (Bateman, 1955; Kachroo et al., et al., 2003). 2002). Multiple independent evolutionary origins of self-compatibility have occurred in Some crucifer species have patterns of diploid crucifers, including Leavenworthia flower and inflorescence development that uniflora and Leavenworthia crassa (Liu et al., differ substantially from A. thaliana. For 1998), Capsella rubella, A. thaliana, Arabidopsis example, most species of Lepidium have cebennensis, Arabidopsis croatica and Boechera reduced petals and/or stamen number (Lee species (Mitchell-Olds, unpublished). et al., 2002). Hybridization and polyploidy Inbreeding is especially common in weedy have played a major role in Lepidium floral crucifers (Appel and Al-Shehbaz, 2003). evolution, apparently because of dominant Theory predicts that self-fertilization should mutations causing loss of lateral stamens. In reduce genetic variation by 50% in compari- addition, Shu et al. (2000) compared the son with complete outcrossing (Pollack, 1987; elongated inflorescence of A. thaliana with Charlesworth, 2003). Further reductions of the rosette-flowering crucifer Ionopsidium genetic variability in inbreeding species may acaule, where flowers are borne singly in the also occur because of effects of background axils of rosette leaves. In situ hybridization selection and genetic hitchhiking (Maynard suggested that orthologues of LEAFY may Smith and Haigh, 1974; Charlesworth et al., control evolutionary changes in inflores- 1993; Nordborg et al., 1996). Comparisons cence architecture. Comparative genomics can identify con- served regulatory elements by comparing

128 T. Mitchell-Olds et al. of nucleotide polymorphism in inbreeding Emasculation of apomictic plants results and outbreeding species of Leavenworthia in in sterility, hence apomictic individuals are North America, as well as Arabidopsis native also pseudogamous (i.e. fertilization is to Europe (Clauss and Mitchell-Olds, 2003; required for embryo development, although Wright et al., 2003; Ramos-Onsins et al., no genetic contribution is made by the 2004) support these predictions. Other male). Böcher (1951) showed that apomictic breeding systems, including wind pollina- diploid and triploid individuals produced tion, monoecy, dioecy and gynodioecy are tetraploid and hexaploid endosperms, occasionally found in the Brassicaceae respectively, and thus pseudogamy can lead (Appel and Al-Shehbaz, 2003). to autonomous endosperm formation. There is evidence that endosperm ploidy Asexual reproduction, or apomixis can be variable (Matzk et al., 2000; Sharbel, (Koltunow and Grossniklaus, 2003; unpublished), and it is likely that Richards, 2003; Spielman et al., 2003), endosperm fertilization in some apomictic occurs in Draba verna, Smelowskia calycina, lineages is also possible. Draba oligosperma and several species of Boechera (Böcher, 1951; Roy, 1995; Sharbel In addition to the chromosomal vari- and Mitchell-Olds, 2001; Appel and Al- ability in apomicts, aneuploid chromosome Shehbaz, 2003). The base chromosome fragments in diploid and triploid individu- number in Boechera is n = 7, and its asexual als of Boechera were found (Böcher, 1951). taxa also exhibit polyploidy (predominantly Subsequent work using flow cytometry and 3n) and aneuploidy (Böcher, 1951; Roy, chloroplast DNA sequencing has demon- 1995; Sharbel and Mitchell-Olds, 2001). strated that the aneuploid chromosome is Studies of polyploidy and apomixis have widely distributed among different geo- concentrated on the Boechera holboellii group, graphical locations and can be found in a complex composed of B. holboellii, B. stricta diverse haplotype backgrounds (Sharbel (syn. Arabis drummondi) and their hybrid and Mitchell-Olds, 2001). Studies using Boechera divaricarpa (Dobes et al., 2003, microsatellite markers have shown that a 2004). Interestingly, both polyploidy and similar chromosome fragment is involved aneuploidy have evolved repeatedly within with aneuploidy in genetically diverse the B. holboellii complex (Sharbel and apomictic lineages (Sharbel et al., unpub- Mitchell-Olds, 2001). In addition, Böcher lished), and karyological and DNA (1951) provided evidence for occasional sequencing work have demonstrated that diploid apomicts, an extremely rare condi- the aneuploid chromosome is a non- tion among asexual plants. recombining B chromosome that may undergo both structural and sequence Species of Boechera reproduce via degeneration. diplosporous apomixis (Böcher, 1951). Böcher (1951) demonstrated that in B. hol- The potential development of apomixis boellii originating from Greenland and technology for crop plant research could Alaska gametogenesis and embryo forma- have substantial impacts on agriculture tion can be extremely variable. Normal (Hoisington et al., 1999; van Dijk and van diploid meiosis, followed by typical dyad Damme, 2000), as it represents a method and pollen formation, occurs in sexual B. through which genetic heterozygosity and stricta (= Arabis drummondii), a predomi- hybrid vigour could be fixed and faithfully nantly selfing species that occasionally out- propagated. Consequently, there is consider- crosses (Roy, 1995). Furthermore, synaptic, able interest in deciphering the molecular partially synaptic and asynaptic microsporo- genetic and/or physiological mechanisms genesis (pollen formation) in diploid and behind apomixis expression in the B. holboel- triploid individuals are also possible lii complex. The genetic diversity and geo- (Böcher, 1951). Variability in chromosome graphic distribution of apomictic individuals structure and unequal sister chromatid within the B. holboellii complex imply that exchange are proposed as mechanisms lead- this group may have some predisposition to ing to the lack of synapsis in some lineages. expressing this form of reproduction, and

Crucifer evolution in the post-genomic era 129 that apomixis has been stable and successful repeated evolution of polyploidy within this during the post-Pleistocene history of North group may also be correlated with repeated America. If apomixis expression is associ- expression of this reproductive mode from ated with polyploid gene expression, then sexual ancestors. References Abe, H., Urao, T., Ito, T., Seki, M., Shinozaki, K. and Yamaguchi-Shinozaki, K. (2003) Arabidopsis AtMYC2 (bHLH) and AtMYB2 (MYB) function as transcriptional activators in abscisic acid signaling. Plant Cell 15, 63–78. Agrawal, A., Conner, J., Johnson, M. and Wallsgrove, R. (2002) Ecological genetics of an induced plant defense against herbivores: additive genetic variance and costs of phenotypic plasticity. Evolution 56, 2206–2213. Al-Shehbaz, I.A. (1973) The biosystematics of the genus Thelypodium Cruciferae. Contributions From the Gray Herbarium of Harvard University, 3–148. Al-Shehbaz, I.A. (1984) The tribes of Cruciferae (Brassicaceae) in the southeastern United States. Journal of the Arnold Arboretum 65, 343–373. Al-Shehbaz, I.A. (1985) The genera of Thelypodieae (Cruciferae; Brassicaceae) in the southeastern United States. Journal of the Arnold Arboretum 66, 95–111. Al-Shehbaz, I.A. (2003a) A synopsis of Tropidocarpum (Brassicaceae). Novon 13, 392–395. Al-Shehbaz, I.A. (2003b) Transfer of most North American species of Arabis to Boechera (Brassicaceae). Novon 13, 381–391. Al-Shehbaz, I.A. and O’Kane, S.J. Jr (2002) Taxonomy and phylogeny of Arabidopsis (Brassicaceae). In: Somerville, C.R. and Meyerowitz, E.M. (eds) The Arabidopsis Book. American Society of Plant Biologists, Rockville, Maryland, doi/10.1199/tab.0001. Available at: www.aspb.org/publications/arabidopsis/ Al-Shehbaz, I.A. and Price, R.A. (2001) The Chilean Agallis and Californian Tropidocarpum (Brassicaceae) are congeneric. Novon 11, 292–293. Al-Shehbaz, I.A., O’Kane, S.L. Jr and Price, R.A. (1999) Generic placement of species excluded from Arabidopsis (Brassicaceae). Novon 9, 296–307. Al-Shehbaz, I.A., Mummenhoff, K. and Appel, O. (2002) Cardaria, Coronopus, and Stroganowia are united with Lepidium (Brassicaceae). Novon 12, 5–11. Angiosperm Phylogeny Group (1998) An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85, 531–553. Appel, O. and Al-Shehbaz, I.A. (2003) Cruciferae. In: Kubitzki, K. and Bayer, C. (eds) The Families and Genera of Vascular Plants. Springer-Verlag, Berlin, pp. 75–174. Avise, J.C., Arnold, J., Ball, R.M., Bermingham, E., Lamb, T., Neigel, J.E., Reeb, C.A. and Saunders, N.C. (1987) Intraspecific phylogeography: the mitochondrial-DNA bridge between population genetics and systematics. Annual Review of Ecology and Systematics 18, 489–522. Bailey, C.D., Price, R.A. and Doyle, J.J. (2002) Systematics of the halimolobine Brassicaceae: evidence from three loci and morphology. Systematic Botany 27, 318–332. Bateman, A.J. (1955) Self-incompatibility systems in angiosperms. 3. Cruciferae. Heredity 9, 53–68. Bert, V., Meerts, P., Saumitou-Laprade, P., Salis, P., Gruber, W. and Verbruggen, N. (2003) Genetic basis of Cd tolerance and hyperaccumulation in Arabidopsis halleri. Plant and Soil 249, 9–18. Bleeker, W. (2003) Hybridization and Rorippa austriaca (Brassicaceae) invasion in Germany. Molecular Ecology 12, 1831–1841. Bleeker, W., Huthmann, M. and Hurka, H. (1999) Evolution of hybrid taxa in Nasturtium R.Br. (Brassicaceae). Folia Geobotanica and Phytotaxonomica 34, 421–433. Bleeker, W., Franzke, A., Pollmann, K., Brown, A.H.D. and Hurka, H. (2002) Phylogeny and biogeography of southern hemisphere high-mountain Cardamine species (Brassicaceae). Australian Systematic Botany 15, 575–581. Bloom, T.C., Baskin, J.M. and Baskin, C.C. (2001) Ecological life history of the facultative woodland biennial Arabis laevigata variety laevigata (Brassicaceae): survivorship. Journal of the Torrey Botanical Society 128, 93–108. Böcher, T.W. (1951) Cytological and embryological studies in the amphi-apomictic Arabis holboellii com- plex. Det Kongelige Danske Videnskabernes Selskab 6, 1–59.

130 T. Mitchell-Olds et al. Boffelli, D., McAuliffe, J., Ovcharenko, D., Lewis, K.D., Ovcharenko, I., Pachter, L. and Rubin, E.M. (2003) Phylogenetic shadowing of primate sequences to find functional regions of the human genome. Science 299, 1391–1394. Bowman, J.L., Bruggemann, H., Lee, J.Y. and Mummenhoff, K. (1999) Evolutionary changes in floral struc- ture within Lepidium L. (Brassicaceae). International Journal of Plant Sciences 160, 917–929. Boyce, J.M., Knight, H., Deyholos, M., Openshaw, M.R., Galbraith, D.W., Warren, G. and Knight, M.R. (2003) The sfr6 mutant of Arabidopsis is defective in transcriptional activation via CBF/DREB1 and DREB2 and shows sensitivity to osmotic stress. Plant Journal 34, 395–406. Bradshaw, H. and Schemske, D. (2003) Allele substitution at a flower colour locus produces a pollinator shift in monkeyflowers. Nature 426, 176–178. Bray, E. (1997) Plant responses to water deficit. Trends in Plant Science 2, 48–54. Bressan, R.A., Zhang, C., Zhang, H., Hasegawa, P.M., Bohnert, H.J. and Zhu, J.-K. (2001) Learning from the Arabidopsis experience. The next gene search paradigm. Plant Physiology 127, 1354–1360. Brochmann, C., Stedje, B. and Borgen, L. (1992) Gene flow across ploidal levels in Draba (Brassicaceae). Evolutionary Trends in Plants 6, 125–134. Bruggemann, H. (2000) Molekulare Systematik und Biogeographie der Gattung Lepidium L. und verwanter Sippen (Brassicaceae). PhD thesis, University of Osnabrück, Osnabrück, Germany. Brumfield, R.T., Beerli, P., Nickerson, D.A. and Edwards, S.V. (2003) The utility of single nucleotide poly- morphisms in inferences of population history. Trends in Ecology and Evolution 18, 249–256. Cabin, R.J. (1996) Genetic comparisons of seed bank and seedling populations of a perennial desert mus- tard, Lesquerella fendleri. Evolution 50, 1830–1841. Charlesworth, B., Morgan, M.T. and Charlesworth, D. (1993) The effect of deleterious mutations on neutral molecular variation. Genetics 134, 1289–1303. Charlesworth, D. (2003) Effects of inbreeding on the genetic diversity of populations. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 358, 1051–1070. Cheong, Y.H., Kim, K.-N., Pandey, G.K., Gupta, R., Grant, J.J. and Luan, S. (2003) CBL1, a calcium sensor that differentially regulates salt, drought, and cold responses in Arabidopsis. Plant Cell 15, 1833–1845. Clauss, M.J. and Mitchell-Olds, T. (2003) Population genetics of tandem trypsin inhibitor genes in Arabidopsis species with contrasting ecology and life history. Molecular Ecology 12, 1287–1299. Comai, L., Tyagi, A.P., Winter, K., Holmes-Davis, R., Reynolds, S.H., Stevens, Y. and Byers, B. (2000) Phenotypic instability and rapid gene silencing in newly formed Arabidopsis allotetraploids. Plant Cell 12, 1551–1568. Comai, L., Madlung, A., Josefsson, C. and Tyagi, A. (2003) Do the different parental ‘heteromes’ cause genomic shock in newly formed allopolyploids? Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 358, 1149–1155. Comes, H.P. and Kadereit, J.W. (1998) The effect of quaternary climatic changes on plant distribution and evolution. Trends in Plant Science 3, 432–438. Cooper, G.M. and Sidow, A. (2003) Genomic regulatory regions: insights from comparative sequence analy- sis. Current Opinion in Genetics and Development 13, 604–610. Crespo, M.B., Lledo, M.D., Fay, M.F. and Chase, M.W. (2000) Subtribe Vellinae (Brassiceae, Brassicaceae): a combined analysis of ITS nrDNA sequences and morphological data. Annals of Botany 86, 53–62. Dobes, C., Mitchell-Olds, T. and Koch, M. (2003) Multiple hybrid formation in natural populations: con- certed evolution of the internal transcribed spacer of nuclear ribosomal DNA (ITS) in North American Arabis divaricarpa (Brassicaceae). Molecular Biology and Evolution 20, 338–350. Dobes, C., Mitchell-Olds, T. and Koch, M. (2004) Extensive chloroplast haplotype variation indicates Pleistocene hybridization and radiation of North American Arabis drummondii, A. ϫ divaricarpa, and A. holboellii (Brassicaceae). Molecular Ecology 13, 349–370. Dvorák, F. (1973) Importance of indumentum for investigation of evolutional relationship in family Brassicaceae. Osterreichische Botanische Zeitschrift 121, 155–164. Ehrendorfer, F. (1980) Polyploidy and distribution. In: Lewis, W.H. (ed.) Polyploidy. Biological Relevance. Plenum Press, New York, pp. 45–60. Erickson, L.R., Straus, N.A. and Beversdorf, W.D. (1983) Restriction patterns reveal origins of chloroplast genomes in Brassica amphiploids. Theoretical and Applied Genetics 65, 201–206. Farmer, E.E., Almeras, E. and Krishnamurthy, V. (2003) Jasmonates and related oxylipins in plant responses to pathogenesis and herbivory. Current Opinion in Plant Biology 6, 372–378. Franzke, A. and Hurka, H. (2000) Molecular systematics and biogeography of the Cardamine pratensis com- plex (Brassicaceae). Plant Systematics and Evolution 224, 213–234.

Crucifer evolution in the post-genomic era 131 Franzke, A., Pollmann, K., Bleeker, W., Kohrt, R. and Hurka, H. (1998) Molecular systematics of Cardamine and allied genera (Brassicaceae): ITS and non-coding chloroplast DNA. Folia Geobotanica 33, 225–240. Galloway, G.L., Malmberg, R.L. and Price, R.A. (1998) Phylogenetic utility of the nuclear gene arginine decarboxylase: an example from Brassicaceae. Molecular Biology and Evolution 15, 1312–1320. Gómez-Campo, C. (1999) Taxonomy. In: Gomez-Campo, C. (ed.) Biology of Brassica coenospecies. Elsevier, Amsterdam, pp. 3–32. Hall, A.E., Fiebig, A. and Preuss, D. (2002a) Beyond the Arabidopsis genome: opportunities for comparative genomics. Plant Physiology 129, 1439–1447. Hall, J.C., Sytsma, K.J. and Iltis, H.H. (2002b) Phylogeny of Capparaceae and Brassicaceae based on chloro- plast sequence data. American Journal of Botany 89, 1826–1842. Harris, S.A. and Ingram, R. (1991) Chloroplast DNA and biosystematics – the effects of intraspecific diversity and plastid transmission. Taxon 40, 393–412. Hauser, L.A. and Crovello, T.J. (1982) Numerical analysis of generic relationships in Thelypodieae (Brassicaceae). Systematic Botany 7, 249–268. Hayek, A. (1911) Entwurf eines Cruciferensystems auf phylogenetischer Grundlage. Beihefte zum Botanischen Centralblatt 27, 127–335. Heenan, P.B. and Mitchell, A.D. (2003) Phylogeny, biogeography and adaptive radiation of Pachycladon (Brassicaceae) in the mountains of South Island, New Zealand. Journal of Biogeography 30, 1737–1749. Heenan, P.B., Mitchell, A.D. and Koch, M. (2002) Molecular systematics of the New Zealand Pachycladon (Brassicaceae) complex: generic circumscription and relationship to Arabidopsis sens. lat. and Arabis sens. lat. New Zealand Journal of Botany 40, 543–562. Hoisington, D., Khairallah, M., Reeves, T., Ribaut, J.-M., Skovmand, B., Taba, S. and Warburton, M. (1999) Plant genetic resources: what can they contribute toward increased crop productivity? Proceedings of the National Academy of Sciences USA 96, 5937–5943. Hong, R.L., Hamaguchi, L., Busch, M.A. and Weigel, D. (2003) Regulatory elements of the floral homeotic gene AGAMOUS identified by phylogenetic footprinting and shadowing. Plant Cell 15, 1296–1309. Howard, D.J. (1999) Conspecific sperm and pollen precedence and speciation. Annual Review of Ecology and Systematics 30, 109–132. Hurka, H. and Neuffer, B. (1997) Evolutionary processes in the genus Capsella (Brassicaceae). Plant Systematics and Evolution 206, 295–316. Inada, D.C., Bashir, A., Lee, C., Thomas, B.C., Ko, C., Goff, S.A. and Freeling, M. (2003) Conserved noncod- ing sequences in the grasses. Genome Research 13, 2030–2041. Janchen, E. (1942) Das System der Cruciferen. Osterreichische Botanische Zeitschrift 91, 1–18. Judd, W.S., Sanders, R.W. and Donoghue, M.J. (1994) Angiosperm family pairs: preliminary phylogenetic analysis. Harvard Papers in Botany/Harvard University Herbaria 5, 1–51. Kachroo, A., Nasrallah, M.E. and Nasrallah, J.B. (2002) Self-incompatibility in the Brassicaceae: receptor–ligand signaling and cell-to-cell communication. Plant Cell 14, S227–238. Kerkeb, L. and Kramer, U. (2003) The role of free histidine in xylem loading of nickel in Alyssum lesbiacum and Brassica juncea. Plant Physiology 131, 716–724. Kliebenstein, D.J., Kroymann, J., Brown, P., Figuth, A., Pedersen, D., Gershenzon, J. and Mitchell-Olds, T. (2001) Genetic control of natural variation in Arabidopsis glucosinolate accumulation. Plant Physiology 126, 811–825. Kliman, R.M., Andolfatto, P., Coyne, J.A., Depaulis, F., Kreitman, M., Berry, A.J., McCarter, J., Wakeley, J. and Hey, J. (2000) The population genetics of the origin and divergence of the Drosophila simulans complex species. Genetics 156, 1913–1931. Koch, M. (2002) Genetic differentiation and speciation in prealpine Cochlearia: allohexaploid Cochlearia bavarica Vogt (Brassicaceae) compared to its diploid ancestor Cochlearia pyrenaica DC. in Germany and Austria. Plant Systematics and Evolution 232, 35–49. Koch, M. (2003) Molecular phylogenetics, evolution and population biology in Brassicaceae. In: Sharma, A.K. and Sharma, A. (eds) Plant Genome Biodiversity and Evolution. Vol. 1, part A. Phanerogams. Science Publishers, Enfield, New Hampshire, pp. 1–35. Koch, M. and Al-Shehbaz, I.A. (2000) Molecular systematics of the Chinese Yinshania (Brassicaceae): evidence from plastid and nuclear ITS DNA sequence data. Annals of the Missouri Botanical Garden 87, 246–272. Koch, M. and Al-Shehbaz, I. (2002) Molecular data indicate complex intra- and intercontinental differentia- tion of American Draba (Brassicaceae). Annals of the Missouri Botanical Garden 89, 88–109.

132 T. Mitchell-Olds et al. Koch, M. and Al-Shehbaz, I. (2004) Taxonomic and phylogenetic evaluation of the American ‘Thlaspi’ species: identity and relationship to the Eurasian genus Noccaea (Brassicaceae). Systematic Botany 29, 375–384. Koch, M. and Bernhardt, K.G. (2004) Comparative biogeography of the cytotypes of annual Microthlaspi perfoliatum (Brassicaceae) in Europe using isozymes and cpDNA data: refugia, diversity centers, and postglacial colonization. American Journal of Botany 91, 115–124. Koch, M. and Hurka, H. (1999) Isozyme analysis in the polyploid complex Microthlaspi perfoliatum (L.) F. K. Meyer: morphology, biogeography and evolutionary history. Flora 194, 33–48. Koch, M. and Mummenhoff, K. (2001) Thlaspi s.str. (Brassicaceae) versus Thlaspi s.l.: morphological and anatomical characters in the light of ITS nrDNA sequence data. Plant Systematics and Evolution 227, 209–225. Koch, M., Hurka, H. and Mummenhoff, K. (1996) Chloroplast DNA restriction site variation and RAPD- analyses in Cochlearia (Brassicaceae): biosystematics and speciation. Nordic Journal of Botany 16, 585–603. Koch, M., Huthmann, M. and Hurka, H. (1998) Molecular biogeography and evolution of the Microthlaspi perfoliatum s.l. polyploid complex (Brassicaceae): chloroplast DNA and nuclear ribosomal DNA restriction site variation. Canadian Journal of Botany 76, 382–396. Koch, M., Bishop, J. and Mitchell-Olds, T. (1999a) Molecular systematics of Arabidopsis and Arabis. Plant Biology 1, 529–537. Koch, M., Hurka, H. and Mummenhoff, K. (1999b) Molecular phylogenetics of Cochlearia L. and allied gen- era based on nuclear ribosomal ITS DNA sequence analysis contradict traditional concepts of their evolutionary relationship. Plant Systematics and Evolution 216, 207–230. Koch, M.A., Haubold, B. and Mitchell-Olds, T. (2000) Comparative evolutionary analysis of chalcone syn- thase and alcohol dehydrogenase loci in Arabidopsis, Arabis, and related genera (Brassicaceae). Molecular Biology and Evolution 17, 1483–1498. Koch, M., Haubold, B. and Mitchell-Olds, T. (2001a) Molecular systematics of the Brassicaceae: evidence from coding plastidic matK and nuclear Chs sequences. American Journal of Botany 88, 534–544. Koch, M.A., Weisshaar, B., Kroymann, J., Haubold, B. and Mitchell-Olds, T. (2001b) Comparative genomics and regulatory evolution: conservation and function of the Chs and Apetala3 promoters. Molecular Biology and Evolution 18, 1882–1891. Koch, M., Al-Shehbaz, I. and Mummenhoff, K. (2003a) Molecular systematics, evolution, and population biology in the mustard family (Brassicaceae). Annals of the Missouri Botanical Garden 90, 151–171. Koch, M., Bernhardt, K.G. and Kochjarová, J. (2003b) Cochlearia macrorrhiza (Brassicaceae): a bridging species between Cochlearia taxa from the Eastern Alps and the Carpathians. Plant Systematics and Evolution 242, 137–147. Koch, M., Huthmann, M. and Bernhardt, K.G. (2003c) Cardamine amara L. (Brassicaceae) in dynamic habi- tats: genetic composition and diversity of seed bank and established populations. Basic and Applied Ecology 4, 339–348. Koltunow, A. and Grossniklaus, U. (2003) Apomixis: a developmental perspective. Annual Review of Plant Biology 54, 547–574. Kowalski, S.P., Lan, T.-H., Feldman, K.A. and Paterson, A.H. (1994) Comparative mapping of Arabidopsis thaliana and Brassica oleracea chromosomes reveals islands of conserved organization. Genetics 138, 499–510. Kropf, M., Kadereit, J.W. and Comes, H.P. (2002) Late Quaternary distributional stasis in the submediter- ranean mountain plant Anthyllis montana L. (Fabaceae) inferred from ITS sequences and amplified fragment length polymorphism markers. Molecular Ecology 11, 447–463. Kropf, M., Kadereit, J.W. and Comes, H.P. (2003) Differential cycles of range contraction and expansion in European high mountain plants during the Late Quaternary: insights from Pritzelago alpina (L.) O. Kuntze (Brassicaceae). Molecular Ecology 12, 931–949. Kroymann, J., Donnerhacke, S., Schnabelrauch, D. and Mitchell-Olds, T. (2003) Evolutionary dynamics of an Arabidopsis insect resistance quantitative trait locus. Proceedings of the National Academy of Sciences USA 100, 14587–14592. Kuittinen, H., de Haan, A., Vogl, C., Oikarinen, S., Leppala, J., Mitchell-Olds, T., Koch, M., Langley, C. and Savolainen, O. (2004) Comparing the maps of close relatives Arabidopsis lyrata and Arabidopsis thaliana. Genetics (in press). Kunkel, B.N. and Brooks, D.M. (2002) Cross talk between signaling pathways in pathogen defense. Current Opinion in Plant Biology 5, 325–331.

Crucifer evolution in the post-genomic era 133 Lagercrantz, U. (1998) Comparative mapping between Arabidopsis thaliana and Brassica nigra indicates that Brassica genomes have evolved through extensive genome replication accompanied by chromo- some fusions and frequent rearrangements. Genetics 150, 1217–1228. Lagercrantz, U. and Lydiate, D. (1996) Comparative genome mapping in Brassica. Genetics 144, 1903–1910. Lan, T.-H. and Paterson, A.H. (2000) Comparative mapping of quantitative trait loci sculpting the curd of Brassica oleracea. Genetics 155, 1927–1954. Lee, J.-Y., Mummenhoff, K. and Bowman, J.L. (2002) Allopolyploidization and evolution of species with reduced floral structures in Lepidium L. (Brassicaceae). Proceedings of the National Academy of Sciences USA 99, 16835–16840. Levin, D.A. (1990) The seed bank as a source of genetic novelty in plants. American Naturalist 135, 563–572. Liu, F., Zhang, L. and Charlesworth, D. (1998) Genetic diversity in Leavenworthia populations with different inbreeding levels. Proceedings of the Royal Society of London Series B: Biological Science 265, 293–301. Lombi, E., Tearall, K.L., Howarth, J.R., Zhao, F.-J., Hawkesford, M.J. and McGrath, S.P. (2002) Influence of iron status on cadmium and zinc uptake by different ecotypes of the hyperaccumulator Thlaspi caerulescens. Plant Physiology 128, 1359–1367. Marhold, K., Lihova, J., Perny, M., Grupe, R. and Neuffer, B. (2002) Natural hybridization in Cardamine (Brassicaceae) in the Pyrenees: evidence from morphological and molecular data. Botanical Journal of the Linnean Society 139, 275–294. Matzk, F., Meister, A. and Schubert, I. (2000) An efficient screen for reproductive pathways using mature seeds of monocots and dicots. Plant Journal 21, 97–108. Mayfield, J.A., Fiebig, A., Johnstone, S.E. and Preuss, D. (2001) Gene Families from the Arabidopsis thaliana pollen coat proteome. Science 292, 2482–2485. Maynard Smith, J. and Haigh, J. (1974) The hitch-hiking effect of a favourable gene. Genetical Research Cambridge 23, 23–35. McKay, J.K., Bishop, J.G., Lin, J.Z., Richards, J.H., Sala, A. and Mitchell-Olds, T. (2001) Local adaptation across a climatic gradient despite small effective population size in the rare sapphire rockcress. Proceedings of the Royal Society of London Series B: Biological Sciences 268, 1715–1721. McKay, J.K., Richards, J.H. and Mitchell-Olds, T. (2003) Genetics of drought adaptation in Arabidopsis thaliana: I. Pleiotropy contributes to genetic correlations among ecological traits. Molecular Ecology 12, 1137–1151. Meyer, F.K. (1973) Conspectus der ‘Thlaspi’-Arten Europas, Afrikas und Vorderasiens. Feddes Repertorium 84, 449–470. Meyer, F.K. (1979) Kritische Revision der ‘Thlaspi’-Arten Europas, Afrikas und Vorderasiens. I. Geschichte, Morphologie und Chorologie. Feddes Repertorium 90, 129–154. Mitchell, A.D. and Heenan, P.B. (2000) Systematic relationships of New Zealand endemic Brassicaceae inferred from nrDNA ITS sequence data. Systematic Botany 25, 98–105. Mitchell-Olds, T. (2001) Arabidopsis thaliana and its wild relatives: a model system for ecology and evolu- tion. Trends in Ecology and Evolution 16, 693–700. Mitchell-Olds, T. and Clauss, M.J. (2002) Plant evolutionary genomics. Current Opinion in Plant Biology 5, 74–79. Mitchell-Olds, T., James, R.V., Palmer, M.V. and Williams, P.H. (1995) Genetics of Brassica rapa (syn. campestris). 2. Selection for multiple disease resistance to three fungal pathogens: Peronospora parasit- ica, Albugo candida, and Leptosphaeria maculans. Heredity 75, 362–369. Mummenhoff, K. (1995) Should Cardaria draba (L.) Desv. be classified within the genus Lepidium L. (Brassicaceae)? Evidence from subunit polypeptide composition of RUBISCO. Feddes Repertorium 106, 25–28. Mummenhoff, K. and Hurka, H. (1990) Evolution of the tetraploid Capsella bursa-pastoris (Brassicaceae): isoelectric-focusing analysis of rubisco. Plant Systematics and Evolution 172, 205–213. Mummenhoff, K. and Hurka, H. (1994) Subunit polypeptide composition of rubisco and the origin of allopolyploid Arabidopsis suecica (Brassicaceae). Biochemical Systematics and Ecology 22, 807–811. Mummenhoff, K. and Hurka, H. (1995) Allopolyploid origin of Arabidopsis suecica (Fries) norrlin – evi- dence from chloroplast and nuclear genome markers. Botanica Acta 108, 449–456. Mummenhoff, K. and Koch, M. (1994) Chloroplast DNA restriction site variation and phylogenetic relation- ships in the genus Thlaspi sensu lato Brassicaceae. Systematic Botany 19, 73–88.

134 T. Mitchell-Olds et al. Mummenhoff, K., Franzke, A. and Koch, M. (1997a) Molecular data reveal convergence in fruit characters used in the classification of Thlaspi s.l. (Brassicaceae). Botanical Journal of the Linnean Society 125, 183–199. Mummenhoff, K., Franzke, A. and Koch, M. (1997b) Molecular phylogenetics of Thlaspi sl (Brassicaceae) based on chloroplast DNA restriction site variation and sequences of the internal transcribed spacers of nuclear ribosomal DNA. Canadian Journal of Botany 75, 469–482. Mummenhoff, K., Bruggemann, H. and Bowman, J.L. (2001) Chloroplast DNA phylogeny and biogeography of Lepidium (Brassicaceae). American Journal of Botany 88, 2051–2063. Nasrallah, M.E., Yogeeswaran, K., Snyder, S. and Nasrallah, J.B. (2000) Arabidopsis species hybrids in the study of species differences and evolution of amphiploidy in plants. Plant Physiology 124, 1605–1614. Neuffer, B. (1989) Leaf morphology in Capsella (Cruciferae): dependency on environments and biological parameters. Beiträge zur Biologie der Pflanzen 64, 39–54. Neuffer, B. (1990) Ecotype differentiation in Capsella. Vegetatio 89, 165–171. Neuffer, B. and Hurka, H. (1999) Colonization history and introduction dynamics of Capsella bursa-pastoris (Brassicaceae) in North America: isozymes and quantitative traits. Molecular Ecology 8, 1667–1681. Neuffer, B. and Koch, K. (1996) Maternal effects on populations of Capsella bursa-pastoris (L.) med from Scandinavia and the Alps – independence of seed weight and the germination behavior of the matura- tion conditions in lowland and mountains. Biologisches Zentralblatt 115, 337–352. Nordborg, M., Charlesworth, B. and Charlesworth, D. (1996) Increased levels of polymorphism surrounding selectively maintained sites in highly selfing species. Proceedings of the Royal Society of London Series B: Biological Sciences 263, 1033–1039. Nurnberger, T. and Brunner, F. (2002) Innate immunity in plants and animals: emerging parallels between the recognition of general elicitors and pathogen-associated molecular patterns. Current Opinion in Plant Biology 5, 318–324. O’Kane, S.L. Jr and Al-Shehbaz, I.A. (1997) A synopsis of Arabidopsis (Brassicaceae). Novon 7, 323–327. O’Kane, S.L. Jr and Al-Shehbaz, I.A. (2003) Phylogenetic position and generic limits of Arabidopsis (Brassicaceae) based on sequences of nuclear ribosomal DNA. Annals of the Missouri Botanical Garden 90, 603–612. O’Kane, S.L. Jr, Schaal, B.A. and Al-Shehbaz, I.A. (1997) The origins of Arabidopsis suecica (Brassicaceae) as indicated by nuclear rDNA sequences. Systematic Botany 21, 559–566. Oono, Y., Seki, M., Nanjo, T., Narusaka, M., Fujita, M., Satoh, R., Satou, M., Sakurai, T., Ishida, J., Akiyama, K., Iida, K., Maruyama, K., Satoh, S., Yamaguchi-Shinozaki, K. and Shinozaki, K. (2003) Monitoring expression profiles of Arabidopsis gene expression during rehydration process after dehydration using ca. 7000 full-length cDNA microarray. Plant Journal 34, 868–887. Palmer, J.D., Shields, C.R., Cohen, D.B. and Orton, T.J. (1983) Chloroplast DNA evolution and the origin of amphidiploid Brassica species. Theoretical and Applied Genetics 65, 181–189. Persans, M.W., Nieman, K. and Salt, D.E. (2001) Functional activity and role of cation-efflux family mem- bers in Ni hyperaccumulation in Thlaspi goesingense. Proceedings of the National Academy of Sciences USA 98, 9995–10000. Pigliucci, M. and Byrd, N. (1998) Genetics and evolution of phenotypic plasticity to nutrient stress in Arabidopsis – drift, constraints or selection. Biological Journal of the Linnean Society 64, 17–40. Pigliucci, M. and Marlow, E.T. (2001) Differentiation for flowering time and phenotypic integration in Arabidopsis thaliana in response to season length and vernalization. Oecologia 127, 501–508. Pigliucci, M., Cammell, K. and Schmitt, J. (1999) Evolution of phenotypic plasticity: a comparative approach in the phylogenetic neighbourhood of Arabidopsis thaliana. Journal of Evolutionary Biology 12, 779–791. Pineros, M.A. and Kochian, L.V. (2003) Differences in whole-cell and single-channel ion currents across the plasma membrane of mesophyll cells from two closely related Thlaspi species. Plant Physiology 131, 583–594. Pollack, E. (1987) On the theory of partially inbreeding finite populations. I. Partial selfing. Genetics 117, 353–360. Pollard, H., Cruzan, M. and Pigliucci, M. (2001) Comparative studies of reaction norms in Arabidopsis. I. Evolution of response to daylength. Evolutionary Ecology Research 3, 129–155. Prantl, K. (1891) Cruciferae. Die natürlichen Pflanzenfamilien III, 145–206. Price, R.A., Al-Shehbaz, I.A. and Palmer, J.D. (1994) Systematic relationships of Arabidopsis: a molecular and morphological approach. In: Meyerowitz, E. and Somerville, C. (eds) Arabidopsis. Cold Spring Harbor Press, Cold Spring Harbor, New York, pp. 7–19.

Crucifer evolution in the post-genomic era 135 Price, R.A., Al-Shehbaz, I.A. and O’Kane, S.L. (2001) Beringia (Brassicaceae), a new genus of Arabidopsoid affinities from Russia and North America. Novon 11, 332–336. Ramos-Onsins, S., Stranger, B., Mitchell-Olds, T. and Aguade, M. (2004) Multilocus analysis of variation and speciation in the closely-related species Arabidopsis halleri and A. lyrata. Genetics 166, 373–388. Reboud, X. and Zeyl, C. (1994) Organelle inheritance in plants. Heredity 72, 132–140. Regvar, M., Vogel, K., Irgel, N., Wraber, T., Hildebrandt, U., Wilde, P. and Bothe, H. (2003) Colonization of pennycresses (Thlaspi spp.) of the Brassicaceae by arbuscular mycorrhizal fungi. Journal of Plant Physiology 160, 615–626. Richards, A.J. (2003) Apomixis in flowering plants: an overview. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 358, 1085–1093. Rieseberg, L.H., Raymond, O., Rosenthal, D.M., Lai, Z., Livingstone, K., Nakazato, T., Durphy, J.L., Schwarzbach, A.E., Donovan, L.A. and Lexer, C. (2003) Major ecological transitions in wild sunflowers facilitated by hybridization. Science 301, 1211–1216. Rodman, J.E., Karol, K.G., Price, R.A. and Sytsma, K.J. (1996) Molecules, morphology, and Dahlgrens expanded order Capparales. Systematic Botany 21, 289–307. Rodman, J.E., Soltis, P.S., Soltis, D.E., Sytsma, K.J. and Karol, K.G. (1998) Parallel evolution of glucosinolate biosynthesis inferred from congruent nuclear and plastid gene phylogenies. American Journal of Botany 85, 997–1006. Rollins, R.C. (1993) The Cruciferae of Continental North America. Stanford University Press, Stanford, California. Roy, B.A. (1993) Patterns of rust infection as a function of host genetic diversity and host density in natural populations of the apomictic crucifer, Arabis holboellii. Evolution 47, 111–124. Roy, B.A. (1995) The breeding systems of six species of Arabis (Brassicaceae). American Journal of Botany 82, 869–877. Roy, B.A. (2001) Patterns of association between crucifers and their flower-mimic pathogens: host jumps are more common than coevolution or cospeciation. Evolution 55, 41–53. Sall, T., Jakobsson, M., Lind-Hallden, C. and Hallden, C. (2003) Chloroplast DNA indicates a single origin of the allotetraplold Arabidopsis suecica. Journal of Evolutionary Biology 16, 1019–1029. Sarret, G., Saumitou-Laprade, P., Bert, V., Proux, O., Hazemann, J.-L., Traverse, A., Marcus, M.A. and Manceau, A. (2002) Forms of zinc accumulated in the hyperaccumulator Arabidopsis halleri. Plant Physiology 130, 1815–1826. Savolainen, O., Langley, C.H., Lazzaro, B.P. and Freville, H. (2000) Contrasting patterns of nucleotide poly- morphism at the alcohol dehydrogenase locus in the outcrossing Arabidopsis lyrata and the selfing Arabidopsis thaliana. Molecular Biology and Evolution 17, 645–655. Schein, M., Yang, Z., Mitchell-Olds, T. and Schmid, K.J. (2004) Rapid evolution of a pollen-specific oleosin- like gene family from Arabidopsis thaliana and closely related species. Molecular Biology and Evolution 21, 659–669. Schemske, D.W. and Bradshaw, H.D. (1999) Pollinator preference and the evolution of floral traits in mon- keyflowers (Mimulus). Proceedings of the National Academy of Sciences USA 96, 11910–11915. Schmid, K.J., Sorensen, T.R., Stracke, R., Torjek, O., Altmann, T., Mitchell-Olds, T. and Weisshaar, B. (2003) Large-scale identification and analysis of genome-wide single-nucleotide polymorphisms for mapping in Arabidopsis thaliana. Genome Research 13, 1250–1257. Schranz, M.E., Quijada, P., Sung, S.-B., Lukens, L., Amasino, R. and Osborn, T.C. (2002) Characterization and effects of the replicated flowering time gene FLC in Brassica rapa. Genetics 162, 1457–1468. Schulz, O.E. (1936) Cruciferae. In: Engler, A. and Harms, H. (eds) Die natürlichen Pflanzenfamilien. Verlag von Wilhelm Engelmann, Leipzig, pp. 227–658. Shah, J. (2003) The salicylic acid loop in plant defense. Current Opinion in Plant Biology 6, 365–371. Sharbel, T.F. and Mitchell-Olds, T. (2001) Recurrent polyploid origins and chloroplast phylogeography in the Arabis holboellii complex (Brassicaceae). Heredity 87, 59–68. Sharbel, T.F., Haubold, B. and Mitchell-Olds, T. (2000) Genetic isolation by distance in Arabidopsis thaliana: biogeography and post-glacial colonization of Europe. Molecular Ecology 9, 2109–2118. Shu, G., Amaral, W., Hileman, L.C. and Baum, D.A. (2000) LEAFY and the evolution of rosette flowering in violet cress (Jonopsidium acaule, Brassicaceae). American Journal of Botany 87, 634–641. Singh, G., Dyer, P. and Ashby, A. (1999) Intra-specific and inter-specific conservation of mating-type genes from the discomycete plant-pathogenic fungi Pyrenopeziza brassicae and Tapesia yallundae. Current Genetics 36, 290–300.

136 T. Mitchell-Olds et al. Soltis, D.E. and Soltis, P.S. (1999) Polyploidy: recurrent formation and genome evolution. Trends in Ecology and Evolution 14, 348–352. Spielman, M., Vinkenoog, R. and Scott, R.J. (2003) Genetic mechanisms of apomixis. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences 358, 1095–1103. Stahl, E.A., Dwyer, G., Mauricio, R., Kreitman, M. and Bergelson, J. (1999) Dynamics of disease resistance polymorphism at the Rpm1 locus of Arabidopsis. Nature 400, 667–671. Stebbins, G.L. (1940) The significance of polyploidy in plant evolution. American Naturalist 74, 54–66. Strauss, S.Y., Siemens, D.H., Decher, M.B. and Mitchell-Olds, T. (1999) Ecological costs of plant resistance to herbivores in the currency of pollination. Evolution 53, 1105–1113. Sunnucks, P. (2000) Efficient genetic markers for population biology. Trends in Ecology and Evolution 15, 199–203. Swanson, W. and Vacquier, V. (2002a) The rapid evolution of reproductive proteins. Nature Review Genetics 3, 137–144. Swanson, W.J. and Vacquier, V.D. (2002b) Reproductive protein evolution. Annual Review of Ecology and Systematics 33, 161–179. Sweeney, P.W. and Price, R.A. (2000) Polyphyly of the genus Dentaria (Brassicaceae): evidence from trnL intron and ndhF sequence data. Systematic Botany 25, 468–478. Takhtajan, A. (1997) Diversity and Classification of Flowering Plants. Columbia University Press, New York. The Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796–815. Tian, D., Araki, H., Stahl, E., Bergelson, J. and Kreitman, M. (2002) Signature of balancing selection in Arabidopsis. Proceedings of the National Academy of Sciences USA 99, 11525–11530. Tian, D., Traw, M., Chen, J., Kreitman, M. and Bergelson, J. (2003) Fitness costs of R-gene-mediated resis- tance in Arabidopsis thaliana. Nature 423, 74–77. Tremetsberger, K., Konig, C., Samuel, R., Pinsker, W. and Stuessy, T.F. (2002) Infraspecific genetic variation in Biscutella laevigata (Brassicaceae): new focus on Irene Manton’s hypothesis. Plant Systematics and Evolution 233, 163–181. U, N. (1935) Genome-analysis in Brassica with special reference to the experimental formation of B. napus and peculiar mode of fertilization. Japanese Journal of Botany 7, 389–452. Ungerer, M.C., Halldorsdottir, S.S., Purugganan, M.D. and Mackay, T.F.C. (2003) Genotype-environment interactions at quantitative trait loci affecting inflorescence development in Arabidopsis thaliana. Genetics 165, 353–365. Urbanska, K.M., Hurka, H., Landolt, E., Neuffer, B. and Mummenhoff, K. (1997) Hybridization and evolu- tion in Cardamine (Brassicaceae) at Urnerboden, central Switzerland – biosystematic and molecular evidence. Plant Systematics and Evolution 204, 233–256. van Dijk, P. and van Damme, J. (2000) Apomixis technology and the paradox of sex. Trends in Plant Science 5, 81–84. Warwick, S.I. and Al-Shehbaz, I.A. (1998) Generic evaluation of Boleum, Euzomodendron, and Vella (Brassicaceae). Novon 8, 321–325. Warwick, S.I. and Black, L.D. (1994) Evaluation of the subtribes Moricandiinae, Savignyinae, Vellinae, and Zillinae (Brassicaceae, Tribe Brassiceae) using chloroplast DNA restriction site variation. Canadian Journal of Botany 72, 1692–1701. Warwick, S.I. and Black, L.D. (1997a) Molecular phylogenies from theory to application in Brassica and allies (Tribe Brassiceae, Brassicaceae). Opera Botanica 132, 159–168. Warwick, S.I. and Black, L.D. (1997b) Phylogenetic implications of chloroplast DNA restriction site varia- tion in subtribes Raphaninae and Cakilinae (Brassicaceae, tribe Brassiceae). Canadian Journal of Botany 75, 960–973. Warwick, S.I., Al-Shehbaz, I.A., Price, R.A. and Sauder, C. (2002) Phylogeny of Sisymbrium (Brassicaceae) based on ITS sequences of nuclear ribosomal DNA. Canadian Journal of Botany 80, 1002–1017. Weinig, C., Ungerer, M.C., Dorn, L.A., Kane, N.C., Toyonaga, Y., Halldorsdottir, S.S., Mackay, T.F.C., Purugganan, M.D. and Schmitt, J. (2002) Novel loci control variation in reproductive timing in Arabidopsis thaliana in natural environments. Genetics 162, 1875–1884. Weinig, C., Dorn, L.A., Kane, N.C., German, Z.M., Halldorsdottir, S.S., Ungerer, M.C., Toyonaga, Y., Mackay, T.F.C., Purugganan, M.D. and Schmitt, J. (2003a) Heterogeneous selection at specific loci in natural environments in Arabidopsis thaliana. Genetics 165, 321–329. Weinig, C., Stinchcombe, J.R. and Schmitt, J. (2003b) QTL architecture of resistance and tolerance traits in Arabidopsis thaliana in natural environments. Molecular Ecology 12, 1153–1163.

Crucifer evolution in the post-genomic era 137 Widmer, A. and Baltisberger, M. (1999) Extensive intraspecific chloroplast DNA (cpDNA) variation in the alpine Draba aizoides L. (Brassicaceae): haplotype relationships and population structure. Molecular Ecology 8, 1405–1415. Wright, S., Lauga, B. and Charlesworth, D. (2003) Subdivision and haplotype structure in natural popula- tions of Arabidopsis lyrata. Molecular Ecology 12, 1247–1263. Xiong, L. and Zhu, J.-K. (2002) Molecular and genetic aspects of plant responses to osmotic stress. Plant, Cell and Environment 25, 131–139. Zunk, K., Mummenhoff, K., Koch, M. and Hurka, H. (1996) Phylogenetic relationships of Thlaspi sl (subtribe Thlaspidinae, Lepidieae) and allied genera based on chloroplast DNA restriction-site variation. Theoretical and Applied Genetics 92, 375–381. Zunk, K., Mummenhoff, K. and Hurka, H. (1999) Phylogenetic relationships in tribe Lepidieae (Brassicaceae) based on chloroplast DNA restriction site variation. Canadian Journal of Botany 77, 1504–1512.

9 Genetic variation in plant populations: assessing cause and pattern David J. Coates and Margaret Byrne Science Division, Department of Conservation and Land Management, Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia Introduction events associated with recent habitat frag- mentation such as reduced population size The assessment of patterns of genetic varia- and increased isolation. tion in plant populations has made critical contributions to many studies in evolution- A major development in plant population ary biology, conservation genetics, plant genetics has been the significant advance, breeding and ecological genetics. The value over the last three decades, in technologies in such assessments not only relates to that allow the direct molecular characteriza- quantification of the amount and distribu- tion of genes and gene products. In many tion of genetic variation in populations but respects the development and use of various also to the investigation of those processes molecular markers that track changes in that influence patterns of genetic variation. individual genes has revolutionized popula- Plants in particular, with their huge diver- tion genetics and broadened its applicability sity in breeding systems and contrasting across many fields in biology. The effective life-histories, provide a rich source of infor- neutrality of molecular markers means they mation in relation to patterns and are ideal for a broad range of plant popula- processes that characterize genetic diver- tion genetic, conservation genetic and evolu- sity in populations. For example, the rela- tionary studies such as investigating patterns tively recent broad-based interest in of gene flow, mating systems, population applying population genetic principles in genetic structure, hybridization and effective the conservation of small fragmented pop- population size. ulations and management of rare and geo- graphically restricted species has resulted In conjunction with these advances there in a plethora of new studies investigating has been a dramatic increase in studies of patterns of genetic variation in plant popu- intraspecific variation often combining pop- lations (see Young and Clarke, 2000). In ulation genetics, phylogenetics and biogeog- attempting to explain the observed pat- raphy. Such integrated approaches have terns, many of these studies have high- seen the development of new fields of study lighted the complex interactions between such as phylogeography (see Avise, 2000), factors such as evolutionary history, breed- which have made significant contributions to ing system, mode of reproduction and our understanding of evolutionary and eco- logical processes in plant populations (see Schaal and Olsen, 2000). © CAB International 2005. Plant Diversity and Evolution: Genotypic and 139 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

140 D.J. Coates and M. Byrne Measuring Genetic Variation probably correct in most instances, may not necessarily always be the case (see Merila Genetic variation in plant populations has and Crnokrak, 2001). In addition, if they been measured using a broad range of are effectively neutral, measures of popula- approaches. These include: (i) the assess- tion genetic variation based on these mark- ment of quantitative (continuous) characters ers will not be expected to reflect the actions such as seed set, growth rate and time to of selection on other parts of the genome. flowering; (ii) observable heritable polymor- phisms such as flower colour and including Instead of molecular markers, genetic recessive lethal alleles; (iii) chromosome variation in populations can be investigated rearrangements such as translocations and by assessing quantitative variation that is inversions; (iv) protein variation, in particu- under polygenic control where many loci, lar isozyme electrophoresis; and (v) nuclear and the environmental effects on those loci, and organelle DNA variation. contribute to the quantitative variation in the traits being investigated. Yet analysing The application of molecular markers to patterns of genetic variation from molecular investigations of genetic variation in popula- markers has become increasingly popular as tions started with the development of pro- molecular techniques become more cost tein electrophoresis and analysis of isozyme effective and less invasive. Unfortunately variation, some three decades ago. evidence for concordance in these two mea- Subsequently a wide array of DNA-based sures of genetic diversity is equivocal, with a markers is now available that have allowed number of studies suggesting it is poor an ongoing refinement of approaches to the (Reed and Frankham, 2001; MacKay and study of population-based variation and Latta, 2002). Therefore a key question is micro-evolutionary change. Extensive data how well are molecular marker and quanti- sets are now available on population genetic tative variation correlated? variation in numerous plant species, for allozyme variation (see Hamrick and Godt, It is often assumed that the various mea- 1989) and more recently DNA-based mark- sures of genetic variation are positively cor- ers such as RAPDs (see Nybom and Bartish, related, yet there are a number of reasons 2000). Not only has the development of why there may be a lack of agreement these molecular markers allowed the visual- between measures of genetic diversity based ization of locus-specific variation in popula- on molecular markers and quantitative tions, it has also been accompanied by the traits. Molecular markers may not necessar- development of methods that allow the ily track quantitative genetic variation due to ready interpretation of this information in non-additive effects, differential selection, the context of population genetic theory (see different mutation rates, environmental Weir, 1996). effects on quantitative variation, and the influence of genetic variation on gene regu- Quantitative and molecular marker variation lation (see Lynch et al., 1999; Reed and Frankham, 2001). There is increasing evi- Before reviewing the range of molecular dence that within populations there is little markers now available for population association between levels of genetic varia- genetic and evolutionary studies in plants it tion estimated by molecular marker het- is important to consider whether single erozygosity and life-history trait locus variation based on these markers is heritabilities. This implies that neutral mole- representative of the entire genome. Most cular markers are unlikely to provide con- studies of genetic variation in populations servation biologists and evolutionary based on molecular genetic markers con- biologists with any clear indication of a pop- sider those markers to be selectively neutral ulation’s evolutionary potential. For exam- or near neutral. This assumption, although ple, in a meta-analysis Reed and Frankham (2001) found no correlation between molec- ular markers and quantitative trait variation over 19 animal and plant studies.

Genetic variation in plant populations 141 In comparison, the relationship between provided contrasting results. In some genetic differentiation among populations species there is a strong correlation between based on molecular markers and quantita- the two estimates of differentiation, while in tive traits is less clear. The critical question of others, such as some forest trees, variation in whether molecular marker differences quantitative traits associates closely with reflect adaptive divergence among popula- environmental gradients but allozyme varia- tions has been raised in a number of studies tion does not (see Hamrick, 1983). Despite (Lynch et al., 1999; Merila and Crnokrak, this, a trend does appear to be emerging 2001; Reed and Frankham, 2001; McKay that supports observations by Hamrick and Latta, 2002). As pointed out by Lynch (1983) that quantitative traits show as much (1996) significant molecular divergence pro- or more differentiation among populations vides strong evidence that adaptive diver- than allozyme markers. This trend has been gence has the opportunity to occur but lack confirmed in recent reviews by Merila and of any molecular divergence is likely to be Crnokrak (2001) and McKay and Latta uninformative. (2002) where population differentiation for quantitative traits (QST) is typically higher Investigations of population differentia- than estimates of neutral molecular diver- tion in plants that have compared estimates gence based on FST (Table 9.1). from allozymes with quantitative traits have Table 9.1. Comparisons of divergence in quantitative traits (QST) and divergence in marker genes (FST) for plant species where QST partitions quantitative genetic variation in an analogous fashion to FST for single gene markers. For neutral traits, FST and QST should be equal, while the level of difference between them can be used to infer directional selection (QST > FST) or selection favouring the same phenotype in different populations (QST < FST). Note that in most cases QST is larger than FST indicating that natural selection is likely to be a significant force in determining patterns of quantitative trait differentiation among plant populations. Species QST FST Marker Reference a Arabidopsis thaliana 0.830 0.890 Allozymes, microsatellites 1 Arabidopsis thaliana 0.885 1.000 Allozymes 2 Arabis fecunda 0.980 0.200 Allozymes 1 Brassica insularis 0.060 0.210 Allozymes 2 Centaurea corymbosa 0.220 0.364 Allozymes 2 Clarkia dudleyana 0.380 0.075 Allozymes 2 Clarkia dudleyana 0.353 0.068 Allozymes 1 Larix laricina 0.490 0.050 Allozymes 2 Larix occidentalis 0.490 0.086 Allozymes 2 Medicago truncatula 0.584 0.330 Allozymes 2 Phlox drummondii 0.250 0.038 Allozymes 2 Picea glauca 0.360 0.035 Allozymes 2 Picea sitchensis 0.290 0.079 Allozymes 2 Pinus brutia 0.250 0.140 Allozymes 2 Pinus contorta 0.120 0.019 Allozymes 1,2 Pinus sylvestris 0.364 0.018 Allozymes, microsatellites, RFLPs 1 Pseudotsuga menziesii 0.420 0.022 Allozymes 2 Quercus petrea 0.310 0.025 Allozymes 2 Salix vimnalis 0.070 0.041 Allozymes 1,2 Scabiosa canescens 0.095 0.164 Allozymes 1,2 Scabiosa columbaria 0.452 0.123 Allozymes 1,2 Sequoiadendron giganteum 0.180 0.097 Allozymes 2 Silene diclinis 0.118 0.052 Allozymes 1,2 a 1 and 2 are reviews that contain references to the original studies: 1. McKay and Latta (2002); 2. Merila and Crnokrak (2001).

142 D.J. Coates and M. Byrne As emphasized by both Merila and range from broad historically based geo- Crnokrak (2001) and McKay and Latta graphical variation (see Thompson, 1999; (2002), key issues that can be addressed in Schaal and Olsen, 2000) to finer scale studies that compare molecular marker vari- metapopulation genetic structure (see Manel ation with quantitative trait variation are the et al., 2003). Popular molecular marker tech- insights they provide into the relative niques for population genetic studies importance of genetic drift and natural include isozymes and an array of DNA- selection as causes of population divergence based techniques such as restriction frag- in specific quantitative traits. Population ment length polymorphism (RFLP), random structure of quantitative traits is not amplified polymorphic DNA (RAPD), ampli- expected to differ from that of single molec- fied fragment length polymorphism (AFLP) ular marker loci if the traits determined by and microsatellites or simple sequence both are selectively neutral. Consequently repeats (SSRs). Some of these markers, such any differences between QST and FST esti- as isozymes, RFLPs and microsatellites, are mates can be attributed to natural selection codominant and can be analysed as single (see Table 9.1). That is, the extent of local locus markers, while others, such as RAPDs adaptation can be assessed for various traits and AFLPs, are dominant multilocus mark- and the QST for those traits experiencing the ers (Table 9.2). strongest local selection will be expected to show the largest difference from the molecu- Isozymes (allozymes) lar marker FST. Isozyme electrophoresis has made an Despite these recent reviews, covering a immense contribution to research in plant range of different plant species, there is still population genetics, systematics and evolu- no clear answer to the question of how well tion (see Soltis and Soltis, 1989), and plant divergence based on neutral markers pre- conservation biology (see Falk and dicts that based on quantitative traits. Merila Holsinger, 1991). Despite the development and Crnokrak (2001) found in their meta- of a range of DNA-based markers, allozymes analysis based on 27 plant and animal continue to be an important and reliable species that the level of differentiation in tool for the study of genetic variation and neutral marker loci is closely predictive of evolutionary processes in plant populations. the level of differentiation in loci encoding There seems little doubt that their wide- quantitative traits. In contrast, McKay and spread and continued use stems from their Latta (2002) found that differentiation in cost effectiveness, technical ease, number of neutral marker loci and quantitative trait available loci and codominant inheritance loci was poorly correlated across their sam- (Arnold and Imms, 1998). ple of 29 species. However, both emphasize the need for further theoretical and empiri- Apart from advantages in terms of cost cal studies to address the relationship and time, allozymes have a number of other between QST and FST. advantages that make them convenient and reliable genetic markers. These include Molecular markers in plant population Mendelian inheritance, codominant expres- genetics sion, and similarity of apparently homolo- gous isozyme loci and their allozyme Over the last few decades, the use of neutral patterns between different species. Other molecular markers has dominated studies advantages lie not so much in the technique, on population genetic structure and geo- analytical methods or properties of graphic patterns of genetic variation in isozymes, but rather in the large number of plants. The range of markers now available studies that have been conducted on a wide allows increased flexibility for investigators range of plant taxa. From this large database to utilize more than one marker at varying of information it is now possible to make spatial and temporal scales. These may quite useful generalizations on the relation- ships between patterns of genetic variation

Genetic variation in plant populations 143 Table 9.2. Common molecular techniques used in assessing patterns of genetic variation in plant populations. Technique Methodology Technical Level of difficulty polymorphism Resolution Reliability Codominant, single-locus markers Allozymes/isozymes Gel electrophoresis and Easy Low to Moderate Very high moderate visualization of cellular Moderate to high enzymes and proteins Very high Restriction fragment Digestion of total Moderate High Very high Moderate length polymorphism genomic DNA with to difficult to high High (RFLP) restriction endonucleases followed by Southern blotting and hybridization with specific DNA fragments Microsatellites or Specifically developed PCR Difficult High High simple sequence primers used to amplify repeats (SSRs) hypervariable tandemly repeated units. Variation at these loci can be investigated in both nuclear and chloroplast genomes Dominant multilocus markers Random amplified Amplification of random DNA Easy Moderate Medium polymorphic DNA segments using arbitrary short (RAPD) sequence primers (≈ 10 nucleotides in length) Amplified fragment Amplification of total genomic Moderate High Medium to high length polymorphism DNA digested with restriction (AFLP) endonucleases, where the subsequent restriction site is then used as a primer binding site for selective amplification using PCR primers that anneal perfectly to target sequences and factors such as life-history, breeding sys- greatest value that isozyme analysis will have tem, geographic distribution and habitat in the future is not necessarily as an alterna- (see Hamrick and Godt, 1989). tive to more recent DNA-based techniques but as a source of supporting data that may Although isozyme analysis has a number indicate directions for further detailed mole- of features that make it an attractive tech- cular studies. nique, it also has a number of well-known limitations. There is the potential for signifi- Nuclear DNA markers cant undetected variation given that only about 20–30% of base substitutions in the RESTRICTION FRAGMENT LENGTH POLYMORPHISMS gene will result in detectable change using standard electrophoresis conditions. The major use of DNA-based markers in Probably the greatest restriction of isozyme plant population studies commenced with markers is the relatively low level of varia- the application of endonuclease methodolo- tion. Compared with markers such as gies and the development of DNA restric- microsatellites, allozymes clearly have much tion fragment length polymorphisms. less resolving power in, for example, pater- Single-copy nuclear RFLPs provide a large nity analysis and gene flow studies. The number of highly variable codominant

144 D.J. Coates and M. Byrne markers. The demonstration that they could where they have been used, they have be readily used in plant breeding studies proved to be extremely robust and infor- and for analysing population genetic varia- mative in assessing patterns of population tion in plants led to their rapid utilization, genetic variation in a range of species cov- particularly in crop species and wild rela- ering a broad range of plant genera and tives (Clegg, 1989). families (see Byrne et al., 1998). In addi- tion, in contrast to microsatellites (see Some of the first applications of RFLPs in below) they can be readily used across plant population studies involved investiga- species and genera. tions of genetic variation at the high-copy ribosomal loci (see Schaal et al., 1991). These MICROSATELLITES Microsatellites or SSRs investigations were based on rDNA inter- genic spacer length and restriction site vari- are now recognized as potentially one of the ation. While some species, such as Clematis fremontii, can show significant rDNA varia- most useful genetic markers in plant popula- tion at the population level, others show lit- tle or no length variation or restriction site tion studies. Microsatellites consist of tan- variation (Schaal et al., 1991). Thus rDNA RFLP data was found to be relatively limited dem repeats of a short ‘motif ’ sequence, in its application to population studies. usually of one to six bases. These regions In contrast, single-copy nuclear RFLP markers have proved to be particularly occur frequently and randomly in plant and informative in the analysis of patterns of genetic variation within and among plant animal genomes and often have large num- populations. Comparisons of genetic varia- tion using RFLPs with that of allozymes bers of moderately frequent alleles. Thus indicate that the level of polymorphism detected with RFLP loci is generally three they show extensive variation between indi- to four times higher than the level detected with isozymes (see Byrne et al., 1998). As viduals within populations and have been with other DNA markers, this higher level of variation can be readily attributed, at developed for a wide range of purposes in least theoretically, to the assaying of all mutational variation compared with only a plant breeding, conservation biology and subset of total variation detectable with isozymes. Although most comparisons of population genetics including forensics, RFLPs versus isozymes show similar levels and patterns of divergence among popula- paternity analysis and gene mapping (Jarne tions, there are notable exceptions such as in Beta vulgaris subsp. maritima (sea beet) and Lagoda, 1996). In particular, in plant (Raybould et al., 1996). These findings have been generally attributed to selection oper- population genetic studies they have proved ating on some isozyme loci, although, as Raybould et al. (1996) point out, an alterna- to be ideal for assessing gene flow among tive scenario is that RFLPs could be under disruptive selection. populations (see Chase et al., 1996), and are Despite their significant resolving power ideally suited to fine-scale analysis of mating in population genetics, single-copy nuclear RFLPs have become less popular in recent within populations. times. Reasons given are that they are time consuming and expensive, they require the Unfortunately the significant benefits of use of radiolabelled probes, and relatively large amounts of DNA are needed. Yet, such hypervariable codominant markers are offset by the time and effort involved in their development (see Squirrell et al., 2003). Sequence information is required to design appropriate primers and such infor- mation is generally only available for a lim- ited number of commercially important species. For most plant species, microsatel- lites can only be developed from clones iso- lated through construction and screening of a genomic library. As pointed out by Squirrell et al. (2003), the development of a working primer set is subject to a consider- able attrition rate compared with the origi- nal number of clones sequenced. This contrasts with, and increases the appeal of, other highly polymorphic markers such as AFLPs and RAPDs (see below) where generic primers are readily available. One approach that has been successfully

Genetic variation in plant populations 145 used in some animal groups involves cross- nificant problem since the lack of specificity species amplification of microsatellite loci. associated with the use of short arbitrary However, this approach is much less success- primers can lead to increased sensitivity to ful in plants where the general patterns indi- PCR conditions resulting in erratic amplifica- cate that cross-species amplification of plant tion. However, this problem can be mini- microsatellites will be restricted to closely mized with strictly controlled and related species or at least congeners (Peakall standardized reaction conditions. Homology et al., 1998; Butcher et al., 2000), although is also an important issue, given that comigra- broader cross-transferability has been tion of products assumes homology, which demonstrated in Vitaceae (Arnold et al., may not necessarily be correct (see Rieseberg, 2002). 1996). This is more likely to be a problem in comparisons where higher levels of genetic RANDOM AMPLIFIED POLYMORPHIC DNA RAPDs are divergence are involved, such as between probably the next most common DNA taxa, and is less likely to be an issue in popu- markers that have been used in plant popu- lation genetic studies. Another limitation is lation genetic studies. These markers were that the majority of RAPD markers are domi- the first of a number of multilocus PCR- nant and there is therefore a significant loss based markers that have been widely of information content compared with applied across plant species. The technique codominant markers such as isozymes, RFLPs uses single arbitrary primer sequences to and microsatellites. In this regard, compar- amplify anonymous regions of the genome isons between RAPDs and allozymes need to and can be used to identify and screen be treated with caution, given the different numerous polymorphic loci. Since no approaches used for dealing with monomor- sequence information is needed, this tech- phic loci and that dominant RAPD loci are nique is particularly applicable in cases unsuited to estimation of population genetic where little molecular genetic information is parameters such as F statistics and GST unless available on the target species. Furthermore, assumptions are made regarding the breed- the assay is very simple and fast, and many ing system and Hardy–Weinberg equilibrium loci can be identified, often with a single (Lynch and Milligan, 1994; Bussell, 1999). reaction. AMPLIFIED FRAGMENT LENGTH POLYMORPHISMS Initially RAPDs were extensively used in plant breeding studies and particularly in AFLPs are a more recently developed mole- genome mapping to identify quantitative cular marker for plant population genetic trait loci. Subsequently they became increas- studies. Like RAPDs, AFLPs have an advan- ingly popular in studies of genetic variation tage over RFLPs because they are applicable in natural populations, often in conjunction to DNA of any origin without prior with, or as an adjunct to, isozyme studies sequence information, primer synthesis or (see Peakall et al., 1995). Patterns of genetic library construction. Also, like RAPDs, most diversity and population genetic structure AFLPs are dominant markers, but unlike using RAPDs have now been investigated in RAPD protocols a single AFLP reaction can a broad range of plant species (Harris, 1999; survey as many as 100–200 loci (see Mueller Nybom and Bartish, 2000). They have and Wolfenbarger, 1999). The polymor- proved to be of particular value where they phisms typically exhibit Mendelian inheri- have revealed useful levels of genetic varia- tance, enabling their use for typing, tion within and/or between populations identification and mapping of genetic char- despite the detection of only minimal acteristics. The potential to produce hun- allozyme variation. dreds of polymorphic loci per individual and the relative ease of applicability to most Although RAPDs provide a fast and cost- organisms makes AFLPs one of the most effi- effective means of investigating genetic varia- cient molecular markers for generating tion, they have a number of limitations (see polymorphic loci in plant population genetic Arnold and Imms, 1998; Harris, 1999). studies (Mueller and Wolfenbarger, 1999). Reproducibility has often been cited as a sig-

146 D.J. Coates and M. Byrne Although AFLP analyses do not suffer sequence conservation in the chloroplast from the reproducibility problems that genome that allows heterologous probes to potentially exist in the use of RAPDs, they be utilized across most plant families. With are still largely dominant markers and the development of PCR, a number of RFLP therefore pose the same difficulties in studies have now been carried out based on reliance on the assumptions of equilibrium PCR-amplified products, thus avoiding the for estimations of population genetic para- time required for Southern hybridization. meters. However, their high resolution and the ease with which one can generate large A limitation of cpDNA analysis in many numbers of genetic markers has seen a species is the relatively low sequence diver- steady increase in their application in the sity, which can be attributed to the haploid analysis of population level variation in status of the genome, reduced rate of muta- plant species. tion and lack of recombination (see Schaal et al., 1998; Ennos et al., 1999). To address this AFLPs and microsatellites are rapidly issue there have been attempts to target sec- becoming the DNA-based markers of choice tions of the chloroplast genome that have for plant population genetic structure, mat- relatively high mutation rates. Useful levels ing system and gene flow studies. A number of intraspecific variation have been detected of recent investigations have compared the in non-coding regions of the cpDNA potential level of resolution and efficiency of genome. For example, Vaillancourt and these markers in plant population studies. Jackson (2000) detected significant levels of As expected, microsatellites are considerably polymorphism for eucalypts in sequence more polymorphic than AFLPs at the locus studies of the JLA junction between the level, but AFLPs are much more efficient at inverted repeats and the large single-copy revealing polymorphic loci. For example, in region, with much of the variation due to Avicennia marina average expected heterozy- complex insertion/deletions. Another useful gosity for AFLPs was only 0.193 compared approach with considerable potential with 0.780 for microsatellites, but all of the involves the analysis of variation using 918 AFLP bands scored were polymorphic chloroplast microsatellites (Powell et al., (Maguire et al., 2002). 1995). A recent assessment of the occur- rence of microsatellites in six species where Chloroplast DNA (cpDNA) variation the cpDNA genome has been completely sequenced detected a total of 505 cpDNA Whereas the molecular markers discussed microsatellites (Provan et al., 1999). previously are based on various portions of Although high levels of polymorphism make the nuclear genome, another marker that cpDNA microsatellites extremely useful has considerable potential in plant popula- markers in plants for gene flow, their high tion studies is cpDNA. Compared with the mutation rate, similar to nuclear microsatel- nuclear genome of higher plants, which con- lites, makes them unlikely to be useful mark- sist of a diploid complement of randomly ers in phylogenetic analysis, primarily segregating biparentally inherited chromo- because of the problems of homoplasy somes, the chloroplast genome is predomi- where the same mutation can arise from nantly uniparentally inherited and consists of independent events. a single circular molecule. In angiosperms the chloroplast genome is generally mater- Despite occasional difficulties in accessing nally inherited, while in most gymnosperms appropriate levels of variation in plant pop- it is paternally inherited. A number of differ- ulation studies, cpDNA has proved to be ent approaches have been utilized to charac- extremely useful in providing insights into terize variation in cpDNA. The first and population and evolutionary processes that probably most widely used involves RFLP could not be delivered by nuclear markers. analysis of the entire chloroplast genome. An important factor assisting this approach One of the most fundamental applica- resides in the relatively high degree of tions of patterns of cpDNA variation has been in the analysis of phylogeographic pat- terns of population variation (see section on

Genetic variation in plant populations 147 ‘Historical associations and phylogeographic ory for small populations occupying a nar- patterns’, below). In addition to uniparental row range of environments and have inheritance and lack of recombination, there received significant attention over the last are a number reasons why cpDNA may be decade (see Karron, 1987; Hamrick and far more suitable for such studies than Godt, 1989; Barrett and Kohn, 1991; nuclear markers. The effective population Ellstrand and Elam, 1993; Gitzendanner size of an organelle gene is theoretically half and Soltis, 2000). Theoretically, geographi- that of a nuclear gene, so cpDNA variation is cally restricted and rare plants might be likely to be more sensitive to population dif- expected to show low levels of genetic varia- ferentiation by genetic drift. Also, because tion within both species and populations the maternally inherited cpDNA will only be because of selection under a narrow range dispersed by seed, the genetic differences of environmental conditions and genetic existing when populations come into contact drift and inbreeding in small, isolated pop- will probably break down much more slowly ulations (Barrett and Kohn, 1991; Ellstrand than for nuclear markers, leaving a signa- and Elam, 1993). Low genetic variation ture of historical relationships for much within geographically restricted species may longer (see Ennos et al., 1999). also be due to founder events associated with recent speciation (Gottlieb, 1981; Significant Determinants of Genetic Loveless and Hamrick, 1988). Broad com- Variation in Plant Populations parisons between geographically restricted and widespread species generally follow Determinants of patterns of genetic variation predicted trends, with allozyme studies in plant populations are extremely varied and showing that geographically restricted often involve complex interactions between species have less genetic variability than plant attributes such as life-form, floral archi- widespread species (Hamrick and Godt, tecture, mode of reproduction, incompatibil- 1989; Table 9.3). However, this association ity system, pollination system, and ecological was not evident in a review of plant popula- and environmental parameters that may tion-based studies using RAPDs (Nybom influence pollination events, population size and Bartish, 2000). and isolation (see Table 9.3). A further level of complexity can be added when one considers A useful demonstration of the relation- the evolutionary history of the species where ship between geographic range and patterns events such as climate change and localized of genetic variation can be found in compar- extinction, contraction to refugia, range isons between closely related taxa of a trig- expansion and fluctuations in population size gerplant (Stylidium) species complex in over time can also have substantial influence south-west Australia (Coates et al., 2003). on the current patterns of genetic variation in The Stylidium caricifolium complex taxa show a plant population (Schaal et al., 1998). Table a range of geographic distributions but are 9.3 provides a summary of attributes and an phylogenetically closely related and are indication of their level of influence on pat- characterized by the same pollination sys- terns of genetic variation within and between tem, similar seed dispersal mechanisms, self- plant populations. Four key themes are compatibility and frequent geitonogamous explored in the following sections that self-pollination. These characteristics make it broadly address those attributes. possible to readily investigate theoretical predictions of lower genetic diversity and Geographical distribution and rarity reduced genetic structure for rare and geo- graphically restricted taxa in this complex, Predictions regarding the genetic conse- while at the same time minimizing any con- quences of restricted geographic range and founding effects that may be associated with rarity in plants generally follow genetic the- phylogenetic differences and differences in life-history attributes and breeding system (see Karron, 1987; Karron et al., 1988; Gitzendanner and Soltis, 2000).

148 D.J. Coates and M. Byrne Table 9.3. Levels of genetic variation at the population level and differentiation among populations based on allozymes (Hamrick and Godt, 1989) and RAPDs (Nybom and Bartish, 2000) for species with different attributes. Allozymes RAPDs Attributes N Hep N GST N Hpop N GST Taxonomic status *** *** *** NS Gymnosperms Monocotyledons 56 0.160 80 0.068 5 0.386 6 0.180 Dicotyledons 80 0.144 81 0.231 9 0.190 6 0.310 Life form Annual 338 0.096 246 0.273 27 0.191 19 0.320 Short-lived perennial Short-lived perennial *** *** NS NS (herbaceous) Short-lived perennial 187 0.105 146 0.357 4 0.125 2 0.470 (woody) Long-lived perennial 13 0.207 13 0.300 Long-lived perennial (herbaceous) 159 0.096 119 0.233 Long-lived perennial (woody) 11 0.094 8 0.088 23 0.242 14 0.230 Geographic range Endemic 4 0.084 2 0.213 Narrow Regional 115 0.149 131 0.076 Widespread *** NS NS NS Breeding system 5 0.190 Selfing 100 0.063 52 0.248 5 0.191 7 0.220 Mixed 9 0.350 Mixed–animal 115 0.105 82 0.242 4 0.233 9 0.330 Mixed–wind Outcrossing 180 0.118 186 0.216 16 0.222 Outcrossing–animal Outcrossing–wind 85 0.159 87 0.210 15 0.208 Seed dispersal *** *** *** *** Gravity Gravity–attached 113 0.074 78 0.510 8 0.091 5 0.590 Attached Explosive 6 0.219 5 0.190 Ingested Wind 85 0.090 60 0.216 Mode of reproduction 10 0.198 11 0.100 Sexual Sexual and asexual 24 0.260 18 0.230 Successional status 164 0.124 124 0.197 Early Mid 102 0.148 134 0.099 Late ** *** NS NS 199 0.101 161 0.277 16 0.212 16 0.300 12 0.127 11 0.124 68 0.137 52 0.257 3 0.165 2 0.470 34 0.062 23 0.243 54 0.129 39 0.223 17 0.228 9 0.170 105 0.123 121 0.143 3 0.261 2 0.230 NS NS 413 0.114 352 0.225 56 0.103 54 0.213 NS *** ** ** 198 0.107 165 0.289 10 0.166 8 0.500 182 0.106 121 0.259 19 0.195 13 0.230 103 0.133 121 0.101 12 0.287 9 0.200 N, number of taxa; Hep, genetic diversity (allozymes); Hpop, genetic diversity (RAPDs); GST, proportion of total genetic diversity among populations. Significance levels: * P < 0.05; ** P < 0.01; *** P < 0.001; NS, not significant.

Genetic variation in plant populations 149 Population-based estimates of genetic mentation and hybridization, respectively. diversity are relatively high for all taxa in This example reinforces issues raised by var- the S. caricifolium complex when compared ious authors in relation to rarity. They stress with other outcrossing animal-pollinated that rarity has multiple origins and that and short-lived herbaceous perennials, but genetic and ecological consequences of rar- indicate significant differences in genetic ity cannot necessarily be generalized in any variability between geographically restricted simplistic fashion (Fiedler and Ahouse, taxa and more widespread taxa. The three 1992; Gaston, 1997). most widespread and common species showed consistently, and in a number of Another generalized assumption is that cases significantly, higher levels of genetic geographically restricted species might also diversity than four of the six geographically show less genetic structure across their restricted species. Lower levels of genetic range than widespread species. The S. carici- diversity in the geographically restricted folium complex study appears to confirm this taxa are likely to be due to a number of fac- prediction with a significant trend from tors. Fluctuations in population size and higher FST values for the more widespread repeated bottlenecks associated with taxa to lower values for the more geographi- extended Pleistocene climatic instability may cally restricted taxa. However, this trend has be primary determinants. Habitat specificity not been observed generally in comparisons is also a likely factor, with three taxa con- between rare and widespread congeners fined to breakaways and rocky slopes associ- (Gitzendanner and Soltis, 2000), or in ated with granite outcrops, banded broader comparisons between species with ironstone and laterites, while the fourth is contrasting geographic ranges (Table 9.3). restricted to coastal dune systems. In addi- Explanations for this include confounding tion, the close phylogenetic relationship effects due to differences in life-history, taxo- between three taxa suggests that lower levels nomic biases, different evolutionary histories of genetic diversity in the two rare and geo- and sampling strategies (see Hamrick and graphically restricted taxa may be due to Godt, 1989; Godt and Hamrick, 1999). founder events associated with relatively recent divergence and isolation. Mode of reproduction and clonality Despite reduced levels of genetic diver- A feature of flowering plants is the vast array sity in geographically restricted taxa this of different modes of reproduction both trend is not consistent across all such taxa in within and across plant families (Richards, the complex. Two other rare and geographi- 1997). In particular, both asexual and sexual cally restricted taxa show comparable or reproduction may vary among species, higher levels of genetic diversity than the among populations within species and within three widespread species. A number of populations (Ellstrand and Roose, 1987; explanations have been given for unexpect- Coates, 1988; Kennington and James, 1998; edly high levels of genetic variation in rare Sydes and Peakall, 1998). In some cases the and geographically restricted species. These reproductive systems may change within include recent origin and retention of high populations or species in response to envi- levels of variation from a widespread prog- ronmental factors (Eckert and Barrett, 1993; enitor (Gottlieb et al., 1985), hybridization Richards, 1997) or with time (Stöcklin, (see below), maintenance of relatively large 1999). Although sexual reproduction is gen- populations (Young and Brown, 1996), erally prevalent, the ability to asexually long-term stability of populations occupying reproduce is common in plants resulting in refugia (Lewis and Crawford, 1995), and clonal or partially clonal populations. relatively recent fragmentation of previously Mechanisms of asexual reproduction can be widespread species (Karron, 1991). The divided into vegetative reproduction and likely causes of the unexpectedly high levels apomixis (Richards, 1997). In the latter, a of genetic variation in these two taxa are variety of mechanisms can result in the pro- large population size prior to recent frag-

150 D.J. Coates and M. Byrne duction of seeds with the same diploid geno- tive reproduction and clonality. The north- type as the maternal genotype. In many ern populations reproduce primarily by cases, fertilization of the endosperm is seed, and individuals within populations required (pseudogamy), indicating that polli- generally show different multilocus geno- nation is just as essential for these species as types and high levels of genetic diversity. In sexual species. In contrast, vegetative repro- contrast, southern populations are charac- duction is achieved by a propagule other terized by a few multilocus genotypes, often than seed and is often associated with exten- with fixed heterozygosity at multiple loci sive clonality in populations of some species. and consequently low levels of genetic diver- sity (Coates, 1988). Assumptions regarding the population genetic consequences of lack of sexual repro- Mating system duction are varied; some suggest that asexual populations will be genotypically depauper- The mating system is expected to be a key ate, while others indicate that asexual popula- factor influencing levels of genetic variation tions may be as genetically variable as sexual within populations and the population ones (see Ellstrand and Roose, 1987). In their genetic structure of a species. Plant mating review, Ellstrand and Roose (1987) found systems are influenced by a number of that the vast majority of clonal plants investi- attributes such as flowering phenology, pre- gated were multiclonal both within and and post-zygotic incompatibility and flower between populations, and that they often pos- structure, as well as a range of ecological fac- sess considerable genetic variability. Similarly, tors. These may include mode of pollina- Hamrick and Godt (1989) found no differ- tion, population size and density, and ences at the population level between sexual population position in the landscape. Plant species and species that reproduce by both species, therefore, exhibit a wide array of sexual and asexual means (Table 9.3). mating systems and this diversity is better Interestingly Hamrick et al. (1992) showed thought of as a continuum rather than as that in long-lived woody shrubs genetic diver- specific categories (Schemske and Lande, sity was significantly higher in populations of 1985; Brown, 1989; Barrett and Eckert, species with combined sexual and asexual 1990). Although this is generally acknowl- reproduction compared with entirely sexual edged, it is important to be able to recognize species. These findings indicate that clonal the major mating system types, given the species may maintain comparable or some- contrasting effects they may have on pat- times higher levels of genetic diversity than terns of genetic variation within plant sexually producing species. species. A useful summary of mating system modes is given by Brown (1989). These are: Geographical patterns of clonal variation predominant self-fertilization, predominant within species will be expected to have a sig- outcrossing, mixed selfing and outcrossing, nificant influence on the partitioning of apomixis and intragametophytic selfing or genetic variation among populations. haploid selfing as may occur in homo- Higher levels of differentiation and lower sporous ferns and allied lower plants. Some levels of genetic variation would be pre- 20% of higher plant species are predomi- dicted for clonal populations where rela- nantly selfing and, as emphasized by Brown tively few clones characterize the (1989), critical factors that will influence the populations. This pattern was clearly outcomes of mating in such species will be observed in Acacia anomala, a small herba- the occurrence and pattern of variation in ceous shrub known from only ten popula- occasional outcrossing events. Conversely, in tions occurring in two disjunct areas some predominantly outcrossing species the pres- 30 km apart (Fig. 9.1). Each population ence of selfing and biparental inbreeding group covers only a few kilometres, with will have a major influence on genetic struc- allozymes indicating high genetic divergence ture both within and between populations. between the two groups. Significantly, this level of divergence is associated with a change from sexual reproduction to vegeta-

Genetic variation in plant populations 151 Fig. 9.1. Geographical distribution and unweighted pair-group method using an arithmetic average (UPGMA) of largely sexual and clonal populations of the rare and endangered grass wattle, Acacia anomala. The northern populations (Chittering) are sexual while the southern populations (Kalamunda) are clonal. Structuring of genetic diversity among populations, based on allozymes, is shown in the UPGMA. The estimated number of plants in each population (N) and gene diversity (He) are shown for each population (UPGMA branch). For clonal populations the estimate of He is based on the number of ramets (Coates, 1988). The significance of mating systems as a breeding system than by any other factor primary determinant of genetic structure (Hamrick and Godt, 1989; Nybom and and levels of genetic variation within plant Bartish, 2000). Allozyme data indicate that populations has been emphasized in a num- selfing species have 51% of their genetic ber of studies. Selfing species are expected variation partitioned among populations, to have less genetic variation within popula- while in outcrossed and mixed mating sys- tions and greater genetic differentiation tem species this is reduced to 10–22% (Table between populations than outcrossing 9.3). For example, in a study on two sym- species. Reviews comparing allozyme, RAPD patric Delphinium species, Williams et al. and quantitative genetic variation across a (2001) found that the many-flowered wide range of species generally support Delphinium barbeyi had a lower outcrossing these expectations (see Table 9.3), with pre- rate through increased geitonogamous self- dominantly outcrossed species having signif- pollination and a tenfold increase in popula- icantly higher levels of genetic diversity than tion subdivision, compared with the selfing species or species with a mixed mat- few-flowered and more highly outcrossing ing system (Hamrick and Godt, 1989; Delphinium nuttallianum. In contrast, Hamrick Schoen and Brown, 1991; Charlesworth and et al. (1992) found that breeding system Charlesworth, 1995; Nybom and Bartish, played a relatively minor role in predicting 2000). levels of genetic diversity among popula- tions of woody plants, although they point It has also been shown in broad compar- out that this may be because of the limited isons between plant species that the parti- range of breeding systems in their sample of tioning of genetic variation among woody plants. populations is generally influenced more by

152 D.J. Coates and M. Byrne Interspecific comparisons of mating sys- ence the mating system and thus patterns of tems by various authors have resulted in the genetic variation remains a key issue in the recognition of a number of key issues associ- conservation of such populations. In addi- ated with mating system change. For exam- tion, populations of such species can provide ple, plant species investigated by Schemske valuable experimental systems for investigat- and Lande (1985) show a strongly bimodal ing forces that affect mating systems. For distribution of outcrossing rates with a signif- example, in a review of mating system varia- icant deficit of species with intermediate out- tion in animal-pollinated rare and endan- crossing rates. They argue that predominant gered populations in Western Australia, selfing and predominant outcrossing are the Sampson et al. (1996) found that outcrossing stable endpoints in mating system evolution, rates from disturbed populations of mixed suggesting that selection on outcrossing rates mating species often differed substantially is strongly directional. Investigations by from those of undisturbed populations Barrett and Eckert (1990) indicate a signifi- (Table 9.4). Disturbance, associated with loss cant association between outcrossing rate of understorey species and weed invasion, and longevity, suggesting that increased out- may influence pollinator type, abundance crossing in long-lived woody species may be and activity, and population density and because of an increased genetic load in such structure. Clearly, as expected, habitat dis- species (see also Ledig, 1986). Pollination turbance has a significant effect on the mat- mode has also been found to influence ing system and levels of inbreeding in patterns of outcrossing rates among species; populations of these species. wind-pollinated species show a clear bimodal distribution while animal-pollinated species In addition to outcrossing rates, another do not (Aide, 1986; Barrett and Eckert, mating system parameter that has proved to 1990). Although such comparative studies be informative in comparative studies provide valuable insight into factors that may among populations is the correlation of out- have significant influences on mating system crossed paternity, that is, the probability that patterns in different species, it is important sibs share the same father. For example, to recognize the limitations in these studies on bird-pollinated, long-lived woody approaches. Comparisons between many shrubs in the family Proteaceae have shown unrelated species may be strongly con- that average paternal diversity in open-polli- founded by their different evolutionary nated sib arrays can be low. In two rare histories (Barrett and Eckert, 1990). woody shrubs, Lambertia orbifolia (Coates and Hamley, 1999) and Grevillea iaspicula Intraspecific variation in outcrossing rates (Hoebee and Young, 2001), the high levels has the potential to be much more valuable of correlated paternity and low estimates of in assessing the causes of mating system neighbourhood size were attributed to mat- change and subsequent changes in patterns ing between small groups of plants. This of genetic variation. Levels of outcrossing interpretation was generally consistent with within populations are the outcome of a pollinator observations, which indicated that complex interaction of the environmental, bird movements are frequently restricted to demographic, life-history and genetic charac- only a few mature plants. teristics of the populations (Barrett and Eckert, 1990). Good examples of this com- Gene flow plexity are provided in a number of recent studies that have focused on the patterns and Gene flow is a key factor in shaping gene causes of mating system variation in rare and pools and the population genetic structure endangered plant populations. A range of of a species, both as a force in maintaining threatening processes such as habitat loss, genetic continuity between populations and degradation and fragmentation are likely to as a means by which genetic diversity can be cause significant changes in population size, enhanced. Thus the gene flow potential of a density, isolation and pollination biology. species will be expected to have a major Understanding how these changes will influ-

Genetic variation in plant populations 153 Table 9.4. Estimates of outcrossing rate in natural and disturbed plant populations of animal-pollinated, mixed mating, rare and endangered Western Australian flora. tm is the multilocus outcrossing rate. Outcrossing rates for disturbed populations often differ substantially from those of undisturbed populations. In some species there appears to be a trend to reduced outcrossing in disturbed populations (Banksia cuneata, Eucalyptus rhodantha, Lambertia orbifolia) while in another there is no clear trend (Verticordia fimbrilepis). Population/size Population status tm Banksia cuneataa High disturbance 0.67 (0.04) 1. 56 Undisturbed 0.95 (0.05) 2. 40 Low disturbance 0.76 (0.05) 3. 86 Low disturbance 0.88 (0.07) 4. 120 Remnant 0.77 (0.01) Banksia triscuspisa Undisturbed 0.74 (0.01) 1. 1 Undisturbed 0.92 (0.01) 2. 3 Undisturbed 0.84 (0.05) 3. 5 Undisturbed 0.79 (0.04) 4. 350 Remnant 1.02 (0.04) 5. 108 Burnt 0.69 (0.04) 6. 24 Undisturbed 0.71 (0.02) 7. 4,140 8. 89 Undisturbed 0.84 (0.03) Undisturbed 0.48 (0.05) Eucalyptus ramelianaa Undisturbed 0.97 (0.03) 1. 83 0.96 (0.02)d 2. 57 3. 200 0.59 (0.04) 0.67 (0.05)d Eucalyptus rhodanthaa 0.26 (0.05) 1. 180 Undisturbed remnant 0.53 (0.08) 0.72 (0.10) 2. 14 High disturbance 0.57 (0.11) 0.41 (0.08) Lambertia orbifoliab Undisturbed 1. 483 Undisturbed 0.62 (0.07) 2. 250 Low disturbance 0.52 (0.13) 3. 100 High disturbance 0.70 (0.13) 4. 56 0.73 (0.10) High disturbance Verticordia fimbrilepisc Undisturbed 1. 90 High disturbance 2. 305 Undisturbed 3. 796 4. 59,270 aData for these species are from Sampson et al. (1996). bData from Coates and Hamley (1999). cData from Sampson (personal communication). dWithin population studies. Standard errors in parentheses. influence on the partitioning of genetic vari- Hamrick and Godt (1989) clearly supports ation among plant populations and levels of this expectation based on their analysis of genetic variation within populations. The the relationship between GST and Hep, and review of allozyme data in plants by species attributes, summarized in Table 9.3.

154 D.J. Coates and M. Byrne Limited gene flow may lead to divergence underestimate gene flow. For example, in between populations as a consequence of Chamaecrista fasciculata, despite the estimates genetic drift or differential selection in local of limited gene dispersal based on pollen environments. Whereas high levels of gene and seed dispersal events, decreased fitness flow enhance homogenization between gene of progeny from near-neighbour matings pools and increase effective population size, and increased fitness from more distant mat- they may also result in the establishment of ings suggest that gene flow will be more unfavourable alleles where populations are widespread (Fenster, 1991). In addition, differentially adapted to localized condi- studies on the distribution of genetic varia- tions. In the latter case, it has been suggested tion based on molecular markers indicate that gene flow may result in outbreeding that pollen-mediated gene flow can be far depression (see Templeton, 1986; Fenster greater than expected from direct estimates and Dudash, 1994), where the progeny of of pollen dispersal. For example, in the bee- distant matings may be less fit than the pollinated Lupinus texensis, Schaal (1980) progeny of near or neighbour matings. demonstrated the effects of pollen carry Genetic causes of outbreeding depression over with significantly greater patterns of may relate to the external environment or to gene movement estimated from isozyme intrinsic factors involving critical interac- markers than pollen movement. tions between groups of genes or coadapted gene complexes. For example, F1 hybrids More recently, direct estimates of the dis- may not be adapted to either parental envi- persal of genes via pollen have been devel- ronment because heterozygosity at major oped through the application of molecular single genes, important in local adaptation, markers in paternity analysis and parentage results in reduced survivorship or reproduc- assignment. These approaches involve com- tion. Alternatively reduced fitness in F1 parisons of segregating alleles in parental hybrids may relate to the break up of co- and progeny cohorts and have been devel- adapted gene complexes (see Fenster and oped to determine paternal contributions Dudash, 1994). Waser (1993) provides evi- through exclusion techniques, maximum- dence for outbreeding depression in a num- likelihood estimates or a combination of ber of different plant species following these methods using fractional paternity (see crossing between geographically distant Devlin et al., 1988). Evidence from a number populations. In contrast to these findings of paternity studies indicates that there can other studies have found that gene migra- be significant gene flow into a population tion between populations or subpopulations from outside sources (see Broyles et al., may result in increased fitness or heterosis 1994; Schnabel and Hamrick, 1995). (Ledig, 1986). Estimates of pollen from outside sources based on allozyme studies can be as high as Gene flow, like the mating system, is 50% in the perennial herb Asclepias exaltata heavily influenced by the pollination system (Broyles et al., 1994) and 30% in the legumi- of the target species but importantly it is also nous tree Gleditsia triacanthos (Schnabel and influenced by seed dispersal. Although vari- Hamrick, 1995). Similarly, estimates based ous indirect measures of gene flow such as on microsatellite markers indicate compara- pollinator flight distances, seed dispersal, ble levels in some tropical tree species dye and pollen analogue dispersal, genetic (Chase et al., 1996). diversity statistics and distribution of alleles among populations (see Broyles et al. 1994; While gene flow in plants has been largely Schnabel and Hamrick, 1995) are useful in investigated in terms of pollen flow, often comparative studies, and as indicators of using nuclear markers, far less attention has potential gene flow within and between pop- been given to gene flow mediated by seed ulations, they do not necessarily provide a dispersal. Although seed movement can true representation of effective gene move- clearly make a significant contribution to ment. As pointed out by Levin (1981), there gene flow, it has proved difficult to measure is good reason to believe that dispersal data and is not well understood in relation to pop- ulation genetic variation in plants. The avail-

Genetic variation in plant populations 155 ability of cpDNA markers combined with the plants are numerous there are certain key development of suitable theoretical method- areas of investigation that highlight the ologies (see Ennos et al., 1999) has provided a many combinations of historical, demo- basis for addressing this important issue in graphic and ecological processes that may plant population studies. Ennos et al. (1999) generate such patterns. These are investi- show that, by the joint measuring of genetic gated in the following sections. differentiation for nuclear and cpDNA mark- ers, the ratio of pollen to seed flow among Disjunct distributions and differentiation populations can be readily estimated and compared across various plant species. Disjunct population systems are common in plant species in many parts of the world and Geographical Patterns of Genetic may reflect geological and edaphic complex- Variation among Plant Populations ities, localized extinction events, contraction to refugia associated with climate change, Geographical patterns of intraspecific and island isolation associated with changes genetic variation have been of fundamental in sea levels. Genetic studies of these popu- interest to plant evolutionary biologists and lation systems can not only give indications population geneticists as they are often of the level and patterns of divergence assumed to represent the initiation of events among populations but can also give valu- that will lead to independent evolutionary able clues to the significance of these lineages and allopatric speciation. This processes in the evolution of individual information is also of particular importance plant species and groups of plant species to those involved in the management and across regions. For example, the significant conservation of genetic resources either for divergence among populations of interconti- species of commercial interest or for species nental acacias found in northern Australia, targeted in conservation programmes. For such as Acacia aulacocarpa (GST = 0.626; example, strategies for the collection and McGranahan et al., 1997), probably reflects a maintenance of genetic resources of com- wider distribution on the Australian geologi- mercial crops place considerable reliance on cal plate in the Tertiary followed by sea-level understanding the pattern and distribution changes, geographic separation and contrac- of genetic variation among source popula- tions due to cycles of aridity during the tions across different eco-geographic Quaternary. Climatic fluctuations during the regions. The value of these data in the man- Quaternary, particularly the Pleistocene, agement of crop genetic resources is have also been important in the evolution of reflected in the large number of species for other tree species such as Abies firs in which variation in the genetic structure of Mexico and Guatemala. Here, the signifi- landrace populations of crops and wild rela- cant differentiation among populations tives of domesticated plants has been investi- appears to be associated with the increased gated using isozyme analysis (see Frankel et isolation of populations as they retreated al., 1995). upwards during the Pleistocene glaciation and the warming period that followed Studies of intraspecific geographical pat- (Aguirre-Planter et al., 2000). terns of genetic variation have been used to investigate the wide array of factors that Some floras found in regions with may be important as evolutionary determi- Mediterranean climates, such as the flora of nants. These include historical associations south-west Australia and the Mediterranean among populations; the role of selection, flora, are typically rich with closely related gene flow and drift; the development of bar- disjunct species or species with disjunct pop- riers to gene flow among populations; and ulations (Thompson, 1999; Coates, 2000). In spatial variation in mating systems both regions, disjunct distributions appear to (Thompson, 1999). Although studies on be associated with a number of key geo- intraspecific patterns of genetic variation in historical phenomena. In south-west Australia

156 D.J. Coates and M. Byrne these include the existence of marine, tions is very high (FST = 0.441). Phylogenetic edaphic and climatic barriers that have sepa- analyses based on either gene frequency data rated this area from the rest of Australia or genetic distance give identical tree topolo- since the Eocene; the formation of nutrient gies and indicate that the two disjunct popu- deficient sands and laterites favouring a lation groups are separate evolutionary shrubland flora that could readily adapt to lineages. The analysis of cpDNA variation the increasing aridity of the late Tertiary confirms this conclusion (Byrne et al., 1999). and Quaternary; and climatic and landscape Inferences based on the proposed long-term instability in the transitional rainfall zone effects of Pleistocene climate change on the (Hopper, 1979). In the Mediterranean flora, south-west Australian flora suggest that the however, disjunct distributions have been current population genetic structure in L. associated with the geological complexity of orbifolia is the result of local extinction of the Mediterranean basin; movement and intervening populations, and extended isola- isolation of tectonic microplates during the tion of the two remnants (Coates and Tertiary; island isolation with change in sea Hamley, 1999). This is supported by studies levels; and dispersal (Thompson, 1999). on large endemic forest eucalypts in areas between the two population groups that Many species in the south-west show patterns of local extinction and range Australian flora are likely to be relictual contraction due to climate change (Wardell- and probably had wider, more continuous Johnson and Coates, 1996). distributions during favourable climatic regimes up to the early Pleistocene. Historical associations and phylogeographic Following the increased aridity and climatic patterns instability during the Pleistocene, these taxa have become locally extinct but have Previous mention has been made of the survived as disjunct remnants, particularly influence of historical events on patterns of through the semi-arid transitional rainfall genetic variation among disjunct plant popu- zone. Recent gene flow between these dis- lations. In those cases, the opportunity for junct population groups, either by long dis- contemporary gene flow among populations tance seed dispersal or pollen movement, is low and any phylogenetic similarity is has probably been limited or absent for probably due to common ancestry rather long periods. As a consequence, significant than any ongoing process of genetic genetic differentiation between populations exchange. Yet in more continuous popula- is typical of many species and is particularly tion systems it is far more difficult to assess evident in rare and geographically the significance of contemporary patterns of restricted species. Relatively high levels of gene flow versus genetic similarity due to population differentiation have been recent common ancestry. Most studies of reported for 22 animal-pollinated, mainly inter-population genetic variation in plants outcrossing, taxa with disjunct population are based on allele frequency data from systems. These taxa cover a range of south- markers such as allozymes, nuclear RFLPs, west Australian genera including long-lived RAPDs, AFLPs and microsatellites where woody shrubs and trees, and herbaceous genetic change over time cannot be directly perennials (Coates, 2000). Thompson inferred (Schaal and Olsen, 2000). As (1999) describes similar patterns of diver- pointed out by Schaal and Olsen (2000), esti- gence among disjunct populations of mates of genetic exchange and the analysis of species in the Mediterranean flora. population genetic structure are generally based on models that assume equilibrium A typical example of these patterns in between genetic drift and gene flow. When south-west Australia can be found in investigating determinants of geographical Lambertia orbifolia, a large, bird-pollinated patterns of genetic variation such estimates woody shrub known only from seven popu- based on, for example, F statistics (Wright, lations that have a significant disjunct distrib- ution (Fig. 9.2). Allozyme studies show that the genetic divergence between all popula-

Genetic variation in plant populations 157 Fig. 9.2. Phylogenetic relationships based on UPGMA of allozyme data and geographical distribution of disjunct populations (Scott River Plains and Narrikup) of Lambertia orbifolia. Population size (number of reproductively mature plants) is given after each population name. The analysis of cpDNA variation based on RFLPs detected eight mutations distributed over five haplotypes (I–V) with the Narrikup populations distinguished by a single haplotype characterized by six unique mutations. 1951) do not distinguish between historical extremely useful source of variation for phy- effects and contemporary patterns of gene logeographic studies in a number of plant flow. Furthermore, gene flow drift equilib- species. The lack of recombination and uni- rium is generally considered to be an unlikely parental mode of inheritance in cpDNA per- scenario for most plant populations, although mits genealogical relationships to be followed it may be more prevalent in ancient floras across populations and the delimitation of where metapopulation systems within species phylogeographically distinct populations. may have remained relatively stable over long Unfortunately, as mentioned previously, periods of time (see Coates et al., 2003). cpDNA may have limited application in some cases because insufficient variation is present Phylogeography provides an approach to allow geographical patterns to be that potentially allows discrimination detected. In particular, the variance in between historical and contemporary pat- cpDNA diversity is likely to be large, with terns of gene exchange (Schaal et al., 1998). some species showing little diversity due to As a sub-discipline of biogeography, phylo- recent selective sweeps (see Ennos et al., geography involves the analysis of the geo- 1999). graphical distribution of genealogical lineages and focuses on the assessment of A number of studies based primarily on historical factors as determinants of evolu- cpDNA variation have now been published tionary patterns among populations (Avise, describing common plant phylogeographic 2000). Chloroplast DNA has proved to be an patterns for certain geographic regions.

158 D.J. Coates and M. Byrne Probably the best examples to date are those Evolutionarily significant units and that show postglacial migration routes from conservation Pleistocene refugia for multiple plant species in both Europe and north-west America. An important factor in assessing priorities for Phylogeographic analyses in European oak the conservation and management of species species indicate that there are three main involves understanding patterns of intraspe- haplotype lineages with postglacial migra- cific genetic variation to identify populations tion inferred to start from three distinct which may be critical for the conservation of southerly refugia in the Balkan, Iberian and genetic resources and evolutionary processes Italian peninsulas (see Ferris et al., 1999). (Hopper and Coates, 1990; Newton et al., Similar patterns of cpDNA haplotype vari- 1999). Although species are generally ability have been shown for two other accepted as primary units for conservation, European tree species, beech (Fagus sylvatica; existing taxonomies may not adequately con- Demesure et al., 1996) and alder (Alnus gluti- sider intraspecific variation. The concept of nosa; King and Ferris, 1998); both studies the ‘evolutionarily significant unit’ (ESU) has indicate a single glacial refugium in the subsequently been introduced to deal with Carpathians. Phylogeographic studies based groups of populations that warrant separate on cpDNA variation in six plant species management for conservation (Ryder, 1986). from the Pacific Northwest also show pat- The concept is based on a sound under- terns concordant with postglacial coloniza- standing of the evolutionary significance of tion. Here the cpDNA phylogenies for five geographically based genetic variation within vascular plants and one fern, representing a a species, with ESUs generally considered to range of different life-histories, indicate two be geographically discrete. clades of populations with a north–south separation. This separation has also been Criteria for defining an ESU have supported by population genetic studies on included significant divergence of allele fre- other plant species (Soltis et al., 1997). quencies, specific levels of genetic distance and phylogenetic differences based on cer- Recent phylogeographic studies on tain genes. A specific approach outlined by species in south-west Australia indicate some Moritz (1994) defined an ESU as an histori- commonality in geographical patterns cally isolated and independently evolving set although the findings here are generally of populations. In animals this was regarded more complex and appear to reflect a much as populations showing reciprocal mono- more ancient pattern of population evolu- phyly for mitochondrial DNA (mtDNA) alle- tion associated with climatic instability since les with significant divergence of alleles at the late Tertiary. Studies on two species nuclear loci. As mentioned previously, using based on cpDNA variation, Santalum spicatum organelle DNA, such as mtDNA or cpDNA, (Western Australian sandalwood) and permits genealogical relationships to be fol- Eucalyptus loxophleba, indicate two geographi- lowed across populations and the delimita- cally distinct haplotype lineages showing a tion of phylogeographically distinct north–south separation. In both species, a population groups and ESUs. For example, nested clade analysis inferred past fragmen- the array of different cpDNA lineages found tation as the most likely cause of the differ- in various European tree species has been entiation between the lineages (Fig. 9.3). considered indicative of separate ESUs. The level of sequence divergence between Perhaps more importantly, however, these lineages was similar in both species and sug- studies have highlighted the value of refu- gests a mid-Pleistocene timeframe for this gial areas in southern Europe. These are divergence. This shared phylogeographic considered key areas for genetic resource pattern is consistent with a hypothesis of sig- conservation because they contain many nificant climatic fluctuations during the unique haplotypes (Newton et al., 1999). Pleistocene and suggests that such climatic instability has resulted in significant frag- Although defining ESUs based on mentation events in the flora of this region. cpDNA variation and phylogeographic analysis has potential in plant conservation

Genetic variation in plant populations 159 Fig. 9.3. Haplotype network showing nested clades for Eucalyptus loxophleba and Santalum spicatum. Haplotypes are represented by letters. Interior haplotypes not detected in the sample are represented by 0. Each line connecting haplotypes represents a single mutational change. One-step clades are indicated by thin-lined boxes, two- and three-step clades by heavier-lined boxes. Differentiation of the three-step clades in E. loxophleba and two-step clades in S. spicatum represent distinct lineages in each species and show a shared phylogeographic pattern indicating a north–south geographic separation. Inferences from a nested clade analysis identified past fragmentation as the most likely cause of the significant geographic association of genealogical lineages for each species. Application of a molecular clock indicates a mid-Pleistocene split and that the past fragmentation is most likely due to significant climate instability during that period in the south-west Australian region (Byrne et al., 2003; Byrne and Hines, 2004). genetics (see Newton et al., 1999), it may analysis of gene frequency data, may still be well be limited, given the low level of extremely valuable in identifying conserva- cpDNA variation found in some plant tion units in plant taxa (Coates, 2000). groups. Where cpDNA is restricted, it will probably be necessary to define conservation The example given previously (L. orbifolia, units in terms of contemporary population Fig. 9.2) clearly indicates the value of such genetic structure based on polymorphic information in setting conservation priorities. nuclear markers, such as allozymes, RAPDs Significant genetic divergence among or microsatellites, rather than historical pop- disjunct populations based on nuclear genes ulation structure inferred from phylogeo- (allozymes) and cpDNA indicates two ESUs. graphic analyses. Whilst these data lack L. orbifolia has an International Union for genealogical information and therefore have Conservation of Nature and Natural limitations in phylogeographic analyses, sig- Resources (IUCN) and Western Australian nificant divergence in allele frequencies at ranking of endangered, but with the large nuclear loci, combined with phylogenetic number of critically endangered plants in that State it has a relatively low priority for

160 D.J. Coates and M. Byrne conservation and recovery actions. Recently, entiation among populations but this will the two ESUs were recognized as distinct con- certainly not always be the case. Likewise, servation units and listed separately under lack of any molecular marker differentiation Western Australian State legislation. Small does not necessarily allow one to infer a lack population size, increased inbreeding and of adaptive differences. In addition, as threats posed by disease and habitat degrada- pointed out by Newton et al. (1999), each tion indicate that the Narrikup ESU (Fig. 9.2) marker has its limitations. While cpDNA has is critically endangered (Coates and Hamley, proved to be extremely valuable, its rela- 1999). Subsequently, this ESU was targeted tively slow mutation rate may mean that for a major re-introduction programme with intraspecific patterns of variation based on a new population established on a nearby shared haplotypes may be confounded by nature reserve. This population of some 500 much more ancient relationships perhaps plants is now successfully recruiting from covering timeframes of millions of years. In seed produced by the plants originally estab- that case, phylogeographic patterns may not lished on the site (L. Monks, Perth, 2003, be readily interpretable in relation to the personal communication). recent history of the populations and geo- graphical patterns (see Ennos et al., 1999). It Despite the value of population genetic would be more appropriate in these situa- and phylogeographic studies in informing tions to combine the cpDNA analyses with genetic resource conservation and setting studies of genetic relatedness based on priorities for conservation, caution is needed nuclear markers that may help elucidate the in interpreting the results from such studies. influence of recent population history on Genetic differences based on molecular current population genetic structure. markers may be indicative of adaptive differ- References Aguirre-Planter, E., Furnier, G.R. and Eguiarte, L.E. (2000) Low levels of genetic variation and high levels of genetic differentiation among populations of species of Abies from southern Mexico and Guatemala. American Journal of Botany 87, 362–371. Aide, T.M. (1986) The influence of wind and animal pollination on variation in outcrossing rates. Evolution 40, 434–435. Arnold, C., Rosetto, M., McNally, J. and Henry, R.J. (2002) The application of SSRs characterized for grape (Vitis vinifera) to conservation studies in Vitaceae. American Journal of Botany 89, 22–28. Arnold, M.L. and Imms, S.K. (1998) Molecular markers, gene flow, and natural selection. In: Soltis, D.E., Soltis, P.S. and Doyle, J.J. (eds) Molecular Systematics of Plants. II. DNA Sequencing. Kluwer Academic Publishers, Boston, Massachusetts, pp. 442–458. Avise, J. (2000) Phylogeography. Harvard University Press, Harvard, Massachusetts. Barrett, S.C.H. and Eckert, C.G. (1990) Variation and evolution of mating systems in seed plants. In: Kawano, S. (ed.) Biological Approaches and Evolutionary Trends in Plants. Academic Press, London, pp. 229–254. Barrett, S.C.H. and Kohn, J.R. (1991) Genetics and evolutionary consequences of small population size in plants: implications for conservation. In: Falk, D.A. and Holsinger, K.E. (eds) Genetics and Conservation of Rare Plants. Oxford University Press, New York, pp. 3–30. Brown, A.H.D. (1989) Genetic characterization of mating systems. In: Brown, A.H.D., Clegg, M.T., Kahler, A.L. and Weir, B.S. (eds) Plant Population Genetics, Breeding and Genetic Resources. Sinauer Associates, Sunderland, Massachusetts, pp. 145–162. Broyles, S.B., Schnabel, A. and Wyatt, R. (1994) Evidence for long-distance pollen dispersal in milkweeds (Asclepias exaltata). Evolution 48, 1032–1040. Bussell, J.D. (1999) The distribution of random amplified polymorphic DNA (RAPD) diversity amongst popu- lations of Isotoma petraea (Lobeliaceae). Molecular Ecology 8, 775–789. Butcher, P.A., Decroocq, S., Gray, Y. and Moran, G.F. (2000) Development, inheritance and cross species amplification of microsatellite markers from Acacia mangium. Theoretical and Applied Genetics 101, 1282–1290.

Genetic variation in plant populations 161 Byrne, M. and Hines, B. (2004) Phylogeographical analysis of cpDNA variation in Eucalyptus loxophleba (Myrtaceae). Australian Journal of Botany 52, 459–470. Byrne, M., Parrish, T.L. and Moran, G.F. (1998) Nuclear RFLP diversity in Eucalyptus nitens. Heredity 81, 225–233. Byrne, M., Macdonald, B. and Coates, D.J. (1999) Divergence in the chloroplast genome and nuclear rDNA of the rare Western Australian plant Lambertia orbifolia (Proteaceae). Molecular Ecology 8, 1789–1796. Byrne, M., Macdonald, B. and Brand, J. (2003) Phylogeography and divergence in the chloroplast genome of Western Australian Sandalwood (Santalum spicatum). Heredity 91, 389–395. Charlesworth, D. and Charlesworth, B. (1995) Quantitative genetics in plants: the effect of the breeding sys- tem on genetic variability. Evolution 49, 911–920. Chase, M.R., Moller, C., Kesseli, R. and Bawa, K.S. (1996) Distant gene flow in tropical trees. Nature 383, 398–399. Clegg, M.T. (1989) Molecular diversity in plant populations. In: Brown, A.H.D., Clegg, M.T., Kahler, A.L. and Weir, B.S. (eds) Plant Population Genetics, Breeding and Genetic Resources. Sinauer Associates, Sunderland, Massachusetts, pp. 98–115. Coates, D.J. (1988) Genetic diversity and population genetic structure in the rare Chittering grass wattle, Acacia anomala Court. Australian Journal of Botany 36, 273–286. Coates, D.J. (2000) Defining conservation units in a rich and fragmented flora: implications for the manage- ment of genetic resources and evolutionary processes in south-west Australia. Australian Journal of Botany 48, 329–339. Coates, D.J. and Hamley, V.L. (1999) Genetic divergence and mating systems in the endangered and geo- graphically restricted species Lambertia orbifolia Gardner (Proteaceae). Heredity 83, 418–427. Coates, D.J., Carstairs, S. and Hamley, V.L. (2003) Evolutionary patterns and genetic structure in localized and widespread species in the Stylidium caricifolium complex (Stylideaceae). American Journal of Botany 90, 997–1008. Demesure, B., Comps, B. and Petit, R.J. (1996) Chloroplast DNA phylogeography of the common beech (Fagus sylvatica L.) in Europe. Evolution 50, 2515–2520. Devlin, B., Roeder, K. and Ellstrand, N.C. (1988) Fractional paternity assignment: theoretical development and comparison to other methods. Theoretical and Applied Genetics 76, 369–380. Eckert, C.G. and Barrett, S.C.H. (1993) Clonal reproduction and patterns of genotypic diversity in Decodon verticillatus (Lythraceae). American Journal of Botany 80, 1175–1182. Ellstrand, N.C. and Elam, D.R. (1993) Population genetic consequences of small population size: implica- tions for plant conservation. Annual Review of Ecology and Systematics 24, 217–242. Ellstrand, N.C. and Roose, M.L. (1987) Patterns of genotypic diversity in clonal plant species. American Journal of Botany 74, 123–131. Ennos, R.A., Sinclair, W.T., Hu, X.-S. and Langdon, A. (1999) Using organelle markers to elucidate the his- tory, ecology and evolution of plant populations. In: Hollingsworth, P.M., Bateman, R.M. and Gornall, R.J. (eds) Molecular Systematics and Plant Evolution. Taylor & Francis, London, pp. 1–19. Falk, D.A. and Holsinger, K.E. (1991) Genetics and Conservation of Rare Plants. Oxford University Press, New York. Fenster, C.B. (1991) Gene flow in Chamaecrista fasciculata (Leguminosae) II. Gene establishment. Evolution 45, 410–422. Fenster, C.B. and Dudash, M.R. (1994) Genetic considerations for plant population restoration and conser- vation. In: Bowles, M.L. and Whelan, C.J. (eds) Restoration of Endangered Species. Cambridge University Press, Cambridge, pp. 34–62. Ferris, C., King, R.A. and Hewitt, G.M. (1999) Isolation within species and the history of glacial refugia. In: Hollingsworth, P.M., Bateman, R.M. and Gornall, R.J. (eds) Molecular Systematics and Plant Evolution. Taylor & Francis, London, pp. 20–34. Fiedler, P.L. and Ahouse, J.J. (1992) Hierarchies of cause: toward an understanding of rarity in vascular plant species. In: Fiedler, P.L. and Jain, S.K. (eds) Conservation Biology: the Theory and Practice of Nature Conservation. Chapman & Hall, New York, pp. 23–48. Frankel, O.H., Brown, A.H.D. and Burdon, J.J. (1995) The Conservation of Plant Biodiversity. Cambridge University Press, Cambridge. Gaston, K.J. (1997) What is rarity? In: Kunin, W.E. and Gaston, K.J. (eds) The Biology of Rarity. Chapman & Hall, London, pp. 30–41. Gitzendanner, M.A. and Soltis, P.S. (2000) Patterns of genetic variation in rare and widespread congeners. American Journal of Botany 87, 783–792.

162 D.J. Coates and M. Byrne Godt, M.J.W. and Hamrick, J.L. (1999) Genetic divergence among infraspecific taxa of Sarracenia purpurea. Systematic Botany 23, 427–438. Gottlieb, L.D. (1981) Electrophoretic evidence and plant populations. Progress in Phytochemistry 7, 1–46. Gottlieb, L.D., Warwick, S.I. and Ford, V.S. (1985) Morphological and electrophoretic divergence between Layia discoidea and L. glandulosa. Systematic Botany 10, 133–148. Hamrick, J.L. (1983) The distribution of genetic variation within and among natural plant populations. In: Schonewald-Cox, C.M., Chambers, S.M., MacBryde, B. and Thomas, L. (eds) Genetics and Conservation. Benjamin-Cummings, London, pp. 343–348. Hamrick, J.L. and Godt, M.J.W. (1989) Allozyme diversity in plant species. In: Brown, A.H.D., Clegg, M.T., Kahler, A.L. and Weir, B.S. (eds) Plant Population Genetics, Breeding and Genetic Resources. Sinauer Associates, Sunderland, Massachusetts, pp. 43–63. Hamrick, J.L., Godt, M.J.W. and Sherman-Broyles, S.L. (1992) Factors influencing levels of genetic diversity in woody plant species. New Forests 6, 95–124. Harris, S.A. (1999) RAPDs in systematics – a useful methodology? In: Hollingsworth, P.M., Bateman, R.M. and Gornall, R.J. (eds) Molecular Systematics and Plant Evolution. Taylor & Francis, London, pp. 211–228. Hoebee, S.E. and Young, A.G. (2001) Low neighbourhood size and high interpopulation differentiation in the endangered shrub Grevillea iaspicula McGill (Proteaceae). Heredity 86, 489–496. Hopper, S.D. (1979) Biogeographical aspects of speciation in the southwest Australian flora. Annual Review of Systematics and Ecology 10, 399–422. Hopper, S.D. and Coates, D.J. (1990) Conservation of genetic resources in Australia’s flora and fauna. Proceedings of the Ecological Society of Australia 16, 567–577. Jarne, P. and Lagoda, P.J.L. (1996) Microsatellites, from molecules to populations and back. Trends in Ecology and Evolution 11, 424–429. Karron, J.D. (1987) A comparison of levels of genetic polymorphism and self-compatibility in geographically restricted and widespread plant congeners. Evolutionary Ecology 1, 47–58. Karron, J.D. (1991) Patterns of genetic variation and breeding systems in rare plant species. In: Falk, D.A. and Holsinger, K.E. (eds) Genetics and Conservation of Rare Plants. Oxford University Press, New York, pp. 87–98. Karron, J.D., Linhart, Y.B., Chaulk, C.A. and Robertson, C.A. (1988) Genetic structure of populations of geo- graphically restricted and widespread species of Astragalus (Fabaceae). American Journal of Botany 75, 1114–1119. Kennington, W.J. and James, S.H. (1998) Allozyme and morphometric variation in two closely related mallee species from Western Australia, Eucalyptus argutifolia and E. obtusifolia (Myrtaceae). Australian Journal of Botany 46, 173–186. King, R.A and Ferris, C. (1998) Chloroplast DNA phylogeography of Alnus glutinosa (L.) Gaertn. Molecular Ecology 7, 1151–1161. Ledig, F.T. (1986) Heterozygosity, heterosis, and fitness in outbreeding plants. In: Soulé, M.E. (ed.) Conservation Biology. Sinauer Associates, Sunderland, Massachusetts, pp. 77–104. Levin, D.A. (1981) Dispersal versus gene flow in plants. Annals of the Missouri Botanical Gardens 68, 233–253. Lewis, P.O. and Crawford, D.J. (1995) Pleistocene refugium endemics exhibit greater allozymic diversity than widespread congeners in the genus Polygonella (Polygonaceae). American Journal of Botany 82, 141–149. Loveless, M.D. and Hamrick, J.L. (1988) Genetic organization and evolutionary history in two American species of Cirsium. Evolution 42, 254–265. Lynch, M. (1996) A quantitative-genetic perspective on conservation issues. In: Avise, J.C. and Hamrick, J.L. (eds) Conservation Genetics: Case Histories from Nature. Chapman & Hall, New York, pp. 471–501. Lynch, M. and Milligan, B.G. (1994) Analysis of population genetic structure with RAPD markers. Molecular Ecology 3, 91–99. Lynch, M., Pfrender, M., Spitze, K., Lehman, N., Hicks, J., Allen, D., Latta, L., Ottene, M., Bogue, F. and Colbourne, J. (1999) The quantitative and molecular genetic architecture of a subdivided species. Evolution 53, 100–110. McGranahan, M., Bell, J.C., Moran, G.F. and Slee, M. (1997) High genetic divergence between geographic regions in the highly outcrossing species Acacia aulacocarpa (Cunn. ex Benth.). Forest Genetics 4, 1–13. McKay, J.K. and Latta, R.G. (2002) Adaptive population divergence: markers, QTL and traits. Trends in Ecology and Evolution 17, 285–291.

Genetic variation in plant populations 163 Maguire, T.L., Peakall, R. and Saenger, P. (2002) Comparative analysis of genetic diversity in the mangrove species Avicennia marina (Forsk.) Vierh. (Avicenniaceae) detected by AFLPs and SSRs. Theoretical and Applied Genetics 104, 388–398. Manel, S., Schwartz, M.K., Luikart, G. and Taberlet, P. (2003) Landscape genetics: combining landscape ecology and population genetics. Trends in Ecology and Evolution 18, 189–197. Merila, J. and Crnokrak, P. (2001) Comparison of genetic differentiation at marker loci and quantative traits. Journal of Evolutionary Biology 14, 892–903. Moritz, C. (1994) Defining ‘evolutionarily significant units’ for conservation. Trends in Ecology and Evolution 9, 373–375. Mueller, U.G. and Wolfenbarger, L.L. (1999) AFLP genotyping and fingerprinting. Trends in Ecology and Evolution 14, 389–393. Newton, A.C., Allnutt, T.R., Gillies, A.C.M., Lowe, A.J. and Ennos, R.A. (1999) Molecular phylogeography, intraspecific variation and the conservation of tree species. Trends in Ecology and Evolution 14, 140–145. Nybom, H. and Bartish, I.V. (2000) Effects of life history traits and sampling strategies on genetic diversity estimates obtained with RAPD markers in plants. Perspectives in Plant Ecology, Evolution and Systematics 3, 93–114. Peakall, R., Smouse, P.E. and Huff, D.R. (1995) Evolutionary implications of allozyme and RAPD variation in diploid populations of dioecious buffalograss Buchloë dactloides. Molecular Ecology 4, 135–147. Peakall, R., Gilmore, S., Keys, W., Morgante, M. and Rafalski, A. (1998) Cross-species amplification of soybean (Glycine max) simple sequence repeats (SSRs) within the genus and other legume genera: implications for the transferability of SSRs in plants. Molecular Biology and Evolution 15, 1275–1287. Powell, W., Morgante, M., NcDevitt, R., Vendramin, G.G. and Rafalski, J.A. (1995) Polymorphism simple sequence repeat regions in chloroplast genomes: applications to population genetics of pines. Proceedings of the National Academy of Sciences USA 92, 7759–7763. Provan, J., Soranzo, N., Wilson, N.J., McNicol, J.W., Morgante, M. and Powell, W. (1999) The use of uni- parentally inherited simple sequence repeat markers in plant population studies and systematics. In: Hollingsworth, P.M., Bateman, R.M. and Gornall, R.J. (eds) Molecular Systematics and Plant Evolution. Taylor & Francis, London, pp. 35–50. Raybould, A.F., Mogg, R.J. and Clarke, R.T. (1996) The genetic structure of Beta vulgaris ssp. maritima (sea beet) populations: RFLPs and isozymes show different patterns of gene flow. Heredity 77, 245–250. Reed, D.H. and Frankham, R. (2001) How closely correlated are molecular and quantitative measures of genetic variation? A meta-analysis. Evolution 55, 1095–1103. Richards, A.J. (1997) Plant Breeding Systems, 2nd edn. Chapman & Hall, London. Rieseberg, L.H. (1996) Homology among RAPD fragments in interspecific comparisons. Molecular Ecology 5, 99–105. Ryder, O.A. (1986) Species conservation and systematics: the dilemma of subspecies. Trends in Ecology and Evolution 1, 9–10. Sampson, J.F., Coates, D.J. and van Leeuwen, S.J. (1996) Mating system variation in animal-pollinated rare and endangered plant populations in Western Australia. In: Hopper, S.D., Chappill, J.A., Harvey, M.S. and George, A.S. (eds) Gondwanan Heritage: Past, Present and Future of the Western Australian Biota. Surrey Beatty and Sons, Chipping Norton, Australia, pp. 187–195. Schaal, B.A. (1980) Measurement of gene flow in Lupinus texensis. Nature 284, 450–451. Schaal, B.A. and Olsen, K.M. (2000) Gene genealogies and population variation in plants. Proceedings of the National Academy of Sciences USA 97, 7024–7029. Schaal, B.A., Leverich, W.J. and Rogstad, S.H. (1991) A comparison of methods for assessing genetic varia- tion in plant conservation biology. In: Falk, D.A. and Holsinger, K.E. (eds) Genetics and Conservation of Rare Plants. Oxford University Press, New York, pp. 123–134. Schaal, B.A., Hayworth, D.A., Olsen, K.M., Rauscher, J.T. and Smith, W.A. (1998) Phylogeographic studies in plants: problems and prospects. Molecular Ecology 7, 465–475. Schemske, D.W. and Lande, R. (1985) The evolution of self-fertilization and inbreeding depression in plants. II. Empirical observations. Evolution 39, 41–52. Schnabel, A. and Hamrick, J.L. (1995) Understanding the population genetic structure of Gleditsia triacan- thos L.: the scale and pattern of pollen gene flow. Evolution 49, 921–931. Schoen, D.J. and Brown, A.H.D. (1991) Intraspecific variation in population gene diversity and effective population size correlates with the mating system in plants. Proceedings of the National Academy of Sciences USA 88, 4494–4497.

164 D.J. Coates and M. Byrne Soltis, D.E. and Soltis, P.S. (1989) Isozymes in Plant Biology. Dioscorides Press, Portland, Oregon. Soltis, D.E., Gitzendanner, M.A., Strenge, D.D. and Soltis, P.S. (1997) Chloroplast DNA intraspecific phylo- geography of plants from the Pacific Northwest of North America. Plant Systematics and Evolution 206, 353–373. Squirrell, J., Hollingsworth, P.M., Woodhead, M., Russell, J., Lowe, A.J., Gibby, M. and Powell, W. (2003) How much effort is required to isolate microsatellites from plants? Molecular Ecology 12, 1339–1348. Stöcklin, J. (1999) Differences in life history traits of related Epilobium species: clonality, seed size and seed number. Folia Geobotanica and Phytotaxonomica Praha 34, 7–18. Sydes, M.A. and Peakall, R. (1998) Extensive clonality in the endangered shrub Haloragodendron lucasii (Haloragaceae) revealed by allozymes and RAPDs. Molecular Ecology 7, 87–93. Templeton, A.R. (1986) Coadaptation and outbreeding depression. In: Soulé, M.E. (ed.) Conservation Biology. Sinauer Associates, Sunderland, Massachusetts, pp. 105–116. Thompson, J.D. (1999) Population differentiation in Mediterranean plants: insights into colonization history and the evolution and conservation of endemic species. Heredity 82, 229–236. Vaillancourt, R.E. and Jackson, H.D. (2000) A chloroplast DNA hypervariable region in eucalypts. Theoretical and Applied Genetics 101, 473–477. Wardell-Johnson, G. and Coates, D.J. (1996) Links to the past: local endemism in four species of forest eucalypts in southwestern Australia. In: Hopper, S.D., Chappill, J.A., Harvey, M.S. and George, A.S. (eds) Gondwanan Heritage: Past, Present and Future of the Western Australian Biota. Surrey Beatty & Sons, Chipping Norton, Australia, pp. 137–154. Waser, N.M. (1993) Population structure, optimal outbreeding, and assortative mating in angiosperms. In: Thornhill, N.W. (ed.) The Natural History of Inbreeding and Outbreeding. University of Chicago Press, Chicago, Illinois, pp. 173–199. Weir, B.S. (1996) Genetic Data Analysis II. Sinauer Associates, Sunderland, Massachusetts. Williams, C.F., Ruvinsky, J., Scott, P.E. and Hews, D.K. (2001) Pollination, breeding system, and genetic structure in two sympatric Delphinium (Ranunculaceae) species. American Journal of Botany 88, 1623–1633. Wright, S. (1951) The genetical structure of populations. Annals of Eugenics 15, 323–354. Young, A.G. and Brown, A.H.D. (1996) Comparative population genetic structure of the rare woodland shrub Daviesia suaveolens and its common congener D. mimosoides. Conservation Biology 10, 374–381. Young, A.G. and Clarke, G.M. (2000) Genetics, Demography and Viability of Fragmented Populations. Cambridge University Press, Cambridge.

10 Evolution of the flower Douglas E. Soltis,1 Victor A. Albert,2 Sangtae Kim,1 Mi-Jeong Yoo,1 Pamela S. Soltis,3 Michael W. Frohlich,4 James Leebens-Mack,5 Hongzhi Kong,5,6 Kerr Wall,5 Claude dePamphilis5 and Hong Ma5 1Department of Botany and the Genetics Institute, University of Florida, Gainesville, FL 32611, USA; 2The Natural History Museums and Botanical Garden, University of Oslo, NO-0318 Oslo, Norway; 3Florida Museum of Natural History and the Genetics Institute, University of Florida, Gainesville, FL 32611, USA; 4Department of Botany, Natural History Museum, London SW7 5BD, UK; 5Department of Biology, The Huck Institutes of the Life Sciences and Institute of Molecular Evolutionary Genetics, The Pennsylvania State University, University Park, PA 16802, USA; 6Laboratory of Systematic and Evolutionary Botany, Institute of Botany, The Chinese Academy of Sciences, Beijing 100093, China Introduction development (evo-devo). Evo-devo offers the best hope for rapid advance in the The origin and evolution of the flower have understanding of flower evolution. To been intensively studied not only because of appreciate this potential one must be cog- the great importance of flowers (and espe- nizant of recent advances in all of these cially the fruits they produce) in providing fields – phylogeny, morphology and devel- human food, but also because of their cru- opmental genetics – that are merging to cre- cial role in angiosperm sexual reproduction ate evo-devo. and many plant–animal interactions. The past centuries of morphologically and taxo- Here we describe recent progress in the nomically based studies of flowers gener- study of floral evolution, beginning with ated much information, but left some of the advances in phylogeny and the reconstruc- most critical questions of flower origin and tion of trends in floral evolution. We evolution unresolved. Recent progress in include a brief comparative review of some understanding angiosperm (and seed plant) of the genes known to regulate flower phylogeny provides a solid framework for development, with an emphasis on recent evaluating evolutionary innovation, and studies relevant to the classic ABC model identifies the taxa that provide the best of flower development. We conclude with a insights into key innovations. The recent perspective on future research on floral growth of developmental genetics provides biology at the genomic level. Throughout exciting new data for understanding flower our discussion we describe how experi- evolution; the interplay of developmental mental genetic and phylogenetic analyses genetics with focused studies of morphol- are together improving our understanding ogy, development and phylogeny has gener- of the evolution of floral architecture and ated a new field of study: the evolution of the molecules regulating floral develop- ment. © CAB International 2005. Plant Diversity and Evolution: Genotypic and 165 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

166 D.E. Soltis et al. Trends in Floral Evolution Inferred from Fossils are critical for inferring the origin Phylogeny and early diversification of angiosperms, but fossil flowers of the earliest angiosperms are Background scarce. None the less, early Cretaceous angiosperm fossils are consistent with the A clear understanding of angiosperm phy- hypothesis that the first flowers were small logeny has recently emerged (e.g. Qiu et al., to moderate in size, with an undifferentiated 1999; Soltis et al., 1999, 2000; Barkman et perianth (Crane, 1985; Friis et al., 1994, al., 2000; Zanis et al., 2002, 2003). These 2000; Crane et al., 1995), although Magnolia- well-resolved and highly concordant DNA- like forms also occurred during the same based phylogenies have important implica- geological time (e.g. Archaeanthus; Dilcher tions for interpreting the morphology of and Crane, 1984). In addition, some early early angiosperms and subsequent patterns angiosperms lacked a perianth (e.g. of floral evolution. Archaefructus; Sun et al., 2002), but these may not be basal within angiosperms (Friis et al., Before the application of explicit phylo- 2003). There are no known fossils repre- genetic methods, several investigators pro- senting unequivocal stem-group angio- posed that the first angiosperms had large, sperms (i.e. angiosperms that attach below Magnolia-like flowers (Arber and Parkin, the basal node leading to Amborella, 1907; Bessey, 1915; Takhtajan, 1969; Nymphaeaceae and all other living Cronquist, 1981). Stebbins (1974), in con- angiosperms). trast, suggested that the earliest flowers were moderate in size. Endress (1987) proposed One way to infer ancestral states is to that the earliest angiosperm was bisexual, employ character-state reconstruction with but that the transition to unisexuality was phylogenetic trees and programs such as relatively easy, the perianth was undifferen- MACCLADE (Maddison and Maddison, 1992). tiated and could be easily lost, and that the Using this approach, the evolution of spe- number of floral parts was labile. cific floral characters in basal angiosperms has been reconstructed (e.g. Albert et al., Early phylogenetic studies focused 1998; Doyle and Endress, 2000; Ronse De attention on several herbaceous lineages Craene et al., 2003; Zanis et al., 2003; Soltis (e.g. Nymphaeaceae, Piperaceae and et al., 2004). We review some of the findings Chloranthaceae; Fig. 10.1) as possible first- of these character-state reconstructions branching extant angiosperms (Donoghue below using the most conservative optimiza- and Doyle, 1989; Doyle et al., 1994). Based tion method (all most parsimonious states; on these results, it was suggested that early Maddison and Maddison, 1992). Other flowers were small, with a trimerous peri- reconstructions, using other optimization anth, and with few stamens and carpels. methods and tree topologies, are provided However, more recent analyses (e.g. in the references noted above. Most of the Mathews and Donoghue, 1999; Parkinson same general conclusions are supported et al., 1999; Qiu et al., 1999; Soltis et al., regardless of optimization. 1999, 2000; Barkman et al., 2000; Doyle and Endress, 2000; Graham and Olmstead, Perianth differentiation 2000; Zanis et al., 2002, 2003; Borsch et al., 2003; Hilu et al., 2003) place Amborella, A differentiated or bipartite perianth has an Nymphaeaceae (including Cabombaceae; outer whorl of sepals clearly differentiated see APG II, 2003) and Austrobaileyales as from the inner whorl(s) of petals. In con- basal to other extant angiosperms (Fig. trast, an undifferentiated perianth lacks 10.2). This topology suggests instead that clear differentiation between the outer and the earliest flowers were small to moderate inner whorls, or the perianth may consist of in size, with an undifferentiated perianth, undifferentiated spirally arranged parts. stamens lacking a well-differentiated fila- These undifferentiated perianth organs ment, and a gynoecium composed of one or more distinct carpels.

Fig. 10.1. Floral diversity in basal angiosperms (a–i) and early-diverging eudicots (j–m). and Igersheim, 2000). (b) A. trichopoda (Amborellaceae), pistillate flower (from Endress 1994b). (d) Trimenia papuana (Trimeniaceae) (from Endress and Sampson, 1983). (e) Tri (f ) Aristolochia (Aristolochiaceae) flower (from Solereder in Engler and Prantl, 1887–191 Takhtajania perrieri (Winteraceae; Canellales) (from Endress et al., 2000). (i) Magnolia ϫ Eupomatia (Eupomatiaceae) flowering shoot (from Uphof in Engler and Prantl, 1959). (k Euptelea polyandra (Eupteleaceae) (from Endress, 1986). (m) Trochodendron (Trochoden Endress, 1986). (o) Buxus balearica (Buxaceae), inflorescence with lateral staminate flow

Evolution of the flower . (a) Amborella trichopoda (Amborellaceae), staminate flower (from Endress 167 s and Igersheim, 2000). (c) Cabomba aquatica (Nymphaeaceae) (from Endress, icyrtis pilosa (Liliaceae), flower (from Engler in Engler and Prantl, 1887–1915). 15). (g) Austrobaileya scandens (Austrobaileyaceae) (from Endress, 1980). (h) ϫ soulangiana (Magnoliaceae; Magnoliales) (from Endress, 1987). ( j) k) Sarcandra chloranthoides (Chloranthaceae) (from Endress, 1987). (l) ndraceae) (from Endress, 1986). (n) Tetracentron (Trochodendraceae) (from wers and terminal carpellate flower (from Von Balthazar and Endress, 2002).

168 D.E. Soltis et al. Fagales Cucurbitales Rosales Fabales Eurosid I Zygophyllales Celastrales Oxalidales Malpighiales Rosids Sapindales Malvales Eurosid II Brassicales Crossosomatales Myrtales Geraniales Saxifragales Lamiales Solanales Euasterid I Gentianales lamiids Garryales Asterids Asterales Dipsacales Euasterid II Apiales campanulids Aquifoliales Cornales Ericales Caryophyllales Berberidopsidales Santalales Gunnerales Buxaceae Trochodendraceae Proteales Sabiaceae Ranunculales Canellales Piperales Magnoliales Magnoliids Laurales Chloranthaceae Ceratophyllaceae Monocots Trimeniaceae Schisandraceae Austrobaileyaceae Nymphaeaceae Amborellaceae Taxaceae Podocarpaceae Pinaceae Ginkgoaceae Gymnosperms Welwitschiaceae Gnetaceae Ephedraceae Fig. 10.2. Summary topology for angiosperms showing general positions of model organisms (in bold). Modified from Soltis et al. (2003). have traditionally been referred to as tepals. and petals) that are not clearly differentiated The term tepal was coined by De Candolle morphologically; thus, the entire perianth (1827) to describe perianth organs (sepals may be petaloid. Takhtajan (1969), in con-

Evolution of the flower 169 trast, used the term ‘tepal’ in a phylogenetic Following Albert et al. (1998), two or more sense such that all monocots have tepals. whorls of perianth parts must be present for Takhtajan’s definition limits the application an unambiguous interpretation of sepals and of tepal to specific groups of angiosperms petals. If only a single perianth whorl is pre- and requires different terms for an undiffer- sent, it may be difficult to interpret as ‘sepals’ entiated perianth in other groups. Following or ‘petals’ (see also Endress, 1994a,b). Is the other recent investigators, we will use the single whorl an undifferentiated perianth, term tepal as defined by De Candolle. composed of neither sepals nor petals, or is the single whorl composed of either sepals or Distinguishing sepals from petals is not petals with the other perianth whorl absent? A always straightforward (Endress, 1994a; single-whorled perianth has traditionally been Albert et al., 1998). Whereas sepals and referred to as being composed of ‘sepals’ as a petals are readily distinguished in most eudi- matter of convention (e.g. Cronquist, 1968). cots (~ 75% of all angiosperms), this is often Families of basal angiosperms that contain not the case in basal angiosperms (Fig. 10.1), taxa with a single-whorled perianth include many of which have numerous undifferenti- nearly all Aristolochiaceae (except Saruma), ated perianth parts arranged in spirals, all Myristicaceae and Chloranthaceae rather than in distinct whorls, a condition (Hedyosmum). In some cases, however, the long considered ancestral (e.g. Bessey, 1915; nature of a single-whorled perianth can be Cronquist, 1968; Takhtajan, 1969). determined through comparison with the perianths of closely related taxa. In The origin of a differentiated perianth of Aristolochiaceae, most taxa have a single- sepals and petals has long been of interest whorled perianth that is considered a calyx (e.g. Eames, 1931; Hiepko, 1965; Kosuge, (Cronquist, 1968, 1981; Tucker and 1994; Albert et al., 1998; Kramer and Irish, Douglas, 1996; Takhtajan, 1997). In con- 1999, 2000). It has been proposed that trast, Saruma has two perianth whorls that petals evolved first and that sepals evolved are differentiated into sepals and petals. later (e.g. Albert et al., 1998) and that petals Furthermore, in some species of Asarum, have evolved multiple times from different petals apparently begin to develop, but the floral organs in different groups (e.g. Eames, only traces are small, thread-like structures 1961; Takhtajan, 1969; Kosuge, 1994; Albert (Leins and Erbar, 1985). et al., 1998; Zanis et al., 2003). In recent reconstructions (Ronse De Takhtajan (1969, 1997) suggested two Craene et al., 2003; Zanis et al., 2003; Soltis origins of petals, one from stamens and one et al., 2004) (Fig. 10.3), the ancestral state for from bracts. Support for multiple, indepen- the angiosperms is an undifferentiated peri- dent origins of petals has come from mor- anth. Amborella and Austrobaileyales have an phological studies showing that ‘petals’ of undifferentiated perianth. In contrast, the various angiosperms exhibit major differ- ancestral state for Nymphaeaceae is recon- ences and can be grouped into two basic structed as equivocal because some classes (e.g. Endress, 1994a; Kramer and Nymphaeaceae (e.g. Cabomba, Brasenia, Irish, 2000). In one group are petals that Nuphar) have a differentiated perianth resemble stamens. The petals are develop- whereas more derived waterlilies (Victoria, mentally delayed and are similar in appear- Nymphaea) have an undifferentiated peri- ance to stamen primordia at inception anth. Above the basal angiosperm grade, the (Endress, 1994a). These petals have some- undifferentiated perianth continues to be times been termed andropetals. The second ancestral for the remaining angiosperms type of petaloid organ (conventionally (Fig. 10.3). Importantly, all reconstructions termed tepals; Cronquist, 1981) is found in indicate that a differentiated perianth undifferentiated perianths and is more leaf- evolved multiple times (see Albert et al., like in general characteristics. These petals 1998). Separate origins include some initiate and mature much earlier than do Nymphaeaceae, monocots, some Mag- the stamens and are generally more leaf-like noliaceae, Annonaceae, Canellaceae, some in appearance than are other petals (Smith, 1928; Tucker, 1960; Takhtajan, 1969, 1997).

Perianth differentiation Amborellaceae unordered Nymphaea Undifferentiated Sepals + petals Victoria Absent Equivocal Nuphar Nymphaeaceae Fig. 10.3. MACCLADE reconstruction using all most parsimonious states optimization (Maddison and Maddison, 1992) of the evolution of perianth differentiation in Brasenia angiosperms, with an emphasis on basal angiosperms and early-diverging eudicots. Topology is based on Zanis et al. (2002, 2003) and Soltis et al. (2000, 2003). Data are from Zanis et al. (2003) and Ronse De Craene et al. (2003). Modified from Soltis et al. (2004). For impact of other optimization methods (ACCTRAN, DELTRAN) Cabomba see Zanis et al. (2003), Ronse De Craene et al. (2003) and Soltis et al. (2004). Illiciaceae Trimeniaceae Austrobaileyaceae Ceratophyllaceae Acorus other monocots Chloranthus Chloranthaceae Hedyosmum Bubbia Winteraceae Drimys Takhtajania Canellaceae Aristolochia Aristolochiaceae Saruma Piperaceae Saururaceae Calycanthaceae Atherospermataceae Gomortegaceae Hernandiaceae Hortonia Kibara Lauraceae Myristicaceae Magnolia Magnoliaceae Liriodendron Degeneriaceae Himantandraceae Eupomatiaceae Annonaceae Papaveraceae Eupteleaceae Menispermaceae Berberidaceae Glaucidium Ranunculaceae Ranunculus Nelumbonaceae Plantanaceae Proteaceae Sabiaceae Buxus Buxaceae Didymeles Tetracentron Trochodendron Gunneraceae Saxifragales caryophyllids rosids asterids D.E. Soltis et al. 170

Evolution of the flower 171 Aristolochiaceae and Siparunaceae with sidered whorled, because of their close additional origins in early-diverging eudi- apposition at maturity. cots (e.g. Papaveraceae, Menispermaceae, Ranunculus, Sabiaceae) and core eudicots. Although spiral phyllotaxis is present in Comparative developmental studies are Amborellaceae and Austrobaileyales, the required to test whether multiple origins of presence of whorled (and irregular) phyl- perianth differentiation were driven by simi- lotaxis in Nymphaeaceae makes the ances- lar changes in gene function and regulation. tral reconstruction for perianth phyllotaxis for the angiosperms dependent on the cod- Phyllotaxis ing of the outgroup. However, outgroup coding is problematic because the immedi- Amborella has spiral phyllotaxis (Fig. 10.1), as ate sister group of the angiosperms is do members of Austrobaileyales. In some unknown. Furthermore, no fossil group is basal families, phyllotaxis is complex. For known to have possessed flowers. If the out- example, in some Nymphaeaceae, phyl- group is coded as lacking a perianth, then lotaxis has been considered spiral, but it either a spiral or whorled phyllotaxis is now appears to be primarily whorled, or in reconstructed as equally parsimonious for some cases irregular (Endress, 2001). In the base of the angiosperms. If the outgroup some Winteraceae (Drimys and Pseudowintera), is coded as having a spirally arranged peri- phyllotaxis is primarily whorled, but occa- anth, then a spiral perianth is reconstructed sionally spiral (Doust, 2000). In Drimys as ancestral for the angiosperms. If the out- winteri, flowers within one tree vary between group is coded as having a whorled peri- spiral and whorled (Doust, 2001). anth, then a whorled perianth is ancestral for the angiosperms with a spiral perianth The distinction between spiral and evolving several times. whorled is not always clear. In Amborella, recent developmental studies indicate that Above the Amborellaceae, Nymph- some floral organs (e.g. carpels) are initiated aeaceae, Austrobaileyales grade, whorled in a nearly whorl-like manner, although they perianth phyllotaxis is reconstructed as are commonly described as spirally ancestral for all remaining angiosperms with arranged (Buzgo et al., 2004b). Studies of multiple shifts to a spiral perianth occurring other basal angiosperms reveal that in some in basal lineages, including Calycanthaceae, cases floral organs that appear to be whorled Atherospermataceae, Gomortegaceae, some in mature flowers actually result from spiral Monimiaceae, Degeneriaceae and some initiation of primordia and a bimodal distri- Magnoliaceae (Fig. 10.4). A possible trans- bution of long and short time intervals formation from whorled to spiral phyllotaxis between the initiation of consecutive organ may have occurred in Drimys and primordia (Tucker, 1960; Leins and Erbar, Pseudowintera (Winteraceae), which have a 1985; Endress, 1994a). Thus, both spiral complex phyllotaxis involving spirals and and whorled phyllotaxis of mature flowers multiple whorls (Doust, 2000, 2001; Endress result from the organs developing in a spiral et al., 2000). Still additional reversals to a sequence (Endress, 1987). For example, spiral perianth are found in the early- Illicium has spiral phyllotaxis in developing diverging eudicots Nelumbo (Proteales) and buds, but in mature flowers the carpels have Xanthorhiza, Caltha and Ranunculus an apparently whorled arrangement. (Ranunculaceae). Thus, perianth phyllotaxis Furthermore, even in some eudicots the is highly labile in basal angiosperms and in sepals initiate in a spiral sequence, with the basal eudicots (Endress, 1994b; Albert et al., later-arising sepals positioned slightly inside 1998; Ronse De Craene et al., 2003; Zanis et the earliest to originate, as reflected in their al., 2003; Soltis et al., 2004). Again, compara- imbricate arrangement at maturity. The tive developmental studies are necessary to inner organs arise in precise whorls, and determine whether unrelated taxa with con- even the sepals have traditionally been con- vergent phyllotaxis share common regula- tory networks for organ initiation.

Perianth phyllotaxis Amborellaceae unordered Nymphaea Spiral Whorled Victoria Polymorphic No perianth Nuphar Nymphaeaceae Equivocal Brasenia Fig. 10.4. MACCLADE reconstruction using all most parsimonious states optimization (Maddison and Maddison, 1992) of the evolution of perianth phyllotaxis in angiosperms, with an emphasis on basal angiosperms and early-diverging eudicots. Topology is based on Zanis et al. (2002, 2003) and Soltis et al. (2000, 2003). Cabomba Data are from Zanis et al. (2003) and Ronse De Craene et al. (2003). Modified from Soltis et al. (2004). For impact of other optimization methods (ACCRAN, DELTRAN) see Zanis et al. (2003), Ronse De Craene et al. (2003) and Soltis et al. (2004). Illiciaceae Trimeniaceae Austrobaileyaceae Ceratophyllaceae Acorus other monocots Chloranthus Chloranthaceae Hedyosmum Bubbia Winteraceae Drimys Takhtajania Canellaceae Aristolochia Aristolochiaceae Saruma Piperaceae Saururaceae Calycanthaceae Atherospermataceae Gomortegaceae Hernandiaceae Hortonia Kibara Lauraceae Myristicaceae Magnolia Magnoliaceae Liriodendron Degeneriaceae Himantandraceae Eupomatiaceae Annonaceae Papaveraceae Eupteleaceae Menispermaceae Berberidaceae Glaucidium Ranunculaceae Ranunculus Nelumbonaceae Platanaceae Proteaceae Sabiaceae Buxus Buxaceae Didymeles Tetracentron Trochodendron Gunneraceae Saxifragales caryophyllids rosids asterids D.E. Soltis et al. 172

Evolution of the flower 173 Merosity including Calycanthaceae (e.g. Calycanthus), the clade of Atherospermataceae and Among basal angiosperms, many lineages Gomortegaceae, Himantandraceae, some have numerous parts, some clades are Monimiaceae (e.g. Hortonia) and some trimerous, and others defy simple coding of Magnoliaceae (Magnolia). A perianth has also merosity. In Winteraceae, the outermost flo- been lost several times (e.g. Eupomatiaceae ral organs are in dimerous whorls, followed (see below), Piperaceae, most Chloranthaceae, by a switch to tetramerous whorls, and and Ceratophyllaceae) (Fig. 10.5). finally (in Takhtajania) a change to pentamer- ous whorls (Endress et al., 2000). Similarly, These reconstructions indicate that peri- in Magnoliaceae, the perianth of some anth merosity is labile in basal angiosperms species of Magnolia is an indeterminate spi- (see also Endress, 1987, 1994b; Albert et al., ral, whereas that of Liriodendron and other 1998; Zanis et al., 2003), a condition that con- species of Magnolia is in three trimerous tinues through the early-diverging eudicots whorls and may represent a transition from (Fig. 10.5). Dimery is often seen in early- spiral to whorled phyllotaxis (Tucker, 1960; diverging eudicots. However, trimery is also Erbar and Leins, 1981, 1983). prevalent (Ranunculales), and pentamery is seen in some taxa. In contrast, in core eudi- Amborella and Austrobaileyales have an cots, pentamery predominates. Interestingly, indeterminate spiral (Fig. 10.1). However, dimery is found in Gunnera, sister group to all within Nymphaeaceae, Cabomba, Brasenia other core eudicots. Thus, reconstructions and Nuphar they are trimerous; other gen- not only indicate that perianth merosity is era (e.g. Victoria, Nymphaea) are trimerous or labile in basal angiosperms and early-diverg- tetramerous (Endress, 2001). As found for ing eudicots, but also suggest that a dimerous phyllotaxis (above), reconstruction of the perianth could be the immediate precursor to ancestral merosity of extant angiosperms is the pentamery characteristic of eudicots dependent on the coding of merosity for the (Soltis et al., 2003). Once more, comparative outgroup. If the outgroup of the developmental studies are required to eluci- angiosperms is coded as having an indeter- date the molecular basis of changes in meros- minate number of perianth parts, then an ity throughout angiosperm history. indeterminate number is also ancestral for the angiosperms. Alternatively, if the ances- Genes Controlling Early Floral tor of the angiosperms is considered to lack Development a perianth, then it is equally parsimonious for the base of the angiosperms to be either The models trimerous or indeterminate in perianth merosity (see Zanis et al., 2003; Soltis et al., Developmental genetic analyses have pro- 2004). vided unprecedented insights into the mole- cular mechanisms that determine identities However, regardless of outgroup coding, of the principal floral organs, at least in the above the basal grade of Amborella, eudicot model organisms used for these Nymphaeaceae and Austrobaileyales, the studies. Arabidopsis thaliana and Antirrhinum ancestral character state for all remaining majus, two derived eudicots, were the first angiosperms is a trimerous perianth (Fig. models studied, and are still the best under- 10.5) (e.g. Ronse De Craene et al., 2003; Zanis stood. Investigations of these models have et al., 2003; Soltis et al., 2004). Thus, although resulted in the identification and under- the trimerous condition is typically associated standing of over 80 genes critical for normal with monocots, these results indicate that floral development, including genes trimery played a major role in the early evolu- involved in flower initiation; however, the tion and diversification of the flower true number is bound to be much larger (Kubitzki, 1987). (Zhao et al., 2001a; Ni et al., 2004) (Fig. 10.6). Careful morphological developmental Following the origin of a trimerous peri- anth, there was a return to an indeterminate spiral perianth in several basal lineages,

Perianth merosity Amborellaceae unordered Nymphaea Indeterminate Dimerous Victoria Trimerous Tetramerous Nuphar Nymphaeaceae Pentamerous Absent Brasenia Uncertain Polymorphic Cabomba Equivocal Illiciaceae Fig. 10.5. MACCLADE reconstruction using all most parsimonious states optimization (Maddison and Maddison, 1992) of the evolution of perianth merosity (merism) in angiosperms, with an emphasis on basal angiosperms and early-diverging eudicots. Topology is based on Zanis et al. (2002, 2003) and Soltis et al. (2000, 2003). Trimeniaceae Data are from Zanis et al. (2003) and Ronse De Craene et al. (2003). Modified from Soltis et al. (2004). For impact of other optimization methods (ACCTRAN, DELTRAN) see Zanis et al. (2003), Ronse De Craene et al. (2003) and Soltis et al. (2004). Austrobaileyaceae Ceratophyllaceae Acorus other monocots Chloranthus Chloranthaceae Hedyosmum Bubbia Winteraceae Basal angiosperms Drimys Takhtajania Canellaceae Aristolochia Aristolochiaceae Saruma Piperaceae Saururaceae Calycanthaceae Atherospermataceae Gomortegaceae Hernandiaceae Hortonia Kibara Lauraceae Myristicaceae Magnolia Magnoliaceae Liriodendron Degeneriaceae Himantandraceae Eupomatiaceae Annonaceae Papaveraceae Eupteleaceae Menispermaceae Early-diverging eudicots Berberidaceae Glaucidium Ranunculaceae Ranunculus Nelumbonaceae Platanaceae Eudicots Proteaceae Sabiaceae Buxus Buxaceae Didymeles Tetracentron Trochodendron Gunneraceae Core eudicots Saxifragales caryophyllids rosids asterids D.E. Soltis et al. 174

Evolution of the flower 175 Fig. 10.6. Genes that have been demonstrated genetically to regulate flowering time, floral meristem and organ identities in Arabidopsis. MADS-box genes are shown in ovals. For genes that encode other types of proteins, only those that play critical roles in floral meristem and organ identities are shown in boxes. The black lines and arrows indicate positive genetic interaction; the dotted lines with a short bar at the end represent negative genetic interactions. The arrows indicate that the specific organ identity gene(s) is (are) required for the identity of the corresponding organ. See Fig. 10.7 for an illustration of the ABC model. Although few genes have been identified that function downstream of the organ identity genes (Sablowski and Meyerowitz, 1998), a number of putative downstream genes for LFY and AP3/PI have been reported recently from microarray analysis (Schmid et al., 2003; Zik and Irish, 2003). Modified from a figure in Soltis et al. (2002), with recent information on the regulation of floral meristem identity genes by CO and FT (Schmid et al., 2003) and regulation of floral organ identity genes by EMF1, EMF2, LUG and SEU (Franks et al., 2002; Moon et al., 2003; Schmid et al., 2003).

176 D.E. Soltis et al. studies (Smyth et al., 1990) provided a foun- Homologous genes from these two models dation for evaluating the effects of mutations sometimes have different names, creating and defining gene functions. This integra- some confusion for newcomers to the field; tion of morphological and developmental we therefore often provide both names in genetic investigations has characterized the our overview. The protein products of work on several other model systems as well, DEFICIENS (DEF = APETALA3 (AP3) in including the derived monocots Zea mays Arabidopsis) and AGAMOUS (AG = PLENA and Oryza sativa (Poaceae), and to a lesser (PLE) in Antirrhinum) were found to be extent Petunia hybrida and Lycopersicon escu- from the same family of transcription fac- lentum (= Solanum lycopersicum), both of tors, which are regulators of the expression Solanaceae (Coen and Meyerowitz, 1991; of other genes (Schwarz-Sommer et al., Meyerowitz et al., 1991; Ma, 1994, 1998; 1990). This family was named MADS-box Weigel and Meyerowitz, 1994; Weigel, 1995; genes after a DNA-binding amino acid Yanofsky, 1995; Ma and dePamphilis, 2000; domain present in MCM1 (mini-chromo- Zhao et al., 2001a; Irish, 2003). some maintenance-1; from yeast), AG, DEF and SRF (serum response factor; from The best-known genes controlling floral humans). MADS-box genes encode a con- organ identity are the A, B and C function served domain that constitutes most of the genes (Coen and Meyerowitz, 1991; DNA-binding domain. Meyerowitz et al., 1991). According to the ABC model, three overlapping gene func- It had been hypothesized from mutant tions, A, B and C, act alone or in combina- phenotypes that the DEF (= AP3) and AG tion to specify the four types of floral (= PLE) genes control floral organ identity organs (Fig. 10.7). In 1990, the genes rep- in a combinatorial, whorl-specific fashion: A resenting deficiens (B class) and agamous (C function directs sepal identity; B function class) mutants were cloned from together with A specifies petals; B plus C Antirrhinum and Arabidopsis, respectively. function designates stamens; and C alone promotes carpel development (Meyerowitz Fig. 10.7. Extended ABC model for floral organ et al., 1991; Ma, 1994; Weigel and specification (modified from Theißen, 2001). Meyerowitz, 1994; see below; Fig. 10.7). The DEF and AG gene products were assigned to the B and C functions, respectively. As noted, in Arabidopsis, the A function genes are AP1 and AP2 (Fig. 10.7), the B function genes are AP3 (= DEF) and PIS- TILLATA (PI = GLO in Antirrhinum), genes that resulted from an ancient duplication event (discussed below), and the C function is specified by AG (= PLE) (reviewed in Ma, 1994; Ma and dePamphilis, 2000). Genetic studies were crucial for the identification of these gene functions, with mutations in each of these genes affecting two adjacent whorls. For example, ap3 mutants produce sepals and carpels instead of petals and stamens, respectively. Double- and triple-mutant analyses in Arabidopsis have further clarified the genetic interactions among A, B, C class genes. Expression studies have also been important in confirming aspects of the ABC model. All of the ABC MADS-box genes are expressed in the regions of the floral meris- tem that they help specify. The model is sup-

Evolution of the flower 177 ported by over-expression studies of the Genes are also known that specify the flo- ABC genes in Arabidopsis, which can place ral character of the apical meristem that any of the four flower organs in any of the forms the flower. The genes FLORICAULA four whorls. (FLO) and LEAFY (LFY) of Antirrhinum and Arabidopsis, respectively, are transcription Recently, in Arabidopsis, the role of the factors of a family unique to land plants. class E genes, SEPALLATA1, SEPALLATA2 FLO/LFY is single copy in diploid and SEPALLATA3, has been demonstrated: angiosperms. FLO/LFY is expressed in a they act redundantly to specify petals, sta- graded manner and acts synergistically with mens and carpels (Pelaz et al., 2000, 2001; the MADS-box gene SQUAMOSA/AP1 to Theißen, 2001). These genes were identified specify the floral character of the apex. through their sequence similarity to AG, These genes integrate signals from multiple rather than through individual mutant phe- pathways involved in the transition to flow- notypes. Triple-null mutants of SEP1-3 pro- ering. Some of the additional genes involved duce ‘flowers’ consisting only of sepal-like in floral specification are shown in Fig. 10.6 organs, suggesting that these related genes (e.g. Coen et al., 1990; Weigel et al., 1992; have redundant functions in controlling the Weigel, 1995; Riechmann and Meyerowitz, identity of petals and reproductive organs 1997; Ma, 1998; Theißen et al., 2000; (Pelaz et al., 2000, 2001). Floral MADS- Theißen, 2001; Zhao et al., 2001a). domain proteins can form homodimers, het- erodimers and tetramers, providing a New model plants mechanism for the interaction of genes within and between the A, B, C and E func- Exhaustive studies of a few key model plants, tions (Theißen, 2001) (see Fig. 10.10). chiefly Arabidopsis and Antirrhinum, have pro- vided enormous insights into the genetic In addition to A, B, C and E function control of flower development. However, a genes, numerous other genes are also regu- key question is, are the models of the genetic lators of normal floral development. control of floral development in these Furthermore, not all floral regulators are derived eudicots applicable to all MADS-box genes. In Arabidopsis, the non- angiosperms? Interestingly, the conservation MADS APETALA2 (AP2) confers A function of A function is unclear in angiosperms other along with the MADS gene APETALA1 than Brassicaceae. Another floral develop- (AP1= SQUAMOSA (SQUA) in Antirrhinum). mental model emphasizing the B and C LEAFY (LFY), which controls the entire flo- functions alone (called at that time A and B) ral developmental programme, codes for a was developed even before the ABC model, previously unknown type of transcriptional and this focus might be more broadly applic- regulator (Weigel et al., 1992). Space does able (Schwarz-Sommer et al., 1990) (Fig. not permit review of all of the numerous 10.8). The genetic architecture of floral genes involved in floral development here. development in angiosperms other than the Readers are encouraged to consult recent well-known models should also be investi- reviews (e.g. Ma, 1998; Zhao et al., 2001a; Ni gated (e.g. Albert et al., 1998; Kramer and et al., 2004; Fig. 10.6). Because most genes Irish, 2000; Soltis et al., 2002). To obtain with known functions in flower development maximal benefit from the enormous have been detected through their single- resources afforded by well-developed models gene mutant phenotypes, genes such as the for floral developmental genetics, it is imper- sepallata genes with redundant function ative that researchers expand their emphasis (Pelaz et al., 2000; Theißen, 2001), or genes to include additional species representing a that are lethal when disrupted, are not usu- wider phylogenetic coverage of angiosperms. ally discovered except through detailed fol- low-up analysis. As a result, even this rapidly The rapid increase in interest in the evo- growing collection of genes of known func- lutionary developmental biology (‘evo-devo’) tion must be considered an underestimate of of the flower has stimulated the investiga- the genes with critical roles in flower devel- opment.

178 D.E. Soltis et al. (a) Original B A C Sepal Petal Stamen Carpel (b) Basal B angio? A C (c) Missing sepal Tepal Tepal Stamen Carpel B A C Petal Missing Stamen Carpel (d) Missing B1 petal A C (e) Sepal Missing Stamen Carpel B1 A C Tepal Tepal Stamen Carpel F B1 (f) AC Sepal Tepal Stamen Carpel (petal) G B1 (g) AC Tepal Petal Stamen Carpel (sepal) B1 (h) FG C Sepal Petal Stamen Carpel Fig. 10.8. The original ABC model (a) with variations that could explain morphological changes. Versions (b–d) simply allow the change of the domains of A and B functions to account for the diversity in the perianth. Version (e) makes the control of the tepal identity similar to that of the sepal identity in derived eudicots, although tepals are often morphologically similar to petals. Versions (f–h) propose ‘F and G functions’ different from the ABC functions in distribution and in consequence to control perianth identities. B1 is used instead of B when the function is only used to control the stamen identity. tion of a number of new ‘model’ plants, and significant phylogenetic positions (Fig. 10.2). many of these are under investigation as Amborella (Amborellaceae) and waterlilies part of genomics initiatives (Soltis et al., (Nymphaeaceae) were chosen because they 2002; De Bodt et al., 2003). New models represent the sister groups to all other have typically been chosen based on their angiosperms. Other basal angiosperms (e.g.

Evolution of the flower 179 Lauraceae and Magnoliaceae) are also the Limits on the generality of floral focus of study, as is Acorus (Acoraceae), the developmental genetics sister to all other monocots. Poppies (Papaver and Eschscholzia) are important choices Through molecular evolutionary and gene because they represent an early-diverging exchange studies, it was determined that eudicot lineage (Fig. 10.2) and provide a AP3 represents the Arabidopsis homologue of critical link between derived eudicot models DEF from Antirrhinum. Similarly, sequence (e.g. Arabidopsis and Antirrhinum) and basal and functional homologies were found angiosperms. between AG and LFY and their Antirrhinum counterparts (PLE and FLO). However, the The growing list of new models not only situation with A function genes is more com- expands the phylogenetic diversity under plex. Mutations in the Antirrhinum study, but also the diversity of floral form SQUAMOSA gene, a likely orthologue of AP1 that is currently under molecular and from Arabidopsis, cause floral meristem genetic investigation (Soltis et al., 2002; De defects similar to ap1 mutants. Flowers of Bodt et al., 2003). In addition to the stan- squamosa mutants also exhibit defects in petal dard whorled arrangements of parts, new development, although the role of models such as Amborella exhibit a spiral SQUAMOSA in controlling petal identity is perianth that is undifferentiated. Gerbera, a thought to be less important than that of derived asterid in the sunflower family, is AP1 in Arabidopsis. Recently, two Antirrhinum also a useful model because of its divergent homologues, LIPLESS1 and LIPLESS2 (LIP1 inflorescence format: multiple flowers of dif- and LIP2), of the Arabidopsis AP2 gene, have ferent phenotypes borne together in a dense been shown to have redundant functions in head (Yu et al., 1999; Kotilainen et al., 2000). controlling sepal and petal development (Keck et al., 2003) in a manner similar to that The new models also have important lim- of AP2. However, unlike AP2, LIP1/2 do not itations. For most, genetic studies are not yet seem to be involved in the negative regula- possible. Although developmental morpho- tion of the C-function gene PLENA (PLE = logical and molecular studies can lead to the AG). Therefore, results from Antirrhinum formulation of useful hypotheses regarding support a critical role for A function in deter- the evolution of gene functions, these await mining perianth identities, but the interac- testing using genetic studies. For basal tions between known genes involved in A angiosperms that are woody (e.g. Amborella, and C functions seem to be different between Persea) and not readily analysed genetically, Arabidopsis and Antirrhinum. AP1- and AP2- definitive conclusions about gene functions like genes have been identified from a will be difficult to achieve. Therefore, herba- diverse array of angiosperms (Litt and Irish, ceous basal angiosperms (e.g. the waterlily 2003); however, whether they play a role in Cabomba) and herbaceous basal eudicots (e.g. A function is not known. Papaver or Eschscholzia) may have the great- est potential as new models because of their Therefore, the existence of a conserved A short life cycles and the transformability of function in angiosperm flower development Papaver (Baum et al., 2002). is still uncertain, although there should be gene functions that specify sepal identity in New technologies might provide effective species that produce a differentiated peri- methods for reverse genetic analysis of new anth. It is possible that different genes serve genetic models. Methods that use viruses to this function in different flowering plant lin- generate small, interfering RNA and to post- eages, or that the determination of sepal transcriptionally silence a gene of interest (Lu and petal identity is more complex than et al., 2003) might be applicable in mature depicted in the ABC model. Based on the plants, even in long-lived perennials. Such presence of B- and C-function MADS-box new methods, if perfected, that allow easy genes in gymnosperms (which lack flowers), elucidation of gene function in diverse plants it has been hypothesized that determination by mutation or by gene silencing, could of flower organ identity has evolved from a become as important for evo-devo studies as PCR has been for molecular phylogenetics.

180 D.E. Soltis et al. more ancient role of these genes in sex the applicability of the ABC model to all determination (Hahn and Somerville, 1988; angiosperms. Certainly much of the ABC Münster et al., 1997; Winter et al., 1999). It framework is conserved in a number of has also been hypothesized that ‘true’ sepals eudicots and grasses, but there are impor- of the kind expressed by Arabidopsis and tant variations on the ABC theme in some Antirrhinum are a relatively recent evolution- flowering plants (Fig. 10.8). For example, in ary innovation, because basal eudicots and contrast to the well-differentiated sepal and monocots characteristically lack discrete petal whorls of eudicots such as Arabidopsis sepals and petals and bear tepals instead. and Antirrhinum, the two outer floral whorls in many members of the monocot family Conservation of control over floral specifi- Liliaceae (lily family) are petaloid and almost cation of the flower apex by FLO/LFY has identical in morphology (Fig. 10.1). been shown for several eudicots, but addi- Importantly, in Tulipa (tulip), the B-class tional functions are also known. For example, genes are expressed in both petaloid whorls, the LFY homologue of Pisum (Fabaceae) con- as well as in stamens (Kanno et al., 2003). trols compound leaf development in addition This situation supports the idea that petals to the transition to flowering. In grasses, LFY and petal-like organs require B function, homologues probably direct the development regardless of the position of these organs of inflorescence meristems rather than only within the flower. floral meristems, as in Arabidopsis. Similarly, in some Nymphaeaceae As noted above, the SEPALLATA (SEP) (waterlilies) such as Nuphar, the outer whorl genes of Arabidopsis provide redundant func- of the flower, sometimes referred to as sepals, tion required for floral organ identity. exhibits B class gene expression, as do the However, research on other organisms such petals, stamens and staminodes (Kim et al., as Gerbera has shown that SEP-class homo- 2004) (Fig. 10.8). In Amborella, which has a logues play divergent roles in development spirally arranged perianth with parts that are of the condensed, head inflorescence as well not differentiated into sepals and petals (Fig. as in the different floral forms that are borne 10.1), a similar pattern is observed, with B by it. Specifically, one SEP-like gene confers class genes expressed throughout the peri- the C function only in the staminal whorl of anth, as well as in the stamens (Fig. 10.8). flowers borne at the periphery of the inflo- Similar expression data have been forthcom- rescence, whereas another SEP homologue ing for basal angiosperms in the magnoliid (the probable duplication partner of the first) clade. In Magnolia (Magnoliaceae), B class appears to confer the C function only in gene expression has been documented carpels (Teeri et al., 2002). This partitioning throughout the perianth whorls, as well as in of genetic function has probably had mor- stamens and staminodes (Kramer and Irish, phological evolutionary consequences 2000; Kim et al., unpublished). A similar pat- because the outer flowers of Gerbera inflores- tern of B-class gene expression has been cences are male-sterile and highly asymmet- observed in basal eudicots such as Papaver rical, with fused, elongate petals, whereas the (Papaveraceae) and various members of inner flowers are bisexual and very close to Ranunculaceae (Kramer and Irish, 2000; symmetrical, with non-elongate petals. The Kramer et al., 2003). The expression of B- Gerbera inflorescence looks very much like a function genes in sepal-like organs suggests single flower, and probably attracts pollinat- that these B-function genes are not sufficient ing insects in the same capacity. to specify petal identity. The ABC Model: New Data, New Views The expression of C-class genes has also been examined in several basal ABCs of basal angiosperms angiosperms, and the results for this gene match the predictions of the ABC model. Recent investigations of basal angiosperms For example, homologues of AGAMOUS have provided an important assessment of have been isolated from Amborella, and these are expressed in carpels, stamens and sta-

Evolution of the flower 181 minodes (Kim et al., unpublished). Data for made between functional and phylogeneti- the expression of A-class genes from the cally based classifications of gene relation- basal-most angiosperms remain fragmentary. ships (Becker and Theißen, 2003). Thus, recent data suggest a modified ABC AP3/PI-like genes: an ancient duplication model for basal angiosperms, with B-class genes expressed and presumably functioning The evolution of MADS-box genes has throughout the perianth and stamens (Fig. involved a series of gene duplications and 10.7) (see Van Tunen et al., 1993; Albert et al., subsequent diversification, as well as losses. 1998) following the original ‘BC model’ idea Several investigators have conducted phylo- put forth by Schwarz-Sommer et al. (1990). genetic analyses of the floral MADS-box From a phylogenetic standpoint, the ABC genes (e.g. Purugganan, 1997; Theißen et model may reflect a more recent programme al., 2000; Johansen et al., 2002; Nam et al., that is important in Arabidopsis and possibly 2003; Becker and Theißen, 2003). For other eudicots. The specification of sepals, example, a duplication yielding the A and E which may have evolved more than once + AGL6 class genes occurred approximately (Albert et al., 1998), may well be encoded by 413 million years ago (mya) (Nam et al., different genes in different angiosperm lin- 2003), and the ages of several other promi- eages. The pattern of B-class gene expres- nent MADS-box gene duplications have also sion observed in basal angiosperms and basal been estimated (e.g. Purugganan et al., eudicots probably represents the ancestral 1995; Purugganan, 1997; Nam et al., 2003). condition, with the model originally pro- posed for Arabidopsis and Antirrhinum a Whereas angiosperms possess two B-class derived modification (Fig. 10.8). paralogues (AP3 = DEF, and PI = GLO), only one certain B-class homologue has been An important evolutionary question now found in gymnosperms, suggesting that an becomes: at which node in the angiosperm ancient duplication led to the presence of the tree did the switch from the more general AP3 and PI homologues. However, the accel- BC model occur? Functional studies in phy- erated rate of evolution of AP3 and PI rela- logenetically critical taxa are required before tive to other MADS-box genes precluded this question can be answered, but the estimation of the age of the AP3/PI duplica- switch probably coincided with the evolution tion by molecular clock-based substitution of the core eudicots (Fig. 10.2). Other rate methods (Purugganan et al., 1995; important changes in floral genes similarly Purugganan, 1997; Kramer et al., 1998; Nam appear to coincide with the origin of core et al., 2003). Tree-based methods using a data eudicots, including duplication of AP3 yield- set of over 20 new AP3 and PI gene ing the euAP3 gene lineage, as well as the sequences for basal angiosperms estimated origin of AP1 (Kramer et al., 1998; Litt and that the AP3/PI duplication occurred approx- Irish, 2003). imately 260 mya (range of 230–290 mya) (Kim et al., 2004). This date places the dupli- Molecular phylogenetic analyses of the cation shortly after the split between extant gene families involved in floral development gymnosperms and angiosperms and on the are elucidating the important role that gene ‘stem’ lineage of extant flowering plants. This duplication has played in the evolution of indicates that the AP3/PI duplication flower development. The gene duplications occurred perhaps 100 million years before and losses evident in gene family phyloge- the oldest fossil flowers (generally placed at nies can confuse discussions of functional 125–131.8 mya; Hughes, 1994). Thus, this evolution when the genes with equivalent suggests that the joint expression of AP3 and function in different species are not ortholo- PI did not immediately result in the forma- gous. At the same time, orthology does not tion of petals, structures for which they con- always coincide with strict functional equiva- trol development in extant angiosperms, lence (e.g. the discussion of LIP genes in because no such structures are present in the Keck et al., 2003). Given the lack of perfect correspondence between gene function and phylogeny, a clear distinction should be

182 D.E. Soltis et al. fossil record at that time. This raises the ques- ation of heterodimerization has been tion: what was the early (pre-angiosperm) hypothesized for the morphologically more role of the AP3 and PI homologues? The co- stereotyped core eudicots (Winter et al., expression of AP3 and PI homologues could 2002). However, the phylogenetic point at reflect an evolutionary innovation of animal- which heterodimerization became enforced attractive, petal-like organs well before the is not yet clear. Hints from sequence com- appearance of angiosperms in the fossil parison suggest that Amborella, and perhaps record. Indeed, some fossil, non-angiosperm some other basal lineages (e.g. seed plants from the appropriate timeframe, Nymphaeaceae), may have retained some such as the glossopterids, had sterile spathe- capacity for B-class homodimerization. like organs attached to male or female repro- ductive structures (e.g. Crane, 1985). Because Amborella proteins may have K- domain heterodimerization signals that differ Transcription-factor complexes: early from those in Arabidopsis and other well-stud- flexibility? ied angiosperms, the data suggest that Amborella B-function proteins may have dif- A striking result of Kim et al. (2004) is the ferent dimerization dynamics from monocots strong similarity between Amborella AP3 and and core eudicots. Two different AP3 genes PI C-domain amino acid sequences (Fig. are present in Amborella (Amborella AP3-1 and 10.9). The C domains, as well as K- and AP3-2). Amborella may be capable of forming MADS-domains, signal the assembly of mul- PI/PI, AP3-1/AP3-1 and AP3-2/AP3-2 homo- timers for several MADS proteins in core dimers and perhaps AP3-1/AP3-2 het- eudicots (Egea-Cortines et al., 1999; Ferrario erodimers. Furthermore, if the amino acid et al., 2003). Indeed, higher-order multi- residues in the K1 subdomain of Amborella mers are probably the active state of B-func- AP3 and PI are not sufficient to prevent het- tion MADS-box proteins (Egea-Cortines et erodimerization, but only weaken it, perhaps al., 1999; Honma and Goto, 2000; Theißen, Amborella can also form AP3-1/PI and AP3- 2001; Ferrario et al., 2003). 2/PI heterodimers (Fig. 10.10). Recent stud- ies using transgenic Arabidopsis plants indicate Heterodimerization of AP3 and PI pro- that the C terminus of AP3 is sufficient to teins is required for DNA binding in the confer AP3 functionality on the paralogous core eudicots that have been studied. PI protein (Lamb and Irish, 2003). This find- However, PI/PI homodimers are possible in ing, when considered in the light of Amborella some monocots, at least in vivo, although it is and its indistinct AP3 and PI C domains, also not clear whether these can bind DNA (Fig. supports the possibility of AP3-1/AP3-2 het- 10.10). (Even the Arabidopsis PI proteins can erodimerization. form homodimers, but these cannot bind DNA (Riechmann et al., 1996).) Selective fix- A simple extension of the Arabidopsis ‘quar- tet model’ for MADS protein function (Fig. 10.10; Theißen, 2001) can accommodate both

Fig. 10.9. AP3/PI gene structure in basal angiosperms. (a) (Opposite.) The size of exon 5 in Amborella and Nuphar compared with that observed in Arabidopsis and other eudicots. (b) Comparison of AP3/PI domain similarities in Amborella and other basal angiosperms (from Kim et al., 2004).

Evolution of the flower 183

184 D.E. Soltis et al. Fig. 10.10. Transcription-factor complexes. (a) Quartet model of floral organ specification in Arabidopsis (Theißen, 2001). (b) Extension of the quartet model for determination of floral organ identity to include Amborella. MADS protein tetramers are shown schematically (as in (a); see Thei␤en, 2001). One possible model is presented with the following assumptions: (i) AP3/PI obligate heterodimerization occurs in core eudicots; (ii) additional MADS proteins X and Y are available in cells; (iii) PI/PI dimers can tetramerize with all three configurations of X and Y in monocots (Winter et al., 2002) and possibly other basal angiosperms, whereas only the XY configuration is possible in core eudicots; (iv) Amborella AP3/AP3 and PI/PI dimers possess similar capacities for C-domain tetramerization specification; and (v) Amborella AP3-1 and AP3-2 proteins are able heterodimerize. Tetramer potential would be 12:4:1 for Amborella, monocots (and perhaps other basal angiosperms) and core eudicots, respectively. If Amborella AP3 and PI can also heterodimerize to some extent, the ratio of possible quartets in Amborella:monocots:core eudicots becomes 18:4:1.

Evolution of the flower 185 the monocot and Amborella cases. In this hypo- control of B-function MADS-box genes in thetical example, the MADS protein tetramer the development of the earliest flowers was AP3/PI/X/Y specifies a particular organ iden- dynamic, with different ‘experiments’ tried. tity in Arabidopsis. Assuming that homodimer- Amborella, which may be the most flexible ization is the ancestral state for B-function living angiosperm in its developmental proteins (Winter et al., 2002), we can invoke a genetics, is the sole surviving representative model whereby PI/PI homodimers, as known of its clade. Some of this same biochemical from and argued to have functional signifi- flexibility may also be present in waterlilies. cance in monocots (Münster et al., 2001; These are testable hypotheses, to be pur- Winter et al., 2002), are more flexible in their sued with more rigorous molecular investi- protein partnerships. This scenario could call gations. None the less, Amborella B-function for the possibility of PI/PI/X/Y, PI/PI/X/X and proteins would have represented a consider- PI/PI/Y/Y tetramers in monocots and most able increase in complexity over the demon- other basal angiosperms (Fig. 10.10). strated B-protein homodimerization known for conifers (Sundström et al., 1999) and Although this hypothesis must be tested Gnetales (Winter et al., 2002). However, the using gel shift and yeast 2-, 3- and 4-hybrid amino acid structural evidence suggests that assays (Winter et al., 2002; Ferrario et al., this flexibility was rapidly lost before the 2003), the implications of this model (Fig. bulk of the angiosperm radiation occurred. 10.10; Kim et al., 2004) are that Amborella The unique phylogenetic position of would have 12 times more tetramer possibili- Amborella and waterlilies, coupled with their ties than a core eudicot and three times apparently ancestral and flexible mode of B- greater tetramer potential than a monocot or gene function, make them model organisms other basal angiosperm with limited homo- that should be studied more intensively. dimerization potential. Given that Amborella may have the capacity to form more different The Early Floral Genome protein quartets for a given number of genes, it should possess more distinct con- Rice and Arabidopsis: similarity in gene copy trols (and therefore flexibility) over organ number identity and development than any other flowering plant. The waterlily Nuphar also Early angiosperms clearly had the basic has considerable C-domain similarity for the framework of B- and C-function genes in AP3 and PI proteins, and this may be suffi- place. However, these genes are only a few cient to provide the Nymphaeaceae with at of those involved in floral organ develop- least some extra tetramerization possibilities. ment and identity (Fig. 10.6) (Zhao et al., By contrast, Illicium, which represents the 2001a). Complete sequencing of the rice and next most basal clade of angiosperms after Arabidopsis genomes has made it possible to the Nymphaeaceae (Austrobaileyales; Fig. conduct comparisons of floral gene homo- 10.2), has lost most of the C-domain AP3/PI logues shared by a derived monocot and a similarity. Furthermore, a deletion in the K derived eudicot. These comparisons reveal a domain of PI (Fig. 10.9) first appears in striking similarity in the number of homo- Illicium (Austrobaileyales) and is fixed in all logues of genes involved in floral identity in other angiosperms. Although most flowering the two species (Fig. 10.11). The similarity in plants are canalized in their possibilities for gene family sizes is surprising given that the heterodimerization and multimer formation genome of rice is four times larger than that (Theißen, 2001), several eudicots (e.g. of Arabidopsis and the predicted number of Ranunculales, Kramer et al., 2003; Petunia, protein-coding genes is just over twice as Ferrario et al., 2003) and monocots (Münster large in rice (Goff et al., 2002; Yu et al., et al., 2001) may have regained some poten- 2002). Similarities in gene family size may be tial for developmental flexibility by a differ- due to conservation of orthologous sets of ent mechanism involving later duplications rice and Arabidopsis genes or conservation of of AP3 homologues, PI homologues, or both. The data suggest that the evolution of the

186 D.E. Soltis et al. Fig. 10.11. Similarity of size of gene families that contain key floral regulators (Fig. 10.6) in two distantly related flowering plant species, Arabidopsis (shaded bars) and rice (open bars). The proteomes of Arabidopsis (26,993 proteins) and rice (62,657 proteins) were gathered into 20,934 ‘tribes’ or putative gene families (Wall et al., unpublished) using the Markov-clustering method of Enright et al. (2002). gene number following independent gene angiosperms; Friis et al., 2001), the data sug- duplications and losses after the split of the gest that early angiosperms possessed a monocot and eudicot lineages at least 125 diverse tool kit of floral genes. mya. Phylogenetic analyses of gene families allow us to test these hypotheses and investi- As more genes are examined phylogeneti- gate the evolution of gene function. cally, it is also clear that there are different types of floral gene histories. In some cases, Basal angiosperms: a diverse tool kit of floral the gene phylogenies roughly track organis- genes mal phylogeny. This is the case for the B-class genes, PI and AP3. The single-copy gene As more data have emerged from major Gigantea also appears to track organismal EST (expressed sequence tag) projects on phylogeny (Chanderbali et al., unpublished). angiosperms, it has become possible to make However, several gene families present in rice broader comparisons of some of the numer- and Arabidopsis exhibit an array of different ous genes and gene families that are evolutionary patterns (see below). involved in normal floral development. Particularly useful have been ESTs obtained AP3/PI-like genes for several basal angiosperms (www. floralgenome.org). Many genes identified in Phylogenetic analyses of AP3 and PI homo- rice and Arabidopsis have clear homologues logues (Kim et al., 2004) resulted in gene in basal angiosperms. Given that extant trees that generally track the organismal basal angiosperms represent old lineages phylogeny (Fig. 10.12) inferred from analy- (the waterlily lineage, for example, is among ses of large data sets of plastid, mitochondr- the oldest in the fossil record of ial and nuclear rDNA sequences. Amborella and Nuphar (Nymphaeaceae) appear as sis- ters to all other angiosperms, in complete

Evolution of the flower 187 Fig. 10.12. AP3 and PI gene trees. Strict consensus of 72 equally most parsimonious trees (shown as a phylogram) using M-, I-, K- and C-domain regions of amino acid sequences. Numbers above branches are bootstrap values; only values above 50% are indicated (from Kim et al., 2004).

188 D.E. Soltis et al. agreement with the organismal phylogeny clades of SHAGGY-like kinase genes that mir- (see references in ‘Background’). Several rored the AtSK subgroups reported for clades of AP3 and PI homologues that cor- Arabidopsis. Importantly, SHAGGY-like kinase respond to well-supported organismal genes from rice and from basal angiosperms clades were consistently recognized by Kim were also represented in these four clades. et al. (2004), including Magnoliales and For example, SHAGGY-like kinase ESTs from monocots. The euAP3 gene clade, which basal angiosperms appeared in all four of the was previously described (Kramer et al., subclades noted (Fig. 10.13). These data indi- 1998), was recovered in most analyses. cate that SHAGGY-like kinase genes diversi- fied into these four well-marked clades early SHAGGY-like kinases in angiosperm evolution. The SHAGGY/GSK3-like kinases are non- SKP1-like proteins receptor Ser-Thr kinases that play numerous roles in plants and animals (Kim and Gene duplications within the angiosperms are Kimmel, 2000). The rice and Arabidopsis pro- also important in the history of the SKP1 gene teomes include 69 and 79 SHAGGY-like family. SKP1 (S-phase kinase-associated pro- kinases, respectively, but these genes can be tein 1) is a core component of Skp1-Cullin-F- subdivided into smaller gene families. For box protein (SCF) ubiquitin ligases and example, ten Arabidopsis genes were identi- mediates protein degradation, thereby regu- fied as forming a clade with SHAGGY itself lating many fundamental processes in eukary- (AtSK genes; Dornelas et al., 2000). The AtSK otes such as cell-cycle progression, family was shown to form four subclades in a transcriptional regulation and signal trans- phylogenetic analysis (Charrier et al., 2003): duction (Hershko and Ciechanover, 1998; (i) AtSK41 and AtSK42 formed a subclade sis- Callis and Vierstra, 2000). Among the four ter to the remaining genes, which were components of the SCF complexes, Rbx1 and weakly supported as a clade (bootstrap sup- Cullin form a core catalytic complex, an F-box port < 50%); (ii) AtSK31 and AtSK32 formed protein acts as a receptor for target proteins, a second subclade sister to the remaining and SKP1 links one of the variable F-box pro- genes, which formed a well-supported (88% teins with a Cullin (Zheng et al., 2002). There bootstrap) clade (this clade was composed of is only one known functional SKP1 protein in the two remaining subclades, each of which human and yeast, and this unique protein is received strong bootstrap support); (iii) able to interact with different F-box proteins AtSK21, AtSK22 and AtSK23 (100%); and (iv) to ubiquinate different substrates (Ganoth et AtSK11, AtSK12, AtSK13 (98%). The AtSK al., 2001). In some plant and invertebrate loci appear to have diverse functions. species, however, there are multiple SKP1 Mutant-based analyses indicate that AtSK11 genes, which have evolved at highly heteroge- and AtSK12 have a role in floral develop- neous rates (Farras et al., 2001; Nayak et al., ment; expression analyses suggest that 2002; Yamanaka et al., 2002; Kong et al., AtSK31 is flower-specific (Charrier et al., 2004). The extreme rate of heterogeneity 2003). The SHAGGY-like kinases are also observed among the 38 rice and 19 Arabidopsis involved in plant responses to stress. SKP1 homologues raised concerns that long- branch attraction may obscure true relation- The Floral Genome Project research con- ships in phylogenetic analyses of the entire sortium has obtained ESTs for a number of gene family. For this reason, Kong et al. (2004) SHAGGY-like kinase genes in basal partitioned the original data set into subsets of angiosperms. Yoo et al. (2005) conducted a genes with slow, medium and rapid rates of phylogenetic analysis of all SHAGGY-like evolution and analysed each group separately. kinase genes available in public databases, as Most SKP1 homologues observed in EST well as the ESTs from basal angiosperms (Fig. databases were included in the set of slowly 10.13). Plant SHAGGY-like kinase genes form evolving genes. In Arabidopsis, the slowly a well-supported clade distinct from those of evolving SKP1 genes were expressed more animals (see also Charrier et al., 2003). Across widely (in more tissues and more develop- all angiosperms, Yoo et al. identified four

Evolution of the flower 189 Fig. 10.13. SHAGGY-like kinase tree. Strict consensus of equally most parsimonious trees (shown as a phylogram) based on phylogenetic analysis of amino acid sequences. Closed triangle represents GSK/SHAGGY-like protein kinase from Arabidopsis and open triangle represents Oryza. Clade designations (I–IV) follow those given to Arabidopsis sequences (see text). ESTs provided by the Floral Genome Project (www.floralgenome.org) are underlined; monocot-specific clades are indicated by a vertical bar. Numbers above branches are bootstrap values; only values above 50% are indicated (from Yoo et al., 2005).

190 D.E. Soltis et al. mental stages) and at higher levels than the tions and function of organs within a flower, more rapidly evolving rice and Arabidopsis from developmental morphology, from phy- SKP1 homologues. In addition, the strength logeny, from developmental genetics, or a of purifying selection was found to be signifi- combination of these approaches (Albert et cantly greater in the slowly evolving al., 1998; Buzgo et al., 2004a). Albert et al. Arabidopsis SKP1-like genes (Kong et al., 2004). (1998) were among the first to explore the Taken together, these results suggest that the topic of using gene expression data as one slowly evolving SKP1 homologues serve the means of determining floral organ identity, most fundamental function(s) to interact with and this application of expression informa- Cullin and F-box proteins. tion continues to be a topic of debate. As an example, there are now divergent definitions The two slowly evolving Arabidopsis SKP1- of perianth organs and interpretations of like 1 and 2 genes, ASK1 and ASK2, are organ identity. Sepals typically are the outer- important for vegetative and flower develop- most organs of the flower, whereas petals are ment and essential for male meiosis (Samach conspicuous organs, typically of the second et al., 1999; Yang et al., 1999; Zhao et al., perianth whorl. The two outer floral whorls 1999, 2001b, 2003). Slowly evolving SKP1 in Tulipa may be positionally homologous to homologues from other plant species usually sepals and petals, respectively. However, both have very similar sequences, suggesting that whorls are morphologically petaloid, and, as they may also serve similar fundamental noted, patterns of B-class gene expression in functions (Kong et al., 2004). Multiple slowly both whorls resemble those of eudicot petals evolving SKP1 homologues have been sam- (Kanno et al., 2003). If gene expression pat- pled in EST studies for a variety of terns are conserved across the broad phylo- angiosperm species, including Liriodendron, genetic distances from Arabidopsis to tulip, Persea, Mesembryanthemum, Vitis, Medicago, then these data suggest homology of both Lotus, Rosa, Arabidopsis, Brassica, Gossypium, whorls to petals. Alternatively, changes in Helianthus and Solanum. While relationships expression patterns of B-function genes may are poorly resolved across the angiosperm have occurred during angiosperm evolution SKP1 phylogeny, it is clear that gene duplica- (e.g. ‘shifting borders’; Kramer et al., 2003); if tion events that have occurred throughout so, similar expression patterns may not indi- angiosperm history have contributed to this cate homology. Extension of expression and set of conserved genes (Fig. 10.14). Recent functional data to homology assessment of duplication has increased SKP1 gene diver- the lodicules of grasses is even more challeng- sity in Brassica, Helianthus, Medicago and ing. Although lodicules occur in the position Triticum. In contrast, conserved paralogues in of petals and exhibit B-class gene expression Liriodendron, Persea, Mesembryanthemum, Vitis, (e.g. Schmidt and Ambrose, 1998), as pre- Lotus, Rosa, Arabidopsis, Gossypium and dicted for petals, their unique morphology Solanaceae are the products of ancient dupli- suggests that they may not be ‘petals’, despite cation events. Interestingly, the basal position their position and gene expression patterns. of the sole Amborella SKP1 homologue sam- Thus, morphology, developmental data and pled from a set of 10,000 ESTs suggests that genetic data may provide conflicting evidence all of these duplications occurred after the of homology (organ identity) and yet ulti- origin of the angiosperms (Fig. 10.14). mately a more complete, and complex, view of a structure (Buzgo et al., 2004a). Homology of Floral Organs: Extending Out from New Models Eupomatia: a case study Can we use expression data to determine Eupomatia (Eupomatiaceae; Fig. 10.1) is a organ identity? genus of two species that possess an unusual structure (a calyptra) that encloses and pre- The homology of characters leading to the sumably protects the flower in bud. The ori- assessment of organ identity can be inferred gin of the calyptra has been debated. Some from the mature phenotype, from the posi-

Evolution of the flower 191 Fig. 10.14. Phylogenetic relationships of slowly evolving SKP1 proteins from select rosid (ROS, ROS1, ROS2), euasterid (AST), monocot (MON), magnoliid (MAG), basal angiosperm (BAS) and gymnosperm (GYM) species suggest that gene duplications have occurred both before and after the origin of major taxa within the angiosperms. The maximum likelihood tree shown is one of >2000 most parsimonious trees. Maximum parsimony bootstrap values higher than 50% are shown above or below the branch. Support for each branch was tested with 500 replicates of bootstrap analysis using random input order for each replicate. Note that most nodes on the tree are not well supported because the regions used for analysis are rather short (149 amino acids) and highly conserved. The taxonomic categories for these species follow Soltis et al. (2000). ESTs generated by the Floral Genome Project are underlined. Closed triangles indicate SKP1 homologues from Arabidopsis and open triangles those from rice. Modified from Kong et al. (2004).

192 D.E. Soltis et al. have considered it to represent a modified Hughes, 1994), but each typically accounts perianth (Cronquist, 1981), but develop- for only a limited range of observations, and mental data indicate that it represents a none is testable, unless revealing fossils hap- modified bract (Endress, 2003). pen to be discovered. Eupomatiaceae are closely related to the Magnolia family, members of which have a The Mostly Male Theory (Frohlich and well-developed perianth of showy tepals, as Parker, 2000; Frohlich, 2001, 2002, 2003) well as bracts that enclose the flower. arose through studies of the LFY gene, in particular from the observations that LFY is Kim et al. (unpublished) examined the single copy in diploid angiosperms, but that expression of A and B class genes in the there are two copies present in all extant calyptra. The B-function genes (AP3 and PI) gymnosperm groups, owing to an ancient isolated from staminodes of Eupomatia duplication predating the divergence of species were strongly expressed in develop- angiosperms from extant gymnosperms. ing stamens, staminodes and carpels, but Angiosperms would have inherited both either not expressed or expressed weakly in copies of LFY, but one of these has been lost. the calyptra, at a level consistent with expres- Expression of the gymnosperm LFYs in pine, sion in leaves. As reviewed, recent studies coupled with the role of LFY in angiosperms, suggest that in basal angiosperms and mono- suggests that one gymnosperm LFY helps to cots an ‘ancestral’ ABC model (compared specify the male reproductive unit while the with Arabidopsis) operates with B-function other helps to specify the female unit. genes expressed throughout the perianth. Angiosperms retain only the male-specifying Following this model (Fig. 10.8), the pattern unit, suggesting that the angiosperm flower of expression of floral genes in the calyptra may derive more from the male reproductive of Eupomatia generally matches the expecta- structure of the gymnosperm ancestor, rather tions for a non-floral organ (such as a leaf or than from the female unit. Other data from bract) rather than predictions for perianth, extant plants and from fossils are consistent consistent with Endress’s (2003) interpreta- with this view, and together suggest that tion based on developmental morphology. bisexuality of the angiosperm reproductive structure may have arisen when the ovule Thus, in some situations such as a clade antecedent became ectopic upon micro- of related taxa, floral gene expression may sporophylls of a reproductive unit resembling be useful in addressing the origins of enig- that of the fossil gymnosperm group, matic structures. There are caveats, how- Corystospermales. The theory is consistent ever. In the case of a basal angiosperm such with the flower antecedent originally consist- as Eupomatia, comparisons are better made ing of stamen- and carpel-like structures, but in light of the proposed ‘ancestral’ ABC without a perianth, and with insect-attractive model, rather than the classic ABC model of features and insect pollination long predating Arabidopsis and other core eudicots. the full elaboration of the flower (Frohlich, 2001), as suggested by the timing of the Gene Evidence and the Origin of Flowers AP3/PI gene duplication described above. An understanding of the nature of the Other recent gene-based hypotheses also flower in basal angiosperms should help in focus on the origin of angiosperm bisexual elucidating the evolutionary origin of the reproductive structures from the unisexual flower itself. Flowers differ so greatly from structures typical of gymnosperms. Albert et the reproductive structures of living and fos- al. (2002) proposed an alternative to Mostly sil gymnosperms that the origin of the Male. Their hypothesis stresses the possible flower has long been a famous question in functional replacement of one copy of LFY evolutionary biology. Numerous hypotheses by the other, resulting in expression of both based on morphological, developmental and male- and female-specific genes in the palaeobotanical studies have been proposed reproductive unit. Theißen et al. (2002) sug- (reviewed in Stebbins, 1974; Crane, 1985; gest that changed expression patterns of B- class genes could have generated bisexual

Evolution of the flower 193 reproductive structures from either male or homeotic evolution of a second corolla (fused female cones of gymnosperms. Both of these petal) whorl in the Hawaiian genus Clermontia hypotheses, but especially that of Theißen et (Campanulaceae) would be a simple issue of al. (2002), suggest relatively equal participa- demonstrating ectopic expression of B-func- tion of male- and female-derived gym- tion genes in the first, normally sepalar, nosperm genes in the organization of the whorl. However, with the generality of the A flower. If the distinction between gym- function now in question, the mechanistic nosperm male and female structures is fully basis for the double-corolla phenotype in determined by differences in B-gene expres- Clermontia might be other than simply out-of- sion, then changes in B-gene expression place B-function gene expression. patterns should bring the full panoply of Furthermore, if altered B-gene expression is male- or female-specific genes into the (for- the cause, various mechanisms could generate merly) unisexual cone of the other gender. such modified expression. The naturally occurring mutation could be within a B-func- The relative contribution of male- and tion coding sequence, or perhaps in its tran- female-specific gymnosperm genes to those scriptional promoter, which might have active in the flower constitutes a direct test elements that fine-tune spatio-temporal of these theories. The gene discovery, gene expression. It could equally well be that the phylogeny and gene expression studies of double-corolla lesion is in a different gene, the the Floral Genome Project (see below) will product of which normally interacts with a B- provide this test. function gene’s promoter to regulate where it expresses. In other words, the gene that nor- Future Prospects mally excludes B-function genes from the first whorl of Clermontia, or any gene upstream of The evolution of flower morphology is being it in its developmental regulatory cascade, elucidated through research on the genetic could be the culprit. Analysing this problem mechanisms of reproductive development in will not be simple, because Clermontia, unlike diverse angiosperms. Arabidopsis has figured weedy Arabidopsis, is a small tree that is much prominently in these studies. Although less tractable to genetic studies that require Arabidopsis was the first plant to have its progeny analysis. Such ‘forward’ genetic stud- nuclear genome completely sequenced (in ies, starting with a phenotypically recogniz- 2000), earlier genetic studies of Arabidopsis able mutation and culminating with the gene (beginning in the late 1980s) paved the way linked to it, may be difficult to accomplish out- for evolutionary interpretations of the mole- side of the model plants. Therefore, investiga- cular processes underlying floral diversity. tors are turning more and more to ‘reverse’ genetic approaches that start with a gene The evolutionary genetics of floral sequence that is suspected to have a particular morphology function (e.g. through a molecular evolution- ary relationship to genes of known function) While the simplicity of the ABC model made and work backwards to establish this function it seem that diversions from ‘normal’ through transgenic experiments that over- sepal/petal/stamen/carpel identity among and/or underexpress the gene’s protein prod- angiosperms might be explored through com- uct. In this way the Gerbera SEP-like genes parative expression studies of A-, B- and C- were characterized. function genes (as has been done analogously with homeobox genes in the segmental evolu- The Floral Genome Project, a large-scale tion of animals), the greater genetic complex- effort to identify genes specific to flower ity now recognized behind flower development and those linked to floral diver- development indicates that this view requires sification, is under way (www. revision. For example, one author of this con- floralgenome.org; overview in Soltis et al., tribution once felt that explaining the 2002). The Floral Genome Project is sequencing genes expressed during the earli- est stages of floral development in diverse

194 D.E. Soltis et al. lineages of angiosperms, particularly basal Acknowledgements angiosperms such as Amborella, waterlilies, tulip tree (Magnoliaceae), avocado This research was supported in part by an (Lauraceae) and Acorus (sister to all other NSF Plant Genome Grant, DBI-0115684. monocots), plus Eschscholzia (poppy, a basal We thank Matyas Buzgo, William eudicot), and will obtain expression data for Farmerie, Lena Landherr, Yi Hu, Marlin genes in a subset of these taxa. Genes in Druckenmeyer, Sheila Plock and John common with or distinct to these species Carlson for technical assistance in building should provide valuable molecular tools for cDNA libraries and generating EST data. the next generation of plant evolutionary developmental research. References Albert, V.A., Gustafsson, M.H.G. and Di Laurenzio, L. (1998) Ontogenetic systematics, molecular develop- mental genetics, and the angiosperm petal. In: Soltis, D.E., Soltis, P.S. and Doyle, J.J. (ed.) Molecular Systematics of Plants II. Kluwer, New York. Albert, V.A., Oppenheimer, D. and Lindqvist, C. (2002) Pleiotropy, redundancy and the evolution of flowers. Trends in Plant Science 7, 297–301. APG II (2003) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants. Botanical Journal of the Linnean Society 141, 399–436. Arber, E.A.N. and Parkin, J. (1907) On the origin of angiosperms. Journal of the Linnean Society, Botany 38, 29–80. Barkman, T.J., Chenery, G., McNeal, J.R., Lyons-Weiler, J. and DePamphilis, C.W. (2000) Independent and combined analysis of sequences from all three genomic compartments converge to the root of flower- ing plant phylogeny. Proceedings of the National Academy of Sciences USA 97, 13166–13171. Baum, D.A., Doebley, J., Irish, V.F. and Kramer, E.M. (2002) Response: Missing links: the genetic architecture of flower and floral diversification. Trends in Plant Science 7, 31–34. Becker, A. and Theißen, G. (2003) The major clades of MADS-box genes and their role in the development and evolution of flowering plants. Molecular Phylogenetics and Evolution 29, 464–489. Bessey, C.E. (1915) The phylogenetic taxonomy of flowering plants. Annals of the Missouri Botanical Garden 2, 109–164. Borsch, T., Hilu, K.W., Quandt, D., Wilde, V., Neinhuis, C. and Barthlott, W. (2003) Non-coding plastid trnT- trnF sequences reveal a well resolved phylogeny of basal angiosperms. Journal of Evolutionary Biology 16, 558–576. Buzgo, M., Soltis, D.E., Soltis, P.S. and Hong, M. (2004a) Toward a comprehensive integration of morpho- logical and genetic studies of floral development. Trends in Plant Science 9, 164–173. Buzgo, M., Soltis, P.S. and Soltis, D.E. (2004b) Floral developmental morphology of Amborella trichopoda (Amborellaceae). International Journal of Plant Sciences (in press). Callis, J. and Vierstra, R.D. (2000) Protein degradation in signaling. Current Opinion in Plant Biology 3, 381–386. Charrier, B., Champion, A., Henry, Y. and Kreis, M. (2003) Expression profiling of the whole Arabidopsis Shaggy-like kinase multigene family by real-time reverse transcriptase-polymerase chain reaction. Plant Physiology 130, 577–590. Coen, E.S. and Meyerowitz, E.M. (1991) The war of the whorls: genetic interactions controlling flower development. Nature 353, 31–37. Coen, E.S., Momero, J.M., Doyle, S., Elliot, R., Murphy, G. and Carpenter, R. (1990) Floricaula: a homeotic gene required for flower development in Antirrhinum majus. Cell 63, 1311–1322. Crane, P.R. (1985) Phylogenetic analysis of seed plants and the origin of angiosperms. Annals of the Missouri Botanical Garden 72, 716–793. Crane, P.R., Friis, E.M. and Pedersen, K.R. (1995) The origin and early diversification of angiosperms. Nature 374, 27–33. Cronquist, A. (1968) The Evolution and Classification of Flowering Plants. Houghton Mifflin, Boston, Massachusetts.

Evolution of the flower 195 Cronquist, A. (1981) An Integrated System of Classification of Flowering Plants. Columbia University Press, New York. De Bodt, S., Raes, J., Van de Peer, Y. and Theißen, G. (2003) And then there were many: MADS goes genomic. Trends in Plant Science 8, 475–483. De Candolle, A.P. (1824–1873) Prodomus systematis naturalis regni vegetabilis. Déterville, Paris. Dilcher, D.L. and Crane, P.R. (1984) Archaeanthus: an early angiosperm from the Cenomanian of the west- ern interior of North America. Annals of the Missouri Botanical Garden 71, 351–383. Donoghue, M.J. and Doyle, J.A. (1989) Phylogenetic analysis of angiosperms and the relationships of Hamamelidae. In: Crane, P.R. and Blackmore, S. (eds) Evolution, Systematics, and Fossil History of Hamamelidae. Clarendon Press, Oxford, pp. 17–45. Dornelas, M.C., Lejeune, B., Dron, M. and Kreis, M. (2000) The Arabidopsis SHAGGY-related protein kinase (ASK) gene family: structure, organization and evolution. Gene 212, 249–257. Doust, A.N. (2000) Comparative floral ontogeny in Winteraceae. Annals of the Missouri Botanical Garden 87, 366–379. Doust, A.N. (2001) The developmental basis of floral variation in Drimys winteri (Winteraceae). International Journal of Plant Sciences 162, 697–717. Doyle, J.A. and Endress, P.K. (2000) Morphological phylogenetic analysis of basal angiosperms: comparison and combination with molecular data. International Journal of Plant Sciences 161, S121–S153. Doyle, J.A., Donoghue, M.J. and Zimmer, E. (1994) Integration of morphological and ribosomal RNA data on the origin of angiosperms. Botanical Review 52, 321–431. Eames, A.J. (1931) The vascular anatomy of the flower with refutation of the theory of carpel polymorphism. American Journal of Botany 18, 147–188. Egea-Cortines, M., Saedler, H. and Sommer, H. (1999) Ternary complex formation between the MADS-box proteins SQUAMOSA, DEFICIENS and GLOBOSA is involved in the control of floral architecture in Antirrhinum majus. EMBO Journal 18, 5370–5379. Endress, P.K. (1980) Ontogeny, function and evolution of extreme floral construction in Monimiaceae. Plant Systematics and Evolution 134, 79–120. Endress, P.K. (1986) Floral structure, systematics and phylogeny in Trochodendrales. Annals of the Missouri Botanical Garden 73, 297–324. Endress, P.K. (1987) The early evolution of the angiosperm flower. Trends in Ecology and Evolution 2, 300–304. Endress, P.K. (1994a) Diversity and Evolutionary Biology of Tropical Flowers. Cambridge University Press, Cambridge. Endress, P.K. (1994b) Floral structure and evolution of primitive angiosperms: recent advances. Plant Systematics and Evolution 192, 79–97. Endress, P.K. (2001) The flowers in extant basal angiosperms and inferences on ancestral flowers. International Journal of Plant Sciences 162, 1111–1140. Endress, P.K. (2003) Early floral development in Eupomatiaceae. International Journal of Plant Sciences 164, 489–503. Endress, P.K. and Igersheim, A. (2000) Gynoecium structure and evolution in basal angiosperms. International Journal of Plant Sciences 161, S211–S223. Endress, P.K. and Sampson, F.B. (1983) Floral structure and relationships of the Trimeniaceae (Laurales). Journal of the Arnold Arboretum 64, 447–473. Endress, P.K., Igersheim, A., Sampson, F.B. and Schatz, G.E. (2000) Floral structure of Takhtajania and its systematic position in Winteraceae. Annals of the Missouri Botanical Garden 87, 347–365. Engler, A. and Prantl, K. (eds) (1887–1915) Die Natürlichen Pflanzenfamilien, 1st edn. W. Engelmann, Leipzig. Engler, A. and Prantl, K. (eds) (1959) Die Natürlichen Pflanzenfamilien, 2nd edn. W. Engelmann, Leipzig. Enright, A.J., Van Dongen, S. and Ouzounis, C.A. (2002) An efficient algorithm for large scale detection of protein families. Nucleic Acids Research 30, 1575–1584. Erbar, C. and Leins, P. (1981) Zur Spirale in Magnolienbluten. Beitrage zur Biologie der Pflanzen 56, 225–241. Erbar, C. and Leins, P. (1983) Zur Sequenz von Blutenorganen bei enigen Magnoliiden. Botanische Jahrbucher für Systematik 103, 433–449. Farras, R., Ferrando, A., Jasik, J., Kleinow, T., Okresz, L., Tiburcio, A., Salchert, K., Del Pozo, C., Schell, J. and Koncz, C. (2001) SKP1–SnRK protein kinase interactions mediate proteasomal binding of a plant SCF ubiquitin ligase. EMBO Journal 20, 2742–2756. Ferrario, S., Immink, R.G., Shchennikova, A., Busscher-Lange, J. and Angenent, G.C. (2003) The MADS box gene FBP2 is required for the SEPALLATA function in Petunia. Plant Cell 15, 914–925.

196 D.E. Soltis et al. Franks, R.G., Wang, C., Levin, J.Z. and Liu, Z. (2002) SEUSS, a member of a novel family of plant regulatory proteins, represses floral homeotic gene expression with LEUNIG. Development 129, 253–263. Friis, E.M., Pedersen, K.R. and Crane, P.R. (1994) Angiosperm floral structures from the Early Cretaceous of Portugal. Plant Systematics and Evolution, Supplement 8, 31–49. Friis, E.M., Pedersen, K.R. and Crane, P.R. (2000) Reproductive structure and organization of basal angiosperms from the early Cretaceous (Barremian or Aptian) of western Portugal. International Journal of Plant Science 161, 5169–5182. Friis, E.M., Pedersen, K.R. and Crane, P.R. (2001) Fossil evidence of waterlilies (Nymphaeales) in the Early Cretaceous. Nature 410, 357–360. Friis, E.M., Doyle, J.A., Endress, P.K. and Leng, Q. (2003) Archaefructus – angiosperm precursor or special- ized early angiosperm? Trends in Plant Science 8, 369–373. Frohlich, M.W. (2001) A detailed scenario and possible tests of the Mostly Male theory of flower evolution- ary origins. In: Zelditch, M.L. (ed.) Beyond Heterochrony: the Evolution of Development. Wiley-Liss, New York, pp. 59–104. Frohlich, M.W. (2002) The Mostly Male theory of flower origins: summary and update regarding the Jurassic pteridosperm Pteroma. In: Cronk, Q.C.B., Bateman, R.M. and Hawkins, J.A. (eds) Developmental Genetics and Plant Evolution, Systematics Association Special Volume Series 65. Taylor & Francis, New York, pp. 85–108. Frohlich, M.W. (2003) An evolutionary scenario for the origin of flowers. Nature Reviews Genetics 4, 559–566. Frohlich, M.W. and Parker, D.S. (2000) The Mostly Male theory of flower evolutionary origins: from genes to fossils. Systematic Botany 25, 155–170. Ganoth, D., Borstein, G., Ko, Y.K., Larsen, B., Tyers, M., Pagano, M. and Hershko, A. (2001) The cell-cycle regulatory protein Cks1 is required for SCF (Skp2)-mediated ubiquitinylation of p27. Nature Cell Biology 3, 321–324. Goff, S.A., Ricke, D., Lan, T.H., Presting, G., Wang, R., Dunn, M., Glazebrook, J., Sessions, A., Oeller, P., Varma, H., Hadley, D., Hutchison, D., Martin, C., Katagiri, F., Lange, B.M., Moughamer, T., Xia, Y., Budworth, P., Zhong, J., Miguel, T., Paszkowski, U., Zhang, S., Colbert, M., Sun, W.L., Chen, L., Cooper, B., Park, S., Wood, T.C., Mao, L., Quail, P., Wing, R., Dean, R., Yu, Y., Zharkikh, A., Shen, R., Sahasrabudhe, S., Thomas, A., Cannings, R., Gutin, A., Pruss, D., Reid, J., Tavtigian, S., Mitchell, J., Eldredge, G., Scholl, T., Miller, R.M., Bhatnagar, S., Adey, N., Rubano, T., Tusneem, N., Robinson, R., Feldhaus, J., Macalma, T., Oliphant, A. and Briggs, S. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science 296, 92–100. Graham, S. and Olmstead, R. (2000) Utility of 17 chloroplast genes for inferring the phylogeny of the basal angiosperms. American Journal of Botany 87, 1712–1730. Hahn, G.W. and Sommerville, C.R. (1988) Genetic control of morphogenesis in Arabidopsis. Developmental Genetics 9, 73–89. Hershko, A. and Ciechanover, A. (1998) The ubiquitin system. Annual Review of Biochemistry 67, 425–479. Hiepko, P. (1965) Vergleichend-morphologische und entwicklungs-geschichtliche Untersuchungen über das Perianth bei den Polycarpicae. Botanische Jarbucher für Systematik 84, 359–508. Hilu, K.W., Borsch, T., Muller, K., Soltis, D.E., Soltis, P.S., Savolainen, V., Chase, M.W., Powell, M., Alice, L., Evans, R., Sauquet, H., Neinhuis, C., Slotta, T., Rohwer, J. and Catrou, L. (2003) Inference of angiosperm phylogeny based on matK sequence information. American Journal of Botany 90, 1758–1776. Honma, T. and Goto, K. (2000) Complexes of MADS-box proteins are sufficient to convert leaves into floral organs. Nature 409, 525–529. Hughes, N.F. (1994) The Enigma of Angiosperm Origins. Cambridge University Press, Cambridge. Irish, V.F. (2003) The evolution of floral homeotic gene function. BioEssays 25, 637–646. Johansen, B., Pedersen, L.B., Skipper, M. and Frederiksen, S. (2002) MADS-box gene evolution–structure and transcription patterns. Molecular Phylogenetics and Evolution 23, 458–480. Kanno, A., Saeki, H., Kameya, T., Saedler, H. and Theißen, G. (2003) Heterotropic expression of class B flo- ral homeotic genes supports a modified ABC model for tulip (Tulipa gesneriana). Plant Molecular Biology 52, 831–841. Keck, E., McSteen, P., Carpenter, R. and Coen, E. (2003) Separation of genetic functions controlling organ identity in flowers. EMBO Journal 22, 1058–1066. Kim, I. and Kimmel, A.R. (2000) GSKs, a master switch regulating cell-fate specification and tumorigenesis. Current Opinion in Genetics and Development 10, 508–514. Kim, S., Yoo, M.-J., Albert, V.A., Farris, J.S., Soltis, P.S. and Soltis, D.E. (2004) Phylogeny and diversification

Evolution of the flower 197 of B-function MADS-box genes in angiosperms: evolutionary and functional implications of a 260-mil- lion-year-old duplication. American Journal of Botany 91 (in press). Kong, H., Leebens-Mack, J., Ni, W., DePamphilis, C.W. and Ma, H. (2004) Highly heterogeneous rates of evolution in the SKP1 gene family in plants and animals: functional and evolutionary implications. Molecular Biology and Evolution 21, 117–128. Kosuge, K. (1994) Petal evolution in Ranunculaceae. Plant Systematics and Evolution 152, 185–191. Kotilainen, M., Elomaa, P., Uimari, A., Albert, V.A., Yu, D. and Terri, T.H. (2000) GRCD1, an AGL2-like MADS box gene, participates in the C function during stamen development in Gerbera hybrida. Plant Cell 12, 1893–1902. Kramer, E.M. and Irish, V.F. (1999) Evolution of genetic mechanisms controlling petal development. Nature 399, 144–148. Kramer, E.M. and Irish, V.F. (2000) Evolution of the petal and stamen developmental programs: evidence from comparative studies of the lower eudicots and basal angiosperms. International Journal of Plant Sciences 161, S29–S40. Kramer, E.M., Dorit, R.L. and Irish, V.F. (1998) Molecular evolution of genes controlling petal and stamen development: duplication and divergence within the APETALA3 and PISTILLATA MADS-box gene lin- eages. Genetics 149, 765–783. Kramer, E.M., Di Stillio, V.S. and Schluter, P.M. (2003) Complex patterns of gene duplication in the Apetala3 and Pistillata lineages of the Ranunculaceae. International Journal of Plant Sciences 164, 1–11. Kubitzki, K. (1987) Origin and significance of trimerous flowers. Taxon 36, 21–28. Lamb, R.S. and Irish, V.F. (2003) Functional divergence within the APETALA3/PISTILLATA floral homeotic gene lineages. Proceedings of the National Academy of Sciences USA 100, 6558–6563. Leins, P. and Erbar, C. (1985) Ein Beitrag zur Blutenentwicklung der Aristolochiaceen, einer Vermittlergruppe zu den Monokotylen. Botanische Jahrbucher für Systematik 107, 343–368. Litt, A. and Irish, V.F. (2003) Duplication and diversification in the APETALA1/FRUITFULL floral homeotic gene lineage: implications for the evolution of floral development. Genetics 165, 821–833. Lu, R., Martin-Hernandez, A.M., Peart, J.R., Malcuit, I. and Baulcombe, D.C. (2003) Virus-induced gene silencing in plants. Methods 30, 296–303. Ma, H. (1994) The unfolding drama of flower development: recent results from genetic and molecular analyses. Genes and Development 8, 745–756. Ma, H. (1998) To be, or not to be, a flower – control of floral meristem identity. Trends in Genetics 14, 26–32. Ma, H. and dePamphilis, C. (2000) The ABCs of floral evolution. Cell 101, 5–8. Maddison, W.P. and Maddison, D.R. (1992) MacClade: analysis of Phylogeny and Character Evolution. Sinauer Associates, Sunderland, Massachusetts. Mathews, S. and Donoghue, M. (1999) The root of angiosperm phylogeny inferred from duplicate phy- tochrome genes. Science 286, 947–950. Meyerowitz, E.M., Bowman, J.L., Brockman, L.L., Drews, G.N., Jack, T., Sieburth, L.E. and Weigel, D. (1991) A genetic and molecular model for flower development in Arabidopsis thaliana. Development Supplement 1, 157–167. Moon, Y.H., Chen, L., Pan, R.L., Chang, H.S., Zhu, T., Maffeo, D.M. and Sung, Z.R. (2003) EMF genes main- tain vegetative development by repressing the flower program in Arabidopsis. Plant Cell 15, 681–693. Münster, T., Pahnke, J., Di Rosa, A., Kim, J.T., Martin, W., Saedler, H. and Theißen, G. (1997) Floral homeotic genes were recruited from homologous MADS-box genes preexisting in the common ances- tor of ferns and seed plants. Proceedings of the National Academy of Sciences USA 94, 2415–2420. Münster, T., Wingen, L.U., Faigl, W., Werth, S., Saedler, H. and Theißen, G. (2001) Characterization of three GLO-BOSA-like MADS-box genes from maize: evidence for ancient parology in one class of floral homeotic B-function genes of grasses. Gene 262, 1–13. Nam, J., dePamphilis, C.W., Ma, H. and Nei, M. (2003) Antiquity and evolution of the MADS-box gene fam- ily controlling flower development in plants. Molecular Biology and Evolution 20, 1435–1447. Nayak, S., Santiago, S.F., Jin, H., Lin, D., Schedl, T. and Kipreos, E.T. (2002) The Caenorhabditis elegans Skp1-related gene family: diverse functions in cell proliferation, morphogenesis, and meiosis. Current Biology 12, 277–287. Ni, W., Li, W., Zhang, X., Hu, W., Zhang, W., Wang, G., Han, T., Zahn, L.M., Zhao, D. and Ma, H. (2004) Genetic control of reproductive development in flowering plants. In: Sharma, R.P. (ed.) Molecular Plant Physiology. Haworth Press, New York (in press). Parkinson, C.L., Adams, K.L. and Palmer, J.D. (1999) Multigene analyses identify the three earliest lineages of extant flowering plants. Current Biology 9, 1485–1488.

198 D.E. Soltis et al. Pelaz, S., Ditta, G.S., Baumann, E., Wisman, E. and Yanofsky, M.F. (2000) B and C floral organ identity func- tions require SEPALLATA MADS-box genes. Nature 405, 200–203. Pelaz, S., Tapia-Lopez, R., Alvarez-Buylla, E.R. and Yanofsky, M.F. (2001) Conversion of leaves into petals in Arabidopsis. Current Biology 11, 182–184. Purugganan, M.D. (1997) The MADS-box floral homeotic gene lineages predate the origin of seed plants: phylogenetic and molecular clock estimates. Journal of Molecular Evolution 45, 392–396. Purugganan, M.D., Sounsley, S.D., Schmidt, R.J. and Yanofsky, M.F. (1995) Molecular evolution of flower development: diversification of the plant MADS-box regulatory gene family. Genetics 140, 345–356. Qiu, Y.-L., Lee, J., Bernasconi-Quadroni, F., Soltis, D.E., Soltis, P.S., Zanis, M., Chen, Z., Savolainen, V. and Chase, M.W. (1999) The earliest angiosperms: evidence from mitochondrial, plastid and nuclear genomes. Nature 402, 404–407. Riechmann, J.L. and Meyerowitz, E.M. (1997) MADS domain proteins in plant development. Biological Chemistry 378, 1079–1101. Riechmann, J.L., Krizek, B.A. and Meyerowitz, E.M. (1996) Dimerization specificity of Arabidopsis MADS domain homeotic proteins Apetala1, Apetala3, Pistillata and Agamous. Proceedings of the National Academy of Sciences USA 93, 4793–4798. Ronse De Craene, L.P., Louis, P., Soltis, P.S. and Soltis, D.E. (2003) Evolution of floral structure in basal angiosperms. International Journal of Plant Sciences 164, S329–S363. Sablowski, R.W. and Meyerowitz, E.M. (1998) A homolog of NO APICAL MERISTEM is an immediate target of the floral homeotic genes APETALA3/PISTILLATA. Cell 92, 93–103. Samach, A., Klenz, J.E., Kohalmi, S.E., Risseeuw, E., Haughn, G.W. and Crosby, W.L. (1999) The UNUSUAL FLORAL ORGANS gene of Arabidopsis thaliana is an F-box protein required for normal patterning and growth in the floral meristem. Plant Journal 20, 433–445. Schmid, M., Uhlenhaut, N.H., Godard, F., Demar, M., Bressan, R., Weigel, D. and Lohmann, J.U. (2003) Dissection of floral induction pathways using global expression analysis. Development 130, 6001–6012. Schmidt, R.J. and Ambrose, B.A. (1998) The blooming of grass flower development. Current Opinion in Plant Biology 1, 60–67. Schwarz-Sommer, Z., Huijser, P., Nacken, W., Saedler, H. and Sommer, H. (1990) Genetic control of flower development by homeotic genes in Antirrhinum majus. Science 250, 931–936. Smith, G.H. (1928) Vascular anatomy of Ranalian flowers. II. Ranunculaceae (continued), Menispermaceae, Calycanthaceae, Annonaceae. Botanical Gazette 85, 152–177. Smyth, D.R., Bowman, J.L. and Meyerowitz, E.M. (1990) Early flower development in Arabidopsis. Plant Cell 2, 755–768. Soltis, D.E., Soltis, P.S., Chase, M.W., Mort, M.E., Albach, D.C., Zanis, M., Savolainen, V., Hahn, W.H., Hoot, S.B., Fay, M.F., Axtell, M., Swensen, S.M., Prince, L.M., Kress, W.J., Nixon, K.C. and Farris, J.S. (2000) Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133, 381–461. Soltis, D.E., Soltis, P.S., Albert, V.A., dePamphilis, C.W., Frohlich, M., Ma, H. and Theißen, G. (2002) Missing links: the genetic architecture of the flower and floral diversification. Trends in Plant Sciences 7, 22–30. Soltis, D.E., Senters, A.E., Zanis, M.J., Kim, S., Thompson, J.D., Soltis, P.S., Ronse De Craene, L.P., Endress, P.K. and Farris, J.S. (2003) Gunnerales are sister to other core eudicots: implications for the evolution of pentamery. American Journal of Botany 90, 463–470. Soltis, D.E., Endress, P.K., Chase, M.W. and Soltis, P.S. (2004) Phylogeny and Evolution of Angiosperms. Smithsonian Institution Press, Washington, DC. Soltis, P.S., Soltis, D.E. and Chase, M.W. (1999) Angiosperm phylogeny inferred from multiple genes as a research tool for comparative biology. Nature 402, 402–404. Stebbins, G.L. (1974) Flowering Plants: Evolution Above the Species Level. Belknap Press, Cambridge, Massachusetts. Sun, G., Li, Q., Dilcher, D.L., Zheng, S., Nixon, K.C. and Wang, X. (2002) Archaefructaceae, a new basal angiosperm family. Science 296, 899–904. Sundström, J., Carlsecker, A., Svensson, M.E., Svenson, M., Johanson, U., Theißen, G. and Engström, P. (1999) MADS-box genes active in developing pollen cones of Norway spruce (Picea abies) are homol- ogous to the B-class floral homeotic genes in angiosperms. Developmental Genetics 25, 253–266. Takhtajan, A. (1969) Flowering Plants: Origin and Dispersal. Smithsonian Institution Press, Washington, DC. Takhtajan, A. (1997) Diversity and Classification of Flowering Plants. Columbia University Press, New York. Teeri, T.H., Albert, V.A., Elomaa, P., Hämäläinen, J., Kotilainen, M., Pöllänen, E. and Uimari, A. (2002) Involvement of non-ABC MADS-box genes in determining stamen and carpel identity in Gerbera

Evolution of the flower 199 hybrida (Asteraceae). In: Cronk, Q.C.B., Bateman, R.M. and Hawkins, J.A. (eds) Developmental Genetics and Plant Evolution. Taylor & Francis, London, pp. 220–232. Theißen, G. (2001) Development of floral organ identity: stories from the MADS house. Current Opinion in Plant Biology 4, 75–86. Theißen, G., Becker, A., Di Rosa, A., Kanno, A., Kim, J.T., Münster, T., Winter, K.-U. and Saedler, H. (2000) A short history of MADS-box genes in plants. Plant Molecular Biology 42, 115–149. Theißen, G., Becker, A., Winter, K.-U., Münster, T., Kirchner, C. and Saedler, H. (2002) How the land plants learned their floral ABCs: the role of MADS box genes in the evolutionary origin of flowers. In: Cronk, Q.C.B., Bateman, R.M. and Hawkins, J.A. (eds) Developmental Genetics and Plant Evolution, Systematics Association Special Volume Series 65. Taylor & Francis, New York, pp. 173–206. Tucker, S.C. (1960) Ontogeny of the floral apex of Michelia fuscata. American Journal of Botany 47, 266–277. Tucker, S.C. and Douglas, A.W. (1996) Floral structure, development, and relationships of Paleoherbs: Saruma, Cabomba, Lactoris, and selected Piperales. In: Taylor, D.W. and Hickey, L.J. (eds) Flowering Plant Origin, Evolution, and Phylogeny. Chapman & Hall, New York, pp. 141–175. Van Tunen, A.J., Eikelboom, W. and Angenet, G.C. (1993) Floral organogenesis in Tulipa. Flower Newsletter 16, 33–38. Von Balthazar, M. and Endress, P.K. (2002) Development of inflorescences and flowers in Buxaceae and the problem of perianth interpretation. International Journal of Plant Sciences 163, 847–876. Weigel, D. (1995) The genetics of flower development: from floral induction to ovule morphogenesis. Annual Review Genetics 29, 19–39. Weigel, D. and Meyerowitz, E.M. (1994) The ABCs of floral homeotic genes. Cell 78, 203–209. Weigel, D., Alvarez, J., Smyth, D.R., Yanofsky, M.F. and Meyerowitz, E.M. (1992) LEAFY controls floral meristem identity in Arabidopsis. Cell 69, 843–859. Winter, K.-U., Becker, A., Münster, T., Kim, J., Saedler, H. and Theißen, G. (1999) MADS-box genes reveal that gnetophytes are much more closely related to conifers than to flowering plants. Proceedings of the National Academy of Sciences USA 96, 7342–7347. Winter, K.-U., Weiser, C., Kaufmann, K., Bohne, A., Kirchner, C., Kanno, A., Saedler, H. and Theißen, G. (2002) Evolution of class B floral homeotic proteins: obligate heterodimerization originated from homodimerization. Molecular Biology and Evolution 19, 587–596. Yamanaka, A., Yada, M., Imaki, H., Koga, M., Ohshima, Y. and Nakayama, K. (2002) Multiple Skp1-related proteins in Caenorhabditis elegans: diverse patterns of interaction with Cullins and F-box proteins. Current Biology 12, 267–275. Yang, M., Hu, Y., Lodhi, M., McCombie, W.R. and Ma, H. (1999) The Arabidopsis SKP1-LIKE1 gene is essen- tial for male meiosis and may control homolog separation. Proceedings of the National Academy of Sciences USA 96, 11416–11421. Yanofsky, M.F. (1995) Floral meristems to floral organs: genes controlling early events in Arabidopsis flower development. Annual Review of Plant Physiology and Plant Molecular Biology 46, 167–188. Yoo, M.-J., Albert, V.A., Soltis, P.S. and Soltis, D.E. (2005) Diversification of glycogen synthase kinase 3/SHAGGY-like kinase genes in plants. BMC Evolutionary Biology (submitted). Yu, D.Y., Kotilainen, M., Pollanen, E., Mehto, M., Elomaa, P., Helariutta, Y., Albert, V.A. and Teeri, T.H. (1999) Organ identity genes and modified patterns of flower development in Gerbera hybrida. Plant Journal 17, 51–62. Yu, J., Hu, S., Wang, J., Wong, G.K., Li, S., Liu, B., Deng, Y., Dai, L., Zhou, Y., Zhang, X., Cao, M., Liu, J., Sun, J., Tang, J., Chen, Y., Huang, X., Lin, W., Ye, C., Tong, W., Cong, L., Geng, J., Han, Y., Li, L., Li, W., Hu, G., Li, J., Liu, Z., Qi, Q., Li, T., Wang, X., Lu, H., Wu, T., Zhu, M., Ni, P., Han, H., Dong, W., Ren, X., Feng, X., Cui, P., Li, X., Wang, H., Xu, X., Zhai, W., Xu, Z., Zhang, J., He, S., Xu, J., Zhang, K., Zheng, X., Dong, J., Zeng, W., Tao, L., Ye, J., Tan, J., Chen, X., He, J., Liu, D., Tian, W., Tian, C., Xia, H., Bao, Q., Li, G., Gao, H., Cao, T., Zhao, W., Li, P., Chen, W., Zhang, Y., Hu, J., Liu, S., Yang, J., Zhang, G., Xiong, Y., Li, Z., Mao, L., Zhou, C., Zhu, Z., Chen, R., Hao, B., Zheng, W., Chen, S., Guo, W., Tao, M., Zhu, L., Yuan, L. and Yang, H. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp. indica). Science 296, 79–92. Zanis, M.J., Soltis, D.E., Soltis, P.S., Mathews, S. and Donoghue, M.J. (2002) The root of the angiosperms revisited. Proceedings of the National Academy of Sciences USA 99, 6848–6853. Zanis, M.J., Soltis, P.S., Qiu, Y.-L., Zimmer, E. and Soltis, D.E. (2003) Phylogenetic analyses and perianth evolution in basal angiosperms. Annals of the Missouri Botanical Garden 90, 129–150. Zhao, D., Yang, M., Solava, J. and Ma, H. (1999) The ASK1 gene regulates development and interacts with the UFO gene to control floral organ identity in Arabidopsis. Developmental Genetics 25, 209–223.

200 D.E. Soltis et al. Zhao, D., Yu, Q., Chen, C. and Ma, H. (2001a) Genetic control of reproductive meristems. In: McManus, M.T. and Veit, B. (eds) Meristematic Tissues in Plant Growth and Development. Sheffield Academic Press, Sheffield, pp. 89–142. Zhao, D., Yu, Q., Chen, C. and Ma, H. (2001b) The ASK1 gene regulates B function gene expression in cooperation with UFO and LEAFY in Arabidopsis. Development 128, 2735–2746. Zhao, D., Ni, W., Feng, B., Han, T., Petrasek, M.G. and Ma, H. (2003) Members of the Arabidopsis-SKP1- like gene family exhibit a variety of expression patterns and may play diverse roles in Arabidopsis. Plant Physiology 133, 203–217. Zheng, N., Schulman, B.A., Song, L., Miller, J.J., Jeffrey, P.D., Wang, P., Chu, C., Koepp, D.M., Elledge, S.J., Pagano, M., Conaway, R.C., Conaway, J.W., Harper, J.W. and Pavletich, N.P. (2002) Structure of the Cul1-Rbx1-Skp1-F boxSkp2 SCF ubiquitin ligase complex. Nature 416, 703–709. Zik, M. and Irish, V.F. (2003) Flower development: initiation, differentiation and diversification. Annual Review of Cell and Developmental Biology 19, 119–140.

11 Diversity in plant cell walls Philip J. Harris School of Biological Sciences, The University of Auckland, Private Bag 92019, Auckland, New Zealand Introduction comparing wall compositions among differ- ent taxa, it is important to appreciate that The cell walls of vascular plants account for because wall composition varies with wall much of the carbon fixed during photosyn- type, there is diversity in wall composition thesis and make up much of their biomass. In even within an individual plant. Two major addition to determining the size, shape and groups of wall types are commonly recog- mechanical strength of plant cells, the walls of nized: primary and secondary (Bacic et al., many species of seed plants (angiosperms and 1988; Carpita and Gibeaut, 1993). Primary gymnosperms) are economically important. walls are deposited while the cells are still They form the bulk of wood for timber or enlarging. By contrast, secondary walls are paper, as well as textile and other fibres; in laid down on the primary wall after cell forages, they are a major source of energy for expansion has stopped, and are usually very ruminants; they make up most of the fibre in much thicker than primary walls. When human diets; and they play a key role in mature, the different cell types can be determining the texture of fruits and vegeta- grouped according to whether they have only bles. Because of their economic importance, primary walls or both primary and secondary much is known about the structure and com- walls. For example, parenchyma cells fre- positions of the walls of many species of quently have only a thin primary wall, angiosperms and some species of coniferous whereas sclerenchyma fibre cells have both a gymnosperms. However, because pterido- primary wall and a very much thicker sec- phytes are of only minor economic impor- ondary wall. In the latter cell type, the sec- tance, information about the composition of ondary wall is laid down over the whole their walls is only fragmentary. surface of the primary wall. However, in xylem tracheary elements, the secondary wall All the walls in vascular plants have a simi- is laid down in a variety of patterns that may lar construction, consisting of two phases: a cover only a portion of the surface area of the fibrillar phase of cellulose microfibrils set in a primary wall. Cell types with secondary walls matrix phase that contains a high proportion can be further classified into those with ligni- of non-cellulosic polysaccharides that vary in fied and those with non-lignified walls, those structure (Figs 11.1–11.3) (Bacic et al., 1988; with lignified walls being the more common Carpita and Gibeaut, 1993). Structural pro- group. In lignified walls, the lignin occurs in teins and glycoproteins, as well as phenolic both the primary and the secondary walls. components including lignin, may also be Cell types with lignified secondary walls present in the wall matrix. However, before © CAB International 2005. Plant Diversity and Evolution: Genotypic and 201 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

202 P.J. Harris Fig. 11.1. Structures of the wall polysaccharides cellulose, callose and xyloglucans. The letters G, X, S, L and F refer to an unambiguous nomenclature for xyloglucan structure developed by Fry et al. (1993) (see text). Xyloglucans in walls are often acetylated, but, for simplicity, the sites of possible acetylation are not shown.

Diversity in plant cell walls 203 Fig. 11.2. Structures of the wall polysaccharides (galacto-) glucomannan, (galacto-) mannan and heteroxylans. For simplicity, sites of possible acetylation are not shown. include sclerenchyma fibres and xylem trac- about 94% cellulose and the walls of pollen heary elements. Cell types with non-lignified tubes contain callose, a (1→3)-␤-D-glucan secondary walls include cotton-seed hairs, (Fig. 11.1) (Bacic et al., 1988; Stone and pollen tubes and the thickened parenchyma Clarke, 1992). The secondary walls of trans- walls of endosperms and cotyledons in some fer cells, which form ingrowths with a diver- seeds. The walls of cotton-seed hairs contain sity of morphologies, are also non-lignified.

204 P.J. Harris Thus, in making comparisons of wall com- comparisons among taxa because the wall positions among different taxa, it is important preparations were from whole organs, or to compare the compositions of the same wall even whole plants, and contained a mixture types. Many of the early studies of wall com- of wall types. Ideally, comparisons should be positions are difficult to interpret and use in made using wall preparations from the same

Diversity in plant cell walls 205 Fig. 11.3. (Above and opposite.) Structures of the pectic polysaccharides that occur in walls of vascular plants. For simplicity, sites of possible acetylation are not shown. The rhamnogalacturonan II (RG-II) structure is that which occurs in grape (wine) (Vitis vinifera); however, the substituents on the arabinopyranose residue of the B side chain vary with taxon, resulting in chains containing 6–9 monosaccharides. cell type. Methods have been described for polyclonal antibodies are now available that isolating walls from several different cell specifically recognize particular wall compo- types, but these are often difficult to carry out nents, particularly polysaccharides (Willats et (Harris, 1983). Therefore, more commonly, al., 2000). These antibodies can be used in wall preparations that contain mostly one conjunction with secondary antibodies wall type, such as primary walls or lignified labelled either with fluorochromes for fluo- secondary walls, are isolated and chemically rescence light microscopy, or with colloidal analysed. Although so far little used, another gold for transmission electron microscopy. way of making comparisons of wall composi- Because it is known that environmental con- tions among different taxa is to use immuno- ditions and pathogens can also affect wall cytochemistry. A range of monoclonal and compositions, plants used for comparative

206 P.J. Harris studies should be pathogen free and ideally groups. However, most of this research has grown in the same environment. Stage of been done on the walls of eudicotyledons, development of an organ can also affect wall rather than the other dicotyledons, often composition: for example, striking changes in referred to as basal angiosperms, compris- wall composition often occur during fruit ing the Amborellaceae, Nymphaeaceae, ripening, and this should be recognized in Austrobaileyales, Chloranthaceae, magnoli- making comparisons among taxa. ids and Ceratophyllales (APG II, 2003) (Fig. 11.5). Second, most of the polymers present Recognition of the considerable diversity in dicotyledon walls are also present in the in wall composition that occurs among differ- other groups, although often with different ent vascular plants is particularly important fine structures and in different proportions, as increasingly attention focuses on the wall and thus most of the wall polymers dis- compositions of just a few model species, cussed throughout the review are intro- such as Arabidopsis thaliana and rice (Oryza duced in this section. sativa), which have sequenced genomes. Studies of wall diversity among vascular Angiosperm Walls plant taxa have been, and will increasingly be, helped by the enormous impact that Dicotyledon walls gene sequences have had on our under- standing of phylogeny. This has been partic- Primary walls of dicotyledons ularly striking for the angiosperms and has resulted in a new classification (APG, 1998; The primary walls of the eudicotyledons are APG II, 2003) that provides an excellent probably the most thoroughly investigated basis for investigating whether variation in a walls in vascular plants and have broadly simi- particular wall component discovered in one lar compositions in all taxa so far examined. taxon is also present in phylogenetically They contain large proportions of pectic poly- related taxa (Figs 11.5 and 11.6). saccharides, smaller proportions of xyloglu- cans, and minor proportions of heteroxylans, In this review of diversity in the walls of glucomannans and/or galactoglucomannans. vascular plants, I have focused particularly They also contain structural proteins and gly- on the compositions of primary walls, ligni- coproteins and may contain phenolic compo- fied secondary walls and non-lignified sec- nents (Bacic et al., 1988; Carpita and Gibeaut, ondary walls of seeds, because these are 1993). Although much less research has been particularly common wall types and most is done on the walls of the other dicotyledons known about their diversity of composition. (basal angiosperms), there is so far no evi- In doing so, I have concentrated on the dence that their wall compositions differ from structures of their component non-cellulosic those of the eudicotyledons. polysaccharides and lignins, because most information is available about these compo- PECTIC POLYSACCHARIDES Pectic polysaccha- nents. Where information is available, struc- tural proteins and glycoproteins will be rides are highly complex polymers compris- discussed. However, walls containing other polymers, including suberin, cutin, cutan ing four domains: homogalacturonan (HG), and sporopollenin, will not be discussed. Walls of the dicotyledons are discussed first, rhamnogalacturonan I (RG-I), rhamno- followed by the monocotyledons, gym- nosperms and pteridophytes. The reason galacturonan II (RG-II) and xylogalacturo- for beginning with the walls of the dicotyle- dons, a phylogenetically advanced group, nan (XGA) (Ridley et al., 2001; Willats et al., and concluding with the pteridophytes, the least advanced group, is twofold. First, more 2001) (Fig. 11.3). HG is composed of linear research has been done on the walls of a greater variety of families in the dicotyle- chains of galacturonic acid residues that may dons than on the walls of any of the other be methyl-esterified and acetylated. The extent of acetylation apparently varies with taxon. The HG of sugarbeet (Beta vulgaris) (Chenopodiaceae) and potato (Solanum tuberosum) (Solanaceae) is highly acetylated

Diversity in plant cell walls 207 (Willats et al., 2001). Furthermore, the HG 2-O-methyl-␣-L-fucose, varies in size in walls of two other Chenopodiaceae species, (Glushka et al., 2003) (Fig. 11.3). The B side Salicornia ramosissima and Chenopodium chain contains nine monosaccharides in gin- quinoa, is also highly acetylated (Renard et seng (Panax ginseng) (Araliaceae) and grape al., 1993, 1999). Variation has also been (wine) (Vitis vinifera) (Vitaceae), seven in A. found in the pattern of acetylation of HG; thaliana and six in sugarbeet and beetroot this variation may be taxonomic, develop- (B. vulgaris). In sycamore (Acer pseudopla- mental or both (Perrone et al., 2002). The tanus) (Aceraceae), this side chain varies in XGA domain has single ␤-D-xylosyl residues length, containing seven, eight or nine attached to C-3 of galacturonic acid residues monosaccharides (Glushka et al., 2003). The of an HG backbone, and is apparently par- differences in the B side chain occur at the ticularly abundant in the walls of seeds and non-reducing end and appear to result fruits (Ridley et al., 2001; Willats et al., 2001). from the presence or absence of sub- stituents linked to C-2 and/or C-3 of the The RG-I domain is composed of alter- arabinopyranose residue (Fig. 11.3). nating galacturonic acid and rhamnose Although these differences probably repre- residues. Many of the rhamnose residues sent real variation among species, at least have polysaccharide or oligosaccharide side some could also result from differences in chains rich in arabinose and/or galactose the isolation procedures. Except for these that include arabinans, galactans and Type I differences in the B side chain, RG-II has a arabinogalactans (Fig. 11.3); small amounts highly conserved structure and occurs as a of Type II arabinogalactans possibly also dimer cross-linked by 1:2 borate-diol esters occur. The structures of the side chains vary involving the apiosyl residue in only side with taxon. For example, RG-I from cab- chain A (Ishii et al., 1999). bage (Brassica oleracea) walls has mostly side chains of arabinans, whereas RG-I from FERULIC ACID ESTER-LINKED TO PECTIC POLYSACCHA- walls of potato (S. tuberosum) tubers has RIDES The hydroxycinnamic acid ferulic mostly galactan side chains (Harris et al., acid (Fig. 11.4) is attached to pectic polysac- 1997). However, whether phylogenetically charides in the primary walls of spinach related taxa have similar side chains is (Spinacia oleracea) and sugarbeet (B. vulgaris) unknown. Moreover, immunolabelling stud- (Chenopodiaceae). It is attached to the ara- ies using monoclonal antibodies that recog- binan and galactan side chains of RG-I, nize arabinans (LM6) and galactans (LM5) where it is esterified via its carboxyl group to have shown that the structures of these side the C(O)2 hydroxyl of arabinofuranosyl chains can also vary during development. residues in the arabinans and to the C(O)6 In suspension-cultured carrot (Daucus hydroxyl of galactopyranosyl residues in the carota) (Apiaceae) cells, arabinans predomi- galactans (Ishii and Tobita, 1993; nate in the walls of meristematic cells, Colquhoun et al., 1994). The primary walls whereas galactans predominate in the walls of spinach also contain small amounts of of elongating cells (Willats et al., 1999). ester-linked p-coumaric acid (Fig. 11.4) that may be linked to RG-I in the same way (Fry, Small proportions of the RG-II domain 1988). In addition to ferulic and p-coumaric are also present in these walls. This low acids, early research showed that the walls of molecular weight, but highly complex spinach also contained 5-5Ј dehydrodiferulic domain, which is attached to HG, contains acid, which was then known simply as ‘difer- 11 different monosaccharide residues, ulic acid’ (Hartley and Harris, 1981; Fry, including some that are not found else- 1983) (Fig. 11.4). More recently, a range of where in plant wall polysaccharides. It has a other dehydrodiferulic acids has been found backbone of at least seven galacturonic acid in the walls of sugarbeet and beetroot (B. vul- residues, which has attached to it four garis) (Waldron et al., 1997). The most abun- structurally different side chains: A, B, C dant dimers are the 8-O-4Ј and 8-5Ј dimers, and D. C and D are disaccharides, A, which with smaller proportions of the 5-5Ј and 8-8Ј contains 2-O-methyl-␣-D-xylose, is an octasaccharide and B, which contains

208 P.J. Harris Fig. 11.4. Structures of some of the phenolic components that occur in walls of vascular plants: (a) p-coumaric acid (R = H) and ferulic acid (R = OCH3); (b–e) examples of dehydrodiferulic acids, (b) 8-5’ dimer (benzofuran form), (c) 8-O-4’ dimer, (d) 5-5’ dimer, (e) 8-8’ dimer (aryltetralin form); (f) examples of substituted cyclobutanes, 4,4’-dihydroxy-␣-truxillic acid (R = R’ = H), 4,4’-dihydroxy-3,3’-dimethoxy-␣-truxillic acid (R = R’ = OCH3), and 4,4’-dihydroxy-3-methoxy-␣-truxillic acid (R = OCH3, R’ = H); (g) structural units in lignin, p- hydroxyphenyl (R = R’ = H), guaiacyl (R = H, R’ = OCH3) and syringyl (R = R’ = OCH3). dimers (Fig. 11.4). Dehydrodiferulic acids in all ten families they examined of the order have also been found in the walls of another Caryophyllales as defined by Cronquist species of Chenopodiaceae, Chenopodium (1981), but in no other families (Fig. 11.5). quinoa, where the 8-O-4Ј dimer also predom- These families were as follows: Aizoaceae, inates (Renard et al., 1999). If the two ferulic Amaranthaceae, Basellaceae, Cactaceae, acid residues of such dimers are attached to Caryophyllaceae, Chenopodiaceae, Didierea- different RG-I molecules, then the dimer ceae, Nyctaginaceae, Phytolaccaceae and forms a cross-link between the RG-I mole- Portulacaceae. Species of the two other cules. families in this order, Achatocarpaceae and Molluginaceae, were not examined. It Ester-linked ferulic acid, however, is not should be noted that the order confined to the primary walls of the Caryophyllales as defined by Cronquist Chenopodiaceae. In a survey of the primary (1981) is the core of the much larger order walls of 251 species in 150 families of Caryophyllales as defined by APG II (2003). dicotyledons, using UV fluorescence Chemical analyses were also carried out on microscopy at two pH values, Hartley and wall preparations, and p-coumaric acid Harris (1981) found ester-linked ferulic acid

Diversity in plant cell walls 209 Angiosperms (except commelinid monocots) Primary walls Lignified secondary walls Asterids P: PPs > XGs > heteroxylans P: 4-O-MGXs Arabinoxyloglucans in & (G-) GMs > GMs Solanaceae & Oleaceae (minor) L: S & G units Acetylated lignins in Hibiscus cannabinus Rosids p-hydroxybenzoic acid ester-linked to lignins in Salicaceae Ferulic acid ester-linked in primary walls of Caryophyllales Eudicots Commelinids Apiogalacturonans in Monocots primary walls of some Alismatales Basal angiosperms Gymnosperms Coniferales Some spp. P: as angiosperms P: GGMs S & G units in lignins >4-O-MGAXs Gnetales L: G units Cycadales Ferulic ester-linked in Ginkgoales primary walls P: unknown P: unknown L: S & G units S & G units in P: unknown lignins of P: unknown P: unknown Stangeria paradoxa L: G units P: as Coniferales L: G units Pteridophytes S & G units in lignins of P: (G-) GMs P: as Coniferales Polypodiales Dennstaedtia bipinnata L: G units > PPs & XGs (Some spp. as Ophioglossales) Equisetales P: as Polypodiales? P: as Coniferales L: G units Ophioglossales P: PPs > P: as Coniferales (G-) GMs L: G units & XGs? Psilotales S & G units in lignins of P: as Polypodiales? P: as Coniferales Lycopodiales Selaginella L: G units 3-O-methylgalactose P: as Polypodiales? P: as Coniferales? in primary walls L: G units, but Selaginella S & G units Fig. 11.5. The phylogeny of vascular plants based on nucleotide sequences of genes. The figure is a combination of trees adapted from Pryer et al. (2001), Soltis et al. (2002) (for the gymnosperms) and APG II (2003) (for the angiosperms). On the right are the non-cellulosic polysaccharides of the primary and lignified secondary walls and the lignin monomers of the lignified secondary walls. Selected wall features of particular taxa are shown arrowed. Abbreviations: 4-O-MGX = 4-O-methylglucuronoxylan; 4-O-MGAX = 4-O-methylglucuronoarabinoxylan; G = guaiacyl; (G-) GM = (galacto-) glucomannans; L = lignins; P = non- cellulosic polysaccharides; PP = pectic polysaccharides; S = syringyl; XG = xyloglucans.

210 P.J. Harris and/or 5-5Ј dehydrodiferulic acid were released. Such analyses have shown that detected in some of these wall preparations fucogalactoxyloglucans contain three major (Hartley and Harris, 1981). However, it is subunits: XXXG, XXFG and XLFG. For not known if the ferulic, p-coumaric and example, the xyloglucan from leaves of A. 5-5Ј dehydrodiferulic acid in the walls of thaliana yielded 45% XXXG, 24% XXFG, these other species are ester-linked to arabi- 16% XLFG, 8% XXLG, 4% XLLG and 3% nans and galactans as in the walls of spinach XLXG (Vanzin et al., 2002). The structure of and sugarbeet. XLFG is shown in Fig. 11.1. However, the proportions of the different xyloglucan sub- More recently, small amounts of ferulic units vary somewhat with the organ from and 8-O-4Ј dehydrodiferulic acids have also which the walls are isolated (Pauly et al., been found ester-linked to the walls of carrots 2001). The xyloglucan from walls of (Parr et al., 1997). These small amounts of sycamore is unusual in that about 3% of the hydroxycinnamic acids, which were not backbone glucose residues have attached to detected in the study of Hartley and Harris them the unusual side chains ␤-D-Xylp- (1981), nevertheless could have an important (1→2)-, ␣-L-Araf-(1→2)- or ␣-L-Araf-(1→3)- influence on the properties of the walls. Small ␤-D-Xylp-(1→2); these glucose residues also amounts of such acids may be of widespread have an ␣-D-Xylp-(1→6)- residue attached to occurrence in dicotyledon primary walls. The them (Vincken et al., 1997). walls of carrots also contain significant amounts of ester-linked p-hydroxybenzoic Nevertheless, not all dicotyledons have acid but, as with the ferulic acid, the polysac- fucogalactoxyloglucans: so far, two families, charide to which it is attached is unknown Solanaceae and Oleaceae, have been identi- (Hartley and Harris, 1981; Parr et al., 1997). fied that have arabinoxyloglucans. These contain no fucose but have S side chains (Fig. XYLOGLUCANS Xyloglucans have a linear 11.1). In contrast to the fucogalactoxyloglu- backbone of (1→4)-linked ␤-D-glucopyra- cans, the arabinoxyloglucans of the nosyl residues substituted at C(O)6 with ␣-D- Solanaceae have an XXGG core structure. xyopyranosyl residues. Other substituents Galactose occurs attached to some of the can also occur on the xylose residues and xylose residues in the xyloglucans of tomato an unambiguous nomenclature has been (Lycopersicon esculentum) and potato (S. tubero- developed by Fry et al. (1993) to describe sum), but not in the xyloglucan of tobacco the structures of xyloglucans. The letters G, (Nicotiana tabacum) (Vincken et al., 1996; York X, S, L and F are used to refer to the fol- et al., 1996; Jia et al., 2003). Another side lowing structures: G = unsubstituted ␤- chain (T = tomato) ␤-Araf-(1→3)-␣-Araf- Glcp; X = ␣-D-Xylp-(1→6)-␤-D-Glcp; S and (1→2)-␣-Xylp occurs in the xyloglucan of L = X with ␣-L-Araf-(1→2)- and ␤-D-Galp- tomato, but not tobacco or potato. The major (1→2)- attached to the non-reducing end, subunits of tobacco xyloglucan are XSGG respectively; and F = L with ␣-L-Fucp- and XXGG and of tomato are XXGG, (1→2)- attached to the non-reducing end XSGG, LSGG and XTGG (Vierhuis et al., (Fig. 11.1). 2001). The structure of XSGG is shown in Fig. 11.1. Unlike the arabinoxyloglucans of The xyloglucans in the primary walls of the Solanaceae, that in the walls of olive (Olea most dicotyledon species examined so far europaea) (Oleaceae) has an XXXG core are substituted with galactose and fucose, structure, with XXXG, XXSG and XLSG as and are referred to as fucogalactoxyloglu- the major subunits (Vierhuis et al., 2001). cans. They have an XXXG core structure: three successive glucose residues are substi- Both the families Solanaceae and tuted with xylose residues, but the fourth Oleaceae are in the asterids (euasterid I) glucose is unsubstituted (Fig. 11.1) (Vincken (Fig. 11.5), the Solanaceae in the order et al., 1997). An excellent way of analysing Solanales and the Oleaceae in the Lamiales their structures is by treating them with a (APG II, 2003). It would be interesting to (1→4)-endo-␤-glucanase and then identify- know the structures of xyloglucans of species ing and quantifying the oligosaccharides in other families in both of these orders and

Diversity in plant cell walls 211 in the other two orders of the euasterid I POLYSACCHARIDES In contrast to the primary clade: Garryales and Gentianales. There is walls, the major non-cellulosic polysaccharides already evidence that the xyloglucan of a are usually 4-O-methylglucuronoxylans, with mint (Mentha arvensis ϫ Mentha spicata) smaller proportions of glucomannans (Fig. (Lamiaceae) (Lamiales) is probably also an 11.2). There are exceptions, however; for arabinoxyloglucan (Maruyama et al., 1996). example in the walls of bast fibres of sunn However, the xyloglucan of burdock, Arctium hemp (Crotalaria juncea) (Fabaceae), gluco- lappa (Asteraceae) (Asterales), in the euas- mannans are the major non-cellulosic polysac- terid II clade is a fucogalactoxyloglucan charides (Bacic et al., 1988). (Vierhuis et al., 2001). PROTEINS AND GLYCOPROTEINS In addition to HETEROXYLANS AND GLUCOMANNANS AND/OR occurring in primary walls, extensins and GALACTOGLUCOMANNANS Small propor- proline-rich proteins occur in lignified sec- tions of heteroxylans occur in these walls, ondary walls. For example, extensins occur and the heteroxylan from the walls of in the walls of sclerenchyma cells in the seed sycamore suspension-cultured cells has been coats of soybean (Glycine max) (Fabaceae) characterized as a glucuronoarabinoxylan (Cassab and Varner, 1988). Proline-rich pro- (GAX) (Darvill et al., 1980). This has a simi- teins may be involved in the process of ligni- lar structure to the GAXs of commelinid fication (Showalter, 1993). Although the monocotyledons (Fig. 11.2), but the side best-characterized glycine-rich protein, GRP chains are all linked at C(O)2 to the xylose 1.8, is present in protoxylem vessels, which residues of the (1→4)-␤-D-xylan backbone. have lignified secondary-wall thickenings, Small proportions of glucomannans and/or immunocytochemistry showed that it is in galactoglucomannans (Fig. 11.2) also occur fact located in the primary walls between the in primary cell walls (Bacic et al., 1988; wall thickenings (Ringli et al., 2001). Schröder et al., 2001). PROTEINS AND GLYCOPROTEINS Three LIGNINS Lignins are unique to the walls of vascular plants; there is no good evi- groups of structural proteins and glycopro- dence for lignins in bryophyte walls (Sarkanen and Hergert, 1971; Chen, 1991). teins are often recognized: extensins, pro- Lignins are conventionally defined as com- plex polymers resulting from the oxidative line-rich proteins and glycine-rich proteins polymerization of three types of hydroxy- cinnamyl alcohol precursors (monolignols) (Cassab and Varner, 1988; Showalter, 1993). that result in p-hydroxyphenyl, guaiacyl and syringyl units in lignins (Fig. 11.4). In addition, arabinogalactan proteins These units are joined by a series of inter- unit linkages, the relative abundance of (AGPs) occur in the apoplast and are which appears to depend largely on the proportions of the different types of unit. secreted in slimes and mucilages; some of The most frequent is the ␤-O-4 (␤-aryl ether) linkage, constituting 50% or more of these AGPs may be ionically bound to walls. all linkages. Dicotyledon lignins consist mostly of guaiacyl and syringyl units, with However, it is now apparent that proline- trace amounts of p-hydroxyphenyl units (Sarkanen and Hergert, 1971; Chen, 1991; rich proteins, extensins and AGPs form a Boerjan et al., 2003). They usually have a syringyl:guaiacyl ratio of 2–2.5:1, which family of proline-/hydroxyproline-rich gly- varies with taxon. Early research indicated low proportions of syringyl units in the coproteins (P/HRGPs) with a continuum of lignins of some species of the Winteraceae, although considerable variation apparently molecules from AGPs, containing 99% car- bohydrate, to proline-rich proteins, which are either not or only minimally glycosy- lated (Gaspar et al., 2001). Lignified secondary walls of dicotyledons As for the primary walls, there are no indi- cations so far that the compositions of the lignified secondary walls of the eudicotyle- dons differ from those of the rest of the dicotyledons (basal angiosperms).

212 P.J. Harris occurs in the syringyl:guaiacyl ratio within the need to broaden the conventional defini- this family (Sarkanen and Hergert, 1971). tion of lignin to include these non-conven- A lower proportion of syringyl units occurs tional monomers. in the lignin of tension wood, formed in hardwoods in regions held under tension Secondary walls of dicotyledon seeds (Chen, 1991). The proportions of syringyl and guaiacyl units also vary with cell type: Thick, non-lignified secondary walls occur lignin in the walls of sclerenchyma fibres is in the cotyledons and endosperms of many enriched in syringyl units, whereas that in dicotyledon seeds. These walls usually con- the walls of xylem vessels is enriched in tain large amounts of one of four polysac- guaiacyl units (Saka and Goring, 1985). charides that function as reserve Moreover, even within the wall of a particu- carbohydrates and are mobilized during lar cell type, the lignin is not homoge- germination (Bacic et al., 1988; Buckeridge neous: p-hydroxyphenyl units are laid et al., 2000). Two of these polysaccharides, down first, then guaiacyl and finally galactomannans and mannans (Fig. 11.2), syringyl (Donaldson, 2001). occur only in seed walls. Galactomannans occur in the endosperms of many species of The lignins of some taxa are acylated. Fabaceae and their structures are of chemo- The lignins of the Salicaceace, which com- taxonomic value: species of the three sub- prises the willows (Salix spp.) and poplars families (Caesalpinioidae, Mimosoideae and (Populus spp.), are p-hydroxybenzoylated Faboideae) can be distinguished by their (Landucci et al., 1992), and the lignin of ratio of mannose to galactose (Buckeridge et kenaf (Hibiscus cannabinus) (Malvaceae) is al., 2000). Galactomannans also occur in the acetylated (Lu and Ralph, 2002) (Fig. 11.5). seeds of some other families, for example Recent evidence indicates that these acylated the Annonaceae and Convolvulaceae. lignins result from the polymerization of Mannans, with little or no galactose, have acylated monolignols; acylation does not been recorded in the seeds of several species occur after the polymerization of the lignin, including Coffea arabica (Rubiaceae) and as was first believed (Lu and Ralph, 2002; Carum carvi (Apiaceae) (Bacic et al., 1988). Boerjan et al., 2003). The other two polysaccharides, (1→4)-␤- galactans and xyloglucans, occur as normal Lignins also contain a variety of minor components of primary walls. (1→4)-␤- components, many of which are produced galactans, similar in structure to those in in much greater quantities by mutants or side chains of RG-I (Fig. 11.3), occur in the transgenics involving lignin pathway genes cotyledons of some species of Lupinus (Boerjan et al., 2003). Such studies have (Fabaceae). However, the xyloglucans, which shown that these components are also syn- occur in the seeds of the Fabaceae (subfamily thesized by the polymerization of non-con- Caesalpinioidae) and many other families ventional monolignols. For example, (Kooiman, 1960), have been particularly transgenic tobacco (Nicotiana tabacum) that extensively studied. had the gene encoding the enzyme cinnamyl alcohol dehydrogenase (CAD) downregu- The xyloglucans of seeds, like most non- lated, incorporated significant amounts of seed xyloglucans, have the XXXG core hydroxycinnamyl aldehyde into its lignin. structure. However, seed xyloglucans, with CAD converts hydroxycinnamyl aldehydes one exception, lack fucose. A comparative to hydroxycinnamyl alcohol, and when this study of seed xyloglucans of the three enzyme is downregulated, the hydroxycin- species Tropaeolum majus (Tropaeolaceae), namyl aldehydes build up and are incorpo- Tamarindus indica and Copaifera langsdorffi rated into lignin (Ralph et al., 2001). (Fabaceae) showed that they all contained the Hydroxycinnamyl aldehydes are minor same major structural units (XXXG, XLXG, components of normal lignins and react XXLG and XLLG), but the proportions var- with the histochemical reagent phlorogluci- ied with the species (Buckeridge et al., 1992). nol-HCl to give a red colour. These studies For the xyloglucan from C. langsdorffi, there illustrate the plasticity of lignification and

Diversity in plant cell walls 213 was also evidence for slight differences in different wall compositions, especially of the proportions between different plant their primary walls, and will be discussed populations. separately. Other variations in the structures of seed Primary walls of monocotyledons xyloglucans have also been reported. The xyloglucan of Hymenaea courbaril (Fabaceae) NON-COMMELINID PRIMARY WALLS Surveys of has a core structure based on XXXG and XXXXG, in about equal proportions the monosaccharide compositions of the pri- (Buckeridge et al., 1997). Unusually, the xyloglucan from the cotyledon walls of jojoba mary walls of a range of non-commelinid (Simmondsia chinensis) (Simmondsiaceae) seeds contains the same major structural units as monocotyledons indicated that the non-cel- dicotyledon fucogalactoxyloglucans: XXXG, XXFG and XLFG (Hantus et al., 1997). lulosic polysaccharide compositions of these However, even more remarkable is the occurrence in this xyloglucan of two further walls are similar to those of the primary major structural units: XXJG and XLJG, where J (jojoba) is ␣-L-Galp-(1→2)-␤-D-Galp- walls of dicotyledons and contain a large (1→2)-␣-D-Xylp-(1→6)-, which is unusual because it contains both L-Gal and D-Gal. An proportion of pectic polysaccharides (Jarvis identical side chain occurs in the xyloglucan of the walls of the mur1 mutant of A. thaliana et al., 1988; Harris et al., 1997; Harris, where XXJG and XLJG structural units occur (Zablackis et al., 1996). In this mutant, 2000). The neutral monosaccharide profiles the ␣-L-Galp-(1→2)- residues replace the ␣-L- Fucp-(1→2)- residues of the wild-type. indicated that RG-I galactans are more Monocotyledon walls abundant than RG-I arabinans (Harris et al., In a survey of the primary walls of 104 species 1997; Harris, 2000). in 52 families of monocotyledons using UV fluorescence microscopy at two pH values, Detailed analyses of the non-cellulosic Harris and Hartley (1980) showed that ester- linked ferulic acid was confined to those polysaccharides, including linkage analyses, of species belonging to a group that was subse- quently also identified using DNA sequences the primary walls of non-commelinid mono- and named the commelinoid group (Chase et al., 1993). However, this name has recently cotyledons have been done on only a few been changed to the commelinid group to avoid confusion with the Commelinoideae, a species. These include onion (Allium cepa) subfamily of the Commelinaceae (APG II, 2003). The commelinid group, which is the (Alliaceae) (Redgwell and Selvendran, 1986) most highly evolved major group in the monocotyledons, comprises the Arecales and asparagus (Asparagus officinalis) (palms), Commelinales, Zingiberales and Poales (Fig. 11.6). The other, non-commelinid (Asparagaceae) (Waldron and Selvendran, monocotyledons comprise the Acorales, Alismatales and lilioids (Asparagales, 1990). In both species, the walls contain large Dioscorales, Liliales and Pandanales) (Chase et al., 2000; APG II, 2003). The commelinid proportions of pectic polysaccharides, com- and non-commelinid groups have quite prising homogalacturonans and rhamnogalac- turonan I. Although RG-II has not been identified in the walls of non-commelinid monocotyledons, its widespread occurrence in other seed plants suggests that it also occurs in this group. The structures of the xyloglucans in onion and garlic (Allium sativa) walls have been investigated and found to be fucogalac- toxyloglucans, identical to those in dicotyle- don walls (Ohsumi and Hayashi, 1994). In addition, the walls of some species of the Alismatales contain pectic polysaccharides rich in the monosaccharide apiose (Harris, 2000; Ridley et al., 2001) (Fig. 11.5). Apiogalacturonan, which is a homogalacturo- nan bearing mono- and di-apiosyl side chains attached to C-2 and C-3 of the back- bone (Fig. 11.3), has been isolated from the walls of Lemna minor (Araceae). Apiose-rich pectic polysaccharides also occur in the walls of vegetative parts of the sea grasses Heterozostera tasmanica, Zostera marina, Zostera pacifica and Phyllospadix sp. (Zosteraceae).

Commelinid monocotyledons p-coumaric acid est to lignin in Poaceae Poales (sensu Dahlgren et a Poales Rest of Poale Zingiberales Commelinale Zingiberales p-coumar to lignin in Commelinales Arecales p-hydroxybenzo linked to lignin sagu Ferulic acid ester-linked in primary walls Fig. 11.6. The phylogeny of the commelinid monocotyledons based on nucleotide seque non-cellulosic polysaccharides of the primary and lignified secondary walls and the ligni particular taxa are shown arrowed. Abbreviations: as for Fig. 11.5 and (1,3),(1,4)-G = (1→

ter-linked Primary walls Lignified secondary 214 P.J. Harris e walls P: GAXs > (1,3),(1,4)-Gs al., 1985) (variable proportions), P: GAXs > (G-) GMs PPs & XGs (minor). (minor) L: G, S & some H units es, P: GAXs (major), PPs P: GAXs & & XGs L: G & S units es [higher proportions than Poales sensu Dahlgren et al. (1985)] ric acid ester-linked P: GAXs ? (some) n Musa textilis L: G & S units oic acid ester- P: PPs > XGs, in Metroxylon GAXs (minor) ences of three genes (re-drawn from Chase et al., 2000). On the right are the in monomers of the lignified secondary walls. Selected wall features of →3),(1→4)-␤-D-glucan; GAX = glucuronoarabinoxylan; H = p-hydroxyphenyl.

Diversity in plant cell walls 215 Small amounts of ester-linked ferulic and but in large proportions in others, for exam- p-coumaric acids, together with a range of ple those of barley (Hordeum vulgare) starchy dehydrodiferulic acids, have recently been endosperm contain 75%. In the walls of reported in the walls of asparagus (Rodriguez- coleoptiles, the proportions vary with stage Arcos et al., 2002). As in the dicotyledons, of development. these small amounts of hydroxycinnamic acids, which were not detected in the study Pectic polysaccharides and xyloglucans of Harris and Hartley (1980), may be of also occur in Poaceae walls, but only in small widespread occurrence in the walls of non- proportions. The pectic polysaccharides commelinid monocotyledons. have similar structures to those in the walls of dicotyledons (Carpita, 1996; Harris, Although virtually nothing is known 2000). For example, RG-Is isolated from the about the structural proteins and glycopro- walls of suspension-cultured cells of maize teins in the walls of this group, there is a (Zea mays) and rice (O. sativa) have similar report of extensins being isolated from the structures to the RG-I from the walls of sus- walls of asparagus (Kieliszewski et al., 1992). pension-cultured sycamore (Thomas et al., 1989). RG-II from the walls of bamboo COMMELINID PRIMARY WALLS: POACEAE Most (Phyllostachys edulis) has an identical structure to RG-IIs from dicotyledons, with the vari- research on the walls of the commelinid able B side chain containing eight monosac- charides (Glushka et al., 2003) (Fig. 11.3). group has focused on the walls of the The xyloglucans are less branched than those of dicotyledons and non-commelinid Poaceae (grasses and cereals) because of monocotyledons; they usually contain no fucose, less xylose and much less galactose their enormous economic importance. (Vincken et al., 1997). However, a fucosyl- ated xyloglucan has been reported in the Indeed, for many years the composition of walls of suspension-cultured tall fescue (Festuca arundincea) cells (McDougall and Fry, Poaceae primary walls, which is quite differ- 1994). It has recently been shown that the xyloglucan in the walls of barley coleoptiles ent from those of dicotyledon and non-com- has a core structure based on XXGG and XXGGG (Gibeaut et al., 2002). melinid walls, was sometimes falsely In addition to ferulic and p-coumaric regarded as being representative of all acids, Poaceae walls contain two series of dimers of hydroxycinnamic acids: dehy- monocotyledon primary walls. In Poaceae drodimers, formed by oxidation, and substi- tuted cyclobutanes, formed photochemically walls, the most abundant non-cellulosic (Fig. 11.4). 5-5Ј Dehydrodiferulic acid (‘diferulic acid’) was the first dehydrodimer polysaccharides are GAXs rather than pectic discovered and, for many years, was thought to be the only dehydrodimer present polysaccharides. GAXs have mostly single ␣- (Harris and Hartley, 1980). However, the full range of dehydrodimers of ferulic acid L-arabinosyl and ␣-D-glucuronosyl (or its 4- that can theoretically be formed has now been discovered (Ralph et al., 1994); the O-methyl derivative) residues linked at most abundant are the 8-5Ј, 8-O-4Ј and 8-8Ј dimers (Fig. 11.4). Additionally, dehy- C(O)3 and C(O)2 respectively to the xylose drodimers of sinapic acid and heterodimers of ferulic and sinapic acids have been found residues of the (1→4)-␤-D-xylan backbone in the walls of several cereal grains, especially wild rice (Zizania aquatica) (Bunzel et al., (Fig. 11.2) (Carpita, 1996; Harris, 2000). 2003), and a 4-O-8Ј, 5Ј-5ЈЈ dehydrotrimer However, more complex side chains also occur (Bacic et al., 1988; Wende and Fry, 1997). Ferulic acid and small amounts of p- coumaric acid (Fig. 11.4) are esterified by their carboxyl groups to the C(O)5 hydroxyl of some of the arabinosyl residues. Poaceae walls also contain variable pro- portions of (1→3),(1→4)-␤-D-glucans, which are linear polysaccharides containing both (1→3)- and (1→4)-glycosidic linkages; 70% of the linkages are usually (1→4) (Bacic et al., 1988; Stone and Clarke, 1992; Smith and Harris, 1999) (Fig. 11.1). These polysac- charides occur in only small proportions in some primary walls, for example those of mesophyll cells of ryegrass (Lolium multiflo- rum and Lolium perenne) leaves contain <3%,

216 P.J. Harris of ferulic acid has been found in the walls of ester-linked. There is also some evidence for maize (Z. mays) bran (Rouau et al., 2003). the xyloglucans being fucogalactoxyloglu- Substituted cyclobutanes formed from fer- cans, as occur in the dicotyledons and non- ulic and/or p-coumaric acid also occur; the commelinid monocotyledons (Carnachan most abundant are 4,4Ј-dihydroxy-␣-truxil- and Harris, 2000a). lic acid, 4,4Ј-dihydroxy-3,3Ј-dimethoxy- ␣-truxillic acid and 4,4Ј-dihydroxy-3- The walls of the rest of the commelinid methoxy-␣-truxillic acid (Fig. 11.4) (Hartley group have compositions intermediate and Morrison, 1991). Both the dehy- between that of the Poales (sensu Dahlgren et drodimers and the substituted cyclobutanes al., 1985) and that of the Arecaceae (Fig. are believed to be esterified to GAXs at both 11.6). As classified by APG II (2003) and carboxyl groups thus cross-linking the GAXs. Chase et al. (2000), these plants comprise the It is also theoretically possible for the dehy- Commelinales, the Zingiberales and the rest drotrimer of ferulic acid to cross-link up to of the Poales. As in the walls of the Poales three GAXs (Rouau et al., 2003). (sensu Dahlgren et al., 1985), GAXs are major components, but more pectic polysac- The same three groups of structural pro- charides and probably more xyloglucans are teins and glycoproteins described in dicotyle- usually present (Harris et al., 1997). A don walls also occur in Poaceae walls; detailed analysis, including linkage analysis however, most is known about the extensins, of the polysaccharides, has been done on the which occur in smaller proportions than in walls of pineapple (Ananas comosus) dicotyledon walls (Carpita, 1996). Two (Bromeliaceae) (Poales sensu APG II, 2003) extensins, one rich in histidine and the other (Smith and Harris, 1995). The ferulic acid in rich in threonine, have been isolated from Z. these walls is esterified to GAXs in exactly mays walls (Kieliszewski and Lamport, 1994). the same way as in the walls of the Poaceae Interestingly, the histidine-rich extensin (Smith and Harris, 2001). The pineapple shares sequence homology and glycosylation walls contained only a small amount of pec- patterns with both extensins and AGPs. tic polysaccharides, but the walls were iso- lated from a ripe fruit and the proportion of COMMELINID PRIMARY WALLS: OTHER FAMILIES pectic polysaccharides may have decreased The composition of the Poaceae primary during ripening. As in the walls of palms, wall is not unique within commelinid mono- there is some evidence that the xyloglucans cotyledons. The walls of species in other in these walls are fucogalactoxyloglucans families within the order Poales (sensu (Smith and Harris, 1995). Dahlgren et al., 1985) have a similar compo- sition, with large proportions of GAXs, vari- In addition to ester-linked ferulic acid, able amounts of (1→3),(1→4)-␤-glucans, the walls of commelinid monocotyledons and small proportions of pectic polysaccha- contain 5-5Ј dehydrodiferulic acid (‘diferulic rides and xyloglucans (Smith and Harris, acid’) (Harris and Hartley, 1980). More 1999; Harris, 2000). The Poales (sensu recently, other dehydrodimers of ferulic acid Dahlgren et al., 1985) is part of the much have been found in the walls of Chinese larger Poales order as defined by Chase et al. water chestnut (Eleocharis dulcis) (Parr et al., (2000) and APG II (2003) (Fig. 11.6). 1996) and chufa (Cyperus esculentus) (Parker et al., 2000), both in the Cyperaceae. In contrast to the walls of the Poales (sensu Dahlgren et al., 1985), the walls of the Lignified secondary walls of monocotyledons Arecaceae (the palms), which is a basal fam- ily in the commelinid group, have a similar NON-COMMELINID LIGNIFIED SECONDARY WALLS composition to the walls of dicotyledons and Only fragmentary information is available non-commelinid monocotyledons, with large about these walls. 4-O-Methylglucuronoxy- proportions of pectic polysaccharides and lans, similar to those in the equivalent walls smaller amounts of xyloglucans (Fig. 11.6). of dicotyledons, have been characterized However, they do contain small proportions from the walls of sisal (Agave sisalana), of GAXs to which ferulic acid is probably Sansevieria trifasciata and Cordyline indivisa

Diversity in plant cell walls 217 (Asparagaceae) (Bacic et al., 1988). The been identified in the walls of Cyperus papyrus lignins of non-commelinid monocotyledons (Cyperaceae), pineapple (A. comosus) are apparently similar to those of dicotyle- (Bromeliaceae) (Bacic et al., 1988) and sago dons, although acylated lignins have not palm (Metroxylon sagu) (Arecaceae) (Ozawa et been reported (Sarkanen and Hergert, al., 1998). 1971; Chen, 1991). The lignins of all other commelinid COMMELINID LIGNIFIED SECONDARY WALLS monocotyledons are probably also com- posed of mostly guaiacyl and syringyl units. The major non-cellulosic polysaccharides in As in Poaceae lignin, p-coumaric acid is the lignified walls of Poaceae are GAXs with ester-linked to the lignin of abaca (Musa tex- similar structures to those in their primary tilis) (Musaceae) (Sun et al., 1999), but in walls, but with a lower degree of substitution sago palm (M. sagu), p-hydroxybenzoic acid by glycosyl residues (Smith and Harris, is ester-linked to lignin (Kuroda et al., 2001) 1999). Ferulic acid and its dimers, which are (Fig. 11.6). ester-linked to GAXs, appear to act as nucle- ation sites for lignin polymerization and Secondary walls of monocotyledon seeds become incorporated into lignin (Boerjan et al., 2003). In addition to GAXs, small pro- As in dicotyledon seeds, the seeds of many portions of xyloglucans, (1→3),(1→4)-␤- monocotyledons have thick, non-lignified glucans, and glucomannans and/or galacto- walls that contain mostly one structural type glucomannans also occur in these walls of polysaccharide that is mobilized as a (Smith and Harris, 1999; Trethewey and reserve carbohydrate during germination Harris, 2002). (Bacic et al., 1988; Buckeridge et al., 2000). In the non-commelinid monocotyledons, the Poaceae lignins are similar to dicotyledon endosperms of many species of Asparagales lignins in containing guaiacyl and syringyl and Liliales contain glucomannans. Among units, but the proportions of p-hydroxy- the commelinid monocotyledons, the phenyl units are slightly higher (Sarkanen endosperms of various species of the and Hergert, 1971; Chen, 1991; Boerjan et Arecaceae (palms) contain mannans or al., 2003). However, in contrast to dicotyle- galactomannans, and the aleurone walls of don lignins, Poaceae lignins are acylated the Poaceae contain arabinoxylans and with p-coumaric acid (Lu and Ralph, 1999). (1→3),(1→4)-␤-glucans. As in the GAXs of As with acylated lignins in dicotyledons, primary walls of Poaceae, the arabinoxylans recent evidence indicates that these are syn- of the aleurone walls of wheat (Triticum aes- thesized by the polymerization of acylated tivum) have ferulic acid and small amounts monolignols (Boerjan et al., 2003). Both of p-coumaric acid esterified by their car- ester-linked p-coumaric acid and the p- boxyl groups to the C(O)5 hydroxyl of some hydroxyphenyl units of lignins are oxidized of the arabinosyl residues; however, only to p-hydroxybenzaldehyde with alkaline minor amounts of dehydrodimers of these nitrobenzene. However, for many years, it hydroxycinnamic acids and no substituted was often assumed that the p-hydroxyben- cyclobutanes occur in these walls (Rhodes et zaldehyde originated only from p-hydroxy- al., 2002). In addition to proline-rich and phenyl units of the lignins and hence the glycine-rich proteins, wheat aleurone walls lignins had a high content of these units. contain proteins with up to 23% serine (Rhodes and Stone, 2002). There is some evidence that the non-cel- lulosic polysaccharide compositions of the Gymnosperm Walls lignified secondary walls of all other com- melinid monocotyledons are similar to those Compared with the 249,000 species of in the Poaceae. The walls of species of other angiosperms, there are only 840 extant species families in the Poales (sensu Dahlgren et al., of gymnosperms, which form a monophyletic 1985) that have been examined have similar polysaccharide compositions to the Poaceae (Smith and Harris, 1999). GAXs have also

218 P.J. Harris group divided into four orders: the amounts of glucomannans and/or galac- Coniferales (conifers); the Ginkgoales, which toglucomannans have also been reported in contains only one species, Ginkgo biloba; the the walls of suspension-cultured cells Cycadales (cycads); and the Gnetales, which (Edashige and Ishii, 1996) and cambium of contains Ephedra (Ephedraceae), Gnetum C. japonica (Edashige et al., 1995). However, (Gnetaceae) and Welwitschia (Welwitschiaceae) it is possible that at least some of these glu- (Mabberley, 1997; Judd et al., 2002; Soltis et comannans and/or galactoglucomannans al., 2002) (Fig. 11.5). Most research has been could result from contaminating lignified done on the walls of the conifers because of secondary walls in the wall preparations. the economic importance of many of these as None of these polysaccharides was found in sources of timber and wood pulp. the walls of suspension-cultured P. menziesii cells (Thomas et al., 1987). The non-cellu- Primary walls of gymnosperms losic polysaccharide compositions of the pri- mary walls of the other three classes of Polysaccharides gymnosperms are unknown. The walls of coniferous gymnosperms have Ester-linked ferulic acid similar non-cellulosic polysaccharide compo- sitions to those of dicotyledons; the most A survey using UV fluorescence microscopy abundant non-cellulosic polysaccharides are at two pH values showed that the primary pectic polysaccharides, followed by xyloglu- walls of species from all 17 extant families of cans (Fig. 11.5). The pectic polysaccharides gymnosperms contain ester-linked ferulic are similar to those of dicotyledons and acid (Carnachan and Harris, 2000b) (Fig. comprise homogalacturonan, RG-I and RG- 11.5). Analyses of primary wall preparations II domains. In addition, an XGA (Fig. 11.3) showed that ester-linked p-coumaric acid has been characterized from the walls of was also present. In addition, three dehy- Pinus mugo (Pinaceae) pollen (Bouveng, drodimers of ferulic acid, two 8-8Ј dimers 1965). RG-Is from the walls of suspension- and a 8-5Ј dimer, have been reported from cultured Pseudotsuga menziesii (Pinaceae) cells hypocotyl walls of Pinus pinaster (Pinaceae) (Thomas et al., 1987) and Cryptomeria japon- (Sánchez et al., 1996). However, the wall ica (Taxodiaceae) cambium (Edashige and polysaccharides to which these hydroxycin- Ishii, 1997) differ only in structural details namic acid monomers and dimers are ester- from RG-Is from dicotyledon walls. RG-II in linked are unknown. the walls of Pinus densiflora (Pinaceae) hypocotyls occurs as a dimer cross-linked by Proteins and glycoproteins 1:2 borate-diol esters and has an identical structure to RG-IIs from dicotyledons, with Structural proteins and glycoproteins, similar the variable B side chain containing six to those in dicotyledon walls, have been iden- monosaccharide residues (Shimokawa et al., tified in gymnosperm walls. Hydroxyproline 1999; Glushka et al., 2003) (Fig. 11.3). The arabinosides have been isolated from the xyloglucans also have identical structures to walls of suspension cultures of the coniferous those from dicotyledon walls. They are fuco- gymnosperm Cupressus sp. (Cupressaceae) galactoxyloglucans and have been best char- and from G. biloba (Ginkgoaceae) and Ephedra acterized in the walls of C. japonica cambium sp. (Ephedraceae), indicating the presence of (Kakegawa et al., 1998) (Fig. 11.1). Small extensins (Lamport and Miller, 1971). amounts of a heteroxylan, probably similar However, unlike dicotyledon extensins, in structure to heteroxylans in dicotyledon more hydroxyproline tri-arabinoside was primary walls, have been reported from the isolated than the tetra-arabinoside. More walls of suspension-cultured P. menziesii cells recently, an extensin and a proline-rich pro- (Thomas et al., 1987) and the cambium of C. tein have been characterized from the walls japonica (Edashige et al., 1995). Significant of suspension-cultured cells of P. menziesii (Kieliszewski et al., 1992; Kieliszewski and

Diversity in plant cell walls 219 Lamport, 1994). In addition, expressed units have been reported in the lignins of all sequence tags encoding proline-rich and three families of the Gnetales: Gnetaceae, glycine-rich proteins have been identified in Ephedraceae and Welwitschiaceae (Sarkanen a zone including the cambium and develop- and Hergert, 1971; Logan and Thomas, ing xylem of Pinus taeda (Pinaceae) (Allona et 1985; Chen, 1991) (Fig. 11.5). al., 1998). However, it is unclear whether the proteins occur in the primary walls of the Proteins and glycoproteins cambium or in the lignified secondary walls of the tracheids. An extensin-like protein has been character- ized from the tracheid walls of P. taeda Lignified secondary walls of gymnosperms (Pinaceae) (Bao et al., 1992). Polysaccharides Pteridophyte Walls The major non-cellulosic polysaccharides in The traditionally recognized group, the the lignified secondary walls of most species pteridophytes, which contains 9800 extant of coniferous gymnosperms examined are species (Mabberley, 1997), is now recognized galactoglucomannans (O-acetyl-galactogluco- as paraphyletic. A phylogenetic analysis mannans), together with smaller amounts of using a combination of morphological data 4-O-methylglucuronoarabinoxylans (Whistler and DNA sequences from four genes identi- and Chen, 1991) (Fig. 11.2). However, there fied two monophyletic lineages: the are exceptions: in the walls of Calocedrus decur- Lycopodiales (lycophytes or lycopsids), com- rens (Cupressaceae) (Whistler and Chen, 1991) prising the club mosses and their allies that and Podocarpus lambertii (Podocarpaceae) form the most basal group of vascular (Bochicchio and Reicher, 2003), the 4-O- plants; and a lineage comprising the Psilotales methylglucuronoarabinoxylans predominate (whisk ferns), the Equisetales (horsetails or over the galactoglucomannans. The polysac- sphenopsids), the Polypodiales (the leptospo- charide composition of the lignified secondary rangiate ferns) and the Ophioglossales (the walls of G. biloba (Ginkgoaceae) is similar to eusporangiate ferns, comprising the that of the walls of most coniferous gym- Marattiaceae and Ophioglossaceae) (Pryer et nosperms (Timell, 1960). al., 2001; Judd et al., 2002) (Fig. 11.5). Compared with the walls of other vascular Lignins plants little is known about the compositions of pteridophyte walls. In the walls of G. biloba and most coniferous gymnosperms and cycads, the lignins con- Primary walls of pteridophytes tain almost all guaiacyl units (Sarkanen and Hergert, 1971; Logan and Thomas, 1985; Polysaccharides Chen, 1991) (Fig. 11.5). In addition to these units, p-hydroxyphenyl units occur in com- In contrast to the primary walls of seed pression wood formed in regions of conifer- plants, acid hydrolysates of the primary ous gymnosperms under compression walls of many pteridophyte taxa contain (Chen, 1991). Significant proportions of high concentrations of mannose, indicating syringyl units have also been reported in the the presence of large proportions of lignins of some coniferous gymnosperms, mannose-containing polysaccharides, proba- including Tetraclinis articulata (Cupressaceae) bly glucomannans and/or galactoglucoman- and certain species of Podocarpus nans (Bailey and Pain, 1971; Popper and (Podocarpaceae), and in the lignin of the Fry, 2003). Pteridophytes with large propor- cycad Stangeria paradoxa (Stangeriaceae) tions of such polysaccharides in their walls (Towers and Gibbs, 1953) (Fig. 11.5). include the lycophytes, the whisk fern Additionally, both syringyl and guaiacyl

220 P.J. Harris Psilotum nudum (Psilotaceae) and the horse- detailed structures of the RG-IIs of S. kraus- tail Equisetum debile (Equisetaceae) (Popper siana and E. hyemale, including those of the B and Fry, 2003). Bailey and Pain (1971) side chains, are apparently identical to that of examined the walls of a wide range of the RG-II of the eudicotyledon A. thaliana species of ferns (both leptosporangiate and (Matsunaga et al., 2004). In the RG-IIs of L. eusporangiate) and divided the species into tristachyum, P. nudum and the leptosporangiate three groups depending on whether their ferns, Ceratopteris thalictroides (Pteridaceae) and walls contained low (0–3%), medium or high Platycerium bifurcatum (Polypodiaceae), the 3-O- (>10%) concentrations of mannose. The methyl-rhamnose is linked to C-3, and some- eusporangiate ferns contained only low con- times also to C-2, of the arabinopyranose centrations of mannose in their walls, but the residue at the non-reducing end of the B side leptosporangiate ferns included species in all chain (Fig. 11.3); otherwise, the structures of three groups, although species with low man- these RG-IIs are identical to that of A. thaliana. nose concentrations in their walls appear to be Popper and Fry (2003) also found 3-O-methyl- confined to more derived species. rhamnose in acid hydrolysates of primary walls Interestingly, however, there is evidence that of lycophyte species in the family walls with different concentrations of these Lycopodiaceae, but not in the family mannose-containing polysaccharides can Selaginellaceae. However, they did not find occur even within a single species (White et al., this monosaccharide in acid hydrolysates of 1986). Acid hydrolysates of walls from suspen- primary walls of other species of vascular sion cultures of the leptosporangiate fern plants, including other pteridophytes. As the Pteridium aquilinum (Dennstaedtiacae) con- proportions of RG-II in the primary walls of tained different concentrations of mannose lycophytes and other pteridophytes are similar depending on whether the cultures were (Matsunaga et al., 2004), it is possible that in generated from the sporophyte or the lycophyte walls other polysaccharides, in addi- gametophyte. The walls of sporophyte cul- tion to RG-II, contain 3-O-methyl-rhamnose. tures contained only 1% mannose, but the walls of the gametophyte culture contained Interestingly, acid hydrolysates of lyco- 11%. Overall, however, the walls of most phyte primary walls, but the walls of no pteridophytes probably contain substantial other vascular plants, contain appreciable proportions of glucomannans and/or galac- quantities of another unusual monosaccha- toglucomannans. ride, 3-O-methyl-D-galactose (Popper et al., 2001; Popper and Fry, 2003) (Fig. 11.5). The There is also evidence that the walls of all wall polysaccharide or polysaccharides con- pteridophytes contain pectic homogalacturo- taining this monosaccharide are unknown, nans (Popper and Fry, 2003). RG-I has so far although there are reports in angiosperms not been reported in pteridophyte walls, but of small quantities of 3-O-methyl-galactose Matsunaga et al. (2004) have recently shown in the side chains of RG-I. that RG-II is present in the primary walls of all extant groups of pteridophytes (lycophytes, Xyloglucans have also been found in the whisk ferns, horsetails, eusporangiate and lep- walls of a range of pteridophytes, including tosporangiate ferns), where, as in seed plants, lycophytes, leptosporangiate ferns and a it occurs as a dimer cross-linked by 1:2 borate- eusporangiate fern, but the structures of diol esters. RG-IIs from the lycophyte these are unknown (Popper and Fry, 2003). Selaginella kraussiana (Selaginellaceae) and the horsetail Equisetum hyemale have identical Proteins and glycoproteins monosaccharide compositions to RG-IIs from seed plants, but RG-IIs from another lyco- Little is also known about the structural pro- phyte, Lycopodium tristachyum (Lycopodiaceae), teins and glycoproteins in pteridophyte walls. the whisk fern Psilotum nudum and four species However, hydroxyproline arabinosides have of leptosporangiate ferns also contain the been isolated from the walls of Equisetum sp. unusual monosaccharide 3-O-methyl-rham- and the leptosporangiate fern Onoclea sensi- nose (trivial name acofriose). Remarkably, the bilis (Dryopteridaceae), indicating the pres- ence of extensins (Lamport and Miller, 1971).

Diversity in plant cell walls 221 Lignified secondary walls of pteridophytes and Thomas, 1985; Chen, 1991) (Fig. 11.5). Taxa with lignins of this type include the Polysaccharides lycophyte genus Isoetes (Isoetaceae), the whisk fern Psilotum, the horsetail Equisetum There is no published information on the and both eusporangiate and leptosporan- non-cellulosic polysaccharides of wall giate ferns. However, exceptions have been preparations solely obtained from cells with reported in which the lignins also contain a lignified secondary walls. However, galacto- significant proportion of syringyl units. Such glucomannans and 4-O-methylglu- lignins occur in the lycophyte Selaginella curonoarabinoxylans have been isolated (although not in Lycopodium) and in the lep- and characterized from wall preparations tosporangiate fern Dennstaedtia bipinnata obtained from whole rachides of the lep- (Dennstaedtiaceae) (White and Towers, tosporangiate ferns Osmunda cinnamomea 1967; Logan and Thomas, 1985). (Osmundaceae) (Timell, 1962a,b) and Pteridium aquilinum (Bremner and Wilkie, Wall Evolution 1966, 1971). These rachides contain a high proportion of lignified secondary walls and It is possible to trace the evolution of the so it is likely that most of these polysaccha- rides were from these walls. In both species, compositions of primary and lignified sec- the 4-O-methylglucuronoarabinoxylans have similar structures to those in the lignified ondary walls (Figs 11.5 and 11.6). From secondary walls of coniferous gym- nosperms, but with a lower proportion of what little is known about the compositions arabinose residues. Galactoglucomannans have also been isolated from wall prepara- of the walls of the lycophytes, the most basal tions from whole organs of Lycopodium clava- tum, P. nudum and Equisetum arvense (Timell, extant group of vascular plants, it is evident 1964). Such wall preparations would have contained both primary and lignified sec- that most of the non-cellulosic polysaccha- ondary walls. Nevertheless, because of their thickness, it is likely that lignified secondary rides found in angiosperm walls, as well as walls would have made a substantial gravi- metric contribution to the isolated walls, lignins, had already evolved at the base of and at least some of the galactoglucoman- nans were probably from these walls. The the vascular plant lineage. These include the limited evidence available indicates that most species in both lineages of the pterido- highly complex, pectic polysaccharide RG-II phytes have lignified secondary walls in which the glucomannans and/or galactoglu- (Matsunaga et al., 2004). However, the rela- comannans occur either in greater amounts than the 4-O-methylglucuronoarabinoxy- tive proportions of the different polysaccha- lans or in about equal amounts. However, it is possible that 4-O-methylglucuronoarabi- rides and the structures of the lignins noxylans predominate in the walls of some leptosporangiate ferns (Timell, 1962c). changed during vascular plant evolution. In Lignins both primary and lignified secondary walls, Although few studies have been done, it there was apparently a marked decrease in appears that most pteridophytes have lignins similar to those of coniferous gym- the proportions of glucomannans and/or nosperms, containing almost all guaiacyl units (Sarkanen and Hergert, 1971; Logan galactoglucomannans, but this decrease occurred more rapidly in the primary walls than in the lignified secondary walls. Thus, in the primary walls of the lycophytes, these polysaccharides are predominant, but in the walls of coniferous gymnosperms, they are only a minor component and pectic polysac- charides are the predominant non-cellulosic polysaccharides, as they are in most angiosperms (dicotyledons and non-com- melinid monocotyledons). Even within the pteridophytes, the eusporangiate and some leptosporangiate ferns have primary walls with only a small proportion of glucoman- nans and/or galactoglucomannans. However, these polysaccharides persisted in lignified secondary walls until the evolution

222 P.J. Harris of angiosperms, when heteroxylans became Despite the major differences in propor- predominant in these walls. tions of different non-cellulosic polysaccha- rides in primary and lignified secondary Within the angiosperms, the most strik- walls and in lignin structure among differ- ing evolutionary changes in wall composi- ent vascular plants, all these walls presum- tion occurred in the primary walls of the ably perform all the functions necessary for commelinid monocotyledons. At the base of plant growth and development. This sug- this group, the Arecaceae (palms) have walls gests that different non-cellulosic polysac- similar to those of dicotyledons and non- charides and lignins with different commelinid monocotyledons, except for the structures play similar and essential roles presence of ester-linked ferulic acid, which is that have been conserved during evolution probably linked to GAXs that occur in these (Harris, 2000). walls in small amounts. During the evolution of the commelinid group, the proportions of Conclusions these GAXs, with ferulic acid, esterified to them, apparently increased, reaching their Although much is known about the compo- highest proportions in the walls of the sitions of the walls of many vascular plants, Poales (sensu Dahlgren et al., 1985), where there are many gaps in our knowledge. In the (1→3),(1→4)-␤-glucans probably first particular, information about the walls of appeared. At what stage in commelinid evo- pteridophytes and non-coniferous gym- lution the simple xyloglucans found in the nosperms is only fragmentary. Phylogenetic walls of the Poaceae evolved from fucogalac- trees, based on nucleotide sequences of toxyloglucans is unknown. genes, which are now available for all groups of vascular plants, provide powerful The evolution of lignin structure can also guides for selecting taxa for wall analyses. be traced from lignins with almost all guai- acyl units in the pteridophytes and conifer- Acknowledgements ous gymnosperms, cycads and G. biloba to lignins with both guaiacyl and syringyl units I thank Professor B.A. Stone and J.A.K. that occur in the angiosperms. However, Trethewey for critically reading the manu- reports of syringyl units in the lignins of script, and Professor A.G. Darvill for provid- some lycophytes, a leptosporangiate fern, ing a pre-print of a paper. the Gnetales, a cycad and a few species of coniferous gymnosperms suggest that syringyl units in lignin probably evolved a number of times. References Allona, I., Quinn, M., Shoop, E., Swope, K., Cyr, S.S., Carlis, J., Reidl, J., Retzel, E., Campbell, M.M., Sederoff, R. and Whetten, R.W. (1998) Analysis of xylem formation in pine by cDNA sequencing. Proceedings of the National Academy of Sciences USA 95, 9693–9698. APG (1998) An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85, 531–553. APG II (2003) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141, 399–436. Bacic, A., Harris, P.J. and Stone, B.A. (1988) Structure and function of plant cell walls. In: Preiss, J. (ed.) The Biochemistry of Plants. Academic Press, San Diego, California, pp. 297–371. Bailey, R.W. and Pain, V. (1971) Polysaccharide mannose in New Zealand ferns. Phytochemistry 10, 1065–1073. Bao, W.L., Omalley, D.M. and Sederoff, R.R. (1992) Wood contains a cell-wall structural protein. Proceedings of the National Academy of Sciences USA 89, 6604–6608. Bochicchio, R. and Reicher, F. (2003) Are hemicelluloses from Podocarpus lambertii typical of gym- nosperms? Carbohydrate Polymers 53, 127–136.

Diversity in plant cell walls 223 Boerjan, W., Ralph, J. and Baucher, M. (2003) Lignin biosynthesis. Annual Review of Plant Biology 54, 519–546. Bouveng, H.O. (1965) Polysaccharides in pollen. II. Xylogalacturonan from mountain pine pollen. Acta Chemica Scandinavica 19, 953–963. Bremner, I. and Wilkie, K.C.B. (1966) The hemicelluloses of bracken. Part I. An acidic xylan. Carbohydrate Research 2, 24–34. Bremner, I. and Wilkie, K.C.B. (1971) The hemicelluloses of bracken. Part II. A galactoglucomannan. Carbohydrate Research 20, 193–203. Buckeridge, M.S., Rocha, D.C., Reid, J.S. and Dietrich, S.M.C. (1992) Xyloglucan structure and post-germi- native metabolism in seeds of Copaifera langsdorfii from savanna and forest populations. Physiologia Plantarum 86, 145–151. Buckeridge, M.S., Crombie, H.J., Mendes, C.J.M., Reid, J.S.G., Gidley, M.J. and Vieira, C.C.J. (1997) A new family of oligosaccharides from the xyloglucan of Hymenaea courbaril L. (Leguminosae) cotyledons. Carbohydrate Research 303, 233–237. Buckeridge, M.S., dos Santos, H.P. and Tine, M.A.S. (2000) Mobilisation of storage cell wall polysaccharides in seeds. Plant Physiology and Biochemistry 38, 141–156. Bunzel, M., Ralph, J., Kim, H., Lu, F.C., Ralph, S.A., Marita, J.M., Hatfield, R.D. and Steinhart, H. (2003) Sinapate dehydrodimers and sinapate–ferulate heterodimers in cereal dietary fiber. Journal of Agricultural and Food Chemistry 51, 1427–1434. Carnachan, S.M. and Harris, P.J. (2000a) Polysaccharide compositions of primary cell walls of the palms Phoenix canariensis and Rhopalostylis sapida. Plant Physiology and Biochemistry 38, 699–708. Carnachan, S.M. and Harris, P.J. (2000b) Ferulic acid is bound to the primary cell walls of all gymnosperm families. Biochemical Systematics and Ecology 28, 865–879. Carpita, N.C. (1996) Structure and biogenesis of the cell walls of grasses. Annual Review of Plant Physiology and Plant Molecular Biology 47, 445–476. Carpita, N.C. and Gibeaut, D.M. (1993) Structural models of primary cell walls in flowering plants: consistency of molecular structure with the physical properties of the walls during growth. The Plant Journal 3, 1–30. Cassab, G.I. and Varner, J.E. (1988) Cell wall proteins. Annual Review of Plant Physiology and Plant Molecular Biology 39, 321–353. Chase, M.W., Soltis, D.E., Olmstead, R.G., Morgan, D., Les, D.H., Mishler, B.D., Duvall, M.R., Price, R.A., Hills, H.G., Qiu, Y.-L., Kron, K.A., Rettig, J.H., Conti, E., Palmer, J.D., Manhart, J.R., Sytsma, K.J., Michaels, H.J., Kress, W.J., Karol, K.G., Clark, W.D., Hedren, M., Gaut, B.S., Jansen, R.K., Kim, K.-J., Wimpee, C.F., Smith, J.F., Furnier, G.R., Strauss, S.H., Xiang, Q.-Y., Plunkett, G.M., Soltis, P.S., Swensen, S.M., Williams, S.E., Gadek, P.A., Quinn, C.J., Eguiarte, L.E., Golenberg, E., Learn, G.H. Jr, Graham, S.W., Barrett, S.C.H., Dayanandan, S. and Albert, V.A. (1993) Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botantical Garden 80, 528–580. Chase, M.W., Soltis, D.E., Soltis, P.S., Rudall, P.J., Fay, M.F., Hahn, W.H., Sullivan, S., Joseph, J., Molvray, M., Kores, P.J., Givnish, T.J., Sytsma, K.J. and Pires, J.C. (2000) Higher-level systematics of the monocotyle- dons: an assessment of current knowledge and a new classification. In: Wilson, K.L. and Morrison, D.A. (eds) Monocots: Systematics and Evolution. CSIRO, Melbourne, pp. 3–16. Chen, C.-L. (1991) Lignins: occurrence in woody tissues, isolation, reactions, and structure. In: Lewin, M. and Goldstein, I.S. (eds) Wood Structure and Composition. Marcel Dekker, New York, pp. 183–261. Colquhoun, I.J., Ralet, M.-C., Thibault, J.-F., Faulds, C.B. and Williamson, G. (1994) Structure and identifica- tion of feruloylated oligosaccharides from sugar-beet pulp by NMR spectroscopy. Carbohydrate Research 263, 243–256. Cronquist, A. (1981) An Integrated System of Classification of Flowering Plants. Columbia University Press, New York. Dahlgren, R.M.T., Clifford, H.T. and Yeo, P.F. (1985) The Families of the Monocotyledons. Springer-Verlag, New York. Darvill, J.E., McNeil, M., Darvill, A.G. and Albersheim, P. (1980) Structure of plant cell walls XI. Glucuronoarabinoxylan. A second hemicellulose in the primary cell walls of suspension-cultured sycamore cells. Plant Physiology 66, 1135–1139. Donaldson, L.A. (2001) Lignification and lignin topochemistry – an ultrastructural view. Phytochemistry 17, 859–873. Edashige, Y. and Ishii, T. (1996) Structures of cell-wall polysaccharides from suspension-cultured cells of Cryptomeria japonica. Mokuzai Gakkaishi 42, 895–900.

224 P.J. Harris Edashige, Y. and Ishii, T. (1997) Rhamnogalacturonan I from xylem differentiating zones of Cryptomeria japonica. Carbohydrate Research 304, 357–365. Edashige, Y., Ishii, T., Hiroi, T. and Fujii, T. (1995) Structural analysis of polysaccharides of primary walls from xylem differentiating zones of Cryptomeria Japonica D. Don. Holzforschung 49, 197–202. Fry, S.C. (1983) Feruloylated pectins from the primary cell wall: their structure and possible functions. Planta 157, 111–123. Fry, S.C. (1988) The Growing Plant Cell Wall: Chemical and Metabolic Analysis. Longman, Harlow, UK. Fry, S.C., York, W.S., Albersheim, P., Darvill, A., Hayashi, T., Joseleau, J.-P., Kato, Y., Lorences, E.P., Maclachlan, G.A., Mort, A.J., Reid, J.S., Seitz, H.U., Selvendran, R.R., Voragen, A.G.J. and White, A.R. (1993) An unambiguous nomenclature for xyloglucan-derived oligosaccharides. Physiologia Plantarum 89, 1–3. Gaspar, Y., Johnson, K.L., McKenna, J.A., Bacic, A. and Schultz, C.J. (2001) The complex structures of ara- binogalactan-proteins and the journey towards understanding function. Plant Molecular Biology 47, 161–176. Gibeaut, D.M., Pauly, M., Bacic, A. and Fincher, G.B. (2002) Cell wall polysaccharide composition of barley coleoptiles. Plant Polysaccharide Workshop 2002, Palm Cove, Queensland, Australia. Glushka, J.N., Terrell, M., York, W.S., O’Neill, M.A., Gucwa, A., Darvill, A.G., Albersheim, P. and Prestegard, J.H. (2003) Primary structure of the 2-O-methyl-␣-L-fucose-containing side chain of the pectic polysaccharide, rhamnogalacturonan II. Carbohydrate Research 338, 341–352. Hantus, S., Pauly, M., Darvill, A.G., Albersheim, P. and York, W.S. (1997) Structural characterization of novel L-galactose-containing oligosaccharide subunits of jojoba seed xyloglucans. Carbohydrate Research 304, 11–20. Harris, P.J. (1983) Cell walls. In: Hall, J.L. and Moore, A.L. (eds) Isolation of Membranes and Organelles from Plant Cells. Academic Press, London, pp. 25–53. Harris, P.J. (2000) Composition of monocotyledon cell walls: implications for biosystematics. In: Wilson, K.L. and Morrison, D.A. (eds) Monocots: Systematics and Evolution. CSIRO, Melbourne, pp. 114–126. Harris, P.J. and Hartley, R.D. (1980) Phenolic constituents of the cell walls of monocotyledons. Biochemical Systematics and Ecology 8, 153–160. Harris, P.J., Kelderman, M.R., Kendon, M.F. and McKenzie, R.J. (1997) Monosaccharide composition of unlignified cell walls of monocotyledons in relation to the occurrence of wall-bound ferulic acid. Biochemical Systematics and Ecology 25, 167–179. Hartley, R.D. and Harris, P.J. (1981) Phenolic constituents of the cell walls of dicotyledons. Biochemical Systematics and Ecology 9, 189–203. Hartley, R.D. and Morrison, W.H. (1991) Monomeric and dimeric phenolic acids released from cell walls of grasses by sequential treatment with sodium hydroxide. Journal of the Science of Food and Agriculture 55, 365–375. Ishii, T. and Tobita, T. (1993) Structural characterization of feruloyl oligosaccharides from spinach-leaf cell walls. Carbohydrate Research 248, 179–190. Ishii, T., Matsunaga, T., Pellerin, P., O’Neill, M.A., Darvill, A. and Albersheim, P. (1999) The plant cell wall polysaccharide rhamnogalacturonan II self-assembles into a covalently cross-linked dimer. Journal of Biological Chemistry 274, 13098–13104. Jarvis, M.C., Forsyth, W. and Duncan, H.J. (1988) A survey of the pectic content of non-lignified monocot cell walls. Plant Physiology 88, 309–314. Jia, Z., Qin, Q., Darvill, A.G. and York, W.S. (2003) Structure of the xyloglucan produced by suspension-cul- tured tomato cells. Carbohydrate Research 338, 1197–1208. Judd, W.S., Campbell, C.S., Kellogg, E.A., Stevens, P.F. and Donoghue, M.J. (2002) Plant Systematics: a Phylogenetic Approach. Sinauer, Sunderland, Massachusetts. Kakegawa, K., Edashige, Y. and Ishii, T. (1998) Xyloglucan from xylem-differentiating zones of Cryptomeria japonica. Phytochemistry 47, 767–771. Kieliszewski, M. and Lamport, D.T.A. (1994) Extensin: repetitive motifs, functional sites, post-translational codes, and phylogeny. The Plant Journal 5, 157–172. Kieliszewski, M., de Zacks, R., Leykam, J.F. and Lamport, D.T.A. (1992) A repetitive proline-rich protein from the gymnosperm Douglas fir is a hydroxyproline-rich glycoprotein. Plant Physiology 98, 919–926. Kooiman, P. (1960) On the occurrence of amyloids in plant seeds. Acta Botanica Neerlandica 9, 208–219. Kuroda, K., Ozawa, T. and Ueno, T. (2001) Characterization of sago palm (Metroxylon sagu) lignin by ana- lytical pyrolysis. Journal of Agricultural and Food Chemistry 49, 1840–1847. Lamport, D.T.A. and Miller, D.H. (1971) Hydroxyproline arabinosides in the plant kingdom. Plant Physiology 48, 454–456.

Diversity in plant cell walls 225 Landucci, L.L., Deka, G.C. and Roy, D.N.A. (1992) 13C NMR study of milled wood lignins from hybrid Salix clones. Holzforschung 46, 505–511. Logan, K.J. and Thomas, B.A. (1985) Distribution of lignin derivatives in plants. New Phytologist 99, 571–585. Lu, F. and Ralph, J. (1999) Detection and determination of p-coumaroylated units in lignin. Journal of Agricultural and Food Chemistry 47, 1888–1992. Lu, F. and Ralph, J. (2002) Preliminary evidence for sinapyl acetate as a lignin monomer in kenaf. Chemical Communications 1, 90–91. Mabberley, D.J. (1997) The Plant-book: a Portable Dictionary of the Vascular Plants. Cambridge University Press, Cambridge. Maruyama, K., Goto, C., Numata, M., Suzuki, T., Nakagawa, Y., Hoshino, T. and Uchiyama, T. (1996) O- Acetylated xyloglucan in extracellular polysaccharides from cell-suspension cultures of Mentha. Phytochemistry 41, 1309–1314. Matsunaga, T., Ishii, T., Matsumoto, S., Higuchi, M., Darvill, A., Albersheim, P. and O’Neill, M.A. (2004) Occurrence of the primary cell wall polysaccharide rhamnogalacturonan II in pteridophytes, lyco- phytes, and bryophytes. Implications for the evolution of vascular plants. Plant Physiology 134, 1–13. McDougall, G.J. and Fry, S.C. (1994) Fucosylated xyloglucans in suspension-cultured cells of the gramina- ceous monocotyledon, Festuca arundinacea. Journal of Plant Physiology 143, 591–595. Ohsumi, C. and Hayashi, T. (1994) The oligosaccharide units of the xyloglucans in the cell walls of bulbs of onion, garlic and their hybrid. Plant Cell Physiology 35, 963–967. Ozawa, T., Ueno, T., Negishi, O. and Masaki, S. (1998) Chemical characteristics of hemicelluloses in the fibrous residue of sago palm. Nettai Nogyo 42, 172–178. Parker, M.L., Ng, A., Smith, A.C. and Waldron, K.W. (2000) Esterified phenolics of the cell walls of chufa (Cyperus esculentus L.) tubers and their role in texture. Journal of Agricultural and Food Chemistry 48, 6284–6291. Parr, A.J., Waldron, K.W., Ng, A. and Parker, M.L. (1996) The wall-bound phenolics of Chinese water chest- nut (Eleocharis dulcis). Journal of the Science of Food and Agriculture 71, 501–507. Parr, A.J., Ng, A. and Waldron, K.W. (1997) Ester-linked phenolic components of carrot cell walls. Journal of Agricultural and Food Chemistry 45, 2468–2471. Pauly, M., Qin, Q., Greene, H., Albersheim, P., Darvill, A. and York, W.S. (2001) Changes in the structure of xyloglucan during cell elongation. Planta 212, 842–850. Perrone, P., Hegwage, C.M., Thomson, A.R., Bailey, K., Sadler, I.H. and Fry, S.C. (2002) Patterns of methyl and O-acetyl esterification in spinach pectins: new complexity. Phytochemistry 60, 67–77. Popper, Z.A. and Fry, S.C. (2003) Primary cell wall composition of bryophytes and charophytes. Annals of Botany 91, 1–12. Popper, Z.A., Sadler, I.H. and Fry, S.C. (2001) 3-O-Methyl-D-galactose residues in lycophyte primary cell walls. Phytochemistry 57, 711–719. Pryer, K.M., Schnelder, H., Smith, A.R., Cranfill, R., Wolf, P.G., Hunt, J.S. and Sipes, S.D. (2001) Horsetails and ferns are a monophyletic group and the closest living relatives to seed plants. Nature 409, 618–622. Ralph, J., Quideau, S., Grabber, J.H. and Hatfield, R.D. (1994) Identification and synthesis of new ferulic acid dehydrodimers present in grass cell walls. Journal of the Chemical Society, Perkin Transactions 1 23, 3485–3498. Ralph, J., Lapierre, C., Marita, J.M., Kim, H., Lu, F., Hatfield, R.D., Ralph, S., Chapple, C., Franke, R., Hemm, M.R., Van Doorsselaere, J., Sederoff, R.R., O’Malley, D.M., Scott, J.T., MacKay, J.J., Yahiaoui, N., Boudet, A.-M., Pean, M., Pilate, G., Jouanin, L. and Boerjan, W. (2001) Elucidation of new struc- tures in lignins of CAD- and COMT-deficient plants by NMR. Phytochemistry 57, 993–1003. Redgwell, R.J. and Selvendran, R.R. (1986) Structural features of cell-wall polysaccharides of onion Allium cepa. Carbohydrate Research 157, 183–199. Renard, C.M.G.C., Champenois, Y. and Thibault, J.-F. (1993) Characterisation of the extractable pectins and hemicelluloses of the cell walls of glasswort, Salicornia ramosissima. Carbohydrate Polymers 22, 239–245. Renard, C.M.G.C., Wende, G. and Booth, E. (1999) Cell wall phenolics and polysaccharides in different tis- sues of quinoa (Chenopodium quinoa Willd). Journal of the Science of Food and Agriculture 79, 2029–2034. Rhodes, D.I. and Stone, B.A. (2002) Proteins in walls of wheat aleurone cells. Journal of Cereal Science 36, 83–101.

226 P.J. Harris Rhodes, D.I., Sadek, M. and Stone, B.A. (2002) Hydroxycinnamic acids in walls of wheat aleurone cells. Journal of Cereal Science 36, 67–81. Ridley, B.L., O’Neill, M.A. and Mohnen, D. (2001) Pectins: structure, biosynthesis, and oligogalacturonide- related signaling. Phytochemistry 57, 929–967. Ringli, C., Keller, B. and Ryser, U. (2001) Glycine-rich proteins as structural components of plant cell walls. Cellular and Molecular Life Sciences 58, 1430–1441. Rodriguez-Arcos, R.C., Smith, A.C. and Waldron, K.W. (2002) Effect of storage on wall-bound phenolics in green asparagus. Journal of Agricultural and Food Chemistry 50, 3197–3203. Rouau, X., Cheynier, V., Surget, A., Gloux, D., Barron, C., Meudec, E., Louis-Montero, J. and Criton, M. (2003) A dehydrotrimer of ferulic acid from maize bran. Phytochemistry 63, 899–903. Saka, S. and Goring, D.A.I. (1985) Localization of lignins in wood cell walls. In: Higuchi, T. (ed.) Biosynthesis and Biodegradation of Wood Components. Academic Press, Orlando, Florida, pp. 57–62. Sánchez, M., Pena, M.J., Revilla, G. and Zarra, I. (1996) Changes in dehydrodiferulic acids and peroxidase activity against ferulic acid associated with cell walls during growth of Pinus pinaster hypocotyl. Plant Physiology 111, 941–946. Sarkanen, K.V. and Hergert, H.L. (1971) Classification and distribution. In: Sarkanen, K.V. and Ludwig, C.H. (eds) Lignins: Occurrence, Formation, Structure and Reactions. Wiley-Interscience, New York, pp. 43–93. Schröder, R., Nicholas, P., Vincent, S.J.F., Fischer, M., Reymond, S. and Redgwell, R.J. (2001) Purification and characterisation of a galactoglucomannan from kiwifruit (Actinidia deliciosa). Carbohydrate Research 331, 291–306. Shimokawa, T., Ishii, T. and Matsunaga, T. (1999) Isolation and structural characterization of rhamnogalac- turonan II–borate complex from Pinus densiflora. Journal of Wood Science 45, 435–439. Showalter, A.M. (1993) Structure and function of plant cell wall proteins. The Plant Cell 5, 9–23. Smith, B.G. and Harris, P.J. (1995) Polysaccharide composition of unlignified cell walls of pineapple [Ananas comosus (L.) Merr.] fruit. Plant Physiology 107, 1399–1409. Smith, B.G. and Harris, P.J. (1999) The polysaccharide composition of Poales cell walls: Poaceae cell walls are not unique. Biochemical Systematics and Ecology 27, 33–53. Smith, B.G. and Harris, P.J. (2001) Ferulic acid is esterified to glucuronoarabinoxylans in pineapple cell walls. Phytochemistry 56, 513–519. Soltis, D.E., Soltis, P.S. and Zanis, M.J. (2002) Phylogeny of seed plants based on evidence from eight genes. American Journal of Botany 89, 1670–1681. Stone, B.A. and Clarke, A.E. (1992) Chemistry and Biology of (1→3)-␤-Glucans. La Trobe University Press, La Trobe University, Victoria, Australia. Sun, R.C., Fang, J.M., Goodwin, A., Lawther, J.M. and Bolton, A.J. (1999) Fractionation and characterization of ball-milled and enzyme lignins from abaca fibre. Journal of the Science of Food and Agriculture 79, 1091–1098. Thomas, J.R., McNeil, M., Darvill, A.G. and Albersheim, P. (1987) Structure of plant cell walls XIX. Isolation and characterization of wall polysaccharides from suspension-cultured Douglas fir cells. Plant Physiology 83, 659–671. Thomas, J.R., Darvill, A.G. and Albersheim, P. (1989) Rhamnogalacturonan I, a pectic polysaccharide that is a component of moncot cell-walls. Carbohydrate Research 185, 279–305. Timell, T.E. (1960) Studies on Ginkgo biloba, L. 1. General characteristics and chemical composition. Svensk Papperstidning 63, 652–657. Timell, T.E. (1962a) Studies on ferns (Filicineae). Part 1. The constitution of a xylan from cinnamon fern (Osmunda cinnamomea L.). Svensk Papperstidning 65, 122–125. Timell, T.E. (1962b) Studies on ferns (Filicineae). Part 2. The properties of a galactoglucomannan and a cellu- lose from cinnamon fern (Osmunda cinnamomea L.). Svensk Papperstidning 65, 173–177. Timell, T.E. (1962c) Studies on ferns (Filicineae) Part 3. General chemical characteristics and comparison with gymnosperms and angiosperms. Svensk Papperstidning 65, 266–272. Timell, T.E. (1964) Studies on some ancient plants. Svensk Papperstidning 67, 356–363. Towers, G.H.N. and Gibbs, R.D. (1953) Lignin chemistry and taxonomy of higher plants. Nature 172, 25–26. Trethewey, J.A.K. and Harris, P.J. (2002) Location of (1→3)- and (1→3), (1→4)-␤-D-glucans in vegetative cell walls of barley (Hordeum vulgare) using immunolabelling. New Phytologist 154, 347–358. Vanzin, G.F., Madson, M., Carpita, N.C., Raikhel, N.V., Keegstra, K. and Reiter, W.-D. (2002) The mur2 mutant of Arabidopsis thaliana lacks fucosylated xyloglucan because of a lesion in fucosyltransferase AtFUT1. Proceedings of the National Academy of Sciences USA 99, 3340–3345.

Diversity in plant cell walls 227 Vierhuis, E., York, W.S., Kolli, V.S.K., Vincken, J.-P., Schols, H.A., Van Alebeek, G.-J.W.M. and Voragen, A.G.J. (2001) Structural analyses of two arabinose containing oligosaccharides derived from olive fruit xyloglucan: XXSG and XLSG. Carbohydrate Research 332, 285–297. Vincken, J.-P., Wijsman, A.J.M., Beldman, G., Niessen, W.M.A. and Voragen, A.G.J. (1996) Potato xyloglu- can is built from XXGG-type subunits. Carbohydrate Research 288, 219–232. Vincken, J.-P., York, W.S., Beldman, G. and Voragen, A.G.J. (1997) Two general branching patterns of xyloglucan, XXG and XXGG. Plant Physiology 114, 9–13. Waldron, K.W. and Selvendran, R.R. (1990) Composition of the cell walls of different asparagus (Asparagus officinalis) tissues. Physiologia Plantarum 80, 568–575. Waldron, K.W., Ng, A., Parker, M.L. and Parr, A.J. (1997) Ferulic acid dehydrodimers in the cell walls of Beta vulgaris and their possible role in texture. Journal of the Science of Food and Agriculture 74, 221–228. Wende, G. and Fry, S.C. (1997) 2-O-␤-D-Xylopyranosyl-(5-O-feruloyl)-L-arabinose, a widespread compo- nent of grass cell walls. Phytochemistry 44, 1019–1030. Whistler, R.L. and Chen, C.-C. (1991) Hemicelluloses. In: Lewin, M. and Goldstein, I.S. (eds) Wood Structure and Composition. Marcel Dekker, New York, pp. 287–319. White, A.R., Elmore, H.W. and Watson, M.B. (1986) Cell wall polysaccharides and carbohydrate composi- tion of fern tissue cultures. 13th International Carbohydrate Symposium, Abstract B 129, Cornell University, Ithaca, New York. White, E. and Towers, G.H.N. (1967) Comparative biochemistry of the lycopods. Phytochemistry 6, 663–667. Willats, W.G.T., Steele-King, C.G., Marcus, S.E. and Knox, J.P. (1999) Side chains of pectic polysaccharides are regulated in relation to cell proliferation and cell differentiation. The Plant Journal 20, 619–628. Willats, W.G.T., Steele-King, C.G., McCartney, L., Orfila, C., Marcus, S.E. and Knox, J.P. (2000) Making and using antibody probes to study plant cell walls. Plant Physiology and Biochemistry 38, 27–36. Willats, W.G.T., McCartney, L., Mackie, W. and Knox, J.P. (2001) Pectin: cell biology and prospects for func- tional analysis. Plant Molecular Biology 47, 9–27. York, W.S., Kolli, V.S.K., Orlando, R., Albersheim, P. and Darvill, A.G. (1996) The structures of arabinoxy- loglucans produced by solanaceous plants. Carbohydrate Research 285, 99–128. Zablackis, E., York, W.S., Pauly, M., Hantus, S., Reiter, W.-D., Chapple, C.C.S., Albersheim, P. and Darvill, A. (1996) Substitution of L-fucose by L-galactose in cell walls of Arabidopsis mur1. Science 272, 1808–1810.

12 Diversity in secondary metabolism in plants Peter G. Waterman Centre for Phytochemistry, Southern Cross University, Lismore, NSW 2480, Australia Introduction variable, amount of resources in producing secondary metabolites. However, a consen- In all higher plants discrete organic com- sus view has now evolved among most inter- pounds are found that are not intimately ested scientists that their presence does, in involved in the primary processes of physiol- some way, confer some advantage to the ogy and development. Collectively such com- producer, usually through their interaction pounds are often referred to as secondary with extrinsic factors in the environment. metabolites, presumably because they have no obvious impact on primary metabolism. One Paramount among these is that they, to a of the most striking features of secondary degree, defend the producer against the metabolites is their enormous structural diver- attacks of predators and pathogens. Such a sity. This is well illustrated by the Dictionary of function would certainly explain why there Natural Products (2003), the most recent edi- would be selection for structural diversity tion of which lists over 140,000 discrete com- in as much as a ‘defence role’ would need pounds, most of which originate from higher to be viewed in terms of a war of adapta- plant sources. They have given rise to a tion and counter-adaptation between plants branch of chemistry (natural products chem- and their pests. The evidence for this role istry) and have been a significant stimulant for is now compelling (Rosenthal and the development of chromatographic, spec- Berenbaum, 1991). troscopic and synthetic techniques that are necessary for their isolation and characteriza- Other secondary metabolites are impli- tion. Secondary metabolites are also notewor- cated in attraction of pollinators (colour and thy for exhibiting a wide range of biological scent) while the complex issue of arranging activities and they have played a critical role in seed dispersal by an animal mediator can the development of our current pharmaceuti- require appreciable manipulation in sec- cal and agrochemical armoury. ondary metabolite profiles in unripe fruit, which needs to be protected, and ripe fruit, Why produce secondary metabolites? where the mediator needs to be encouraged to take the fruit but avoid eating the seed. Unsurprisingly there is no simple answer to why plants invest a significant, although Such a view of secondary metabolites dif- fers somewhat from that described by Jarvis (2000). He recognizes compounds that act as internal messengers as secondary metabo- lites, although he expressly avoids the use of © CAB International 2005. Plant Diversity and Evolution: Genotypic and 229 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

230 P.G. Waterman that term because of a blurred distinction The Building Blocks of Secondary from primary metabolites. In this discussion Metabolism emphasis is placed on those metabolites that are of limited distribution, and conserved It is one of the wonders of evolution that, messenger molecules, plant hormones and despite the enormous diversity in end prod- so on are excluded. However, that does not ucts, the carbon skeletons of most secondary eliminate the possibility, indeed the likeli- metabolites can be traced back to three hood, that some secondary metabolites are building blocks: ‘acetate’, ‘mevalonate’ and multifunctional and have both intrinsic and ‘shikimate’ (Fig. 12.1). Furthermore, none of extrinsic roles to play. these building blocks is used specifically for Fig. 12.1. Derivation of the three building blocks from primary processes.

Diversity in secondary metabolism in plants 231 secondary metabolites but they are critical IPP is clearly independent of mevalonic acid components of primary processes. Some and it is probably better in future to refer to amino acids feature as precursors for nitro- the collective pathway and its outcomes as gen-containing secondary metabolites, the terpenoid pathway. notably the alkaloids. Simple sugars are often involved, giving rise to compounds Shikimate known collectively as glycosides. Acetate The shikimic acid pathway, originating in the formation of the cyclic (six-membered The acetate or polyketide pathway is based ring) acid from erythrose-4-phosphate and on a two-carbon unit attached to co-enzyme phosphoenolpyruvate, spins off relatively A (acetyl-coA; Fig. 12.1). Polymerization of few secondary metabolites. The addition of a the acetate units builds carbon chains that further three-carbon side chain by the inclu- can either remain linear or cyclize. By varia- sion of further phosphoenolpyruvate to tion in the length and the mode of cycliza- yield the C6–C3 phenylpropane is, in con- tion of the polyketide chain, the array of trast, the starting point for many classes of products that can be formed is vast. The secondary metabolite (Fig. 12.1). Seemingly highpoint of polyketide secondary metabo- there are once again parallel pathways to be lite biosynthesis is probably found in the found in plastids and cytosol, but in this case macrolide and polyether antibiotics pro- the precursors appear to be identical duced in Streptomyces, fungi and some algae (Hrazdina and Jensen, 1992). (Tsantrizos and Yang, 2000), but it remains a vital player in secondary metabolism of Amino acids higher plants. There is an interesting dichotomy among Mevalonate natural amino acids with a few (phenylala- nine, tyrosine, tryptamine, lysine, ornithine) Mevalonic acid is a six-carbon intermediate commonly incorporated into secondary that, by decarboxylation, gives rise to a five- metabolites whilst others are rarely involved, carbon branched chain (isoprene) unit that particularly in higher plants. The alkaloids occurs in the activated (through phosphoryla- are the largest group of amino acid-derived tion) isomeric forms isopentenyl diphosphate metabolites and are often classified in (IPP) and dimethylallyl diphosphate respect of the precursor amino acid, which is (DMAPP) (see Fig. 12.1). Mevalonic acid is commonly one of the five amino acids noted itself formed by the combination of three above. Other amino acid-derived metabo- acetate units. The activated five-carbon units lites include cyanogenic compounds and are able to polymerize to give rise to a class of glucosinolate. compounds generally referred to as terpenes. The products of the terpenoid pathway are Sugars the most numerous and, even when small (C15 or C20), can be exceedingly complex. A range of hexose and pentose sugars are often added to a preformed carbon skeleton Quite recently an alternative pathway to (an aglycone) to form a class of compounds IPP and DMAPP has been found to occur in known collectively as glycosides. The most plastids (Rohmer, 1999), whereas the meval- widely exploited is ␤-D-glucose in the pyra- onate pathway occurs in the cytosol. The nose form linked to the aglycone through an penultimate product in the new route is oxygen (ether) bridge. Rhamnose, galactose, deoxyxylulose phosphate (DXP) and the pri- xylose and arabinose also occur widely, while mary precursors are pyruvate and glycer- many others are recorded less commonly. aldehyde-3-phosphate. The DXP route to

232 P.G. Waterman On occasion disaccharides right through to the Leguminosae–Papilionaceae, notably decasaccharides made up of mixtures of dif- Derris and Tephrosia. In Fig. 12.2, the origins ferent sugars can be added. of the carbon skeleton of rotenone are high- lighted. The main events are as follows: The addition of the sugar profoundly changes the solubility profile of the glyco- 1. Linkage of a phenylpropane (shikimate) side in comparison with the aglycone, mak- and polyketide (acetate) units to form the ing it far more water soluble, more readily flavonoid skeleton. transportable, and capable of storage in the 2. Modifications to the structure of that cell vacuole. skeleton to generate an isoflavone with the appropriate oxygenation pattern for forma- Integration and ornamentation tion of the rotenoid nucleus. 3. Addition of methyl (from methionine?) In the sections above, the various key build- and subsequent cyclization to the rotenoid. ing blocks have been introduced in isolation. 4. Addition of IPP to the rotenoid skeleton However, we cannot view them in isolation followed by cyclization to give the furan ring. as, in reality, many secondary metabolites are formed by linking subunits derived from Thus the rotenone incorporates elements different pathways. Take, for example, the from acetate, terpenoid and phenylpropanoid flavonoid derivative rotenone, an insectici- pathways. It also involves addition of methyl dal product found widely in some genera of units from a methyl donor and modification of the isoflavone skeleton, one methyl and the Fig. 12.2. Steps in the formation of rotenone.

Diversity in secondary metabolism in plants 233 terpenoid element through oxidative and/or cell membranes and in the production of reductive processes. This process involves a fixed oils (Forkmann and Heller, 1999). The series of specific enzymes to guide the various starter unit is most often ‘acetate’, leading to steps and is clearly controlled by the genome. linear chains with an even number of carbon atoms, but other options can occur. Between Role of Acetate each addition the carbonyl closest to the starter unit is usually eliminated and The acetate pathway extends by the addi- replaced with either a fully saturated carbon tion of two-carbon units to an existing or an olefinic group (unsaturated fatty acids). ‘starter’ unit and this will progress until a predetermined number of units have been Far more important in terms of a contri- added. Current evidence points to the bution to secondary metabolite diversity and adding unit being malonyl-CoA (HOOC- also very instructive with respect to the evo- CH2-CO-CoA) rather than acetyl-CoA (CH3- lutionary steps involved in creating diversity CO-CoA) with decarboxylation assisting in are the cyclic (polyketide) products of the driving the condensation reaction. acetate pathway. These are characteristically generated from chains of between three The simplest outcome of this process is (triketide) and seven (heptaketide) ‘acetate’ the building up of long chains to produce units. In Fig. 12.3, an example of a simple the ubiquitous fatty acids that are involved in tetraketide is shown, where the starter unit is acetyl-CoA. This simple eight-carbon Fig. 12.3. Alternative cyclizations of a tetraketide (R = CH3) bound on to chalcone synthase (CHS)- and stilbene synthase (STS)-related proteins.

234 P.G. Waterman chain can yield two distinctly different aro- Fig. 12.4. Formation of 6’-deoxychalcone from a matic (homocyclic) products depending on modified CHS cluster; * = extender carbons. which carbons are involved in the cyclization reactions. Which of these two cyclizations CHS. Both normal CHS and the 6Ј-deoxy- actually happens depends on a series of pro- CHS systems can operate in the same teins that are responsible for the condensa- species. tions to the tetraketide and then folding that tetraketide in space so that the required ring While the CHS system has been strongly closure reactions can occur. conserved in chalcone synthesis it has, it seems, proved to be adaptable, in an evolu- The chalcone synthase (CHS; Fig. 12.3) tionary sense, to substitution of the building family of proteins is a particularly important blocks, in particular the starter unit. For one in higher plants because if the starter example, Fig. 12.5a gives the structure for a unit (i.e. R in Fig. 12.3) is the CoA enzyme of typical acridone alkaloid, a type found widely a cinnamic acid variously substituted on the in the Rutaceae. From classical feeding experi- aromatic nucleus then the resulting cyclized ments it has long been known that this origi- product is of the class of compounds known nates from anthranilic acid and a triketide. It as chalcones. Chalcones are the precursors of has now been demonstrated that the combina- a huge group of metabolites that are known tion is catalysed in a manner analogous to that collectively as the flavonoids and which have of chalcones but with an enzyme system capa- been reported to serve a number of func- ble of accepting anthranoyl-CoA as the starter tions that are typical of secondary metabo- and which has a homology with CHS > 65% lites (for reviews, see Harborne, 1989, 1996). (Junghanns et al., 1995). Evidence that homol- ogous systems exist for the biosynthesis of the Flavonoids are known to occur in all xanthone skeleton (Fig. 12.5b) and ␤-tri- major plant phyla back to the Bryophyta. ketone derivatives such as humulone (Fig. The initial steps in their formation, which 12.5c) is significant but as yet not incontro- are mediated by CHS proteins, appear to vertible. A number of other likely starter units have remained the same throughout the can be considered in the formation of deriva- timescale that implies for the existence of tives of the ␤-triketone type. Among both the these enzymes. Diversity in flavonoids has acridone and the xanthone compounds iso- arisen through post-CHS changes that have lated are a few examples devoid of oxygena- modified the skeleton by a series of oxida- tion in the anticipated positions. It is not tion/reduction reactions (Petersen et al., unreasonable to presume that in such cases a 1999; Stafford, 2000) and through supple- modification similar to that proposed for 6Ј- mental additional oxidation of the skeleton, deoxychalcones (Fig. 12.4) is in operation. O-methylation and glycosylation, addition of terpenoid units (usually five-carbon) that may subsequently cyclize. The one major modification that has occurred with CHS and which is seen partic- ularly in parts of the Leguminosae is the absence of one of the hydroxyls of the phloroglucinol ring. This occurs through the intervention of an additional enzyme that uses NADPH (nicotinamide adenine dinucleotide phosphate, reduced) to reduce the keto group of the first ‘malonyl’-derived unit to an alcohol (Schröder, 2000). Subsequent to ring closure, that hydroxyl is lost in the formation of the aromatic ring (Fig. 12.4). The isoflavonoid compound rotenone (Fig. 12.2) will have been formed through the involvement of this modified

Diversity in secondary metabolism in plants 235 Fig. 12.5. Exploitation of the CHS system to produce structurally diverse secondary metabolites: (a) acridone alkaloids (starter = anthranilic acid); (b) xanthones (starter = benzoic acid); (c) ␤-triketones of the humulone type (starter = isovaleric acid); (d) C-methyl flavones (extender = methylmalonyl-CoA). * = extender carbons, starter is circled. C-methyl flavonoids occur in significant tion through the intervention of a reduction numbers. It has been demonstrated stage in chain condensation paralleling the (Schröder et al., 1998), at least in Pinus formation of 6Ј-deoxychalcones. strobus, that these are generated using the CHS assemblage with methylmalonyl-CoA as In CHS and STS systems, the cyclization the substrate for chain extension rather than of the polyketide involves formation of a car- malonyl-CoA (Fig. 12.5d). There is every bon–carbon bond. Comparable cyclizations reason to suspect that this will be true for of polyketides formed in the same way can most if not all C-methylated flavonoids and also involve condensation involving the ter- could also be envisaged for other groups of minal carbon (visualized as a carboxylic acid) ketides where C-methylation is found, such and one of the internal oxygens of the as the ␤-triketones. ketide. Such condensations lead to the for- mation of lactones, such as parasorboside In Fig. 12.3, a second cyclization of the (Fig. 12.7). tetraketide was demonstrated that gives rise to a dihydroxy aromatic acid. The combina- In assessing the acetate pathway, it is strik- tion of enzymes giving rise to these conden- ing that the various systems that have evolved sations and subsequent cyclization is known are well suited for achieving high diversity. collectively as the stilbine synthase (STS) First, each system seems to be able to modify assemblage. In the STS mechanism of in comparable ways to accept a range of cyclization, the terminal carbon of the exten- starter units, most of which are generated sion is not included and is left as a carbonyl, from other metabolic pathways. Secondly, which is generally lost as carbon dioxide. As each assemblage seems to have developed with the CHS system, there is now evidence some capacity to vary the number of conden- that diversity is achieved in the STS system sation units, use different condensation units, through the use of a range of different and perform reduction reactions on conden- starter units (Fig. 12.6) and loss of oxygena- sation units. The way in which these enzyme assemblages parallel each other in their capa-

236 P.G. Waterman (a) (b) (c) (d) Fig. 12.6. Exploitation of the STS system to produce structurally diverse secondary metabolites: (a) stilbene alkaloids (starter = cinnamic acid); (b) bibenzyl (starter = dihydrocinnamic acid); (c) biphenyl (starter = benzoic acid); (d) cannabinoids (starter = hexanoic acid). * = extender carbons, starter is circled. The carboxylic acid (in rectangular box) is always lost during the biosynthesis. Fig. 12.7. Parasorboside, a ‘lactone cyclization up of a multiple of C5 and the branched- product’ derived from a triketide. Glucose is added chain isopentene units. Condensation occurs subsequent to cyclization. through a series of enzymes known as prenyl transferases, which add IPP units to a starter bilities suggests that they have to be viewed as or intermediate chain that is structurally maximizing the synthetic potential of a single comparable to DMAPP. Most of the metabo- theme using a framework that can evolve lites generated fall into the following four with a minimum of need to invent or rein- categories: monoterpenes (C10), sesquiter- vent the wheel. As we shall see, this ‘mecha- penes (C15), diterpenes (C20) and triterpenes nistic parallelism’ repeats itself in the (C30), the latter giving rise to phytosterols exploitation of each of the major building (Fig. 12.8). It is apparent that both the classi- blocks of secondary metabolism. cal mevalonate and the DXP pathways con- tribute IPP and DMAPP for the synthesis of Role of Mevalonate secondary metabolites but there is some evi- dence that monoterpenes and diterpenes Secondary metabolites arising from the originate predominantly from DXP and ‘mevalonate’ pathway are condensates of the sesquiterpenes and triterpenes from the five-carbon precursors IPP and DMAPP (see mevalonate route (Bohlmann et al., 2000). above) and as a consequence they, for the most part, exhibit a carbon skeleton made Conversion of the diphosphate interme- diates into secondary metabolites is rela- tively easy to rationalize from a chemical perspective as it can be envisaged in terms of classical carbonium ion chemistry (Geissman and Crout, 1969). A simple example of this is shown in Fig. 12.9 in which the geranyl diphosphate (GPP)- derived monoterpenyl cation is converted by

Diversity in secondary metabolism in plants 237 Fig. 12.8. Major classes of secondary metabolites originating from IPP and DMAPP. GPP = geranyl diphosphate; FPP = farnesyl diphosphate; GGPP = geranylgeranyl diphosphate. a series of carbonium ion shifts and subse- The triterpenes are formed by the initial quent elimination of the carbonium ion condensation of two units of FPP to give through either hydration or deprotonation squalene that, in the form of squalene 2,3- to give a series of well-known monoter- epoxide, initiates through rupture of the penes. Of course, in real life, each of these epoxide and subsequent carbonium ion for- interactions is governed in a stereospecific mation a train of rearrangements that even- manner by a terpene synthase. In Fig. 12.9 tually leads to the tetracyclic and pentacyclic just a few of the many possibilities for carbo- triterpene skeletons and the plant sterols nium ion-induced structural modification of (Fig. 12.10). In the triterpenes the carbo- the GPP-type precursor are shown. In the nium ion-driven ring closures generally vast majority of cases, the eventual elimina- involve methyl migrations so that a number tion of the carbonium ion is based on one of of different methylation patterns can occur the two mechanisms shown in Fig. 12.9. on the triterpene skeleton. The larger farnesyl diphosphate (FPP) With both diterpenes and triterpenes and geranylgeranyl diphosphate (GGPP) much of the structural diversity arises intermediates leave even more scope for car- through subsequent modification of the bonium ion-driven structural diversity and terpene skeleton, usually through oxida- this is reflected in the numbers of sesquiter- tive and reductive reactions. One example pene and diterpene compounds that have of this is to be found in the Citrus family been recorded (>11,900 sesquiterpenes in (Rutaceae) and its close allies the Meliaceae, 149 structural classes; >11,100 diterpenes in Simaroubaceae and Cneoraceae where the 119 structural classes; Dictionary of Natural tirucallol tetracyclic triterpene skeleton has Products, 2003). been subjected to a series of oxidative ring

238 P.G. Waterman Fig. 12.9. Proliferation of monoterpene structures from a single precursor facilitated by carbonium ion formation and subsequent rearrangements. openings and ring closures to lead to the tones (Fig. 12.11) have evolved and limonoid and quassinoid bitter principles. adapted to allow changes to be made to In these plants, enzymatic processes to oxi- each of the homocyclic rings of the triter- dize, ring open and then reclose as lac- pene skeleton. These can lead to far-rang-

Diversity in secondary metabolism in plants 239 Fig. 12.10. Some key points in the synthesis of triterpenes and plant sterols. ing changes in the final structure to the by a ‘head-to-head’ condensation, while the extent that some compounds are difficult pyrethroids require a ‘head-to-body’ con- to recognize as having arisen from a triter- densation (Fig. 12.12). The iridoid-produc- pene. Oxidation processes are widespread ing families are now widely considered to be but rarely achieve the level seen in the monophyletic and it can be presumed that Rutaceae and its allies (see review in their evolution in higher plants has been a Waterman and Grundon, 1983). unique event. Iridoids differ from the classi- cal mainstream monoterpenes arising from Alternative modes of cyclization GPP in that they tend to be quite highly oxi- dized and often occur as glycosides. They While the ‘head-to-tail’ condensation of are also an integral part of one of the major DMAPP- and IPP-type diphosphates is by far groups of plant alkaloids (see section the most common method of skeleton build- ‘Alkaloids’, below). ing in terpenes there are alternative path- ways that are found in taxonomically Terpenes as alkylating agents restricted groups. For example, the iridoid group of monoterpenes, exemplified by The diphosphate unit, with its capacity to loganin and seco-loganic acid, are generated give a positively charged species through

240 P.G. Waterman Fig. 12.11. Structural modification of tetracyclic triterpenes involving multiple oxidative ring openings and closures. elimination of the phosphate, is predisposed hydroxyl groups or carbons adjacent to aro- to act as an alkylating agent on to elec- matic hydroxyls. This environment is met trophilic carbons and other elements. In perfectly in the alternately oxygenated rings nature these are usually either aromatic that originate from the CHS and STS sys-

Diversity in secondary metabolism in plants 241 Fig. 12.12. Products of unusual terpenoid pathways. tems in polyketides and the aromatic rings than to in vivo synthesis by the producer. of oxygenated phenylpropanes (see section Notwithstanding this issue, the addition of ‘Role of shikimate’, below). The normal the DMA unit leads to an increase not only addition unit is the 3,3-dimethylallyl (DMA) in structural diversity but also in solubility moiety and there is now substantial evidence profile within the skeletal type, because of that specific enzymes direct the DMA unit the lipophilic nature of the DMA unit. on to the carbon skeleton. This DMA unit is often subsequently oxidized to give the cor- Further modifications often occur responding epoxide (Fig. 12.13). Isolation through cyclization of the DMA side chain studies on plants containing 2Ј,3Ј- leading to either furan or pyran ring sys- epoxyDMA will often produce a series of tems. These units are very widely distrib- diols and related dehydration products. It is uted in higher plants and, like the quite likely that these are due to the acid- open-chain precursor, can modify the catalysed opening of the epoxide ring rather properties of the structure to which they are added. Fig. 12.13. Addition of dimethylallyl diphosphate to a ‘non-terpenoid’ structural skeleton and the consequent cyclization to furano and pyrano derivatives.

242 P.G. Waterman While use of the simple five-carbon DMA pivotal enzyme phenylalanine ammonia- unit is by far the most common, there are lyase (PAL). A series of enzymes lead to the significant numbers of metabolites in which step-by-step oxidation and O-methylation GPP, FPP and even GGPP are added. The of cinnamic acid to coumaric acid (4- greater length of the resulting side chain hydroxy), caffeic acid (3,4-dihydroxy), fer- and the presence of more sights for possible ulic acid (3-methoxy, 4-hydroxy), oxidation can lead to more complex ring 5-hydroxyferulic acid (3-methoxy, 4,5-dihy- closures. This is exemplified by a recent droxy) and finally sinapic acid (3,5- report (Sultana et al., 2003) on the structural dimethoxy, 4-methoxy). A family of diversity found in the coumarins of the coumarate:CoA ligase enzymes occurs that Australian genus Philotheca (Rutaceae). have been shown to be able to convert most of these acids to the corresponding CoA Role of Shikimate product, which is ready to add to a wide range of other products, while a second The formation of the phenylpropanes (com- series of hydroxylases and methyltrans- pounds with a C6C3 skeleton) via the ferases have been found that will catalyse shikimic acid pathway and the aromatic the same oxidation/methylation reactions amino acid phenylalanine is well docu- on a cinnamic acid-CoA precursor. These mented (Fig. 12.1). Once again there is evi- two options lead to a very versatile system dence for separate pathways operating in to allow changes in ring ornamentation to plastids and cytosol, but in this case they occur at different stages of metabolite pro- appear to rely on identical resources and duction (Petersen et al., 1999). comparable enzyme assemblages (Hrazdina and Jensen, 1992). A spin-off from the main The two most common groups of sec- phenylpropanoid route occurs at the ondary metabolites with wholly phenyl- shikimic acid stage (C6C1) via the formation propanoid skeletons are the coumarins and of anthranilic acid and subsequently the lignans/neolignans. Despite the relative amino acid tryptophan, both of which play a simplicity of the coumarin structure and role in the formation of alkaloids (see section the widespread occurrence of simple ‘Alkaloids’, below). It is still not entirely clear coumarins based on umbelliferone (Fig. whether the C6C1 product, gallic acid, 12.14), the exact mechanism of formation important in hydrolysable tannin formation from phenylalanine remains uncertain (see section ‘Tannins’, below), is also spun (Petersen et al., 1999). The simple off at the shikimic acid stage or occurs via ␤- coumarin skeleton does not offer great oxidation and chain reduction of a phenyl- scope for diversification other than propane. Current evidence indicates that through oxidation and O-methylation on gallic acid could originate from both sources the homocyclic ring and most of the consid- (Petersen et al., 1999). erable structural diversity that is found in some families, notably the Apiaceae and Metabolites arising from phenylpropanoid Rutaceae, arises through addition of ter- precursors penyl units and subsequent cyclization, usually to the furan or pyran derivatives Secondary metabolites with a carbon skele- (cf. Fig. 12.13). Furthermore, there is struc- ton attributable solely to phenylpropanoid tural convergence to the coumarin skele- precursors are comparatively few in num- ton, which can also be achieved from ber. The first products are the series of sim- acetate (as in C-methyl coumarins) or from ple trans-cinnamic acids that proliferate a combination of phenylpropane and from cinnamic acid itself, which has no aro- acetate (neoflavonoids, 3- or 4-aryl- matic ring substitution and arises from coumarins in which the coumarin skeleton phenylalanine via the intervention of the is partly derived from both pathways). A full range of coumarins, most of which originate from a phenylpropanoid precur- sor, is given by Murray (1997).

Diversity in secondary metabolism in plants 243 Fig. 12.14. Phenylpropanoid metabolites. In the dimeric lignans (pinoresinol and matairesinol) and neolignans one of the monomers is shown circled. Lignans and neolignans are dimers of metabolites known as flavonoids, formed by cinnamyl alcohols, usually coniferyl alcohol. the CHS route, are among the most wide- The name lignan is generally restricted to spread compounds in the plant kingdom, compounds where there is a link between C- present in all major phyla other than algae. 8 of the participating units (see for example The number of flavonoids identified is now pinoresinol and matairesinol; Fig. 12.14), in excess of 4000 (Petersen et al., 1999). while in neolignans the primary linkages are most often 3-3Ј, 8-3Ј, 8-O-4Ј or, as 8-1Ј The enzymology and genetics of almost (shown in Fig. 12.14). While lignans are all of the ‘flavonoid’ network of compounds quite widespread in nature and a number of is now well established (Forkmann and skeletal structures arise from the 8-8Ј bond Heller, 1999) and there is, through current (Ayres and Loike, 1990), the diversification distribution, evidence to show when each that has occurred in this group of metabo- enzyme evolved. Most evolutionary steps lites is relatively unspectacular. Neolignans involve changes in oxidation and reduction are of a somewhat restricted distribution but levels (see Fig. 4.13 in Peterson et al., 1999). where they do occur they are prone to The most radical skeletal modification is somewhat greater skeletal diversification that in which the phenyl group migrates (Davin and Lewis, 1992). from C-2 to C-3 (Fig. 12.2) to form the isoflavone skeleton. Isoflavone synthase Products from phenylpropanoid and acetate seems to have evolved in the Bryophyta pathways (Stafford, 2000) but it is in the Fabaceae that it is developed to form a major class of The incorporation of a cinnamoyl-CoA metabolites that undergo a series of further starter into a tetraketide by means of the cyclizations to form, for example, CHS and STS enzyme groupings has rotenones, pterocarpans and 3-phenyl- already been discussed above. The class of coumarins. An excellent review of structural variation in flavonoids and isoflavonoids can be found in Bohm (1998).

244 P.G. Waterman Tannins Alkaloids The term tannin embraces compounds of Of all the various classes of secondary different structural types but with a common metabolites, it is alkaloids that have received function: the capacity, under some condi- by far the most attention. The dominating tions, to form insoluble complexes with pro- reason for this is that so many of them teins and polysaccharides (Waterman and exhibit profound biological activities and a Mole, 1994). The three most important tan- significant number have found their way nin types are: into traditional and modern materia medica (e.g. morphine, codeine, vincristine, cocaine, 1. Phlorotannins: polymers of cyclic tri- ergotamine, atropine, hyoscine, physostig- ketides in which individual rings are linked mine, strychnine, nicotine). to one another by C–C bonds. Thus, phlorotannins are entirely acetate in origin. The structures of alkaloids contain, by They are found only in algae. definition, nitrogen and the vast majority 2. Condensed tannins: polymers of polyhy- of nitrogenous secondary metabolites droxy flavan-3-ol flavonoid units, which are would be classified as alkaloids. The key found in pteridophytes, gymnosperms and in defining factor for a metabolite to be woody species of angiosperms. Based on the included in the class of alkaloids is that the flavonoid nucleus, they are of mixed phenyl- source of the nitrogen is an amino acid and propane/acetate origin (see section above). that all or most of the amino acid is incor- 3. Hydrolysable tannins: glycosides in which porated into the structure. Essentially ter- a central monosaccharide unit, most com- penoid (steroidal alkaloids, diterpene monly glucose, is esterified by gallic acid alkaloids) or polyketide (coniine) metabo- (3,4,5-trihydroxybenzoic acid) or ellagic acid lites into which nitrogen is inserted share (a cyclized dimeric gallic acid). A single hex- the likelihood for significant biological ose sugar can link with up to five of these activity but are referred to as pseudoalka- acid units. Hydrolysable tannins occur loids. Other classes of nitrogen-containing within a fairly restricted group of families in secondary metabolites that are usually not higher plants. considered to be alkaloids are the isothio- cyanates and cyanogenic glycosides. All of In comparison with most classes of sec- these various groups are reviewed in ondary metabolites, the tannins are struc- Waterman (1993). None of them shares the turally somewhat conservative. This is to coherent biosynthetic themes that tie the some extent inevitable because there are true alkaloids together as a subset of sec- fewer avenues open for structural modifica- ondary metabolites, which are derived tion that do not reduce their capacity to through a ‘shared’ route in which the parts bind to their target molecules. However, if may change but the mechanisms uniting you accept that their prime roles are as them to form alkaloids basically remain the digestibility reducing and perhaps anti-oxi- same. As noted previously (Waterman, dant compounds and that in the former role 1998), this ‘shared’ theme in alkaloid for- their target is primarily a general class of mation requires the bringing together of a compounds (proteins), then there is proba- nitrogen source (the NH2 of an amino acid) bly far less evolutionary pressure to maxi- and a ketone source (usually in the form of mize diversification than is found with most an aldehyde) and the formation of a C–N other classes of secondary metabolite. The (imino) bond involving the two entities, last statement is a simplification of the situa- together with subsequent cyclization involv- tion, both in terms of tannin structural ing this imino group. How this can occur is diversity and the variability of the demonstrated in the formation of the tannin–protein interaction (Waterman and pyrrolizidine alkaloids (Fig. 12.15). In this Mole, 1994), but holds true for comparison example, ornithine contributes both of the with structural diversification in many key functional groups through different groups of secondary metabolites. modifications. The components contribut-

Diversity in secondary metabolism in plants 245 Fig. 12.15. Formation of pyrrolizidine alkaloids from two ornithine precursor units. * = the ketone-bearing carbon. ing to the basic heterocyclic skeletons of the access the genes responsible for the key more widespread classes of alkaloids are imino bond-generating step it is going to be listed in Table 12.1. fascinating to see how it differs between such diverse higher plants as the daffodil, opium In the biosynthesis of alkaloids, there- poppy and lupin. fore, the key process would appear to be bringing the two participating functional Once the initial alkaloid skeleton is gen- groups into close proximity in an environ- erated, then further diversification uses ment where formation of the C=N (imine) much the same types of modification already bond can be generated. The ability to diver- described for other classes (according to the sify the alkaloid skeleton then depends on Dictionary of Natural Products (2003) there are the evolution of modifications that allow the currently 224 structural classes of alkaloids substrate to be changed. Unfortunately known). In particular the 1-benzylisoquino- identification of the genes involved in alka- line and indole-monoterpene types diversify loid biosynthesis has been slow (Saito and through a wide range of possible secondary Murakoshi, 1998). When we eventually cyclizations (see Chapters 9 and 10 in Table 12.1. Origins of major subclasses of alkaloids. Alkaloid type N source Ketone source Amaryllidaceae Tyrosine/phenylalanine Phenylbenzaldehyde derived from tyrosine/phenylalanine 1-Benzylisoquinoline Tyrosine/phenylalanine Phenylacetaldehyde derived from tyrosine/phenylalanine Emetine type Tyrosine Seco-loganin (see Fig. 12.12) Indole-monoterpene Tryptophan Seco-loganin Anthranilate Anthranilic acid ‘Acetate’ chains Betalains Tyrosine or proline Tyrosine (betalamic acid) Tropane Ornithine Ornithine (same molecule) and then diketide Pyrrolizidine Ornithine Aminoaldehyde derived from ornithine Quinolizidine Lysine Lysine-derived aldehyde(s) From Hrazdina and Jensen, 1992.

246 P.G. Waterman Waterman, 1993). Subsequent to these among phytochemists that we would be able refinements in the actual carbon skeleton to predict phylogenetic relationships there are then all the classic through the array of secondary metabolites oxidation/methylation type ornamentations produced by any given species. Subsequently, that produce further diversity, which, while we have found this not to be the case and relatively subtle, can have significant impact that, while it is certainly true that we can link on biological activity (Wink, 1998). some classes of compounds with individual families or related groups of families, sec- Evolution of Secondary Metabolites: ondary metabolites generally show distribu- Some Concluding Comments tions which contain disjunct elements. The most striking feature of secondary This may well be due, in part, to paral- metabolism in higher plants is not that there lelism and convergence in biosynthetic is such staggering diversity in end products pathways, given the apparent adaptability but that the level of diversity is attained on of many of the individual enzymes involved the back of so few building blocks and on and the limited number of building blocks. the use of a relatively small number of Probably more important is that the genes chemical reactions to bring about skeletal for secondary metabolite synthesis are formation and modification and a similarly highly conserved in the genomes of higher small number for skeleton ornamentation. It plants and may become dormant for con- seems apparent that, in the process of evolu- siderable phylogenetic distances before tion, individual enzymes have proved adapt- being reactivated. That this is the case for able in taking on new substrates and have quinolizidine alkaloids was demonstrated been integrated into pathways at different some 20 years ago (Wink and Witte, 1983) points to assist in the generation of struc- but, unfortunately, the phenomenon turally altered end products. remains largely explored. Confirming con- servation and establishing how the genes In the 1960s, with the advent of chemical are involved in the generation of secondary taxonomy as a science, there was real hope metabolites is, today, a major challenge fac- ing molecular biology. References Ayres, D.C. and Loike, J.D. (1990) Lignans: Chemical, Biological and Clinical Properties. Cambridge University Press, Cambridge. Bohlmann, J., Gershenzon, J. and Aubourg, S. (2000) Biochemical, molecular genetic and evolutionary aspects of defense-related terpenoid metabolism in conifers. In: Romeo, J.T., Ibrahim, R., Varin, L. and De Luca, V. (eds) Evolution of Metabolic Pathways. Elsevier, Oxford, pp. 109–150. Bohm, B.A. (1998) Introduction to Flavonoids. Harwood Academic Publishers, Amsterdam. Davin, L.B. and Lewis, N.G. (1992) Phenylpropanoid metabolism: biosynthesis of monolignols, lignans, neolignans, lignins and suberins. In: Stafford, H.A. and Ibrahim, R.K. (eds) Phenolic Metabolism in Plants. Plenum Press, New York, pp. 325–375. Dictionary of Natural Products (2003) Version 11.2. Chapman & Hall through CRC Press, Boca Raton, Florida, CD-ROM. Forkmann, G. and Heller, W. (1999) Biosynthesis of flavonoids. In: Barton, D.H.R., Nakanishi, K. and Meth- Coon, O. (eds) Comprehensive Natural Products Chemistry, Vol. 1, Polyketides and Other Secondary Metabolites Including Fatty Acids and Their Derivatives (Sankawa, U., volume ed.). Elsevier Press, Oxford, pp. 713–748. Geissman, T.A. and Crout, D.H.G. (1969) Organic Chemistry of Secondary Plant Metabolism. Freeman Cooper & Co., San Francisco, California. Harborne, J.B. (ed.) (1989) Plant phenolics. In: Dey, P.M. and Harborne, J.B. (eds) Methods in Plant Biochemistry, Vol. 1. Academic Press, London. Harborne, J.B. (ed.) (1996) The Flavonoids: Advances in Research Since 1986. Chapman & Hall, London.

Diversity in secondary metabolism in plants 247 Hrazdina, G. and Jensen, R.A. (1992) Spatial organization of enzymes in plant metabolic pathways. Annual Review of Plant Physiology and Plant Molecular Biology 43, 241–267. Jarvis, B.B. (2000) The role of natural products in evolution. In: Romeo, J.T., Ibrahim, R., Varin, L. and De Luca, V. (eds) Evolution of Metabolic Pathways. Elsevier, Oxford, pp. 1–24. Junghanns, K.T., Kneusel, R.E., Baumert, A., Maier, W., Gröger, D. and Matern, U. (1995) Molecular cloning and recombinant expression of acridone synthase from elicited Ruta graveolens L. cell suspension cul- tures. Plant Molecular Biology 27, 681–692. Murray, R.D.H. (1997) Naturally occurring plant coumarins. Progress in the Chemistry of Organic Natural Products 72, 1–119. Petersen, M., Stracke, D. and Matern, U. (1999) Biosynthesis of phenylpropanoids and related compounds. In: Wink, M. (ed.) Biochemistry of Plant Secondary Metabolism. Sheffield Academic Press, Sheffield, pp. 151–221. Rohmer, M. (1999) A mevalonate-independent route to isopentenyl pyrophosphate. In: Cane, D.A. (ed.) Comprehensive Natural Products: Isoprenoids, Including Carotenoids and Steroids, Vol. 2. Pergamon Press, New York, pp. 45–67. Rosenthal, G.A. and Berenbaum, M.R. (eds) (1991) Herbivores: Their Interactions with Secondary Metabolites, 2nd edn. Academic Press, San Diego, California. Saito, K. and Murakoshi, I. (1998) Genes in alkaloid metabolism. In: Roberts, M.F. and Wink, M. (eds) Alkaloids: Biochemistry, Ecology and Medical Applications. Plenum Press, New York, pp. 147–157. Schröder, J. (2000) The family of chalscone synthase-related proteins: functional diversity and evolution. In: Romeo, J.T., Ibrahim, R., Varin, L. and De Luca, V. (eds) Evolution of Metabolic Pathways. Elsevier, Oxford, pp. 55–89. Schröder, J., Raiber, S., Berger, T., Schmidt, A., Schmidt, J., Soares-Sello, A.M., Bardshiri, E., Strack, D., Simpson, T.J., Veit, M. and Schröder, G. (1998) Plant polyketide synthases: a chalcone synthatase-type enzyme which performs a condensation reaction with methylmalonyl-CoA in the biosynthesis of C- methylated chalcones. Biochemistry 37, 8417–8425. Stafford, H.A. (2000) The evolution of phenolics in plants. In: Romeo, J.T., Ibrahim, R., Varin, L. and De Luca, V. (eds) Evolution of Metabolic Pathways. Elsevier, Oxford, pp. 25–54. Sultana, N., Sarker, S.D., Armstrong, J.A., Wilson, P.G. and Waterman, P.G. (2003) The coumarins of Philotheca sensu lato: distribution and systematic significance. Biochemical Systematics and Ecology 31, 681–691. Tsantrizos, Y.S. and Yang, X.-S. (2000) Macrolide and polyether polyketides: biosynthesis and molecular diversity. In: Romeo, J.T., Ibrahim, R., Varin, L. and De Luca, V. (eds) Evolution of Metabolic Pathways. Elsevier, Oxford, pp. 91–107. Waterman, P.G. (ed.) (1993) Alkaloids and sulphur compounds. In: Dey, P.M. and Harborne, J.B. (eds) Methods in Plant Biochemistry, Vol. 8. Academic Press, London. Waterman, P.G. (1998) Alkaloid chemosystematics. In: Cordell, G.A. (ed.) The Alkaloids: Chemistry and Biology, Vol. 50. Academic Press, Boston, Massachusetts, pp. 537–565. Waterman, P.G. and Grundon, M.F. (eds) (1983) Chemistry and Chemical Taxonomy of the Rutales. Academic Press, London. Waterman, P.G. and Mole, S. (1994) Analysis of Phenolic Plant Metabolites. Blackwell Scientific Publications, Oxford. Wink, M. (1998) Modes of action of alkaloids. In: Roberts, M.F. and Wink, M. (eds) Alkaloids: Biochemistry, Ecology and Medical Applications. Plenum Press, New York, pp. 301–326. Wink, M. and Witte, L. (1983) Evidence for a widespread occurrence for the genes of quinolizidine biosyn- thesis. Induction of alkaloid accumulation in cell suspension cultures of alkaloid ‘free’ species. FEBS Letters 159, 196–200.

13 Ecological importance of species diversity Carl Beierkuhnlein1 and Anke Jentsch2 1Universität Bayreuth, Lehrstuhl fur Biogeografie, D-95440 Bayreuth, Germany; 2UFZ Centre for Environmental Research Leipzig, Conservation Biology and Ecological Modelling, Permoserstr. 15, D-04318 Leipzig, Germany Introduction the most important biodiversity hypotheses, which are still under debate. Concluding, we Understanding the ecological importance of point to emerging challenges related to key biodiversity functions, historical contingency, cross-scale and cross-system research, and the implica- Understanding the ecological importance of tions of spatio-temporal dynamics for the biodiversity for ecosystem functioning and performance of biodiversity under changing ecological services to mankind requires us to environmental conditions. relate the diversity of ecosystem properties to the diversity of species performances in Currently, an extensive and controversial space, in time, in biotic interaction and debate is questioning the effects that are under changing environmental conditions. expected to follow the decline of plant Before discussing ecosystem functioning, we species diversity (Mooney et al., 1996; therefore explore three basic properties of Grime, 1998; Kaiser, 2000; Schmid, 2002). ecological systems related to energy, matter This debate stimulated ecological theory and and information, three fundamental aspects methodology (e.g. Risser, 1995; Wardle et al., of biodiversity related to quantitative, quali- 1997; Lawton, 1999; Loreau, 2000; tative and functional aspects, and knowledge Bednekoff, 2001; Naeem and Wright, 2003). on functional traits of species. We then Initially, opposing standpoints were devel- review emerging theory on the role of biodi- oped concerning functional implications of versity for ecosystem functioning, including species richness (Huston, 1997; Hector et al., functions such as productivity, stability, 1999, 2000b). The progress in biodiversity nutrient retention, resistance against inva- research during recent years is thus a conse- sion, and the temporal performance of com- quence of the engagement of various inter- munities. We further extend our scope to national research groups with differing the benefits and services of biodiversity for approaches and perspectives. Scientists human societies at large and discuss possible began to realize that no general unified implications of losing biodiversity. Finally, mechanism can be found that could be we present prominent experimental meth- applied to every ecosystem (or site), nor are ods, modelling and conceptual approaches individualistic restrictions per se responsible in biodiversity science, thereby reviewing for patterns and processes. Today, para- digms are shifting (Loreau et al., 2001; © CAB International 2005. Plant Diversity and Evolution: Genotypic and 249 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

250 C. Beierkuhnlein and A. Jentsch Naeem et al., 2002). Based on the critical bilities. Organisms have differentiated their analyses of data and experiments, new chal- functional traits and niche occupation dur- lenges and research issues are emerging in ing speciation (e.g. Cody, 1991). Their coex- biodiversity science (Loreau et al., 2001). In istence is a reflection of functional addition, the recent insights contribute to a specialization and niche complementarity. better understanding of the potential effects Although redundancy of functions may of global changes for human society. occur in various species at a certain focus of interest, each species generally performs It is obvious that local and regional biodi- unique mechanisms and functions within an versity is strongly influenced by human land ecological system. Therefore, a correlation use and its alteration in many landscapes between species diversity and functional (Machlis and Forester, 1996). The threat to diversity is probable but is not necessarily a biodiversity during the next century will be causal explanation (Tilman et al., 1997b) caused mainly by changes of land use. At the (Fig. 13.2). global scale, climate change, depositions of nutrients and toxic compounds, and inva- Still, the ecological implications of species sive species will be less important (Sala et al., diversity are more complex. Ecological sys- 2000). This adds the human factor to the tems and species assemblages are influenced complexity of systems as humans act within by stochastic processes. Species combination ecosystems and control many functions and diversity is not deterministic and also directly. Anthropogenic action may influ- not directly connected to a given environ- ence biodiversity as well as key ecosystem ment. Additionally, plant species that con- functions. If we are interested in the effects tribute to the diversity and functioning of of biodiversity loss on ecosystem function- ecosystems differ in many aspects: for exam- ing, direct and indirect effects of human ple in size, longevity and metabolisms. This influences will have to be considered. indicates that the ecological importance of species diversity must be related to specific Properties of ecological systems communities and ecosystems. For mankind, it has been shown early that there is a rela- In principle, there are always three different tionship between biodiversity and the rise of aspects or properties of ecological systems highly developed ancient cultures (Vavilov, that are controlled or maintained by the 1935). For instance, the centres of old cul- assemblage and diversity of organisms: the tures and the origins of many crops are flow and cycling of (i) energy, (ii) matter and closely linked to the global ‘hot spots’ of bio- (iii) information. In addition, storage and diversity (Myers, 1988; Barthlott et al., transformation occurs. Organisms are influ- 1996). The decline of such cultures is very encing these flows, transformations and stor- probably an effect of non-sustainable use of ages in a non-stochastic, directed way. They resources and biodiversity. Kim and Weaver regulate ecological processes and functions. (1994) even predict that the survival of This regulation cannot be predicted on the mankind depends on the preservation of basis of physical laws or chemical processes biodiversity. only. The genetic information of species becomes ecologically effective as regards, for The diversity of biodiversity example, life-history traits, metabolisms and their plasticity under changing environmen- Initiated by Wilson’s (1985) alert on the ‘cri- tal conditions (Fig. 13.1). sis of biodiversity’ and the Rio Conference, intensive research on biodiversity topics As there is a limited range of ecological emerged, followed up by an incredible num- niches in any ecological system, species ber of publications (‘The diversity of publi- diversity is believed to be limited too cations on diversity is overwhelming’, van (Cornell and Lawton, 1992). The diversity der Maarel, 1997). Public and political of coexisting species can probably be under- awareness occupied the theme as well. stood by considering their functional capa-

Ecological importance of species diversity 251 Fig. 13.1. Hypothetical consequences of environmental and land-use changes. The subsequent loss of biodiversity is likely to be followed by functional changes and shifts of ecological complexity. Some of these changes result in a decline in human benefits. Note that there are many unclear connections and unsolved questions between these levels. Looking closer, many research projects are diversity is just one part of biodiversity. Yet, continuing traditional approaches under the it does not inform about abundance, domi- label of biodiversity just to gain funding. nance patterns or equitability. Perhaps more problematic is the lack of the- ory and concepts, which is a source of confu- Generally we can distinguish: (i) qualita- sion and misinterpretation of results. tive variability from (ii) quantitative richness of a community, an ecosystem or an area. In Different opinions and views of biodiver- addition, different degrees of (iii) functional sity research simply reflect the fact that the interactions create varying ecological com- concept of biodiversity summarizes and inte- plexity. With the focus on plants, this means grates various aspects of biotic variability at that phytodiversity integrates the variability different levels of organization (Bowman, between plants, their number and their 1993). Organisms are just one of these lev- functional differences. Most attention is els. Other levels are genes, populations, concentrated on the number of species, communities or ecosystems. Thus, species because this is easy to measure. However,

252 C. Beierkuhnlein and A. Jentsch Fig. 13.2. Four aspects of biotic variability. Biodiversity occurs at all categories (spatial, temporal, phylogenetic and functional). Even if the performance of functional variability or diversity is influenced by spatio-temporal restrictions and reflects genetically fixed traits, it cannot merely be explained or predicted on the basis of one single criterion such as species diversity, but has its own quality. we should keep in mind that taxonomic erally more prominent in animal ecology units such as species are just one possibility (Hawkins and MacMahon, 1989). Another for the classification of plants. In this case, functional perspective in plant ecology is the types are based on phylogenetic related- derived from population biology, where the ness. Other criteria could be applied as well, regeneration of species is seen as a key fac- such as growth form or seasonality, and tor for the maintenance of species diversity then other classes and units result. This in (Grubb, 1977). This approach concentrates turn would influence the number of types only on the reaction of populations to func- to be counted. tional processes. There is no single index or value for all The concept of plant functional types different aspects of biodiversity. Conse- (PFTs) or functional groups deals explicitly quently, effects of biodiversity on ecosystem with functional diversity (Smith et al., 1993; functioning have to be clearly related to the Woodward and Cramer, 1996; Westoby and particular aspects of biodiversity considered Leishman, 1997; Woodward and Kelly, in a study. 1997). This classification approach is based on functional traits (Walker et al., 1999). In Functional traits and types turn, the classification of individual species depends on those functional criteria Based on approaches that concentrate on applied. PFTs are very helpful if global fore- the functional response of species to certain casts of effects of climate change are under- environments (Grime, 1977), different func- taken, because they are both specific and tional groupings have been developed in coarse enough to show global patterns and vegetation science (Körner, 1993). However, processes (Smith et al., 1993; Diaz and guilds and functional groups have been gen- Cabido, 1997). It is not realistic to work at this level with species diversity. The same is

Ecological importance of species diversity 253 true for changes in global biogeochemical important ecophysiological metabolic or cycles and land-use changes. mutualistic properties (C3/C4 grasses, nitro- gen fixers) are not evident in many cases. Functional attributes and traits focus on properties of plant organs or their metabo- Functional traits can be defined not only lism. This approach is more flexible and by the optimum conditions for species clearly related to criteria such as fluxes of responses, but also by the range of their tol- carbon or storage of water. Several func- erance and the shape of their response tional species traits have been used to clas- curves to a particular factor. Indeed, most sify plants in ways that relate either effect or species have plastic responses to the envi- response to the environment (Noble and ronment, and their role in, for example, Slayter, 1980; Pavlovic, 1994). Now, the post-disturbance recovery is a function not functional perspective concentrates on the only of their optima but also of the competi- effects of species, in particular of species tive environment they encounter. For exam- diversity, on ecological functions (Lamont, ple, even shade-tolerant, slow-growing 1995; Grime, 1997; Hector et al., 1999). It is species respond to added light with acceler- evident that the response to changes in the ated growth, but at a slower rate than light- environment is likely to be shifts in the demanding species (Brokaw, 1985; White et functional composition of a community al., 1985). The problem is even more com- (Aiguar et al., 1996). If the ‘importance’ of plex: the species of a particular ecosystem, plant diversity in a changing world is ques- and thus the range of responses in that tioned, the concept of functional types is ecosystem, have functional traits that were obviously successful (Boutin and Keddy, shaped by past exposures to environmental 1993; Chapin, 1993; Golluscio and Sala, processes. Thus, there is a twofold historical 1993; Diaz, 1995; Box, 1996; Chapin et al., contingency in responses of species diversity. 1996; Diaz and Cabido, 1997; Gitay and First, in ecological time, only those species Noble, 1997; Diaz et al., 1998). However, with access to the site can participate in the numerical diversity of functional types recovery (this access can be influenced by (alpha-diversity) depends on the selection of prior disturbance) and, second, in evolution- certain key functions; it differs for the same ary time, species adaptations reflect previous real community with different criteria evolution. Both determine the diversity of applied. It seems very much more concise functional responses within an ecosystem. to concentrate directly on specific functional traits (Leishman and Westoby, 1992; Whittaker’s classification (1972) into Solbrig, 1993). alpha-, beta- and gamma-diversity can only be partly applied to this concept, because it Functional traits are properties of an does not consider functional diversity organism that are considered to be impor- (Hooper et al., 2002) or – in a systemic per- tant according to the response to or to the spective – ecological complexity. Alpha- and effect on the environment. Functional traits gamma-diversity are just an index for the may be reflected in the morphology of plant number of objects (species) in a certain sub- organs (e.g. leaf size, seed structure) or in set. They depend on the scale of observation morphological capabilities (e.g. resprouting, and the number of records. Beta-diversity clonal growth) (Kindscher and Wells, 1995). may be seen as an index for qualitative dif- Other traits or attributes are related to the ferences between objects. However, it is life cycles of plants. Their ‘vital attributes’ mainly applied at the level of communities (Noble and Slayter, 1980) or ‘life-history and then characterizes the resemblance or attributes’ (McIntyre et al., 1995) and the floristic distance or turn-over between sam- quantitative contribution of such strategies ples. It can be applied to identify spatial het- to communities help to predict their erogeneity and temporal trends. New responses to disturbance events. The com- genetic techniques would allow the calcula- position of temporal traits reflects the long- tion of the similarity or dissimilarity (con- term disturbance regime at the community trast) between organisms. However this level (White and Jentsch, 2001). However, approach is as yet uncommon.

254 C. Beierkuhnlein and A. Jentsch Ecosystem Functioning species are primarily the ones affected by disturbances, other species may increase Processes, mechanisms and functions after a disturbance, even if their functional traits are similar to the previously dominant Functional aspects of biotic communities are species. This has been expressed by the characterized as ‘ecosystem processes’ resilience hypothesis (Walker et al., 1999). (Odum, 1993), ‘biogeochemical processes’ Dominant and minor species in the same (Schlesinger, 1991) or ‘ecosystem functions’ functional groups are similar with respect to (Schulze and Mooney, 1993). We would like the contribution to ecosystem function, but to make a distinction between biotic func- they differ in their environmental require- tions and processes. Processes are mecha- ments and tolerances and, thus, in their abil- nisms such as photosynthesis, pollination or ity to respond to disturbances. Dominant nitrogen fixation. Their properties do not and less dominant species switch in abun- depend on the object. In contrast, functions dance under changing environmental condi- are a relation between processes and objects. tions allowing functional stability. Thus, We may find functions of an object accord- species diversity including functional redun- ing to a certain process, for example high dancy is important in ensuring the persis- capability of nitrogen fixation for nutrient tence of ecosystem function under changing retention in soils, or identify a function for environmental conditions and in ensuring an object via the same or another mecha- resilience in response to a disturbance. nism, for example nutrient retention in stor- Moreover, apparently redundant species age organs of plants. may operate on different spatial and tempo- ral scales (Peterson et al., 1998), thereby As already mentioned, there is a differ- reinforcing function across scales. ence between functions that affect an organism and those that are the effect of Both complementarity and redundancy an organism. Some morphological proper- can be mechanisms that contribute to over- ties of an organism may not easily be all ecosystem stability. For example, Marks attributed to one or the other way of func- (1974) showed that fast-growing, early-suc- tional interaction. In plants, for instance, cessional trees are able to take up dissolved the reaction to a given environment nitrogen after a disturbance, thus prevent- (drought) may lead to certain growth. ing nitrogen export to groundwater and These structural properties, however, may streams. Vitousek’s (1984) general theory of also be genetically fixed and the occur- forest nutrient dynamics suggested that rence of a certain species or ecotype will early-successional species immobilize limit- just indicate the competitiveness of certain ing nutrients quickly after a disturbance. functional traits under these site condi- tions. Functions may also result from dif- Ecosystem functioning as a system prop- ferent processes. erty will be the integral of all different processes going on between the members of To ensure the persistence of ecological the community. Some of these functions functions in plant communities in the face of (e.g. carbon cycling) may be relevant to disturbance, functional adaptations of objects (e.g. humans) outside the system species generally underlie two mechanisms (Reich et al., 2001). of ecosystem response: complementarity and redundancy (Loreau and Hector, 2001). Initiated by DiCastri and Younès (1990) First, species have evolved a diverse spec- and then strongly supported by Chapin et al. trum of abilities relative to disturbance. (1992) and Schulze and Mooney (1993), After a particular disturbance, some species functional aspects became a major focus of increase or invade, while others decrease or biodiversity research from the 1990s retreat (Vogl, 1974). Thus, ecosystem onwards (Baskin, 1994; Mooney et al., response is, in part, a result of niche com- 1995a,b; Chapin et al., 1997, 1998; Tilman et plementarity. Second, when dominant al., 1997c, 1998; Schläpfer et al., 1999; Wall, 1999; Loreau et al., 2001; Kinzig et al., 2002; Mooney, 2002; Schmid et al., 2002b). The

Ecological importance of species diversity 255 identification and description of species and Temporal performance of diverse plant species diversity and its loss were the promi- communities nent concerns related to expected global changes (May, 1986, 1988, 1990; Soulé, It is assumed that species-rich communities 1991; Pimm et al., 1995). will have the capability to react to a variety of events and disturbances, ensuring func- The functioning of ecosystems is hard to tions and ongoing dynamics despite distur- tackle scientifically. The perception of the bance. The probability that some members functional diversity or complexity of eco- of a community will be able to cope with logical systems remains unclear and varies extremes increases with species richness. In among authors (Franklin, 1988; Lawton, the face of an increase in extreme events, 1996; Martinez, 1996; Lavorel and which is expected during global climate Richardson, 1999). The connections change (IPCC, 2001), diverse communities between species diversity and ecological that are adapted to an intensive disturbance complexity are also controversial (Wilson, regime might react flexibly to trends and 1992; Lawton, 1994; Naeem et al., 1994; events. With increasing diversity, plant Tilman and Downing, 1994; Tilman et al., species that are able to tolerate extremes are 1996, 1997a; Hooper, 1998; Hooper and likely to occur. If environmental conditions Vitousek, 1998; Symstad et al., 1998), even swing back to former states in the future, when functioning is related to evident such surviving species would be already on parameters of diversity such as spatial vege- the spot. tation structures (van der Maarel, 1986, 1988; Pacala and Deutschmann, 1996), Nevertheless, current biodiversity plant species composition (Hooper and research is facing the dilemma that chang- Vitousek, 1997; Tilman, 1997b) or domi- ing environmental conditions are accelerat- nance patterns (Grime, 1987; Smith and ing, and that anthropogenic pressures on Knapp, 2003). biodiversity are of global extent (Sala et al., 2000). Obviously, the organismic potential to As animals depend directly on vegetation change its current location by large-distance structure and composition, herbivores and migration or fast alteration of life-history other trophic groups have been correlated cycle and growth form will offer survival to plant species diversity (Asteraki et al., mechanisms in the face of global change for 1995; Siemann et al., 1998; Koricheva et al., a certain fraction of species (Higgins et al., 2000). Further on, there are correlations 2003). Still, the slow alteration of distribu- between plant diversity and the diversity tion areas (decades, centuries) will most and functioning of soil bacteria and fungi likely cover only short distances, and evolu- (Spehn et al., 2000b; Stephan et al., 2000). tionary adaptation mechanisms will most There is evidence that plant species diversity likely take many generations (millennia). positively affects key ecosystem processes The expected spatio-temporal dynamics of such as decomposition via its influence on global change clearly exceed such low-speed microbial functioning (Hector et al., 2000a; and short-range developments of most Knops et al., 2001; Mikola et al., 2002). species. Hence, there is a sensitive threshold Another important type of interaction is to the ongoing speed of change. As soon as direct mutualism between plants and fungi. species cannot cope with the speed or the Here as well, hints of positive correlations spatial extent of environmental change, they between plant diversity and fungal diversity are likely to become extinct. A decrease in are found, but methodological constraints local biodiversity is to be expected if the spa- have hindered further insights (van der tio-temporal mechanisms of migration, phe- Heijden and Cornelisson, 2002). Generally, notypic plasticity and dispersal, the linkage between above-ground diversity meta-population dynamics or evolutionary and dynamics and below-ground processes development do not meet the scales of and aspects (Wardle and van der Putten, global change (Jentsch et al., 2002; Jentsch 2002) is one of the important research gaps and Beierkuhnlein, 2003). Consequently, that has to be filled.

256 C. Beierkuhnlein and A. Jentsch ecosystem functioning may be permanently This debate has to be seen as a modern altered with respect to biotic feedback, mate- reflection of the idea of the balance of rial and energy cycles. nature, which has been a general paradigm since the 19th century. Now it has shifted to High species diversity is likely to go along whether and how plant diversity influences with a diversity of ecological rhythms, which ecosystem functioning. can also contribute to stability of ecosystem functioning. For instance, in high mountain Stability includes the persistence of func- ecosystems, a few dominating species reach tions despite disturbance or despite change very long life spans of several hundred to of environmental conditions. Examples are several thousand years. Most of them are persistence of productivity, nutrient reten- trees, dwarf shrubs or clonal grasses such as tion, carbon sequestration, air and water Betula nana, Pinus longaeva or Carex curvula. purification and slope stability, and the Such long-lived, slow-growing species may avoidance of erosion, desertification, and neither be able to react to changing environ- other forms of degradation. The constancy mental conditions, nor die owing to some of species composition and abundance pat- unfavourable decades or centuries. They terns, and finally of biodiversity, then represent ecological inertia in the face of reflects the functional continuity of an altering conditions or competitive pressure ecosystem in the face of disturbance impacts by new species (Jentsch and Beierkuhnlein, (Tilman, 1993; McIntyre et al., 1995). 2003). This can mean both risk and poten- McGrady-Steed et al. (1997) demonstrate the tial for the future of biodiversity. positive role of biodiversity for the control and predictability of ecosystems. Naeem The risk of not being able to cope with (1998) points at the role of redundant changing environmental conditions by adap- species in changing ecosystems for maintain- tation or migration is simply the fate of ing their functioning (‘reliability’). extinction (Ehrlich and Ehrlich, 1981). On the other hand, evolutionary inertia or the Based on the hypotheses from Connell ‘persistent niche’ (Bond and Midgley, 2001) and Orias (1964) about the expected feed- may provide temporal refuges in cyclically back between species diversity and stability, changing environments. The potential of a theoretical consideration of such mecha- enduring novel conditions via long-term nisms developed (e.g. Margalef, 1969; survival is an option for ‘a better future’, in Goodman, 1975). May (1972) pointed early which conditions may become favourable on at the restrictions according to the con- again, although this strategy does not seem nection of the state of a system (diversity) currently adequate. Species with very long and its ongoing processes (stability) and sug- lifespans exhibit genetic stability through gested replacing diversity by complexity. He time. They do not respond to novel condi- postulated that with increasing complexity tions. When trends of alteration return to the variability of whole systems will be lower past conditions, their particular traits may than the variability of species populations. be most successful and even ensure the per- Lehman and Tilman (2000) also found that sistence of these species through cyclic alter- diversity stabilizes the community but desta- ations. Populations of some tree species bilizes individual populations. perform prolific resprouting after being cut or blown down. To date much research has been directed towards the interrelation between stability Diversity and the stability of functions and diversity, because this connection is both theoretically and practically attractive One of the ‘evergreen’ topics of ecology is (McNaughton, 1977, 1978; Thierry, 1982; the relationship between diversity and stabil- Kikkawa, 1986; Frank and McNaughton, ity of ecosystem functions (Tilman et al., 1991; Johnson et al., 1996; Tilman, 1996; 1994, 2002a; Levine and D’Antonio, 1999; Doak et al., 1998; Tilman et al., 1998; Loreau et al., 2001, 2002; Tilman, 2001). Loreau and Behara, 1999; White and Jentsch, 2001). Decreases in population size as a consequence of resource partitioning in

Ecological importance of species diversity 257 diverse stands may lead to higher sensitivity higher stability of diverse stands because of against stochastic fluctuations. their dependence on certain site conditions and nutrient availability. This indicates that There is fundamental significance of in natural ecosystems both diversity and site multi-trophic dynamics for ecosystem effects will have to be considered. Huston processes such as stability, with primary pro- and McBride (2002) also hint at the relative ducers in a key role (Raffaelli et al., 2002). importance of both factors for the control of Increasing the number of species without ecosystems. Wardle et al. (2000) suggest the increasing the food-web linkages within an importance of above-ground functional ecosystem is not likely to increase the stabil- group richness and composition, which may ity (Leigh, 1965). dominate stability effects. The purpose of species diversity for sta- Species in turn are capable of changing bility and maintenance of functions within their own environment. Mutualistic interre- ecosystems has been discussed in a commu- lationships between legumes and rhizobia nity-based approach by Walker et al. (1999). strongly modify the nutrient status of a site This paper proposed that persistence in (Spehn et al., 2002). Such species are indi- ecosystem function under changing envi- rect ecosystem engineers, such as termites or ronmental conditions and resilience against ants (Jones et al., 1994). disturbance are ensured by functional simi- larities among dominant and minor species. Species are idiosyncratic in their response According to the resilience hypothesis, to environmental constraints or disturbance major and minor species switch in abun- regimes. Some species are keystone species dance during times of stress or disturbance, that influence ecosystem dynamics more thus maintaining ecosystem function. than others (Naeem et al., 2002). For Abundant species contribute to ecosystem instance, the fuel provided by a dominant performance at any particular time (and are understorey grass is critical to the fire functionally dissimilar from each other). regime, species diversity and pine regenera- However, minor species contribute to tion in longleaf pine forests in the south- ecosystem resilience during times of stress eastern United States (Christensen, 1981). (Mulder et al., 2001) or disturbance (and are functionally similar to dominant species Diversity and productivity and could increase in abundance to main- tain function if dominant species decline or The effects of biodiversity on the productiv- disappear). Peterson et al. (1998) indicated ity of stands are of crucial importance and that apparently redundant species operate are closely linked to economic perspectives at different scales and thus reinforce func- (Tilman, 1999). Early studies found a posi- tion across scales. It may be shown with the tive correlation (Connel and Orias, 1964; help of model ecosystems, that diversity– Pianka, 1966; MacArthur, 1969). Others stability relationships are likely to occur came to opposing conclusions (Margalef, (Doak et al., 1998). 1969). It depends on the system whether effects of diversity per se or of certain parts The answer to the question of whether of the community such as productive species diversity and stability are related varies with emerge even if an increase in productivity is the community or ecosystem that is dealt theoretically to be expected (Tilman et al., with. On the other hand, stability concepts 1997b). Guo and Berry (1998) could not differ. Grimm and Wissel (1997) identify find clear effects of species number and bio- four primary stability concepts (persistence, mass at different sites. Nevertheless, many resistance, resilience and constancy). studies detect a positive relationship Diversity will influence such different quali- between plant species richness and ecosys- ties specifically. tem processes, especially regarding above- ground primary production, and especially Tilman and Downing (1994) demonstrate in species-poor communities (Schläpfer and a linear relationship between species diver- sity and the recovery of grassland after severe drought. Givnish (1994) doubts the

258 C. Beierkuhnlein and A. Jentsch Schmid, 1999; Schwartz et al., 2000; Spehn ments that are otherwise vulnerable to et al., 2000a; Troumbis and Memtas, 2000; leaching. The greater the productivity, the Bergamini et al., 2001; Hector, 2002; Schmid greater is the differentiation of successional et al., 2002b; Tilman et al., 2002b). Not many roles and the greater the amount of succes- studies are able to separate site effects that sional turnover during assembly (White and are controlling both biodiversity and pro- Jentsch, 2004). This is reflected in changes ductivity. There is also an interrelation in life-history traits: resource use efficiency, between productivity and functional stability longevity and age at sexual maturity (Lehman and Tilman, 2000; Pfisterer and increase, while relative investment in repro- Schmid, 2002). duction decreases. As resources become immobilized in biomass and organic detri- The probability of the occurrence of tus, present diversity creates a filter for fur- highly productive species grows with increas- ther establishment. ing species richness (Aarssen, 1997). Additionally species that promote nutrient In this context, it is remarkable that the availability, such as legumes (Spehn et al., greater the resource supply in a diverse 2002), or influence the ecosystems as ecosys- community, the greater the importance of tem engineers in a positive way are likely to disturbance to increase turnover by removal occur in species-rich stands. In Hooper’s of inhibition (White and Jentsch, 2004). (1998) experiment biomass varied more Whereas, the greater the stress or distur- within certain levels of diversity than between bance, the greater the importance of facilita- them. This also hints at the individual func- tion and mutualism within the species tional importance of certain species. Huston community (Temperton et al., 2004). (1997) criticizes experimental approaches that ignore the possibility of ‘hidden treat- Nutrients, soil and relief ments’ through the ecophysiological differ- ences in species assemblages. Hooper and In the context of global climate change it is Vitousek (1998) find a better use of resources assumed that diverse ecosystems will have in species-rich stands. This reflects the para- better capabilities to adapt to novel condi- digm of higher efficiency of diverse commu- tions and environmental constraints by shift- nities due to niche partitioning. ing dominances (Peters and Lovejoy, 1992; Peters, 1994). Not many approaches are able to sepa- rate such effects from mere biodiversity Following mechanical ground distur- effects. Van Ruijven and Berendse (2003) bances, the mineralization of nutrients find positive effects of plant species richness would lead to nutrient leaching, as demon- on the productivity of communities even in strated for dry acidic grasslands in the low- the absence of legumes. land area of central Europe (Jentsch, 2004). If nutrients become available after distur- In restoration ecology, it is of crucial bance, temporal aspects of species diversity importance to determine both the presence increase in significance. Early colonists are of particular functional traits in plant com- able to store those resources rapidly. When munities and the species diversity as a pool resources become available after distur- of complementary regeneration mechanisms bance, such as in forest blowdowns, coloniz- for community assembly. For instance, diver- ing ability and growth rate are important sity can affect initial productivity correlated and can have a lasting impact on ecosystem with soil resources in several ways: the composition and structure. Rapid establish- greater the diversity of present response ment supports rapid uptake of resources groups to a disturbance, the higher the ini- and stabilization of soil. Such mechanisms tial rates of establishment, growth and pro- have been shown to be important in the ductivity. The greater the abruptness and tropics. However, generally, biodiversity and magnitude of an increase in resources, the its temporal performance may play a deci- greater the initial selection for rapid colo- sive role in nutrient cycling. nization and the higher the initial growth rates, leading to critical uptake of soil ele-

Ecological importance of species diversity 259 Hooper and Vitousek (1998) could not Biodiversity and the invasibility of prove lower leaching of nutrients in diverse communities stands. In contrast, Scherer-Lorenzen (1999) and Scherer-Lorenzen et al. (2003) Biodiversity is not constant in time. At the found lower levels of nitrate in the leachate regional scale and within the temporal scales under grasslands when their species diver- of ecosystems there are fluctuations. sity was high, although plots with legumes Dispersal and succession occur. Species com- showed higher nitrate values and the proba- position may shift. Biodiversity is also influ- bility of the occurrence of legumes increased enced by invasive species and vice versa with species diversity. Even in non-fertilized (Palmer and Maurer, 1997; Prieur-Richard plots without legumes, high concentrations and Lavorel, 2000). Due to the increasing of nitrate occurred due to atmospheric connectivity of isolated habitats by anthro- deposition and mineralization. This was only pogenic vectors, species become introduced the case in species-poor communities. In and thereby extend their former distribu- rich stands, critical levels of nitrate in the tion. Competitive neophytes are initially leachate could not be measured. adding to the flora of a region. As soon as they contribute to the local extinction of sev- On steep slopes, soil stability is an impor- eral less competitive indigenous species, tant property and also a service for the pro- negative effects on species diversity can fol- tection of human settlements and low, especially in biodiversity hot spots infrastructure. Soil stability is highly depen- (Stohlgren et al., 1999). dent on plant cover and rooting patterns. The more diverse the root growth forms, High diversity was found to act as a bar- the less likely it is that extreme events will rier against or at least have a negative influ- promote soil erosion. The loss of diversity ence on ecological invasions and delay them could alter the sensitivity to soil erosion and in some communities (e.g. Tilman, 1997a; slope stability in high mountains (Körner, Crawley et al., 1999; Naeem et al., 2000; 1999). This can also be perceived in terms Prieur-Richard et al., 2000; Kennedy et al., of the insurance hypothesis (Yachi and 2002). Field experiments support the role of Loreau, 1999). diversity in controlling invading plants (Knops et al., 1999; Hector et al., 2001a). Combating desertification and degrada- However, there are also some contraindica- tion is another of the most important inter- tions in dynamic systems with high turnover national activities in the scope of sustainable and short-lived species (Robinson et al., development (UNEP, 1992). The problem is 1995; Palmer and Maurer, 1997). Wardle mainly caused by increasing human popula- (1999, 2001) is sceptical about findings that tions and overexploitation in developing support diversity effects because the control countries. In semi-arid climates with high of invasion can be species specific and then natural variability of drought and rainfall, is related to diversity via the sampling effect. sites could be negatively influenced by species loss. Species-rich plant communities Even if Dukes (2001) did not find an are likely to be able to shift in abundance or effect of species diversity in grassland micro- dominance patterns and to preserve ecosys- cosms on the establishment of alien tem functioning under stress. At higher Centaurea solstitialis, he observed a stronger scales, multi-patch vegetation patterns are suppressed growth of species-poor stands by likely to control the process of desertification this invasive species. This means that biodi- (von Hardenberg et al., 2001). Spatial het- versity did not prevent invasion but affected erogeneity or evenness becomes effective for the stability of previous ecosystem proper- the maintenance of ecosystem functions in ties. On longer timescales, this might pro- terms of promoting or hindering small-scale duce negative feedback. Meiners et al. (2004) reactions to ecosystem changes (Wilsey and stress the fact that species-specific aspects Potvin, 2000; Wilsey and Polley, 2002). and sampling effects are important and However, this broad topic cannot be com- overlay diversity effects. Furthermore, it is pletely covered in this review. important to understand species-specific

260 C. Beierkuhnlein and A. Jentsch interactions and the mechanisms of competi- ber per se and effects of functional diversity tion and mutualism that occur in a given (Prieur-Richard and Lavorel, 2000), the lat- community. ter being more likely to be important (Symstad, 2000). Stohlgren et al. (1999) and van Ruijven et al. (2003) point out that there is a scale Human Threats and Benefits dependence of diversity effects. Such mech- anisms are likely to occur only at the com- Crisis of biodiversity munity level. At larger scales, other factors (e.g. disturbances, heterogeneity of resource The current species-extinction period is availability) are more decisive (see also mainly caused by human impact, and is esti- Levine, 2000; Wardle, 2001). mated to happen at a rate 1000 times greater than the natural rate of extinction However, also at the scale of the commu- (Primack, 1993). Recently, various global nity, there is an influence of short-term distur- change scenarios have been developed that bances that will affect invasibility (Rejmánek, address the effects on biodiversity caused by 1989). The problem is that such disturbances atmospheric warming, altered precipitation and their effects have a short duration. Thus, patterns, land-use changes, increased frag- it is absolutely necessary to look at spatio-tem- mentation, urban expansion and other poral patterns. This is true not only for com- human activities (Sala et al., 2000). Ongoing petition-free space but also for the discussions among natural and social scien- performance of biodiversity that may differ tists (Jentsch et al., 2003) have been further over short distances and time steps (e.g. sea- alerted by the last report of the sons). In some ecological zones, disturbances Intergovernmental Panel on Climate and temporal variability can foster invasions Change (IPCC, 2001), stating an accelerated owing to environmental constraints. In the speed of change of environmental condi- Mediterranean, temporal variability strongly tions. The upcoming IPCC report will controls the diversity of plant communities. emphasize the crucial role of extreme events The co-occurrence of therophytes, geophytes, for driving biodiversity patterns. dwarf-shrubs and bushes simply reflects the fact that there are temporal niches that are This ‘crisis of biodiversity’ as a conse- occupied. Even there, an effect of diversity on quence of human impacts has brought much invasibility can be detected (Lavorel et al., attention and many repercussions in society, 1999). However, it is important to realize that not only because of ethical responsibility and disturbance may promote diversity. In many aesthetic interests. More than that, the fear cases species-rich stands are frequently dis- of losing ecosystem functions and especially turbed. The invasibility then is controlled by those that are of societal importance disturbances but these promote the initial (Ehrlich and Wilson, 1991) characterize such diversity as well. This is the case in floodplains functions as ‘ecological services’ (Daily, 1997; and riparian sites (McIntyre et al., 1988; Daily et al., 1997). This perception is the rea- Planty-Tabacchi et al., 1996). son for political attention and societal aware- ness. Society is afraid that benefits could be The most sensitive phase is the establish- lost that are delivered from nature for free. ment of invaders and thus the existence of This would mean economic constraints as a competition-free safe sites. Such conditions follow-up of species decline (Jentsch et al., can be delivered by disturbances (Burke and 2003). The direct use of natural resources Grime, 1996), and this explains why (plants) or the direct protection that is given dynamic systems are more prone to invasion by them (pure water, preservation of soil, than stable ecosystems. As we have seen for protection against avalanches and land- the Mediterranean vegetation, dynamic slides, etc.) is not the only concern; there is communities are often also rich in species. It also the loss of potential benefits (e.g. phar- will be important to separate effects of diver- maceutical traits; Cragg and Newman, 2002) sity and effects of temporal variability in invasion research. Diversity effects have to be differentiated into effects of species num-

Ecological importance of species diversity 261 that have not yet been identified Meanwhile, the socio-economic value of (Farnsworth, 1988). In this perspective bio- ecosystem services is widely acknowledged diversity per se is regarded as a resource (Costanza et al., 1997). Biodiversity, especially (Plotkin, 1988; Nader and Mateo, 1998). functional biodiversity, is increasingly recog- nized as decisive for maintaining these ser- Biodiversity, however, may also con- vices (Hooper and Vitousek, 1997; Hector et tribute to threats to human health and wel- al., 2001b). Still, it remains a fundamental fare (Dobson, 1995). Vectors may distribute challenge to assign economic and ethical toxic plants as well as diseases. Public aware- attributes to particular species, communities ness concentrates on human pathogenic or to ecological functions, in order to propose microorganisms. Pathogenic microorganisms conservation measures where the obtained and insects that are distributed by trade and benefits exceed the costs necessary for action traffic can affect plants as well. Then species (Jentsch et al., 2003). Nevertheless, there are diversity and genetic variability within popu- some approaches to calculate services and lations may contribute to the regulation of goods derived from ecosystem functioning at outbreaks of disease and the severity of such a monetary scale (Huston, 1993; Buongiorno outbreaks (Mitchell et al., 2002). It is mainly et al., 1994; Pearce and Moran, 1994; because of such assumed capabilities, which Montgomery and Pollack, 1996; Costanza et are not easy to prove, that biodiversity has a al., 1997; Pimentel et al., 1997; Rickleffs, positive image. 1997). As money is an efficient tool to make things comparable and to communicate the Goods, Services and Values value of a subject, it may serve as a powerful argument for the preservation of biodiversity The benefits that society gains from biodi- (Perrings, 1995). It may also contribute to versity are differentiated into use values finding solutions in conflicts between ecology (extractive benefits) and non-use values (e.g. and economy (Gowdy and Daniel, 1995). ethical, non-extractive benefits). Goods and services represent direct or indirect use val- Balmford et al. (2002) estimate, based on ues. Goods are directly related to an eco- conservative assumptions and using a broad nomic profit. Services represent the range of evaluation techniques such as functioning of ecosystems (van Wilgen et al., hedonic pricing, contingent valuation and 1996; Williams et al., 1996). Services of bio- replacement cost methods, that there is a diversity include preservation and renewal tremendous underinvestment in nature of soil fertility, air and water purification, reserves. After their calculations and review, nutrient recycling, carbon uptake, waste the benefit:cost ratio of reserve systems is detoxification and decomposition, modera- around 100:1. Still, the value of biodiversity tion of disturbances and maintenance of is most appreciated in a crisis, and in crises genetic diversity for agricultural improve- its value is extraordinary. ments, as well as control of agricultural pests and human diseases (e.g. Randall, 1994). To determine monetary values of biodi- versity that are not reflected in current mar- It is clear that the value of biodiversity ket prices is an important task from an to mankind has many aspects. The ethical economic perspective (Jentsch et al., 2003). value and the heritage that we must pre- These values include: (i) non-consumptive serve for future generations represent a use values, such as the benefits that species moral duty for society. The aesthetic values richness provides to tourism; (ii) indirect use are evident as well (Heerwagen and Orians, values for ecosystem stability or functions, 1993); some of them may be economically such as the provision of clean water; (iii) important. However, these values may only option values, such as future use in pharma- be identified and evaluated in the course of ceuticals; (iv) existence; and (v) bequest val- an inter-subjective participatory discussion. ues (Perrings, 1995; Costanza et al., 1997; Biologists are no more competent in these Fromm, 2000; Heal, 2000; Dasgupta, 2001). fields than other groups of society. A debate has evolved that elaborates the conditions under which economic valuation

262 C. Beierkuhnlein and A. Jentsch of biodiversity is sensible (Hampicke, 1999; Vitousek and Hooper (1993) initially identi- Seidl and Gowdy, 1999; Nunes and van den fied three major possible relationships Bergh, 2001). Limiting factors include the between biological diversity and ecosystem- non-substitutability of the natural resource, level biogeochemical functions: no effect, the fact that societal preferences for the nat- linear correlation and asymptotic approxi- ural good to be valued cannot be repre- mation of a maximum level. The third was sented by individual preferences, and that regarded as reflecting species redundancy institutional structures significantly influ- by Lawton and Brown (1993). It has to be ence the results of the monetarization. stressed that these concepts focus on ‘species richness’ only, even if graphical representa- As biodiversity conservation typically tions are often simplified to ‘biodiversity’ causes costs at the local level while produc- (Naeem et al., 2002). Functional diversity, ing a global public good, special attention is which is not necessarily correlated to species paid to developing well-defined mechanisms richness, is rarely explicitly addressed for compensating local communities and (Wardle et al., 2000). land users (Ring, 2002). Hence, research on global biodiversity governance further Since those initial theoretical concepts, includes the investigation of incentive struc- many other views and hypotheses were pub- tures and policy instruments (OECD, 1999; lished (review in Schläpfer and Schmid, Barbier, 2000). 1999). Within the group of hypotheses that describes the occurrence of positive effects, We can only briefly mention that, even if perhaps the linear hypothesis (complemen- most authors focus on biodiversity-related tarity hypothesis) and the idiosyncratic economic evaluations on positive aspects of hypothesis represent extreme positions. The diversity and genetic resources and benefits first assumes that each species adds a com- (e.g. ten Kate and Laird, 2002), invasive parable part to ecosystem processes. The non-indigenous species – which may add to reason for such a pattern is seen in the species diversity – are causing enormous assumed complementarity of species costs and hazards (e.g. Pimentel et al., 2000). (Hector, 1998; Hooper, 1998). We have Some invasive aliens such as giant hogweed already pointed out that, because of their (Heracleum mantegazzianum) in Central ecological traits, species perform specifically. Europe (Pysek and Pysek, 1995) may even An equal contribution of each species to cause severe health problems. ecosystem functioning is unlikely. If invasibility is reduced by biodiversity, Nevertheless, there is an indication that a as has been indicated above, then the risks close relationship between particular ecosys- that are related to non-indigenous plants tem processes and species diversity occurs. will be reduced in species-rich stands. This does not imply general mechanisms. However, we have demonstrated that factors Species diversity might be relevant only other than diversity decide whether a com- within a given frame. Most experiments munity is prone to invasion. In the case of have been conducted within rather species- giant hogweed, no effect of species diversity poor communities, and it is not surprising could be demonstrated but the disturbance that within a set of just a few species a corre- regime was of crucial importance. lation of increasing number of species to functioning will be proven. The response of Heuristic Methods and Approaches ecosystem functions to diversity is often con- trolled by restrictions of resource availability. Theoretical considerations and diversity Such restrictions will be effective in one case hypotheses but not in another. If effective, increasing diversity would not influence the function- Theoretical considerations on the impor- ing of the system. tance and mechanisms of species diversity are a challenging field in ecology (Naeem et Biodiversity effects, such as higher bio- al., 2002; Tilman and Lehman, 2002). mass production with increasing species diversity in experimental grasslands (e.g.

Ecological importance of species diversity 263 Hector et al., 1999), can occur per se as an powerful (e.g. productive) species will occur effect of ecological complexity and functional and contribute to the function of interest complementarity (Hector, 1998) (Fig. 13.3). (Huston, 1997). Those species that are lack- However, they can also be a result of proba- ing in species-poor stands can be key species, bility. A critical perspective on biodiversity which play decisive roles in diverse communi- effects, the sampling hypothesis, says that ties. This problem is hard to tackle. However, with increasing numbers of species (or other it seems possible to separate sampling and objects) the probability arises that single biodiversity effects (Loreau, 1998b). Fig. 13.3. According to the complementarity, rivet or linear hypotheses, the loss of species diversity is assumed to lead to a loss of functioning or output (dark arrows) of the community (dashed lines). This example simulates a loss from 11 (a) to 3 (c) species (objects) that belong to three functional groups (symbols). Every species contributes to the functioning of the community.

264 C. Beierkuhnlein and A. Jentsch The rivet hypothesis (Ehrlich and (‘functional analogues’ after Barbault et al., Ehrlich, 1981) says that some species con- 1991). The taxonomic similarity between tribute additionally to the functioning of species (e.g. their assignment to a genus or ecosystems. The hypothetic trajectory along family) does not necessarily hint at func- a diversity gradient is then more stepwise tional resemblance. Nevertheless, closely than linear. However, it is inconsistently related species or representatives of one life interpreted. Some views are close to the form are quite often regarded to be func- complementarity hypothesis rather than the tionally redundant. This is due to the fact keystone hypothesis. that some morphological or ecological traits are restricted to a limited set of genetically A modification of the complementarity or related taxa. linear theory is the redundancy hypothesis (Walker, 1992; Lawton and Brown, 1993; The idiosyncratic hypothesis takes into Gitay et al., 1996). This consideration pre- account that the response of an additional dicts that the cumulative contribution of species depends on the complexity of the species in diverse communities will show an community that is already established. asymptotic distribution. When transformed Unpredictable interactions occur (Fig. 13.6). to log scale, the data become linear. For a There is no clear linear or non-linear trend given diversity, each additional species con- but more or less chaotic behaviour of the tributes or adds in a different way to the system. This cannot be explained by key- functioning of this system. Just recording stone species because here the functioning the number of species therefore ignores the of a species is not regarded as independent. individual response capabilities of species This theory does not state that there is no depending on the number of species that effect, but that the effects are individualistic are already there. In any ecosystem, there and not to be predicted only by the number are limits to the performance of certain of species. functions. There are maximum values or thresholds that can be achieved. In less- The unequal contribution of species to diverse communities, any additional species functioning is reflected by other hypotheses. is important, but after a certain threshold The fundamental difference is that the fol- new species will not add remarkably to lowing hypotheses assume that the contribu- ecosystem functioning (Fig. 13.4). This tion of a species is genetically fixed. Hence, its hypothesis has been criticized from a nature functional performance does not depend on conservation perspective because it implies the diversity of the community. Consequently, that some species are unnecessary and their some species would be more efficient than extinction would not cause negative effects. others under certain site conditions. The importance of redundancy effects If only a few or one single effective depends on the dynamic trend of the com- species occur, they can be regarded as ‘key- munity. The increase or decrease in func- stone’ or ‘key species’. Functioning of the tioning can be influenced by the direction community will more or less exclusively of the development (Naeem et al., 2002). It depend on this species. In most cases, there matters whether species are added or lost. will be more than one species that is strongly Non-linearity and hysteresis is possible and effective. With increasing species diversity even likely to occur. The response of the probability of the occurrence of effective ecosystem functioning will differ for the species increases as well (‘sampling effect’) same level of species diversity (Fig. 13.5). (Fig. 13.7). Another source of redundancy may be rareness. If organisms are rare or afford lit- Hector et al. (2002b) have tested the sam- tle space, they will not interact. Then, it is pling effect hypothesis (Aarsen, 1997; possible that the same ecological niche is Huston, 1997; Tilman et al., 1997b). occupied by different species. They will Although diverse communities are strongly contribute to the species diversity of the influenced by some dominant plant species, community but not increase its complexity it could not be shown that species with high- est biomass in monocultures were also most efficient in mixtures. Yields from mixtures

Ecological importance of species diversity 265 Fig. 13.4. The redundancy hypothesis predicts that there are redundant species within a functional group, which contribute only to species diversity but not remarkably to the functioning of a community (dashed lines) (a, b). Only if complete functional groups are lost (c) will reductions of functionality occur and the output of key functions (dark arrows) decline. Species diversity is reflected by the number of objects and functional diversity (functional groups) by different symbols. were generally higher than the monoculture from stochastic mixtures, ‘overyielding’ is yield of dominant species within these mix- detected. For instance, there are higher tures. Pacala and Tilman (2002) support a common values than might result from shift from the sampling hypothesis to the adding the values for single species derived complementarity hypothesis. from monocultures. In agricultural com- munities, it has also been found that certain If there is a stronger effect of an combinations of species in polycultures had increase in species diversity than expected

266 C. Beierkuhnlein and A. Jentsch Fig. 13.5. The direction of species loss and gain is of species within such a group, the probabil- likely to cause different repercussions at the same ity grows that the functions that are attrib- level of species diversity. Hysteretic processes and uted to this group will be maintained in a non-linear response curves are mainly relevant in changing environment (Chapin et al., 1996). communities with low turnover rates and for species This means that redundancy under certain with high longevity. At the same level of diversity, conditions will deliver the potential to react saturated communities will then differ strongly from to new conditions (Fonseca and Ganade, pioneer stages. Initial mechanisms such as the 2001). The ‘reliability’ of communities selection of functionally relevant species (e.g. in seed increases with species diversity (Naeem and mixtures) will differ from extinction of key species. In Li, 1997; Naeem, 1998). Based on such communities with high turnover rates, both curves thoughts, Yachi and Loreau (1999) have will be rather close. As ecosystem functions are not developed the insurance hypothesis (Fig. directly connected to species diversity but influenced 13.8). more by ecological complexity, the redundancy of species is a matter of the direction of temporal The above-mentioned theories do not processes such as invasion or extinction and of the consider the influences of site conditions time that is available to strengthen functional and especially of resource availability. interactions such as competition. Complementarity is most likely to occur when the participating species are not lim- higher biomass production than expected ited, for example by nutrient availability. At when reducing initial species diversity poor or dry sites, individuals of different (Vandermeer et al., 2002). However, species may exist in low abundances with- Vandermeer et al. point out that overyield- out any interaction. Adding or losing ing does not necessarily imply that inter- species will produce stochastic reactions specific facilitation occurs. It could be within the limited frame of the environ- related to species-specific capabilities in the ment. On the other hand, it can be use of resources. observed that when resource availability is high, for example on fertilized or wet sites, Biodiversity is considered to be a poten- only few specialists will be competitive and tial resource that might become effective in abundant. There, redundancy is common. the future. Rare species of today could play In conclusion, we emphasize that the type more important roles under the expected of response depends very much on the new environments of tomorrow. The differ- availability of resources such as light, nutri- ent abilities of species to tolerate and react, ents or water (Fig. 13.9). to survive and to disperse are some sort of insurance against changing conditions in Experimental Approaches heterogeneous landscapes (Loreau et al., 2003). They could become relevant even Removal experiments when no evident functional diversity within a group of plants can be detected under It is strikingly simple to follow this recent conditions. With increasing numbers approach and to exclude some species from formerly diverse communities in order to simulate the loss of plant species diversity. Most of the removal experiments in the lit- erature are applied to animals and focus on food-web complexity. A critical point about removal experiments with plant communi- ties is the impact on nutrient cycling. This is a problem even if no soil disturbances are affected or if no dead biomass or litter is left. After cutting a plant or destroying it with herbicides, its remaining root biomass

Ecological importance of species diversity 267 Fig. 13.6. The idiosyncratic response hypothesis implies that it is almost impossible to predict the effects of a decline of functional complexity that accompany species loss. Non-linear chaotic responses may occur. An initial decrease in the functioning (dark arrows) of a community (dashed lines) (a, b) can be followed by increases due to changing interactions (e.g. the removal of competitive but slow-growing species) during ongoing losses of species (objects) (c). will be mineralized. This leads to temporar- competition whereas in fact it is a hidden ily enhanced nutrient values in the soil fertilizing effect. Such processes are hard to (Jentsch, 2004). In consequence, these avoid and even harder to quantify. This nutrients will promote the remaining speci- may explain why species-removal experi- mens from other species to be more pro- ments have been widely neglected, whereas ductive. This effect then could be synthetic experiments have had a strong interpreted as a positive signal due to less impact in the scientific community.

268 C. Beierkuhnlein and A. Jentsch Fig. 13.7. The sampling hypothesis points at the lower probability of the occurrence of strongly influential species (represented by larger objects), according to the function of interest (e.g. biomass production) (dark arrows), with decreasing species diversity (a, b). It is difficult to avoid such effects in synthetic/additive experiments. Only some key species contribute substantially to the functioning of the community (dashed lines). If key species representing important functional groups (symbols) are preserved, there will be no negative effects of species loss to functioning (c). Symstad and Tilman (2001) showed, in a unequally distributed across functional types. 5-year removal experiment on abandoned With this approach, traits have been identified agricultural fields, that there is a strong effect that might become important in the course of of the functional groups remaining in the species extinctions. Synthetic experiments community. The ability to occupy the space with all species starting at the same time after and to make use of the resources that have the experiment has been installed are not been required by the former competitors is able to simulate such spatio-temporal aspects.

Ecological importance of species diversity 269 Fig. 13.8. The insurance hypothesis hints at the higher probability of flexible functional responses and adaptations to novel environments in species-rich communities (dashed lines). Shifts in abundance or dominance of species and in their relative contribution to the functioning may compensate for restrictions or the decline of sensitive species (a, b). If a negative threshold of diversity is surpassed, further changes in the environment will not be answered adequately (c). Wootton and Downing (2003) point out could be helpful to find out which effect is that the results of species removal are related to key species (Mills et al., 1993) highly idiosyncratic and therefore impossi- and how biodiversity contributes via com- ble to forecast. They suggest combining plex interactions between species to ecosys- targeted species removals with general tem functioning or buffers environmental diversity manipulation. This approach extremes (Hughes et al., 2002).

270 C. Beierkuhnlein and A. Jentsch Fig. 13.9. Controversial findings from different experiments and field research can probably be explained by underlying effects of resource availability. An increase or loss of species diversity will hardly have any effect (e.g. on biomass production) if the community is strongly limited in resource supply (e.g. nutrients, water and energy). High resource availability on the other hand will support the predominance of a few or one specialized and competitive species (K species). Additional species will not be important. Only under intermediate conditions (for the ecological scale of the community) will complementarity be found and is species diversity likely to be functionally effective. Synthetic experiments Most experiments with plants focus on short-lived species (grasses, forbs) or are car- During the search for interrelationships ried out on artificial substrates to reduce site between biodiversity and ecosystem functions, heterogeneity and noise. The simplification of experiments in simple model ecosystems were the approaches leads to a gap in the validation much supported. The reduction of complex- of gained results versus natural communities. ity in these systems as well as their strongly Critical voices point at many other shortcom- controlled environmental conditions (see also ings and restrictions of such experiments Fukami et al., 2001) would allow us to identify (Grime, 1997; Huston, 1997; Fridley, 2002). effects of changing variables such as species diversity (Schmid et al., 2002a). In addition The heuristic value of experiments is such experiments can be carried out within clear. Functional consequences of biodiver- rather short timeframes. Their results can be sity loss and causal effects can be detected – validated by means of replicates. Further- if just under the restricted conditions of an more, specific experimental designs can be experiment. This helps to support or falsify repeated at different sites (Hector et al., hypotheses that have been developed theo- 2002a,c). This helps to identify generality retically. In nature many factors are interact- behind individual experimental results. ing. It is impossible to separate them and to relate observed phenomena to selected site One of the first influential projects in this conditions directly. field was the ECOTRON microcosm experi- ment (Naeem et al., 1994, 1995, 1996; Many recent experiments and manipula- Hodgson et al., 1998; Lawton et al., 1998; tions that deal with biodiversity effects con- Thompson and Hodgson, 1998). This centrate on plants because they are approach was still very artificial. It was carried non-mobile and in many cases easy to con- out in isolated chambers. However, it pre- trol and to establish. In addition, they repre- pared the way for more natural field experi- sent only one trophic level. Most of these ments. research projects find positive correlations between species diversity of plants and key

Ecological importance of species diversity 271 ecosystem functions (in most cases biomass relations between biodiversity and ecosystem production) (Hector et al., 1999; Schläpfer functions must be related to temporal and and Schmid, 1999; Schwartz et al., 2000; spatial scales (Oksanen, 1996; Rapson et al., Hector, 2002; Schmid et al., 2002b; Tilman et 1997; Bengtsson et al., 2002; Levine et al., al., 2002a,b). 2002). Small-scale effects are not necessarily valid at larger scales (Waide et al., 1999; In grasslands, two comprehensive field Weiher, 1999; Chase and Leibold, 2002). experiments had a strong impact. One experiment featured many replicates on one To apply a functional perspective and to site in Minnesota, USA (Cedar Creek identify the repercussions of changes in Experiment; Tilman et al., 1996) and the species diversity on the complexity of eco- other followed biogeographical gradients systems will be almost impossible in most across Europe (Biodiversity and Ecosystem natural ecosystems. This is why reductionist Processes in Terrestrial Herbaceous models and experiments with defined con- Ecosystems – BIODEPTH; Diemer et al., ditions and environmental interactions 1997; Hector et al., 1999). Grasslands do became prominent during the past decade, have the advantages of being spatially as when forecasts on the effects of species loss well as economically important and they are on key ecosystem functions were discussed. easy to establish in a short time. Experiments have also been installed in for- Modelling approaches est communities (e.g. in Finland and in Germany). These will continue until stable Another promising heuristic approach for conditions have developed, then the results investigating the relationship between diver- will be published. sity and functioning is the application of ecological models (Loreau, 1998a). Models Due to the complexity of natural and allow us to simulate interactions and multi- anthropogenic ecosystems and landscapes, species diversity effects without being prone the monitoring of the loss of diversity is to the restrictions and noise of field investi- rather time consuming and will only deliver gations and experiments. In addition, they good results for selected examples and are not restricted to short-lived species. areas. There, it will be necessary to prove However, most models are still extremely the generality of the results and to identify simple and cannot cope with the reality of the causes of the decline, if it occurs at all. ecological complexity. However, local species diversity may even increase because of new vectors and invad- Doak et al. (1998) found that statistical ing species at the same time as global extinc- averaging would result in greater stability of tions occur, even in the same area, when ecosystem functioning at high levels of rare species are lost. diversity. Tilman et al. (1998) demonstrated that statistical averaging is not a necessary Some important insights into ecological consequence of high diversity, but depends functions of biodiversity and driving factors on the system that is investigated. The ‘port- for the decline of endangered species have folio effect’ may lead to a limited stabiliza- been gained through the recent use of tion of the community due only to statistical experimental designs at the landscape level grounds. The term is derived from econom- (e.g. Caughley and Gunn, 1996). Results ics, where experience shows that a diversi- from experimental manipulations predict fied portfolio will be less endangered by that high biodiversity will enhance ecosys- stochastic market processes. tem responses to elevated carbon dioxide and nitrogen deposition (Reich et al., 2001; Lehman and Tilman (2000) compared Catovsky et al., 2002). Reich et al. (2001) different types of ecological models (mecha- show that current trait-based functional clas- nistic models, phenomenological models sifications alone might not be sufficient for and statistical models). They showed that, understanding ecosystem responses to ele- even if the models perform differently, the vated carbon dioxide. general reactions of the simulated systems Nevertheless, experimentally proven cor-

272 C. Beierkuhnlein and A. Jentsch according to the stability of the communities constraints such as soils, relief and climate are are comparable: the variability of the entire strongly influencing ecosystem processes. The communities decreased and the variability of direct effect of such mechanisms will be much the contributing populations increased. more important in many cases than biodiver- sity effects. In addition, direct human impact Yachi and Loreau (1999) formulated the (e.g. pollution) may have consequences for insurance hypothesis based on biodiversity biodiversity and ecosystem functioning. Cause models. Their model proves that the main- and effect of changes in ecosystem functioning tenance of key ecosystem functions as a reac- are then difficult to separate. tion to temporal variability of the environment is more likely to occur in Nevertheless, plant species diversity plays species-rich stands. Other ecological models a significant role for the control of ecosystem that are dealing with insurance and related processes and overall functioning. In some research questions have produced compara- cases, the effects will be related to comple- ble results (Fonseca and Ganade, 2001; mentarity of functional traits of species, in Petchey and Gaston, 2002). others just to the occurrence of key species (e.g. productive ones or ecosystem engi- Generally, ecological modelling can be neers). Today, the impact of biodiversity on used as a tool for integrating scientific ecosystem functioning can be neither pre- results from various experimental and dicted nor neglected. observational analyses as well as scenarios of changing environmental or socio-economic Species are not similar. The historical and conditions. Spatially explicit grid-based evolutionary background of each species models (De Angelis and Gross, 1992; may have a strong influence on the per- Grimm, 1999) show that spatio-temporal formance of entire ecosystems. Traits control correlations are a key to understanding sys- the reaction pattern and metabolic capabili- tem dynamics, their vulnerability or ties of plant species. It is not possible to con- resilience (stability). There is growing evi- clude general principles simply from the dence that such correlations are the cur- response of a given set of species. Mooney rency to understand not only (2002) points at the fact that there is no spatio-temporal dynamics of ecological sys- simple solution to the controversial stand- tems, but also general mechanisms of biodi- points of whether the number of species or versity and species-specific functions and the variability of functional traits determines traits (Wiegand et al., 1999). ecological functioning. Recently developed ecological–economic From a methodological point of view, it is models are promising techniques for the absolutely necessary to link controlled but integration of social and natural sciences. artificial experiments not only with models They are pioneering approaches for eco- but also with standardized monitoring tech- nomic assessments of different ecological niques. It is promising that the communica- management options (e.g. Frank and Ring, tion and exchange of results gained with 1999; Johst et al., 2002). For instance, the different approaches will contribute to a bet- modelling approach establishes the relation- ter understanding of ecological mechanisms. ship between economic parameters of dis- In this field of cross-cutting research, enor- turbance-management alternatives and the mous gaps still have to be filled. ecological effects on biodiversity properties. Another challenge is the transfer or the Conclusion validation of results across systems and com- munities. What has been found for the rela- Looking at certain key functions that are tionship between biodiversity and thought to be important, we do have to keep productivity in Central European grasslands in mind that primary ecological factors such (Minns et al., 2001; Hector, 2002) may be as water and nutrient cycling, energy in- and true for North American grasslands (Tilman output and secondary or integrated abiotic et al., 2001, 2002a,b) but will be hard to apply to deciduous forests or even subtropi- cal or tropical ecosystems.

Ecological importance of species diversity 273 Most of the diverse and threatened Biodiversity is a complex resource that is ecosystems of the world are poorly produc- hard to define and to analyse with respect to tive (e.g. the South African fynbos; Davis et its functional effects. However, there is al., 1994). Other mechanisms and functional strong support for the idea that it con- interrelations will be important in these tributes to the maintenance of ecosystem communities and may be reduced or modi- functioning, which is fundamentally impor- fied by biodiversity loss. The key functions tant for human beings. Some of the poten- (e.g. inflammability and proliferation of fire tial uses of biodiversity have not yet been as a key disturbance for the maintenance of discovered because of the large number of diversity) are largely to be identified. On the unknown species and our limited knowledge other hand, rather species-poor ecosystems of the functional traits of plant species at the such as mangroves might suffer severe func- global scale. tional restrictions with plant species losses (Field et al., 1998). Implicitly, the loss of biodiversity has been regarded as an indication of the loss of qual- The ecological importance of biodiversity ity of life since the publication of Rachel can be subdivided into aspects that are rele- Carson’s book in 1962. In more recent vant for ecosystem functioning and others decades this development has been perceived that are, in addition, important to human to be negative. However, it has actually society. Some aspects are exclusively relevant speeded up in recent years. Short-term eco- to humans (aesthetical and ethical values), nomic interests are more prominent and the but these have to be discussed at a broad survival of growing human populations in societal level. marginal habitats has to be ensured. Obviously, those socio-economic forces are The paradox of depending on biodiver- powerful drivers for the loss of biodiversity. sity and threatening it at the same time is On the other hand, this ongoing loss is very one of the phenomena in complex human likely to be followed up by violent negative societies that are difficult to cope with. The feedback. Profits from the use of the global human contribution to the processes that stock of biotic resources and ecosystems could maintain and threaten biodiversity works at be endangered in the future. Then, short- different scales. The regional diversity in term individual economic gains would be fol- Europe is to a major part dependent on lowed by long-term societal economic losses. anthropogenic disturbances and structures. If land use was stopped, biodiversity would Natural scientists tend to ignore norma- be lost. Some species have even evolved with tive social or economic values. In the case of close dependence on land-use techniques biodiversity it would be foolish not to coop- and crops. At the global scale, land-use erate with socio-economic scientists in order change today is the major driver for the to both identify the driving forces of extinc- irreversible loss of genetic variability (Sala et tions and forecast and evaluate their effects. al., 2000). This is due to the speed of transi- This chapter concentrates on the ecological tions and to the technical and chemical part of the problem. Economists have devel- intensity that is applied to fulfil social and oped methods to scrutinize the value that is economic requirements. given to natural subjects by people. One method is to ask for the ‘willingness to pay’. The same is true for the benefits that can However, this willingness is strongly influ- be derived from biodiversity. Services at one enced by knowledge, and knowledge on bio- spatial or temporal scale may be accompa- diversity is still fragmentary. nied by a non-sustainable use of such bene- fits. The awareness of the risk of economic Ecological complexity, meaning the func- restrictions in connection with the loss of tional interactions between biota and their biodiversity might strongly support action to abiotic environment, can be seen as the most slow down or even stop this development. important aspect of biodiversity for human society as it controls the services and goods Sustainable, long-term use and develop- that can be derived from ecosystems. The ment is only possible if crucial ecological biota does not exist in isolation from abiotic compartments and objects are maintained.

274 C. Beierkuhnlein and A. Jentsch site conditions (extrinsic factors) but influ- interactions between species under specific ences them and is influenced by them. This site conditions and disturbance regimes is crucially important for basic ecological resulting in comparable assemblages and research as well. The functional interactions structures. Such structures and their between single elements of ecological sys- resource character from the human perspec- tems such as organisms or soil or water bod- tive are nothing less than the effect of eco- ies are still not completely understood. The logical complexity. Thus, only a better web of interactions and reactions is woven at understanding of ecological complexity will different levels of time (from osmotic help us to save and manage biotic resources responses to evolutionary patterns) and and functioning. Former experiences and space (from cells to landscapes). Systemic knowledge have to be adapted to novel cir- features emerge from the non-stochastic cumstances in the face of climate change. References Aarssen, L.W. (1997) High productivity in grassland ecosystems: effected by species diversity or productive species. Oikos 80, 183–184. Aiguar, M.R., Paruelo, J.M., Sala, O.E. and Lauenroth, W.K. (1996) Ecosystem responses to changes in plant functional type composition: an example from Patagonian steppe. Journal of Vegetation Science 7, 381–390. Asteraki, E.J., Hanks, C.B. and Clements, R.O. (1995) The influence of different types of grassland field mar- gin on carabid beetle (Coleoptera, Carabidae) communities. Agriculture, Ecosystems and Environment 54, 195–202. Balmford, A., Bruner, A., Cooper, P., Costanza, R., Farber, S., Green, R.S., Jenkins, M., Jefferiss, P., Jessamy, V., Madden, J., Munro, K., Myers, N., Naeem, S., Paavola, J., Rayment, M., Rosendo, S., Roughgarden, J., Trumper, K. and Turner, R.K. (2002) Economic reasons for conserving wild nature. Science 297, 950–953. Barbault, R., Colwell, R.K., Dias, B., Hawksworth, D.L., Huston, M., Laserre, P., Stone, D. and Younes, T. (1991) Conceptual framework and research issues for species diversity at the community level. In: Solbrig, O.T. (ed.) From Genes to Ecosystems: a Research Agenda for Biodiversity. IUBS, Cambridge, Massachusetts, pp. 37–71. Barbier, B.E. (2000) How to allocate biodiversity internationally? In: Siebert, H. (ed.) The Economics of International Environmental Problems. Mohr, Tübingen, pp. 79–106. Barthlott, W., Lauer, W. and Placke, A. (1996) Global distribution of species diversity: towards a world map of phytodiversity. Erdkunde 50, 317–327. Baskin, Y. (1994) Ecosystem function of biodiversity. BioScience 44, 657–660. Bednekoff, P.A. (2001) Contributions – Commentary – A baseline theory of biodiversity and ecosystem func- tion. Bulletin of the Ecological Society of America 82, 205. Bengtsson, J., Engelhardt, K., Giller, P., Hobbie, S., Lawrence, D., Levine, J., Vilà, M. and Wolters, V. (2002) Slippin’ and slidin’ between the scales: the scaling components of biodiversity–ecosystem functioning relations. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 209–220. Bergamini, A., Pauli, P., Peintinger, M. and Schmid, B. (2001) Relationships between productivity, number of shoots, and number of species in bryophytes and vascular plants. Journal of Ecology 89, 920–929. Bond, W.J. and Midgley, J.J. (2001) Ecology of sprouting in woody plants: the persistence niche. Trends in Ecology and Evolution 16, 45–51. Boutin, C. and Keddy, P.A. (1993) A functional classification of wetland plants. Journal of Vegetation Science 4, 591–600. Bowman, D.M.J.S. (1993) Biodiversity: much more than biological inventory. Biodiversity Letters 1, 163. Box, E.O. (1996) Plant functional types and climate at the global scale. Journal of Vegetation Science 7, 309–320. Brokaw, N.V.L. (1985) Gap-phase regeneration in a tropical forest. Ecology 66, 682–687. Buongiorno, J., Dahir, S., Lu, H.-C. and Lin, C.-R. (1994) Tree size diversity and economic returns in uneven-aged forest stands. Forest Science 40, 83–103.

Ecological importance of species diversity 275 Burke, M.J.W. and Grime, J.P. (1996) An experimental study of plant community invasibility. Ecology 77, 776–790. Carson, R. (1962) Silent Spring. Houghton Mifflin, Boston, Massachusetts. Catovsky, S., Bradford, M.A. and Hector, A. (2002) Biodiversity and ecosystem productivity: implications for carbon storage. Oikos 97, 443–448. Caughley, G. and Gunn, A. (1996) Conservation Biology in Theory and Practice. Blackwell Science, Oxford. Chapin, F.S. (1993) Functional role of growth forms in ecosystems and global processes. In: Ehleringer, J.R. and Field, C.B. (eds) Scaling Physiological Processes: Leaf to Globe. Academic Press, San Diego, California, pp. 287–312. Chapin, F.S., Schulze, E.-D. and Mooney, H.A. (1992) Biodiversity and ecosystem processes. Trends in Ecology and Evolution 7, 107–108. Chapin, F.S., Bret-Harte, M.S., Hobbie, S. and Zhong, H. (1996) Plant functional types as predictors of the transient response of arctic vegetation to global change. Journal of Vegetation Science 7, 347–357. Chapin, F.S., Walker, B.H., Hobbs, R.J., Hooper, D.U., Lawton, J.H., Sala, O.E. and Tilman, D. (1997) Biotic control over the functioning of ecosystems. Science 277, 500–504. Chapin, F.S., Sala, O.E., Burke, I.C., Grime, J.P., Hooper, D.U., Lauenroth, W.K., Lombard, A., Mooney, H.A., Mosier, A.R., Naeem, S., Pacala, S.W., Roy, J., Steffen, W.L. and Tilman, D. (1998) Ecosystem consequences of changing biodiversity. BioScience 48, 45–52. Chase, J.M. and Leibold, M.A. (2002) Spatial scale dictates the productivity–diversity relationship. Nature 415, 427–430. Christensen, N.L. (1981) Fire regimes in Southeastern ecosystems. In: Mooney, H.A., Bonnicksen, J.M., Christensen, N.L. and Reiners, W.A. (eds) Fire Regimes and Ecosystem Properties. USDA Forest Service general technical report WO-26. United States Department of Agriculture, Washington, DC, pp. 112–136. Cody, M.L. (1991) Niche theory and plant growth form. Vegetatio 97, 39–55. Connell, J.H. and Orias, E. (1964) The ecological regulation of species diversity. American Naturalist 98, 399–414. Cornell, H.V. and Lawton, J.H. (1992) Species interactions, local and regional processes and limits to the richness of ecological communities: a theoretical perspective. Journal of Animal Ecology 61, 1–12. Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P. and van den Belt, M. (1997) The value of the world’s ecosys- tem services and natural capital. Nature 387, 253–260. Cragg, G.M. and Newman, D.J. (2002) Chemical diversity: a function of biodiversity. Trends in Pharmacological Sciences 23, 404–405. Crawley, M.J., Brown, S.L., Heard, M.S. and Edwards, G.R. (1999) Invasion-resistance in experimental grass- land communities: species richness or species identity? Ecology Letters 2, 140–148. Daily, G.C. (ed.) (1997) Nature’s Services. Island Press, Washington, DC. Daily, G.C., Alexander, S., Ehrlich, P.R., Goulder, L., Lubchenco, J., Matson, P.A., Mooney, H.A., Postel, S., Schneider, S.H., Tilman, D. and Woodwell, G.M. (1997) Ecosystem services: benefits supplied to human societies by natural ecosystems. Issues in Ecology 1, 1–16. Dasgupta, P. (2001) Economic value of biodiversity, overview. In: Levin, S.A. (ed.) Encyclopedia of Biodiversity, Vol. 2. Academic Press, San Diego, California, pp. 291–303. Davis, G.W., Midgley, G.F. and Hoffman, M.T. (1994) Linking biodiversity to ecosystem function – a chal- lenge to Fynbos ecology. South African Journal of Science 90, 319–321. De Angelis, D.L. and Gross, L.J. (1992) Individual-based Models and Approaches in Ecology. Chapman & Hall, London. Diaz, S. (1995) Elevated CO2 responsiveness, interactions at the community level, and plant functional types. Journal of Biogeography 22, 289–295. Diaz, S. and Cabido, M. (1997) Plant functional types and ecosystem function in relation to global change: a multiscale approach. Journal of Vegetation Science 8, 463–474. Diaz, S., Cabido, M. and Casanoves, F. (1998) Plant functional traits and environmental filters at a regional scale. Journal of Vegetation Science 9, 113–122. DiCastri, F. and Younès, T. (1990) Ecosystem function of biological diversity. Biology International Special Issue 22, 1–20. Diemer, M., Joshi, J., Schmid, B., Körner, C. and Spehn, E. (1997) An experimental protocol to assess the effects of plant diversity on ecosystem functioning utilised by a European research network. Bulletin des Geobotanischen Instituts ETH Zürich 63, 95–107.

276 C. Beierkuhnlein and A. Jentsch Doak, D.F., Bigger, D., Harding, E.K., Marvier, M.A., O’Malley, R.E. and Thomson, D. (1998) The statistical inevitability of stability–diversity relationships in community ecology. American Naturalist 151, 264–276. Dobson, A. (1995) Biodiversity and human health. Trends in Ecology and Evolution 10, 390–391. Dukes, J.S. (2001) Biodiversity and invasibility in grassland microcosms. Oecologia 126, 563–568. Ehrlich, P.R. and Ehrlich, A.H. (1981) Extinction: the Causes and Consequences of the Disappearance of Species. Random House, New York. Ehrlich, P.R. and Wilson, E.O. (1991) Biodiversity studies: science and policy. Science 253, 758–762. Farnsworth, N.R. (1988) Screening plants for new medicines. In: Wilson, E.O. and Peter, F.M. (eds) BioDiversity. National Academy Press, Washington, DC, pp. 83–97. Field, C.B., Osborn, J.G., Hoffmann, L.L., Polsenberg, J.F., Ackerly, D.D., Berry, J.A., Bjorkman, O., Held, Z., Matson, P.A. and Mooney, H.A. (1998) Mangrove biodiversity and ecosystem function. Global Ecology and Biogeography Letters 7, 3–14. Fonseca, C. and Ganade, G. (2001) Species functional redundancy, random extinctions and the stability of ecosystems. Journal of Ecology 89, 118–125. Frank, D.A. and McNaughton, S.J. (1991) Stability increases with diversity in plant communities: empirical evidence from the 1988 Yellowstone drought. Oikos 62, 360–362. Frank, K. and Ring, I. (1999) Model-based criteria for the effectiveness of conservation strategies – an evalu- ation of incentive programmes in Saxony, Germany. In: Ring, I., Klauer, B., Wätzold, F. and Månsson, B.Å. (eds) Regional Sustainability. Applied Ecological Economics Bridging the Gap Between Natural and Social Sciences. Physica-Verlag, Heidelberg, pp. 91–106. Franklin, J.F. (1988) Structural and functional diversity in temperate forests. In: Wilson, E.O. and Peter, F.M. (eds) BioDiversity. National Academy Press, Washington, DC. Fridley, J.D. (2002) Resource availability dominates and alters the relationship between species diversity and ecosystem productivity in experimental plant communities. Oecologia 132, 271–277. Fromm, O. (2000) Ecological structure and functions of biodiversity as elements of its total economic value. Environmental and Resource Economics 16, 303–328. Fukami, T., Naeem, S. and Wardle, D. (2001) On similarity among local communities in biodiversity experi- ments. Oikos 95, 340–348. Gitay, H. and Noble, I.R. (1997) What are plant functional types and how should we seek them? In: Smith, T.M., Shugart, H.H. and Woodward, F.I. (eds) Plant Functional Types: Their Relevance to Ecosystem Properties and Global Change. Cambridge University Press, Cambridge, pp. 3–19. Gitay, H., Wilson, J.B. and Lee, W.G. (1996) Species redundancy: a redundant concept? Journal of Ecology 84, 121–124. Givnish, T.J. (1994) Does diversity beget stability? Nature 371, 113–114. Golluscio, R.A. and Sala, O.E. (1993) Plant functional types and ecological strategies in Patagonian forbs. Journal of Vegetation Science 4, 839–846. Goodman, D. (1975) The theory of diversity–stability relationships in ecology. The Quarterly Review of Biology 50, 237–266. Gowdy, J.M. and Daniel, C.N. (1995) One world, one experiment: addressing the biodiversity – economics conflict. Ecological Economics 15, 181–192. Grime, J.P. (1977) Evidence for the existence of three primary strategies in plants and its relevance to eco- logical and evolutionary theory. American Naturalist 111, 1169–1194. Grime, J.P. (1987) Dominant and subordinate components of plant communities: implications for succes- sion, stability and diversity. In: Gray, A.J., Edwards, P. and Crawley, M. (eds) Colonisation, Succession and Stability. Blackwell, Oxford, pp. 413–428. Grime, J.P. (1997) Biodiversity and ecosystem function: the debate deepens. Science 277, 1260–1261. Grime, J.P. (1998) Benefits of plant diversity to ecosystems: immediate, filter and founder effects. Journal of Ecology 86, 902–910. Grimm, V. (1999) Ten years of individual-based modelling in ecology: what have we learned, and what could we learn in the future? Ecological Modelling 115, 129–148. Grimm, V. and Wissel, C. (1997) Babel, or the ecological stability discussions: an inventory and analysis of terminology and a guide for avoiding confusion. Oecologia 109, 323–334. Grubb, P.J. (1977) The maintenance of species richness in plant communities: the importance of the regen- eration niche. Biological Reviews of the Cambridge Philosophical Society 52, 107–145. Guo, Q. and Berry, W.L. (1998) Species richness and biomass: dissection of the hump-shaped relationships. Ecology 79, 2555–2559.

Ecological importance of species diversity 277 Hampicke, U. (1999) The limits to economic valuation of biodiversity. International Journal of Social Economics 26, 158–173. Hawkins, C.P. and MacMahon, J.A. (1989) Guilds: the multiple meanings of a concept. Annual Review of Entomology 34, 423–451. Heal, G. (2000) Nature and the Marketplace: Capturing the Value of Ecosystem Services. Island Press, Washington, DC. Hector, A. (1998) The effect of diversity on productivity: detecting the role of species complementarity. Oikos 82, 597–599. Hector, A. (2002) Biodiversity and the functioning of grassland ecosystems: multi-site comparisons. In: Kinzig, A.P., Pacala, S.W. and Tilman, D. (eds) The Functional Consequences of Biodiversity. Princeton University Press, Princeton, New Jersey, pp. 71–95. Hector, A., Schmid, B., Beierkuhnlein, C., Caldeira, M.C., Diemer, M., Dimitrakopoulos, P.G., Finn, J., Freitas, H., Giller, P.S., Good, J., Harris, R., Högberg, P., Huss-Danell, K., Joshi, J., Jumpponen, A., Körner, C., Leadley, P.W., Loreau, M., Minns, A., Mulder, C.P.H., O’Donovan, G., Otway, S.J., Pereira, J.S., Prinz, A., Read, D.J., Scherer-Lorenzen, M., Schulze, E.-D., Siamantziouras, A.-S.D., Spehn, E., Terry, A.C., Troumbis, A.Y., Woodward, F.I., Yachi, S. and Lawton, J.H. (1999) Plant diversity and pro- ductivity of European grasslands. Science 286, 1123–1127. Hector, A., Beale, A., Minns, A., Otway, S. and Lawton, J.H. (2000a) Consequences of loss of plant diversity for litter decomposition: mechanisms of litter quality and microenvironment. Oikos 90, 357–371. Hector, A., Schmid, B., Beierkuhnlein, C., Caldeira, M.C., Diemer, M., Dimitrakopoulos, P.G., Finn, J., Freitas, H., Giller, P.S., Good, J., Harris, R., Högberg, P., Huss-Danell, K., Joshi, J., Jumpponen, A., Körner, C., Leadley, P.W., Loreau, M., Minns, A., Mulder, C.P.H., O’Donovan, G., Otway, S.J., Pereira, J.S., Prinz, A., Read, D.J., Scherer-Lorenzen, M., Schulze, E.-D., Siamantziouras, A.-S.D., Spehn, E., Terry, A.C., Troumbis, A.Y., Woodward, F.I., Yachi, S. and Lawton, J.H. (2000b) No consistent effect of plant diversity on productivity. Science 289, 1255. Hector, A., Dobson, K., Minns, A., Bazeley-White, E. and Lawton, J.H. (2001a) Community diversity and invasion resistance: an experimental test in a grassland ecosystem and a review of comparable studies. Ecological Research 16, 819–831. Hector, A., Joshi, J., Lawler, S., Spehn, E.M. and Wilby, A. (2001b) Conservation implications of the link between biodiversity and ecosystem functioning. Oecologia 129, 624–628. Hector, A., Schmid, B., Beierkuhnlein, C., Caldeira, M.C., Diemer, M., Dimitrakopoulos, P.G., Finn, J., Freitas, H., Giller, P.S., Good, J., Harris, R., Högberg, P., Huss-Danell, K., Joshi, J., Jumpponen, A., Körner, C., Leadley, P.W., Loreau, M., Minns, A., Mulder, C.P.H., O’Donovan, G., Otway, S.J., Pereira, J.S., Prinz, A., Read, D.J., Scherer-Lorenzen, M., Schulze, E.-D., Siamantziouras, A.-S.D., Spehn, E.M., Terry, A.C., Troumbis, A.Y., Woodward, F.I., Yachi, S. and Lawton, J.H. (2002a) Biodiversity and the functioning of grassland ecosystems: multisite studies. In: Kinzig, A.P., Pacala, S.W. and Tilman, D. (eds) Functional Consequences of Biodiversity: Experimental Progress and Theoretical Extensions. Princeton University Press, Princeton, New Jersey, pp. 71–95. Hector, A., Bazeley-White, E., Loreau, M., Otway, S. and Schmid, B. (2002b) Overyielding in grassland communities: testing the sampling effect hypothesis with replicated biodiversity experiments. Ecology Letters 5, 502–511. Hector, A., Loreau, M., Schmid, B., Caldeira, M.C., Diemer, M., Dimitrakopoulos, P.G., Finn, J., Freitas, H., Giller, P.S., Good, J., Harris, R., Högberg, P., Huss-Danell, K., Joshi, J., Jumpponen, A., Körner, C., Leadley, P.W., Minns, A., Mulder, C.P.H., O’Donovan, G., Otway, S.J., Pereira, J.S., Prinz, A., Read, D.J., Scherer-Lorenzen, M., Schulze, E.-D., Siamantziouras, A.-S.D., Spehn, E., Terry, A.C., Troumbis, A.Y., Woodward, F.I., Yachi, S. and Lawton, J.H. (2002c) Biodiversity manipulation experiments: studies replicated at multiple sites. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 36–46. Heerwagen, J.H. and Orians, G.H. (1993) Humans, habitats, and aesthetics. In: Kellert, S.R. and Wilson, E.O. (eds) The Biophilia Hypothesis. Island Press, Washington, DC, pp. 138–172. Higgins, S.I., Clark, J.S., Nathan, R., Hovestadt, T., Schurr, F., Fragosoa, J.M.V., Aguiar, M.R., Ribbens, E. and Lavorel, S. (2003) Forecasting plant migration rates: managing uncertainty for risk assessment. Journal of Ecology 91, 341–347. Hodgson, J.G., Thompson, K., Wilson, P.J. and Bogaard, A. (1998) Does biodiversity determine ecosystem function? The Ecotron experiment reconsidered. Functional Ecology 12, 843–848. Hooper, D.U. (1998) The role of complementarity and competition in ecosystem responses to variation in plant diversity. Ecology 79, 704–719.

278 C. Beierkuhnlein and A. Jentsch Hooper, D.U. and Vitousek, P.M. (1997) The effect of plant composition and diversity on ecosystem processes. Science 277, 1302–1305. Hooper, D.U. and Vitousek, P.M. (1998) Effects of plant composition and diversity on nutrient cycling. Ecological Monographs 68, 121–149. Hooper, D.U., Solan, M., Symstad, A., Díaz, S., Gessner, M.O., Buchmann, N., Degrange, V., Grime, P., Hulot, F., Mermillod-Blondin, F., Roy, J., Spehn, E. and van Peer, L. (2002) Species diversity, functional diversity, and ecosystem functioning. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 195–208. Hughes, J.B., Ives, R. and Norberg, J. (2002) Do species interactions buffer environmental variation (in the- ory)? In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 92–101. Huston, M.A. (1993) Biological diversity, soils, and economics. Science 262, 1676–1680. Huston, M.A. (1997) Hidden treatments in ecological experiments: re-evaluating the ecosystem function of biodiversity. Oecologia 110, 449–460. Huston, M.A. and McBride, A.C. (2002) Evaluating the relative strengths of biotic versus abiotic controls on ecosystem processes. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 47–60. IPCC (Intergovernmental Panel on Climate Change) (2001) Climate Change 1999: the Scientific Basis. Cambridge University Press, Cambridge. Jentsch, A. (2004) Disturbance driven vegetation dynamics. Concepts from biogeography to community ecology, and experimental evidence from dry acidic grasslands in central Europe. Dissertationes Botanicae 384, 1–218. Jentsch, A. and Beierkuhnlein, C. (2003) Global climate change and local disturbance regimes as interacting drivers for shifting altitudinal vegetation patterns. Erdkunde 57, 216–231. Jentsch, A., Beierkuhnlein, C. and White, P. (2002) Scale, the dynamic stability of forest ecosystems, and the persistence of biodiversity. Silva Fennica 36, 393–400. Jentsch, A., Wittmer, H., Jax, K., Ring, I. and Henle, K. (2003) Biodiversity – emerging issues for linking nat- ural and social sciences. Gaia 2, 121–128. Johnson, K.H., Vogt, K.A., Clark, H.J., Schmitz, O.J. and Vogt, D.J. (1996) Biodiversity and the productivity and stability of ecosystems. Trends in Ecology and Evolution 11, 372–377. Johst, K., Drechsler, M. and Wätzold, F. (2002) An ecologic–economic modelling procedure to design effective and efficient compensation payments for the protection of species. Ecological Economics 41, 37–49. Jones, C.G., Lawton, J.H. and Shachak, M. (1994) Organisms as ecosystem engineers. Oikos 69, 373–386. Kaiser, C. (2000) Rift over biodiversity divides ecologists. Science 289, 1282–1283. Kennedy, T.A., Naeem, S., Howe, K.M., Knops, J.M.H., Tilman, D. and Reich, P. (2002) Biodiversity as a bar- rier to ecological invasion. Nature 417, 636–638. Kikkawa, J. (1986) Complexity, diversity, and stability. In: Kikkawa, J. and Anderson, D.J. (eds) Community Ecology: Pattern and Process. Blackwell, Melbourne, pp. 41–62. Kim, K.C. and Weaver, R.D. (1994) Biodiversity and humanity: paradox and challenge. In: Kim, K.C. and Weaver, R.D. (eds) Biodiversity and Landscapes. Cambridge University Press, Cambridge, pp. 3–27. Kindscher, K. and Wells, P.V. (1995) Prairie plant guilds: a multivariate analysis of prairie species based on ecological and morphological traits. Vegetatio 117, 29–50. Kinzig, A.P., Pacala, S.W. and Tilman, D. (2002) The Functional Consequences of Biodiversity: Empirical Progress and Theoretical Extensions. Princeton University Press, Princeton, New Jersey. Knops, J.M.H., Tilman, D., Haddad, N.M., Naeem, S., Mitchell, C.E., Haarstad, J., Ritchie, M.E., Howe, K.M., Reich, P.B., Siemann, E. and Groth, J. (1999) Effects of plant species richness on invasion dynam- ics, disease outbreaks, insect abundances and diversity. Ecological Letters 2, 286–293 Knops, J.M.H., Wedin, D. and Tilman, D. (2001) Biodiversity and decomposition in experimental grassland ecosystems. Oecologia 126, 429–433. Koricheva, J., Mulder, C.P.H., Schmid, B., Joshi, J. and Huss-Danell, K. (2000) Numerical responses of differ- ent trophic groups of invertebrates to manipulations of plant diversity in grasslands. Oecologia 125, 271–282. Körner, C. (1993) Scaling from species to vegetation: the usefulness of functional groups. In: Schulze, E.-D. and Mooney, H.A. (eds) Biodiversity and Ecosystem Function, Ecological Studies 99. Springer-Verlag, Berlin, pp. 117–140. Körner, C. (1999) Alpine Plant Life. Springer, Berlin. Lamont, B.B. (1995) Testing the effect of ecosystem composition/structure on its functioning. Oikos 74, 283–295.

Ecological importance of species diversity 279 Lavorel, S. and Richardson, D.M. (1999) Diversity, stability and conservation of Mediterranean-type ecosys- tems in a changing world: an introduction. Diversity and Distributions 5, 1–2. Lavorel, S., Prieur-Richard, A.-H. and Grigulis, K. (1999) Invasibility and diversity of plant communities: from pattern to process. Diversity and Distributions 5, 41–49. Lawton, J.H. (1994) What do species do in ecosystems? Oikos 71, 367–374. Lawton, J.H. (1996) The role of species in ecosystems: aspects of ecological complexity and biological diversity. In: Abe, T., Levin, S.A. and Higashi, M. (eds) Biodiversity: an Ecological Perspective. Springer, New York, pp. 215–228. Lawton, J.H. (1999) Biodiversity and ecosystem processes: theory, achievements and future directions. In: Kato, M. (ed.) The Biology of Biodiversity. Springer-Verlag, Tokyo, pp. 119–131. Lawton, J.H. and Brown, V.K. (1993) Redundancy in ecosystems. In: Schulze, E.D. and Mooney, H.A. (eds) Biodiversity and Ecosystem Function, Ecological Studies 99. Springer, Berlin, pp. 255–270. Lawton, J.H., Naeem, S., Thompson, L.J., Hector, A. and Crawley, M.J. (1998) Biodiversity and ecosystem function: getting the Ecotron experiment in its correct context. Functional Ecology 12, 848–852. Lehman, C.L. and Tilman, D. (2000) Biodiversity, stability, and productivity in competitive communities. The American Naturalist 156, 534–552. Leigh, E.G. (1965) On the relationship between productivity, biomass, diversity and stability of a commu- nity. Proceedings of the National Academy of Sciences USA 53, 777–783. Leishman, M.R. and Westoby, M. (1992) Classifying plants into groups on the basis of associations of indi- vidual traits – evidence from Australian semi-arid woodlands. Journal of Ecology 80, 417–424. Levine, J.M. (2000) Species diversity and biological invasions: relating local process and community pattern. Science 288, 852–854. Levine, J.M. and D’Antonio, C.M. (1999) Elton revisited: a review of evidence linking diversity and stability. Oikos 87, 15–26. Levine, J.M., Kennedy, T. and Naeem, S. (2002) Neighbourhood scale effects of species diversity on biologi- cal invasions and their relationship to community patterns. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 114–124. Loreau, M. (1998a) Biodiversity and ecosystem functioning: a mechanistic model. Proceedings of the National Academy of Sciences USA 95, 5632–5636. Loreau, M. (1998b) Separating sampling and other effects in biodiversity experiments. Oikos 82, 600–602. Loreau, M. (2000) Biodiversity and ecosystem functioning: recent theoretical advances. Oikos 91, 3–17. Loreau, M. and Behara, N. (1999) Phenotypic diversity and stability of ecosystem processes. Theoretical Population Biology 56, 29–47. Loreau, M. and Hector, A. (2001) Partitioning selection and complementarity in biodiversity experiments. Nature 412, 72–76. Loreau, M., Naeem, S., Inchausti, P., Bengtsson, J., Grime, J.P., Hector, A., Hooper, D.U., Huston, M.A., Raffaelli, D., Schmid, B., Tilman, D. and Wardle, D.A. (2001) Biodiversity and ecosystem functioning: current knowledge and future challenges. Science 294, 804–808. Loreau, M., Downing, A., Emmerson, M., Gonzalez, A., Hughes, J., Inchausti, P., Joshi, J., Norberg, J. and Sala, O. (2002) A new look at the relationship between diversity and stability. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 79–91. Loreau, M., Mouquet, N. and Gonzalez, A. (2003) Biodiversity as spatial insurance in heterogeneous land- scapes. Proceedings of the National Academy of Sciences USA 100, 12765–12770. MacArthur, R.H. (1969) Patterns of communities in the tropics. In: Lowe-McConnell, R.H. (ed.) Speciation in Tropical Environments. Academic Press, New York, pp. 19–30. Machlis, G.E. and Forester, J.F. (1996) The relationship between socio-economic factors and the loss of bio- diversity: first efforts at theoretical and quantitative models. In: Szaro, R.C. and Johnston, D.W. (eds) Biodiversity in Managed Landscapes: Theory and Practice. Oxford University Press, New York, pp. 121–146. Margalef, R. (1969) Diversity and stability: a practical proposal and a model of interdependence. Brookhaven Symposium in Biology 22, 25–37. Marks, P.L. (1974) The role of pin cherry (Prunus pennsylvanica L.) in the maintenance of stability in north- ern hardwood ecosystems. Ecological Monographs 44, 73–88. Martinez, N.D. (1996) Defining and measuring functional aspects of biodiversity. In: Gaston, K.J. (ed.) Biodiversity: a Biology of Numbers and Difference. Blackwell, Oxford, pp. 114–148. May, R.M. (1972) Will a large complex system be stable? Nature 238, 413–414.

280 C. Beierkuhnlein and A. Jentsch May, R.M. (1986) How many species are there? Nature 324, 514–515. May, R.M. (1988) How many species are there on earth? Science 324, 1441–1449. May, R.M. (1990) How many species? Philosophical Transactions of the Royal Society of London Series B 330, 293–304. McGrady-Steed, J., Harris, P.M. and Morin, P.J. (1997) Biodiversity regulates ecosystem predictability. Nature 390, 162–165. McIntyre, S., Ladiges, P.Y. and Adams, G. (1988) Plant species-richness and invasion by exotics in relation to disturbance of wetland communities on the Riverine Plain, NSW. Australian Journal of Ecology 13, 361–373. McIntyre, S., Lavorel, S. and Tremont, R.M. (1995) Plant life-history attributes: their relationship to distur- bance response in herbaceous vegetation. Journal of Ecology 83, 31–44. McNaughton, S.J. (1977) Diversity and stability of ecological communities: a comment on the role of empiricism in ecology. American Naturalist 111, 515–525. McNaughton, S.J. (1978) Stability and diversity of ecological communities. Nature 274, 251–253. Meiners, S.J., Cadenasso, M.L. and Pickett, S.T.A. (2004) Beyond biodiversity: individualistic controls of invasion in a self-assembled community. Ecology Letters 7, 121–126. Mikola, J., Bardgett, R.D. and Hedlund, K. (2002) Biodiversity, ecosystem functioning and soil decomposer food webs. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 169–180. Mills, L.S., Soulé, M.E. and Doak, D.F. (1993) The keystone-species concept in ecology and conservation. BioScience 43, 219–224. Minns, A., Finn, J., Hector, A., Caldeira, M., Joshi, J., Palmborg, C., Schmid, B., Scherer-Lorenzen, M., Spehn, E., Troumbis, A., Beierkuhnlein, C., Dimitrakopoulos, P.G., Huss-Danell, K., Jumpponen, A., Loreau, M., Mulder, C.P.H., O’Donovan, G., Pereira, J.S., Schulze, E.-D., Terry, A.C. and Woodward, F.I. (2001) The functioning of European grassland ecosystems: potential benefits of biodiversity to agri- culture. Outlook on Agriculture 30, 179–185. Mitchell, C., Tilman, D. and Groth, J.V. (2002) Effects of grassland plant species diversity, abundance, and composition on foliar fungal disease. Ecology 83, 1713–1726. Montgomery, C.A. and Pollack, R.A. (1996) Economics and biodiversity: weighing the benefits and costs of conservation. Journal of Forestry 94, 34–38. Mooney, H.A. (2002) The debate on the role of biodiversity in ecosystem functioning. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 12–17. Mooney, H.A., Lubchenko, J., Dirzo, R. and Sala, O.E. (1995a) Biodiversity and ecosystem functioning: ecosystem analyses. In: Heywood, V.H. and Watson, R.T. (eds) Global Biodiversity Assessment. Cambridge University Press, Cambridge, pp. 327–452. Mooney, H.A., Lubchenko, J., Dirzo, R. and Sala, O.E. (1995b) Biodiversity and ecosystem functioning: basic principles. In: Heywood, V.H. and Watson, R.T. (eds) Global Biodiversity Assessment. Cambridge University Press, Cambridge, pp. 275–326. Mooney, H.A., Cushman, J.H., Medina, E., Sala, O.E. and Schulze, E.-D. (eds) (1996) Functional Roles of Biodiversity – a Global Perspective. Wiley & Sons, Chichester. Mulder, C.P.H., Uliassi, D.D. and Doak, D.F. (2001) Physical stress and diversity-productivity relationships: the role of positive species interactions. Proceedings of the National Academy of Sciences USA 98, 6704–6708. Myers, N. (1988) Threatened biotas: ‘hotspots’ in tropical forests. Environmentalist 8, 1–20. Nader, W. and Mateo, N. (1998) Biodiversity – resource for new products, development and self-reliance. In: Barthlott, W. and Winiger, M. (eds) Biodiversity. A Challenge for Development Research and Policy. Springer, Berlin, pp. 181–197. Naeem, S. (1998) Species redundancy and ecosystem reliability. Conservation Biology 12, 39–45. Naeem, S. and Li, S. (1997) Biodiversity enhances ecosystem reliability. Nature 390, 507–509. Naeem, S. and Wright, J.P. (2003) Disentangling biodiversity effects on ecosystem functioning: deriving solutions to a seemingly insurmountable problem. Ecology Letters 6, 567–579. Naeem, S., Thompson, L.J., Lawler, S.P., Lawton, J.H. and Woodfin, R.M. (1994) Declining biodiversity can alter the performance of ecosystems. Nature 368, 734–737. Naeem, S., Thompson, L.J., Lawler, S.P., Lawton, J.H. and Woodfin, R.M. (1995) Empirical evidence that declining species diversity may alter the performance of terrestrial ecosystems. Philosophical Transactions of the Royal Society of London Series B 347, 249–262.

Ecological importance of species diversity 281 Naeem, S., Håkansson, K., Lawton, J.H., Crawley, M.J. and Thompson, L.J. (1996) Biodiversity and plant productivity in a model assemblage of plant species. Oikos 76, 259–264. Naeem, S., Knops, J.M.H., Tilman, D., Howe, K.M., Kennedy, T. and Gale, S. (2000) Plant diversity increases resistance to invasion in the absence of covarying extrinsic factors. Oikos 91, 97–108. Naeem, S., Loreau, M. and Inchausti, P. (2002) Biodiversity and ecosystem functioning: the emergence of a synthetic ecological framework. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 3–11. Noble, I.R. and Slayter, R.O. (1980) The use of vital attributes to predict successional changes in plant com- munities subject to recurrent disturbances. Vegetatio 43, 5–21. Nunes, P. and van den Bergh, J. (2001) Economic valuation of biodiversity: sense or nonsense? Ecological Economics 39, 203–222. Odum, E.P. (1993) Ideas in ecology for the 1990s. BioScience 42, 542–545. OECD (1999) Handbook of Incentive Measures for Biodiversity. Design and Implementation. OECD, Paris. Oksanen, J. (1996) Is the humped relationship between species richness and biomass an artifact due to plot size? Journal of Ecology 84, 293–296. Pacala, S.W. and Deutschmann, D.H. (1996) Details that matter: the spatial distribution of individual trees maintains forest ecosystem function. Oikos 74, 357–365. Pacala, S. and Tilman, D. (2002) The transition from sampling to complementarity. In: Kinzig, A.P., Pacala, S.W. and Tilman, D. (eds) The Functional Consequences of Biodiversity: Empirical Progress and Theoretical Extensions. Princeton University Press, Princeton, New Jersey, pp. 151–166. Palmer, M.W. and Maurer, T.A. (1997) Does diversity beget diversity? A case study of crops and weeds. Journal of Vegetation Science 8, 235–240. Pavlovic, N.B. (1994) Disturbance-dependent persistence of rare plants: anthropogenic impacts and restora- tion implications. In: Boels, M.L. and Whelan, C. (eds) Recovery and Restoration of Endangered Species. Cambridge University Press, Cambridge, pp. 159–193. Pearce, D. and Moran, D. (eds) (1994) The Economic Value of Biodiversity. Earthscan, London. Perrings, C. (ed.) (1995) Biodiversity Loss – Economic and Ecological Issues. Cambridge University Press, Cambridge. Petchey, O.L. and Gaston, K.J. (2002) Extinction and the loss of functional diversity. Proceedings of the Royal Society of London Series B 269, 1721–1727. Peters, R.L. (1994) Conserving biological diversity in the face of climate change. In: Kim, K.C. and Weaver, R.D. (eds) Biodiversity and Landscapes. Cambridge University Press, Cambridge, pp. 105–132. Peters, R.L. and Lovejoy, T.E. (eds) (1992) Global Warming and Biological Diversity. Yale University Press, New Haven, Connecticut. Peterson, G., Allen, C.R. and Holling, C.S. (1998) Ecological resilience, biodiversity and scale. Ecosystems 1, 6–18. Pfisterer, A.B. and Schmid, B. (2002) Diversity-dependent production can decrease the stability of ecosystem functioning. Nature 416, 84–86. Pianka, E.R. (1966) Latitudinal gradients in species diversity: a review of concepts. American Naturalist 100, 33–46. Pimentel, D., Wilson, C., McCullum, C., Huang, R., Dwen, P., Flack, J., Tran, Q., Saltman, T. and Cliff, B. (1997) Economic and environmental benefis of biodiversity. BioScience 47, 747–755. Pimentel, D., Lach, L., Zuniga, R. and Morrison, D. (2000) Environmental and economic costs of nonindige- nous species in the United States. BioScience 50, 54–65. Pimm, S.L., Russell, G.J., Gittelman, J.L. and Brooks, T. (1995) The future of biodiversity. Science 269, 347–350. Planty-Tabacchi, A.-M., Tabacchi, E., Naiman, R.J., DeFerrari, C. and Decamps, H. (1996) Invasibility of species-rich community in riparian zones. Conservation Biology 10, 598–607. Plotkin, M.J. (1988) The outlook for new agricultural and industrial products from the tropics. In: Wilson, E.O. and Peter, F.M. (eds) BioDiversity. National Academy Press, Washington, DC, pp. 106–116. Prieur-Richard, A.-H. and Lavorel, S. (2000) Invasions: the perspective of diverse plant communities. Austral Ecology 25, 1–7. Prieur-Richard, A.-H., Lavorel, S., Grigulis, K. and Dos Santos, A. (2000) Plant community diversity and invasibility by exotics: invasion of Mediterranean old fields by Conyza bonariensis and Conyza canadi- ensis. Ecology Letters 3, 412–422. Primack, R.B. (1993) Essentials of Conservation Biology. Sinauer Associates, Sunderland, Massachusetts. Pysek, P. and Pysek, A. (1995) Invasion by Heracleum mantegazzianum in different habitats in the Czech Republic. Journal of Vegetation Science 6, 711–718.

282 C. Beierkuhnlein and A. Jentsch Raffaelli, D., van der Putten, W.H., Persson, L., Wardle, D.A., Petchey, O.L., Koricheva, J., van der Heijden, M., Mikola, J. and Kennedy, T. (2002) Multi-trophic dynamics and ecosystem processes. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 147–154. Randall, A. (1994) Thinking about the value of biodiversity. In: Kim, K.C. and Weaver, R.D. (eds) Biodiversity and Landscapes. Cambridge University Press, Cambridge, pp. 271–285. Rapson, G.L., Thompson, K. and Hodgson, J.G. (1997) The humped relationship between species richness and biomass – testing its sensitivity to sample quadrat size. Journal of Ecology 85, 99–100. Reich, P.B., Knops, J., Tilman, D., Craine, J., Ellsworth, D., Tjoelker, M., Lee, T., Wedin, D., Naeem, S., Bahauddin, D., Hendrey, G., Jose, S., Wrage, K., Goth, J. and Bengston, W. (2001) Plant diversity enhances ecosystem responses to elevated CO2 and nitrogen deposition. Nature 410, 809–812. Rejmánek, M. (1989) Invasibility of plant communities. In: Drake, J.A., Mooney, H.A., Di Castri, F., Groves, R.H., Kruger, F.J., Rejmánek, M. and Williamson, M. (eds) Biological Invasions: a Global Perspective. John Wiley & Sons, Chichester, UK, pp. 369–388. Rickleffs, R.E. (1997) The Economy of Nature. Freeman, New York. Ring, I. (2002) Ecological public functions and fiscal equalisation at the local level in Germany. Ecological Economics 42, 415–427. Risser, P.G. (1995) Biodiversity and ecosystem function. Conservation Biology 9, 742–746. Robinson, G.R., Quinn, J.F. and Stanton, M.L. (1995) Invasibility of experimental habitat islands in a California winter annual grassland. Ecology 76, 786–794. Sala, O.E., Chapin, F.S. III, Armesto, J.J., Berlow, R., Bloomfield, J., Dirzo, R., Huber-Sanwald, E., Huenneke, L.F., Jackson, R.B., Kinzig, A., Leemans, R., Lodge, D., Mooney, H.A., Oesterheld, M., Poff, N.L., Sykes, M.T., Walker, B.H., Walker, M. and Wall, D.H. (2000) Global biodiversity scenarios for the year 2100. Science 287, 1770–1774. Scherer-Lorenzen, M. (1999) Effects of plant diversity on ecosystem processes in experimental grassland communities. Bayreuther Forum Ökologie 75, 1–195. Scherer-Lorenzen, M., Palmborg, C., Prinz, A. and Schulze, E.-D. (2003) The role of plant diversity and com- position for nitrate leaching in grasslands. Ecology 84, 1539–1552. Schläpfer, F. and Schmid, B. (1999) Ecosystem effects of biodiversity: a classification of hypotheses and exploration of empirical results. Ecological Applications 9, 893–912. Schläpfer, F., Schmid, B. and Seidl, I. (1999) Expert estimates about effects of biodiversity on ecosystem processes and services. Oikos 84, 346–352. Schlesinger, W.H. (ed.) (1991) Biogeochemistry: an Analysis of Global Change. Academic Press, San Diego, California. Schmid, B. (2002) The species richness–productivity controversy. Trends in Ecology and Evolution 17, 113–114. Schmid, B., Hector, A., Huston, M.A., Inchausti, P., Nijs, I., Leadley, P.W. and Tilman, D. (2002a) The design and analysis of biodiversity experiments. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 61–75. Schmid, B., Joshi, J. and Schläpfer, F. (2002b) Empirical evidence for biodiversity–ecosystem functioning relationships. In: Kinzig, A.P., Pacala, S.W. and Tilman, D. (eds) The Functional Consequences of Biodiversity: Empirical Progress and Theoretical Extensions. Princeton University Press, Princeton, New Jersey, pp. 120–150. Schulze, E.-D. and Mooney, H.A. (1993) Ecosystem function and biodiversity: a summary. In: Schulze, E.-D. and Mooney, H.A. (eds) Biodiversity and Ecosystem Function, Ecological Studies 99. Springer-Verlag, Berlin, pp. 497–510. Schwartz, M.W., Brigham, C.A., Hoeksema, J.D., Lyons, K.G., Mills, M.H. and van Mantgem, P.J. (2000) Linking biodiversity to ecosystem function: implications for conservation ecology. Oecologia 122, 297–305. Seidl, I. and Gowdy, J. (1999) Monetäre Bewertung von Biodiversität: Grundannahmen, Schritte, Probleme und Folgerungen. GAIA 8, 102–112. Siemann, E., Tilman, D., Haarstad, J. and Ritchie, M. (1998) Experimental tests of the dependence of arthro- pod diversity on plant diversity. American Naturalist 152, 738–750. Smith, M.D. and Knapp, A.K. (2003) Dominant species maintain ecosystem function with non-random species loss. Ecology Letters 6, 509–517. Smith, T.M., Shugart, H.H., Woodward, F.I. and Burton, P.J. (1993) Plant functional types. In: Solomon, A.M. and Shugart, H.H. (eds) Vegetation Dynamics and Global Change. Chapman & Hall, New York, pp. 272–292.

Ecological importance of species diversity 283 Solbrig, O.T. (1993) Plant traits and adaptive strategies: their role in ecosystem function. In: Schulze, E.-D. and Mooney, H.A. (eds) Biodiversity and Ecosystem Function, Ecological Studies 99. Springer-Verlag, Berlin, pp. 97–116. Soulé, M.E. (1991) Conservation: tactics for a constant crisis. Science 253, 744–750. Spehn, E.M., Joshi, J., Schmid, B., Diemer, M. and Körner, C. (2000a) Aboveground resource use increases with plant species richness in experimental grassland ecosystems. Functional Ecology 14, 326–337. Spehn, E., Joshi, J., Schmid, B., Alphei, J. and Körner, C. (2000b) Plant diversity effects on soil heterotrophic activity in experimental grassland ecosystems. Plant and Soil 224, 217–230. Spehn, E.M., Scherer-Lorenzen, M., Schmid, B., Hector, A., Caldeira, M.C., Dimitrakopoulos, P.G., Finn, J.A., Jumpponen, A., O’Donnovan, G., Pereira, J.S., Schulze, E.-D., Troumbis, A.Y. and Körner, C. (2002) The role of legumes as a component of biodiversity in a cross-European study of grassland bio- mass nitrogen. Oikos 98, 205–218. Stephan, A., Meyer, A.H. and Schmid, B. (2000) Plant diversity affects culturable soil bacteria in experimen- tal grassland communities. Journal of Ecology 88, 988–998. Stohlgren, T.J., Binkley, D., Chanog, G.W., Kalkhan, M.A., Schell, L.D., Bull, K.A., Otsuki, Y., Newman, G., Bashkin, M. and Son, Y. (1999) Exotic plant species invade hot spots of native plant diversity. Ecological Monographs 69, 25–46. Symstad, A.J. (2000) A test of the effects of functional group richness and composition on grassland invasi- bility. Ecology 81, 99–109. Symstad, A.J. and Tilman, D. (2001) Diversity loss, recruitment limitation and ecosystem functioning: lessons learned from a removal experiment. Oikos 92, 424–435. Symstad, A.J., Tilman, D., Willson, J. and Knops, J. (1998) Species loss and ecosystem functioning: effects of species identity and community composition. Oikos 81, 389–397. Temperton, V.M., Hobbs, R., Fattorini, M. and Halle, S. (eds) (2004) Assembly Rules in Restoration Ecology – Bridging the Gap between Theory and Practice. Island Press Books, Washington, DC, and Sinauer Associates, Sunderland, Massachusetts. ten Kate, K. and Laird, S.A. (2002) The Commercial Use of Biodiversity. Earthscan, London. Thierry, R.G. (1982) Environmental instability and community diversity. Biological Reviews 57, 671–710. Thompson, K. and Hodgson, J.G. (1998) Biodiversity and ecosystem function: getting the Ecotron experi- ment in its correct context – response. Functional Ecology 12, 852–853. Tilman, D. (1993) Community diversity and succession: the roles of competition, dispersal and habitat mod- ification. In: Schulze, E.-D. and Mooney, H.A. (eds) Biodiversity and Ecosystem Function, Ecological Studies 99. Springer-Verlag, Berlin, pp. 327–344. Tilman, D. (1996) Biodiversity: population versus ecosystem stability. Ecology 77, 350–363. Tilman, D. (1997a) Community invasibility, recruitment limitation, and grassland biodiversity. Ecology 78, 81–92. Tilman, D. (1997b) Distinguishing between the effects of species diversity and species composition. Oikos 80, 185–185. Tilman, D. (1999) Diversity and production in European grasslands. Science 286, 1099–1100. Tilman, D. (2001) Effects of diversity and composition on grassland stability and productivity. In: Press, M.C., Huntly, N.J. and Levin, S. (eds) Ecology: Achievement and Challenge. Blackwell Science, Oxford, pp. 183–207. Tilman, D. and Downing, J.A. (1994) Biodiversity and stability in grasslands. Nature 367, 363–365. Tilman, D. and Lehman, C. (2002) Biodiversity, composition, and ecosystem processes: theory and con- cepts. In: Kinzig, A., Pacala, S. and Tilman, D. (eds) Functional Consequences of Biodiversity: Empirical Progress and Theoretical Extensions. Princeton University Press, Princeton, New Jersey, pp. 9–41. Tilman, D., Downing, J.A. and Wedin, D.A. (1994) Does diversity beget stability? Nature 371, 113–114. Tilman, D., Wedin, D. and Knops, J. (1996) Productivity and sustainability influenced by biodiversity in grassland ecosystems. Nature 379, 718–720. Tilman, D., Knops, J., Wedin, D., Reich, P., Ritchie, M. and Siemann, E. (1997a) The influence of functional diversity and composition on ecosystem processes. Science 277, 1300–1302. Tilman, D., Lehman, C.L. and Thomson, K.T. (1997b) Plant diversity and ecosystem productivity: theoretical considerations. Proceedings of the National Academy of Sciences USA 94, 1857–1861. Tilman, D., Naeem, S., Knops, J., Reich, P., Siemann, E., Wedin, D., Ritchie, M. and Lawton, J. (1997c) Biodiversity and ecosystem properties. Science 278, 1866–1867. Tilman, D., Lehman, C.L. and Bristow, C.E. (1998) Diversity–stability relationships: statistical inevitability or ecological consequence? The American Naturalist 151, 277–282.

284 C. Beierkuhnlein and A. Jentsch Tilman, D., Reich, P.B., Knops, J., Wedin, D., Mielke, T. and Lehman, C. (2001) Diversity and productivity in a long-term grassland experiment. Science 294, 843–845. Tilman, D., Knops, J., Wedin, D. and Reich, P. (2002a) Experimental and observational studies of diversity, productivity and stability. In: Kinzig, A.P., Pacala, S.W. and Tilman, D. (eds) The Functional Consequences of Biodiversity: Empirical Progress and Theoretical Extensions. Princeton University Press, Princeton, New Jersey, pp. 42–70. Tilmann, D., Knops, J., Wedin, D. and Reich, P. (2002b) Plant diversity and composition: effects on produc- tivity and nutrient dynamics of experimental grasslands. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 21–35. Troumbis, A. and Memtas, D. (2000) Observational evidence that diversity may increase productivity in Mediterranean shrublands. Oecologia 125, 101–108. UNEP (1992) World Atlas of Desertification. UNEP, London. van der Heijden, M.G.A. and Cornelissen, J.H.C. (2002) The critical role of plant–microbe interactions on biodiversity and ecosystem functioning: arbuscular mycorrhizal associations as an example. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 181–192. van der Maarel, E. (1986) Vegetation structure and its relations with dynamics and diversity. Journale Botanico Italiano 120, 59–60. van der Maarel, E. (1988) Diversity in plant communities in relation to structure and dynamics. In: During, H.J., Werger, M.J.A. and Willems, J.H. (eds) Diversity and Pattern in Plant Communities. SPB, The Hague, pp. 1–14. van der Maarel, E. (1997) Biodiversity: From Babel to Biosphere Management. Opulus Press, Uppsala, Sweden. van Ruijven, J. and Berendse, F. (2003) Positive effects of plant diversity on productivity in the absence of legumes. Ecology Letters 6, 170–175. van Ruijven, J., De Deyn, G.B. and Berendse, F. (2003) Diversity reduces invasibility in experimental plant communities: the role of plant species. Ecology Letters 6, 910–918. van Wilgen, B.W., Cowling, R.M. and Burgers, C.J. (1996) Valuation of ecosystem services. BioScience 46, 184–189. Vandermeer, J., Lawrence, D., Symstad, A. and Hobbie, S. (2002) Effect of biodiversity on ecosystem func- tioning in managed ecosystems. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 221–233. Vavilov, N.I. (1935) The phytogeographic basis of plant breeding. In: The Origin, Variation, Immunity and Breeding of Cultivated Plants: Selected Writing of N.I. Vavilov. (transl. by S. Chester) 13 (1949/50). Chronica Botanica Co., Waltham, Massachusetts, pp. 13–54. Vitousek, P.M. (1984) A general theory of forest nutrient dynamics. In: Agren, G.I. (ed.) State and Change of Forest Ecosystems – Indicators in Current Research, Report 13. Swedish University of Agricultural Sciences, Uppsala, pp. 121–135. Vitousek, P.M. and Hooper, D.U. (1993) Biological diversity and terrestrial ecosystem biogeochemistry. In: Schulze, E.-D. and Mooney, H.A. (eds) Biodiversity and Ecosystem Function. Springer-Verlag, Berlin, pp. 3–14. Vogl, R.J. (1974) Effects of fire on grasslands. In: Kozlowski, T.T. and Ahlgren, C.E. (eds) Fire and Ecosystems. Academic Press, New York, pp. 139–194. von Hardenberg, J., Meron, E., Shachak, M. and Zarmi, Y. (2001) Diversity of vegetation patterns and deser- tification. Physical Review Letters 87, 198101. Waide, R.B., Willig, M.R., Steiner, C.F., Mittelbach, G., Gough, L., Dodson, S.I., Juday, G.P. and Parmenter, R. (1999) The relationship between productivity and species richness. Annual Review of Ecology and Systematics 30, 257–300. Walker, B.H. (1992) Biodiversity and ecological redundancy. Conservation Biology 6, 18–23. Walker, B.H., Kinzig, A. and Langridge, J. (1999) Plant attribute diversity, resilience, and ecosystem function: the nature and significance of dominant and minor species. Ecosystems 2, 95–113. Wall, D.H. (1999) Biodiversity and ecosystem functioning. BioScience 49, 107–108. Wardle, D.A. (1999) Is sampling effect a problem for experiments investigating biodiversity–ecosystem func- tion relationships? Oikos 87, 403–407. Wardle, D.A. (2001) Experimental demonstration that plant diversity reduces invasibility – evidence of a biological mechanism or a consequence of sampling effect? Oikos 95, 1–170.

Ecological importance of species diversity 285 Wardle, D.A. and van der Putten, W.H. (2002) Biodiversity, ecosystem functioning and above- ground–below-ground linkages. In: Loreau, M., Naeem, S. and Inchausti, P. (eds) Biodiversity and Ecosystem Functioning. Oxford University Press, Oxford, pp. 155–168. Wardle, D.A., Bonner, K.I. and Nicholson, K.S. (1997) Biodiversity and plant litter: experimental evidence which does not support the view that enhanced species richness improves ecosystem function. Oikos 79, 247–258. Wardle, D.A., Bonner, K.I. and Barker, G.M. (2000) Stability of ecosystem properties in response to above- ground functional group richness and composition. Oikos 89, 11–23. Weiher, E. (1999) The combined effects of scale and productivity on species richness. Journal of Ecology 87, 1005–1011. Westoby, M. and Leishman, M.R. (1997) Categorizing plant species into functional types. In: Smith, T.M., Shugart, H.H. and Woodward, F.I. (eds) Plant Functional Types: Their Relevance to Ecosystem Properties and Global Change. Cambridge University Press, Cambridge, pp. 104–121. White, P.S. and Jentsch, A. (2001) The search for generality in studies of disturbance and ecosystem dynam- ics. Progress in Botany 62, 399–449. White, P. and Jentsch, A. (2004) Disturbance, succession and community assembly in terrestrial plant com- munities. In: Temperton, V.M., Hobbs, R., Fattorini, M. and Halle, S. (eds) Assembly Rules in Restoration Ecology – Bridging the Gap between Theory and Practise. Island Press Books, Washington, DC, pp. 371–399. White, P.S., MacKenzie, M.D. and Busing, R.T. (1985) A critique of overstory/understory comparisons based on transition probability analysis of an old growth spruce–fir stand in the Appalachians. Vegetatio 64, 37–45. Whittaker, R.H. (1972) Evolution and measurement of species diversity. Taxon 12, 213–251. Wiegand, T., Moloney, K.A., Naves, J. and Knauer, F.E. (1999) Finding the missing link between landscape structure and population dynamics: a spatially explicit perspective. American Naturalist 154, 605–627. Williams, P.H., Humphries, C.J., Vane-Wright, R.I. and Gaston, K.J. (1996) Value in biodiversity, ecological services and consensus. Trends in Ecology and Evolution 11, 385. Wilsey, B.J. and Polley, H.W. (2002) Reductions in grassland species evenness increase dicot seedling inva- sion and spittle bug infestation. Ecology Letters 5, 676–684. Wilsey, B.J. and Potvin, C. (2000) Biodiversity and ecosystem functioning: the importance of species even- ness in an old field. Ecology 81, 887–892. Wilson, D.S. (1992) Complex interactions in meta-communities, with implications for biodiversity and higher levels of selection. Ecology 73, 1984–2000. Wilson, E.O. (1985) The biological diversity crisis. BioScience 35, 700–706. Woodward, F.I. and Cramer, W. (1996) Plant functional types and climatic changes: introduction. Journal of Vegetation Science 7, 306–308. Woodward, F.I. and Kelly, C.K. (1997) Plant functional types: towards a definition by environmental con- straints. In: Smith, T.M., Shugart, H.H. and Woodward, F.I. (eds) Plant Functional Types: Their Relevance to Ecosystem Properties and Global Change. Cambridge University Press, Cambridge, pp. 47–65. Wootton, T.W. and Downing, A. (2003) Understanding the effects of reduced biodiversity: a comparison of two approaches. In: Kareiva, P. and Levin, S.A. (eds) The Importance of Species: Perspectives on Expendability and Triage. Princeton University Press, Princeton, New Jersey. Yachi, S. and Loreau, M. (1999) Biodiversity and ecosystem productivity in a fluctuating environment: the insurance hypothesis. Proceedings of the National Academy of Sciences USA 96, 57–64.

14 Genomic diversity in nature and domestication Eviatar Nevo Institute of Evolution, University of Haifa, Mt Carmel, Haifa, Israel Problems of Genomic Diversity in nature and domestication determined by the Nature and Domestication mating system of selfing and outcrossing, hitch-hiking, ecological stresses, adaptive Genomic diversity is the basis of evolution- origin and maintenance, linkage disequilib- ary change in nature underlying the evolu- ria or epistasis? How much genomic change tion of biodiversity at the level of genes, occurred during domestication evolution? Is genomes, populations, species and ecosys- genetic erosion threatening cultivars, their tems, and the evolution of genome– genetic diversity, the future of agriculture phenome diversity under environmental and the stable supply of food to a constantly stress (Nevo, 2001a,b). It is the prime mover increasing human population? If so, how of molecular-genotypic and organismal-phe- should the genetic resources of the progeni- notypic evolution. Genetic diversity and tors of cultivated plants be protected and selection drive the twin evolutionary utilized in crop improvement? processes of adaptation and speciation and the fast human evolutionary processes of Evidence and Theory of Genetic domestication, agriculture and medicine. Diversity in Nature Nevertheless, despite the cardinal role of Substantial evidence derived from the analy- genomic diversity in the evolutionary sis of protein and DNA genomic diversity and processes in nature and domestication, its divergence in nature locally, regionally and evolutionary origin, dynamics and mainte- globally indicate the following (Nevo, nance remain largely enigmatic, notwith- 2001a,b): (i) an abundant genotypic and phe- standing the dramatic discoveries of notypic diversity exists in nature; (ii) the molecular biology of abundant diversity in organization and evolution of molecular- nature. For example, what is the relative genetic and organismal diversity in nature at importance of deterministic (natural selec- all geographic scales are non-random and tion) and partly stochastic (mutation, migra- structured; (iii) genetic and genomic diversity tion, recombination, interspecific lateral display regularities across life and are posi- gene transfer and random events) forces in tively correlated with, and partly predictable driving the patterns, levels and dynamics of by, abiotic and biotic environmental hetero- genomic diversity of both coding and non- geneity and stress; and (iv) biodiversity evo- coding regions, but especially of the latter? How much is genetic differentiation in © CAB International 2005. Plant Diversity and Evolution: Genotypic and 287 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

288 E. Nevo lution, even in small isolated populations, is environment and their local adaptations and primarily driven by natural selection includ- genetic diversity resources are extractable ing diversifying, balancing, cyclical and puri- for crop improvement. Plant and animal fying selective regimes, interacting with, but domestication revolutionized human cul- ultimately overriding, the effects of muta- tural evolution and is the major factor tion, migration and stochasticity (Nevo, underlying human civilization, but most 1978, 1988, 1998a,b, 2001a,b; Nevo et al., emphatically, domestication is an extraordi- 1984, 2002). nary demonstration of active evolution of differential stages of speciation and adapta- Evolution under Domestication tion to the human-created agricultural niche. So, if the current presumed predica- How does the aforementioned evidence and ment of domestication due to genetic ero- theory relate to one of the most dramatic sion is verified by critical analysis, protecting and rapid evolutionary phenomena in this precious base of civilization becomes an nature, the domestication of plants and ani- urgent concern. This is particularly true in mals by humans, which is considered the view of the rapid disappearance of centres of most important human development in the diversity because of human developments past 13,000 years (Diamond, 2002)? How do (Harlan, 1975a,b). It should be mentioned, agricultural practices create new selection however, that studies exist (e.g. Reeves et al., pressures, in numerous traits, against those 1999) indicating no quantitative changes in favoured by natural selection? How do culti- genetic diversity of UK cereal crops over the vars and domesticates introduced by past 60–70 years. humans adapt to new environmental condi- tions? Diversity in characters associated with Origin of Agriculture crop plant and animal domestication evolu- tion is largely absent in the wild progenitors A post-Pleistocene global rise in temperature and much of it evolved under domestica- following the ice age (i.e. climatic change) tion, either unconsciously or consciously may have induced the expansion of econom- (Ladizinsky, 1998; Peng et al., 2003). What is ically important thermophilous plants and, the taxonomic status of cultivated plants vis- in turn, promoted complex foraging and à-vis their wild progenitors? Subspecies? plant cultivation. The shift from foraging to New species? Did humans cause rapid speci- steady production led to an incipient agricul- ation hundreds or thousands of years ago? ture varying in time in various parts of the world. Vavilov (1926) proposed eight centres Domestication, as already noted by of origin for most world-cultivated plants Darwin in his first chapter of ‘The Origin’ (Fig. 1 in Harlan, 1975a; see also Fig. 2 in (1859), is mankind’s most dramatic imitation Diamond, 2002). Agricultural origins (follow- of the evolutionary process of speciation. ing Harlan, 1975a) consist of centres and Darwin was so impressed by animal and non-centres including the Near East centre, plant domesticates that he regarded the cul- African non-centre, North Chinese centre, tivated derivatives more variable than their South-east Asia and South Pacific non- progenitors. Cultivated varieties for Darwin centres, Meso-American centre, and South tend to be converted into new and distinct American non-centre (Fig. 2 in Harlan, species. If so, human domestication of wild 1975a). In the Levant, agriculture developed plants and animals during the last ten mil- out of an intensive specialized exploitation of lennia is doubtlessly one of the best demon- plants and animals (Zohary, 1996). Natufian strations of fast evolution in action. It is a sedentism, followed by rapid population process that has been condensed in a short growth and resource stress induced by the time and advanced by artificial rather than expanding desert and coupled with available natural selection. Luckily, the original plant grinding technology, may have triggered and animal progenitors of many cultivars plant domestication (Bar-Yosef, 1998). and domesticates still live in their natural

Genomic diversity in nature and domestication 289 The earliest signs of domestication found distribution and ecological ranges of the so far in the archaeological records (Smith, wild relatives have been established. 1995; Bar-Yosef, 1998; Ladizinsky, 1998; Furthermore, comparisons between wild Badr et al., 2000; Lev-Yadun et al., 2000; types and their cultivated counterparts have Zohary and Hopf, 2000; Diamond, 2002; revealed the evolutionary changes brought Gopher et al., 2002; Kislev, 2002; Salamini et about by domestication (Peng et al., 2000, al., 2002) appear 10,000–12,000 years ago 2003) and taxonomic status of domesticates in the Fertile Crescent of the Near East, (Styles, 1986). For example, research in this Central America and southern China, field has enabled us to assess the relative involving different crops and independent importance of the evolutionary forces dri- cradles of domestication. Each was founded ving wheat evolution, hybridization, migra- on cereal staples whose reliability, yield and tion, drift and natural selection, interacting suitability for storage may have been crucial in generating the final wheat genotype. Our for early commitment to an agricultural way studies suggest that besides hybridization, of life and the domestication of other which contributed to the tetraploidy of plants. It is generally believed that plant Triticum dicoccoides, natural selection played a domestication first took place in the Jordan significant role, and it has oriented wheat Valley and in adjacent areas of the southern evolution primarily through the mecha- Levant (present-day Israel and Jordan). nisms of diversifying and balancing selection However, botanical, genetic and archaeolog- regimes (Nevo et al., 2002). ical evidence point to a small core area within the Fertile Crescent – near the upper The evolution of domestication can be reaches of the Tigris and Euphrates Rivers, considered in four contexts: in present-day south-eastern Turkey/north- ern Syria – as the cradle of Old World agri- 1. The evolution of new species of crop plants culture (Bar-Yosef, 1998; Pasternak, 1998; by humans through strong artificial selection Badr et al., 2000; Lev-Yadun et al., 2000; leading to adaptive complexes fitting human Zohary and Hopf, 2000; Salamini et al., demands. This provides fascinating insight 2002). Further evidence is needed to clarify into the evolutionary process (e.g. Hilton and when and where wheat and barley domesti- Baut, 1998; Peng et al., 2000, 2003). cation and agriculture originated, driving 2. The evolution of human civilization and modern civilization. Was it mono- or poly- the current consequences of population phyletic in origin (Zohary, 1999)? For date explosion and increasing world hunger in estimates on microsatellite genetic analysis, developing countries of the southern conti- which overlap the archaeological findings, nents (Diamond, 2002). see Ramachandran et al. (2005). 3. How much genetic change is involved in domestication (e.g. Lauter and Doebly, The genetic changes required for domesti- 2002; Peng et al., 2003)? cation to occur were relatively straight- 4. Did domestication involve a decline in forward and rapid, most of them arising from genetic diversity threatening their future direct and indirect selection for higher yield evolution? and quality traits. In domestication evolution, the relative importance of the evolutionary This chapter primarily focuses on point 4. driving forces including hybridization, migra- tion, drift and selection can best be assessed Darwin’s View on Variation under with a major emphasis on selection and Domestication hybridization. Considerable progress has been achieved in the field of the wild ancestry Darwin first realized that evolutionary of Old World crops including wild cereals change is a two-step process: (i) the produc- (Zohary and Hopf, 2000). The wild progeni- tion of variation; and (ii) the sorting out of tors of most of these cultivated plants have this variability by natural selection. This now been satisfactorily identified by compar- complementarity of variation and selection ative morphology and genetic analysis. The is the essence of the evolutionary process

290 E. Nevo both in nature and domestication. Darwin compared with that of their wild progenitors observed that the abundant variability in clearly indicate the superiority under nat- nature resulted not from major saltations ural conditions of the latter over the former but from the accumulation of small changes (Yamashita, 1965). occurring at random with respect to envi- ronmental conditions. For Darwin, however, Domestication Evolution and its natural selection was the direction-giving Predicaments force leading to adaptations. Moreover, Darwin claimed that a Domestication assembles from nature wild animals and plants and keeps them under careful study of domesticated animals and cultivation. Domesticated quantitative trait cultivated plants would offer the best chance … loci defining the genetic changes accompany- to gain a clear insight into the means of ing utilization of plants by humans have been modification and co-adaptation … our studied in rice, maize, sorghum, millet and knowledge, imperfect though it be, of variation wheat (Peng et al., 2003, and references under domestication, afforded the best and therein). Domesticates have been genetically safest clue to understanding the co-adaptations changed by human selection either con- for organic beings to each other and to their sciously or unconsciously. In general, domes- physical conditions of life… Therefore, … I ticates lose some of their fitness components shall devote the first chapter of the Origin to (reproductive and dispersal power as well as Variation Under Domestication… We shall see abiotic and biotic resistances to stress) adapt- that a large amount of hereditary modification ing them to their natural habitats; hence, is at least possible and how great the power of they largely cannot survive outside the agri- man in accumulating by his selection successive culturally protected environment (see below). slight variations. (Darwin, 1859, Introduction, phrases rearranged) Domesticates grow outside the natural distribution range of their progenitors and For Darwin, the have changed morphologically, physiologi- cally and phenologically, with a genetic vast diversity of the plants and animals which change establishing a close and evolving have been cultivated and which have varied symbiosis between humans and their domes- during all ages under the most different climates ticates in which humans protect and dis- and treatments are simply due to our domestic perse the domesticates, and they provide us productions having been raised under with food and other resources (Ladizinsky, conditions of life not so uniform as, and 1998). Domestication is an evolutionary somewhat different from, those to which the process in which new types are constantly parent species have been exposed under nature being selected to meet human ecology, econ- … we may safely conclude that very many of the omy and culture, thereby distancing them most strongly marked domestic varieties could genotypically and phenotypically from their not possibly live in a wild state. (Darwin, 1859) wild progenitors and making them human- dependent, surviving only under cultivation. Whereas the astounding varieties of cultivars and domesticates generated by humans Genetic Erosion across the globe are evident, the environ- mentally limited area of each cultivar and Humans protect crops (e.g. weed, pest and the vulnerability to biotic and abiotic stresses pathogen control) through cultivation, includ- are notorious, and their turnover including ing ploughing, fertilizing, irrigating, spraying, extinction is extraordinary (Hutchinson and and providing a more uniform environment Weiss, 1999). Only under the favourable for growth. Consequently, crops differentially conditions of human agriculture of fertiliz- and gradually lose, throughout domestication ers, cultivation, spraying, dusting and so on evolution, their natural genetic resistance can they survive under human-created arti- ficial environments. Left alone, they will most probably competitively perish by local superiorly adapted wild species. Our genetic insights into the structure and genomic architecture of the cultivars

Genomic diversity in nature and domestication 291 against ecological abiotic and biotic stresses. tapped in future plant breeding for a variety This is reflected in their increasing vulnerabil- of agronomically useful traits. ity, susceptibility and loss of genetic diversity. Domestication resulted in the dramatic Agricultural Predicament and Plant impoverishment of the gene pools of cultivars Genetic Conservation (Frankel and Bennett, 1970; Frankel and Hawkes, 1975; Harlan, 1975a,b; Frankel and Human agriculture has revolutionized the Soule, 1981; Hawkes, 1983, 1991; IBPGR, planet and human evolution. However, this 1985; Plucknett et al., 1987; Brown et al., unmatched change in the planet and human 1989, 1990; Holden et al., 1993; Frankel et al., societies has had the dramatic cost of envi- 1995; ICPGR, 1995; Gaut et al., 1997; Maxted ronmental deterioration and loss of biodi- et al., 1997; Valdes et al., 1997; Jones et al., versity through extinction (Nevo, 1995b), 1998; Nevo, 1998b; Poncet et al., 1998; including the genetic erosion of domesti- Hammer et al., 1999; Hoisington et al., 1999; cates. In an attempt to save the planet and Xu et al., 2000; Hernandez-Verdugo et al., humankind from the predicament of annihi- 2001; Ellis et al., 2003). (See also Brush lation, massive directed conservation of bio- (1999) for a critical evaluation of the concept diversity and genetic diversity is imperative. of genetic erosion of cultivars.) In contrast, wild plants and wild progenitors and relatives Conservation of biodiversity in nature of crops have been under constant ecological became more intensive during recent decades, stresses, both abiotic (climatic such as solar in an attempt to alleviate the pending extinc- radiation, temperature, drought) and biotic tion of the biosphere by humans (Soule, 1986, (pathogens, pests, competitors). They have 1987; Ehrlich and Wilson, 1991; Wilson, 1992, evolved throughout millions of years of evolu- 2002; Fiedler and Jain, 1992). The salvation of tion adaptive genetic strategies for fitness and biodiversity is very relevant to crop plants and survival. Moreover, generally, they are richer their progenitors. Domestication resulted in a in genetic diversity of proteins and DNA than dramatic impoverishment of the gene pools of their cultivated derivatives (see below). cultivars (see references above). Feeding a still growing world population (with >90% in Therefore, wild relatives of crop plants developing countries) will require an astonish- constitute invaluable gene resources for crop ing increase in food production. Forecasts call improvement and sustain agricultural pro- for wheat to become the most important world ductivity (Vavilov, 1951; Nevo et al., 1979, cereal with maize close behind. Together these 1986, 2002; Plucknett et al., 1987; Nevo, crops will account for 80% of developing coun- 1992; Tanksley and McCouch, 1997). The tries’ cereal import requirements. Access to a next agricultural revolution in the 21st cen- range of genetic diversity is critical to the suc- tury for increasing food production to cope cess of a breeding programme. The global with the exploding world population and effort to assemble and use the genetic increasing starvation in the southern conti- resources is enormous and genetic diversity in nents will be primarily genetic. The best hope nature and gene banks is critical in the world’s for crop improvement resides in the prog- fight against hunger (Hoisington et al., 1999). enitors of cultivated plants that harbour rich genetic resources for tolerance against The accumulated information on biodi- stressful abiotic and biotic stresses. Likewise, versity provides partial answers to the fol- the progenitors harbour rich genetic lowing questions: resources for diverse agronomic traits, pho- tosynthetic yield, high-quality proteins and 1. How much and what types of genetic amino acids, among others. For genetic change were involved in domestication? resources of wild emmer wheat – the prog- 2. How fast was this change? enitor of most cultivated wheats – see Nevo 3. Can the crops be regarded as new species (1983, 1988, 1995a) and Feldman and Sears created by humans? (1981). It is this rich genetic resource that is 4. Could they survive without the ecological so valuable to agriculture and should be niche created by humans?

292 E. Nevo 5. Can they be further improved to rein- research programme); 675 natural popula- force their qualitative and quantitative agri- tions and 15,500 genotypes have been col- cultural traits? lected in central, peripheral and marginal populations plus several Israeli microsites, The following section focuses primarily on the which have been partly analysed genetically first question in relation to genetic erosion of and agronomically (Table 14.1). In all, the wheat and barley during domestication, based programme has revealed a broad view of on genetic studies conducted thus far. In par- wild barley and wild emmer in a substantial ticular, I will give a brief overview of the part of the Fertile Crescent (Nevo et al., 1986, research programme on wild progenitors of 2002; Nevo and Beiles, 1989; Nevo, 1992). crop plants, wild cereals and wild lettuce con- We have also conducted a series of microgeo- ducted at the Institute of Evolution. This pro- graphic population genetics and ecology gramme focuses on genetic diversity in nature studies on allozymes and DNA poly- and the decline under domestication, provid- morphisms of wild cereals in Israel. In ing detailed evidence of genetic erosion in addition, we have studied agronomic traits, these crops and some others. This may bear disease resistance polymorphisms (Moseman directly on the available wild genetic resources et al., 1984; Nevo et al., 1985; Fetch et al., that can be utilized to increase diversity and 2000, 2003a,b; see list on wild cereals at productivity of their cultivars. http://research.haifa.ac.il~evolut), protein content, amino acid composition, herbicide Multidisciplinary Research Programme of resistance, photosynthetic yield, and drought Wild Cereals for Crop Improvement at and salt tolerance. A methodology has been the Institute of Evolution, University of developed, based on ecological factors and Haifa, Israel allozyme markers, for assessing wild cereal genetic resources, enabling us to screen sin- The following review leans heavily on a long- gle- and multiple-trait elite genotypes in term multidisciplinary research programme nature (Nevo, 1987). of wild emmer wheat, T. dicoccoides (Nevo, 1986, 1989, 2001c; Nevo et al., 2002), and In our detailed study of domestication wild barley, Hordeum spontaneum (Harlan and evolution in wheat, we studied wild emmer Zohary, 1966; Nevo, 1992), that has been wheat, T. dicoccoides, which is the progenitor conducted at the Institute of Evolution, of modern tetraploid and hexaploid culti- University of Haifa, Israel, since 1975. The vated wheats (Nevo et al., 2002). Our objec- project is aimed at studying the ecological- tive was to map domestication-related genetic, genomic and domestication evolu- quantitative trait loci (QTL) in T. dicoccoides tion of wild cereals, the progenitors of wheat (Peng et al., 2000, 2003). The studied traits and barley, across the Near East Fertile included brittle rachis, heading date, plant Crescent centre of origin and diversity. The height, grain size, yield and yield compo- multidisciplinary programme has focused on nents. Our mapping population was derived the patterning and maintenance of genetic from a cross between T. dicoccoides and diversity and divergence, genome organiza- Triticum durum. Approximately 70 domestica- tion and agronomically important traits in tion QTL effects were detected, non-ran- natural populations, attempting to highlight domly distributed among and along their adaptive evolutionary processes and chromosomes. Seven domestication syn- potential genetic resources for crop improve- drome factors were proposed, each affecting ment (Fig. 14.1). The programme has 5–11 traits. We showed: (i) clustering and involved macrogeographic population genet- strong effects of some QTLs; (ii) remarkable ics and ecology studies in the Near East genomic association of strong domestication- Fertile Crescent in several countries: Israel, related QTLs with gene-rich regions; and Turkey, Iran and Jordan (including 24 other (iii) unexpected predominance of QTL countries added to our recent wild Lactuca effects in the A genome. The A genome of wheat may have played a more important role than the B genome during domestica-

Genomic diversity in nature and domestication 293 Fig. 14.1. Multidisciplinary long-term research programme of wild cereals at the Institute of Evolution, University of Haifa, Israel. tion evolution. The cryptic beneficial alleles Genetic Resources of Wild Barley in the at specific QTL derived from T. dicoccoides Near East: Structure, Evolution and may contribute to wheat and cereal improve- ment (Peng et al., 2000, 2003, and Fig. 14.2). Comparison with Cultivated Barley and We conducted similar studies in wild barley, Breeding H. spontaneum (Guoxiong et al., 2002). Genetic diversity and structure of popula- Since 1994, we also embarked upon a tions of the wild progenitor of barley, H. spon- world collection and analysis of wild lettuce taneum, from three countries (Israel, Turkey species, primarily the progenitor of lettuce, and Iran) in the Near East Fertile Crescent Lactuca serriola, in collaboration with were compared and contrasted (Nevo et al., Richard Michelmore of Davis, California. 1986). The analysis was based on allozymic Our current collection of wild lettuce variation in proteins encoded by 27 shared includes 252 populations and 2876 geno- loci in 2125 individuals representing 52 pop- types from 21 countries (Sicard et al., 1999; ulations of wild barley. The results indicated Kuang et al., 2004a,b).

294 E. Nevo Table 14.1. Gene bank collections at the Institute of Evolution, University of Haifa, Israel, used for genetic and agronomic research programmes. Species Countries Populations Genotypes Hordeum spontaneum (wild barley) 5 133a 3,393 Triticum dicoccoides (wild emmer wheat) 6b 32c 1,650 Aegilops species 3 6,600 Lactuca species (wild lettuce) 21 220 2,876 Avena (wild hexaploid oats) 1 252 Total 36 (24)d 985 38 15,504 675 aIncluding three populations of Tabigha, four populations of Newe Yaar and seven populations of ‘Evolution Canyon’ (see Nevo, 2001b). bIncluding Iran, Syria and Iraq, which are represented by a few genotypes each. cExcluding Iran, Syria and Iraq. dCountries involved in the overall count (36) and the actual number of countries considering overlap of species (24). that: (i) H. spontaneum in the Near East Fertile tors and molecular markers as effective pre- Crescent is very variable genetically; (ii) dictive guidelines. genetic differentiation of populations includes some clinal but primarily regional An intriguing question is: what is the pat- and local patterns often displaying sharp geo- tern of DNA polymorphisms of H. sponta- graphic differentiation over short distances; neum? Are they neutral as widely presumed, (iii) the average relative genetic differentia- or do they display, like allozymes, ecological tion (Gst) was 54% within populations, 39% correlates and are they also subjected to nat- among populations and 8% among the three ural selection? We elucidated this issue by countries; (iv) allele distribution is character- studying microsatellite, amplified fragment ized by a high proportion of unique alleles length polymorphisms (AFLP) and riboso- (51%) and a high proportion of common alle- mal DNA (rDNA) diversities in wild cereals. les that are distributed either locally or spo- To answer this question we analysed radically; (v) discriminant analysis by allele microsatellite molecular markers, also frequencies successfully clustered wild barley known as simple sequence repeats (SSRs). of each of the three countries (96% correct SSRs are short (1–6 bp), tandemly repeated classification); (vi) a substantial portion of the DNA sequences and are highly polymorphic patterns of allozyme variation in the wild as a result of frequent variation in the num- gene pool is significantly correlated with the ber of times the core sequence is repeated environment and is predictable ecologically, (Li et al., 1999). chiefly by a combination of humidity and temperature variables; (vii) natural popula- Ecological-genomic Diversity of tions of wild barley are, on average, more Microsatellites in Wild Barley, H. variable than two composite crosses and land- spontaneum, Populations in Israel and races of cultivated barley (Fig. 14.2). The spa- tial patterns and environmental correlates Jordan and predictors of genetic variation of H. spon- taneum in the Fertile Crescent indicate that Israel genetic variation in wild barley populations is not only rich, but at least partly adaptive and Microsatellite diversity at 18 loci was analysed predictable by ecology and allozyme markers. in 94 individual plants of ten wild barley, H. Consequently, conservation and utilization spontaneum (C. Koch) Thell., populations sam- programmes should optimize sampling pled from Israel across a southward transect of strategies by following ecological-genetic fac- increasing aridity (Turpeinen et al., 2001). Allelic distribution in populations was non-

Genomic diversity in nature and domestication 295 Fig. 14.2. Comparison of genetic indices among wild barley, H. spontaneum (from Iran, Turkey and Israel), and landraces (from Iran and Europe) and composite crosses of H. vulgare, based on 19 shared loci. Abbreviations: A = average number of alleles per locus; P = proportion of polymorphic loci per population; He = genic diversity. Twenty-nine accessions from the collection of Landbrugets Plantekulturs Sortssamling, Denmark, that includes European landraces, plus one from India and two from Egypt. random. Estimates of mean gene diversity were graphical variables to three principal compo- highest in stressful arid-hot environments. nents: water factors, temperature and geogra- Sixty-four per cent of the genetic variation was phy. At three loci, stepwise multiple regression partitioned within populations and 36% analysis significantly explained the gene diver- between populations. Associations between sity by a single principal component (water fac- ecogeographical variables and gene diversity, tors). Based on these observations, it was He, were established for nine microsatellite suggested that simple sequence repeats are at loci. By employing principal component least partly subjected to natural selection. (For analysis, we reduced the number of ecogeo- adaptive random amplified polymorphic DNA

296 E. Nevo (RAPD) variation in H. spontaneum, see also Near East Fertile Crescent Volis et al., 2001.) Similar patterns of H. sponta- neum, in Israel, were obtained by AFLP analy- AFLP diversity in H. spontaneum in the Fertile sis (Turpeinen et al., 2003). Crescent was based on 39 genotypes, 268 AFLP loci, of which 204 proved to be poly- Jordan morphic. Genotypes were grouped accord- ing to the area of origin. Shoot Na+ content We also analysed the ecological-genomic and carbon isotope (␦13) reflecting drought diversity of microsatellites of wild barley, H. resistance was associated with site of origin spontaneum, at 18 loci in 306 individuals of 16 ecogeographic data (Pakniyat et al., 1997). populations from Jordan across a southward transect of increasing aridity (Baek et al., Genetic Diversity of Wild Emmer Wheat 2003). The 18 microsatellites revealed a total in Israel and Turkey of 249 alleles, with an average of 13.8 alleles per locus (range 3–29), with non-random dis- Structure, evolution and application in tribution. The proportion of polymorphic loci breeding per population averaged 0.91 (range 0.83–1.00); gene diversity, He, averaged 0.512 Allozyme variation in the tetraploid wild (range 0.38–0.651). We compared the num- emmer wheat, T. dicoccoides, the progenitor ber of alleles of the 18 loci with those found in of all cultivated wheats, was studied for pro- Israel populations by Turpeinen et al. (2001). teins encoded by 42 gene loci in 1815 plants Out of the 280 alleles, 138 (49.3%) were representing 37 populations – 33 from Israel unique (i.e. occurred in only one of the coun- and four from Turkey – sampled in 33 locali- tries). The percentage of unique alleles in ties from 1979 to 1987 (Nevo and Beiles, Jordan and Israel populations was 43.0% and 1989). The results showed the following: 17.9%, respectively, suggesting that Jordan together with Israel is an important centre of 1. Six loci (14%) were monomorphic in all origin and diversity of wild barley (see also the populations, 15 loci (36%) were locally poly- 400 AFLP polymorphic loci analysis of Badr et morphic and 21 loci (50%) were regionally al., 2000). Estimates of mean gene diversity polymorphic. These results are similar to were highest in the populations collected near those obtained earlier on 12 Israeli popula- the Golan Heights, such as Shuni North, tions (Nevo et al., 1982). All polymorphic Shuni South and Jarash. Sixty-nine per cent loci (except four) displayed high local levels of the microsatellite variation was partitioned of polymorphism (у10%). within populations and 31% between popula- 2. The mean number of alleles per locus, A, tions. Associations between ecogeographical was 1.252 (range: 1.050–1.634); the propor- variables and gene diversity values were estab- tion of polymorphic loci per population lished for eight microsatellite loci. averaged 0.220 (range: 0.050–0.415); gene Remarkably, the cluster produced by SSR diversity, He, averaged 0.059 (range: data of wild barley coincides with the result of 0.002–0.119). the dendrogram of the Spalax ehrenbergi 3. Altogether there were 119 alleles at the superspecies of blind subterranean mole rats 42 putative loci tested, 114 (96%) of these in in Jordan based on allozyme gene loci. The Israel. major soil type in the wild barley habitat of 4. Genetic differentiation was primarily each ecological group was different. Stepwise regional and local, not clinal; 70% of the multiple regression analysis indicated that the variant alleles were common (у10%) and variance of gene diversity was explained by not widespread, but rather localized or spo- altitude (i.e. temperature; R2 = 0.362, P < radic, displaying an ‘archipelago’ population 0.01). These observations suggest that genetics and ecology structure. The coeffi- microsatellites are at least partly adaptive and cients of genetic distance between popula- subject to natural selection. tions were high and averaged D = 0.134 (range: 0.018–0.297), an indication of sharp genetic differentiation over short distances.

Genomic diversity in nature and domestication 297 5. Discriminant analyses differentiated found to be highly effective in distinguishing Israel into central and three marginal genotypes of T. dicoccoides originating from regions as well as those with different soil- diverse ecogeographical sites in Israel and type populations (Fig. 14.3). Turkey, with 95.5% of the 100 genotypes 6. Allozymic variation was 40% within and correctly classified into sites of origin by dis- 60% between populations. criminant analysis based on RAPD genotyp- 7. Gametic-phase disequilibria were abun- ing (Fig. 14.4a). However, interpopulation dant, their number being positively corre- genetic distances showed no association with lated (rs = 0.60, P<0.01) with the humidity. geographic distance between the population 8. Multilocus organization (Brown et al., sites of origin, negating a simple isolation by 1980) was substantive and also positively distance model. Spatial autocorrelation of correlated with humidity. RAPD frequencies suggests that migration is 9. Allozyme diversity, overall and at single not influential. Our present RAPD results loci, was significantly correlated with, and are non-random and in agreement with the partly predictable by, climatic and edaphic previously obtained allozyme patterns (Nevo factors. and Beiles, 1989) although the genetic 10. The distribution of the significant posi- diversity values obtained with RAPDs are tive and negative values and the absence of much higher than the allozyme values. autocorrelations in the correlogram Significant correlates of RAPD markers with revealed no similar geographic patterns various climatic and soil factors suggest that, across loci, eliminating migration as a prime as in the case of allozymes, natural selection factor of population genetic differentiation. causes adaptive RAPD ecogeographical dif- ferentiation, also in non-coding sequences. These results suggest that: (i) during the The results obtained suggest that RAPD evolutionary history of wild emmer, diversi- markers are useful for the estimation of fying natural selection, through climatic and genetic diversity in wild material of T. dicoc- edaphic factors, was a major agent of genetic coides and the identification of suitable par- structure and differentiation at both the sin- ents for the development of mapping gle and multilocus levels; and (ii) wild populations for tagging agronomically emmer harbours large amounts of genetic important traits derived from T. dicoccoides. diversity exploitable as genetic markers in Similar differentiation in T. dicoccoides, in sampling and abundant adaptive genetic Israel, was achieved by microsatellite analy- resources utilizable for wheat improvement. sis (Fahima et al., 1999; Fig. 14.4b). In addition, we conducted DNA studies (RAPD and SSR) on wild emmer wheat Microgeographic Critical Tests in Nature: (Fahima et al., 1999, 2002). Evaluation of Levels and Divergence of RAPD Polymorphism of Wild Emmer Genetic Diversity Wheat Populations, T. dicoccoides, in Microsite ecological contrasts are excellent Israel critical tests for evaluating the dynamics of genome and phenome evolution and assess- Genetic diversity in RAPDs was studied in ing the relative importance for adaptation 110 genotypes of the tetraploid wild progen- and speciation of the evolutionary forces itor of wheat, T. dicoccoides, from 11 popula- causing differentiation (Nevo, 2001b). The tions sampled in Israel and Turkey (Fahima latter involve mutation (in the broadest et al., 1999). Our results show a high level of sense, including recombination), migration, diversity of RAPD markers in wild wheat chance and selection. At a microsite, muta- populations in Israel. The ten primers used tion, which is usually considered a clockwise in this study amplified 59 scoreable RAPD neutral process, is expected to be similar loci of which 48 (81.4%) were polymorphic across the microsite. Migration, which oper- and 11 monomorphic. RAPD analysis was ates for any organism at the microsite, even

Fig. 14.3. Discriminant analysis between ecogeographical, edaphic and population size populations in Israel and Turkey; (d–f) the pattern of 29 populations from Israel (excludin four types of marginal and one type of central populations: 1 = cold steppe; 2 = west m centroid. (b) and (e) discriminate between small-, medium- and large-size populations: p and (f) discriminate between populations growing on three soil types: 1 = terra rossa; 2

298 E. Nevo e categories in T. dicoccoides in Israel and Turkey; (a–c) the pattern of 37 ng the four micropopulations of Tabigha). (a) and (d) discriminate between margin; 3 = east and south margin; 4 = Turkey; 5 = central population; * = population size: 1 = small; 2 = medium; 3 = large; * = group centroid. (c) = rendzina; and 3 = basalt; * = group centroid (from Nevo and Beiles, 1989).

Genomic diversity in nature and domestication 299 (a) (b) Fig. 14.4. (a) Plot of canonical discriminant functions 1, 2 and 3 based on the 48 polymorphic RAPD loci of T. dicoccoides. Symbols: 1 = Mt Hermon; 2 = Gamla; 3 = Rosh Pinna; 4 = Tabigha; 5 = Mt Gilboa; 6 = Mt Gerizim; 7 = Gitit; 8 = Kokhav Hashahar; 9 = Bat-Shelomo; 0 = Givat-Koach; A = Diyarbakir. The proportion of the eigen values of functions 1, 2 and 3 from the sum of the eigen values of the ten discriminant functions are 36.9%, 16% and 12%, respectively (Fahima et al., 1999). (b) Plot of canonical discriminant functions 1, 2 and 3 based on ten polymorphic microsatellite loci of 15 populations of wild emmer wheat, T. dicoccoides, from Israel and Turkey. Symbols: H = Hermon; Y = Yehudiyya; M = Gamla; R = Rosh Pinna; T = Tabigha; B = Mt Gilboa; Z = Mt Gerizim; G = Gitit; K = Kokhav Hashahar; J = J’aba; A = Amirim; B = Bet-Oren; S = Bat-Shelomo; V = Givat-Koach; D = Diyarbakir (Fahima et al., 2002).

300 E. Nevo in sessile organisms, is expected to homoge- repetitive DNA elements, hence providing nize allele frequencies. Stochasticity is not comprehensive coverage of the wild wheat expected to result in repetitive, ecologically and barley genomes. At each microsite, we correlated patterns. Selection seems to be identified non-random divergence of the only evolutionary force expected to allozyme, RAPD and SSR diversities. result in repeated ecologically correlated Extensive studies examined allozyme, RAPD and parallel patterns and contrasts. and SSR and rDNA interslope diversity of H. spontaneum in ‘Evolution Canyon’ I in Mt In 1977, at the Institute of Evolution, we Carmel, Israel, where a mesic ‘European’ embarked on a series of microsite studies slope faces the xeric ‘African’ slope at an comparing diverse, sharply contrasting eco- average distance of 200 metres. Significant logical alternative patterns including tem- niche-specific (high frequency in niche type) peratures, aridity index, lithology, soil types, and niche-unique (limited to a niche type) topography, and chemical and organic pol- alleles and linkage disequilibria abounded, lutants in marine organisms. The aforemen- in all microsites in both wild cereals, barley tioned studies demonstrated differential and wheat, allowing classification into niches viability of allozyme and DNA genotypes of either coded or non-coded markers (ref- where diversity and divergence were erences 29–33 in Nevo, 2001b, and addi- selected at a microscale or under critical tional studies in 2002 and 2003; see our empirically contrasting conditions and website http://research.haifa.ac.il/~evolut/). ecologies. Will the non-coding genome also For adaptive RAPD variation in H. sponta- display ecological correlates at regional and neum, see also Volis et al. (2001, 2002). local scales? The answer is, emphatically, yes for outbreeding mammals (Nevo, 2001b) In wild emmer wheat at Ammiad, the and inbreeding wild cereals (reviewed here). three marker systems used in wild wheat showed dramatically different levels of gene Microscale Molecular Population diversity (He) and genetic distance (D): SSR > Genetics of Wild Cereals at Four Israeli RAPD > allozymes. The gene differentiation (Gst) order was allozymes > SSR > RAPD. Microsites Remarkably, the three marker systems revealed similar trends of diversity and diver- We used three molecular marker systems gence. All three molecular markers displayed that included allozymes, RAPDs and non-random allele distributions, habitat-spe- microsatellites (SSRs) to detect molecular cific and habitat-unique alleles, and linkage diversity and divergence in three popula- disequilibria. The subpopulations in the drier tions of wild emmer wheat (T. dicoccoides). habitats showed higher genetic diversities in These populations were from Ammiad, the three marker systems. The genetic dis- Tabigha and Yehudiyya microsites in north- tances among the four subpopulations ern Israel, and these microsites displayed tended to increase with the difference in soil topographic, edaphic and climatic ecological moisture after the early rain of the growing contrasts, respectively (references 29–33 in season. These results may suggest that eco- Nevo, 2001b). Likewise, we examined mole- logical selection, probably through aridity cular diversity with allozymes, RAPD and stress, acts both on structural protein coding SSR markers in wild barley, H. spontaneum, and on presumably partially regulatory non- in the Tabigha microsite north of the Sea of coding DNA regions (SSR, RAPD and AFLP) Galilee and in Newe Yaar, Lower Galilee resulting in microscale adaptive patterns. (Gupta et al., 2002); the latter microsite rep- Similar microscale molecular (allozymes, resented a mosaic of microniches of sun, RAPDs and SSRs) divergence was found in shade, rock, deep soil and their combina- three populations of wild barley (H. spontaneum) tions. The three marker systems represented at Tabigha, Newe Yaar (Owuor et al., 2003) and protein-coding (allozyme) and DNA non- for ‘Evolution Canyon’ (Owuor et al., 1997). coding and coding (most of RAPDs) regions Ribosomal DNA diversity displayed dramatic plus short (most SSRs) and long (rDNAs) differences in two microsites associated with

Genomic diversity in nature and domestication 301 soil and microclimatic divergence (Gupta et resents the allelic diversity maintained in the al., 2002). A dramatic genetic separation in domestic gene pool, is compared. When this wild emmer wheat occurred in the shady and comparison is made for the four esterase sunny microniches, separated by only a few loci, Est 1–5, the differences between the metres, as shown in Fig. 14.5. wild and cultivated gene pools in number of alleles are: Est-1 (8,7); Est-2 (17,12); Est-4 Genetic Diversity in Wild Barley versus (10,4); Est-5 (6,6). These results emphatically Cultivated Barley demonstrate the potential role of the wild progenitor as well as that of the landraces as Allozymes valuable genetic resources for plant breed- ing, and the depletion of allelic diversity in A comparison was made of total allozymic the cultivated germplasm. Importantly, a diversity of 19 shared loci of wild barley in comparison between wild barley from Israel Iran, Turkey and Israel with landraces of and two composite crosses of cultivated bar- cultivated barley from Iran and a collection ley indicates more multilocus organization in of European landraces. The comparison of the wild than in the cultivated germplasm two accessions from Egypt and one from (Brown et al., 1980). India, and two composite crosses is shown in Table 14.2 and Fig. 14.2. Allozymic diversity Microsatellite Comparisons in Wild and was lowest in the composite crosses, inter- Cultivated Barley mediate in the landraces and highest in the wild germplasm in the order: Israel > Maroof et al. (1994) examined the extent of Turkey > Iran (Fig. 14.2, Tables 14.2 and genetic variation in barley SSRs and studied 14.3) for an analysis of three overall genetic the evolutionary dynamics of SSR alleles. SSR indices, A (allelic diversity), P (genetic poly- polymorphisms were resolved by the poly- morphism) and He (genetic diversity) as well merase chain reaction of four pairs of primers. as for the analysis of diversity in each of the In total, 71 variants were observed in a sample 19 gene loci (Table 14.3). These results of 207 accessions of wild and cultivated barley. appear to hold even if the world collection Analyses of wheat–barley addition lines and of cultivated barley, which presumably rep- barley-doubled haploids identified these vari- Fig. 14.5. Histogram of frequencies of canonical scores for wild emmer wheat, T. dicoccoides, and sunny niches in Yehudiyya according to 25 polymorphic RAPD loci (Li et al.,1999).

302 E. Nevo Table 14.2. Comparison of genetic indices among H. spontaneum, landraces and composite crosses of H. vulgare based on 19 shared loci (from Nevo et al., 1986). NA P He H. vulgare Composite cross XXI (F17) 30 1.58 0.42 0.099 Composite cross XXXIV (F4) 30 1.68 0.37 0.139 Landraces Europeana 178 2.26 0.47 0.146 Iranian 290 2.21 0.74 0.164 H. spontaneum Iran 309 2.58 0.68 0.218 Turkey 437 2.63 0.74 0.226 Israel 1179 4.11 1.00 0.228 aTwenty-nine accessions from the collection of Landrugets Plantekulturs Sortssamling, Denmark, which include European landraces, plus one from India and two from Egypt. Table 14.3. Comparison of numbers of alleles and total diversity (Ht) among H. spontaneum and H. vulgare (from Nevo et al., 1986). H. spontaneum H. vulgare (wild barley) (cultivated barley) Israel Turkey Iran Near East Iranian European Ht Allele Ht Allele Ht Allele Ht Allele Ht Allele Ht Allele Locus 0.86 15 0.72 6 0.61 7 0.83 17 0.38 3 0.28 6 Est-2 0.66 4 0.63 5 0.48 4 0.73 8 0.57 4 0.47 4 Acph-3 0.70 7 0.63 4 0.62 4 0.70 0.45 3 0.58 4 Est-4 0.53 6 0.64 5 0.57 4 0.59 10 0.01 2 0.40 4 Est-5 0.54 4 0.48 2 0.29 3 0.47 6 0.50 3 0.01 2 Nadh-1 0.24 3 0.52 3 0.17 2 0.29 4 0.04 2 0.00 1 Adh-1 0.13 6 0.18 3 0.44 4 0.25 3 0.44 3 0.48 3 Est-1 0.03 2 0.28 2 0.30 2 0.17 8 0.00 1 0.00 1 Aat-1 0.13 3 0.17 4 0.22 4 0.16 3 0.19 3 0.00 1 Pgt 0.18 3 0.02 2 0.004 2 0.10 5 0.01 2 0.00 1 Pgm 0.11 3 0.04 2 0.11 3 0.10 3 0.00 1 0.00 1 Adh-2 0.01 2 0.01 2 0.34 2 0.09 5 0.15 2 0.00 1 Cat 0.08 4 0.02 3 0.004 2 0.05 3 0.00 1 0.00 1 Mdh-2 0.07 3 0.00 1 0.00 1 0.04 7 0.05 2 0.00 1 Gdh 0.03 4 0.00 1 0.00 1 0.02 3 0.02 2 0.04 2 Aat-2 0.02 3 0.00 1 0.00 1 0.02 4 0.21 3 0.46 4 Gpgd-2 0.01 2 0.00 1 0.00 1 0.01 3 0.00 1 0.00 1 Acph-1 0.01 3 0.00 1 0.00 1 0.00 2 0.00 1 0.07 6 Gpgd-1 0.01 2 0.002 2 0.00 1 0.00 3 0.11 2 0.00 1 Mdh-1 2 Twenty-nine accessions from the collection of Landbrugets Plantckulturs Sortssamling, Denmark, which include European landraces plus one Indian and two from Egypt. ants (alleles) with four loci, each located on a sequences. The numbers of alleles at two loci different chromosome. The numbers of alleles were 28 and 37. Three alleles were resolved by detected at a locus corresponded to the num- each of the other two loci. Allelic diversity was ber of nucleotide repeats in the microsatellite greater in wild than in cultivated barley and in

Genomic diversity in nature and domestication 303 surveys of two generations (F8 and F53) of Genetic Diversity of Wild Emmer versus Composite Cross II (Table 2 in Maroof et al., Cultivated Wheat 1994). An experimental population of culti- vated barley showed that few of the alleles pre- In the regional wild wheat study (Fahima et sent in the 28 parents survived into generation al., 2002) using 20 GWM (Gatersleben wheat F53 whereas some infrequent alleles reached microsatellites), 363 alleles were revealed high frequencies (Table 15.3 in Maroof et al., among 135 wild wheat accessions, an average 1994). Such changes in frequency indicate that of 18 alleles per GWM. In a previous study, the chromosomal segments marked by the also using GWM markers, Plaschke et al. SSR alleles are under the influence of natural (1995) revealed a total of 142 alleles among 40 selection. The SSR variants allow specific DNA cultivated wheat lines, an average of 6.2 alleles sequences to be followed through generations. per GWM (but see Huang et al., 2002, where Thus, the great resolving power of SSR assays an extensive study showed high SSR diversity may provide clues regarding the precise tar- varying among A, B and D genomes). The gets of natural and human-directed selection results obtained in the wild emmer study (Maroof et al., 1994). See also the comparison demonstrate the high diversity in microsatel- of rDNA in wild and cultivated barley in lite sequences among T. dicoccoides accessions Maroof et al. (1984) and Gupta et al. (2002), compared with the cultivated germplasm cor- and in wild and cultivated rice in Liu et al. roborating RAPDs (Fahima et al., 1999) and (2002). allozyme markers (Nevo and Beiles, 1989). Clearly, genetic diversity was eroded across Using extensive chromosomally mapped both the coding and non-coding genome dur- multiallelic microsatellites, the SCRI ing the domestication of major cereal crops, as (Scottish Crop Research Institute, www.scri. was displayed by allozymes (Nevo et al., 1986), sari.ac.uk) showed, likewise, that allelic RAPD and SSR loci, paralleling the genetic diversity in H. spontaneum far exceeds that of erosion shown earlier in wild barley proteins Hordeum vulgare. A major bottleneck associ- and DNA. This contributed to the susceptibil- ated with the domestication of barley (H. ity and vulnerability of the wheat cultivars to vulgare) from wild barley (H. spontaneum) has abiotic and biotic stresses. been identified (see Fig. 14.6 and other SCRI web figures). Large-scale sequencing The dendrogram presented in Fig. 15.4.6 projects can now compare and contrast in Nevo et al. (2002) demonstrates the ability thousands of gene sequences (Ests) in wild of microsatellites developed from Triticum aes- and cultivated germplasm. Wild relatives Cultivated gene pool SSR 1 2 3 4 Old cultivars Modern cultivars SSR 1 2 3 4 SSR 1 2 3 4 Fig. 14.6. Domestication: reduction of SSR alleles from wild germplasm to old and modern cultivars (from www.scri.sari.ac.uk).

304 E. Nevo tivum sequences to detect a large amount of (RGC2) in wild species confirmed the pres- genetic diversity in wild emmer wheat and to ence of numerous chimeric RGC2 genes in identify intergroup differences clearly identi- nature (Kuang et al., 2004a,b) and the high fying central vs. peripheral populations. All AFLP polymorphism of the progenitor of the examined T. dicoccoides populations were cultivated lettuce, Lactuca serriola (Kuang et distinguishable by the GWM markers used, al., 2004a,b). even among closely related populations originating from close geographic locations. Evidence on Domestication Genetics and Our results demonstrated that the SSR DNA Evolution of other Crops polymorphism of wild emmer wheat was cor- related with macro- and microscale ecogeo- Additional evidence on domestication genet- graphic factors. In particular, the Israeli ics and evolution and the depletion of populations exhibited high interpopulation genetic diversity under domestication has and interregional polymorphism. These been reported, generally, in: observations are consistent with previous results obtained with isozymes and different ● Plants (Tripp and van der Heide, 1996; DNA markers for different collections of wild Tanksley and McCouch, 1997; emmer wheat covering a much wider geo- Hoisington et al., 1999; Sandhu and Gill, graphic range. They demonstrate geographic 2002; Feuillet and Keller, 2002; and genomic congruence and continuity www.gramene.org) from macro- to microscales and from coding (protein) to non-coding (presumably partly ● Oilseed sunflower, Helianthus annuus regulatory) DNA (Nevo et al., 1982; Nevo and (Burke et al., 2002; Tang and Knapp, Beiles, 1989; Li et al., 2002; and all our 2003) microsite studies in wild emmer; Li et al., 1999; see citations in Nevo et al., 2002, and ● Barley, H. vulgare (Parzies et al., 2000), also in http://research.haifa.ac.il/~evolut). and wild barley, landraces and cultivars (Lin et al., 2001; SCRI at Molecular Diversity at the Major Cluster www.scri.sari.ac.uk; Neale et al., 1988) of Disease-resistance Genes in Cultivated ● Wheat, T. aestivum (Huang et al., 2002) and Wild Lactuca Species ● Maize, Zea mays (Matsuoka et al., 2002; We conducted the first study on diversity at Vigouroux et al., 2002 and Zhang et al., the major resistance gene cluster of Lactuca 2002), and Cucurbita (Sanjur et al., 2002) species germplasm using molecular markers ● Rice, Oryza sativa (Morishima and Oka, derived from resistance genes of the NBS– 1970; Yang et al., 1994; Liu et al., 1996; LRR type (nucleotide binding site (NBS) Akimoto et al., 1999; Xiong et al., 1999; and a leucine-rich repeat region (LRR)) to Zhu et al., 2000) downy mildew caused by the fungus Bremia ● Cereals (Gale and Devos, 1998; Buckler et lactucae (Sicard et al., 1999). Large numbers al., 2001; Kellogg, 2001; Appels et al., of haplotypes were detected in three wild 2002) Lactuca species indicating the presence of ● Diverse crops (Zohary, 1999) numerous resistance genes in wild species ● Melon, Cucumis melo (Christopher J. compared with cultivated lettuce, Lactuca Duxler at www.genglab.ucdavis.edu/chris) sativa (see Table 15.5 in Sicard et al., 1999), ● Beans, Phaseolus lunatus (Mimura et al., exposing the distinct variation that was 2000; Beebe et al., 2001; Fofana et al., found within wild populations. The large 2001) number of wild haplotypes recommends the ● Cowpea, Vigna unguiculata (Coulibaly et use of wild germplasm in protecting culti- al., 2002) vated lettuce against pathogens. Additional ● Soybeans (Maughan et al., 1995, 1996) in-depth studies of resistant gene candidates and millet landraces, Pennisetum glaucum (Bhattacharjee et al., 2002) ● Potato, Solanum tuberosum (Ortiz and Huaman, 2001).

Genomic diversity in nature and domestication 305 Conclusions and Prospects ural populations at Qazrin, Ammiad, Tabigha and Yehudiyya microsites (wild Wheat and barley as model organisms in emmer wheat), and three natural popula- domestication tions of wild barley at Tabigha, Newe Yaar and ‘Evolution Canyon’ in northern Israel, Wheat and barley are important model display parallel ecological-genetic pattern- organisms for testing various aspects of evo- ing. The regional and local results demon- lutionary theory (both for speciation and strated significant spatial and temporal adaptation) and a major source of human molecular divergence at the DNA and pro- nutrition. Wheat speciation involves a poly- tein levels in the wild cereal populations and ploid series (2x, 4x, 6x). The origin of most subpopulations. Specifically, these patterns wheat is wild emmer, T. dicoccoides (genome revealed the following: AABB). Wild barley is a diploid (2n = 14) and the progenitor of all barleys. In 1975, 1. Significant genetic diversity and diver- the Institute of Evolution at the University gence exist at single-, two- and multilocus of Haifa established a long-term multidisci- structures of allozymes, RAPD, AFLP, rDNA plinary research programme to study the and SSRs both regionally within and genomics of wild cereal (http://research. between populations, but, most importantly, haifa.ac.il/~evolut) in the origin and diver- also locally, over very short distances of sev- sity centre of Old World agriculture in the eral to a few dozen metres in the six Near East Fertile Crescent (Zohary and microsites. Hopf, 2000). The programme includes evo- 2. The rich genetic patterns across coding lutionary ecological genetics and barley (allozymes) and largely non-coding (RAPDs, genomics and genetic mapping coupled with AFLP and SSRs, but see Morgante et al., the exploration of genetic resources for 2002) genomic regions are correlated with, wheat and barley improvement. Both and predictable by, environmental stress (cli- aspects, the theoretical and the applied, matic, edaphic, biotic) and heterogeneity have proved to be of great importance for (the niche-width variation hypothesis), dis- studying evolutionary theory and cereal playing significant niche-specific and -unique crop improvement. alleles and genotypes. 3. The genomic organization of wild cereals Conclusion: Genetic Diversity of Wild is non-random, heavily structured and, at Barley and Wild Emmer for Barley and least partly if not largely, adaptive. It defies explanation by genetic drift, neutrality, or Wheat Improvement near neutrality models as the primary dri- ving forces of wild cereal molecular evolu- The present review of wild cereals supports tion. The only viable model to explain the the idea that the highest hope for future genomic organization of wild cereals is nat- crop improvement lies in rationally and ural selection, primarily diversifying, balanc- effectively exploring and exploiting the rich ing and cyclical selection over space and gene pool of the plant’s wild relatives. What time according to the two- or multiple-niche is the role that wild relatives can play in crop ecological models (see Nevo et al., 2002). improvement? This question is of great Spatial models are complemented by tempo- importance in view of the dramatic reduc- ral models of genetic diversity and change tion in genetic diversity and consequent (Kirzhner et al., 1995, 1996, 1999). Natural increased vulnerability to abiotic and biotic selection may interact with mutation, migra- stresses of some of the prime food crops for tion and stochastic factors but it overrides humans. them in orientating wild cereal evolutionary processes. The molecular diversity and divergence of wild emmer wheat, regionally in the Near Based on mathematical modelling, we East Fertile Crescent and locally in four nat- established that stabilizing selection with a cyclically moving optimum could efficiently

306 E. Nevo protect polymorphism for linked loci, addi- 2000; Zohary and Hopf, 2000; Gopher et al., tively affecting the selected trait (Kirzhner et 2002; Salamini et al., 2002). Particularly in al., 1996; Korol et al., 1998). In particular, Israel with its extraordinary biotic and phys- unequal gene action and/or dominance ical diversity, wild emmer (Nevo et al., 2002) effects may lead to local polymorphism sta- and wild barley in the Near East (Nevo et al., bility with substantial polymorphism attract- 1986; Nevo, 1992) developed, both within ing domain. Moreover, under strong cyclical and between populations, a wide range and selection, complex dynamic patterns were rich adaptive diversity to multiple diseases, revealed including ‘supercycles’ (with peri- pests and ecological stresses over a long evo- ods comprising hundreds of environmental lutionary history. Most importantly, this oscillation periods) and ‘deterministic chaos’ diversity is neither random nor neutral. In (Kirzhner et al., 1995, 1996, 1999; Korol et contrast, it displays at all levels, adaptive al., 1998). These patterns could substantiate genetic diversity for biochemical, morpho- polymorphism in natural populations of logical and immunological characteristics wild cereals and increase genetic diversity which contribute to the species’ ability to over long periods, thereby contributing to adapt to widely diverse climatic and edaphic overcoming massive extinctions in natural conditions by diverse and complex fitness populations (Nevo, 1995b). syndromes. The long-lasting co-evolution of wild emmer with parasites and the ecologi- Unique population genetic structure and cally heterogeneous abiotic nature of Israel centre of origin of wild emmer wheat and and the Near East Fertile Crescent led to the development of single multiallelic genes, wild barley multilocus structures and abiotic/biotic stress genomes locally and regionally co-adapted Primarily, wild emmer wheat and, secondar- for both short- and long-term survival. ily, wild barley have unique ecological- genetic structure. Emmer wheat central Genetic Resources for Cereal Improvement populations in the catchment area of the upper Jordan Valley, and wild barley popu- Wild barley and wild emmer wheat are rich lations in the Golan Heights, eastern Galilee in genetic resources and represent the best and Jordanian Mountains, are massive and hopes for enriching the genetically impov- lush and represent their centre of origin erished cultivars and advancing cereal and diversity. However, southwards in Israel improvement. These include abiotic (e.g. and northwards into Turkey, wild emmer drought, cold, heat and salt) tolerances and becomes fragmented into sporadic semi-iso- biotic (viral, bacterial and fungal) and herbi- lated and isolated populations that are char- cide resistances, high quantity and quality acterized by an archipelago genetic storage proteins, hordeins (glutenins and structure in which alleles are built up locally gliadins), amylase and photosynthetic yield in high frequency, but are often missing in (all descriptions and citations in Nevo, neighbouring localities. This phenomenon 1992; Nevo et al., 2002). A small fraction of may even occur in the central continuous these resources have already been used for populations in which alternative fixation of generating disease resistance cultivars in the up to eight alleles was described over tens to USA and Europe. Most of these resources hundreds of metres in the Golan Heights are as yet untapped and provide potentially between Qazrin and Yehudiyya (see Section precious sources for barley and wheat 5.1 in Nevo et al., 2002). improvement (Nevo, 1989, 2001c). The cur- rent rich genetic map of T. dicoccoides with The centre of origin and diversity of wild 549 molecular markers and 48 significant emmer and wild barley, the progenitors of QTLs for 11 traits of agronomic importance most wheat and barleys, and that of other (Peng et al., 2000, 2003), as well as the pre- progenitors of cultivated plants is the Near viously established association between mol- East Fertile Crescent (Nesbitt and Samuel, 1998; Badr et al., 2000; Lev-Yadun et al.,

Genomic diversity in nature and domestication 307 ecular markers and disease resistance (Nevo including Hordeum, Triticum and Aegilops 1987, 1992 and references therein; Fetch et species, and other Triticeae, the rich reper- al., 2003a,b), permit the unravelling of ben- toire of R-genes in wild lettuce, and other eficial alleles of candidate genes that are diverse progenitors of cultivars. One exam- otherwise hidden. These beneficial alleles ple of QTL mapping as a basis for identify- could be introduced into cultivated barley ing candidate genes for wheat improvement and wheat (simultaneously eliminating is shown in Fig. 14.7. For a general review, agronomically undesirable alleles) by using see Mauricio (2001). the strategy of marker-assisted selection. The genetic programmes of wild cereals Specifically, we could divide the prospects conducted at the Institute of Evolution, somewhat arbitrarily into theoretical and University of Haifa, and elsewhere, con- applied perspectives (Nevo et al., 2002). For a firmed that H. spontaneum and T. dicoccoides general review, see Kush (2001). are very valuable wild germplasm resources for future barley and wheat improvement. Theoretical Perspective This programme highlighted the evolution of wheat domestication (Peng et al., 2003) 1. Highlighting the genetic structure, func- and expedited the genomic analysis of wild tion, regulation and evolution at macro- and cereal relatives. It can thus provide a solid microgeographical scales of natural popula- basis for introgression or cloning and trans- tions and the corresponding cultivars, bridg- formation of agronomically important ing multilocus marker structures with genes and QTLs from the wild to the culti- fitness-related traits in order to get direct vated cereals and for advancing cereal estimates of the adaptive fitness differentia- improvement. This is particularly important tion within and between populations (by in a world whose population is exploding, using transplant experiments, mapping where hunger is prevalent, desertification analysis, microarray methodology and selec- and salinization are dramatically increasing, tion-based mapping of fitness components and water and fertile land resources are and domesticated contributing genes over limited because of constant pollution and the entire life cycle). environmental degradation. 2. Exposing the large-scale genome organi- zation of the progenitors and their corre- Prospects for Crop Improvement sponding cultivars (wheat, soybeans, rice, maize, barley and others) using bacterial What will be the next step in the research artificial chromosome (BAC) clones for mol- into wild cereals and other progenitors of ecular complete (Arabidopsis and rice) or par- cultivated plants in the genomic and post- tial sequence analysis of expressed sequence genomic era in an attempt to improve tags (ESTs), and open reading frames crops? Conceptually, in-depth probing of (ORFs) in the transcribed strand; molecular comparative genome structure and function cytogenetic methods could complementarily are the major challenges, in particular, the probe the structure and interactions of the intimate relationship of the coding and non- nuclear, mitochondrial and chloroplast coding genomes and the focus on genomic genomes. Sequences upstream and down- regulation. This may be particularly aided stream of selected ORFs (5Ј and 3Ј untrans- by the discovery that microsatellites are pref- lated regions (UTRs), respectively) could be erentially associated with non-repetitive probed for functional regulatory polymor- DNA in plant genomes (Morgante et al., phisms and their function tested by genetic 2002) and that the regulatory functions may transformation. be of great importance (Li et al., 2002). Such 3. Analysis of the progenitors’ genetic system, studies will unravel genome evolution and or the ‘transmission system’, which deter- highlight rich genetic potentials for wheat mines the genetic flexibility of the species in improvement residing in the progenitors diverse ecological contexts including:

308 E. Nevo Fig. 14.7. Map locations of domestication syndrome factors (DSFs) and their involved QTLs in L version maps of wild emmer wheat, T. dicoccoides. (Upper) Short arms of chromosomes. (Right) DSFs and corresponding QTLs: , DSF; , Kernal number/spike (KNS); , kernel number/spikelet (KNL); , grain yield/plant (YLD); , plant height (HT); , spikelet number/spike (SLS); , single spike weight (SSW); , spike weight/plant (SWP); , kernel number/plant (KNP); , spike number/plant (HD); , GWH; , spike number/plant (SNP). The regular trait name represents a single QTL; the italic trait name represents a single QTL (Q2) detected by linked-QTL analysis; the regular trait name tailed with Q1 means the first QTL and tailed with Q2, the second QTL in a pair of linked QTLs. A tailed trait name (5) means that the QTL effect is not significant at the level of 5% of false discovery rate (FDR) but is significant at FDR 10%; (10) means that the effect is not significant at FDR 10% (from Peng et al., 2003, where details can be found). ● Breeding system (reaction norm and ● Genome evolution in the polyploidization genetic variation of the outcrossing rate). process. ● Mutation rate in different elements of the Applied Perspective genome, both coding and non-coding. 1. Optimization of predictive sampling ● Recombination properties of the genome, strategies based on ecological heterogeneity their genetic and ecological control. and molecular markers as guidelines, both regionally and locally, for in situ and ex situ ● Genomic distribution of structural genes, conservation and utilization. primarily abiotic and biotic stress genes 2. Genetic fine-mapping and dissection of and their regulatory function. ● Interface between ecological and genomic spatio-temporal dynamics and adaptive systems.

Genomic diversity in nature and domestication 309 the collected unique wild genetic resources the progenitors across their ecogeographic of agricultural importance by molecular ecological spectrum and their cultivars across markers including SSRs and single diverse climates, soils and cultures and across nucleotide polymorphisms (SNPs) and coding and non-coding genomic regions to sequencing followed by transformation of assess comparative diversities, and exploit the detected genes/alleles into elite cultivars the rich wild germplasm for crop improve- via marker-assisted selection. ment through genetic reinforcement. This 3. Molecular identification of adaptation could substantially optimize food production genes based on integrated genomic strate- and meet the world’s demands for food in a gies, and novel methodologies including second expanded green revolution. genetic and physical mapping (Fig. 14.7), molecular markers, ESTs, mapping, cloning Acknowledgements and sequencing, microarray expression analysis and genetic transformation of the I thank my colleagues A. Beiles, A. Korol defined target genes/alleles. and G. Song for reading and commenting 4. Comparative genetics/genomics of cereal on the manuscript. plants aimed at deciphering the common specific ways of domestication evolution. 5. Critical comparative genetics/genomics of References Akimoto, M., Shimamoto, Y. and Morishima, H. (1999) The extinction of genetic resources of Asian wild rice, Oryza rufipogon Griff.: a case study in Thailand. Genetic Resources and Crop Evolution 46, 419–425. Appels, R., Francki, M. and Chibbar, R. (2002) Advances in cereal functional genomics. Functional and Integrative Genomics 3, 1–24. Badr, A., Müller, K., Schafer-Pregl, R., El Rabey, H., Effgen, S., Ibrahim, H.H., Pozzi, C., Rohde, W. and Salamini, F. (2000) On the origin and domestication history of barley. Molecular Biology Evolution 17, 499–510. Baek, H.J., Beharav, A. and Nevo, E. (2003) Ecological-genomic diversity of microsatellites in wild barley, Hordeum spontaneum, populations in Jordan. Theoretical and Applied Genetics 106, 397–410. Bar-Yosef, O. (1998) The Natufian culture in the Levant, threshold of the origin of agriculture. Evolutionary Anthropology 6, 159–177. Beebe, S., Rengifo, J., Gaitan, E., Duque, M.C. and Tohme, J. (2001) Diversity and origin of Andean landraces of common bean. Crop Science 41, 854–862. Bhattacharjee, R., Bramel, P.J., Hash, C.T., Kolesnikova-Allen, M.A. and Khairwal, I.S. (2002) Assessment of genetic diversity within and between pearl millet landraces. Theoretical and Applied Genetics 105, 666–673. Brown, A.H.D., Feldman, M.W. and Nevo, E. (1980) Multilocus structure of natural populations of Hordeum spontaneum. Genetics 96, 523–536. Brown, A.H.D., Frankel, O.H., Marshall, D.R. and Williams, J.T. (eds) (1989) The Use of Plant Genetic Resources. Cambridge University Press, Cambridge. Brown, A.H.D., Clegg, M.T., Kahler, A.L. and Weir, B.S. (1990) Plant Population Genetics, Breeding, and Genetic Resources. Sinauer Associates Inc., Sunderland, Massachusetts. Brush, S. (1999) Genetic erosion of crop populations in centers of diversity: a revision. FAO, Proceedings of the Technical Meeting on the Methodology of the FAO World Information and Early Warning System on Plant Genetic Resources, 21–23 June 1999, Research Institute of Crop Production, Prague, Czech Republic. Buckler, E., Thornsberry, J. and Kresovich, S. (2001) Molecular diversity, structure and domestication of grasses. Genetic Resources, Cambridge 77, 213–218. Burke, J., Tang, S., Knapp, S. and Rieseberg, L. (2002) Genetic analysis of sunflower domestication. Genetics 161, 1257–1267.

310 E. Nevo Coulibaly, S., Pasquet, R., Papa, R. and Gepts, P. (2002) AFLP analysis of the phenetic organization and genetic diversity of Vigna unguiculata L. Walp. reveals extensive gene flow between wild and domesti- cated types. Theoretical and Applied Genetics 104, 358–366. Darwin, C. (1859) On the Origin of Species. Atheneum, New York. Diamond, J. (2002) Evolution, consequences, and future of plant and animal domestication. Nature 418, 700–706. Ehrlich, P.R. and Wilson, E.O. (1991) Biodiversity studies: science and policy. Science 253, 758–762. Ellis, R., Russell, J., Ransay, L., Waugh, R. and Powell, W. (2003) Barley domestication – Hordeum sponta- neum, a source of new genes for crop improvement. SCRI at www.scri.sari.ac.uk Fahima, T., Sun., G.L., Beharav, A., Krugman, T., Beiles, A. and Nevo, E. (1999) RAPD polymorphism of wild emmer wheat population, Triticum dicoccoides, in Israel. Theoretical and Applied Genetics 98, 434–447. Fahima, T., Röder, M.S., Wendehake, K., Kirzhner, V.M. and Nevo, E. (2002) Microsatellite polymorphism in natural populations of wild emmer wheat, Triticum dicoccoides, in Israel. Theoretical and Applied Genetics 104, 17–29. Feldman, M. and Sears, E.R. (1981) The wild gene resources of wheat. Scientific American 244, 102–112. Fetch, T., Steffenson, B. and Nevo, E. (2000) Multiple disease resistance in Hordeum spontaneum accessions from Israel and Jordan. 8th International Barley Genetics Symposium, 22–27 October, Adelaide, South Australia. Fetch, T., Gold, J. and Nevo, E. (2003a) Evaluation of wild oat germplasm for stem rust resistance. 8th International Congress Plant Pathology, 2–8 February, Christchurch, New Zealand (abstract). Fetch, T.G. Steffenson, B.J. and Nevo E. (2003b) Diversity and sources of multiple disease resistance in Hordeum spontaneum. Plant Disease 87, 1439–1448. Feuillet, C. and Keller, B. (2002) Comparative genomics in the grass family: molecular characterization of grass genome structure and evolution. Annals of Botany 89, 3–10. Fiedler, P.L. and Jain, S.K. (eds) (1992) Conservation Biology: the Theory and Practice of Nature Conservation Preservation and Management. Chapman & Hall, New York. Fofana, B., Jardin, P. and Baudoin, J. (2001) Genetic diversity in the lima bean (Phaseolus lunatus L.) as revealed by chloroplast DNA (cpDNA) variations. Genetic Resources and Crop Evolution 45, 437–445. Frankel, O.H. and Bennett, E. (1970) Genetic Resources in Plants – Their Exploration and Conservation. Blackwell, Oxford. Frankel, O.H. and Hawkes, J.G. (1975) Crop Genetic Resources for Today and Tomorrow. Cambridge University Press, Cambridge. Frankel, O.H. and Soule, M.E. (1981) Conservation and Evolution. Cambridge University Press, Cambridge. Frankel, O.H., Brown, A.H.D. and Burdon, J.J. (1995) The Conservation of Plant Biodiversity. Cambridge University Press, Cambridge. Gale, M. and Devos, K. (1998) Comparative genetics in the grasses. Proceedings of the National Academy of Sciences USA 95, 1971–1974. Gaut, B., Le Thierry d’Ennequin, M., Peek, A. and Sawkins, M. (1997) Maize as a model for the evolution of plant nuclear genomes. Proceedings of the National Academy of Sciences USA 97, 7008–7015. Gopher, A., Abbo, S. and Lev-Yadun, S. (2002) The ‘when’, ‘where’ and the ‘why’ of the Neolithic revolution in the Levant. Documenta Praehistorica 28, 49–62. Guoxiong, C., Krugman, T., Fahima, T., Korol, A.B. and Nevo, E. (2002) Comparative study of morphological and physiological traits related to drought resistance between xeric and mesic Hordeum spontaneum lines in Israel. Barley Genetic Newsletter 32, 22–33. Gupta, P.K., Sharma, P.K., Balyan, H.S., Roy, J.K., Sharma, S., Beharav, A. and Nevo, E. (2002) Polymorphism at rDNA loci in barley and its relation with climatic variables. Theoretical and Applied Genetics 104, 473–481. Hammer, K., Diederichsen, A. and Spahillari, M. (1999) Basic studies toward strategies for conservation of plant genetic resources. FAO, Proceedings of the Technical Meeting on the Methodology of the FAO World Information and Early Warning System on Plant Genetic Resources, 21–23 June 1999, Research Institute of Crop Production, Prague, Czech Republic. Harlan, J.R. (1975a) Crops and Man. American Society of Agronomy, Crop Science Society of America, Madison, Wisconsin. Harlan, J.R. (1975b) Our vanishing genetic resources. Science 188, 618–621. Harlan, J.R. and Zohary, D. (1966) Distribution of wild wheats and barley. Science 153, 1074–1080. Hawkes, J.G. (1983) The Diversity of Crop Plants. Harvard University Press, Cambridge, Massachusetts.

Genomic diversity in nature and domestication 311 Hawkes, J.G. (1991) International workshop in dynamic in situ conservation of wild relatives of major culti- vated plants: summary of final discussion and recommendations. Israel Journal of Botany 40, 529–536. Hernandez-Verdugo, S., Luna-Reyes, R. and Oyama, K. (2001) Genetic structure and differentiation of wild and domesticated populations of Capsicum annuum (Solanaceae) from Mexico. Plant Systematics and Evolution 226, 129–142. Hilton, H. and Baut, B. (1998) Speciation and domestication in maize and its wild relatives: evidence from the globulin-1 gene. Genetics 150, 863–872. Hoisington, D., Khairallah, M., Reeves, T., Ribaut, J.M., Skovmand, B., Taba, S. and Warburton, M. (1999) Plant genetic resources: what can they contribute toward increased crop productivity? Proceedings of the National Academy of Sciences USA 96, 5937–5943. Holden, J., Peacock, J. and Williams, T. (1993) Genes, Crops, and the Environment. Cambridge University Press, Cambridge. Huang, X.Q., Börner, A., Röder, M.S. and Ganal, M.W. (2002) Assessing genetic diversity of wheat (Triticum aestivum L.) germplasm using microsatellite markers. Theoretical and Applied Genetics 105, 699–707. Hutchinson, C. and Weiss, E. (1999) Remote sensing contribution to an early warning system for genetic erosion of agricultural crops. FAO, Proceedings of the Technical Meeting on the Methodology of the FAO World Information and Early Warning System on Plant Genetic Resources, 21–23 June 1999, Research Institute of Crop Production, Prague, Czech Republic. IBPGR (1985) Ecogeographical Surveying and in situ Conservation of Crop Relatives. IBPGR Secretariat, Rome. ICPGR (1995) Outline of the Report on the State of the World’s Plant Genetic Resources. Item 7 of the Provisional Agenda, Sixth Session in Rome, 19–30 June 1995, Commission on Plant Genetic Resources. Jones, M.K., Allaby, R.G. and Brown, T.A. (1998) Wheat domestication. Science 279, 302–303. Kellogg, E. (2001) Evolutionary history of the grasses. Plant Physiology 125, 1198–1205. Khush, G. (2001) Green revolution: the way forward. Nature Reviews Genetics 2, 815–822. Kirzhner, V.M., Korol, A., Turpeinen, T. and Nevo, E. (1995) Genetic supercycles caused by cyclical selec- tion. Proceedings of the National Academy of Sciences USA 92, 7130–7133. Kirzhner, V.M., Korol, A.B. and Nevo, E. (1996) Complex dynamics of multilocus systems subjected to cycli- cal selection. Proceedings of the National Academy of Sciences USA 93, 6532–6535. Kirzhner, V.M., Korol, A.B. and Nevo, E. (1999) Abundant multilocus polymorphisms caused by genetic interaction between species on trait-for-trait basis. Journal of Theoretical Biology 198, 61–70. Kislev, M. (2002) Origin of annual crops by agro-evolution. Israel Journal of Plant Sciences 50, 85–88. Korol, A.B., Kirzhner, V.M. and Nevo, E. (1998) Dynamics of recombination modifiers caused by cyclical selection: interaction of forced and auto oscillations. Genetic Resources 72, 135–147. Kuang, H., Woo, S.S., Meyers, B., Nevo, E. and Michelmore, R. (2004a) Multiple genetic processes result in heterogeneous rates of evolution within the major cluster disease resistance genes in lettuce. The Plant Cell 16 (in press). Kuang, H., Ochoa, O., Nevo, E. and Michelmore, R. (2004b) The resistance gene Dm3 is infrequent in nat- ural populations of Lactuca serriola due to deletions and gene conversions (in press). Ladizinsky, G. (1998) Plant Evolution Under Domestication. Kluwer Academic Publishers, Dordrecht, The Netherlands. Lauter, N. and Doebly, J. (2002) Genetic variation for phenotypically invariant traits detected in teosinte: implications for the evolution of novel forms. Genetics 160, 333–342. Lev-Yadun, S., Gopher, A. and Abbo, S. (2000) The cradle of agriculture. Science 288, 1602–1603. Li, Y.C., Fahima, T., Beiles, A., Korol, A.B. and Nevo, E. (1999) Microclimatic stress and adaptive DNA differ- entiation in wild emmer wheat (Triticum dicoccoides). Theoretical and Applied Genetics 98, 873–883. Li, Y.C., Korol, A.B., Fahima, T., Beiles, A. and Nevo, E. (2002) Microsatellites: genomic distribution, putative functions and mutational mechanisms: a review. Molecular Ecology 11, 2453–2465. Lin, J., Brown, A. and Clegg, M. (2001) Heterogeneous geographic patterns of nucleotide sequence diversity between two alcohol dehydrogenase genes in wild barley (Hordeum vulgare subspecies spontaneum). Proceedings of the National Academy of Sciences USA 98, 531–536. Liu, K.D., Yang, G.P., Zhu, S.H., Zhang, Q., Wang, X.M. and Saghai Maroof, M.A. (1996) Extraordinarily polymorphic ribosomal DNA in wild and cultivated rice. Genome 39, 1109–1116. Liu, Z., Sun, Q., Ni, Z., Nevo, E. and Yang, T. (2002) Molecular characterization of a novel powdery mildew resistance gene Pm30 in wheat originating from wild emmer. Euphytica 123, 21–29. Maroof, M.A., Soliman, K.M., Jorgensen, R.A. and Allard, R.W. (1984) Ribosomal DNA spacer-length poly-

312 E. Nevo morphisms in barley: Mendelian inheritance, chromosomal location and population dynamics. Proceedings of the National Academy of Sciences USA 81, 8014–8018. Maroof, M.A., Biyashev, R.M., Yang, G.P., Zhang, Q. and Allard, R.W. (1994) Extraordinarily polymorphic microsatellite DNA in barley: species diversity, chromosomal locations, and population dynamics. Proceedings of the National Academy of Sciences USA 92, 5466–5470. Matsuoka, Y., Vigouroux, Y., Goodman, M., Sanchez, J., Buckler, E. and Doebley, J. (2002) A single domesti- cation for maize shown by multilocus microsatellite genotyping. Proceedings of the National Academy of Sciences USA 99, 6080–6084. Maughan, P.J., Saghai Maroof, M.A. and Buss, G.R. (1995) Microsatellite and amplified sequence length polymorphisms in cultivated and wild soybean. Genome 38, 715–723. Maughan, P.J., Saghai Maroof, M.A., Buss, G.R. and Huestis, G.M. (1996) Amplified fragment length poly- morphism (AFLP) in soybean: species diversity, inheritance, and near-isogenic line analysis. Theoretical and Applied Genetics 93, 392–401. Mauricio, R. (2001) Mapping quantitative trait loci in plants: uses and caveats for evolutionary biology. Nature Reviews 2, 370–381. Maxted, J., Ford-Lloyd, B.V. and Hawkes, J.G. (eds) (1997) Plant Genetic Conservation. The In Situ Approach. Chapman & Hall, London, pp. 144–159. Mimura, M., Yasuda, K. and Yamaguchi, H. (2000) RAPD variation in wild, weedy and cultivated azuki beans in Asia. Genetic Resources and Crop Evolution 47, 603–610. Morgante, M., Hanafey, M. and Powell, W. (2002) Microsatellites are preferentially associated with non- repetitive DNA in plant genomes. Genetics 30, 194–200. Morishima, H. and Oka, H. (1970) A survey of genetic variations in the populations of wild Oryza species and their cultivated relatives. Japanese Journal of Genetics 45, 371–385. Moseman, J.G., Nevo, E., El Morshidy, M. and Zohary, D. (1984) Resistance of Triticum dococcoides to infection with Erysiphe graminis tritici. Euphytica 33, 41–47. Neale, D., Saghai-Maroof, M., Allard, R., Zhang, Q. and Jorgensen, R. (1988) Chloroplast DNA diversity in populations of wild and cultivated barley. Genetics 120, 1105–1110. Nesbitt, M. and Samuel, D. (1998) Wheat domestication: archaeological evidence. Science 279, 1433. Nevo, E. (1978) Genetic variation in natural populations: patterns and theory. Theoretical Population Biology 13, 121–177. Nevo, E. (1983) Genetic resources of wild emmer wheat: structure, evolution and application in breeding. In: Sakamoto, S. (ed.) Proceedings of the 6th International Wheat Genetics Symposium. Kyoto University, Kyoto, pp. 421–431. Nevo, E. (1986) Genetic resources of wild cereals and crop improvement: Israel, a natural laboratory. Israel Journal of Botany 35, 255–278. Nevo, E. (1987) Plant genetic resources: prediction by isozyme markers and ecology. In: Rattazi, M., Scandalios, J. and Whitt, G.S. (eds) Isozymes: Current Topics in Biological Research. Agriculture, Physiology and Medicine 16, 247–267. Nevo, E. (1988) Genetic diversity in nature: patterns and theory. Evolutionary Biology 23, 217–247. Nevo, E. (1989) Genetic resources of wild emmer wheat revisited: genetic evolution, conservation and uti- lization. In: Miller, T.E. and Koebner, R.M.D. (eds) Proceedings of the Seventh International Wheat Genetics Symposium, 13–19 July 1998, Institute of Plant Science Research, Cambridge, pp. 121–126. Nevo, E. (1992) Origin, evolution, population genetics and resources for breeding of wild barley, Hordeum spontaneum, in the Fertile Crescent. In: Shewry, P.R. (ed.) Barley: Genetics, Biochemistry, Molecular Biology and Biotechnology. CAB International, Wallingford, UK, pp. 19–43. Nevo, E. (1995a) Genetic resources of wild emmer, Triticum dicoccoides, for wheat improvement: News and Views. Proceedings for the 8th International Wheat Genetic Symposium, 20–25 July 1993, Beijing, pp. 79–87. Nevo, E. (1995b) Evolution and extinction. In: Nierenberg, W.N. (ed.) Encyclopedia of Environmental Biology, Vol. 1. Academic Press, New York, pp. 717–745. Nevo, E. (1998a) Molecular evolution and ecological stress at global, regional and local scales: the Israeli perspective. Journal of Experimental Zoology 282, 95–119. Nevo, E. (1998b) Genetic diversity in wild cereals: regional and local studies and their bearing on conserva- tion ex-situ and in-situ. Genetic Resource and Crop Evolution 45, 355–370. Nevo, E. (2001a) Genetic diversity. In: Levin, S. (ed.) Encyclopedia of Biodiversity, Vol. 3. Academic Press, New York, pp. 195–213.

Genomic diversity in nature and domestication 313 Nevo, E. (2001b) Evolution of genome–phenome diversity under environmental stress. Proceedings of the National Academy of Sciences USA 98, 6233–6240. Nevo, E. (2001c) Genetic resources of wild emmer, Triticum dicoccoides, for wheat improvement in the third millennium. Israel Journal of Plant Sciences 49, 162. Nevo, E. and Beiles, A. (1989) Genetic diversity of wild emmer wheat in Israel and Turkey: structure, evolu- tion and application in breeding. Theoretical and Applied Genetics 77, 421–455. Nevo, E., Zohary, D., Brown, A.H.D. and Haber, M. (1979) Genetic diversity and environmental associa- tions of wild barley, Hordeum spontaneum, in Israel. Evolution 33, 815–833. Nevo, E., Golenberg, E., Beiles, A., Brown, A.H.D. and Zohary, D. (1982) Genetic diversity and environ- mental associations of wild wheat, Triticum dicoccoides, in Israel. Theoretical and Applied Genetics 62, 241–254. Nevo, E., Beiles, A. and Ben-Shlomo, R. (1984) The evolutionary significance of genetic diversity: ecologi- cal, demographic and life history correlates. Lecture Notes in Biomathematics 53, 13–213. Nevo, E., Moseman, J.G., Beiles, A. and Zohary, D. (1985) Patterns of resistance of Israeli wild emmer wheat to pathogens. I. Predictive method by ecology and allozyme genotypes for powdery mildew and leaf rust. Genetica 67, 209–222. Nevo, E., Beiles, A. and Zohary, D. (1986) Genetic resources of wild barley in the Near East: structure, evo- lution and application in breeding. Biological Journal of the Linnean Society 27, 355–380. Nevo, E., Korol, A.B., Beiles, A. and Fahima, T. (2002) Evolution of Wild Emmer Wheat Improvement. Population Genetics, Genetic Resources, and Genome Organization of Wheat’s Progenitor, Triticum dicoccoides. Springer-Verlag, Berlin, pp. 364. Ortiz, R. and Huaman, Z. (2001) Allozyme polymorphisms in tetraploid potato gene pools and the effect on human selection. Theoretical and Applied Genetics 103, 792–796. Owuor, E.D., Fahima, T., Beiles, A., Korol, A.B. and Nevo, E. (1997) Population genetics response to microsite ecological stress in wild barley Hordeum spontaneum. Molecular Ecology 6, 1177–1187. Owuor, E.D., Beharav, A., Fahima, R., Kirzhner, V.M., Korol, A.B. and Nevo, E. (2003) Microscale ecological stress causes RAPD molecular selection in wild barley, Neve Yaar microsite, Israel. Genetic Resources and Crop Evolution 50, 213–224. Pakniyat, H., Powell, W., Baird, E., Handley, L.L., Robinson, D., Scrimgeour, C.M., Nevo, E., Hackett, C.A., Caligari, P.D. and Forster, B.P. (1997) AFLP variation in wild barley (Hordeum spontaneum C. Koch) with reference to salt tolerance and associated ecogeography. Genome 40, 332–341. Parzies, H., Spoor, W. and Ennos, R. (2000) Genetic diversity of barley landrace accessions (Hordeum vulare ssp. vulgare) conserved for different lengths of time in ex situ gene banks. Heredity 84, 476–486. Pasternak, T. (1998) Investigations of botanical remains from Nevali Cori PPNB, Turkey: a short interim report. In: Damania, A.B., Valkoun, J., Wilcox, G. and Qualset, C.O. (eds) The Origins of Agriculture and Crop Domestication. ICARDA, Aleppo, Syria, pp. 170–176. Peng, J.H., Korol, A.B., Fahima, T., Röder, M.S., Li, Y.C. and Nevo, E. (2000) Molecular genetic maps in wild emmer wheat, Triticum dicoccoides: genome-wide coverage, massive negative interference, and puta- tive quasi-linkage. Genome Research 10, 1509–1531. Peng, J.H., Ronin, Y.I., Fahima, T., Roder, M., Li, Y., Nevo, E. and Korol, A.B. (2003) Domestication quantita- tive trait loci in Triticum dicoccoides, the progenitor of wheat. Proceedings of the National Academy of Sciences USA 100, 2489–2494. Plaschke, J., Ganal, M.W. and Röder, M.S. (1995) Detection of genetic diversity in closely related bread wheat using microsatellite markers. Theoretical and Applied Genetics 91, 1001–1007. Plucknett, D.L., Smith, N.J., Williams, J.T. and Anishetty, N.M. (1987) Gene Banks and the World’s Food. Princeton University Press, Princeton, New Jersey. Poncet, V., Lamy, G., Enjalbert, J., Jolys, H., Sarr, A. and Robert, T. (1998) Genetic analysis of the domestica- tion syndrome in pearl millet (Pennisetum glaucum L., Poaceae): inheritance of the major characters. Heredity 81, 648–658. Ramachandran, S., Fahima, T., Krugman, T., Roder, M., Nevo, E. and Feldman, M. (2005) Domestication times of wheat and barley using microsatellite polymorphisms (in review). Reeves, J.C., Law, J.R., Donini, P., Koebner, R. and Cooke, R. (1999) Changes over time in the genetic diver- sity of UK cereal crops. FAO, Proceedings of the Technical Meeting on the Methodology of the FAO World Information and Early Warning System on Plant Genetic Resources, 21–23 June, Research Institute of Crop Production, Prague, Czech Republic. Salamini, F., Ozkan, H., Brandolini, A., Schafer-Pregl, R. and Martin, W. (2002) Genetics and geography of wild cereal domestication in the Near East. Nature Reviews 3, 429–441.

314 E. Nevo Sandhu, D. and Gill, K.S. (2002) Gene-containing regions of wheat and the other grass genomes. Plant Physiology 128, 803–811. Sanjur, O., Piperno, D., Andres, T. and Wessel-Beaver, L. (2002) Phylogenetic relationships among domesti- cated and wild species of Cucurbita (Cucurbitaceae) inferred from a mitochondrial gene: implications for crop plant evolution and areas of origin. Proceedings of the National Academy of Sciences USA 99, 535–540. Sicard, D., Woo, S.-S., Arroyo-Garcia, R., Ochoa, O., Nguyen, D., Korol, A.B., Nevo, E. and Michelmore, R. (1999) Molecular diversity at the major cluster of disease resistance genes in cultivated and wild Lactuca spp. Theoretical and Applied Genetics 99, 405–418. Smith, B.D. (1995) The Emergence of Agriculture. Scientific American Library, New York. Soule, M.E. (ed.) (1986) Conservation Biology: the Science of Scarcity and Diversity. Sinauer Association Inc., Sunderland, Massachusetts. Soule, M.E. (1987) Viable Populations for Conservation. Cambridge University Press, Cambridge. Styles, B.T. (ed.) (1986) Infraspecific Classification of Wild and Cultivated Plants. Systematics Association Special Volume No. 29, Clarendon Press, Oxford. Tang, S. and Knapp, S. (2003) Microsatellites uncover extraordinary diversity in Native American land races and wild populations of cultivated sunflower. Theoretical and Applied Genetics 106, 990–1003. Tanksley, S.D. and McCouch, S.R. (1997) Seed banks and molecular maps: unlocking genetic potential from the wild. Science 277, 1063–1066. Tripp, R. and van der Heide, W. (1996) The Erosion of Crop Genetic Diversity: Challenges, Strategies and Uncertainties. Natural Resource Perspectives, No. 7. Overseas Development Institute, London. Turpeinen, T., Tenhola, T., Manninen, O., Nevo, E. and Nissila, E. (2001) Microsatellite diversity associated with ecological factors in Hordeum spontaneum populations in Israel. Molecular Ecology 10, 1577–1591. Turpeinen, T., Vanhala, V., Nevo, E. and Nissila, E. (2003) AFLP genetic polymorphism in wild barley (Hordeum spotaneum) populations in Israel. Theoretical Applications of Genetics 106, 1333–1339. Valdes, B., Heywood, V.H., Raimondo, F. and Zohary, D. (eds) (1997) Proceedings of the Workshop on Conservation of the Wild Relatives of European Cultivated Plants, 1–6 September 2003, Palermo, Italy. Vavilov, N.I. (1926) Studies on the Origin of Cultivated Plants. Institute of Botanique Applique et d’Amelioration des Plants, Leningrad. Vavilov, N.I. (1951) The origin, variation and breeding of cultivated plants. Cronica Botanica 13, 1–364 (translated from Russian). Vigouroux, Y., McMullen, M., Hittinger, C.T., Houchins, K., Schulz, L., Kresovich, S., Matsuoka, Y. and Doebley, J. (2002) Identifying genes of agronomic importance in maize by screening microsatellites for evidence of selection during domestication. Proceedings of the National Academy of Sciences USA 99, 9650–9655. Volis, S., Yakubov, B., Shulgina, E., Ward, D., Zur, V. and Mendlinger, S. (2001) Tests for adaptive RAPD vari- ation in population genetic structure of wild barley, Hordeum spontaneum Koch. Biological Journal of the Linnean Society 74, 289–303. Volis, S., Mendlinger, S. and Ward, D. (2002) Differentiation in populations of Hordeum spontaneum along a gradient of environmental productivity and predictability: life history and local adaptation. Biological Journal of the Linnean Society 77, 479–490. Wilson, E.O. (1992) The Diversity of Life. Harvard University Press, Cambridge, Massachusetts. Wilson, E.O. (2002) The Future of Life. Alfred A. Knopf, New York. Xiong, L., Liu, K., Dai, X., Xu, C. and Zhang, Q. (1999) Identification of genetic factors controlling domesti- cation-related traits of rice using an F2 population of a cross between Oryza sativa and O. rufipogon. Theoretical and Applied Genetics 98, 243–251. Xu, R., Tomooka, N., Vaughan, D. and Doi, K. (2000) The Vigna angularis complex: genetic variation and relationships revealed by RAPD analysis, and their implications for in situ conservation and domestica- tion. Genetic Resources 47, 123–134. Yamashita, K. (1965) Cultivated Plants and Their Relatives. Scientific Expedition to the Karakoram and Hindukush, 1955, Vol. I. Yang, G.P., Saghai Maroof, M.A., Xu, C.G., Zhang, Q. and Biyashev, R.M. (1994) Comparative analysis of microsatellite DNA polymorphism in landraces and cultivars of rice. Molecular and General Genetics 245, 187–194. Zhang, L., Peek, A., Dunams, D. and Gaut, B. (2002) Population genetics of duplicated disease-defense

Genomic diversity in nature and domestication 315 genes, hm1 and hm2, in maize (Zea mays ssp. Mays L.) and its wild ancestor (Zea mays ssp. Parviglumis). Genetics 162, 851–860. Zhu, Y., Chen, H., Fan, J., Wang, Y., Li, Y., Chen, J., Fan, J.X., Yang, S., Hu, L., Leung, H., Mew, T., Teng, P., Wang, Z. and Mundt, C. (2000) Genetic diversity and disease control in rice. Nature 17, 718–722. Zohary, D. (1996) The mode of domestication of the founder crops of Southwest Asian agriculture. In: Harris, D.R. (ed.) The Origins and Spread of Agriculture and Pastoralism in Eurasia. UCL Press, London, pp. 142–158. Zohary, D. (1999) Monophyletic vs. polyphyletic origin of the crops on which agriculture was founded in the Near East. Genetic Resources and Crop Evolution 46, 133–142. Zohary, D. and Hopf, J. (2000) Domestication of Plants in the Old World, 3rd edn. Oxford University Press, Oxford.

15 Conserving genetic diversity in plants of environmental, social or economic importance Robert J. Henry Centre for Plant Conservation Genetics, Southern Cross University, PO Box 157, Lismore, NSW 2480, Australia Introduction Rare and Endangered Species The case for conserving plant diversity was Large numbers of plant species are consid- made in Chapter 1. This chapter further ered to be rare or endangered. The security describes what needs to be conserved and of wild plant populations is usually evalu- how. Plants and plant-derived products are ated in the context of current land use. A central features of human society (Table 15.1). species is threatened when extinction Species that are used as sources of food, fibre, becomes a likely outcome if current land use shelter or medicine may have added social sig- is continued unaltered. Documentation of nificance beyond their direct practical or eco- rare species continues to be a major chal- nomic use. Diverse groups of plants are lenge. Many rare species have probably currently used by humans (Tables 15.2 and become extinct before there was a chance to 15.3). Future uses of plants may involve using describe them. An example of an approach biotechnology to deliver vaccines or other to the identification and classification of rare high-value medicinal products or producing plants is given in Table 15.4. The develop- precursors for plastics or other industrial ment of a recovery plan for rare species is materials. Furthermore, biotechnology allows the next step beyond initial discovery and for the potential use of any or all plant species recognition of rarity. Implementation of as a source of useful genes. This provides a recovery plans in most cases lags well behind convincing argument in support of efforts to these earlier stages of discovery, evaluation conserve all higher plant diversity. and planning. All plant species are of importance from Two approaches to plant conservation an environmental perspective, in that loss of can be considered: in situ (Prance, 2004) and diversity in any species reduces plant diver- ex situ (Thormann, 2004). Conservation in sity. However, most interest is directed situ aims to protect wild plant populations towards plants that are rare or in danger of and ensure diversity is preserved and that extinction in the short or medium term. evolution of the species can continue. This is This should not distract effort from the the most common approach to conservation related issue of preserving diversity in more of environmentally important species. common or widespread species and espe- Conservation of plants in seed banks, or as cially the importance of conserving rare living collections such as those in botanic variants of more common species. gardens, may be necessary for critically © CAB International 2005. Plant Diversity and Evolution: Genotypic and 317 Phenotypic Variation in Higher Plants (ed. R.J. Henry)

318 R.J. Henry Table 15.1. Human uses of higher plants. Category of use Examples Food Cereal, pulses, fruit, vegetables, oilseeds and sugar Beverage Wine, beer, tea, coffee Animal feed Pastures, fodder Fibre Cotton, hemp and paper Fuel Firewood Construction Housing and furniture Medicine Pharmaceutical products, traditional medicines Ornament Cut flowers, pot plants, garden plants and turf grass Industry Ethanol for fuel, electricity Environment Environmental restoration, greenhouse gas sequestration Other Perfumes, cosmetics Table 15.2. Groups of flowering plants defined by DNA analysis (see Chapter 2, Fig. 2.1, this volume). Table 15.3. The uses of flowering plants (grouped as listed in Table 15.2) by humans. Based upon economic uses mainly as defined by Heywood (1978). Amborellaceae Amborella only New Caledonia Nymphaeaceae Ornamental water lilies Austrobaileyales Food (seeds and rhizomes) Chloranthaceae Cosmopolitan (fresh water) Canellales One genus, two species only in Australia, no known uses Piperales Ornament (Chloranthus glaberi) Beverage (Chloranthus officinalis) Laurales Medicine (Hedyosmum brasiliense) Ornament Magnoliales Food white cinnamon (Canella winterana) Alismatales Medicine Asparagales Food pepper (Piper nigrum) Beverage kava (Piper methysticum) Dioscoreales Ornament Liliales Medicine Food avocado (Persea americana) Ornament cinnamon, bay leaves Construction Other perfume (Doryphora sassafras) Medicine Food nutmeg (Myristica fragrans), custard apple (Annona) Ornament Magnolia Construction Ornament Food Sagittaria sagittifolia (tubers) Food onions, garlic, leek, vanilla, asparagus Ornament Gladiolus, Iris, Freesia, daffodils, orchids Medicine Other saffron (Crocus sativus) Food yams Medicine Beverage sarsaparilla Ornament Lilium, Tulipa (tulip) Mediine Continued

Conserving genetic diversity 319 Table 15.3. Continued. Pandanales Food (starchy fruits) Ornament Pandanus Dasypogonaceae Other perfume, baskets Arecales Ornament Xanthorrhoea Other varnishes Poales Food coconuts, copra, dates, sago, palm oil Fibre coir, raffia Commelinales Ornament Zingiberales Food rice, wheat, maize, barley, sorghum, millet, sugarcane, bamboo, pineapple Animal feed pastures Ceratophyliales Ornament lawns, water plants Ranunculales Other baskets, brooms, thatching Ornament wandering Jew, water hyacinth Sabiaceae Food banana, ginger, cardamom, turmeric, arrowroot Proteales Fibre manila, hemp Ornament Strelitzia, Canna Buxaceae Other perfume Trochodendraceae Other protects fish in freshwater Gunnerales Food fruits Aextoxicaceae Ornament Clematis, Ranunculus (buttercups) Berberidopsidaceae Medicine opium (Papaver somniferum) Dilleniaceae Ornament Caryophyliales Food Macadamia integrifolia Santalales Ornament Banksia, Grevillea, Telopea, Protea, Leucadendron, planes Saxifragales Construction timber Ornament Crossosomatales Construction Other birdlime (seed to catch birds) Gereniales Ornament Gunnera Other tanning and dyeing Myrtales Ornament (scared bamboo) Celastrales Medicine Ornament Hibbertia, Dillenia Food Arraranthus Ornament cockscombs (Celosia cristata), Ptilotus Medicine Food grapes (Vitis vinifera), gooseberries, currants (Ribes) Ornament Hydrangeas, Kalanchoe Construction timber Other perfume Ornament Construction Medicine Ornament Geranium, Pelargonium Construction timber Medicine Other perfume Food cloves (Syzgium aromatieun), lilly pilly Ornament bottlebrushes, Tibouchina, Fuschias Construction Eucalyptus Medicine Other essential oils Beverage Arabia tea (Catha edulis) Medicine Other essential oils Continued

320 R.J. Henry Table 15.3. Continued. Malpighiales Food cassava (Manihot glaziovii), passionfruit (Passiflora) Ornament Poinsettia (Euphorbia), violets Oxalidales Industry castor oil (Ricinus communis) Fabales Other rubber (Hevea brasiliensis) Ornament flycatcher plant (Cephalotus follicularis) Rosales Construction timber Food peas, beans, groundnut (peanut), soybean Cucurbitales Animal feed clover, lucerne Fagales Other nitrogen fixing Brassicales Construction timber Malvales Ornament Acacia Food fruits (apple, plum, pear, cherry, mulberries, fig, raspberries, strawberries) Sapindales Fibre hemp (Cannabis sativa) Ornament roses Cornales Beverage hops (Humulus lupulus) Ericales Construction elms Garryales Food cucumber, pumpkin, melon Gentianales Ornament Begonia Lamiales Food chestnut (Castanea sativa), walnut, pecan Solanales Construction beeches, oaks, birches Food oilseed rape, mustard, vegetables (cabbage, cauliflower), papaya (Carica Aquifoliales papaya) Apiales Animal feed fodder Asterales Fibre cotton (Gossypium) Dipsacales Ornament Hibiscus Construction Other chocolate Food orange, lemon, lime, mango, cashew, pistachio, lychee (Litchi chinensis), maple sugar Ornament maples Medicine Construction mahoganies Other perfume, poison ivy Ornament dogwoods Beverage tea (Camellia sinensis) Ornament Camellia Construction Medicine Beverage coffee (Coffea) Ornament Gentian, oleandas (Nerium), Gardenia Medicine quinine (Cinchona) Food olives (Olea europaca) Food potato, aubergine, tomato, pepper, sweet potato Ornament morning glory (Ipomoea purpurea) Medicine Other tobacco Ornament holly Construction Food carrot, celery, parsley, fennel, dill Medicine Other perfume Food sunflower, lettuce, chicory, Jerusalem artichoke Ornament Dahlia, Gerbera Beverage elderberry (wine) Ornament honeysuckles (Lonicera) Medicine

Conserving genetic diversity 321 Table 15.4. Classification of rare and endangered plant species (Environment Protection and Biodiversity and Conservation Act, 1999, Australia). Extinct – no reasonable doubt that the last member of the species has died Extinct in the wild – species exists only in cultivation Critically endangered – extremely high risk of extinction in the immediate future Endangered – very high risk of extinction in the medium term Vulnerable – high risk of extinction in the medium term Conservation dependent – species is dependent on a specific conservation programme without which it could become vulnerable, endangered or critically endangered within 5 years endangered species or in cases where the tion of many species will remain important. natural habitat is no longer available or suit- The relative security of wild populations and able. This ex situ conservation is much more ex situ collections such as those provided by commonly used for species of economic seed banks varies. Seed collections may be importance such as the major crop species. lost or destroyed. The wild populations should provide at the very least a backup for Species of Economic Importance even the best-collected species. Conservation in situ should allow continued evolutionary The sustainable production of food in agri- development of the species in the longer culture requires the effective use of plant term while seed collections or other ex situ genetic resources (Henry, 2001). Increasing collections represent the genetic resource at and even maintaining agricultural produc- the time of collection. tion is reliant upon continued genetic improvement of crop species. The available Cultivated plants genetic resources for major crops are the key resource for global food supply. The main source of genetic resources for Humans rely on a very small number of many domesticated species is plants in crops plant species for a large proportion of their on farms or in gardens. The domesticated diets. Just three cereal species, wheat, maize gene pool can often be considered to be dis- and rice, may account for as much as half of tinct from the wild gene pool. The increased all human food. Genetic improvement and adoption of higher performing plant vari- the effective use of available genetic diversity eties produced by modern plant breeding in these species in essential to continued techniques may displace the production of world food supply (Henry, 2004). Many older varieties or landraces resulting in loss other species are of importance to humans of this genetic resource. This does not neces- as a source of food or other products. This sarily need to result in a loss of diversity in chapter will briefly review the diversity of crop species in production if efforts are genetic resources available for genetic made to ensure improved varieties from improvement of higher plants of economic genetically diverse sources are released for importance. production. Private and public gardens rep- resent the main source of germplasm for Wild populations ornamentals and for many minor crops. Plants in wild populations are an important Genetic resource collections source of genetic resources for economically important plants. The importance of wild The conservation of plant genetic resources populations may be reduced by the estab- ex situ may take different forms depending lishment of ex situ collections from these on the biology of the species. Seed collec- populations. However, the in situ conserva- tions are important for species with long-

322 R.J. Henry lived seeds. Some plants do not have seed Other living collections that can be stored for any significant period of time. These must be maintained by vege- Species with seeds that are not suitable for tative propagation, usually as living collec- storage or that require vegetative propaga- tions in the field or glasshouse. Species that tion to retain genetic integrity are usually require clonal propagation to maintain maintained as living collections. Recalcitrant genetic characteristics may also be stored as species have seeds that cannot be dried with- living collections even if their seed is suitable out loss of viability. These species require for storage. Very large numbers of geno- alternatives to seed storage for germplasm types are held in germplasm collections for collections. Options include growth in the the major species (Table 15.5). field, in a plant house or in tissue culture. Techniques such as cyropreservation may be Seed banks used for some species. A range of different explants can be stored by cyropreservation. Large collections of seed are stored for These include cell suspensions, calluses, each of the major field crop species. These shoot tips and embryos (Thormann, 2004). collections are deliberately duplicated in different collections in different countries Plants may be kept in tissue culture at to increase global security of the resource. low temperatures to reduce the cost of This approach is only suitable for plants maintaining the cultures. Low temperatures with long-lived seeds. DNA analysis tools allow a lower frequency of media change. are becoming increasingly important in Slow-growing collections are used for managing these collections. Confirming species such as banana, plantain, cassava, sample identity and establishing relation- potato and sweet potato (Thormann, 2004). ships between accessions in collections allows more targeted efforts to conserve Botanic gardens diversity and supports more effective uti- lization of these genetic resources in plant Botanic gardens are widespread, usually improvement (Nagamine, 2004). These public institutions that provide a repository benefits also apply to other types of genetic usually including a very wide range of plant resource collection. species. Some private gardens also con- tribute in this way but may provide a lower Table 15.5. Numbers of accessions held in level of long-term security to the genetic germplasm collections of some major crop resource. The added value of botanic gar- species. For more information see FAO (1996). dens is usually associated with the documen- tation of the material in the collections Species Number of accessions (Makinson, 2004) and links to associated in collections herbaria. (Godwin, 2003) DNA banks Wheat 784,500 The storage of plant genomic DNA and Barley 485,000 even individual genes or gene libraries is an Rice 420,500 additional, more recent option for the con- Maize 277,000 servation of plant genetic resources (Rice, Oat 222,500 2004). This approach is not currently suit- Sorghum 168,500 able as an alternative to other methods for Soybean 174,500 conserving species because living plants can- Groundnut not be recovered from the stored DNA. Tomato 81,000 However, specific genes may be protected Apple 78,000 and reintroduced into plants. DNA banks Grape 97,500 47,000

Conserving genetic diversity 323 for plants are currently being established of crop species and important relatives has and expanded in several key locations inter- received comparatively little attention in nationally. many regions (Meilleur and Hodgkin, 2004). The gene pool for these species may Genetic resources available for different be considered to include the primary gene plant groups pool of the species itself, the secondary gene pool of closely related species that can be The quantity (numbers of samples) and used in breeding and a tertiary gene pool of quality (diversity of samples and standard of more distant relatives that have genes that storage conditions) of genetic resources might be accessed using more novel tech- available for plants varies. Some species are niques (Fig. 15.1). In many cases relatively reliant completely on wild populations. little effort has been given to conservation of Others have very large numbers of samples germplasm outside the primary gene pool. in seed banks (Table 15.5) and are distrib- Many crop plants of economic importance uted around the world. are listed in Table 15.3. Crop plants Recently Meilleur and Hodgkin (2004) defined key objectives for advancing the Genetic resources available for major crop conservation of crop wild relatives. The species are critical to world food security. A main recommendations included: very small number of species account for most human food. Cereals are the stable ● Development of an agreed definition of food sources for most human diets (Henry, crop wild relatives. 2001). The very large collections of these species contrast with the relatively small ● Preparation of comprehensive lists of genetic resource collections available for the species. minor regional crop species. Wild popula- tions of crop species or their relatives ● Identification and flagging of wild rela- remain very important genetic resources for tives of crops in existing germplasm data- most species. Protection of wild populations bases. ● Expansion of data on these entries. ● Conducting targeted research to identify conservation needs. ● Greater coordination of conservation efforts. Tertiary (genus/tribe/family) Secondary (genus) Primary (species) Fig. 15.1. The gene pool of cultivated species (adapted from Henry, 2000).

324 R.J. Henry Ornamental plants mercial forest species in the southern hemi- sphere and poplars are grown widely in Ornamental plants have often been highly northern regions. Other less widely available selected, and cultivated forms may appear species have very high value (e.g. oaks very different from wild ancestors. In some (Quercus robur and Quercus petrea), walnut cases the wild progenitors can no longer be (Jugulans regia) and mahogany (Swietinia identified. Often the apparent large pheno- humilis)). Genetic resources for most forest typic differences may be explained by rela- species are mainly to be found in situ in wild tively small genetic changes in loci populations and in forest plantations. contributing to key morphological traits. Dedicated ex situ forest germplasm collec- Gardens are important centres of conserva- tions are relatively limited. tion of plant diversity for ornamental and edible species. In some regions species Future Opportunities in Plant Genetic diversity in gardens exceeds that in the Resources wider environment. Ornamental plant con- servation is subject to the vagaries of fashion Advances in plant genetics are continuing to in gardens. Very high values may be placed expand the options for the recombination of on some rare plants. The high price may genes and genomes to generate novel assist in conservation of the species but often germplasm. Technologies for transfer of has a negative impact, generating collection uncharacterized genes from distant wild rel- pressure on wild genetic resources. The atives have been described (Abedinia et al., final outcome may be serious depletion of 2000). Recombinant DNA technologies allow wild populations and if this is followed by new variation of all types to be generated. loss of ex situ collections in gardens because Effective use of this capability is likely to of difficulty in cultivation or reduction in continue to rely on an understanding of the horticultural interest in the species, species arrangement and variation in plant diversity may be reduced and long-term sur- genomes and on the processes of genome vival of the species threatened. evolution. Forest plants Genomics has resulted in large amounts of genetic information becoming available Forest species are not yet well domesticated for many important plant species. The compared with crop plants (Campbell et al., expansion of DNA banks and rapid 2003). Most forest plantations use genetic improvements in the efficiency of DNA resources that have been subjected to rela- analysis tools promise the potential to better tively little selection or genetic improve- characterize and utilize plant germplasm in ment. A relatively small number of plant the future. Plant diversity conservation will species dominate forest planting worldwide. be greatly enhanced by these continuing These include gymnosperms (specifically the technological developments (especially in pines (Pinus)) and angiosperms (poplars molecular biology and bioinformatics), (Populus) and eucalypts (Eucalyptus)). The allowing very large amounts of biological eucalypts are the most widely planted com- and especially genetic data to be collected, stored and analysed.

Conserving genetic diversity 325 References Abedinia, M., Henry, R.J., Blakeney, A.B. and Lewin, L. (2000) Accessing genes in the tertiary gene pool of rice by direct introduction of total DNA from Zizania palustris (wild rice). Plant Molecular Biology Reporter 18, 133–138. Campbell, M.M., Brunner, A.M., Jones, H.M. and Strauss, S.H. (2003) Forestry’s fertile crescent: the applica- tion of biotechnology to forest trees. Plant Biotechnology Journal 1, 141–154. FAO (1996) The State of the World’s Plant Genetic Resources for Food and Agriculture. FAO, Rome. Godwin, I. (2003) Plant germplasm collections as sources of useful genes. In: Newbury, H.J. (ed.) Plant Molecular Breeding. Blackwell, Oxford, pp. 134–151. Henry, R.J. (2000) Technical advances in plant transformation providing opportunities to expand the cereal gene pool. In: O’Brien, L. and Henry, R.J. (eds) Transgenic Cereals. AACC, St Paul, Minnesota, pp. 252–276. Henry, R.J. (2001) Exploiting cereal genetic resources. Advances in Botanical Research 34, 23–57. Henry, R.J. (2004) Genetic improvement of cereals. Cereal Foods World 49, 122–129. Heywood, V.H. (ed.) (1978) Flowering Plants of the World. Oxford University Press, Oxford. Makinson, R. (2004) Botanic gardens and plant conservation. In: Henry, R.J. (ed.) Plant Conservation. Haworth Press, Binghampton, New York. Meilleur, B.A. and Hodgkin, T. (2004) In situ conservation of crop wild relatives: status and trends. Biodiversity and Conservation 13, 663–684. Nagamine, T. (2004) The role of genetic resources held in seedbanks. In: Henry, R.J. (ed.) Plant Conservation. Haworth Press, Binghampton, New York. Prance, G. (2004) Strategies for in situ plant conservation. In: Henry, R.J. (ed.) Plant Conservation. Haworth Press, Binghampton, New York. Rice, N. (2004) Conservation of plant genes and the role of DNA banks. In: Henry, R.J. (ed.) Plant Conservation. Haworth Press, Binghampton, New York. Thormann, I. (2004) Techniques for ex situ plant conservation. In: Henry, R.J. (ed.) Plant Conservation. Haworth Press, Binghampton, New York.

Index Abelia 21 Anthriscus 20 Acanthaceae 20 Antirrhinum 177 Acetate 231 AP3/PI-like genes 186 Acoraceae 11 Aphanopetalum 21 Acorales 11 Aphyllanthaceae 12 Acorus 11 Apiaceae 20 Adoxaceae 21 Apiales 20 Aextoxicaceae 21 Apocynaceae 20 Agathis 37 Apomixis 3, 128 Agavaceae 12 Aponogetonaceae 12 Agave 12 Aquifoliaceae 21 Agriculture 288 Aquifoliales 20, 21 Alismataceae 12 Arabidopsis thaliana 110, 119 Alismatales 12 Araceae 11, 12 Alkaloids 244 Aralia 20 Alliaceae 12 Araliaceae 20 Allium 12, 13 Araucaria 37 Allopolyploidy 99, 100 Araucariaceae 34, 37 Allozymes 142, 301 Archaeanthus 166 Aloe 12 Archaefructus 166 Alstroemeriaceae 13 Arecaceae 13 Altingiaceae 15 Arecales 13 Amaranthaceae 14 Arenosa 110 Amborella 10, 11, 166 Asparagaceae 12 Amborellaceae 10 Asparagales 12, 13 Ambrosia 21 Asteraceae 7, 12, 21, 54 Amino acids 231 Asterales 21 Amplified fragment length polymorphisms Asteridae 20 Asterids 19 (AFLP) 145 atpB 10 Anacardiaceae 19, 20 Atropa 20 Anethum 20 Aubergine 20 Angiosperm 2, 9 Aucuba 19 Angiosperm phylogeny group (APG) 7 Austrobaileyaceae 11 Angiosperm walls 206 Austrobaileyales 11, 166 Anthericaceae 12 Autopolyploidy 99, 100 Anthocyanins 14 327

328 Index Avena 13 Cell walls 201 Azaleas 19 Cellulose 202 Celtidaceae 18 Basal angiosperms 180 Centrospermae 14 Beans 17 Cephalotaceae 15 Begoniceae 17 Cephalotaxaceae 36 Behniaceae 12 Cephalotus 14 Bellis 21 Ceratophyllaceae 11, 22 Bennetitales 32 Ceratophyllum 11 Berberidopsidales 21 Cheirolepidaceae 36 Berberidosidaceae 21 Chicorium 21 Betelain 14 Chicory 21 Betulaceae 16, 17 Chloranthaceae 11, 22, 166 Bombacaceae 18 Chloroplast 45 Borage 19 Chloroplast DNA 145 Boraginaceae 19 Chloroplast genes 62 Botanic gardens 322 Chloroplast genome 46, 48 Bowenia 31 Chrysanthemum 21 Brassica napus 105, 110 Cleomaceae 18 Brassicaceae 16, 18, 119 Cleomoideae 119 Brassicales 18, 19 Climate change 85 Brochinnia 14 Clubmosses 2 Bromeliaceae 13, 14 Clusiaceae 16, 17 Broomrape 20 Coffea 20 Brugmannsia 20 Coffee 20 Bruniaceae 20 Colchicaceae 13 Bryophytes 2 Colchicum 13 Burmanniaceae 13 Combretaceae 17 Butomaceae 12 Commelinaceae 13 Commelinales 13 Cabombaceae 166 Compositae 21 Cactaceae 14 Coniferae 25 Calceolariaceae 20 Conifers 33 Calendula 21 Convallaria 12 Callistophyton 38 Convallariaceae 12 Callitris 35 Convolvulaceae 20 Callose 202 Cordaianthus 38 Campanula 21 Cordaites 32, 38 Campanulaceae 21, 54 Cordyline 12 Canellales 11 Cornaceae 19 Cannabaceae 18 Cornales 19 Caper 18 Cosmos 21 Capparaceae 18, 119 Crassulaceae 15 Caprifoliaceae 21 Crocus 12 Capsicum 20 Crop plants 323 Caricaceae 18, 19 Crossosomatales 16 Carrots 20 Crucifer 119 Caryophyllaceae 14 Cruciferae 18, 119 Caryophyllales 13, 14, 16 Cucurbitaceae 16, 17 Caryophyllids 14 Cucurbitales 17 Cashews 19 Cultivated plants 321 Castileja 20 Cunoniaceae 17, 21 Casuarinaceae 17 Cupressaceae 34, 36 Catalpa 20 Cycadaceae 25, 30, 31 Caytonia 38 Cycadae 25 Celastraceae 17, 19 Cycadales 25 Celastrales 17 Cycadinae 25

Index 329 Cycadophyta 25, 29 Fagaceae 16 Cycadophytina 25 Fagales 16, 17 Cycads 27, 29 Fennel 20 Cyclanthaceae 13 Feoniculum 20 Cymodoceaceae 12 Ferns 2 Cynara 21 Ferulic acid 207 Floral evolution 166 Dacrydium 36 Flower 165 Dahlia 21 Food 3 Dandelion 21 Forest plants 324 Daphniphyllaceae 16 Frankeniaceae 14 Daucus 20 Frankenias 14 Delphinium 14 Diapensiales 19 Galacto 203 Dicotyledon walls 206 Galactoglucomannans 211 Digitalis 20 Galanthus 12 Dilleniaceae 15, 16 Garrya 19 Dilleniidae 15, 16 Garryaceae 19 Dioscorea 12, 13 Garryales 16, 19 Dioscoreaceae 13 Gene copy number 185 Dioscoreales 12, 13 Gene flow 152 Diplosporous apomixis 128 Genetic assimilation 92 Dipsacales 21 Genetic erosion 290 Dipsacus 21 Genetic resource collections 321 Dipterocarpaceae 19 Genome 1 Dispersal 30 Gentianaceae 20 Diversity, importance of 1 Gentianales 20 Diversity, and productivity 257 Geraniaceae 16 DNA banks 322 Geraniales 16 DNA sequencing 59 Geranium 16 Domestication 287 Gesneriaceae 20 Dracaena 12 Ginkgo 27, 32 Dracaenaceae 12 Ginkgophyta 32 Droseraceae 14 Gladiolus 12 Glucomannan 203, 211 Ebenaceae 19 Glycine 104 Ebenales 19 Glycoproteins 211 Ecological systems 250 Gnetaceae 25, 38 Ecosystem 3, 254 Gnetales 27 Elaeocarpaceae 17 Gnetophyta 25 Endangered species 317 Gnetum 38 Ephedra 32, 38 Goodeniaceae 21 Ephedraceae 25, 38 Gossypium 109 Ericaceae 16, 19 Graminae 13 Ericales 19 Grossulariaceae 15 Ericas 19 Guttiferae 17 Eriospermaceae 12, 20 Gymnosperm walls 217 Eucommia 19 Gymnospermae 25, 27 Eucommiaceae 16 Gymnospermophyta 25 Eucommoniaceae 19 Gymnosperms 2, 25 Euphorbiaceae 17 Eupomatia 190 Haemodoraceae 13 Eupomatiaceae 191 Hamamelidaceae 15, 16 Evolutionarily significant units (ESU) 158 Hamamelidae 15, 16, 20 Hamamelidales 8 F1 hybrid 81 Hedera 20 Fabaceae 7, 8 Helianthus 21 Fabales 17, 18

330 Index Helwingiaceae 21 Lilium 13 Hemerocallis 12 Limnocharitaceae 12 Hesperocallidaceae 12 Liverwort 2, 45 Heteroxylans 203, 211 Loasaceae 19 Heuchera grossulariifolia 101 Lonicera 21 Hibiscus 18 Loranthaceae 15 Hippeastrum 12 Lycophytes 2 Homologous 106 Hordeum 13 Macrozamia 31 Hornworts 2 Madders 20 Horsetails 2 MADS-box 175 Hosta 12 Magnolia 166 Hostaceae 12 Magnoliales 11 Hyacinthaceae 12 Malpighiales 17 Hyacinthus 12 Malvaceae 16, 18 Hybrid fitness 84 Malvales 18, 19 Hybrid speciation 86 Mango 19 Hybridization 83 Mannan 203 Hydrangeaceae 19 Marchantia 45 Hydrocharitaceae 12 Marjoram 20 Mating system 150 Internal transcribed spacer of nuclear Medullosa 33 ribosomal DNA 120 Melanthiaceae 12 Melastomataceae 17 Introgression 88 Meliaceae 19 Inversions 52 Menyanthaceae 21 Inverted repeat 51 Merosity 173 Ipomoea 20 Metabolome 1 Iridaceae 12 Mevalonate 231 Isozymes 142 Microsatellite 144, 301 Milkweed 20 Jerusalem artichoke 21 Milkworts 18 Joinvilleaceae 54 Mints 20 Juglandaceae 17 Mitochondrial genome 69 Jujube 18 Mitochondrial transcript processing 73 Juncaginaceae 12 Monocotyledon walls 213 Juniper 35 Monophyly 8 Moraceae 18 Labiatae 20 Mosses 2 Lactuca 21, 304 Musaceae 13 Lamiaceae 20 Myrtaceae 17 Lamiales 20 Myrtales 16 Laurales 11 Lavender 20 Najadaceae 12 Lavendula 20 Narcissus 12, 13 Laxmanniaceae 12 Nelumbo 13 Lebachia 37 Nelumbonaceae 8, 13 Lecythidales 19 Nepenthaceae 14 Legume 53 Nicotiana 20, 45, 109 Legumes 17 Nilssonia 32 Lentibulariaceae 15 Nolinaceae 12 Leptocycas 32 Nothofagaceae 17 Lettuce 21 Nuclear DNA markers 143 Lignified secondary walls 211 Nuytsia 15 Lignins 211 Nymphaeaceae 11, 166 Liliaceae 12, 13 Nymphaeales 11 Liliales 12, 13

Index 331 Ocimum 20 Ploidy 87 Oleaceae 20 Plumbaginaceae 14 Onagraceae 12, 17 Poaceae 7, 13, 54 Orchidaceae 7, 12, 13, 21 Podocarpaceae 34, 36 Oregano 20 Podocarpus 35, 36 Oreganum 20 Polemoniaceae 19 Ornamental plants 324 Polemoniales 19 Orobanchaceae 20 Pollen 34 Orthologues 106 Polygonaceae 14, 18 Oryza 13, 109 Polyploidy, prevalence of 97 Oxalidaceae 17 Polyploidy, types of 99 Oxalidales 17, 21 Pontederiaceae 13, 14 Posidoniaceae 12 Paeoniaceae 15, 16 Potamogetonaceae 12 Palmae 13 Potato 20 Panax 20 Primary walls 206 Pandanaceae 13 Primulaceae 14, 16, 19 Pandanales 12, 13 Primulales 19 Pandanus 13 Proteaceae 8, 13 Papaveraceae 13 Proteales 8, 13, 16 Paralogous 106 Proteome 1 Paralogues 106 Pseudotsuga 35 Passifloraceae 16, 17, 19 Pteridophyte walls 219 Pastinaca 20 Pulses 17 Pectic polysaccharides 205, 206 Pedicularis 20 Random amplified polymorphic DNA Pelargonium 17 (RAPD) 145 Pentatricopeptide repeat (PPR) 73 Peppers 20 Ranunculaceae 13, 14, 54 Perianth 166 Ranunculales 13 Peridiscaceae 15 rbcL 10 Petrosaviales 12 RDNA (18S) 10 Petroselinum 20 Restoniaceae 54 Petunia 20 Restriction fragment length polymorphisms Phenome 1 Phenylpropanoid precursors 242 (RFLP) 143 Phormium 12 Reticulate evolution 81 Phyllocladus 37 Rhamnaceae 18, 19 Phyllonomaceae 21 Rhododendrons 19 Phyllotaxis 171 Rosaceae 18 Phylogeography 125 Rosales 15, 18 Physalis 20 Rose 18 Pinaceae 34, 35 Rosemary 20 Pines 34 Rosidae 20 Pinophyta 25, 33 Rosids 16 Pinus 34 Rosmarinus 20 Piperaceae 166 Rubiaceae 20 Piperales 11 Ruppiaceae 12 Pistachios 19 Ruscaceae 12 Pittosporaceae 20 Rutaceae 19 Plant populations 139 Plantaginaceae 20 Sage 20 Platanaceae 8, 16 Salvia 20 Platanales 8 Sambucus 21 Platanidae 8 Santalaceae 15 Platanus 8 Santalales 15 Platycodon 21 Sapindaceae 19 Sapindales 18, 19 Sapotaceae 19

332 Index Sarraceniaceae 15, 19 Thelypodieae 119 Sarraceniales 19 Themidaceae 12 Saxifragaceae 14, 15, 19 Thymelaeaceae 18 Saxifragales 15, 16, 21 Tiliaceae 18 Scabiosa 21 Tobacco 20, 45 Scaevola 21 Tofieldiaceae 12 Schisandraceae 11 Tomatillo 20 Schizanthus 20 Tomato 20 Sciadopityaceae 36 Tragopogon 104 Sciadopitys verticillata 36 Transcriptome 1 Scrophulariaceae 20 Trillium 12 Secondary metabolism 229 Trimeniaceae 11 Secondary walls 212 Triteleia 12 Seed banks 322 Triticum 13 Self-incompatibility 127 Trochodendraceae 16 Senecio cambrensis 104 Tropaeolaceae 18 SHAGGY-like kinases 188 Tulipa 13 Shikimate 231 Simondsiaceae 14 Ulmaceae 18 Simple sequence repeats (SSRs) 144 Umbelliferae 20 SKP1-like proteins 188 Urticaceae 18, 19 Smilacaceae 13 Snapdragon 20 Vanilla 12 Soil 258 Verbanaceae 20 Solanaceae 20 Verbena 20 Solanales 20 Veronica 20 Solanum 20 Viburnum 21 Soybean 104 Violaceae 17 Spartina anglica 104 Vitaceae 20 Speciation 123 Species diversity 266 Walchia 37 Spermatophyta 25 Wall evolution 221 Spermopteris 32 Welwitschia 38 Stangeria 31 Welwitschiaceae 38 Stangeriaceae 31 Wild barley 293 Stanleyeae 119 Wild emmer wheat 296 Stemonuraceae 21 Wild populations 321 Sterculiaceae 18 Wollemia 37 Sugars 231 Sunflower 21 Xanthorrhoea 12 Sweet potato 20 Xanthorrhoeaceae 12 Xyloglucans 202, 210 Tacca 13 Taeniopteris 32 Yucca 12 Tamaricaceae 14 Tannins 244 Zamiaceae 31 Taraxacum 21 Zea 13 Taxaceae 25, 35 Zingiberaceae 13 Taxus 35 Zizania 109 Teak 20 Zosteraceae 12 Tectonia 20 Zygophyllaceae 8 Theaceae 16, 19 Zygophyllales 8 Theales 19