Biology of Root Formation and Development (Basic Life Sciences)

Preface to the Third Edition

Evolution of roots was a fundamental development that enabled plants to migrate from aquatic to terrestrial habitats. Eventually it led to a division of functions between the carbohydrate-supplying shoot and the waterand mineral-supplying roots. Thus, it is not only the function of the individual shoots that constitutes the basis for the subsistence of global primary productivity and supports all animal and microbial life in terrestrial ecosystems, but also the concerted function of the shoots and roots. In spite of the great importance of roots for basic and applied scientific aspects and a long history of root research, there remains much to be learned about root development and functions. Roots are buried in the soil and any attempt to investigate their activity causes them irreversible damage. In addition, because roots live in very close contact with all other soil biota, the relationships may be so consolidated that it is hard to distinguish between a root as a plant organ and a root as a symbiont. Plant roots have remained an exciting and intriguing field of science. In the past five years, an exceptional proliferation of interest in root biology has developed, associated with intensive research activity in this field and contemporary developments in the understanding of root function and development. The new biological research involving the cloning and identification of genes and the control of their expression has deeply penetrated into root biology. More and more root-specific genes are being identified and the role of such genes in the control of root structure and function is being discovered. New methods and tools have been applied to root research, while old ideas and interpretations are being reexamined. As a result, it became necessary to update and expand our viewpoints. The chapters of this book are not intended to provide complete overviews. Our purpose in this expanded edition is to present a cross-section of the accomplishments of the past in and the future direction of root research. The third edition covers not only the numerous subjects of the previous editions but also the additional fields of genetics, molecular biology, growth-substance physiology, pH effects, biotechnology, and biomechanics. The book also tackles ecological problems and the multitude of interactions between roots and various types of soil organisms. We added some specific topics such as micropropagation, root signals, environmental sensing and direction finding, and expanded the scope of most of the other sections. Obviously, the chapters express the points of view of the writers. In some cases they are provocative. The presentation of a range of opinions introduces the reader to the current pressing questions in root studies, and points out new bearings for future research. The third edition serves as a major source of information for root scientists, botanists, plant physiologists, microbiologists, soil scientists, and those engaged in related professions. This book summarizes the previous information and designates the present frontiers of our knowledge in this field. The important questions that should be investigated in the future are pointed out. It presents a multidisciplinary view of the field of plant roots and its iii

iv

Preface to the Third Edition

state of the art. It covers various aspects of root origin, root structure, development and behavior, the interactions between roots and their environment, and the various uses of roots. The book contains 59 chapters, examining a wide range of critical topics and exploring potential applications and future directions. The following themes are covered: The origin and characteristics of roots Structure and development of root systems Root genetics Research techniques for root studies Regulation of root growth Physiological aspects of root systems Root growth under stress Root–rhizosphere interactions Roots of various ecological groups Root of economic value Some of the topics were given more attention than others, especially those in which the literature has proliferated recently or new techniques have been introduced. While we have made every effort to reach uniformity in terminology and style, the presented results, the expressed ideas, and the final shape of the chapters retain the personal imprint of the individual authors. We wish to express our sincere and warmest gratitude to all the eminent contributors, for their scholarly contribution and enthusiastic cooperation. Yoav Waisel Amram Eshel Uzi Kafkafi

Preface to the Second Edition

Plant roots have remained an exciting and intriguing field of science. During the years since the first edition of Plant Roots: The Hidden Half was published, an exceptional proliferation of interest in root biology has developed. This has been associated with the intensive research activity in this field and contemporary innovations in the understanding of root function and development. New methods have been applied, and old ideas and interpretations reexamined. Altogether, it became necessary to both update and expand the coverage of the first edition. We remain loyal to our original goals: to present the cardinal information about plant roots, to designate the frontiers of our knowledge in this field, and to point out the pressing questions that should be investigated. This book presents the state-of-the-art, multidisciplinary view of plant roots. It covers various aspects of root structure, development, and function; the interactions between roots and their environment; and the important benefits of roots to mankind. This is the story of the frontiers of the root sciences. It is a comprehensive volume, examining a wide range of critical topics, newly explored areas, potential applications, and future directions. The second edition covers nine aspects of the root sciences: Structure and development of the root system Methods of root studies Growth and metabolism Growth under stress Ion and water relations The rhizosphere Root pests Roots of various ecological groups Roots of economic value Because of the vast increase in information, this second edition is not only revised, but also expanded. Some of the topics were totally revised, especially those for which the literature has amply proliferated during the last five years or where new techniques of molecular genetics have been introduced. Notable among these are chapters related to root-pest interactions. Where necessary, some of the topics of the first edition were rewritten by new teams of root scientists. Thirty new contributors have joined those who participated in the first edition. Only one chapter was reproduced without change.

v

vi

Preface to the Second Edition

Three new themes are included in the second edition: 1. 2. 3.

Innovative methods of root studies Modeling and simulation Economic aspects of root biology

We have added some of the new topics to existing sections, whereas others are grouped into the following two new sections: Methods of Root Studies (Part II) includes chapters on modern techniques, fractal geometry, and simulation of root system architecture. Roots of Economic Value (Part IX) contains chapters addressing roots as a source of food, the biosynthetic potential of roots, and medicinal uses. Other new topics, including cell biology and development of root hairs, modeling of water uptake, simulation of ion uptake, drought rhizogenesis, and contractile roots, are distributed among the existing sections, according to their specific subject. The division of the 49 chapters into parts was rather difficult. In certain cases a chapter could fit equally well in more than one category. For example, Chapter 48, ‘‘Underground Plant Metabolism: The Biosynthetic Potential of Roots,’’ in Part IX, Roots of Economic Value, would have also been appropriate for Part III, Growth and Metabolism. Readers looking for a specific topic are therefore advised to browse through the entire volume. The numerous selected references, listed at the end of each chapter, direct the reader toward wider reading. These have been updated, many references cited in the first edition were replaced by new ones. Thus, the reader has the chance to unveil an entire epoch of root research, from 1672 to 1995. We wish to express our sincere gratitude to each of the eminent contributors for their scholarly contributions and for their enthusiastic and speedy cooperation. Yoav Waisel Amram Eshel Uzi Kafkafi

Preface to the First Edition

Roots, the ‘‘hidden half’’ of plants, serve a multitude of functions. They are responsible for anchorage, supply the plants with water and with nutrients, and exchange various growth substances with the shoots. Roots perform these functions in most ferns and in all seed plants, whereas additional traits (e.g., formation of storage organs, determination of the depth of the regenerating buds, or aeration of inundated organs) are characteristic of roots of exclusive groups of plants. The root–soil interface is the site where most interactions between the plants and their environment occur. Roots constitute a major source of organic material for the soil and thus affect its structure, aeration, and biological activities. While organic chemicals move out of the roots into the soil, inorganic ones move in: some of the entering materials are needed for normal metabolsim of the plants and are actively sought. Others are not required but are either neutral or toxic. Insufficient or excessive accumulation of most elements would damage plants, and therefore their uptake is controlled at the root surface. Our interest in the development and function of plant roots stems from the academic desire to understand their role in plant life, as well as from the important practical aspects they have. Most agricultural investment (i.e., plowing, seedbed preparation, irrigation, and fertilization) is spent to provide conducive conditions for the growth of roots of crop plants. Functional and healthy plant roots are essential for production of many of the resources on which human prosperity depends. The objectives of the present monograph are multiple: to review the recent contributions to the knowledge of the structure and function of roots, to outline the frontiers of root sciences, to point out the areas where gaps in knowledge exist, and to indicate the direction toward which basic and applied root research should proceed in the future. Plant Roots: The Hidden Half consists of 40 chapters that are grouped into the following seven sections. The first section deals with the structure and development of roots and with their assemblage into root systems. This section also tackles some of the genetic and physiological bases that control the development of individual roots and that eventually determine the structure, position, and function of roots systems. It also discusses in detail the individual root anatomy, with emphasis on root meristems, root caps, root hairs, and the cambium. The second section covers several aspects of the growth and metabolism of roots. It starts with the delicate hormonal relationships of roots and their effects on root growth and gravireaction, touches on various specific metabolic processes, and concludes with a discussion of root turnover and senescence. Studying roots, one can neither ignore the shoots nor disregard the constraints of the environment. The relationships between roots and shoots are, therefore, discussed here. The third section details with one of the very critical aspects of growth of plant roots—their encounter with stressing environments. This section covers the behavior of roots under temperature, low oxygen, heavy metal, and salinity stresses, as well as under the mechanical constraints of the soil. vii

viii

Preface to the First Edition

The fourth group of chapters discusses the uptake of minerals and water by roots and their subsequent transport into all other plant organs. Some chapters deal with nutrient availability, movement of ions across the soil–root interface, and the mechanisms of ion uptake. Three chapters discuss several aspects of the water relations of roots, and one chapter investigates an exceptional function of some roots: uptake of CO2. The various aspects of interactions that occur at the rhizosphere level are compiled in the fifth and sixth sections. Those include discussion of the interrelationship of roots with the biotic and abiotic components of their environment. In some cases the biotic interactions in the rhizosphere are obligatory and neither the roots nor the microorganisms can survive independently in nature. The application of these phenomena in agricultural practice is discussed in several of the chapters. In other cases the root–microorganism relationships are facultative. Still they may exert positive effects on the roots even if the soil organisms feed exclusively on root excretions and on their sloughed cells, without contacting live root cells. Other organisms exploit the roots as pathogens or as parasites and cause great damage to many crop plants. An understanding of the adaptation of plants to their respective habitats would be far from complete without a thorough knowledge of the role of their roots. Thus the last section discusses some of the aspects of plant adaptation at the ‘‘underground’’ level. Out of the great variety of ecological groups of plants that exist in nature, we have selected three unique ones, which present different trends of adaptation: desert plants, epiphytes, and submerged aquatic plants. An attempt to combine various topics of root science into one comprehensive treatise raises numerous difficulties, because of the complexity of integration of several aspects and because of the past division among different disciplines. Such a division has sometimes resulted in different and conflicting approaches and in a multiform terminology. For example, the term exudation is used by soil microbiologists to describe the release of organic materials by roots into their environment. The same term is used by plant physiologists for the completely different process of xylem sap flow out of the cut end of excised roots. The term growth is much less controversial. Still, some investigators use it to describe a general increase in root mass, whereas others employ it for description of the extension of a tap root or for initiation of primordia of lateral roots. These processes are intercorrelated but are certainly not interconvertible. Ion uptake by roots is a third example of the ambiguous use of terms. It was used by some scientists with the meaning of membrane transport, that is, a process that occurs between two solutions: one inside root cells and one outside them. These physiologists tended to disregard the roots’ environment, the existence of a boundary layer, and the contact with the charged surfaces of soil particles. Other investigators have referred to ion uptake in a much broader approach and included in this term all aspects of ion movement along and across root cells, including diffusion and mass flow through the free space in addition to the accumulation in the osmotic space. Soil scientists tend to look at the same process from their own point of view. To them, roots are organic cylinders of uniform quality that attract the soil solution. They describe ion uptake as a soil process that transports ions from the bulk of the soil into physiocochemical structures that by other people are named roots. In assembling this volume, our hope was to overcome the differences in approach to ‘‘roots’’ and to combine the intricate interrelationships of the complex quartet shoots–roots–soil–microorganisms into a single functional system. We have tried, as much as possible, to unify the vocabulary. Some of the contributors have agreed to use the more universal wording, whereas other adhered to the traditional sectorial terms. The wishes of those contributors were respected, and the battle for universal terms was left for future generations. We do hope that readers will not be annoyed by the multiform use of some terms, and we shall appreciate their patience and understanding. We have tried to keep each chapter complete and independent of the others. Therefore, some repetitions were unavoidable. To minimize overlapping without impairing the comprehensive nature of each chapter, such intersections were cross-referenced. The monograph is addressed to several groups of readers: professional scientists in plant sciences, agronomy, forestry, pathology, and soil science who seek to expand their knowledge to the related fields of root biology. The book is also directed to teachers at the university level and toward students of the rhizosphere sciences who intend to join the ‘‘avant garde’’ of root biology and would like to discover the pressing questions of this field. One of the unique features of this monograph is the attempt not only to present the objective state of the art in each domain, but also to emphasize the personal view of each contributor. How do the experts perceive the cardinal points in their own field? We felt that a treatise of this kind should form a rostrum for all manner of ideas, even for those that might seem a bit weird at present. After all, the unorthodox ideas are those that pave the way for progress in science. Although this monograph does not cover all aspects of the structure and function of roots, its backbone

Preface to the First Edition

ix

is composed of an interesting assortment of views combined with high-quality scientific information. We do hope that the result of our endeavors will ignite the imagination of readers and increase their interest in root biology. Yoav Waisel Amram Eshel Uzi Kafkafi

Contents

Preface to the Third Edition Preface to the Second Edition Preface to the First Edition Contributors I.

II.

iii v vii xv

The Origin and Characteristics of Roots 1.

The Origin of Roots ..............................................................................................................................1 Paul Kenrick

2.

Characteristics and Functions of Root Systems ................................................................................. 15 Alastair Fitter

The Root System: Structure and Development 3.

The Root Cap: Structure and Function.............................................................................................. 33 Andreas Sievers, Markus Braun, and Gabriele B. Monshausen

4.

Cellular Patterning in Root Meristems: Its Origins and Significance ................................................. 49 Peter W. Barlow

5.

Root Hairs: Hormones and Tip Molecules......................................................................................... 83 Robert W. Ridge and Masayuki Katsumi

6.

Secondary Growth of Roots: A Cell Biological Perspective............................................................... 93 Nigel Chaffey

7.

The Kinematics of Primary Growth ................................................................................................. 113 Wendy Kuhn Silk

8.

Lateral Root Initiation...................................................................................................................... 127 Pedro G. Lloret and Pedro J. Casero

9.

Functional Diversity of Various Constituents of a Single Root System .......................................... 157 Yoav Waisel and Amram Eshel xi

xii

Contents

III.

IV.

V.

10.

Biomechanics of Tree Root Anchorage ............................................................................................ 175 Alexia Stokes

11.

Root Systems of Arboreal Plants...................................................................................................... 187 Hans A˚. Persson

12.

Root–Shoot Relations: Optimality in Acclimation and Adaptation or the ‘‘Emperor’s New Clothes’’? ............................................................................................................... 205 Peter B. Reich

13.

Root Life Span, Efficiency, and Turnover ........................................................................................ 221 David M. Eissenstat and Ruth D. Yanai

Root Genetics 14.

Maize Root System and Genetic Analysis of Its Formation............................................................ 239 Gu¨nter Feix, Frank Hochholdinger, and Woong June Park

15.

Root Architecture—Wheat as a Model Plant................................................................................... 249 Gu¨nther G. B. Manske and Paul L. G. Vlek

16.

Banana Roots: Architecture and Genetics ........................................................................................ 261 Xavier Draye

17.

Molecular Root Bioengineering ........................................................................................................ 279 Marcel Bucher

Research Techniques for Root Studies 18.

Root Research Methods.................................................................................................................... 295 Janina Polomski and Nino Kuhn

19.

Aeroponics: A Tool for Root Research Under Minimal Environmental Restrictions..................... 323 Yoav Waisel

20.

Use of Microsensors for Studying the Physiological Activity of Plant Roots.................................. 333 D. Marshall Porterfield

21.

Rooting of Micropropagules ............................................................................................................. 349 Geert-Jan de Klerk

22.

Modeling Root System Architecture................................................................................................. 359 Loı¨c Page`s

The Regulation of Root Growth 23.

Auxins in the Biology of Roots ........................................................................................................ 383 Thomas Gaspar, Jean-Franc¸ois Hausman, Odile Faivre-Rampant, Claire Kevers, and Jacques Dommes

24.

Gibberellins ....................................................................................................................................... 405 Eiichi Tanimoto

25.

Roots and Cytokinins ....................................................................................................................... 417 R. J. Neil Emery and Craig A. Atkins

26.

Abscisic Acid in Roots—Biochemistry and Physiology.................................................................... 435 Eleonore Hose, Angela Sauter, and Wolfram Hartung

Contents

VI.

VII.

xiii

27.

Role of Ethylene in Coordinating Root Growth and Development ................................................ 449 Ahmed Hussain and Jeremy A. Roberts

28.

Root Signals ...................................................................................................................................... 461 Mark A. Bacon, William J. Davies, Darren Mingo, and Sally Wilkinson

29.

Environmental Sensing and Directional Growth of Plant Roots ..................................................... 471 D. Marshall Porterfield

30.

Root Growth and Gravireaction: A Critical Study of Hormone and Regulator Implications ...................................................................................................................... 489 Paul-Emile Pilet

31.

Calcium and Gravitropism................................................................................................................ 505 B. W. Poovaiah, Tianbao Yang, and A. S. N. Reddy

Physiological Aspects of Root Systems 32.

Respiratory Patterns in Roots in Relation to Their Functioning..................................................... 521 Hans Lambers, Owen K. Atkin, and Frank F. Millenaar

33.

Root pH Regulation.......................................................................................................................... 553 Jo´ska Gerenda´s and R. George Ratcliffe

34.

Nutrient Absorption by Plant Roots: Regulation of Uptake to Match Plant Demand .................. 571 Anthony D. M. Glass

35.

Dynamics of Nutrient Movement at the Soil–Root Interface .......................................................... 587 Albrecht O. Jungk

36.

Root-Induced Changes in the Availability of Nutrients in the Rhizosphere.................................... 617 Gu¨nter Neumann and Volker Ro¨mheld

37.

Simulation of Ion Uptake from the Soil........................................................................................... 651 Moshe Silberbush

38.

Soil Water Uptake and Water Transport Through Root Systems ................................................... 663 John S. Sperry, Volker Stiller, and Uwe G. Hacke

39.

Ecological Aspects of Water Permeability of Roots ......................................................................... 683 Andrea Nardini, Sebastiano Salleo, and Melvin T. Tyree

40.

Inorganic Carbon Utilization by Root Systems................................................................................ 699 Michael D. Cramer

Root Growth Under Stress 41.

Temperature Effects on Root Growth .............................................................................................. 717 Bobbie L. McMichael and John J. Burke

42.

Root Growth and Metabolism Under Oxygen Deficiency ............................................................... 729 William Armstrong and Malcolm C. Drew

43.

Trace Element Stress in Roots .......................................................................................................... 763 Ju¨rgen Hagemeyer and Siegmar-W. Breckle

44.

Root Growth Under Salinity Stress .................................................................................................. 787 Nirit Bernstein and Uzi Kafkafi

xiv

Contents

VIII.

IX.

X.

45.

High Soil Strength: Mechanical Forces at Play on Root Morphogenesis and in Root:Shoot Signaling ............................................................................................................. 807 Josette Masle

46.

Plant Roots Under Aluminum Stress: Toxicity and Tolerance ........................................................ 821 Hideaki Matsumoto

Root–Rhizosphere Interactions 47.

Root–Bacteria Interactions: Symbiotic N2 Fixation ......................................................................... 839 Carroll P. Vance

48.

Plant Growth Promotion by Rhizosphere Bacteria .......................................................................... 869 Yoram Kapulnik and Yaacov Okon

49.

Fungal Root Endophytes .................................................................................................................. 887 Thomas N. Sieber

50.

Mycorrhizae—Rhizosphere Determinants of Plant Communities .................................................... 919 Ingrid Kottke

51.

Root–Nematode Interactions: Recognition and Pathogenicity......................................................... 933 Hinanit Koltai, Edna Sharon, and Yitzhak Spiegel

52.

Interactions of Soilborne Pathogens with Roots and Aboveground Plant Organs .......................... 949 Jaacov Katan

Roots of Various Ecological Groups 53.

Ecophysiology of Roots of Desert Plants, with Special Emphasis on Agaves and Cacti ........................................................................................................................................... 961 Park S. Nobel

54.

Contractile Roots .............................................................................................................................. 975 Norbert Pu¨tz

55.

Roots of Banksia spp. (Proteaceae) with Special Reference to Functioning of Their Specialized Proteoid Root Clusters ......................................................................................... 989 John S. Pate and Michelle Watt

56.

Ecophysiology of Roots of Aquatic Plants..................................................................................... 1007 Craig Beyrouty

Roots of Economic Value 57.

Roots as a Source of Food ............................................................................................................. 1025 Daniel F. Austin

58.

Underground Plant Metabolism: The Biosynthetic Potential of Roots.......................................... 1045 Jorge M. Vivanco, Rejane L. Guimara˜es, and Hector E. Flores

59.

Roots as a Source of Metabolites with Medicinal Activity ............................................................ 1071 Zohara Yaniv and Uriel Bachrach

Index of Organism Names Subject Index

1093 1103

Contributors

William Armstrong, Ph.D., D.Sc., F.I. Biol., C. Biol. England Owen K. Atkin, Ph.D.

Department of Biology, University of York, York, England

Craig A. Atkins, Ph.D., D.Sc. Australia Daniel F. Austin, Ph.D.

Department of Biological Sciences, University of Hull, Hull,

Department of Botany, University of Western Australia, Perth, Western Australia,

Arizona-Sonora Desert Museum, Tucson, Arizona

Uriel Bachrach, Ph.D. Jerusalem, Israel

Department of Molecular Biology, The Hebrew University–Hadassah Medical School,

Mark A. Bacon, Ph.D.

Department of Biology, Lancaster University, Lancaster, Lancashire, England

Peter W. Barlow, D.Phil., D.Sc. Bristol, England

IACR-Long Ashton Research Station, University of Bristol, Long Ashton,

Nirit Bernstein, Ph.D. Institute of Soil, Water and Environmental Sciences, Agricultural Research Organization, The Volcani Center, Bet-Dagan, Israel Craig Beyrouty, Ph.D.

Department of Agronomy, Purdue University, West Lafayette, Indiana

Markus Braun, Dr. rer. nat. Siegmar-W. Breckle, Ph.D. Germany

Institute of Botany, University of Bonn, Bonn, Germany Department of Ecology, Faculty of Biology, University of Bielefeld, Bielefeld,

Marcel Bucher, Ph.D. Federal Institute of Technology, Institute of Plant Sciences, Plant Biochemistry and Physiology, Zurich, Switzerland

xv

xvi

Contributors

John J. Burke, Ph.D. Plant Stress and Germplasm Development Research Unit, U.S. Department of Agriculture– Agricultural Research Service, Lubbock, Texas Pedro J. Casero, Ph.D. Department of Morphological Science and Cellular and Animal Biology, Faculty of Science, Universidad de Extremadura, Badajoz, Spain IACR-Long Ashton Research Station, University of Bristol, Long Ashton, Bristol,

Nigel J. Chaffey, Ph.D. England

Department of Botany, University of Stellenbosch, Stellenbosch, South Africa

Michael D. Cramer, Ph.D.

Department of Biology, Lancaster University, Lancaster, Lancashire, England

William J. Davies, Ph.D.

Centre for Plant Tissue Culture Research, Lisse, The Netherlands

Geert-Jan de Klerk, Ph.D.

Department of Plant Biology, University of Lie`ge, Lie`ge, Belgium

Jacques Dommes, Ph.D. Xavier Draye, Ph.D. Belgium

Department of Applied Biology, Universite´ catholique de Louvain, Louvain-la-Neuve,

Department of Horticultural Sciences, Texas A & M University, College Station, Texas

Malcolm C. Drew, D.Phil. David M. Eissenstat, Ph.D. Pennsylvania

Department of Biology, Trent University, Peterborough, Ontario, Canada

R. J. Neil Emery, Ph.D. Amram Eshel, Ph.D.

Department of Horticulture, The Pennsylvania State University, University Park,

Department of Plant Sciences, Tel Aviv University, Tel Aviv, Israel

Odile Faivre-Rampant, Ph.D. Gu¨nter Feix, Dr. rer. nat. Alastair Fitter, Ph.D.

Department of Plant Biology, University of Lie`ge, Lie`ge, Belgium

Institute of Biology III, University of Freiburg, Freiburg, Germany

Department of Biology, University of York, York, England

Hector E. Flores, Ph.D. Pennsylvania

Department of Plant Pathology, The Pennsylvania State University, University Park,

Thomas Gaspar, Ph.D.

Department of Plant Biology, University of Lie`ge, Lie`ge, Belgium

Jo´ska Gerenda´s, Ph.D.

Institute for Plant Nutrition and Soil Science, University of Kiel, Kiel, Germany

Anthony D. M. Glass, Ph.D. Columbia, Canada

Department of Botany, University of British Columbia, Vancouver, British

Rejane L. Guimara˜es Pennsylvania

Department of Plant Pathology, The Pennsylvania State University, University Park,

Uwe G. Hacke, Ph.D.

Department of Biology, University of Utah, Salt Lake City, Utah

Ju¨rgen Hagemeyer, Priv-Doz.Dr. Germany

Department of Ecology, Faculty of Biology, University of Bielefeld, Bielefeld,

Contributors

xvii

Julius von Sachs Institut, University of Wu¨rzburg, Wu¨rzburg, Germany

Wolfram Hartung, Ph.D.

Jean-Franc¸ois Hausman, Ph.D.

Crebs Research Unit, CRP-Gabriel Lippman, Luxembourg

Frank Hochholdinger, Dr. rer. nat.

Julius von Sachs Institut, University of Wu¨rzburg, Wu¨rzburg, Germany

Eleonore Hose, Ph.D. Ahmed Hussain, Ph.D.

Division of Plant Science, School of Biosciences, University of Nottingham, England Institute for Agricultural Chemistry, Georg-August University of Go¨ttingen, Go¨ttingen,

Albrecht O. Jungk, Ph.D. Germany Uzi Kafkafi, Ph.D. Rehovot, Israel

Institute of Biology III, University of Freiburg, Freiburg, Germany

Department of Field Crops, Vegetables and Genetics, The Hebrew University of Jerusalem,

Yoram Kapulnik, Ph.D. Department of Agronomony and Natural Research Institute of Field and Garden Crops, Agricultural Research Organization, The Volcani Center, Bet-Dagan, Israel Masayuki Katsumi, Ph.D. Tokyo, Japan

Department of Plant Pathology and Microbiology, The Hebrew University of Jerusalem,

Jaacov Katan, Ph.D. Rehovot, Israel Paul Kenrick, Ph.D.

Division of Natural Sciences, Biology Department, International Christian University,

Department of Paleontology, The Natural History Museum, London, England

Claire Kevers, Ph.D.

Department of Plant Biology, University of Lie`ge, Lie`ge, Belgium

Hinanit Koltai, Ph.D. Dagan, Israel

Department of Nematology, Agricultural Research Organization, The Volcani Center, Bet-

Ingrid Kottke, Ph.D. Department of Systematic Botany, Mycology and Botanical Garden, Botanical Institute, University of Tu¨bingen, Tu¨bingen, Germany Nino Kuhn, Dr. sc. techn. Department of Forest and Environmental Protection, Swiss Federal Research Institute, Birmensdorf, Switzerland Hans Lambers, Ph.D. Australia, Australia

Department of Plant Sciences, University of Western Australia, Crawley, Western

Pedro G. Lloret, Dr.Sc. Department of Morphological Science and Cellular and Animal Biology, Faculty of Science, Universidad de Extremadura, Badajoz, Spain Gu¨nther G. B. Manske, Ph.D.

Center for Development Research, University of Bonn, Bonn, Germany

Josette Masle, Ph.D. Research School of Biological Sciences, Institute of Advanced Studies, Australian National University, Canberra, Australia Hideaki Matsumoto, Ph.D.

Research Institute for Bioresources, Okayama University, Okayama, Japan

xviii

Contributors

Bobbie L. McMichael, Ph.D. Plant Stress and Germplasm Development Research Unit, U.S. Department of Agriculture–Agricultural Research Service, Lubbock, Texas Frank F. Millenaar, Ph.D. Darren Mingo, B.Sc.

Department of Plant Ecophysiology, Utrecht University, Utrecht, The Netherlands

Department of Biology, Lancaster University, Lancaster, Lancashire, England

Gabriele B. Monshausen, Dr. rer. nat.

Department of Biology, University of Trieste, Trieste, Italy

Andrea Nardini, Ph.D.

Institut fu¨r Pflanzenerna¨hrung, University of Hohenheim, Stuttgart, Germany

Gu¨nter Neumann, Ph.D. Park S. Nobel, Ph.D. Angeles, California

Institute of Botany, University of Bonn, Bonn, Germany

Department of Organismic Biology, Ecology and Evolution, University of California, Los

Yaacov Okon, Ph.D. Department of Phytopathology and Microbiology, Faculty of Agriculture, The Hebrew University of Jerusalem, Rehovot, Israel Loı¨ c Page`s, Ph.D.

Unite´ Plantes et Syste`mes de Culture Horticoles, INRA, Centre d’Avignon, Avignon, France Institute of Biology III, University of Freiburg, Freiburg, Germany

Woong June Park, Ph.D.

John S. Pate, D.Sc., F.R.S., F.A.A., F.L.S. Western Australia, Australia

Department of Botany, University of Western Australia, Nedlands,

Hans A˚. Persson, Ph.D. Department of Ecology and Environmental Research, Swedish University of Agricultural Sciences, Uppsala, Sweden Institute of Biology and Plant Physiology, University of Lausanne, Lausanne, Switzerland

Paul-Emile Pilet, Ph.D.

Janina Polomski, M.Agr. Department of Forest and Environmental Protection, Swiss Federal Research Institute, Birmensdorf, Switzerland B. W. Poovaiah, Ph.D. Department of Horticulture and Program in Plant Physiology, Washington State University, Pullman, Washington D. Marshall Porterfield, Ph.D. Norbert Pu¨tz, Dr. rer. nat

Department of Biological Sciences, University of Missouri-Rolla, Rolla, Missouri

Institute for Environmental Education, University of Vechta, Vechta, Germany

R. George Ratclifffe, D. Phil.

Department of Plant Sciences, University of Oxford, Oxford, England

A. S. N. Reddy, Ph.D. Department of Biology and Program in Cell and Molecular Biology, Colorado State University, Fort Collins, Colorado Peter B. Reich, Ph.D.

Department of Forest Resources, University of Minnesota, St. Paul, Minnesota

Robert W. Ridge, Ph.D. Tokyo, Japan Jeremy A. Roberts, Ph.D.

Division of Natural Sciences, Biology Department, International Christian University,

Division of Plant Science, School of Biosciences, University of Nottingham, England

Contributors

xix

Volker Ro¨mheld, Dr. agr. ing.

Department of Biology, University of Trieste, Trieste, Italy

Sebastiano Salleo, Ph.D. Angela Sauter

Institut fu¨r Pflanzenerna¨hrung, University of Hohenheim, Stuttgart, Germany

Julius von Sachs Institut, University of Wu¨rzburg, Wu¨rzburg, Germany Agricultural Research Organization, The Volcani Center, Bet-Dagan, Israel

Edna Sharon, Ph.D.

Department of Forest Sciences, Swiss Federal Institute of Technology, Zurich,

Thomas N. Sieber, Ph.D. Switzerland Andreas Sievers, Dr. rer. nat.

Institute of Botany, University of Bonn, Bonn, Germany

Moshe Silberbush, Ph.D. Wyler Department of Dryland Agriculture, Jacob Blaustein Institute for Desert Research, Ben-Gurion University of the Negev, Sede-Boker, Israel Wendy Kuhn Silk, Ph.D. California

Department of Land, Air, and Water Resources, University of California, Davis,

John S. Sperry, Ph.D.

Department of Biology, University of Utah, Salt Lake City, Utah

Yitzhak Spiegel, Ph.D. Dagan, Israel

Professor of Nematology, Agricultural Research Organization, The Volcani Center, Bet-

Volker Stiller, Ph.D.

Department of Biology, University of Utah, Salt Lake City, Utah

Alexia Stokes, Ph.D.

Laboratoire de Rhe´ologie du Bois de Bordeaux, Cestas Gazinet, France

Eiichi Tanimoto, Dr.Sci. Department of Information and Biological Sciences, Graduate School of Natural Sciences, Nagoya City University, Nagoya, Japan Melvin T. Tyree, Ph.D. Aiken Forestry Sciences Laboratory, U.S. Department of Agriculture Forest Service, South Burlington, Vermont U.S. Department of Agriculture–Agricultural Research Service, University of Minnesota,

Carroll P. Vance, Ph.D. St. Paul, Minnesota Jorge M. Vivanco, Ph.D. Fort Collins, Colorado Paul L. G. Vlek, Ph.D. Yoav Waisel, Ph.D.

Department of Horticulture and Landscape Architecture, Colorado State University,

Center for Development Research, University of Bonn, Bonn, Germany

Department of Plant Sciences, Tel Aviv University, Tel Aviv, Israel

Michelle Watt, Ph.D. Plant Cell Biology Group, Research School of Biological Science, Australian National University, Canberra, Australia Sally Wilkinson, Ph.D. Ruth D. Yanai, Ph.D. New York

Department of Biology, Lancaster University, Lancaster, Lancashire, England College of Environmental Science and Forestry, State University of New York, Syracuse,

xx

Tianbao Yang, Ph.D.

Contributors

Department of Horticulture, Washington State University, Pullman, Washington

Zohara Yaniv, Ph.D. Department of Genetic Resources and Seed Research, Agricultural Research Organization, The Volcani Center, Bet-Dagan, Israel

1 The Origin of Roots Paul Kenrick The Natural History Museum, London, England

I.

INTRODUCTION

the silent and unobtrusive cycling of minerals from soil to biosphere on a truly prodigious scale (Epstein, 1977), and root-mediated weathering of silicates is thought to have had worldwide consequences for the carbon cycle (Berner, 1998). This chapter reviews the origin and early evolution of roots in the major groups of vascular plants, focusing on new data from the fossil record, which is interpreted using modern phylogenetic methods. In addition to providing a narrative of what happened and when, the evolution of root systems and the homologies of roots are considered in detail. One aim is to provide the experimental biologist with a rational for choosing model organisms for research in root developmental biology. A second aim is to provoke the reader into testing some of the ideas of homology, which have consequences that could be examined at the molecular developmental level.

Roots originated during the early phase of the diversification of plants on land in the Devonian Period, some 363 million to 409 million years ago. This was a time of enormous changes in plant life that were to have far-reaching consequences for the evolution of land animals and for the chemical economy of life on earth (Epstein, 1977; DiMichele et al., 1992; Kenrick and Crane, 1997b; Algeo and Scheckler, 1998; Bateman et al., 1998; Berner, 1998). From small rootless organisms a few centimeters tall, plants evolved into large shrubs and trees with a range of specialized rooting structures. Roots combined with a fully integrated vascular system were essential to the evolution of large plants, enabling them to meet the requirements of anchorage, water, and nutrient acquisition. Large roots with secondary wood were widespread by the Late Devonian (378Ma) (Retallack, 1986; Algeo and Scheckler, 1998), possibly earlier (Elick et al., 1998), and this development coincided with the appearance of the earliest forest ecosystems. The impact of roots on the evolution of soils was enormous. Physical effects, such as the fracturing of rock, the binding of loose particles, and the introduction of large quantities of organic material, combined with the chemical consequences of actively pumping solutes through the system, led to the development of soils with modern profiles. Plants ploughed, tilled, and fertilized the land, and in so doing had a lasting influence on Earth geochemistry. The origin of roots set in motion

II.

ROOTS IN THE FOSSIL RECORD

Paleobotanists are continually hampered in their attempts to reconstruct plants from the past because much of the fossil record comprises broken or disarticulated fragments preserved in sediments deposited in rivers or lakes. Piecing together plants to form a conceptual whole is a difficult and painstaking task, and frequently one must be satisfied with an incomplete organism, simply because parts are missing. These missing parts are often the roots. Despite the difficul1

2

Kenrick

ties inherent in reassembling whole plants, roots are occasionally preserved in fossils that have literally been uprooted and transported some distance. Also, it is possible to find environments where plants have been fossilized at their site of growth. Under circumstances such as these, preservation of roots can be exceptionally good. A.

Early Rooting Systems

1. The Rhynie Chert—An Early Terrestrial Biota The earliest examples of rooting structures preserved in growth position come from the 396-million-year-old Rhynie Chert from Scotland. This sequence of fossiliferous cherts—rocks composed of finely crystalline silica—provides a window onto an early terrestrial environment, capturing a period when plant life on land was at an early stage of development. Studies of the depositional environment of the chert show that plants grew on sandy substrates in and around the margins of pools on a plain that was receiving sediment from surrounding higher ground. Hydrothermal activity, driven by local volcanism, caused periodic flooding of the low-lying areas with hot water saturated in silicates, resulting in inundation and preservation of whole plants and the formation of the cherts (Trewin, 1994; Rice et al., 1995). Petrographic thin sections are the most widely used method of investigating and reconstructing the plants, and they reveal amazing details of cell structure. The first description of four vascular plants by Kidston and Lang (1921) revolutionized our understanding of the early land flora. Since then, the Rhynie Chert has continued to yield remarkable new information on the morphology and life cycles of the plants and arthropods of the early Devonian (Kevan et al., 1975; Rolfe, 1980; Edwards and Lyon, 1983; Lyon and Edwards, 1991; Taylor et al., 1992a,b; Remy et al., 1993; Kenrick, 1994; Remy et al., 1994; Taylor et al., 1995, 1997, 1999). 2. Stems Functioning as Roots The Rhynie Chert plants were all small and there is strong evidence that several species lacked roots. In these plants, absorptive and perhaps anchoring functions were performed by nonseptate rhizoids borne on prostrate stems (rhizomes) (Fig. 1). Rhizoids of this type have been documented in the extinct protracheophytes Horneophyton (Kidston and Lang, 1920a, 1921) and Aglaophyton (Kidston and Lang, 1920a). In Aglaophyton, rhizoid structure and development have

Figure 1 Unicellular rhizoids on the rooting stem of a fossil plant from the 396-million-year-old Rhynie Chert, Scotland. The Rhynie plants are silicified and have been fossilized in their growth position.

been studied in some detail (Edwards, 1986; Remy and Hass, 1996). Rhizoids developed from stomatal subsidiary cells. In rhizomes, these epidermal cells underwent division to produce a series of rectangular rhizoid mother cells. Simultaneously, cell division in the hypodermis produced a raised area on the stem. Mature rhizoids were therefore borne on small mounds on the lower surface of the rhizome and probably also on basal regions of erect stems. Stomates were interspersed among the rhizoids. Another specialization included the differentiation of some hypodermal cells into transfusion tissue that was connected with the vascular system. The evidence from Aglaophyton indicates that in protracheophytes (Fig. 2a) there was no clear developmental distinction between stem and rhizome. The rhizome was simply an older portion of the stem that underwent a limited amount of further differentiation to function as a rooting system. The development of rhizomes from prostrate stems resembling aerial branches was also a characteristic of many early vascular plants. The rhizome was much branched in groups such as zosterophylls—early relatives of living clubmosses (Lang, 1927; Lele and Walton, 1961; Gerrienne, 1988; Hao, 1989a) (Fig. 2b). In some species, multicellular trichomelike appendages were borne on both rhizomes and erect stems

Origin of Roots

3

(e.g., Hao, 1989b). The function of these appendages is unknown. Rhizoid development on rhizomes and at the basal regions of stems was probably widespread, but details such as these are seldom preserved in compression fossils formed in sandstone and shale. Exceptionally, preservation can occur where there has been substantial cutinization that extends to the rhizoids themselves, a phenomenon that has been observed in both bryophytes and vascular plants. One such example is the zosterophyll Serrulacaulis (Hueber and Banks, 1979). Preserved cuticles show that large deltoid spines were borne on stems of all sizes and that nonseptate heavily cutinized rhizoids developed on and among spines in the basal regions. 3.

Figure 2 Phylogenetic relationships among basal land plants. y ¼ extinct. A general summary is presented in part (a) and details of the two major groups within vascular plants in (b) (lycopsids) and (c) (euphyllophytes). (a) Relationships among land plants. Most phylogenetic analyses resolve bryophytes as paraphyletic to vascular plants, but relationships among the three major groups of bryophytes depicted are poorly resolved (for recent reviews and alternative hypotheses see Duff and Nickrent, 1999; Hedderson et al., 1998; Kenrick, 2000; Renzaglia et al., 2000). Note that nearly all known early fossil land plants are vascular plants or stem group vascular plants (Kenrick, 2000; Kenrick and Crane, 1997a,b). (b) Relationships among lycopsids with aspects of root evolution depicted (boxed text) (adapted from Kenrick and Crane, 1997a). For a detailed discussion of root homologies in lycopsids, see Rothwell and Erwin (1985). (c) Relationships among euphyllophytes from Rothwell (1999). For alternatives ideas see Kenrick and Crane (1997a), Pryer et al. (1995), and Stevenson and Loconte (1996).

Branches with Uniquely Rooting Function

An increased level of differentiation occurred in the early vascular plant Asteroxylon, also from the Rhynie Chert. Unlike the aerial stems, the rhizome was leafless and there were differences in the vascular system, which in stems was stellate but in the rhizoids was terete or elliptical. This difference is probably a consequence of the absence of vascular appendages (microphyllous leaves) in the rhizome. All branching appears to have been exogenous and there is no evidence of rhizoids or root hairs (Kidston and Lang, 1920b, 1921). Asteroxylon is related to lycopsids (Fig. 2b), and broadly similar roots have been documented in other less well preserved compression fossils of closely related plants such as Bathurstia (Kotyk and Basinger, 2000) and Drepanophycus (Rayner, 1984). These fossils document the early appearance of multicellular vascularized axes that are uniquely rooting in function. Dormant meristematic regions that have the potential to develop into new stems and perhaps roots were common to many early vascular plants (e.g., zosterophylls). These were most frequent on rhizomes but were also present on erect stems where they were usually confined to branch axils (Kenrick and Crane, 1997a). Sometimes they extended over a greater portion of the erect system (Edwards and Kenrick, 1986). These branches were exogenous, and they are known to have developed into new aerial branches in some species, whereas in others it is possible that they developed into roots (Edwards and Kenrick, 1986; Remy et al., 1986; Li, 1992). The pattern of distribution and the developmental plasticity of these enigmatic meristematic regions are notable similarities to the widely discussed rhizophore of living Selaginellaceae. Phylogenetic evidence indicates, however, that homol-

4

Kenrick

ogy is highly unlikely (Kenrick and Crane, 1997a) (Fig. 2b). The reason for this is that Selaginellaceae are phylogenetically remote from those zosterophylls with dormant meristems. Assuming homology of these structures would therefore be highly unparsimonious because it would imply multiple losses of dormant meristems/rhizophores in intervening groups. It has been suggested that the function of dormant meristems in the life cycle of the plant was to facilitate vegetative growth or reproduction or perhaps to enable regeneration following catastrophic burial by flood. 4. Bipolar Growth Plants with upright trunks and larger rooting systems are present in the fossil record by the Mid-Devonian (ca 380Ma). On the whole, these plants were still small and most had no secondary tissues. The Cladoxylopsida were a group of fernlike plants of uncertain systematic position. They are clearly euphyllophytes (i.e., members of the group that contains modern seed plants, ferns, and horsetails; Fig. 2c), but their precise relationships are poorly understood. This group has been implicated in the origins of both ferns and horsetails (Stein et al., 1984; Rothwell, 1999). One uprooted specimen, Lorophyton goense, provides an insight into the morphology of the whole plant. L. goense had a 2-cm-diameter trunk that bore tufts of leaflike branches from the apex and numerous bifurcating roots (maximum 1 cm diameter) from a slightly flared base. The whole plant is estimated to have been about 40 cm in height (Fairon-Demaret and Li, 1993). Schweitzer and Li (1996) have documented a similar growth form in the 50-cm-tall lycopsid Chamaedendron by the early part of the Late Devonian (Frasnian; 377Ma). The roots of Chamaedendron bifurcated at least four times and were at least 15 cm long (Fig. 3). These fossils mark an early departure from essentially unipolar to a form of bipolar growth. Roots were clearly differentiated as such, and they did not function as stems at any time during their ontogeny. B.

Roots of Trees

Trees first appear in the fossil record in the Late Devonian, and arborescence was widespread in progymnosperms—the free-sporing ancestors of living seed plants—and in clubmosses and horsetails, groups whose living relatives are predominantly herbaceous (Mosbrugger, 1990). This dramatic increase in plant

Figure 3 Uprooted fossil. A 377-million-year-old clubmoss (lycopsid) Chamaedendron multisporangiatum from China (Schweitzer and Li, 1996). The specimen illustrated is 35 cm long and shows a bifurcating root systems (R), a bifurcating stem (S) bearing numerous microphyllous leaves, and a slightly expanded root/stem junction (R/S).

Origin of Roots

size was a direct consequence of the evolution of the cambium, an innovation that also led to the development of much more extensive root systems. 1.

Horsetails

Paleozoic tree horsetails had roots that are very similar to those of their living relatives, which are all herbaceous (Eggert, 1962). The rhizome of living Equisetum is a modified stem that produces aerial branches and adventitious roots. During the Carboniferous the related tree horsetail Calamites grew to a height of 20 m. The rooting system was comparatively shallow and comprised an enormous, 40-cm-diameter horizontal rhizome that bore adventitious roots in whorls at nodes, much like branches on the erect stems. Roots also developed at the base of upright branches. Anatomically, the calamite rhizome was very similar to the stem, and it contained a large amount of secondary wood. The older and somewhat smaller horsetail Archaeocalamites (Early Carboniferous; Tournaisian; 363Ma) was of similar construction but had a greatly expanded woody rootstock at the base of a stem that bore numerous adventitious roots (Bateman, 1991). Scrambling or climbing forms such as Sphenophyllum produced adventitious roots at nodes, often with leaves. In this plant, it seems likely that as stem length increased basal portions assumed a horizontal position (Stewart and Rothwell, 1993). The rhizome-based rooting system of horsetails probably originated with plants such as the Mid-Devonian Metacladophyton, in which roots (8 cm long; 2.5 mm wide) were borne along one side of a horizontal rhizome (1 cm diameter) that developed into an upright stem (Wang and Geng, 1997). 2.

Clubmosses

Paleozoic tree clubmosses had highly distinctive roots that are termed rhizomorphs. The rhizomorph was a determinate structure that branched dichotomously to form a shallow but laterally extensive root system that would have exceeded 10 m in diameter in larger trees (Frankenberg and Eggert, 1969). Root apices terminated abruptly and bore a rimlike apical groove (Rothwell, 1984). Rootlets were cylindrical and helically arranged, developed near the apical meristem, and are thought to have been exogenous (Rothwell and Erwin, 1985). Each rootlet contained a large, airfilled cavity. No evidence of root caps or root hairs has been documented, but this may reflect poor preservation and lack of data from rootlet apices. On older portions, secondary growth of cortex and stele caused

5

rootlets to be abscised. Other Late Paleozoic arborescent clubmosses had a rather different rooting structure. Instead of the extended branched rhizomorph, plants such as Chaloneria and Paurodendron had unbranched rootstocks that were comparatively small and rounded, and that had a lobed cormose base (Pigg and Rothwell, 1983; Rothwell and Erwin, 1985). Rootlets of both the rhizomorphic and cormose club mosses had a unique type of stele located at the periphery of a large cavity. This is one of several striking anatomical similarities with the roots of their living relatives, Isoetes (Stewart, 1947). 3.

Tree Ferns

Tree ferns buttressed by extensive root mantles first appeared in the Early Carboniferous and became abundant during the Late Carboniferous and Permian. In general habit the earliest tree ferns resembled modern Dicksoniaceae and Cyatheaceae, but early fossils such as Psaronius are in fact most closely related to the Marattiales (Fig. 2c). Trunks of Psaronius belong to a plant that is estimated to have grown to 10 m in height. An extensive mantle of bifurcating roots that was widest at the base of the plant surrounded a comparatively narrow stem comprising exclusively primary tissues. Leaves were produced from a narrow crown. Roots arose endogenously from the peripheral vascular bundle of the stem and grew downward and outward to the periphery of the trunk. The roots of Psaronius are among the largest known for ferns, growing up to 2 cm in diameter (Ehret and Phillips, 1977). In some families of ferns and possibly also in some early seed plants, root mantles provided a ready alternative to cambial activity as a means of structural support in the evolution of the tree habit. 4.

Progymnosperms

Progymnosperms are a grade of extinct, free sporing plants that are closely related to seed plants (Fig. 2c). In many aspects of their morphology, progymnosperms are intermediate between pteridophytes and gymnosperms. Plants bore fernlike foliage and they ranged from small or medium-size homosporous shrubs (Aneurophytales) to large heterosporous trees (Archaeopteridales). Unlike ferns, progymnosperms produced gymnospermous wood from a bifacial cambium—one that produces both secondary xylem and phloem (Beck, 1971; Scheckler and Banks, 1971; Beck and Wight, 1988). Similarities with gymnosperms extended also to the roots (Scheckler, 1995).

6

Aneurophytales—the first progymnosperms to appear in the Mid-Devonian (Eifelian)—had bifurcating roots that were probably determinate. Most roots were endogenous, and some were adventitious. Cambial activity is evident in roots and rootlets of all sizes. Roots possessed cortical lacunae, periderm, persistent root hairs and possibly an exodermis. There is no evidence of mycorrhizae (Scheckler, 1995). Recent data from petrified trunks indicate that in addition to a well-developed root system with secondary wood, Archaeopteris had adventitious latent primordia similar to those produced by some living trees, which eventually develop into roots on stem cuttings (MeyerBerthaud et al., 1999). In the progymnosperm Eddya sullivanensis—possibly a juvenile Archaeopteris—there was a complex tap root system, comprising a strong main root that gave rise to a profusion of small, much branched, probably endogenous laterals (Beck, 1967).

Kenrick

(Delevoryas, 1955). Early seed plants occupied a broad range of environments including peat-forming swamps, flood plains, extra basin lowlands, and probably even higher ground (DiMichele et al., 1992). In some swamp-dwelling forms there is evidence for adaptation to growth under very wet conditions. Cordaites was a group of scrambling and upright shrubs or small trees that is related to conifers (Rothwell, 1988). Some Carboniferous swamp species had highly branched stilt roots that supported the stem (Cridland, 1964). Anatomically, these roots had an exarch actinostele with secondary xylem similar to that in the stem, well-developed periderm, lenticels that formed in the periderm, and a broad zone of aerenchymatous phelloderm that formed around the stele. The combination of aerenchyma, medullated protosteles, periderm, and lenticels has been likened to the stilt roots of plants growing in modern mangrove environments (Cridland, 1964).

5. Seed Plants The earliest seed plants were small to medium-size shrubs, many of which had a scrambling or semi-selfsupporting habit (Rowe et al., 1993). In this respect they were very similar to aneurophytalean progymnosperms. Adventitious roots borne along the stem are known to have been a feature of many species. In Callistophytaceae roots on aerial stems were axillary to buds or branches (Rothwell, 1975), they were also closely associated with leaves in Calamopityaceae (Meyer-Berthaud and Stein, 1995). However, in Lyginopteridaceae such roots were sometimes organized into vertical rows without apparent relation to other appendages (Taylor and Taylor, 1993). In all species there were many anatomical similarities between stems and roots. The stele was well defined and in some species possessed endodermis and pericycle. Secondary xylem and periderm were present in larger roots. The primary xylem was usually protostelic and diarch. Structures resembling secretory cells and resin canals have been documented in some species (Taylor and Taylor, 1993). Lateral roots were endogenous. Other early seed plants such as Medullosa were small trees (
4–5 m tall) with an upright trunk, and large, bifurcating, fernlike fronds (Stidd, 1981). The lower portion of the stem bore a periderm and numerous adventitious roots, and these roots are a common component of Carboniferous petrifactions (Rothwell and Whiteside, 1974). Roots turned downward between leaf bases and emerged near the base of the stem. Larger roots produced secondary xylem

III.

DISCUSSION

The early fossil record of land plants documents a continuum of variation in the evolution of rooting systems that began with simple rhizoid-bearing stems and developed, over a period of 40 million years, into a broad range of complex multicellular organs specialized in anchorage and nutrient acquisition. To define the term ‘‘root’’ one must draw an arbitrary line in this morphological continuum, and in applying the term to fossils further complications arise from a lack of anatomical information. Frequently, data on the presence/ absence of root cap and root hairs and on root development (endogenous or exogenous) are not available. Here, roots are defined as multicellular vascular organs without leaves or other multicellular appendages that function in anchorage and the acquisition of solutes and nutrients. A.

Root Origins

Fossil evidence demonstrates that roots evolved on land and that they were an early innovation of plant life. Roots, as defined here, are unique to vascular plants, but simpler functionally equivalent systems are more widespread. Rhizoid-bearing stems occur in living bryophytes and even some green algae. Evidence from complete plants fossilized in growth position in the Rhynie Chert confirms the presence of similar systems in the earliest vascular plants. In these plants, rhizoids probably developed in response to physical

Origin of Roots

contact of the stem with a substrate. The widespread occurrence of simple rhizoid-based systems in land plants indicates that this form of rooting system preceded the evolution of true roots and may well have been a legacy inherited from green algal ancestors. Shallow, rhizoid-bearing stems are adequate for small (30–50 cm tall) plants with low transpiration and a simple vascular system, but also present in the Rhynie Chert were plants with organs that were clearly and uniquely specialized for rooting. Asteroxylon and its cosmopolitan relative Drepanophycus possessed small roots that were probably positively geotropic. These plants are related to clubmosses, but other basal fossils in this clade are known to be rootless. Viewed in a phylogenetic context (Fig. 2b), the roots of lycopods are uniquely derived and are therefore not homologous with those of other vascular plants (Kenrick and Crane, 1997a).

B.

Root Diversity in the Late Devonian and Carboniferous

By the Late Devonian and Early Carboniferous an enormous variety of rooting structures had evolved and roots had been co-opted for an equally broad range of additional functions. The evolution of large erect plants and in particular trees placed increasing demands upon rooting systems, and these were solved in a variety of ways. From the outset, differentiation of tissues in the roots of arborescent plants followed closely patterns of tissue differentiation in the erect trunk. One of the major sources of divergence in root and stem anatomy resulted from the influence of lateral appendages in the stem. Arborescent horsetails maintained a rhizomatous rooting system that simply increased in girth through the addition of secondary tissues. Progymnosperms developed bipolar growth, where shoot and root were probably determined early in ontogeny as in their living relatives, the seed plants. Unlike horsetails, progymnosperms and seed plants had the ability to produce deeply penetrating roots. Bipolar growth also occurred in the extinct tree lycopods, but phylogenetic and ontogenetic studies clearly demonstrate that this growth form evolved independently of that in seed plants (Frankenberg and Eggert, 1969; Stubblefield and Rothwell, 1981; Rothwell and Erwin, 1985; Bateman et al., 1992). Roots were co-opted as buttresses or mantles in the formation of the false trunks of tree ferns. These provided a substrate for vascular plant epiphytes and root climbers, which are first observed in the fossil record

7

on tree ferns in the Carboniferous Period (Rothwell, 1991; Krings and Kerp, 1997; Ro¨ßler, 2000). C.

Mycorrhizae—An Early Symbiosis

Root-mediated fungal symbioses are characteristic of most living land plants. Recent inferences from phylogenetic studies and direct observations from the fossil record demonstrate that this form of symbiosis is ancient. Most fungal symbioses of roots involve zygomycetes in the Glomales, which form arbuscular mycorrhizae in the rooting systems of bryophytes and most vascular plants (Smith and Read, 1997). The distribution of arbuscular mycorrhizae among living plants is consistent with an ancient origin of this plant/fungal relationship (Pirozynski and Malloch, 1975; chapter by Read in this volume). This hypothesis is borne out by two recent studies. A phylogenetic analysis of living members of the Glomales based on SSU rRNA reported an origin of endomycorrhizal associations in plants between 462 million and 353 million years ago (Simon et al., 1993). Ages were estimated based on the application of a molecular clock. These results were confirmed by direct observations of fossil endomycorrhizae in plants from the Rhynie Chert (Remy et al., 1994). Nonseptate hyphae and arbuscules have been observed in specialized meristematic regions of the cortex of Aglaophyton and Rhynia. In these plants, the fungal infection is widespread throughout the rhizome and the erect stems. The presence of arbuscular mycorrhizae in living bryophytes and in extinct plants such as Aglaophyton and Rhynia demonstrates that this fungal symbiosis preceded the evolution of roots. D.

Homologies of Roots in Clubmosses (Lycopods)—Observational Consequences in Living Plants

The origin of roots provides some remarkable examples of the co-option of organ systems from one role to a completely different one. Perhaps the most spectacular of these is the hypothesized origin of the root system in extinct tree lycopods. It has long been suspected that the unique rooting system of tree lycopods—the so-called rhizomorph—is in fact a shoot modified for rooting (Frankenberg and Eggert, 1969). The modified shoot hypothesis is based on similarities in anatomy, organization of appendages (rootlets/leaves), and embryology. Rhizomorphs, like stems, had a pith and an endarch xylem; rootlets were borne and abscised in a similar manner to leaves (Rothwell and

8

Erwin, 1985); and evidence from fossil embryos shows that the rooting system developed from an early dichotomy of the shoot apex (Stubblefield and Rothwell, 1981). Under this interpretation ‘‘true’’ roots in tree lycopods have been lost and were substituted by a novel shoot system co-opted to a rooting function—developmental changes that may have accompanied the evolution of arborescence during the Devonian Period. The main trunk root is therefore interpreted as a transformed stem and the rootlets are modified leaves. This hypothesis is potentially testable because it has observational consequences in living plants. The closest living relatives of the tree lycopods are Isoetaceae (Bateman et al., 1992; Pigg, 1992; Kenrick and Crane, 1997a) (Fig. 2b). This ancient group comprises
130 species of terrestrial or aquatic plants with elongate linear leaves (sporophylls) borne, in most species, on a very short trunk (Jermy, 1990). Despite the short trunk length—which is interpreted as highly reduced—there are many similarities between Isoetaceae and their arborescent fossil relatives (Stewart, 1947). The interpretation of root origins in Isoetaceae, however, remains controversial. Noting some fundamental differences in embryogeny, Paolillo (1982) concluded that the root of Isoetaceae is homo-

Kenrick

logous with the embryonic roots of Lycopodiaceae and Selaginellaceae. In other words, these are true roots and not homologs of the rhizomorph. Rothwell and Erwin (1985) suggested an alternative hypothesis. Based on a reanalysis of the embryological data and on new information from fossils they concluded that the roots of Isoetaceae are in fact homologous with the rootlets in the rhizomorphic system. This conclusion is supported by the close anatomical similarities between Isoetes roots and those of extinct Stigmaria. If roots in these two plants are homologous, then by implication thoses of Isoetes also originated as rooting leaves. Modern molecular biology could provide a decisive test of the two hypotheses. If the root/leaf homology is correct, then the roots of Isoetaceae should share a greater similarity at the molecular developmental level with their own leaves than with the roots of close relatives in Lycopodiaceae or Selaginellaceae. E.

Roots, Soils, and the Devonian Environment

The Devonian Period witnessed a progressive increase in root biomass (Fig. 4) that had an enormous impact on various aspects of the environment, notably the development of soils and the evolution of the atmo-

Figure 4 Relative size, morphology, and penetration depth of root systems from selected Devonian plants. (From Algeo and Scheckler, 1998.)

Origin of Roots

sphere. Prior to the Devonian, soils, if developed at all, are thought to have been predominantly thin and of microbial origin (Retallack, 1990, 1992). The rhizomatous root systems of the earliest land plants were shallow and essentially superficial. They are unlikely to have contributed much to soil formation or to the weathering of the land surface. By the MidDevonian, the development of positively geotropic roots coupled with a progressive increase in plant size led to more extensive and penetrating rooting systems. These nurtured an environment suitable for the stabilization of shallow soils. A further significant step in the development of soils was the evolution of the cambium, which from the outset was expressed in both roots and stems. Cambial activity led rapidly to significant increase in plant size and the extent and depth of roots. Soil penetration depths were shallow (< 20 cm) during the Eifelian-Givetian (386–377Ma), but increased to 80–100 cm as archaeopterids spread during the Frasnian-Famenian (377–363Ma) (Algeo and Scheckler, 1998; Fig. 4). The diversity of soils also increased during the Devonian Period. This change was brought about by root-induced weathering and mixing. Roots increase rates of weathering by facilitating the pumping of solutes through soils and through the release of humic acids. Translocation of clays and iron was first recorded in wet lowland soils in the early Frasnian (Retallack, 1985). The effectiveness of soil mixing was boosted through anatomical innovations in roots (Algeo and Scheckler, 1998). The evolution of the cambium initiated continuous perennial root growth and long-term survival of roots in soils. The development of endogenous and adventitious roots, particularly in progymnosperms, enabled repeated penetration of a given soil volume. By the end of the Devonian there was an increase in soil clay content, structure, and profile maturity that correlates with increases in the depth of root penetration. Soils with modern profiles are recognizable at this time (Retallack, 1990, 1992). Deeper soils and larger plants go hand in hand, but the evolution of both may have been more intimately connected than previously suspected. One of the most remarkable phenomena of the Late Devonian is the independent but probably concurrent evolution of large trees in several plant groups. Mosbrugger (1990) argued that this phenomenon may in fact be caused by positive feedback between soil and plant, mediated by the roots. In essence, the idea is based on the observations that large plants generally require more stable and comparatively deeper soils, and that plants themselves contribute to soil formation. Under

9

these circumstances, increments in plant size are expected to lead to increments in the quality and depth of the soil profile, which in turn would provide soils suitable for even larger plants. Plant size and soil depth therefore increased in parallel until other limiting factors came into play. The broader relevance of the root/soil relationship to the early evolution of terrestrial and marine ecosystems is widely appreciated (Retallack, 1990, 1992; Algeo and Scheckler, 1998; Berner, 1998). One area of particular importance is the effect of root-mediated weathering on the atmosphere. In the long-term or geochemical carbon cycle, Ca and Mg silicate minerals in soils react with atmospheric CO2 in solution (HCO 3 ) to produce CaCO3 and MgCO3. These carbonates are transferred to the oceans by rivers, where they are precipitated as minerals. Carbonate formation combined with the burial of organic carbon are responsible for most of the draw down of CO2 from the atmosphere. Carbon dioxide is eventually returned to the atmosphere over long time scales by the weathering of organic carbon and by the thermal breakdown of carbonates at depth (Berner, 1998). The evolution of roots is thought to have had a major influence on this process by increasing the rate of weathering of Ca/Mg silicates. This phenomenon is a byproduct of the release of humic acids by roots, the channeling of solutes through the soil profile, and the physical effects of root penetration. Observations on modern root systems indicate that roots can increase weathering by a factor of 7 (Drever and Zobrist, 1992; Cochran and Berner, 1993; Berner, 1994; Cochran and Berner, 1997). By including a root effect factor, models of atmospheric gases through the Phanerozoic predict a massive decline in CO2 concentrations during the Paleozoic from 4–20PAL (pre-Devonian) to 1PAL (mid Carboniferous) (Berner, 1994), and this is broadly consistent with data from other sources (Yapp and Poths, 1992; Mora et al., 1996; McElwain, 1998).

IV.

LEADING TOPICS, GAPS, AND THE FUTURE

The investigation of root origins depends critically on information from the fossil record and on the development of a well-corroborated phylogenetic framework for interpreting the data. Here, I suggest three areas of special relevance to future progress in understanding root origins: 1. Fossils are important because they bridge the huge morphological gaps that separate the major

10

groups of living land plants. In this respect, the early fossil record of vascular plants is of enormous value and it is yielding data relevant to a raft of issues concerning root origins. The study of plants fossilized in their growth position is providing the most significant results, and the exceptional Rhynie Chert biota is a key source of data (e.g., Remy et al., 1994; Remy and Hass, 1996). More high-quality and detailed data on root morphology are needed from vascular plants in the euphyllophyte and lycopsid stem groups. 2. The integration of morphological data from living and fossil plants into phylogenetic research is having a major impact on our understanding of plant morphology. One consequence is that botanists are able to construct more rigorous hypotheses of homology. In discussing the origins of organs, this helps to sharpen the focus of an investigation by narrowing it down to a small number of plausible alternative hypotheses. Progress in our understanding of land plant phylogeny has been spectacular in the past 10 years, but further integration of molecular phylogenies with those that include fossils is necessary to fully exploit the paleobotanical data. 3. Recent advances in the field of molecular developmental biology provide a powerful new set of tools that could be applied to the investigation of root origins. Although the focus of this chapter has been on the fossil record, some of the hypotheses on root origins discussed here have observational consequences for living plants. REFERENCES Algeo TJ, Scheckler SE. 1998. Terrestrial-marine teleconnections in the Devonian: links between the evolution of land plants, weathering processes, and marine anoxic events. Phil Trans R Soc Lond B Biol Sci 353:113–130. Bateman RM. 1991. Palaeobiological and phylogenetic implications of anatomically-preserved Archaeocalamites from the Dinantian of Oxroad Bay and Loch Humphrey Burn, southern Scotland. Palaeontographica B223:1–59. Bateman RM, DiMichele WA, Willard DA. 1992. Experimental cladistic analysis of anatomically preserved lycoposids from the Carboniferous of Euramerica: an essay on paleobotanical phylogenetics. Ann Mo Bot Gdn 79:500–559. Bateman RM, Crane PR, DiMichele WA, Kenrick P, Speck T, Rowe NP, Stein WE. 1998. Early evolution of land plants: phylogeny, physiology, and ecology of the primary terrestrial radiation. Annu Rev Ecol Syst 29:263– 292.

Kenrick Beck CB. 1967. Eddya sullivanensis, gen. et sp. nov., a plant of gymnospermic morphology from the Upper Devonian of New York. Palaeontographica B121:1–21. Beck CB. 1971. On the anatomy and morphology of lateral branch systems of Archaeopteris. Am J Bot 58:758-784. Beck CB, Wight DC. 1988. Progymnosperms. In: Beck CB, ed. Origin and Evolution of Gymnosperms. New York; Columbia University Press, pp 1–84. Berner RA. 1994. GEOCARB II: a revised model of atmospheric CO2 over Phanerozoic time. Am J Sci 294:56– 91. Berner RA. 1998. The carbon cycle and CO2 over Phanerozoic time: the role of land plants. Phil Trans R Soc Lond B Biol Sci 353:75–82. Cochran MF, Berner RA. 1993. Enhancement of silicate weathering rates by vascular land plants: quantifying the effect. Chem Geol 107:213–215. Cochran MF, Berner RA. 1997. Promotion of chemical weathering by higher plants: field observations on Hawaiian basalts. In: Stillings LL, ed. Chemical and Biological Control on Mineral Growth and Dissolution Kinetics. Amsterdam; Elsevier, pp 71–77. Cridland AA. 1964. Amyelon in American coal balls. Palaeontology 7:186–209. Delevoryas T. 1955. The Medullosae—structure and relationships. Palaeontographica B97:114–167. DiMichele WA, Hook RW, Beerbower R, Boy JA, Gastaldo RA, Hotton III N, Phillips TM, Scheckler SE, Shear WA, Sues H-D. 1992. Paleozoic terrestrial ecosystems. In: Behrensmeyer AK, Damuth JD, DiMichele WA, Potts R, Sues H-D, Wing SL, eds. Terrestrial Ecosystems Through Time: Evolutionary Paleoecology of Terrestrial Plants and Animals. Chicago; University of Chicago Press, pp 205–325. Drever JI, Zobrist J. 1992. Chemical weathering of silicate rocks as a function of elevation in the southern Swiss Alps. Geo Cos Act 56:3209–3216. Duff JR, Nickrent DL. 1999. Phylogenetic relationships of land plants using mitochondrial small-subunit rDNA sequences. Am J Bot 86:372–386. Edwards DS. 1986. Aglaophyton major, a non-vascular landplant from the Devonian Rhynie Chert. Bot J Lin Soc 93:173–204. Edwards D, Kenrick P. 1986. A new zosterophyll from the Lower Devonian of Wales. Bot J Lin Soc 92:269–283. Edwards DS, Lyon AG. 1983. Algae from the Rhynie Chert. Bot J Lin Soc 86:37–55. Eggert DA. 1962. The ontogeny of Carboniferous arborescent sphenopsida. Palaeontographica B110:99–127. Ehret DL, Phillips TL. 1977. Psaronius root systems—morphology and development. Palaeontographica B161:147–164. Elick JM, Driese SG, Mora CI. 1998. Very large plant and root traces from the Early to Middle Devonian; implications for early terrestrial ecosystems and atmospheric pCO2. Geology 26:143–146.

Origin of Roots Epstein E. 1977. The role of roots in the chemical economy of life on earth. Bioscience 27:783–787. Fairon-Demaret M, Li C-S. 1993. Lorophyton goense gen. et sp. nov. from the Lower Givetian of Belgium and a discussion of the Middle Devonian Cladoxylopsida. Rev Pal Pal 77:1–22. Frankenberg JM, Eggert DA. 1969. Petrified Stigmaria from North America. I. Stigmaria ficoides, the underground portions of Lepidodendraceae. Palaeontographica B128:1–47. Gerrienne P. 1988. Early Devonian plant remains from Marchin (north of Dinant Synclinorium, Belgium). I. Zosterophyllum deciduum sp. nov. Rev Pal Pal 55:317335. Hao S-G. 1989a. Gumuia zyzzata—a new plant from the Lower Devonian of Yunnan, China. Act Bot Sin 31:954–961. Hao S-G. 1989b. A new zosterophyll from the Lower Devonian (Siegenian) of Yunnan, China. Rev Pal Pal 57:155–171. Hedderson TA, Chapman RL, Cox CJ. 1998. Bryophytes and the origins and diversification of land plants: new evidence from molecules. In: Bates JW, Ashton NW, Duckett JG, eds. Bryology for the Twenty-first Century. Leeds, UK: Maney & Son, pp 65–77. Hueber FM, Banks HP. 1979. Serrulacaulis furcatus gen. et sp. nov., a new zosterophyll from the lower Upper Devonian of New York State. Rev Pal Pal 28:169–189. Jermy AC. 1990. Isoetaceae. In: Kramer KU, Green PS, eds. Pteridophytes and Gymnosperms. Berlin; SpringerVerlag, pp 26–31. Kenrick P. 1994. Alternation of generations in land plants: new phylogenetic and palaeobotanical evidence. Biol Rev Camb Phil Soc 69:293–330. Kenrick P. 2000. The relationships of vascular plants. Phil Trans R Soc Lond B Biol Sci 355:847–855. Kenrick P, Crane PR. 1997a. The Origin and Early Diversification of Land Plants: A Cladistic Study. Washington; Smithsonian Institution Press. Kenrick P, Crane PR. 1997b. The origin and early evolution of plants on land. Nature 389:33–39. Kevan PG, Chaloner WG, Savile DBO. 1975. Interrelationships of early terrestrial arthropods and plants. Palaeontology 18:391–417. Kidston R, Lang WH. 1920a. On Old Red Sandstone plants showing structure, from the Rhynie Chert Bed, Aberdeenshire. Part II. Additional notes on Rhynia gwynne-vaughani, Kidston and Lang; with descriptions of Rhynia major, n. sp., and Hornea lignieri, n. g., n. sp. Trans R Soc Edin 52:603–627. Kidston R, Lang WH. 1920b. On Old Red Sandstone plants showing structure, from the Rhynie Chert Bed, Aberdeenshire. Part III. Asteroxylon mackiei, Kidston and Lang. Trans R Soc Edin 52:643–680. Kidston R, Lang WH. 1921. On Old Red Sandstone plants showing structure, from the Rhynie Chert Bed, Aberdeenshire. Part IV. Restorations of the vascular

11 cryptogams, and discussion on their bearing on the general morphology of the pteridophyta and the origin of the organization of land-plants. Trans R Soc Edin 52:831–854. Kotyk ME, Basinger JF. 2000. The early Devonian (Pragian) zosterophyll Bathurstia denticulata Hueber. Can J Bot 78:193–207. Krings M, Kerp H. 1997. Cuticles of Lescuropteris genuina from the Stephanian (Upper Carboniferous) of central France: evidence for a climbing growth habit. Bot J Lin Soc 123:73–89. Lang WH. 1927. Contributions to the study of the Old Red Sandstone flora of Scotland. VI. On Zosterophyllum myretonianum, Penh., and some other plant-remains from the Carmyllie Beds of the Lower Old Red Sandstone. VII. On a specimen of Pseudosporochnus from the Stromness beds. Trans R Soc Edin 55:443– 456. Lele KM, Walton J. 1961. Contributions to the knowledge of ‘‘Zosterophyllum myretonianum’’ Penhallow from the Lower Old Red Sandstone of Angus. Trans R Soc Edin 64:469–475. Li C-S. 1992. Hsua robusta, an early Devonian plant from Yunnan Province, China and its bearing on some structures of early land plants. Rev Pal Pal 71:121– 147. Lyon AG, Edwards D. 1991. The first zosterophyll from the Lower Devonian Rhynie Chert, Aberdeenshire. Trans R Soc Edin Earth Sci 82:323–332. McElwain JC. 1998. Do fossil plants signal palaeoatmospheric CO2 concentration in the geological past? Phil Trans R Soc Lond B Biol Sci 353:83–96. Meyer-Berthaud B, Stein WE. 1995. A reinvestigation of Stenomyelon from the Late Tournaisian of Scotland. Int J Plant Sci 156:863–895. Meyer-Berthaud B, Scheckler SE, Wendt J. 1999. Archaeopteris is the earliest known modern tree. Nature 398:700–701. Mora CI, Driese SG, Colarusso LA. 1996. Middle to late Paleozoic atmospheric CO2 levels from soil carbonate and organic matter. Science 271:1105–1107. Mosbrugger V. 1990. The tree habit in land plants. Lecture Notes in Earth Science 28:1–161. Paolillo DJ. 1982. Meristems and evolution: developmental correspondance among the rhizomorphs of the lycopsids. Am J Bot 69:1032–1042. Pigg KB. 1992. Evolution of Isoetalean lycopsids. Ann Mo Bot Gdn 79:589–612. Pigg KB, Rothwell GW. 1983. Chaloneria gen. nov.; heterosporous lycophytes from the Pennsylvanian of North America. Bot Gaz 144:132–147. Pirozynski KA, Malloch DW. 1975. The origin of land plants: a matter of mycotropism. Biosystems 6:153– 164. Pryer KM, Smith AR, Skog JE. 1995. Phylogenetic relationships of extant ferns based on evidence from morphology and rbcL sequences. Am Fern J 85:205–282.

12 Rayner RJ. 1984. New finds of Drepanophycus spinaeformis Go¨ppert from the Lower Devonian of Scotland. Trans R Soc Edin: Earth Sci 75:353–363. Remy W, Hass H. 1996. New information on gametophytes and sporophytes of Aglaophyton major and inferences about possible environmental adaptations. Rev Pal Pal 90:175–194. Remy W, Hass H, Schultka S. 1986. Anisophyton potoniei nov. spec. aus den Ku¨hlbacher Schichten (Emsian) vom Steinbruch Ufersmu¨hle, Wiehltalsperre. Argumenta Palaeobotanica 7:123–138. Remy W, Gensel PG, Hass H. 1993. The gametophyte generation of some early Devonian land plants. Int J Plant Sci 154:35–58. Remy W, Taylor TN, Hass H, Kerp H. 1994. Four hundredmillion-year-old vesicular arbuscular mycorrhizae. Proc Natl Acad Sci USA 91:11841–11843. Renzaglia KS, Duff RJ, Nickrent DL, Garbary DJ. 2000. Vegetative and reproductive innovations of early land plants: implications for a unified phylogeny. Phil Trans R Soc Lond B Biol Sci 355:769– 793. Retallack GJ. 1985. Fossil soils as grounds for interpreting the advent of large plants and animals on land. Phil Trans R Soc Lond B Biol Sci 309:105–142. Retallack GJ. 1986. The fossil record of soils. In: Wright VP, ed. Paleosols: Their Recognition and Interpretation. Oxford, UK: Blackwell, pp 1–57. Retallack GJ. 1990. Soils of the Past. London; UnwinHyman. Retallack GJ. 1992. Paleozoic paleosols. In: Martini IP, Chesworth W, eds. Weathering, Soils and Paleosols. Amsterdam; Elsevier, pp 543–564. Rice CM, Ashcroft WA, Batten DJ, Boyce AJ, Caulfield JBD, Fallick AE, Hole MJ, Jones E, Pearson MJ, Rogers G. 1995. A Devonian auriferous hot spring system, Rhynie, Scotland. J Geol Soc Lond 152:229– 250. Rolfe WDI. 1980. Early invertebrate terrestrial faunas. In: Panchen AL, ed. The Terrestrial Environment and the Origin of Land Vertebrates. London; Academic Press, pp 117–157. Ro¨ßler R. 2000. The late Palaeozoic tree fern Psaronius; an ecosystem unto itself. Rev Pal Pal 108:55–74. Rothwell GW. 1975. The Callistophytaceae (Pteridospermopsida): I. Vegetative structures. Palaeontographica B151:171–196. Rothwell GW. 1984. The apex of Stigmaria (Lycopsida), rooting organ of Lepidodendrales. Am J Bot 71:1031–1034. Rothwell GW. 1988. Cordaitales. In: Beck CB, ed. Origin and Evolution of Gymnosperms. New York; Columbia University Press, pp 273–297. Rothwell GW. 1991. Botryopteris forensis (Botryopteridaceae), a trunk epiphyte of the tree fern Psaronius. Am J Bot 78:782–788.

Kenrick Rothwell GW. 1999. Fossils and ferns in the resolution of land plant phylogeny. Bot Rev 65:188–218. Rothwell GW, Erwin DM. 1985. The rhizomorph apex of Paurodendron: implications for homologies among the rooting organs of Lycopsida. Am J Bot 72:86–98. Rothwell GW, Whiteside KL. 1974. Rooting structures of the Carboniferous medullosan pteridosperms. Can J Bot 52:97–102. Rowe NP, Speck T, Galtier J. 1993. Biomechanical analysis of a Palaeozoic gymnosperm stem. Proc R Soc Lond 252:19–28. Scheckler SE. 1995. Progymnosperms have gymnospermous roots. In: Hemsley AR, Kurmann MH, eds. Proceedings Royal Botanic Gardens, Kew London (abstract). Scheckler SE, Banks HP. 1971. Anatomy and relationships of some Devonian progymnosperms from New York. Am J Bot 58:737–751. Schweitzer H-J, Li C-S. 1996. Chamaedendron nov. gen., eine multisporangiate Lycophyte aus dem Frasnium su¨dchinas. Palaeontographica B238:45–69. Simon L, Bousquet J, Le´vesque C, Lalonde M. 1993. Origin and diversification of endomycorrhizal fungi with vascular plants. Nature 363:67–69. Smith SE, Read DJ. 1997. Mycorrhizal Symbiosis. London; Academic Press. Stein WE, Wight DC, Beck CB. 1984. Possible alternatives for the origin of Sphenopsida. Syst Bot 9:102–118. Stevenson DW, Loconte H. 1996. Ordinal and familial relationships of pteridophyte genera. In: Camus JM, Gibby M, Johns RJ, eds. Pteridology in Perspective. Kew, UK: Royal Botanic Gardens, pp 435–467. Stewart WN. 1947. A comparative study of stigmarian appendages and Isoetes roots. Am J Bot 34:315–324. Stewart WN, Rothwell GW. 1993. Paleobotany and the evolution of plants. Cambridge, UK: Cambridge University Press. Stidd BM. 1981. The current status of the medullosan seed ferns. Rev Pal Pal 32:63–101. Stubblefield SP, Rothwell GW. 1981. Embryogeny and reproductive biology of Bothrodendrostrobus mundus (Lycopsida). Am J Bot 68:625–634. Taylor TN, Taylor EL. 1993. The Biology and Evolution of Fossil Plants. Englewood Cliffs, NJ: Prentice Hall. Taylor TN, Hass H, Remy W. 1992a. Devonian fungi: interactions with the green alga Palaeonitella. Mycologia 84:901–910. Taylor TN, Remy W, Hass H. 1992b. Parasitism in a 400million-year-old green alga. Nature 357:493–494. Taylor TN, Hass H, Remy W, Kerp H. 1995. The oldest fossil lichen. Nature 378:244. Taylor TN, Hass H, Kerp H. 1997. A cyanolichen from the Lower Devonian Rhynie Chert. Am J Bot 84:992– 1004. Taylor TN, Hass H, Kerp H. 1999. The oldest fossil ascomycetes. Nature 399:648.

Origin of Roots Trewin NH. 1994. Depositional environment and preservation of biota in the Lower Devonain hot-springs of Rhynie, Aberdeenshire, Scotland. Trans R Soc Edin Earth Sci 84:433-442.

13 Wang Z, Geng B-Y. 1997. A new Middle Devonian plant: Metacladophyton tetraxylum gen. et sp. nov. Palaeontographica B243:85–102. Yapp CJ, Poths H. 1992. Ancient atmospheric CO2 pressures inferred from natural goethites. Nature 355:342–344.

2 Characteristics and Functions of Root Systems Alastair Fitter University of York, York, England

I.

FUNCTIONS OF ROOT SYSTEMS

Malloch, 1975; Lewis, 1987). Modern plants with similar underground parts, such as those with ‘‘magnolioid’’ roots, i.e., thick, little-branched root systems typified by the primitive family Magnoliaceae (Baylis, 1975), or achlorophyllous orchids (e.g., Neottia, Epipogium), are habitually or even obligately mycorrhizal. It is almost certain that Aglaophyton also was mycorrhizal, since arbuscules, the diagnostic structures of the most abundant group of mycorrhizal fungi, have been identified in its fossils (Remy et al., 1994). Mycorrhizal fungal spores have been found in Silurian deposits that certainly predate the evolution of root systems (Redecker et al., 2000). The origins of the diversity of the root systems of modern plants can be seen as achieving the more effective performance of these primary functions. In this chapter, I review the range of variation in the gross morphology and architecture of root systems and suggest that these can be related to the evolutionary optimization of at least the primary functions. I do not deal with metabolism or with the storage and propagation functions, important though these may be in particular cases (e.g., tubers, root buds, and suckers). The acquisition of resources is, in contrast, both well understood at a physiological level and clearly related to the overall behavior of root systems, and this will be the underlying concept throughout this chapter. It is important to bear in mind the distinction between roots (i.e., individual members of a root system) and their integration as a root system.

The root systems of terrestrial plants perform two primary functions: the acquisition of soil-based resources (principally water and ions), and anchorage. Other root system functions, such as storage, synthesis of growth regulators, propagation, and dispersal, can be seen as secondary. Little is known of the early evolution of the nonaerial parts of plants, although root traces have been found in fossil soils from Silurian deposits (Retallack, 1997), but it is certain that most of the first land plants, such as Cooksonia, Aglaophyton, and Rhynia, dating from the late Silurian and early Devonian periods, had poorly developed root systems (Collinson and Scott, 1987; see also Chapter 1 by Kenrick in this volume). Such fragments as preserved are of large diameter, and sparingly and often dichotomously branched. Early land plants were small and lived in very wet environments; evidence for plants with developed root systems growing in welldrained soils does not appear until the late Devonian (Driese et al., 1997). Neither anchorage nor acquisition of water is therefore likely to have been a serious problem: the tree habit, necessitating deep rooting for anchorage, did not develop until the mid-Devonian, when the evolution of the seed further freed plants from dependence on wet environments (Algeo and Scheckler, 1998; Bateman et al., 1998). It is likely that the most difficult function for early root systems to perform was the acquisition of poorly mobile resources, especially phosphate (Pirozynski and 15

16

II.

Fitter

CHARACTERISTICS OF ROOTS

Whereas plant species can be identified readily from flowers and, within local floras, usually without great difficulty from leaves, roots show few distinctive external features that would permit identification. Variation in leaf form can be described as adaptive to particular environmental influences such as radiation and herbivory. Leaf shape and size influence the transfer of heat and water vapor through the boundary layer (Grace, 1977), but there is no parallel phenomenon below ground. Herbivorous insects can be deterred from attacking leaves by spines and other modifications to the leaf margin (Merz, 1959; Gilbert, 1971), but there seem to be no similar structures on roots despite the fact that root-feeding insects are important and widespread in many communities (Brown and Gange, 1991; see also Chapter 51 by Koltai et al. and Chapter 52 by Katan in this volume). The scarce variation in external features of roots is presumably related to the limited range of variation of the root environment. Such variations as occur may be interpretable in terms of particular root functions. The main features of individual roots that show systematic variation are diameter, color, growth potential, surface texture, and various physiological traits such as transport capability, hormone content, and membrane composition (see Chapter 9 by Waisel and Eshel in this volume).

A.

Root Diameter

The diameter of individual roots varies widely both within and between species. In many taxonomic groups of plants, especially grasses, rushes, and sedges (Poaceae, Juncaceae, and Cyperaceae), root systems have very fine terminal branches (< 100
m diameter; Table 1). These species seem to have roots approaching an effective minimum diameter, determined by the need for a central stele and surrounding tissues (endodermis, cortex, epidermis), to provide transport to and from the root tip and the absorbing cells (Lyford, 1975; Fitter, 1987). At the other extreme, families such as Alliaceae and Magnoliaceae, although remote from each other both taxonomically and ecologically, have typically very coarse terminal roots, often
0:5–1.0 mm in diameter, and many trees approach such figures. Root diameter determines the length of root that the plant can produce for unit input of resources to the system. Many trees that normally grow in association with ectomycorrhizal fungi (especially Pinaceae) have thick roots and very low root length densities, that is, low total lengths of roots per unit soil volume. The fine-rooted species tend to have a lesser tendency to be mycorrhizal (Peat and Fitter, 1994; Fig. 1). Equally, many plants produce finer roots when grown at low nutrient supply rates. This can be demonstrated either by measuring root diameter directly (e.g., Christie and Moorby, 1975) or by determining specific

Table 1 Radius of the Finest Elements of the Root Systems of Selected Plant Species Radius (
m) > 500 250–500 150–250

100–150

35–100

Species

Family

Podocarpus totara Glycyrrhiza lepidota Lygodesmia juncea Sporobolus longifolius Histiopteris incisa Andropogon nutans Smilacina stellata Solanum nigrum Yucca glauca Triticum aestivum Andropogon scoparius Elymus canadensis Holcus lanatus Zea mays Carex coriacea Quercus rubra Senecio aureus

Podocarpaceae Fabaceae Asteraceae Poaceae Dennstaedtiaceae Poaceae Liliaceae Solanaceae Agavaceae Poaceae Poaceae Poaceae Poaceae Poaceae Cyperaceae Fagaceae Asteraceae

Source Baylis (1975) Weaver (1919) Weaver (1919) Weaver (1919) Baylis (1975) Weaver (1919) Weaver (1919) Baylis (1975) Weaver (1919) Hackett (1972) Weaver (1919) Weaver (1919) McGonigle (1987) Miller (1981) Baylis (1975) Lyford (1980) Weaver (1919)

Characteristics and Functions

Figure 1 Numbers of plant species in the British flora that are, respectively, never or rarely, occasionally, and normally mycorrhizal in each of four root diameter classes. (Data from Peat and Fitter, 1993.)

root length (length of root per unit root weight; Fitter, 1985), but it is not a universal response, particularly when parts of a single root system exposed to a heterogeneous nutrient supply are examined (Fransen et al., 1999; Hodge et al., 1998). The tendency to produce fine roots in such conditions can largely be explained in terms of ion mobility and the volume of soil exploited (Baldwin, 1975; Robinson and Rorison, 1988). Fine roots of some species (specifically roots with a thin cortex) also tend to have a relatively greater hydraulic conductivity, increasing transport efficiency (Rieger and Litvin, 1999). The countervailing benefits of coarse roots to plants in nutrient-rich soils or to mycorrhizal plants with an adequate supply of immobile ions are uncertain. Thicker roots may be able to exert greater forces on soil and might possibly have greater ability to penetrate compacted soil (Goss, 1977); it is conceivable, too, that they are more resistant to herbivores. Thick roots are more likely to persist, branch, and contribute to the long-term development of the root system. Fine roots turn over more rapidly than coarse roots (Eissenstat and Yanai, 1997) and also have higher nutrient concentrations (Gordon and Jackson, 2000). Rapid turnover of resource-rich roots may impose a large burden on the plant. B.

Texture

The surfaces of roots vary in terms of the number, size, and duration of root hairs; the persistence of the epidermis and cortex; and the nature of the bark that eventually covers woody roots. The last feature was

17

covered briefly by Cutler et al. (1987), who recorded, however, much greater variation in internal anatomy, leading one to wonder whether an equally detailed study of external features might be profitable. Root hairs are considered in depth by Ridge and Katsumi (Chapter 5, volume). The epidermis is often an ephemeral structure in field-grown roots covering < 30% of the root surface of Citrus roots (Eissenstat and Achor, 1999). Roots in soils are associated with an extensive rhizosphere microflora, and although this offers no visible features, rhizosphere soil frequently adheres to roots removed from soil (McCully, 1995), partly because of the production of mucigel by this association. The extreme case of rhizosphere development is the ectomycorrhizal sheath, an important visible feature of the terminal roots of many (but by no means all) woody plants. Ectomycorrhizae also alter root system characters, such as branching patterns, as described in the chapters by Kottke (50) and Sieber (49) in this volume.

C.

Color

Roots vary surprisingly in color. Old roots are usually colored one of various shades of brown, but young roots can be unpigmented (‘‘white’’) or variously tinged with pink or orange, possibly because pigmented zones contain feeding deterrents. Roots colonized by arbuscular mycorrhizal fungi are often yellow; the physiological significance of this deposition of pigments is unclear (Klinger et al., 1995). Roots of some plants (Lemna, Datura, Triticum) are capable of forming chlorophyll when exposed to light; aerial roots of epiphytic orchids may also be green. Although it is well established that roots can grow autotrophically when transformed in hairy root culture (Flores et al., 1993), the assimilatory potential of green roots in intact plants is unknown.

D.

Growth Potential and Longevity

A root system can be thought of as a population of meristems which vary greatly in the duration of their activity. Some survive for only a few days or weeks and give rise to very short roots; others may continue growing for months or years, and generally produce much longer roots. Oak seedlings (Quercus robur) produce some laterals that grow rapidly at first but soon cease growth, and others that grow slowly initially but continue for a longer time (Page´s, 1995); the difference is

18

apparently related to the diameter of the apex, with the short-lived roots having thinner apices. The growth potential of the meristem and the longevity of the root it gives rise to seem therefore to be related. Some roots have very short lives: Garwood (1967) estimated half-lives of grass roots to be as low as 10 days, and the use of video cameras and rhizotrons has ensured that longevity patterns are increasingly understood. In sugar maple (Acer saccharum) forests in Michigan, Hendrick and Pregitzer (1993) measured half-lives for summer cohorts of roots of 107 cm2 s1 , herringbone topologies have the highest exploitation efficiency (Fig. 10). This result arises because such a topology produces fewer high-order laterals (secondary and tertiary); these typically have lower growth rates and fail to extend out of the depletion zone of the parent root. For less mobile resources (e.g., H2 PO4 , whose diffusion coefficient in soil is usually < 108 cm2 s1 ), topology is unimportant, because depletion zones are so narrow. Such ions may be largely acquired by mycorrhizal hyphae in many plants (see Chapter by 35 in this volume). This analysis therefore suggests that whereas root systems with dichotomous topology are both cheaper to construct and maintain and more efficient at transporting immobile materials, they are less efficient in the acquisition of mobile resources. This would lead one to

Figure 10 The exploitation efficiency of a root system (defined as the volume of soil exploited per unit root tissue volume) increases as the topological index increases, in a computer simulation. High values ( 1) of the topological index represent little-branched (herringbone) root systems; low values represent dichotomous branching patterns. In this simulation, exploitation was a resource with a diffusion coefficient in soil of 107 cm2 s1 , similar to that of nitrate ions. The simulation was performed at a range of values of other architectural characteristics, namely, branching density and the relationship between radius and magnitude, which explains the scatter of the points. (From Fitter et al., 1991.)

Characteristics and Functions

predict that ‘‘expensive’’ herringbone systems would be favored by slow-growing perennials found in resourcepoor soils where acquisition of such resources is likely to limit growth and where allocation of resources to root growth will be favored. Conversely, annuals and other plants for which soil-based resources are less likely to limit growth should produce more nearly dichotomous systems. This prediction is borne out by data available for the limited number of species that have already been studied (Fitter et al., 1988; Fitter and Stickland, 1991; Taub and Goldberg, 1996; Fig. 11). D.

Anchorage

Many features of root systems contribute to their role in providing anchorage for the plant. The part played by stilt, prop, and buttress roots has already been mentioned; only the role of subterranean roots is considered here. Two situations (or a combination) are likely to be important in natural situations: (1) where a simple upward force is exerted on the plant (e.g., by a grazing animal), and (2) where a horizontal force is exerted on the stem (usually by wind) (see Fig. 12). In practice, the first of these situations is most probable for small, mainly nonwoody plants, and the

Figure 11 The topological index of root systems produced by slow-growing species is greater than that of faster-growing species, in accordance with the predictions based on the relationship in Fig. 10. Topological index is calculated from the slope of log altitude on log magnitude (as in Fig. 10) and ranges from approaching zero (in very large, dichotomously branched systems) to 1 (in a herringbone system). (From Fitter and Strickland, 1991.)

27

upward force may be at almost any angle and may include twisting movements. There is certainly variation between species and stages of development in the ease with which they can be uprooted, in the relative strengths of stem base and root system. Stellaria media, for example, shears at the stem base if pulled, leaving the root system and stem base with buds intact. Other species may be uprooted more easily, and in some cases (e.g., Holcus lanatus), this appears to be because they are shallow rooted (Fitter, 1986). It is also likely that the quantity of roots may influence strength, but there seems to be no information on this for herbaceous plants. The nature of the bond between individual roots and soil (determined by root surface area, root hair density, extent of mucigel production, and nature of the rhizosphere among other things) has received little attention (see Chapter 5 by Ridge and Kutsumi in this volume). Nevertheless, Bailey (1998) showed that the root-hair-deficient mutant (rhd1) of Arabidopsis thaliana had the same anchorage characteristics as the wildtype, suggesting that root hairs play a limited role in anchorage. In addition to root distribution and quantity, it is certain that the branching pattern of root systems must influence the strength of anchorage (Coutts, 1983; see Chapter 10 by Stokes in this volume). Where roots are stressed under tension, an analogy can be made with a problem in material science involving the pulling of a reinforcing fibre out of a matrix (Kelly and Macmillan, 1986). Again, it is necessary to know something of the mechanical characteristics of both roots and soil. This has been studied in relation to the effects of roots on the shear resistance of soils (Waldron, 1977), and tension imposed on roots will be transferred by shear to the soil matrix. One important feature is that the

Figure 12 Diagrammatic representation of forces involved in resistance of a tree to windthrow. Windward roots are in tension, those at the hinge in compression. (From Stokes et al., 1995b.)

28

mechanical properties of roots (e.g., stiffness and strength) alter considerably as they develop secondary thickening, whereas those of the soil matrix are less variable at any one site. Woody and herbaceous plants may therefore behave very differently. Ennos (1989, 1990) and Fitter and Ennos (1989) have shown that tension is progressively transferred to soil as one moves distally along a root. In consequence, it is the proximal parts of main roots that are responsible for anchorage strength. If a force is applied that is too great to be dissipated in this way, the root will break long before the tension can be transmitted to the distal parts. Those finely branched regions therefore play little role in anchorage. This is an important result, since it means that the form of fine root systems can be interpreted in terms of resource acquisition ignoring anchorage. E. Plasticity of Root Systems Much of the discussion of root system form and function in this chapter has centered on the concept of optimization. Since the environment of root systems is highly heterogeneous both in time and space, it would appear essential that root systems would have the ability to react to that heterogeneity; in other words, they should possess phenotypic plasticity. In practice, the degree of plasticity is enormous and is largely responsible for the failure of attempts to classify root systems. All the important components of root system architecture are susceptible to environmental modification (see also Chapter 22 by Page´s in this volume). Nevertheless, there is some evidence that metric aspects of geometry (e.g., lengths and radii of links) are more plastic than topology. Root radius is certainly highly variable (Goss, 1977; Christie and Moorby, 1975), but link lengths are much more responsive to variations in nutrient supply than is topology (Fitter et al., 1988), although there is large interspecific variation in topology. One of the best-characterized plastic responses is that to local patches of high nutrient supply in soil. Almost 140 years ago, Nobbe (1862) drew attention to this phenomenon, which has been repeatedly investigated in natural soils (Sprague, 1933), in fertilizer bands in agricultural soils (Duncan and Ohirogge, 1958), and in water culture (Drew et al., 1973). As so often, however, most of this work has been performed on a limited set of species, mostly crop plants, which share certain ecological characteristics. These show rapid growth, a strong response to increased resource supply, and a low root/shoot ratio. Such plants

Fitter

evolved initially as opportunists, exploiting briefly available pulses and patches of high resource availability and were subsequently selected in agriculture for even greater responsiveness. One would expect them, therefore, to show a high degree of phenotypic and physiological plasticity. Many other species, however, normally experience lower and less variable levels of resources; it is likely that these would not have evolved such a high degree of plasticity. That plasticity of root development is itself an adaptive variable is suggested by an ingenious set of experiments by Campbell et al. (1991). By dripping nutrient solutions onto four spots on the surface of a bowl filled with sand, they were able to maintain four distinct quadrants, alternating high and low nutrient concentrations, but without any barriers to root growth. A single plant of each of a range of species was grown in each bowl and root growth in each quadrant measured. All species had greatest root growth in the nutrient-rich quadrants. However, in some species the difference between rich and poor quadrants was slight, whereas other species concentrated 80–90% of all root growth in the nutrient-rich quadrants. The species that showed greatest precision at placing roots in rich patches were also the smallest and least competitive (Fig. 13). Large, competitive species such as Urtica dioica were less selective, although because of their much faster growth rate, they achieved much higher root length densities in both rich and poor patches than the more precise species. Campbell et al. (1991) emphasized the contrast between scale and precision highlighted by these results, and that contrast needs to be viewed in relation to the pattern of heterogeneity that roots encounter in soil. In contrast, Einsmann et al. (1999) found no correlation between scale and precision of response. However, their set of species ranged from annuals to trees, making interpretation difficult. Patchiness may occur in both space and time, and patches may differ in the distribution (pattern in space, frequency in time), size of duration, and intensity. Farley and Fitter (1999) showed that each of seven coexisting herbaceous perennial had a unique set of responses, including changes in SRL, in topology (branching pattern), in mycorrhizal colonization, and in responsiveness to patch size and concentration. These studies demonstrate that plant species display a range of responses to heterogeneity. An apparent paradox is that several studies (Van Vuuren et al., 1996; Fransen et al., 1998; Hodge et al., 1998) using 15N as a label have shown that proliferation does not lead to increased N capture from N-

Characteristics and Functions

29

Figure 13 The precision with which roots are placed in patches of high nutrient concentration in the experiment of Campbell et al., (1991) is inversely related to plant size, as estimated from relative growth rate (R) and seed mass (Wsd) in a compound variable created by multiple regression. Large plants such as Urtica dioica grow roots at more or less random spatially. (From Fitter, A. H. [1994]. In Exploitation of Environmental Heterogeneity by Plants [M. M. Caldwell and R. W. Pearcy, Eds.]. Academic Press, New York, pp. 305–323.)

rich patches even though nitrate is a powerful trigger for localized proliferation (Zhang and Forde, 1998). This result is due to the mobility of nitrate in soil, which means that a low root length density can achieve a high N capture rate. The explanation for this paradox was provided by an experiment by Hodge et al. (1999) in which ryegrass (Lolium perenne) and smooth meadow grass (Poa pratensis) competed for N from an organic patch. In those circumstances the species with the highest root length density captured most N. Hence, proliferation must be seen as a response that promotes competitive ability, a result in striking contrast to the view of Campbell et al. (1991). Plasticity, as exemplified here by the proliferation response, is a fundamental characteristic of the growth of root systems. Their modular construction means that the fates of individual meristems in space and time can each be unique, giving rise to a large network of possible structures. Plasticity is perhaps the most important adaptive feature of roots to their environment, ensuring that they achieve their primary functions of resource acquisition and anchorage.

REFERENCES Abrahamson WG, Caswell H. 1982. On the comparative allocation of biomass, energy and nutrients inplants. Ecology 63:982–991. Algeo TJ, Scheckler SE. 1998. Terrestrial-marine teleconnections in the Devonian: links between the evolution of land plants, weathering processes, and marine anoxic events. Phil Trans R Soc Lond. Series B-Biol Sci 353:113–128. Amthor JS. 1984. The role of maintenance respiration in plant growth. Plant Cell Environ 7:561–569. Baldwin JP. 1975. A quantitative analysis of the factors affecting plant nutrient uptake from some soils. J Soil Sci 26:195–206. Bailey PBH. 1998. Vertical pulling resistance in herbaceous plants. D.Phil. Thesis, University of York. Barley K. 1970. The configuration of the root system in relation to nutrient uptake. Adv Agron 22:159–201. Bateman RM, Crane PR, DiMichele WA, Kenrick PR, Rowe NP, Speck T, Stein WE. 1998. Early evolution of land plants: phylogeny, physiology, and ecology of the primary terrestrial radiation. Annu Rev Ecol Syst 29:263–292.

30 Baylis GTS. 1975. The magnolioid mycorrhiza and mycotrophy in root systems derived from it. In: Sanders FE, Mosse B, Tinker PB, eds. Endomycorrhizas. London; Academic Press, pp 373–390. Berntson GM. 1995. The characterization of topology: a comparison of four topological indices for rooted binary trees. J Theor Biol 177:271–281. Berntson GM, Bazzaz FA. 1996. The allometry of root production and loss in seedlings of Acer rubrum (Aceraceae) and Betula papyrifera (Betulaceae): implications for root dynamics in elevated CO2. Am J Bot 83:608–616 Bray RH. 1954. A nutrient mobility concept of soil-plant relationships. Soil Sci 78:9–22. Brown VK, Gange AL. 1991. Effects of root herbivory on vegetation dynamics. In: Atkinson D, ed. Plant Root Growth. Oxford, UK: Blackwell Scientific: pp 453– 470. Campbell BD, Grime JP, Mackey JM. 1991. A trade-off between scale and precision in resource foraging. Oecologia 87:532–538. Cannon WA. 1911. Root Habits of Desert Plants. Publication No. 131. Washington: Carnegie Institution. Cannon WA. 1949. A tentative classification of root systems. Ecology 30:452–458. Charlton WA. 1975. Distribution of lateral roots and pattern of lateral initiation in Pontaderia cordata L. Bot Gaz 136:225–235. Charlton WA. 1983. Patterns of distribution of lateral root primordia. Ann Bot 51:417–427. Christie EK, Moorby J. 1975. Physiological responses of arid grasses: 1. The influence of phosphorus supply on growth and phosphorus absorption. Aust J Agric Res 26:423–436. Collinson ME, Scott AC. 1987. Factors controlling the organisation and evolution of ancient plant communities. In: Gee JHR, Oiler PS, eds. Organisation of Communities. Oxford, UK: Blackwell Scientific, pp 399–420. Coutts MP. 1983. Root architecture and tree stability. Plant Soil 71:171–188. Coutts MP. 1986. Components of tree stability in Sitka spruce or peaty gley soil. Forestry 59:173–197. Cutler DF, Rudall PJ, Gasson P. 1987. Root Identification Manual of Trees and Shrubs. London: Chapman & Hall. Drew MC, Saker LR, Ashley TW. 1973. Nutrient supply and the growth of the seminal root system in barley: 1. The effect of nitrate concentration on the growth of axes and laterals. J Exp Bot 24:1189–1202. Driese SG, Mora CI, Elick JM. 1997. Morphology and taxonomy of root and stump casts of the earliest trees (Middle to Late Devonian), Pennsylvania and New York, USA. Palaios. 12:524–537.

Fitter Duncan WG, Ohlrogge A. 1958. Principles of nutrient uptake from fertiliser bands: II. Root development in the band. Agron J 50:605–608. Einsmann JC, Jones RH, Pu M, Mitchell RJ. 1999. Nutrient foraging traits in 10 co-occurring plant species of contrasting life forms. Ecology 87:609–619. Eissenstat DM, Achor DS. 1999. Anatomical characteristics of roots of citrus rootstocks that vary in specific root length. New Phytol 141:309–321. Eissenstat DM, Yanai R. 1997. The ecology of root life-span. Adv Ecol Res 27:1–60. Ennos AR. 1989. The mechanics of anchorage in seedlings of sunflower, Helianthus annuus L. New Phytol 113:185– 192. Ennos AR. 1990. The anchorage of leek seedlings: the effect of root length and soil strength. Ann Bot 65:409–416. Ennos AR. 1993. The function and formation of buttresses. Trends Ecol Evol 8:350–351. Eshel A. 1998. On the fractal dimensions of a root system. Plant Cell Environ 21:247–251. Eshel A, Schick I, Waisel Y. 2000. The efficiency of water conducting system of tomato roots. In: Stokes A, ed. The Supporting Roots of Trees and Woody Plants: Form, Function and Physiology. Dordrecht, Netherlands: Kluwer, pp 371–375. Farley RA, Fitter AH. 1999. The responses of seven cooccurring woodland herbaceous perennials to localized nutrient-rich patches. Ecology 87:849–859. Fitter AH. 1985. Functional significance of root morphology and root system architecture. In: Fitter AH, Atkinson D, Read DJ, Usher MB, eds. Ecological Interactions in Soil. Oxford, UK: Blackwell Scientific, pp 87–106. Fitter AH. 1986. Spatial and temporal patterns of root activity in a species rich alluvial grassland. Oecologia 69:594–599. Fitter AH. 1987. An architectural approach to the comparative ecology of plant root systems. New Phytol 106(suppl):61–77. Fitter AH. 1993. Architectural analysis of plant root systems. In: Hendry, GAF, Grime JP, eds. Methods in Comparative Plant Ecology: A Laboratory Manual. London; Chapman & Hall, pp 165–170. Fitter AH. 1994. Architecture and biomass allocation as components of the plastic response of root systems to soil heterogeneity. In: Caldwell, MM, Pearcy RM, eds. Exploitation of Environmental Heterogeneity by Plants. New York; Academic Press, pp 305–323. Fitter AH. 1999. Roots as dynamic systems: the developmental ecology of roots and root systems. In: Press MC, Scholes JD, Barker MG, eds. Plant Physiological Ecology. London; Blackwell Scientific, pp 115–131. Fitter AH, Ennos RA. 1989. Architectural constraints to root system function. In: Robinson D, ed. Roots and the Soil Environment. Aspects Appl Biol 2:15–22.

Characteristics and Functions Fitter AH, Setters NL. 1988. Vegetative and reproductive allocation of phosphorus and potassium in relation to biomass in six species of Viola. J Ecol 76:617–636. Fitter AH, Stickland TR. 1991. Architectural analysis of plant root systems II. Influence of nutrient supply on architecture in contrasting plant species. New Phytol 118:383–389. Fitter AH, Stickland TR. 1992. Fractal characterization of root-system architecture. Funct Ecol 6:632–635. Fitter AH, Nichols R, Harvey ML. 1988. Root system architecture in relation to life history and nutrient supply. Funct Ecol 2:345–351. Fitter AH, Stickland TR, Harvey ML, Wilson GW. 1991a Architectural analysis of plant root systems. I. Architectural correlates of exploitation efficiency. New Phytol 118:375–382. Fitter AH, Stickland TR, Harvey ML, Wilson GW. 1991b. Architectural analysis of plant root systems. II. Influence of nutrient supply on architecture in contrasting plant sciences. New Phytol 118:383–389. Fitter AH, Self GK, Wolfenden J, Van Vuuren M, Brown TK, Williamson L, Graves JD, Robinson D. 1997. Root production and mortality under elevated atmospheric carbon dioxide. Plant Soil 187:299–306. Fitter AH, Graves JD, Self GK, Brown TK, Bogie D, Taylor K. 1998. Root production, turnover and respiration under two grassland types along an altitudinal gradient: influence of temperature and solar radiation. Oecologia 114:20–30. Fitter AH, Self GK, Brown TK, Bogie D, Graves JD, Benham D, Ineson P. 1999. Root production and turnover in an upland grassland subjected to artificial soil warming respond to radiation flux and nutrients, not temperature. Oecologia 120:575–581. Flores HE, Dai YR, Cuello JL, Maldona-Mendoza IE, Loyola-Vargas VM. (1993). Green roots—photosynthesis and photoautotrophy in an underground plant organ. Plant Physiol 101:363–371. Fransen B, de Kroon H, Berendse F. 1998. Root morphological plasticity and nutrient acquisition of perennial grass species from habitats of different nutrient availability. Oecologia 115:351–358. Fransen,B, Blijenberg J, de Kroon H. 1999. Root morphological and physiological plasticity of perennial grass species and the exploitation of spatial and temporal heterogeneous nutrient patches. Plant Soil 211:179– 189. Garwood EA. 1967. Seasonal variation in appearance and growth of grass roots. J Br Grassland Soc 22:121–130. Gibbs RJ, Reid JB 1992. Comparison between net and gross root production by winter wheat and perennial ryegrass. NZ J Crop Hort Sci 20:483–487. Gilbert LE. 1971. Butterfly-plant coevolution: has Passiflora adenopoda won the selectional race with heliconid butterflies? Science 172:585–586.

31 Gordon WS, Jackson RB. 2000. Nutrient concentrations in fine roots. Ecology 81:275–280. Goss MJ. 1977. Effect of mechanical impedance on growth of seedlings. J Exp Bot 28:96–111. Grace J. 1977. Plant Response to Wind. London; Academic Press. Hackett C. 1972. A method of applying nutrients locally under controlled conditions and some morphological effects of locally applied nitrate on the branching of wheat roots. Aust J Biol Sci 25:1169–1180. Head GC. 1973. Shedding of roots. In: Kozlowski TT, ed. Shedding of Plant Parts. New York; Academic Press, pp 237–293. Headley AD, Callaghan TV, Lee JA. 1988. Water uptake and movements in the evergreen clonal plants Lycopodium annotinum L. and Diphasiastrum complantum (L.) Holub. New Phytol 110:487–495. Hendrick RL, Pregitzer KS. 1992. Patterns of the root mortality in two sugar maple forests. Nature 361:59–61. Hodge A, Stewart J, Robinson D, Griffiths BS, Fitter AH. 1998. Root proliferation, soil fauna and plant nitrogen capture from nutrient-rich patches in soil. New Phytol 139:479–494. Hooker JE, Atkinson D. 1996. Arbuscular mycorrhizal fungi-induced alteration to tree-root architecture and longevity. Z Pflanzenernhr Bodenk 159:229–234. Hunt ER, Zakir NJD, Nobel PS. 1987. Water costs and water revenues for established and rain-induced roots of Agave deserti. Funct Ecol 1:125–130. Kelly A, Macmillan NH. 1986. Strong Solids. Oxford, UK: Oxford University Press. Klinger A, Bothe H, Wray V, Marschner H. 1995. Identification of a yellow pigment found in maize roots upon mycorrhizal colonisation. Phytochemistry 38:53–55. Knuth DE. 1968. The Art of Computer Programming. Vol. 1. Reading MA; Addison-Wesley. Kosola KR, Eissenstat DM, Graham JH. 1995. Root demography of mature citrus trees: the influence of Phytophthora nicotianae. Plant Soil 171:283–288. Ko¨stler JN, Bruckner E, Bibelriether H. 1968. Die Wurzeln der Waldba¨ume. Hamburg, Germany: Parey. Kozlowski TT. 1971. Growth and Development of Trees. London; Academic Press. Lambers H, Szaniawski RK, de Visser R. 1983. Respiration of the growth, maintenance and ion uptake: an evaluation of concepts, methods, values and their significance. Physiol Plan 58:556–563. Lewis DH. 1987. Evolutionary aspects of mutualistic associations between fungi and photosynthetic organisms. In: Rayner ADM, Brasier CM, Moore D, eds. Evolutionary Biology of the Fungi. Cambridge, UK: Cambridge University Press, pp 161–178. Lyford WH. 1975. Rhizography of non-woody roots of trees in the forest floor. In: Torrey JG, Clarkson DT, eds.

32 The Development and Function of Roots. New York; Academic Press, pp 179–196. Lyford WH. 1980. Development of the Root System of Northern Red Oak (Quercus rubra L.). Harvard Forest Paper No. 21. Cambridge, MA: Harvard University Press. Masi CEA, Maranville JW. 1998. Evaluation of sorghum root branching using fractals. J Agric Sci 131:259–265. McCully ME. 1995. How do real roots work? Plant Physiol 109:1–6. McGonigle TP. 1987. Vesicular arbuscular mycorrhizas and plant performance in a semi-natural grassland. Ph.D. dissertation, University of York, York, UK. Merryweather JW, Fitter AH. 1995. Arbuscular mycorrhiza and phosphorus as controlling factors in the life history of Hyacinthoides non-scripta (L.) Chouard ex Rothm. New Phytol 129:629–636. Merz E. 1959. Pflanzen und Raupen: u¨ber einigen Prinzipien der Futterwahl bei Grossschmetterlingsraupen. Biol Zentralbl 78:152–158. Nielsen KL, Lynch JP, Weiss HN. 1997. Fractal geometry of bean root systems: correlations between spatial and fractal dimension. Am J Bot 84:26-33. Nobbe F. 1862. Uber die feinere Verastelung der Pflanzenwurzeln. Landwirtsch Versuchs 4:212–224. Nye PH, Tinker PB. 1977. Solute Movement in the Soil Root System. Oxford, UK: Blackwell Scientific. Ozier-Lafontaine H, Lecompte F, Sillon JF. 1999. Fractal analysis of the root architecture of Gliricidia sepium for the spatial prediction of root branching, size and mass: model development and evaluation in agroforestry. Plant Soil 209:167–180. Page´s L. 1995. Growth patterns of the lateral roots of young oak (Quercus robur L) tree seedlings: Relationship with apical diameter. New Phytol 130:503–509. Peat HJ, Fitter AH. 1994. The distribution of mycorrhizas in the British flora. New Phytol 125:845–854. Pirozynski KA, Malloch DW. 1975. The origin of land plants: a matter of mycotrophism. Biosystems 6:153– 164. Plawsky JL. 1993. Transport in branched systems 1: steadystate response. Chem Eng Comm 123:71–86. Pomerleau R, Lotie N. 1962. Relationships of dieback to the rooting depth of white birch. For Sci 8:219–224. Pregitzer KS, Hendrick RL, Fogel R. 1993. The demography of fine roots in response to patches of water and nitrogen. New Phytol 125:575–580. Redecker D, Kodner R, Graham LE. 2000. Glomalean fungi from the Ordovician. Science 289:1920–1921. Reekie EG, Bazzaz FA. 1987. Reproductive effect in plants: 2. Does carbon reflect the allocation of other resources: Am Nat 129:876–896. Remy W, Taylor TN, Hass H, Kerp H. 1994. Four hundred million year old vesicular arbuscular mycorrhizae. Proc Natl Acad Sci USA 91:11841–11843.

Fitter Retallack GJ. 1997. Early forest soils and their role in Devonian global change. Science 276:583–585. Rieger M, Litvin P. 1999. Root system hydraulic conductivity in species with contrasting root anatomy. J Exp Bot 50:201–209. Robinson D, Rorison IH. 1988. Plasticity in grass species in relation to nitrogen supply. Funct Ecol 2:249–257. Rose DA. 1983. The description of the growth of root systems. Plant Soil 75:405-415 Ryser P. 1996. The importance of tissue density for growth and life-span of leaves and roots: a comparison of five ecologically contrasting grasses. Funct Ecol 10:717– 723. Shaver GR, Billings WD. 1975. Root production and root turnover in a wet tundra ecosystem at Barrow, Alaska. Flora 165:247–267. Sprague HB. 1933. Root development of perennial grasses and its relation to soil condition. Soil Sci 36:189–209. Stokes A, Fitter AH. Coutts MP. 1995a. Responses of young trees to wind: effect on root growth. In: Coutts MP, Grace J, eds. Trees and Wind. Cambridge, UK: Cambridge University Press, pp 264–275. Stokes A, Fitter AH, Coutts MP. 1995b. Responses of young trees to wind and shading: effects on root architecture. J Exp Bot 46:1139–1146. Tatsumi J, Yamauchi A, Kono Y. 1989. Fractal analysis of plant root systems. Ann Bot 64:499–503. Taub DR, Goldberg D. 1996. Root system topology of plants from habitats differing in soil resource availability. Funct Ecol 10:258–264. Van Noordwijk M, Spek LY, de Willigen P. 1994. Proximal root diameter as predictor of total root size for fractal branching models. I. Theory. Plant Soil 164:107–117. Van Vuuren MMI, Robinson D, Griffiths BS. 1996. Nutrient inflow and root proliferation during the exploitation of a temporally and spatially discrete source of nitrogen in soil. Plant Soil 178:185–192. Waldron LJ. 1977. The shear resistance of root-permeated homogeneous and stratified soil. Soil Sci Soc Am J 41:843–849. Weaver JE. 1919. The ecological relations of roots. Carnegie Inst Wash Pub No. 286. Weaver JE. 1958. Classification of root systems of forms of grassland and a consideration of their significance. Ecology 39:393–401. Werner C, Smart JS. 1973. Some new methods of topologic classification of channel networks. Geogr Anal 5:271– 295. West GB, Brown JH Enquist BJ. 1999. A general model for the structure and allometry of plant vascular systems. Nature 400:664–667. Zhang H, Forde BG. 1998. An Arabidopsis MADS box gene that controls nutrient-induced changes in root architecture. Science 279:407–409.

3 The Root Cap: Structure and Function Andreas Sievers, Markus Braun, and Gabriele B. Monshausen University of Bonn, Bonn, Germany

I.

INTRODUCTION

II.

ROOT CAP IN EMBRYOGENESIS

In monocots, the different cell types of the root cap originate from the same cell during embryogenesis (Von Guttenberg, 1968), whereas in many dicots, the primary root cap is a composite of cells derived from two different origins. In the embryogenesis of Arabidopsis, the first asymmetric division of the zygote forms a terminal and a basal cell, both of which contribute to the generation of the root cap (Dolan et al., 1993; Scheres et al., 1994). Derivatives of the terminal cell give rise to the epidermal/lateral root cap initials. The initials of the central columella originate from the hypophysis, the uppermost derivative of the basal cell. In the triangular stage of embryogenesis the hypophysis divides to produce two daughter cells, one of which gives rise to the quiescent center, while the other—at the heart stage—forms the columella initials (Scheres et al., 1994). In garden cress (Lepidium sativum) amyloplast starch is accumulated within the prospective statocytes (Fig. 2A). These are the only root cells to do so. During embryogenesis columella cells are not structurally polarized, and the embryonic root, under in situ conditions, does not respond to gravity. In root cap cells of fully developed cress embryos, the transition to dormancy is accompanied by a loss of amyloplast starch (Fig. 2B) and by an accumulation of storage lipids and protein vacuoles (Friedrich and Sievers, 1985).

The root cap covers the outermost tip of roots, from the outgrowing primary root up to the outermost branches of the root system.Whatever the developmental stage is, the root cap protects the apical meristems, senses the direction of gravity and other environmental signals, and generates the rhizosphere. In contrast to most other plant tissues, the root cap (Fig. 1), as a persisting structural entity, consists of cells determined to eventually separate from the cap tissue. Renewed by meristematic divisions, statocytes (i.e., gravity-perceiving cells; Fig. 1C) develop and are transformed into mucilage producing secretion cells (Fig. 1D), which finally slough off from the cap, often as metabolically active border cells. The time for a single cell to proceed through the root cap depends on the cycle time, the size of the root cap, and the species. In general, only a few days are required to renew the cap completely, from 1 day in Zea mays to 6–9 days in Convolvulus (Barlow, 1975). In roots of open-type construction, as in most dicotyledonous species, there is no sharp boundary between the root cap and the root proper. In closed construction, as in grasses, there is a distinct separation. In spite of these anatomical features, the functions of the root caps of both types are the same. In the following sections we focus on the functional aspects of the root cap in the light of specific structural features.

33

34

Sievers et al.

Figure 1 (A) Median longitudinal section of cress root cap showing the regular pattern of cell arrangement. All cells shown here originate from the embryonic root cap, but during further development of the root, every cell generated by meristematic division will proceed through these stages. g— direction of gravity. Bar ¼ 10
m. Schematic representations of the different cell types indicated by solid lines are shown in panels (B–D). (B) Meristematic cell from the dermatocalyptrogen with a central nucleus (n) taking up much of the cell volume and proplastids (p) with some starch. (C) The nucleus (n) of a statocyte is placed at the proximal cell pole and a complex of ER at the distal cell pole. Amyloplasts (a; starch indicated in black) lie on this ER complex. Small vacuoles (v) and mitochondria (m) are seen in the cell center. (D) Secretion cell whose polarity was changed according to its differing function: at the distal (outer) cell pole dictyosomes (d) have accumulated, producing vesicles filled with mucilage. Fusion of these vesicles with the plasma membrane and the release of mucilage is shown by arrowheads. The nucleus (n) is found at the cell center and the vacuoles (v) have increased in size. (A, Original by Ulrike Wellenbrock.)

III.

ROOT STATOCYTES DURING GERMINATION

At the onset of imbibition of dry cress seeds, cells of the root cap become hydrated and starch is synthesized in the amyloplasts (Fig. 2C). At that phase, membranes of the endoplasmic reticulum (ER) are formed in the vicinity of the storage lipids. This is caused mainly by peripheral location of the lipid bodies that line up along the cell periphery (Sargent and Osborne, 1980; Hensel, 1986). Coincidentally, microtubules

appear near the distal cell edges of the statocytes, and an increasing amount of ER at the distal cell pole occurs about 10–15 h after the onset of imbibition. After 2 h of soaking, the plastids begin to synthesize starch, a process that increases during the next 8– 13 h. The increased weight and a reduction of the viscosity of the cytoplasm cause the amyloplasts to sediment into the physically lowermost part of the statocytes. Amyloplasts rest on the developing complex of distal ER in downward growing roots (cf. Fig. 1C, Fig. 3). The development of statocyte polarity is caused neither by the action of gravity per se nor by the gravity-dependent amyloplast sedimentation. This was shown by developing roots that were kept on the rotating horizontal axis of clinostats from the dry seed stage onward (Sievers et al., 1976; Hensel and Sievers, 1981) and by seed germination in a microgravity experiment on a Spacelab mission (Volkmann et al., 1986; Perbal et al., 1987). Instead, the polar differentiation is the result of a genetically prepatterned developmental program. The same argument holds true for the proximal position of the nucleus, which in embryonic cells and in very young (2–8 h after the onset of soaking) statocytes appears to be centrally located (Friedrich and Sievers, 1985). The distance between the nucleus and the proximal periclinal cell wall remains constant during the longitudinal growth of the statocytes, leading to the proximal nuclear position mentioned earlier (Hensel, 1986). Furthermore, microgravity influences starch metabolism and ER development, since less starch and more ER were found under microgravity (Moore et al., 1986, 1987; Volkmann et al., 1986) and under conditions of simulated weightlessness (Hoson et al., 1997). In addition, the diameter of lipid bodies increases under both conditions (Volkmann et al., 1986; Hoson et al., 1997). IV.

STATOCYTES IN MATURE ROOTS

A.

Nucleus

The nucleus occupies a position near the proximal periclinal cell wall of the statocytes of most plants. This has been shown for Arabidopsis thaliana Heynh. (Olsen et al., 1984), Hordeum vulgare (Moore, 1985), Lens culinaris (Perbal et al., 1987), Lepidium sativum (Fig. 3; see also Sievers and Volkmann, 1972), Pisum sativum (Olsen and Iversen, 1980), and Zea mays (e.g., Moore, 1983). Data for nuclei sedimenting in the direction of gravity have been reported (Hestnes and Iversen, 1978; Ransom and Moore, 1983; Lorenzi

Root Cap

35

Figure 2 Development of the root apex of cress. (A) During embryogenesis, the prospective statocytes of the embryonic root cap have accumulated starch (s, statenchyma filled with starch grains, i.e., amyloplasts). The embryo is still connected to the mother plant by the suspensor (su). (B) Starch is hydrolyzed before the seed enters dormancy, hence, between embryogenesis (panel A) and germination (panel C). m, meristem; I–VI denote the storeys of the root cap. (C) Starch grains (amyloplasts) are found during germination of the seed. This root cap was taken from a 10-h old seedling. Note that the root had been inverted; amyloplasts, which always sediment into the physically lowermost part of the cells, rest in the proximal area of the statocytes. Bars ¼ 10
m. (Originals by Ulrike Friedrich.)

and Perbal, 1990). The distance between the nucleus and proximal periclinal cell wall of cress statocytes remains constant irrespective of the increasing length of the statocytes. Since treatment with cytochalasin B (a drug known to destroy actin microfilaments) led to sedimentation of the nuclei, a role of actin microfila-

ments in anchoring the nucleus was suggested (Hensel, 1985; Lorenzi and Perbal, 1990). B.

Proplastids of the meristematic layer(s) accumulate starch and, because of this transformation into the amyloplast stage, sediment according to gravity (cf. Barlow et al., 1984). Amyloplasts from coleoptiles have negative zeta potentials of about 19:4 mV, as measured by rate of transport in an electric field, a feature that may also be valid for amyloplasts of root statocytes (Sack et al., 1983). C.

Figure 3 Statocyte of a cress root from a position near the median plane of the root. See Fig. 1C for explanation and abbreviations. Marker bar ¼ 1
m. (From Busch and Sievers, 1990.)

Amyloplasts

Endoplasmic Reticulum

Root statocytes of all plants investigated so far have only cortical ER (the distal ER complex included) but no ‘‘endoplasmic’’ ER. The location, amount, and even the internal organization of the ER in statocytes are reported to vary depending on the species being studied. Therefore, we shall introduce the well-studied cress statocytes (Fig. 3) as a model system and then describe some of the properties of other systems. In the first studies of cress root caps, Sievers and Volkmann (1972) and Volkmann (1974) noted that the number of ER layers in the distal ER complex increases with statocyte development. By means of inhibitor experi-

36

ments, it was shown that ER movement within statocytes and anchorage at the distal cell pole depend on the action of the cytoskeleton. The ER in mature statocytes is produced by an outgrowth of the outer membrane of the nuclear envelope (Hensel, 1985). An actin microfilament-dependent process translocates the newly formed ER membranes into the cortical cytoplasm of the cells (Hensel, 1988). Thereafter, coordinated action of plasma membrane-bridged microtubules and actin microfilaments translocates the ER into the distal cell pole (Hensel, 1987). The retranslocation of ER, displaced by centrifugation from the distal cell pole, is also driven by actin microfilament action (Wendt and Sievers, 1986). The distal ER complex is anchored in its position by microtubules arranged in an interlaced layer near the plasma membrane of the distal cell edges (Hensel, 1984). Actin microfilaments, however, also contribute to this anchorage, since only 7 min of treatment with cytochalasin D (an actin microfilament-destroying drug, more potent in action than cytochalasin B) led to disintegration of the complex (Hensel, 1987). It was suggested that cortical actin microfilaments may hold the ER complex under tension. This contributes to the regular appearance of the distal ER complex in control statocytes. Although the ER transported along the anticlinal cell walls may consist of smaller elements of the tubular type (Stephenson and Hawes, 1986; Hensel, 1987), the organization of the distal ER complex in cress was shown by freeze-etching to be cisternal (Sievers and Volkmann, 1977). Several species do not evolve an ER complex at the distal cell pole. Nevertheless, examples for distal ER complexes in plants other than cress were presented (Volkmann, 1974; Olsen and Iversen, 1980). An increase in ER during statocyte development appears to be a general feature for root caps (cf. Juniper and Clowes, 1965; Barlow and Sargent, 1978; Stephenson and Hawes, 1986; Hensel, 1987). Values that vary from 240
m2 ER area per statocyte up to > 9000
m2 for maize have been reported. The ER seems to be stable in its original location, possibly because of the anchoring action of the cytoskeleton. Only a few exceptions (e.g., Griffiths and Audus, 1964; Juniper and French, 1970, 1973) of an ER redistribution on gravitropic stimulation have been reported. Under certain experimental conditions, in particular inversion of the roots, formation of additional ER is induced beneath the amyloplasts at the proximal, then physically lowermost, part of the statocytes (Volkmann and Sievers, 1975). This indicates that gravity, obviously by an action of the sedimenting amyloplasts, may induce

Sievers et al.

the formation of ER at cell sites differing from those of normally grown roots. If roots grown in the normal vertical direction are rotated for 20 h on the horizontal axis of a slow (two rotations per minute) clinostat, the polar arrangement of the organelles is destroyed. The ER appears at different sites in the statocytes, often forming whorls or aggregates. As pointed out by Hensel and Sievers (1980), this overstimulation by continuous omnilateral gravistimulation leads to the selfdestruction of the statocytes. This is further indicated by a confluence of lipid bodies, the appearance of autophagosomes, the loss of amyloplast starch, and digestion of anticlinal cell walls. The effect of the self-destruction of statocytes was twofold: (1) the graviresponse of the overstimulated roots was drastically reduced, and (2) the root meristem responded to this damage by reducing the period of the cell cycle, thereby causing a faster repair of the statocytes (Hensel and Sievers, 1980). It should be mentioned that the unique lack of ‘‘endoplasmic’’ ER in mature statocytes—in conjunction with the lack of prominent endoplasmic cytoskeletal elements—facilitates the sedimentation of statoliths, thereby optimizing graviperception. The function of the remarkable amount of ER in mature statocytes can be interpreted in connection with the existence and the role of the Ca2þ -ATPase in plant ER membranes which was discovered in cress roots (Buckhout, 1983; see also the chapters of Pilet [30], Porterfield [20, 29], and of Poovaiah et al. [31] in this volume). D.

Other Organelles

Other organelles in statocytes, such as mitochondria, dictyosomes, vacuoles, and microbodies, are distributed randomly, although their distribution is limited to the area between the proximal nucleus and the distal ER complex with the sedimented amyloplasts. Lipid bodies remain, in general, in the vicinity of ER membranes. E.

Cytoskeleton

The cytoskeleton is a highly dynamic system of filamentous proteins with many functions in cell shaping, organization, motility, and signaling. Extensive tissuespecific arrangements of cortical and endoplasmic actin microfilaments and microtubules have been reported in the various cell types of roots and shoots. In statocytes, however, all labeling techniques applied so far have failed to visualize prominent endoplasmic actin microfilament bundles and microtubules

Root Cap

(Balusˇ ka et al., 1997). Soon after termination of the mitotic division in the prospective statocytes, distinct actin microfilament bundles can no longer be visualized and microtubules are limited to dense cortical arrays (Hensel, 1984; Balusˇ ka et al., 1997). This unique cytoskeletal arrangement is considered responsible for the cell polarity, the exclusion of larger organelles (e.g., nucleus, plastids, ER membranes) from the interior of the cells, and the absence of cytoplasmic streaming. In statocytes of other plant organs—coleoptiles and hypocotyls, where cytoplasmic streaming interferes with sedimentation of the amyloplast statoliths— actin cables have been observed (White and Sack, 1990). Gravity-directed sedimentation of the starchcontaining amyloplasts in root cap statocytes may also be related to the specific cytoskeletal properties which may be due to the specific calcium concentration in the cytosol. The diffuse actin labeling found at the cell periphery and close to the sedimented amyloplasts (Balus˘ ka et al., 1997; Blancaflor and Hasenstein, 1997) suggests that actin is organized in the form of delicate meshworks of oligomers, as is indicated by GFP-talin-transformed Arabidopsis seedlings (M. Jaideep, unpublished results). Interestingly, actin microfilament bundles were visualized in enzymatically extracted root cap tissues (Hensel, 1988; White and Sack, 1990), which indicates a potential ability of the actin cytoskeleton to rapidly change its organization. Inhibitor experiments indicate that the translocation and the polar distribution of the ER as well as the movements of the amyloplasts are based on the actomyosin system. The motor protein myosin was immunocytochemically detected in root tips of Allium cepa (Parke et al., 1986) and in statocytes of cress roots (Balusˇ ka and Hasenstein, 1997) by an antibody against animal myosin which cross-reacted on Western blots with a plant polypeptide of approximately 200,000 molecular weight. Myosin immunofluorescence was found in the vicinity of the sedimented amyloplasts in statocytes. Interconnections between actin microfilaments and plastids were described and are involved in the light-dependent movement of chloroplasts (Grolig and Wagner, 1988; Grolig, 1990). F.

Importance of Statocytes’ Structural Polarity

The structural polarity of statocytes appears to be a precondition for graviperception. This conclusion can be drawn from the fact that the polar differentiation is the result of an endogenous developmental program

37

and from the fact that the lack of endoplasmic structures like ER and prominent cytoskeletal elements facilitates sedimentation of statoliths. In addition, especially the results of two experiments support this conclusion: 1. The biogenic polarity was changed to a physical stratification by root-tip-directed centrifugation (20 min at 50–2000 g; Sievers and Heyder-Caspers, 1983). Within the following 8–10 min at 1 g, the structural polarity at the distal cell pole was reestablished in most statocytes, regardless of their orientation to the gravity vector. The lag phase of graviresponse was also increased by 8–10 min in centrifuged roots as compared to controls and independent of the applied centrifugation dose. The kinetics of the response were identical to controls. That means that some reorganization of the statocyte after stratification is necessary and sufficient for graviperception. 2. Treatment of roots with gibberellic acid and kinetin causes not only complete destarching of amyloplasts but also a total loss of structural polarity in statocytes and graviresponsiveness of roots (Busch and Sievers, 1990). Twenty-two hours after removal of the hormones, the polar arrangement of cell organelles was restored and starch was resynthesized so that the roots again responded gravitropically.

V.

MUCILAGE SECRETION BY ROOT CAPS

The outer layers of the root cap consist of mucilagesecreting cells (Fig. 4) which in maize may occupy as much as 40% of the cap volume (Moore, 1984). While in maize most of the mucilage is produced by lateral cap cells, one to two layers beneath the tissue surface, secretion also takes place in the most peripheral cap cells (Vermeer and McCully, 1982). The transition of statocytes to secretion cells is accompanied by a decrease in amyloplast starch and by an increase in the number of vacuoles and dictyosomes (Moore and McClelen, 1983). The trans-cisternae of the dictyosomes become hypertrophied and mucilage-containing Golgi vesicles are formed (Fig. 4D,E; Staehelin et al., 1990), which are transferred to the plasma membrane in an active, actin microfilament-dependent process for exocytosis (Mollenhauer and Morre´, 1976). In addition, accumulation of ER near the nucleus of secreting cells indicates a role of ER in the membrane flow (Volkmann and Czaja, 1981). Both membrane traffic from the ER to the dictyosomes and exocytosis of the mucilage-containing vesicles seem to be Ca2+-regulated processes. After treatment with cyclopiazonic

38

Figure 4 (A) Root tip of maize surrounded by a drop of mucilage. Bar ¼ 1 mm. (B) The mucilage contains numerous isolated cells which are sloughed off from the periphery of the root cap. Bar ¼ 0:1 mm. (C) The final fate of root cap cells of a Phleum root, shown in toto, the sloughing of the root cap cells is seen. Bar ¼ 10
m. (Original by Hanna Zieschang.) Corresponding images of secretion cells of cress root caps (D) after rapid freezing and freeze-etching and (E) after fixation and thin sectioning. Dictyosomes (d), Golgi vesicles (gv), ER, plasma membrane (pm), cell wall (cw), and the incorporation of mucilage-containing vesicles (asterisks) are shown by surface view or in thin section, respectively. At the arrow (panel D) a late fusion stage of a Golgi vesicle with the plasma membrane can be seen (note the various intramembranous particles). Bars ¼ 0:1
m. (Originals by Dieter Volkmann.)

acid, an inhibitor of the Ca2+-ATPase in the ER, dense aggregates of ER and inactive dictyosomes were formed in the secretory cells (Busch and Sievers, 1993). Exocytosis was stimulated even by small increases in cytosolic [Ca2+] above the normal resting [Ca2+] (Carroll et al., 1998). Dialysis of the protoplasts with annexins (which were shown to be expressed in root cap cells as well as differentiating vascular tissue and elongating cells) also stimulated exocytosis at resting cytosolic [Ca2+], though there was no additive effect at very high [Ca2+] (Carroll et al., 1998; see also Chapter 31 by Poovaiah et al. in this volume).

Sievers et al.

Few investigations have studied the transition from statocytes to secretion cells. However, since cutting root caps into halves has initiated mucilage secretion in the then outer cells, Barlow (1984) has suggested a positional determination of cell function. In intact roots, the transition may also be a consequence of the time/differentiation sequence caused by the meristematic cycle time. The products of mucilage secretion are not the same in the various species investigated, but the main components of the mucilage are carbohydrates (94% [w/w] in maize), with fucose being one of the most prominent sugars (Bacic et al., 1986). A glycine-rich protein may be the major component of the maize mucilage protein fraction (Matsuyama et al., 1999a), which makes up
6% of the mucilage dry weight (Bacic et al., 1986). Only 3% (w/w) uronic acids were found in maize mucilage, indicating that maize root mucilage contains a rather low proportion of acidic pectic polysaccharides and that its enormous capacity for gel formation may be due, at least in part, to interactions of various other polysaccharides (Bacic et al., 1986). Even when fully hydrated, a drop of maize root mucilage retains its integrity and the associated water (99.9% of wet weight) is entrapped by a polymer network of relatively minute dry weight (Guinel and McCully, 1986). The water potential of the fully expanded gel is very low (7:3 kPa), indicating that the mucilage would lose water to all but the wettest soils (Guinel and McCully, 1986). Interestingly, the physical properties of the root cap mucilage change considerably with changing degrees of hydration: upon loss of water, the surface tension of the mucilage is strongly reduced, whereas its viscosity increases (Read and Gregory, 1997). These findings support the idea that mucilage plays a major role in the maintenance of root–soil contact in drying soil: as the surface tension decreases, the ability of the mucilage to wet the surrounding soil particles becomes greater, and as the viscosity and elasticity increase, the rhizosphere structure is stabilized and hydraulic continuity is upheld (Read and Gregory, 1997). The mucilage secreted by the root cap cells forms sizable pockets in the periplasmic space and is exuded into the extracellular space even under conditions of water stress, though it is not known how the large polymer passes through the wall (Guinel and McCully, 1986). The pattern of ice crystal formation in freeze-substituted material suggests that the content of the secretory vesicles and the mucilage in the periplasmic space are much less hydrated than the extracellular mucilage, possibly because physical pressure (turgor, compres-

Root Cap

sion) prevents water uptake until the mucilage is free to expand (Guinel and McCully, 1986). The intercellular connection between peripheral secretory cells is maintained at primary pit fields where mucilage does not accumulate. The plasmodesmata are apparently occluded prior to separation of the cells from the root cap (Vermeer and McCully, 1982). The latter process is dependent on the expression of a pectinmethylesterase (Hawes et al., 1998). Though some of the detached cells (Fig. 4B,C) are destined to die quickly (Matsuyama et al., 1999b), many remain alive and metabolically active for extended periods (Vermeer and McCully, 1982). The differentiation of root cap cells into these border cells (which can be induced in culture to divide and differentiate into organized tissue) is accompanied by a dramatic switch in gene expression and many of the synthesized proteins are quickly exported into the extracellular environment (Hawes et al., 1998). Apart from affecting the rhizosphere, border cells also seem to regulate the turnover of the root cap. The mitotic activity of the root cap meristem is apparently continuous when roots are grown under conditions where border cells are constantly dispersed. However, root cap mitosis is inhibited if border cells do not separate but remain appressed to the root periphery in the absence of free water (Brigham et al., 1998, and references cited therein). Upon removal of the border cells, mitosis is quickly (within 5–15 min) induced in the meristem, and the expression of genes involved in physiological processes specific for different cap regions (starch synthesis in columella, cell wall degradation at cap periphery) is increased concomitantly. These findings support the notion that root cap differentiation is a coordinated and regulated continuous process (Barlow, 1984). Nevertheless, individual cells may remain for some time in intermediate stages, depending on the state of accumulation of the border cells (Brigham et al., 1998).

VI.

ROLE OF ROOT CAP IN SENSING ENVIRONMENTAL SIGNALS

The basic functions of the root system are anchorage of the plant in the soil, uptake of water and mineral nutrients, and delivery of growth substances. In order to perform these functions successfully, roots must be able to respond to their surroundings, e.g., by adjusting growth and actively modifying the rhizosphere. This necessitates the capacity to sense environmental signals. As the outermost tissue that covers the grow-

39

ing root apex, the root cap plays a central role in the perception of such signals. A.

Water Gradients

Near-surface soils (
upper 0.3 m) are subject to intensive drying by direct water evaporation as well as by root water extraction. At very low water contents, steep humidity gradients develop in the soil (Wraith and Wright, 1998). These are sufficient to induce hydrotropic growth in roots (0:5 MPa mm1 for the agravitropic pea mutant Pisum sativum ageotropum; Takano et al., 1995). Though the hydrosensing cells have not yet been identified, it has been shown that perception of a water gradient takes place in the root cap; decapped roots fail to curve hydrotropically (Takahashi and Scott, 1993), and applying a very localized water potential gradient directly to the root cap via sorbitol-containing agar blocks caused the root to grow toward the higher water potential (Takano et al., 1995). The reaction time required for the commencement of the hydrotropic response is
3–4 h in roots of the pea mutant ageotropum. Curvature is initiated in the apical region of the elongation zone (Takano et al., 1995) and is preceded by a decrease in cell wall extensibility on the prospective concave side of the root (Hirasawa et al., 1997). For additional details see Chapter 20 by Porterfield in this volume. B.

Mechanical Impedance

All roots growing in soil experience mechanical impedance to varying degrees. While most studies have concentrated on the long-term effects of strong mechanical stress on root growth and morphogenesis (Bengoun and Mullins, 1990; Chapter 45 by Masle, in this volume), the observation of mechanically induced membrane potential changes (Monshausen and Sievers, 1998) and increases in cytosolic [Ca2+] (Legue´ et al., 1997) has shown that roots are capable of perceiving weak mechanical stimuli of short duration, and that peripheral cells of the root cap are particularly sensitive (Legue´ et al., 1997). These results support earlier findings by Goss and Russell (1980) that a slight (persistent) mechanical stimulus has a remarkably large initial effect on root elongation. Upon encountering small, loosely packed glass beads, the elongation rate of an intact maize root was rapidly reduced by more than two-thirds for a period of 10 min, after which it quickly returned to the original, unimpeded value. No change in the rate of elongation was observed in maize roots which had been decapped

40

Sievers et al.

45 min prior to contact with the beads. Apparently, the root cap must have a central role in regulating the root response to very low mechanical resistance. The root cap may also reduce mechanical stress by sloughing off cells, thereby reducing the frictional resistance experienced by the growing root (Bengough and McKenzie, 1997). C.

Gravity

The requirement of the root cap for graviperception by roots is well established; experimental evidence for this goes back to the pioneering root tip-removal experiments of Charles Darwin (1880). The unequivocal proof was provided by the careful decapping experiments of Juniper et al. (1966). Decapped roots did not respond to gravitropic stimulation, whereas their growth was unaltered. More recent evidence for the statocyte function of the root cap columella cells comes from laser ablation experiments (Blancaflor et al., 1998) specifying the innermost columella cells as the most important sites for graviperception. The mechanism of graviperception is far from being understood. However, since gravity can only work on (deform or displace) masses, biological gravisensors must be equipped with receptors which are able to perceive the information resulting from the physical process of deformation or displacement known as susception. For graviperception by higher plants the starch–statolith theory published by Ne˘mec (1900) and Haberlandt (1900) is widely accepted. In gravisensitive tissues of shoots and roots, starch-containing sedimentable amyloplasts were observed in specialized cells, the statocytes, and are believed to act as susceptors or statoliths (cf. Volkmann and Sievers, 1979; Wilkins, 1984; Bjo¨rkman, 1988; Sack, 1991, 1997; Chen et al., 1999; Chapters by Pilet [30] and by Porterfield [20] in this volume). A similar phenomenon was found in rhizoids and protonemata of the green alga Chara, where vesicles filled with barium-sulfate crystals were identified as statoliths whose gravitydirected sedimentation promotes gravitropic curvature (Sievers et al., 1996; Braun, 1997). In higher plants, a good correlation was found between gravisensitivity and amyloplasts of different starch content (Iversen, 1969; Perbal and Rivie`re, 1976; Busch and Sievers, 1990; Sack, 1991; Kiss et al., 1996, 1997; Weise and Kiss, 1999). Even starchless amyloplasts produced high signal-to-noise ratios to activate the hypothetical receptor molecules. Furthermore, Arabidopsis mutants lacking the endodermal parenchyma in shoots and roots were reported to show no shoot but root gravi-

tropism (Fukaki et al., 1998). Since amyloplast-containing statocytes are present in the shoot endodermis but not in the root endodermis, these mutants strongly support the proposed hypothesis. Nonstatolith theories have suggested the possibility that amyloplast sedimentation might not be the sole mechanism of gravity sensing (Sack, 1997). However, there is evidence from high-gradient magnetic field experiments that amyloplast sedimentation in statocytes of roots and shoots is sufficient to initiate gravitropic curvature. Root curvature occurred in the direction in which the amyloplasts were displaced by the magnetic field (Kuznetsov and Hasenstein, 1996). Molecules involved in the mechanism of graviperception, the transduction of the physical process of sedimentation into a physiological signal, are still to be characterized. The presentation time is very short; for cress roots it is 12 s (Iversen and Larsen, 1973). By intermittent stimulation, the minimum time to be perceived (perception time) by Avena coleoptiles and cress roots was determined to be 0.5 s (cf. Hejnowicz et al., 1998). During the perception time, the amyloplasts are displaced
8 nm. Thus, attention should be paid to small displacements of statoliths in order to understand how gravity effects on statoliths are transferred to competent cellular structures (Sievers et al., 1991b). Cytoskeletal elements of the statocytes are the most likely candidates to be involved in the transduction of the statolith sedimentation into a physiological signal which is transmitted to the responding target cells in the root elongation zone (Sievers et al., 1991b). Circumstantial evidence for this model is provided by the finding that treatment of Phleum roots with the actin inhibitor cytochalasin D inhibited gravityinduced pH changes at the root surface (Monshausen et al., 1996). Eight minutes after tilting, the surface pH of gravistimulated control roots starts to decrease at the meristem and apical elongation zone of the physically upper root side (Zieschang et al., 1993; Monshausen et al., 1996) On the other hand, cytochalasin D had no effect on the graviresponse in roots of some other species (Staves et al., 1997). The unique lack of prominent actin bundles and microtubules in the statocyte interior probably facilitates unconstrained sedimentation of their statoliths (Balus˘ ka and Hasenstein, 1997; Volkmann and Balus˘ ka, 1999). A proposed delicate meshwork of oligomeric actin surrounding the statoliths and myosins localized in the vicinity of the sedimented statoliths (Balus˘ ka and Hasenstein, 1997) are the most likely basis for statolith movements and the displacement of statoliths under microgravity conditions (Fig. 5; Volkmann et al.,

Root Cap

1991; Driss-Ecole et al., 2000). This was also shown in single-cell systems (Braun, 1996, 1997). However, inhibitors affecting the polymerization of actin (cytochalasin, phalloidin) and tubulin (colchicine, taxol) cause significant differences in the sedimentation rate of amyloplasts (Sievers et al., 1989). These results also support the idea that actin microfilaments or microtubules are involved in statolith-mediated graviperception. Cytoskeletal filaments could be interconnected with mechanosensitive ion channels either in the cortical ER or other membranes (Fig. 6; Falke et al., 1988; Schroeder and Hedrich, 1989; Pickard and Ding, 1992; Hwang et al., 1997). Even the energy of small displacements of statoliths amplified by the cytoskeleton would allow dramatic changes in ion fluxes including that of calcium which appears to play a crucial role in signal transduction (Bjo¨rkman, 1988; Sievers et al., 1991a; Sievers and Busch, 1992; Chapter 31 by Poovaiah et al. in this volume). This is in full agreement with the tensegrity model of cytoskeletonmediated mechanotransduction (Ingber, 1993). In the characean rhizoids and protonemata a crucial but different role of actin in gravitropism was demonstrated (Braun and Wasteneys, 2000). Further support for a participation of the cytoskeleton in graviperception comes from a genetic approach. A gene has been identified that affects the graviperception phase in Arabidopsis mutants. The ARG1 locus encodes for a 45-kDa DnaJ-like protein

41

Figure 5 Light microscope photographs of median longitudinal sections through the statenchyma of cress roots. (A) Control cress root which was fixed with potassium permanganate at 1 g. (B) Cress root which was fixed at the end of the 6-min microgravity phase of the parabolic flight of a rocket (TEXUS). Microgravity conditions resulted in a considerable displacement of the statolith amyloplasts in the direction opposite to that of the gravity vector. Bar ¼ 10
m. (From Volkmann et al., 1991.)

Figure 6 Scheme of a central root statocyte of cress illustrating the difference in the tension of the proposed oligomeric actinnetwork and the asymmetrical activities of ion channels, both induced by gravistimulation. (A) Normal vertical orientation with most actin microfilaments in tension. (B) Horizontal orientation with asymmetrically stretched and relaxed actin microfilaments due to the gravity-induced displacement of the amyloplasts. (a) amyloplast (statolith); (ER) endoplasmic reticulum; (g) direction of gravity; (mf) actin microfilament; (mt) microtubulus; (n) nucleus; (PD) plasmodesma; (PM), plasma membrane. Open and closed stars symbolize postulated channels in two different activity states. (From Sievers et al., 1991b.)

42

containing a coiled-coil domain which is typical for cytoskeleton-interacting proteins (Sedbrook et al., 1999). A cytoskeleton membrane anchoring function seems possible because of the presence of several hydrophobic amino acids in the middle, and a putative transmembrane region. ARG1 may represent a component of the early cytoskeleton-mediated gravity signal transduction chain. However, because ARG1-protein is expressed in all plant tissues and is related to a conserved signal transduction molecule, a more general function of ARG1 in signal transduction, protein folding, or protein trafficking can not be ruled out. Recently, rapid changes in cytosolic pH have been demonstrated within Arabidopsis columella cells upon gravistimulation, suggesting that pH gradients might be involved at a very early stage of the gravitropic signal transduction pathway (Scott and Stro¨mgren Allen, 1999). Calcium and phosphoinositides possibly act as second messengers in the signal transduction pathway (Perera et al., 1999). High concentrations of calcium were detected in statocyte amyloplasts and membranes (Chandra et al., 1982; Busch et al., 1993), and calmodulin concentrations in statocytes are higher than in all other cell types (Allan and Trewavas, 1985), though the cytoplasmic [Ca2+] in statocytes is the same as in other root tissues (Legue´ et al., 1997). Furthermore, the degree of the polar arrangement of organelles in statocytes and the gravisensitivity of roots was reduced or eliminated after application of blockers of stretch-activated calcium channels and inhibitors of calmodulin or Ca2+ATPase activities (Biro et al., 1982; Bjo¨rkmann and Leopold, 1987; Stinemetz et al., 1987; Wendt and Sievers, 1989; Sievers and Busch, 1992). Lu and Feldman (1997) discussed the involvement of a Ca2+/calmodulin-dependent protein kinase in the light-dependent gravitropism of maize roots. These data indicate the involvement of calcium–calmodulin activity in the gravitropic signal-transduction pathway, but only of touch-induced changes (Legue´ et al., 1997). In this context it is noteworthy that a root is not only gravistimulated by tilting (dynamic gravistimulation; Sievers et al., 1991b). Vertically oriented roots commonly used as controls are also permanently (statically) stimulated. Statocytes of roots should be investigated in a stimulus-free microgravity environment— such as in space (Sievers, 1999). Nevertheless, calcium seems to act as a second messenger interfering with the polar transport of auxin within the root cap, and could mediate the redirection of auxin to the physically lower root flank (Lee et al., 1984; Lee and Evans, 1985;

Sievers et al.

Evans and Hasenstein, 1987). Additional information on gravitropic response mechanisms involving calcium and auxin is presented in this volume by Pilet (Chapter 31), by Porterfield (20), and by Poovaiah et al. (31). D.

Aluminum

Root exposure to toxic concentrations of aluminum causes a rapid decrease in the rate of root elongation. While this decrease is chiefly due to the inhibition of extension in the main elongation zone, the apical millimeter of the maize root (comprising the meristem and the apical part of the distal elongation zone) is the most Al-sensitive region of the root (Kollmeier et al., 2000). As aluminum treatment of this region inhibits extension in the more proximal part of the elongation zone, Kollmeier et al. (2000) suggest that a signaling pathway in the root apex mediates the Al signal between the meristem/distal elongation zone and the elongation zone proper (possibly via alterations in basipetal auxin transport). Aluminum binding by root cap mucilage may be one way of protecting the sensitive root apex from the toxic effects of Al (see also Chapter 46 by Matsumoto in this volume). Aluminum accumulates in the mucilage of maize roots, where it is tightly bound and thereby rendered nonphytotoxic (Li et al., 2000). However, Al binding by the mucilage does not confer effective protection to maize roots growing in hydroponic culture. This may reflect the relative amount of mucilage-bound Al to total Al in the bulk solution. In the soil, where Al movement is more restricted and where a root can modify its immediate environment via the activity of border cells (Hawes et al., 1998), binding of Al by mucilage might be more effective. Another explanation for this difference might be that in hydroponics the secreted mucilage is continuously washed away, leaving the apex exposed to the Al containing solution. Incubation of individual border cells with Al induces the production of a thick mucilage layer around each cell, which is correlated with the cessation of Al-induced border cell death (Hawes et al., 2000). E.

Pathogens

In the soil, the root is constantly exposed to a variety of pathogenic organisms. Protection of the root apex is especially vital, as the continued exploitation of the soil for water and nutrients depends on the root’s ability to move into new surroundings. Root cap border cells seem to play a major role in such protective

Root Cap

measures, ranging from temporarily immobilizing nematodes and the production of bacteria-repelling mucilage to acting as decoys for pathogenic fungi (Hawes et al., 2000).

VII.

REGENERATION OF THE ROOT CAP

Removal of the entire root cap without damaging cells of the root body, which is possible in roots of the closed type, provides an artificial system to use in studying the development of statocytes (Barlow, 1974a,b; Barlow and Grundwag, 1974; Hillman and Wilkins, 1982). After decapping, the cells of the quiescent center start to divide, initiating a new root cap. Prior to this, proplastids within the outermost apical cell layer of the quiescent center are transformed into amyloplasts and sediment in the direction of gravity. The onset of gravisensitivity is dependent on the existence of those sedimenting amyloplasts (Hillman and Wilkins, 1982). Twelve hours after decapping, amyloplasts are not displacable even by centrifugation (25 g), whereas after 24 h displacement of amyloplasts by centrifugation occurred. Hillman and Wilkins (1982) suggested a change in the characteristics of the cytoplasm (i.e., rearrangement of the actin cytoskeleton). By using a root cap-specific promoter to express a diphtheria toxin gene in Arabidopsis, several layers of the root caps were genetically ablated continuously throughout the plant’s life cycle (Tsugeki and Fedoroff, 1999). This had severe effects on root development. In these transgenic root caps two layers of columella cells and some of the lateral root cap cells were missing, and the remaining layers were increasingly disorganized. The transgenic roots were agravitropic and showed severely inhibited growth. Whereas the overall mitotic activity of the root meristem seemed decreased in the transgenic plants, cell division was observed in the normally mitotically inactive quiescent center. Epidermal root hair development, vascularization, and vacuolization commenced much closer to the apex than in wild-type roots and the transgenic roots were more branched (Tsugeki and Fedoroff, 1999). The authors conclude that essential components of a signaling system (specifically an auxin redistribution system) that determines root architecture reside in the root cap and disruption of this system by genetic ablation of the cap causes the observed changes in root development.

43

VIII.

CONCLUSIONS

As a consequence of its multitude of functions, such as protection of the meristem, production of mucilage, sensing gravity and other environmental signals, the root cap has attracted the attention of cell biologists and physiologists for over a century. Some of the aspects most thoroughly studied were the structure and function of the Golgi apparatus, mechanisms of membrane flow, exocytosis, development of cytoskeleton-dependent cell polarity, and cytoskeleton-mediated transduction of the gravistimulus. Future research will focus mainly on molecular characterization of elements involved in the manifold signal transduction pathways and on the generation and transmission of signals from the root cap directed at the apical elongation zone of the root. REFERENCES Allan E, Trewavas AJ. 1985. Quantitative changes in calmodulin and NAD kinase during early cell development in the root apex od Pisum sativum. Planta 165:493–501. Bacic A, Moody SF, Clarke AE. 1986. Structural analysis of secreted root slime from maize (Zea mays L.). Plant Physiol 80:771–777. Balus˘ ka F, Hasenstein KH. 1997. Root cytoskeleton: its role in perception of and response to gravity. Planta 203:S69–S78. Balus˘ ka F, Kreibaum A, Vitha S, Parker JS, Barlow PW, Sievers A. 1997. Central root cap cells are depleted of endoplasmic microtubules and actin filament bundles: implications for their role as gravity-sensing statocytes. Protoplasma 196:212–223. Barlow PW. 1974a. Regeneration of the cap of primary roots of Zea mays. New Phytol 73:937–954. Barlow PW. 1974b. Recovery of geotropism after removal of the root cap. J Exp Bot 25:1137–1146. Barlow PW. 1975. The root cap. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. London; Academic Press, pp 21–54. Barlow PW. 1984. Positional controls in root development. In: Barlow PW, Carr DJ, eds. Positional Controls in Plant Development. Cambridge, UK: Cambridge University Press, pp 281–318. Barlow PW, Grundwag M. 1974. The development of amyloplasts in cells of the quiescent centre of Zea roots in response to removal of the root cap. Z Pflanzenphysiol 73:56–64. Barlow PW, Sargent JA. 1978. The ultrastructure of the regenerating root cap of Zea mays L. Ann Bot 42:791–799. Barlow PW, Hawes CR, Horne JC. 1984. Structure of amyloplasts and endoplasmic reticulum in the root caps of

44 Lepidium sativum and Zea mays observed after selective membrane staining and by high-voltage electron microscopy. Planta 160:363–371. Bengough AG, Mullins CE. 1990. Mechanical impedance to root growth: a review of experimental techniques and root growth responses. J Soil Sci 41:341–358. Bengough AG, McKenzie BM. 1997. Sloughing of root cap cells decreases the frictional resistance to maize (Zea mays L.) root growth. J Exp Bot 48:885–893. Biro RL, Hale CC II, Wiegand OF, Roux SJ. 1982. Effects of chlorpromazine on gravitropism in Avena coleoptiles. Ann Bot 50:735–747. Bjo¨rkman T. 1988. Perception of gravity by plants. Adv Bot Res 15:1–41. Bjo¨rkman T, Leopold AC. 1987. Effect of inhibitors of auxin transport and of calmodulin on a gravisensing-dependent current in maize roots. Plant Physiol 84:847–850. Blancaflor EB, Hasenstein K-H. 1997. The organization of the actin cytoskeleton in vertical and graviresponding primary roots of maize. Plant Physiol 113:1447–1455. Blancaflor EB, Fasano JM, Gilroy S. 1998. Mapping the functional roles of cap cells in the response of Arabidopsis primary roots to gravity. Plant Physiol 116:213–222. Braun M. 1996. Immunolocalization of myosin in rhizoids of Chara globularis Thuill. Protoplasma 191:1–8. Braun M. 1997. Gravitropism in tip-growing cells. Planta 203:S11–S19. Braun M, Wasteneys GO. 2000. Actin in characean rhizoids and protonemata. Tip growth, gravity sensing and photomorphogenesis. In: Staiger CJ, Balusˇ ka F, Volkmann P, Barlow P, eds. Actin: a Dynamic Framework for Multiple Plant Cell Functions. Dordrecht, Netherlands: Kluwer Academic Publishers, pp 237–258. Brigham LA, Woo H-H, Wen F, Hawes MC. 1998. Meristem-specific suppression of mitosis and a global switch in gene expression in the root cap of pea by endogenous signals. Plant Physiol 118:1223–1231. Buckhout TJ. 1983. ATP-dependent Ca2+-transport in endoplasmic reticulum isolated from the roots of Lepidium sativum L. Planta 159:84–90. Busch MB, Sievers A. 1990. Hormone treatment of roots causes not only a reversible loss of starch but also of structural polarity in statocytes. Planta 181:358–364. Busch MB, Sievers A. 1993. Membrane traffic from the endoplasmic reticulum to the Golgi apparatus is disturbed by an inhibitor of the Ca2+-ATPase in the ER. Protoplasma 177:23–31. Busch MB, Ko¨rtje KH, Rahmann H, Sievers A. 1993. Characteristic and differential calcium signals from cell structures of the root cap detected by energy-filtering electron microscopy (EELS/ESI). Eur J Cell Biol 60:88–100. Carroll AD, Moyen C, Van Kesteren P, Tooke F, Battey NH, Brownlee C. 1998. Ca2+, annexins, and GTP

Sievers et al. modulate exocytosis from maize root cap protoplasts. Plant Cell 10:1267–1276. Chandra S, Chabot JF, Morrison GH, Leopold AC. 1982. Localization of Ca2þ in amyloplasts of root cap cells using ion microscopy. Science 216:1221–1223. Chen R, Rosen E, Masson PH. 1999. Gravitropism in higher plants. Plant Physiol 120:343–350. Darwin C. 1880. The Power of Movement in Plants. London; John Murray. Dolan L, Janmaat K, Willemsen V, Linstead P, Poethig S, Roberts K, Scheres B. 1993. Cellular organisation of the Arabidopsis thaliana root. Development 119:71–84. Driss-Ecole D, Jeune B, Prouteau M, Julianus P, Perbal G. 2000. Lentil root statoliths reach a stable state in microgravity. Planta 211:396–405. Evans ML, Hasenstein K-H. 1987. Stimulus-response coupling in the action of auxin and gravity on roots. In: Cosgrove DJ, Knievel DP, eds. Physiology of Cell Expansion During Plant Growth. Bethesda, MD: American Society of Plant Physiologists, pp 202–214. Falke LC, Edwards KL, Pickard BG, Misler S. 1988. A stretch-activated anion channel in tobacco protoplasts. FEBS Lett 237:141–144. Friedrich U, Sievers A. 1985. Ontogeny of cell polarity in root statocytes of Lepidium sativum L. in the developing embryo and during germination. Eur J Cell Biol 36(suppl 7):18. Fukaki H, Wysocka-Diller J, Kato T, Fujisawa H, Benfey PN, Tasaka M. 1998. Genetic evidence that the endodermis is essential for shoot gravitropism in Arabidopsis thaliana. Plant J 14:425–430. Goss MJ, Russell RS. 1980. Effects of mechanical impedance on root growth in barley (Hordeum vulgare L.). J Exp Bot 31:577–588. Griffiths HJ, Audus LJ. 1964. Organelle distribution in the statocyte cells of the root-tip of Vicia faba in relation to geotropic stimulation. New Phytol 63:319–332. Grolig F. 1990. Actin-based organelle movements in interphase Spirogyra. Protoplasma 155:29–42. Grolig F, Wagner G. 1988. Light dependent chloroplast reorientation in Mougeotia and Mesotaenium:Biased by pigment-regulated plasmalemma anchorage sites to actin filaments. Bot Acta 101:2–6. Guinel FC, McCully ME. 1986. Some water-related physical properties of maize root-cap mucilage. Plant Cell Environ 9:657–665. Haberlandt G. 1900. U¨ber die Perzeption des geotropischen Reizes. Ber Dtsch Bot Ges 18:261–272. Hawes MC, Brigham LA, Wen F, Woo HH, Zhu Y. 1998. Function of root border cells in plant health: pioneers in the rhizosphere. Annu Rev Phytopathol 36:311327. Hawes MC, Gunawardena U, Miyasaka S, Zhao X. 2000. The role of root border cells in plant defense. Trends Plant Sci 5:128–133.

Root Cap Hejnowicz Z, Sondag C, Alt W, Sievers A. 1998. Temporal course of graviperception in intermittantly stimulated cress roots. Plant Cell Environ 21:1293–1300. Hensel W. 1984. A role of microtubules in the polarity of statocytes from roots of Lepidium sativum L. Planta 162:404–414. Hensel W. 1985. Cytochalasin B affects the structural polarity of statocytes from cress roots (Lepidium sativum L.). Protoplasma 129:178–187. Hensel W. 1986. Cytodifferentiation of polar plant cells: use of antimicrotubular agents during the differentiation of statocytes from cress roots. Planta 169:293–303. Hensel W. 1987. Cytodifferentiation of polar plant cells: formation and turnover of endoplasmic reticulum in root statocytes. Exp Cell Res 172:377–384. Hensel W. 1988. Demonstration by heavy-meromyosin of actin microfilaments in extracted cress (Lepidium sativum L) root statocytes. Planta 173:142–143. Hensel W, Sievers A. 1980. Effects of prolonged omnilateral gravistimulation on the ultrastructure and on the graviresponse of roots. Planta 150:338–346. Hensel W, Sievers A. 1981. Induction of gravity-dependent plasmatic responses in root statocytes by short time contact between amyloplasts and the distal endoplasmic reticulum complex. Planta 153:303–307. Hestnes A, Iversen T-H. 1978. Movement of cell organelles and the geotropic curvature in roots of Norway spruce (Picea abies). Physiol Plant 42:406–414. Hillman SK, Wilkins MB. 1982. Gravity perception in decapped roots of Zea mays. Planta 155:267–271. Hirasawa T, Takahashi H, Suge H, Ishihara K. 1997. Water potential, turgor and cell wall properties in elongating tissues of the hydrotropically bending roots of pea (Pisum sativum L.). Plant Cell Environ 20:381–386. Hoson T, Kamisaka S, Masuda Y, Yamashita M, Buchen B. 1997. Evaluation of the three-dimensional clinostat as a simulator of weightlessness. Planta 203:S187– S197. Hwang J-U, Suh S, Yi H, Kim J, Lee Y. 1997. Actin filaments modulate both stomatal opening and inward K+-channel activities in guard cells of Vicia faba L. Plant Physiol 115:335–342. Ingber DE. 1993. Cellular tensegrity: defining new rules of biological design that govern the cytoskeleton. J Cell Sci 104:613–627. Iversen T-H. 1969. Elimination of geotropic responsiveness in roots of cress (Lepidium sativum) by removal of statolith starch. Physiol Plant 22:1251–1262. Iversen T-H, Larsen P. 1973. Movement of amyloplasts in the statocytes of geotropically stimulated roots: the pre-inversion effect. Physiol Plant 28:172–181. Juniper BE, Clowes FAL. 1965. Cytoplasmic organelles and cell growth in root caps. Nature 208: 864–865. Juniper BE, French A. 1970. The fine structure of the cells that perceive gravity in the root tip of maize. Planta 95:314–329.

45 Juniper BE, French A. 1973. The distribution and redistribution of endoplasmic reticulum (ER) in geoperceptive cells. Planta 109:211–224. Juniper BE, Groves S, Landau-Schachar B, Audus LJ. 1966. Root cap and the perception of gravity. Nature 209:93–94. Kiss JZ, Wright JB, Caspar T. 1996. Gavitropism in the roots of intermediate-starch mutants of Arabidopsis. Physiol Plant 97: 237–244. Kiss JZ, Katembe WJ, Edelmann RE. 1997. Gravitropism and development of wild-type and starch-deficient mutans of Arabidopsis during spaceflight. Physiol Plant 102:493–502. Kollmeier M, Felle HH, Horst WJ. 2000. Genotypical differences in aluminum resistance of maize are expressed in the distal part of the transition zone. Is reduced basipetal auxin flow involved in inhibition of root elongation by aluminum? Plant Physiol 122:945–956. Kuznetsov O, Hasenstein KH. 1996. Intracellular magnetophoresis of amyloplasts and induction of root curvature. Planta 198:87–94. Kuznetsov O, Hasenstein KH. 1997. Magnetophoretic induction of curvature in coleoptiles and hypocotyls. J Exp Bot 48:1951–1957. Lee JS, Evans ML. 1985. Polar transport of auxin across gravistimulated roots of maize and its enhancement by calcium. Plant Physiol 77:824. Lee JS, Mulkey TJ, Evans ML. 1984. Inhibition of polar calcium movement and gravitropism in roots treated with auxin-transport inhibitors. Planta 160:536–543. Legue´ V, Blancaflor E, Wymer C, Perbal G, Fantin D, Gilroy S. 1997. Cytoplasmic free Ca2+ in Arabidopsis roots changes in response to touch but not gravity. Plant Physiol 114:789–800 Li XF, Ma JF, Hiradate S, Matsumoto H. 2000. Mucilage strongly binds aluminum but does not prevent roots from aluminum injury in Zea mays. Physiol Plant 108:152–160. Lorenzi G, Perbal G. 1990. Actin filaments responsible for the location of the nucleus in the lentil statocyte are sensitive to gravity. Biol Cell 68:259–263. Lu Y-T, Feldman LJ. 1997. Light-regulated root gravitropism: a role for, and characterization of, a calcium/calmodulin-dependent protein kinase homolog. Planta 203:S91–S97. Matsuyama T, Satoh H, Yamada Y, Hashimoto T. 1999a. A maize glycine-rich protein is synthesized in the lateral root cap and accumulates in the mucilage. Plant Physiol 120:665–674. Matsuyama T, Yasumura N, Funakoshi M, Yamada Y, Hashimoto T. 1999b. Maize genes specifically expressed in the outermost cells of root cap. Plant Cell Physiol 40:469–476. Mollenhauer HH, Morre´ DJ. 1976. Cytochalasin B, but not colchicine, inhibits migration of secretory vesicles in root tips of maize. Protoplasma 87:39–45.

46 Monshausen GB, Sievers A. 1998. Weak mechanical stimulation causes hyperpolarisation in root cells of Lepidium. Bot Acta 111:303–306. Monshausen GB, Zieschang HE, Sievers A. 1996. Differential proton secretion in the apical elongation zone caused by gravistimulation is induced by a signal from the root cap. Plant Cell Environ 19:1408–1414. Moore R. 1983. A morphometric analysis of the ultrastructure of columella statocytes in primary roots of Zea mays L. Ann Bot 51:771–778. Moore R. 1984. Cellular volume and tissue partitioning in caps of primary roots of Zea mays. Am J Bot 71:1452– 1454. Moore R. 1985. A morphometric analysis of the redistribution of organelles in columella cells in primary roots of normal seedlings and agravitropic mutants of Hordeum vulgare. J Exp Bot 36:1275–1286. Moore R, McClelen CE. 1983. A morphometric analysis of cellular differentiation in the root cap of Zea mays. Am J Bot 70:611–617. Moore R, Fondren WM, Koon EC, Wang C-L. 1986. The influence of gravity on the formation of amyloplasts in columella cells of Zea mays L. Plant Physiol 82:867– 868. Moore R, Fondren WM, McClelen CE, Wang C-L. 1987. Influence of microgravity on cellular differentiation in root caps of Zea mays. Am J Bot 74:1006–1012. Ne˘mec B. 1900. U¨ber die Art der Wahrnehmung des Schwerkraftreizes bei den Pflanzen. Ber Dtsch Bot Ges 18:241–245. Olsen GM, Iversen T-H. 1980. Ultrastructure and movements of cell structures in normal pea and an ageotropic mutant. Physiol Plant 50:275–284. Olsen GM, Mirza JI, Maher EP, Iversen T-H. 1984. Ultrastructure and movements of cell organelles in the root cap of agravitropic mutants and normal seedlings of Arabidopsis thaliana. Physiol Plant 60:523–531. Parke J, Miller C, Anderton BH. 1986. Higher plant myosin heavy-chain identified using a monoclonal antibody. Eur J Cell Biol 41:9–13. Perbal G, Rivie`re S. 1976. Relation entre reaction ge´otropique et e´volution due statenchyme dans la racine d’asperge. Physiol Plant 38:39–47. Perbal G, Driss-Ecole D, Rutin J, Salle G. 1987. Graviperception of lentil seedling roots grown in space (Spacelab Dl mission). Physiol Plant 70:119-126. Perera IY, Heilmann I, Boss WF. 1999. Transient and sustained increases in inositol-1,2,5-trisphosphate precede the differential growth response in gravistimulated maize pulvini. Proc Natl Acad Sci USA 96:5838–5843. Pickard BG, Ding JP. 1992. Gravity sensing by higher plants. Adv Comp Environ Physiol 10:81–110. Ransom JS, Moore R. 1983. Geoperception in primary and lateral roots of Phaseolus vulgaris (Fabaceae): 1. Structure of columella cells. Am J Bot 70:1048–056.

Sievers et al. Read DP, Gregory PJ. 1997. Surface tension and viscosity of axenic maize and lupin root mucilages. New Phytol 137:623–628. Sack FD. 1991. Plant gravity sensing. Int Rev Cytol 127:193– 252. Sack FD. 1997. Plastids and gravitropic sensing. Planta 203:S63–S68. Sack FD, Priestley DA, Leopold AC. 1983. Surface charge on isolated maize-coleoptile amyloplasts. Planta 157:511–517. Sargent JA, Osborne DJ. 1980. A comparative study of the fine structure of coleorhiza and root cells during the early hours of germination of rye embryos. Protoplasma 104:91–103. Scheres B, Wolkenfelt H, Willemsen V, Terlouw M, Lawson E, Dean C, Weisbeek P. 1994. Embryonic origin of the Arabidopsis primary root and root meristem initials. Development 120:2475–2487. Schroeder JI, Hedrich R. 1989. Involvement of ion channels and active transport in osmoregulation and signaling of higher plant cells. TIBS 14:187–192. Scott AC, Stro¨mgren Allen N. 1999. Changes in cytosolic pH within Arabidopsis root columella cells play a key role in the early signaling pathway for root gravitropism. Plant Physiol 121:1291–1298. Sedbrook J, Chen R, Masson P. 1999. ARG1 (Altered Response to Gravity) encodes for a novel DnaJ-like protein which potentially interacts with the cytoskeleton. Proc Natl Acad Sci USA 96:1140–1145. Sievers A. 1999. Gravitational biology in Bonn. Am Soc Gravit Space Biol Newslett 15(3):15–22. Sievers A, Busch MB. 1992. An inhibitor of the Ca2+ATPases in the sarcoplasmic and endoplasmic reticula inhibits transduction of the gravity stimulus in cress roots. Planta 188:619–622. Sievers A, Heyder-Caspers L. 1983. The effect of centrifugal acceleration on the polarity of statocytes and on the graviperception of cress roots. Planta 157:64–70. Sievers A, Volkmann D. 1972. Verursacht differentieller Druck der Amyloplasten auf ein komplexes Endomembransystem die Geoperzeption in Wurzeln? Planta 102:160–172. Sievers A, Volkmann D. 1977. Ultrastructure of gravity-perceiving cells in plant roots. Proc R Soc Lond B 199:525–536. Sievers A, Volkmann D, Hensel W, Sobick V, Briegleb W. 1976. Cell polarity in root statocytes in spite of simulated weightlessness. Naturwissenschaften 63:343. Sievers A, Kruse S, Kuo-Huang L-L, Wendt M. 1989. Statoliths and microfilaments in plant cells. Planta 179:275–278. Sievers A, Kramer-Fischer M, Braun M, Buchen B. 1991a. The polar organization of the growing Chara rhizoid and the transport of statoliths are actin-dependent. Bot Acta 104:103–109.

Root Cap Sievers A, Buchen B, Volkmann D, Hejnowicz Z. 199lb. Role of the cytoskeleton in gravity perception. In: Lloyd CW, ed. The Cytoskeletal Basis of Plant Growth and Form. London; Academic Press, pp 169–182. Sievers A, Buchen B, Hodick D. 1996. Gravity sensing in tipgrowing cells. Trends Plant Sci 1:273–279. Sinclair W, Trewavas AL. 1997. Calcium in gravitropism. A re-examination. Planta 203:S85–S90. Staehelin LA, Giddings TH Jr, Kiss JZ, Sack FD. 1990. Macromolecular differentiation of Golgi stacks in root tips of Arabidopsis and Nicotiana seedlings as visualized in high pressure frozen and freeze-substituted samples. Protoplasma 157:75–91. Staves MP, Wayne R, Leopold AC. 1997. Cytochalasin D does not inhibit gravitropism in roots. Am J Bot 84:1530–1535. Stephenson JLM, Hawes CR. 1986. Stereology and stereometry of endoplasmic reticulum during differentiation in the maize root cap. Protoplasma 131:32–46. Stinemetz CL, Kuzmanoff KM, Evans ML, Jarrett HW. 1987. Correlation between calmodulin activity and gravitropic sensitivity in primary roots of maize. Plant Physiol 84:1337–1342. Takahashi H, Scott TK. 1993. Intensity of hydrostimulation for the induction of root hydrotropism and its sensing by the root cap. Plant Cell Environ 16:99–103. Takano M, Takahashi H, Hirasawa T, Suge H. 1995. Hydrotropism in roots: sensing of a gradient in water potential by the root cap. Planta 197:410–413. Tsugeki R, Fedoroff NV. 1999. Genetic ablation of root cap cells in Arabidopsis. Proc Natl Acad Sci USA 96:12941–12946. Vermeer J, McCully ME. 1982. The rhizosphere in Zea: new insight into its structure and development. Planta 156:45–61. Volkmann D. 1974. Amyloplasten und Endomembranen: das Geoperzeptionssystem der Prima¨rwurzel. Protoplasma 79:159–183. Volkmann D, Balusˇ ka F. 1999. Actin cytoskeleton in plants: from transport networks to signalling networks. Microsc Res Technique 47:135–154. Volkmann D, Czaja AWP. 1981. Reversible inhibition of secretion in root cap cells of cress after treatment with cytochalasin B. Exp Cell Res 135:229–236.

47 Volkmann D, Sievers A. 1975. Wirkung der Inversion auf die Anordnung des Endoplasmatischen Reticulums und die Polarita¨t von Statocyten in Wurzeln von Lepidium sativum. Planta 127:11–19. Volkmann D, Sievers A. 1979. Graviperception in multicellular organs. In: Haupt W, Feinleib ME, eds. Encyclopedia of Plant Physiology, New Series. Vol. 7, Physiology of Movements. Berlin; Springer-Verlag, pp 573–600. Volkmann D, Behrens HM, Sievers A. 1986. Development and gravity sensing of cress roots under microgravity. Naturwissenschaften 73:438–441. Volkmann D, Buchen B, Hejnowicz Z, Tewinkel M, Sievers A. 1991. Oriented movement of statoliths studied in a reduced gravitational field during parabolic flights of rockets. Planta 185:153–161. Von Guttenberg H. 1968. Der Prima¨re Bau der Angiospermenwurzel. In: Linsbauer K, ed. Handbuch der Pflanzenanatomie, 2nd ed., Vol. VIII, Part 5. Berlin; Borntraeger. Weise S, Kiss JZ. 1999. Gravitropism of inflorescence stems in starch-deficient mutants of Arabidopsis. Int J Plant Sci 160:521–527. Wendt M, Sievers A. 1986. Restitution of polarity in statocytes from centrifuged roots. Plant Cell Environ 9:17– 23. Wendt M, Sievers A. 1989. The polarity of statocytes and the gravisensitivity of roots are dependent on the concentration of calcium in statocytes. Plant Cell Physiol 30:929–932. White RG, Sack FD. 1990. Actin microfilaments in presumptive statocytes of root caps and coleoptiles. Am J Bot 77:17–26. Wilkins MB. 1984. Gravitropism. In: Wilkins MB, ed. Advanced Plant Physiology. London: Pitman, pp 163–185. Wraith JM, Wright CK. 1998. Soil water and root growth. Hort Sci 33:951–959. Zieschang HE, Koehler K, Sievers A. 1993. Changing proton concentrations at the surfaces of gravistimulated Phleum roots. Planta 190:546–554.

4 Cellular Patterning in Root Meristems: Its Origins and Significance Peter W. Barlow IACR–Long Ashton Research Station, University of Bristol, Long Ashton, Bristol, England

I.

INTRODUCTION

directly on root systems (e.g., Fitter, 1987), whereas others argue that the unit of selection is the whole plant (e.g., Peterson, 1992) on the grounds that it is the whole organism, rather than any particular organ, whose fitness is tested by selective pressures. However, it can also be argued that, on the one hand, populations are the subject of selection or, on the other hand, that it is the hereditary material (genes) which is selected. All arguments have some validity, particularly within the context of a hierarchically organized plant kingdom in which the levels of organization range from the molecular to the individual and the population (Barlow, 1999). Nevertheless, it cannot be denied that the root, like any other organ, contributes to fitness and, thus, is one of the elements of the plant which is offered for natural selection. Likewise, the cellular construction of roots also features somewhere in this general problem of fitness and plant evolution. Plant forms evolve as a result of an interaction between the energy levels inherent to the biotic and abiotic components of the total environment. Roots and their cells have undergone selection not only to resist potentially fatal encounters with the environment (e.g., heavy metals, drought), but also to serve the needs of the whole organism, or organ, of which they are, respectively, a part. Thus, they have characteristics that are relevant to both the evolutionary and functional aspects of plant biology. Each plant form is a

Roots are an integral component of nearly all plants. Where they are absent, as in certain water plants (Ceratophyllum, Salvinia, Utricularia, and Wolffia spp., for example), this is probably the result of evolutionary loss with the usual root function being performed by other organs. This last statement points to two important areas of biological enquiry which have been differentiated as evolutionary and functional biology (Mayr, 1961). In plants, evolutionary biology deals with problems of population genetics, ecology, and palaeontology, whereas functional biology includes the disciplines of anatomy, biochemistry, and embryology. In the present discussion of cellular patterning within root tissues, it is useful to keep these two avenues of enquiry in mind, for they should lead to an understanding not only of the advantages that cellular patterns confer in enhancing the fitness of the individual and hence of the population of which the individual is a member, but also the means by which such patterns are achieved. However, in considering the evolutionary significance of the cellular structure of roots, the question arises as to how such a property relates to natural selection. The units of living matter on which selection acts are topics of debate in the context of the neo-Darwinian interpretation of evolution. Some scholars suggest that natural selection acts 49

50

Barlow

member of a set of archetypes, these being particular energetically favourable biological constructions (cf. Hill, 1990). Certain archetypes might be favoured by natural selection on account of their positive contribution to fitness in a given environment. Examples, at the organism level, are the architectural ‘‘models’’ proposed for the shoot systems of tropical trees by Halle´ et al. (1978) and the rather similar models for root systems adopted by Jenı´ k (1978). Likewise, at the level of individual organs, the forms of shoots and roots represent another set of archetypes. Comparison of the forms of terrestrial roots with, for example, those of the river-dwelling members of the Podostemaceae, the cellular patterns of whose curious thalloid roots were described by Schnell (1967), illustrates the range of rhizogenetic possibilities that have been selected by the environment (Barlow, 1986, 1994a). By the same token, the characteristic pattern of cells that construct organs have also been determined by particular sets of conditions which reside at the various levels of organization whose processes impinge upon the level of the cell (Barlow, 1993, Fig. 1). These conditions include those which are inherent to (1) the cells themselves, (2) the organ to which the cells belong, and (3) the plant and its external environment. In this respect, it would not be surprising if the cellular patterns of roots and other organs also comprised a set of archetypes (Lu¨ck and Lu¨ck, 1986, 1993).

II.

SIGNIFICANCE OF CELLULAR PATTERNS

From time to time the role of cells in the development of plant organs and plant form is reappraised (Sinnott, 1960; Kaplan and Hagemann, 1991; Sitte, 1992; Barlow, 1994b). Some of the persistent concern surrounding this topic is a response to observations on the morphogenetic processes of animals. Here, cell migrations not only establish the sites of stem cells, which are responsible for tissue renewal, but also help generate the form of the growing embryo. Plant development, by contrast, is accompanied by much less movement of cells relative to one another: examples where this regularly occurs are in the processes of fertilization and intrusive growth of fibers in secondary tissues. However, the location of stem cells is a crucial feature of plant morphogenesis (Barlow, 1995a). Observations on the germination and growth of -irradiated seedlings, where cell divisions are largely abolished, are often mentioned as supporting the idea that the form of a plant organ (e.g., a root) develops with

only the minimal involvement of cell division (Haber, 1968). However, this evidence is relatively weak because both the cells comprising the organ and the polarity of cell and organ growth were formed prior to the experimental irradiation event. Since few or no cells are formed postirradiation, this system reveals only that the growth and polarity of preformed cells are not disturbed by irradiation. Nevertheless, in primary roots of -irradiated maize plantlets, it is possible to find that the cells of the quiescent center are provoked into division by the -rays (Barlow, unpublished), just as in the case of young, growing root apices irradiated with x-rays (Clowes, 1964). Other arguments against the necessity of cells for morphogenesis rely on examples from marine plants (Sinnott, 1960). Here, complex forms can be developed by organisms which are coenocytes, where one giant cell may contain thousands of nuclei (Jacobs, 1994). The aqueous environment is one in which plants are free-floating and buoyant. As a result, tissues and organs experience little of their own mass. On land, however, plant tissues and organs do experience their mass and also the forces exerted on them by adjoining organs. Furthermore, once plants grow away from the relatively calm conditions of the soil–air interface in which their seeds germinated, their shoots are subject to buffetting by the aerial environment and their roots experience the impedance of the medium in which they grow. In both these environments, the interpolation of internal mechanical struts and baffles helps counteract these externally applied physical forces which might otherwise compromise the integrity of the organism (Niklas, 1989). Cell walls provide some of the requisite resistance to these forces. The protoplasts also develop a force (osmotic pressure) which augments the support of the enclosing walls and contributes some rigidity to the organ (this is significant only when wall thickness is < 20% the radius of the cell). The processes of mitosis and cytokinesis which take place within a root meristem may therefore be viewed as a means of establishing a network of walls whose pattern has been optimized for its mechanical, force-resisting properties. Moreover, walls differentiate in specialized ways, especially in cells located deep within the interior of the



Sometimes, multinucleate cells of plant tissues are incorrectly described as ‘‘syncytia’’ (e.g., Olsen et al., 1995) when in fact they are coenocytes. There are two types of multinucleate cell. Multinucleate syncytial cells arise from cell fusion events, whereas coenocytic cells, such as those of algae and higher-plant endosperm, arise from a series of nuclear divisions within one cytoplasm.

Cellular Patterning in Root Meristems

organ whereby they contribute to the construction of conductive elements. The significance of the wall network is highlighted in dicots by the derivatives of the cambium, where secondary xylem and phloem simultaneously provide support as well as solute conduction. That is, the wall network confers both structure and function on the tissues. External boundary walls sometimes facilitate organ fusions. A further advantage of walls is that they can neutralize the pressures that might otherwise be exerted on the protoplast and which, unless attenuated, could lead to undesirable physiological disturbances of a thigmomorphogenic nature (Jaffe, 1985). An optimal wall network provides this attenuation. It also permits processes with low mass-sensing thresholds, originating within the protoplast, to initiate responses such as gravitropism (Barlow, 1995b). Thigmomorphogenesis and gravimorphogenesis require higher thresholds for their initiation and use the walls to transduce their respective stimuli (touch and mass perception) to the interior of the cell. As mentioned, initiation of gravitropism requires a low mass-sensing threshold. Although the walls are not directly used for gravity

51

perception, the walls in the zone of bending may, nevertheless, limit the intensity of the graviresponse. Division walls which limit contact between daughter cytoplasms may enable the differentiation of domains, which then function autonomously as tissues. Within these domains, particular metabolic processes can be elaborated without interference from processes in neighboring domains. Histological studies indicate that tissue differentiation begins within the meristematic region of the root and that tissue maturation continues to completion in the older zones behind the meristem. In the youngest zones of a meristem, the boundaries which demarcate tissues of stele, cortex, epidermis, and so on, are often already prefigured in the arrangement of the cell walls (Fig. 1). Presumably, the autonomy of each tissue becomes increasingly evident as (1) the number of cytoplasmic contacts (via plasmodesmata) across the longitudinal walls of the cell files diminishes (Juniper and Barlow, 1969; Cooke et al., 1996), and (2) the specific chemistry of the cell walls and the organization of the cytoplasm, by which tissues are recognized in histological studies, develop to their final states (Barlow, 1982). It is even

Figure 1 Three types of root apex construction based on cellular and histological patterns. (A) Closed type of meristem (Zea mays), where the tissues of cap, stele, and cortex/epidermis are discrete (410). (B) Open type of meristem (Vicia faba), where the tissues of cap, stele, and cortex appear to be in continuity (240). (C) Gymnosperm (Pinus sylvestris), where the stele is separate from the other tissues of the apex. Cap initials are sited upon its distal border (compare this with A, where the cap initials are sited upon the border of the epidermis) (320). (Modified from Clowes, 1959; with permission of Cambridge University Press.)

52

Barlow

possible that the cells themselves can regulate their plasmodesmatal permeability and hence construct physiological domains (Mezitt and Lucas, 1996). Thus, dilution by wall extension need not be the only way to modulate the degree of intercellular communication via plasmodesmata; plasmodesmata can perhaps be sealed and rendered inoperative and thus contribute to the establishment of positional information (PI). In contrast, it seems that secondary plasmodesmata can be inserted into walls (Seagull, 1983), at sites where the demand for rapid fluxes of solutes is likely to be great. III.

CELLULAR PATTERNS THAT ESTABLISH THE ROOT APEX

A.

The Embryo

An embryo develops within a polarized embryo sac and, in time, develops its own polarity. This then enables a root to form at the basal (micropylar) end of the proembryo. On the basis of recent research, it seems probable that not only does the internal organization and polarization of the zygotic cell’s cytoplasm have a profound influence on the initiation of embryonic organs, but so, too, do its boundary wall and membrane (Mordhorst et al., 1997; Vroemen et al., 1999). In angiosperms, the zygotic cell initially has a thicker wall at its micropylar end, whereas its opposite, chalazal end is usually wall-less and bounded only by a plasma membrane (Natesh and Rau, 1984). A particular cytoplasmic domain, together with specific wall or membrane properties at the micropylar region of the zygotic cell, determine that the cells which form here develop as suspensor. Conversely, conditions at the chalazal end may be conducive to embryonic development only, and the eventual development of a wall here, which finally encloses the embryo, may also be needed to maintain embryonic polarity. Indeed, Pennell et al. (1991) have shown that, in embryos of rape (Brassica napus), the plasma membrane of cells derived from the basal cell (cb) of the two-celled zygote have a positive reaction to the monoclonal antibody, JIM8, which recognizes an arabinogalactan protein epitope, whereas derivatives of the apical cell (ca) are unreactive. When a carrot cell with embryogenic potential divides, the daughter cell which becomes the embryonic initial can no longer be marked by JIM8, whereas its sister cell continues to display this protein (McCabe et al., 1997). A characteristic cell wall composition and cytoplasmic status are thus important for initiating the orderly development of both zygotic

embryos (Lu et al., 1996) as well as adventive embryoids (Williams and Maheswaran, 1986). That certain types of wall constituents actively promote cell division (Binns et al., 1987; Kreuger and van Holst, 1995; Toonen et al., 1997) may also be relevant in this respect. Patterns of division associated with early embryogenesis vary according to the species. Six basic embryogenic patterns (or types) have been described (Johansen, 1945; Wardlaw, 1955; Natesh and Rau, 1984). Correlated with these patterns is variation in the origin of root tissues (Fig. 2). One of the major differences is whether or not descendents of the cb, which lies at the micropylar end of the two-celled embryo, contribute to the root. In the onagrad (also called cruciferad), asterad, and solanad types of embryogeny they do so, whereas in the caryophyllad type they do not, and all root tissues arise from descendents of the ca of the two-celled embryo. In many cases, the patterns of early embryonic divisions appear to be regular, and so it is possible to deduce the genealogies of the cells and examine how these comply with the fates of the cells as they differentiate to form the major tissues of the embryonic root. In a number of cases enough of the requisite detail is available for such an analysis, but descriptions in the literature are often incomplete, especially in their three-dimensional detail and in the recording of the frequencies of alternative division sequences. Fortunately, the current interest in the development and genome of Arabidopsis thaliana has resulted in a complete description of the origin of the embryonic root (Ju¨rgens and Mayer, 1994). Arabidopsis conforms to the onagrad type of embryogeny. Numerous mutations have been discovered which affect various aspects of its embryogenesis (Mayer et al., 1991; Ju¨rgens et al., 1997), and many of these suggest that the orientation of new division walls at successive generations within the genealogical descendance are important for the establishment of the correct patterns of embryonic development. From descriptions and illustrations presented by Mayer et al. (1993) and Ju¨rgens and Mayer (1994) for A. thaliana, by Tykarska (1976, 1979) for Brassica napus, and by Swamy and Padmanabhan (1961) for Sphenoclea zeylanica, the genealogy and fate map for roots of three species with an onagrad embryogeny can be proposed. That for Arabidopsis is shown in Fig. 3; that for S. zeylanica was published earlier (Barlow, 1994b). The genealogies deriving from cells ca and cb are similar in all three species except for the number of cells that become suspensor (two cells

Cellular Patterning in Root Meristems

53

Figure 2 Four types of angiosperm embryogeny. (A) Asterad. (B) Onagrad (or cruciferad). (C) Solanad. (D) Caryophyllad. In all four cases, the first division of the zygote is transverse (this division may be unequal [A] or equal [B, C, D]) and divides the cell into an apical cell (stippled) and a basal cell (clear). The root (indicated by r against the relevant tier of cells) may have a different origin in each embryogenic type. A thicker boundary wall encloses the embryo, the remaining group of cells is assigned to the suspensor. (From Sporne, 1974; with permission of Chapman and Hall.)

comprise the suspensor in Sphenoclea, eight or nine in Arabidopsis, up to 14 in Brassica). An important feature to notice in Arabidopsis is that among the descendants of cb, there are four cells descended from cell d1 which initiate the calyptrogen, and four cells which descend from cell h2 constitute a quiescent center (QC), the cells of which become proliferatively inert (Dolan et al., 1993). However, the QC of Petunia hybrida, with solanad embryogeny, has a different origin (Fig. 4). It is descended from ca (Vallade, 1970) via cell l00 6:1. The cap tissue of the root, along with the suspensor, are both descended from cb (Vallade, 1972).

It is possible that, in Petunia, additional cells also become quiescent later in embryogeny and thus increase the size of the QC. The ‘‘adding’’ of cells to the QC was also proposed by Clowes (1958, 1978). These results suggest that the QC could be composed of two parts. One is a ‘‘constitutive’’ group of cells directly formed as a consequence of cell genealogy (Q cells in Figs. 3 and 4), the other is a ‘‘facultative’’ group which forms later in root development as a response to PI. The facultative QC cells in Petunia, which may also descend from cell l00 6:1, serve as the most distal cells of the histogenetic

54

Barlow Figure 3 Genealogies of cells, and an indication of their eventual fates as tissues, during early embryogenesis of Arabidopsis thaliana (onagrad embryogeny). The zygote, z, divides transversely to give equally sized apical and basal cells, ca and cb, respectively. The embryo proper descends from ca which then divides longitudinally to give ql.l and q1.2; descendents of one of these two daughter cells (ql.l) are shown. Each cell continues to divide (although cells contributing to the suspensor undergo a limited number of divisions); their descendants are designated by various letters and numbers. The cell types and future tissues to which the cells will belong are as follows: su1-7, suspensor (a file of 7 cells); c, cap; Q, quiescent center; E, epidermis: Co, cortex; St, stele. These last three tissues become located within either root (R), hypocotyl (H), or shoot (S). Divisions may be of three types with reference to the long axis or to the surface of the embryo: transversal ( ), radial ( ), or periclinal ( ). A radial division may be at right angles to the direction of the previous one; this rotation of the division plane is indicated by the arrow ( ). The genealogy is based on results of investigators mentioned in the text (see also Ju¨rgens and Mayer, 1994).

plerome and periblem. They would be equivalent to the functional initials proposed by Barlow (1994c). Direct evidence for two types of cells in the QC is given by the reponse of root apices to feeding with 0.1 mM ascorbic acid solution for up to 64 h (Innocenti et al., 1990). Within 28 h, the outer, facultative cells of the QC had reentered the DNA-synthetic (S) phase of the mitotic cycle; the innermost, constitutive QC group remained held in the G1 phase of the cycle. In adventitious roots of Allium cepa, the number of these constitutive QC cells comprises

about 10% of the whole QC. However, using primary roots of Zea mays, Kerk and Feldman (1995) found that a 48-h exposure to 0.1 mM ascorbic acid induced all cells of the QC to enter S phase. As indicated in Figs. 3 and 4, the patterns of division and the lineages created in the onagrad and solanad embryogenies relate to the subsequent fates of the cells. The caryophyllad embryogeny may present yet another pathway for the origin of cap and QC. This pathway seems to be evident during embryogeny of apple (Malus pumilla cv. McIntosh) studied by Meyer (1958) (although he classified the embryogeny as solanad). Here, the cap and QC develop in the heart of the embryo and are derived from the fourth tier of embryonic cells (Fig. 5). This tier of cells descends from the ca. The supensor of apple embryos is particularly well developed. Its cells are derived mainly from cb but there is also a small contribution to it by cells descended from ca. A contrasting pattern of cell division vis-a`-vis tissue differentiation is found in embryos of graminaceous species—in Poa annua, for example, whose embryogeny conforms to the asterad type. Although less obvious than the examples mentioned above, the cellular pattern in embryos of P. annua also anticipates the pattern of tissue differentiation (Fig. 6). Root and cap tissues become evident when a thicker cell wall appears between these two areas (Fig. 6D,E). This wall was formed at the second division of the embryo. The base of the root seems to be formed from walls

Cellular Patterning in Root Meristems

55 Figure 4 Genealogies of cells and an indication of their eventual fates as tissues during early embryogenesis of Petunia hybrida (solanad embryogeny). The conventions used to depict this genealogy and the cell fates are similar to those shown in Fig. 3, although some of the letters used are different. The division of the zygote, z, results in a smaller ca cell and a larger cb cell (cf. Fig. 2). In the present scheme, however, the basal daughter cell (l 0 1:1) of the transverse division of ca gives rise to the root and hypocotyl and the apical daughter (l l.l) gives rise to the shoot (these descendants of l1.1 are not shown). Only one of the four clones of cells arising from cells l 00 1:1 and l 0 2:1 are followed in detail; the other three ðl 0 3:2–l 0 3:4) develop similarly. One of the descendents, l00 6:1, gives rise to a quiescent centre (Q) cell. There will be a total of four Q cells by the ninth division. The basal cell, cb, gives rise to suspensor (via cells cl, n, n 0 , m and f) and root cap (via cell d2). The cap can be further subdivided into central cap (cc) and cap flank (cf). (Based on observations of Vallade, 1970, 1972.)

laid down at the fifth division of the proembryo. Careful analysis was also given to embryos of Triticum aestivum by Batygina (1969). When the division sequence depicted by this author is closely scrutinized, it seems that, as is evident in P. annua, certain division walls consistently partition the embryo into regions which later correspond to various organs (coleorhiza, root, shoot) and tissues (epiblast, scutellum). Although Guignard (1961) assembled much useful comparative information for embryogenesis within

the Gramineae, the embryonic division patterns have not been studied with the attention they merit. Because of the ‘‘difficult’’ nature of their embryonic cell patterns, the Gramineae would be a useful system in which to apply molecular techniques for defining the extent of embryonic root tissue, such as Scheres et al. (1994) were able to do for germinating Arabidopsis embryos. Transgenic plants were used in which a marker gene was expressed upon the excision of a transposon. Excision occured at random and the gene product, highlighted by a coloured histochemical reaction for glucuronidase, distinguished clones of cells in which the gene had been activated. In this way, the proximal limit of hypocotylar tissue was identified. It coincided with the anatomical boundary between the shoot and the hypocotyl and, following the embryogenic scheme in Fig. 3, would have coincided with the wall which divided cell l1.1 from cell l 0 1:1 following the division of cell q2.1. Shorter stained segments of tissue corresponded to clones of cells derived from later divisions and these helped to pinpoint the divisions that delimited hypocotyl and root; two of these divisions were in the transversely dividing descendants of cells l 0 2:1 and l0 2:1 The longest segments, resulting from early excision events, ran from root tip (excluding cap and QC) into the cotyledons and marked a clone of cells whose distal boundary was probably the first division wall, between ca and cb. In certain species there is a free nuclear stage during early embryogenesis. Most gymnosperms (Roy Chowdhury, 1962; Singh, 1978), as well as Paeonia spp. (Cave et al., 1961) among the angiosperms, provide examples of this coenocytic state. Even embryoids

56

Barlow

Figure 5 Embryogenesis of apple (Malus pumila cv. McIntosh). (A, B) Early division of the zygote creates two unequally sized cells, ca and cb, and then four tiers of embryonic cells (1, 2, 3, and 4) are formed; they surmount the suspensor (su). (C–H) The four tiers can be recognized throughout embryogeny and are delimited by walls which have been more heavily inked. The root cap, quiescent center, and some of the root develop from tier 4. (From Meyer, 1958; with permission of American Journal of Botany.)

of carrot seem occasionally to be initiated following a multinucleate stage (Steward et al., 1958). Nuclear divisions occur without new walls being inserted between the daughter nuclei, as usually occurs at cytokinesis. In the gymnosperms, the walls may be inserted following the first two or three nuclear divisions (Cupressaceae) or after hundreds of nuclei have been produced (cycads), as is also the case for Paeonia. However, a coenocytic zygote need not be unstructured (Owens et al., 1995). Plant cell bodies, each one of which is composed of nuclear chromatin enclosed within a membrane-plus-cytoskeleton-organizing apparatus (Balus˘ ka et al., 1998), may well direct themselves to specific locations within the coenocytic chamber. From there, each cell body governs a certain cytoplasmic domain. In such circumstances, therefore, nuclear genealogies, or cytoplasmic-domain genealogies, could substitute for cellular genealogies in the

formalization of embryogeny. In gymnosperm and Paeonia embryos, once cellularization has occurred, tissue differentiation begins. Again, PI within the embryo, coupled with an interpretation of embryonic polarity, are probably important for this differentiation. Moreover, if nuclei have been associated with particular domains within the coenocytic cytoplasm which can impart histogenetic imprinting (Lyko and Para, 1999), the effects on differentiation might become manifest during the subsequent embryonic development. The simplest view of embryonic tissue differentiation with respect to cellular patterns is that, in all embryogenic types, the patterns of gene activity which establish tissue identity are specified relatively late, perhaps at the late globular stage, but certainly after many tens of cells have been generated. In all cases, differentiation is the result of the specifications

Cellular Patterning in Root Meristems

57

Figure 6 Meristem differentiation within embryos of Poa annua. (A-C) Third, fourth, and fifth division walls (as judged from a median longitudinal section in each case) are in place. Successively formed walls are labelled 1, 2, 3, etc. (D, E) At later stages of development, it becomes clear that the root is delimited by walls 1 and 5. Wall 2 separates root proper from root cap. Wall 3 approximately bisects the root longitudinally. The internal wall 5 similarly bisects the root cap. The root cap initials (identified by dots) are now evident. Another set of division walls (internal walls from the fourth division of the basal end of the embryo; not labelled, but drawn thicker in E) defines the outer wall of the root. (Modified from Guignard, 1961, who had reproduced drawings prepared by R Soue`ges, 1924.)

of PI in three planes, parallel to the embryonic axis (axial plane), and in two orthogonal directions at right angles to it (radial and tangential planes) (see Liu et al., 1993). In onagrad embryogeny, say, a specific pattern of differentiation appears to be superimposed upon the pattern of cell files, and there may be a

similar colocalization of these features in asterad embryogeny. However, separation of the division and differentiation processes could have adaptive significance because chance variations in the development of the cell lineage would have less serious consequences for tissue differentiation. Such a modus operandi

58

would be characteristic of a regulative type of embryogenesis. But tending to argue against this are findings from Hypericum spp. where defects in the cell division program lead to embryo abortion (Bugnicourt, 1983) and thus suggest some type of lineage-dependent cell differentiation (mosaic, or determinate, type of embryogenesis). In these Hypericum spp., many of the aberrant divisions were found in the hypophysis, signifying an important regulatory role for this group of cells in normal embryogenesis. A comparative study of root meristem formation in zygotic and microspore-derived embryos of Brassica napus revealed that when the hypophysis failed to form correctly (as was the case in the microsporic embryos), the root then failed to grow (Yeung et al., 1996). This failure was associated with the absence of structural initials at the pole of the root. The systematic search for root meristem mutants in Arabidopsis has clarified not only these earlier observations of Bugnicourt (1983) but also sheds light on the conditions that initiate the hypophysis and all cells and tissues which are derived from it. For example, the HOBBIT (HBT) gene is required for the correct development of the hypophysis; in strong hbt mutants both the QC and cap columella are lacking (Willemsen et al., 1998). This effect was not confined to embryos; similar phenotypes relating to cap and QC were found in adventitious roots formed from callus. Additional hypophyseal genes, such as ORC and GREMLIN (Scheres et al., 1996), have been identified in Arabidopsis, but their effects remain to be examined in detail. Other evidence against a strict mosaic, determinative type of embryogenesis is that suspensor cells, which usually do not proliferate, can develop into additional proembryos when growth at the embryonic pole is caused to fail (Haccius, 1978). This can also occur during normal embryogenesis (Masand, 1963). A number of mutants of Arabidopsis also show this feature, notably the mutations at the TWIN locus (Vernon and Meinke, 1994), where it is associated with abnormalities in the primary embryo. Because the embryo develops in the nonrestrictive space of the former embryo sac, the cell patterns associated with the early stages of embryogenesis probably have significance for the internal mechanical stability of the embryo only (see Thomson and Hull, 1934). Any more sophisticated physical property of the wall pattern arising out of the polarity of the embryo and its cells is held in readiness for when the radicle begins to penetrate the soil. An interesting question is whether the division program in the embryo runs without any regulation by the external environment; such would

Barlow

accord with the suggestion of Lintilhac (1974) that the embryo sac is a compression-free space within the ovule and, hence, division patterns are regulated internally. However, in the fern Phlebodium aureum, Ward and Wetmore (1954) showed that relief of stresses imposed by prothallial tissues on the developing embryo disturbed the pattern of divisions during early embryogeny, suggesting that in this case the normal course of development did depend on external mechanical influences. The distinction between the precise, even stereotyped, embryonic cell division patterns of Petunia hybrida (solanad embryogeny) and the more irregular pattern of Vitis vinifera (asterad embryogeny), at least in the early embryogenic stages (Fig. 7A,B), was remarked upon by Vallade (1989). He suggested that

Figure 7 Contrasting cellular patterns in embryos of (A) Petunia hybrida (cf. Fig. 4) and (B) Vitis vinifera. Although the number of cells is different in the two cases, the disposition of cells appears less ordered in B. (From Vallade, 1989; with permission.) (C, D) A similar pair of embryos (C, Brassica napus, and D, Gossypium hirsutum) with contrastingly ordered cellular patterns. (From Tykarska, 1979, and Reeves and Beasley, 1935.)

Cellular Patterning in Root Meristems

the difference between the two embryogenic patterns had its basis in the relationship between embryonic volume growth and the rate of cell proliferation. In Petunia—and the same seemed to hold for Arabidopsis and Brassica (Fig. 7C)—the embryo grew little in size during the early proliferative stages, whereas in Vitis there was relatively more volumetric growth (Fig. 8). A similar relationship is evident in the early embryogeneses of Capsella bursa-pastoris and Gossypium hirsutum (Fig. 8). The Gossypium embryo was initially larger and had a less regular pattern of cell division (Fig. 7D) than did the smaller embryo of Capsella. From this point of view, it may be that the cellular patterns are the outcome of adherence to Errera’s rule for cell division—i.e., that new division walls should be of minimal area (Errera, 1886; see also Steward, 1958; Korn and Spalding, 1973). Given this rule, together with an additional rule defining where the new division wall will be attached (e.g., Hofmeister’s rule of equal volume partitioning of the

Figure 8 Relationship between growth of the embryo (log10 volume,
m3 ) and cell number (plotted as log2 values, but with the actual values also inserted on the inside of the ordinate for reference). Data points are mean values of each variable for five species: A, Arabidopsis thaliana (from Mansfield and Briarty, 1991); B, Brassica napus (from Tykarska, 1976); G, Gossypium hirsutum (cell numbers estimated from Pollock and Jensen, 1964); P, Petunia hybrida; and V, Vitis vinifera (data for these last two species from Vallade, 1989). Note that embryonic volume is greater per cell number, especially at the 4- to 16-cell stage, in species V and G than it is in species A, B, or P. This may enable alternative orientations of future divisions (cf. Fig. 7).

59

dividing cell), it is possible to arrive at predictable and invariant cellular patterns, such as are evident in Petunia and Brassica. In Vitis and Gossypium, on the other hand, the division patterns are more variable simply because, owing to their more rapid embryonic growth, there are alternative positions at which the new division walls can be inserted without the two mentioned division rules being violated. The relationship between the rates of growth and division of the various embryogenic cellular groups may be influenced by the physicochemical properties of the boundary wall of the embryo (Ward and Wetmore, 1954). This could explain some of the irregular embryonic division patterns found in certain mutants of A. thaliana. The effects of gnom (Mayer et al., 1993) and fass (Torres-Ruiz and Ju¨rgens, 1994) were evident at the earliest stages of embryogenesis; here, the usual asymmetry of the first zygotic division was upset. Another interesting mutant is monopteros (Berleth and Ju¨rgens, 1993), in which the divisions of cells descended from h2 and d2.1 and d 0 2:1 (and their homologs in the other half of the root) were switched from their usual longitudinal orientation to a transverse orientation (see Fig. 3). The result was that a column of cells penetrated the basal, or root, end of the embryo, and that the cap and QC failed to develop normally. Cell lineages have been dismissed as irrelevant to plant tissue differentiation (Dawe and Freeling, 1991; Irish, 1991, 1993). But this conclusion was derived in relation to plant structures formed late in development, even though these structures could have already been determined by a lineage-based differentiation system at an earlier stage. The last-mentioned system may operate in embryos, where the structural plan for the future adult tissues is being established. Many of the bifurcations in the embryogenic cell genealogies of Figs. 3 and 4 correspond to quantal mitoses (Holtzer et al., 1975) in which there is supposed to be a partitioning or segregation of differentiated cytoplasmic domains and areas of preexisting cell wall. Both of these features could have the potential to influence the course of subsequent differentiation in the dissimilar daughter cells. Another possibility is that, at certain times during embryogeny, epigenetic modifications occur in the nuclei of sister cells. This would also manifest as a quantal mitotic event. Both situations would lead to the establishment of cell lineages with unique and heritable characteristics. It is known, for instance, that nuclei can inherit specific proteins through a succession of mitoses (Raff et al., 1994) and, moreover, within a growing root, cells of

60

each tissue have characteristic patterns of nuclear chromatin (Barlow et al., 1982; Balus˘ ka, 1990). Thus, each of the four histogens of the root, as well as the QC, may, in the process of their establishment, have come to possess (and then maintain) different cytoplasmic complements and/or epigenetic modifications. The transposon excision in Arabidopsis embryos mentioned earlier (Scheres et al., 1994) is one such epigenetic event, except that it occurs at random (or so it was assumed). A more usual type of epigenetic change associated with tissue differentiation is the methylation of the cytosine bases of nuclear DNA. Methylation results in nuclei which vary in the amounts of 5-methylcytosine (Holliday, 1987; Cedar and Razin, 1990). During the development of somatic carrot embryoids from a progenitor cell type, for example, the degree of DNA methylation increases (LoSchiavo et al., 1989; Munksgaard et al., 1995). So far, this rather general finding correlates neither with the changing spectrum of cellular proteins known to occur concurrently in this system (Racusen and Schiavone, 1988) nor with the development of specific tissues. However, LoSchiavo et al. (1989) found that a small number of embryoids were able to develop in the presence of the hypomethylating drug 2-amino5ethoxy-carbonyl-pyrimidine (ECP), and that, although hypocotyl and shoot development took place, these embryoids lacked roots. It seems that ECP and the associated low level of DNA methylation affected the establishment of proper root–shoot polarity in the embryoid. Division of a cell with a predetermined asymmetry has been proposed as crucial for the initiation of cortex and endodermis in Arabidopsis roots (Gallagher and Smith, 1997). Evidence gathered from the mutants scarecrow and shortroot showed that the corresponding dominant alleles are regulatory elements in this quantal division process, and are of especial importance for the establishment of endodermis (Fig. 9). This tissue normally originates from the inner daughter cell of a periclinally-divided cortex–endodermis progenitor cell (CEP) and may be derived soon after the division which gives birth to the cells labelled CoR in Fig. 3. The CEP is itself the product of unequal transverse division of a mother cell (an autoreproductive stem cell). The two mentioned mutants permit the unequal transverse division in the stem cell, but not the unequal periclinal division in the CEP. The lack of the periclinal division is hard to understand, unless this particular division orientation is dependent upon a predivisional asymmetry of the CEP. If so, then both mutants fail to support this asymmetry. Interestingly,

Barlow

Figure 9 Scheme for the division of a mother cell that intiates files of cortex and endodermis in roots of wild-type and mutant (scarecrow and shortroot) seedlings of Arabidopsis thaliana. (A) In the wild-type, an autoreproducing stem cell (a) divides unequally at step (i) to produce a cortex-endodermis initial (CEP). At step (ii), the CEP divides periclinally to produce endodermis and cortex cells. The two divisions at steps i and ii continue to increase the number of endodermis and cortex cells; the a cell maintains itself at each division. (B) In the mutants, step ii is abolished. As a result, in scarecrow a chimerical cortical-endodermal daughter cell is produced, while in shortroot the daughter shows no characteristics of endodermis. These mutants indicate that some type of asymmetry of either wall or peripheral cytoplasm becomes established in the a cell and in the CEP. (Modified from Gallagher and Smith, 1997; with permission from Elsevier Science.)

the CEP and its descendents are revealed by immunofluorescence binding of the CCRC-M2 antibody (which recognises a cell wall carbohydrate epitope) to be chimeras for characteristics of both cortex and endodermis (Di Laurenzio et al., 1996). Another type of asymmetric division is that which initiates the differentiation of nonhair and hair cells in the root epidermis. The Arabidopsis mutant werewolf abolishes this asymmetry of cell fate and accordingly all cells become hair cells (Lee and Schiefelbein, 1999). Both the WEREWOLF and SCARECROW proteins are transcription factors. B.

Lateral Roots

The earliest stages of lateral root initiation on a parent root (or shoot) axis (see also Chapter 8 by Lloret and Casero in this volume) appear to trace the same crucial rhizogenetic steps that occurred earlier in the zygotic

Cellular Patterning in Root Meristems

embryo. This must also be true of the pattern of gene activity necessary for imparting identity to the various root tissues which differentiate within a growing lateral root primordium. Whether the patterns of divisions are similar in the two systems is unclear, but presumably each must construct a cellular template which enables rhizogenesis to become self-maintaining (Malamy and Benfey, 1997; Barlow et al., 2000). Rhizogenetic regions within cellular aggregates grown in vitro also share the same steps of cell patterning. Reorientation of cell growth and the establishment of four histogenetic tiers of cells seem to be key events in establishing the precursor of a primordial apex (Tylicki et al., 2000). Because lateral root primordia of higher plants have a multicellular origin, no single pair of cells in the pericycle is directly equivalent to the cell pair ca and cb of the embryonic lineage. It is possible that the new axis of a primordium, oriented in the radial plane with respect to the longitudinal axis of the parent root, could commence with cells that are equivalent to cells h1, q2.1, q2.2, q2.3, and q2.4 (in species with onagrad embryogeny) (see Fig. 3). The four quadrants arising from the q2 cells are identifiable in lateral roots of tomato and primary roots of Brassica napus (Kuras, 1980). Transverse divisions occur in pericycle cells before the periclinal divisions which accompany the

61

outgrowth of the primordium (Lloret et al., 1989; Casero et al., 1993; Malamy and Benfey, 1997) (Fig. 10). They may be needed to provide these small primordial q2.1–q2.4 cells. Also, recognition of these early transverse divisions raises the possibility that synthesis of new proteins required for lateral root initiation may occur earlier than anticipated hitherto (e.g., Keller and Lamb, 1989; Neuteboom et al., 1999). When JIM antibodies were applied to transverse sections of roots of carrot, one of them, JIM4, an anti-arabinogalactan protein, was shown to recognize pericycle cells opposite xylem where lateral root primordia would be expected to form (Knox et al., 1989). In soybean, a hydroxylproline-rich glycoprotein is associated with the xylem arcs (Ye and Varner, 1991), and in Arabidopsis parenchyma cells neighboring the xylem arcs express transcripts of a cyclin A gene which is involved in cell cycle regulation (Burssens et al., 2000). However, in all cases, the relevance of these proteins to primordium development is unknown, but does suggest a degree of preformation of the sites for potential primordia. The conditions leading to the origin of lateral root primordia in angiosperms will no doubt become better known when the problem is examined more closely. In the fern Ceratopteris thalictroides, however, the cellular origin of each lateral root primordium is strikingly

Figure 10 Scheme of division of a pair of elongated pericycle cells in the root of Allium cepa during the initiation of a lateral root primordium. These cells would be located opposite a xylem pole. (A) Nuclei of two contiguous cells move toward each other. (B) The cells divide transversely (with respect to the cell and root axis) and unequally. (C, D) Unequal divisions continue until four cells are present within the walls of each original mother cell. (E) The first periclinal division occurs in a central cell of this group of eight cells. This division represents a change in the orientation of the growth axis which permits a primordium to form. This sequence of divisions, A–E, occurs over a distance of 15 mm, commencing 7 mm behind the tip. Although only two cells are shown, neighboring pericycle cells also undergo some unequal divisions. However, the pair of cells that underwent the first of such divisions has precedence and serves as the focus for primordium development opposite the xylem, but some of the other unequally dividing cells also contribute to this process. (From Casero et al., 1993; with permission of Springer-Verlag and of Prof. P. J. Casero.)

62

clear. A cell in the endodermis enlarges and then immediately undergoes a sequence of four divisions in different planes that thereby create three root sectors and one cap sector (Lachmann, 1907). This sequence of divisions repeats, and the derivatives (merophytes) continue to divide in planes characteristic of the homologous merophytes in the parent root apex. The fivecelled primordium thus has already the cellular pattern of a new root (Fig. 11). The parent root develops in a similar way from superficial cells in stem and petiole (Lachmann, 1907; see also Gunning et al., 1978a, for details of the origin of shootborne roots of Azolla pin-

Barlow

nata). Thus, throughout the ontogeny of the plant, once a lateral root-forming apical cell is created, it immediately commences functioning in a root-specific fashion. According to Vladesco (1935), who worked mainly with embryos of Gymnogramme sulphurea, apical cells arise but do not necessarily persist. Then, a new generation of apical cells comes into existence. It is as though, in this species, a number of ‘‘trial runs’’ of cell patterning are required before a stable pattern of organogensis with an apical cell is instituted. The apical cells that structure the lateral roots of leptosporangiate ferns may be determined to develop

Figure 11 Development of a lateral root primordium in a shootborne root of the water fern, Ceratopteris thalictroides. (A, B) The precursor of an apical cell (a) enlarges in the endodermal layer. (C–F) The apical cell then divides in regular sequence and generates a tetrahedral form. The apical cell and the resulting primordium are here emphasized by thicker boundary walls, but in reality these walls are not so prominent. A–E, Transverse sections; F, longitudinal section. (From Lachmann, 1907.)

Cellular Patterning in Root Meristems

in their characteristic fashion in accordance with both PI and a lineage program (Barlow, 1984, 1989a). PI specifies which endodermal cell lying on the circumference of the stele will develop as an apical cell, whereas lineage specifies which group of cells within a merophyte will respond to PI. Whether such a means of primordium specification applies to angiosperm roots is not known, but observations on the first periclinal divisions in the pericycle of tomato roots shows a strict relationship between the order number of the cells in the primordial row and the vasculature of the parent root (Barlow et al., 2000). Most theories of lateral root formation implicate PI (Torrey, 1986; Vuylsteker et al., 1998) and favor a dependency upon critical levels of cytokinin and auxin (Zhang and Hasenstein, 1999); but lineages may also play a part, in the way suggested by Barlow and Adam (1988). In fact, it is probable that the centre of the primordium lies at the junction of four pericyclic cellular complexes (cf. Fig. 10), where the four oldest walls join to form a cross (Barlow et al., 2000). IV.

CELLULAR PATTERNS IN GROWING ROOT APICES

A.

Patterns of Meristematic Cell Division

Most embryos undergo a period of dormancy or quiescence which interrupts their growth shortly after their organogenetic phase has been completed. Imbibition cancels this interruption and revitalizes the growth process. When the radicle protrudes, the cells in the apical meristem may already be actively dividing. This pattern of activity applies to species with epigeal germination. In species with hypogeal germination, cell division commences after the rootlet has emerged (Schatt et al., 1985; Obroucheva, 1999). In both epigeal and hypogeal species, the onset of mitoses seems dependent upon the presence of a critical number of elongating cells (Obroucheva, 1999). Growth and mitotic division within the germinating radicle commence at different locations and at different times (Pukhal’skaya, 1949). The sequence of activation probably reflects the changing pattern of water availability within the radicle and this can be marked indirectly by soluble radioactive tracers of DNA and RNA synthesis (Stein and Quastler, 1963; Payne et al., 1978). Refined molecular probes now enable temporal and spatial patterns of the onset of various processes to be monitored: the pattern of nuclear DNA synthesis, for example, has been shown to coincide with the re-establishment of the cortical microtubular cytoskeleton (Jing et al.,

63

1999), which is a major regulator of cell wall synthesis. New tubulin synthesis and tubulin gene activation occur concurrently with the onset of DNA replication, though some of the microtubules seen early in germination may be elaborated from a granular form of tubulin stored within the dry seed. The number of dividing cells along the length of the meristem of the germinating radicle is small at first and reflects the size of the meristem at the end of embryogenesis prior to dormancy. This number is increased during germination by new rounds of transverse, proliferative divisions until a meristem of stable size is achieved (Deltour et al., 1989). In Arabidopsis thaliana, the resumption of root meristem activity is controlled by two genes, RML1 and RML2, which map to different chromosomes (Cheng et al., 1995). Mutation at either of these loci results in determinate root growth in which the potentially meristematic cells at the apex become completely differentiated. Usually, all preformed cells of the embryonic meristem are replaced by postformed cells produced by divisions within the emerging rootlet, but in the RML mutants it seems as though the switch between preformed and postformed cell growth and division is blocked or subject to a checkpoint. A similar checkpoint seems to occur naturally in determinate root growth systems, such as those of cactus (Dubrovsky, 1997) and in the nodule roots of Myrica gale (Torrey and Callaham, 1978). In roots with indeterminate growth, where the preformed/postformed checkpoint is passed, the question arises as to how it comes about that there are characteristically different numbers of proliferative cells along the cell files of the various tissues (Luxova´, 1975). It is possible that a repressor of mitotic activity passes into the proliferative cells from the older, nondividing and maturing cells. Some of the latter cells would already have been present within the embryonic radicle. In dicot seedlings, preformed, mitorepressive cells might also be located within the hypocotyl. As the meristem becomes increasingly active at germination, these mature (or maturing) cells are replaced by a new population of mitorepressive cells. Thus, a stable feedback system is set up between the meristem and the maturing compartment (Fig. 12). Other feedback systems have been proposed to regulate division and differentiation in the root apex (Barlow, 1984; Van den Berg et al., 1995) while yet another, governed by the CLAVATA gene, operates upon central and rib meristem cells in the shoot apex (Schoof et al., 2000). Action of the CLAVATA gene should be borne in mind in investigations of the interrelationship between QC and the rest of the meristem. And with regard to

64

Figure 12 Proposed negative feedback between the growing, but nondividing, cellular compartment and the meristematic compartment where successive rounds of cell division occur (indicated by tailed arrows). A putative mitorepressive factor is produced by the nondividing cells and inhibits (angled arrow and ‘‘x’’) division at the basal end of the meristem. Cells so affected enter (straight arrow) the nondividing compartment where they in turn acquire the putative mitorepressive property. As long as the rate of mitorepressive action keeps pace with the rate of cell production, the meristem remains a constant size.

the QC, upon germination a variable proportion of cells in this zone traverse the cell cycle from G1, through S phase, to mitosis (most QC cells in Pisum sativum make this traverse [Jones, 1977], fewer do so in Malva sylvestris [Byrne and Heimsch, 1970a]). The significance of this cell division event within the QC is not clear. However, in Picea glauca, germination brings about a slight but significant alteration to the cellular configuration within the root apex (Yeung et al., 1998). Prior to germination the pole of the Picea root is occupied by a layer of nondividing cells. Following germination, these cells divide once and the daughters then comprise a larger, multilayered group of cells. It is as though the mitotic event ensures the genetic and metabolic activation of these ‘‘preinitial’’ cells, and also finalizes their spatial arrangement as true functional initials. Transgenic Arabidopsis plants with additional cyclin gene DNA (cyc1At) controlled by the cdc2aAT promotor were shown to have enhanced root growth (Doerner et al., 1996). Whether this was due to an increased rate of cycling in all meristematic cells or to an increased number of cycling cells was not revealed. It might have even been a result of the relaxation (due to the additional cyclin genes) of the mitorepressive activity of the new maturing cells (see Fig. 12), just as the RML1 and RML2 phenotypes could be due to an enhancement of this repressive system. Whatever the mechanism, it is possible that the relative abundance of cyclin gene product enables flexible growth control (Doerner et al., 1996). This plastic response might naturally be initiated by the level of

Barlow

available internal auxin or external nutrients. Nitrate (NO 3 ) supply is one such exogenous root growth regulator which can stimulate the rate of root elongation (Zhang et al., 1999) and might achieve this through an effect on meristem size and/or degree of mitotic repression, apparently acting in the same manner as (or as a substitute for) auxin (Poddar et al., 1997). The longitudinal, formative divisions which initiate all the cell files of the root occur within the apex of the meristem (Fig. 1). As already discussed, the pattern of these cell divisions is established during embryogeny; the actively growing meristem then perpetuates this pattern as the root continues its growth. However, as the primary root lengthens, the numbers of cells which initiate cell files can either increase (Hayat, 1963), remain the same, or decrease (Litinskaya, 1993). There can also be some moderate alteration to the overall pattern, as indicated by the change from open to closed meristematic organizations in the roots of certain species of Asteraceae as they grow older (Armstrong and Heimsch, 1976). The same occurs in primary roots of Malva sylvestris as they lengthen from 0.3 cm to 33 cm (Byrne and Heimsch, 1970b). A QC also becomes evident (it is absent until the root is 3–6 cm long, perhaps because of the nuclear DNA synthesis that accompanies germination) and increases in size and cell number (up to 700 cells) as root diameter increases (Byrne and Heimsch, 1970a). There need not be a direct positive correlation between the number of initial cells and the diameter of the root: roots that become thinner as they grow do not necessarily show a corresponding decrease in the number of initial cells; in some cases the number can actually show an increase (e.g., Hayat, 1963; Chiang and Tsou, 1974). In primary roots of Phaseolus radiatus, the initial zone was reported as being four or five cells wide at germination (0 h), while 24–48 h later, three additional cells were present (Fig. 13), indicating the occurrence of approximately one new longitudinal division of each cell in this zone (Chiang and Tsou, 1974). The changing number of initials was also associated with an increasing complexity of the histogenetic program (Fig. 14). During the emergence of lateral roots the number of initial cells may increase (Popham, 1955a) and the roots become thicker. Unfortunately, the corresponding behavior of the QC is unknown in these roots. Conversely, the QC in lateral roots of Vicia faba was found to develop shortly after their emergence (MacLeod and McLachlan, 1974) but, in this case, no information was given concerning the number of initial cells. The relaxation of quiescence during germination might permit the increase in the number

Cellular Patterning in Root Meristems

65

Figure 13 Changing cellular patterns at the apex of a primary root of Phaseolus radiatus during germination and early growth. In the dormant radicle (0 h), there are no mitoses; the preformed central initial cells are marked with open circles. By 12 h, some of these cells have begun to divide (daughter cells are stippled) and initiate a central calyptrogen. Later, other central initial cells divide and contribute to the plerome and thence to the stele. At this time, a set of peri-initials (stippled circles) has also become established around the central group. By 48 h and 120 h, further divisions have occurred and the peri-initials have increased in number. They constitute a periblem and a lateral portion of the calyptrogen, and accordingly contribute to cortex and epidermis and peripheral root cap, respectively. For further detail, see Fig. 14. P marks the pericycle cell layer; the outer wall of P is also drawn thicker. (Modified from Chiang and Tsou, 1974; with permission.)

of initials, but this increase may slow once the QC has become established. Any condition that alters the relative rates of growth and cell division in the three planes of the root (axial, radial, and tangential) will affect the proportions of the respective types of division (Barlow and Adam, 1989). This suggests that selection of division planes is a response to Hofmeister’s rule that cells should divide perpendicular to the plane of maximum growth. Probably the effect on cell shape arose owing to rearrangements of the cytoskeleton in response to altered levels of endogenous growth regulators (Balus˘ ka et al., 1999). Similar effects of growth regulators may also operate in the dividing zones of root meristems, being especially significant in regions where formative divisions are taking place. Both the proliferative and formative postformed cells of the growing apex continue to obey the same division rules as applied earlier during root embryogenesis. As mentioned, these rules relate to the rates and directions

of wall growth of a mother cell (Hofmeister’s rule) as well as to the relative dimensions of the cell (Errera’s rule). All this was realized long ago by Sachs (1887; 447). He believed that the laws of embryonic growth, once established, were perpetuated in the meristems of roots and other organs, and that it is the formative zone of apical meristems that retains these embryonic properties. A further description of the pattern of cell growth within a root apical meristem can be achieved by means of a growth tensor which makes use of the meristem’s natural pattern of anticlinal and periclinal walls in specifying a coordinate system. Two types of tensors have been devised to simulate two types of root apex (Hejnowicz, 1989; Hejnowicz and Hejnowicz, 1991). One of them leads to a root with a QC, the other to a root in which there is an active growth site in place of a QC. The resulting pattern of proliferative activity in the latter case corresponds to that found in fern roots in which there is a mitotically active apical cell. Part of

66

Figure 14 Schematic interpretation of the changing population of initial cells of germinating roots of Phaseolus radiatus and their histogenic derivatives (cf. Fig. 12). (A) In the dormant root (0 h). (B) At the 12th hour of germination. Initials on only one side of the root are represented. (Modified from Chiang and Tsou, 1974; with permission.)

the problem of simulating stable patterns of division within an embryonic apex is to develop a suitable growth field. How this comes about in a real meristem is unknown, but presumably it is regulated by the fluxes of growth regulators which specify the relative amounts of cell wall extension in each plane. The importance of hormone fluxes in maintaining the closed apical division pattern of maize roots is reinforced by the findings of Kerk and Feldman (1994). They fed the auxin transport inhibitor, triiodobenzoic acid, to roots which, as a consequence, showed many breaks in the cap boundary and signs of cell divisions in the QC. Additional layers of cortical cells were also found at the pole of the root. However, these two reported anomalies occur occasionally in untreated roots of maize, the first-mentioned type occurring upon germination of certain cultivars, the last-mentioned during the ageing of primary roots (Barlow and Rathfelder, 1984). Variation in auxin transport may account for interspecific variation in the number of tiers of cortical initials—for example,

Barlow

in primary roots of species within the genus Linum (Byrne and Heimsch, 1968). Interestingly, experimental inhibition of the auxin-transporting system in pine roots resulted in extensive dichotomous and coralloid branching (Kaska et al., 1999). Probably these two types of branching resulted from the activation of growth and division at the lateral edge of the quiescent center. Activation of groups of cells on either side of an inactive zone at the tip of the rhizophore of Selaginella spp. resulted in the formation of two or three triangular apical initials which, in turn, led to di- or trichotomous branching and subterranean root formation (Lu and Jernstedt, 1996). In fact, our unpublished observations from Selaginella suggest an evolutionary rationale for the establishment of a QC in higher plants: in some ancestral plant(s) the onset of quiescence in cells at the apex during root growth promoted root branching from primordia which had, as a consequence of this quiescence, been initiated at the apex. The evolutionary fixation of a permanent quiescent zone in roots of modern-day plants then enabled the changeover from dichotomous to herringbone type of root branching. Despite modifications to the cellular pattern which occur in the roots of some species during the immediate postgermination period, the patterns at this stage are descended largely from the pregermination embryonic patterns. Voronin (1969) described a set of seven cellular patterns in roots from a range of taxa within the pteridophytes, gymnosperms, and angiosperms (Fig. 15). Voronin’s classification, however, was intended to reveal phylogenetic relationships by showing how one pattern could have been transformed into another. The idea of a phylogeny of cellular patterns has been developed more generally by taking into account the number of wall faces (3, 4, 5, etc.) displayed by the structural initial cells from which cell files and tissues ultimately arise (Barlow, 1994c). Differing cellular patterns issue from initial cells with increasing grades of geometrical complexity. Each pattern would constitute an archetype of cellular construction (Lu¨ck and Lu¨ck, 1993). Each archetype might be associated with a distinct energy costs (Barlow, 1994d). Then, if the organ is taken as the reference level, the paraboloidal form of angiosperm root apices (see Barlow and Rathfelder, 1984) would correspond to an archetype of organ construction. The distinction between open and closed meristems has been discussed elsewhere (Clowes, 1981; Barlow, 1995a). It relates to the dynamics of cells at the junction between root and cap tissue. In terms of the

Cellular Patterning in Root Meristems

67

Figure 15 Schematic drawings of cellular patterns in root apices from a range of taxa with indications of the putative evolutionary/morphological transformations (arrows) that might convert one type of apex into another. (A) Marratiaceous fern. (B) Leptosporangiate fern. (C) Gymnosperm. (D) Dicotyledon (open type of meristem). (E) Monocotyledon (closed type of meristem). (F) Dicotyledon (closed type of meristem). Filled arrowheads indicate pericycle; open arrowheads indicate the boundary between cap and root tissue. Small arrows indicate the directions in which new cells are produced by the initial cells, each of which are marked with a dot. (Redrawn from Voronin, 1969.)

growth tensor and the growth field which it specifies, whether a meristem is open or closed depends on the stability of the growth field. When the growth field is stationary, a closed meristem develops, but if it shifts in an axial direction, the meristem does not assume closedness and hence remains open. A shifting growth field can be inferred from certain experimental results (Nakielski and Barlow, 1995). Roots of tomato, mutant at the GIB-1 locus, when cultured in vitro, exhibit cycles of growth and shrinkage of the QC which are also associated with cycles of root cap regeneration (Barlow, 1992). Wild-type tomato roots maintain a stable closed structure. Their growth field is thus inferred to be fixed. In roots of the gib-1 mutant, the QC is active on its acropetal face and a new cap meristem begins to replace the old one (Nakielski and Barlow, 1995); hence, the growth field is inferred to shift axially in a basal direction. The GIB-1 locus of tomato regulates gibberellin biosynthesis, and mutant plants have a lower gibberellin content than wild-type plants. How this relates to the stability of the cellular patterns within their roots is unknown. A shifting

growth field may also account for the change from closed to open root apical organization during germination. Another situation in which the growth field is inferred to shift axially is after excision of the root cap (Nakielski, 1992). The center of the field then moves a small distance from the apex, whereupon the growth field is reestablished in a new site. The original pattern of cell growth and division is eventually regained, but only after it has been redirected to regenerate the cellular pattern associated with the root cap. If slightly more tissue than just cap is removed, subsequent regeneration of the field takes place with the development of a new meristem (or sometimes two new meristems) within the remaining basal portion of the original meristem (Reihman and Rost, 1990). It might be conjectured that the initiation of each new lateral root primordium within the pericycle and endodermis also depends on the periodic redevelopment of a new growth field in these two tissues. The above lines of evidence suggest that the patterns of cell enlargement and division during embryogenesis,

68

Barlow

radicle development, and lateral root primordium formation are the consequence of establishing a particular growth field. These patterns conform to a growth tensor with either a natural or an orthogonal coordinate system (Hejnowicz, 1989). The tensor, though based on actual rates of cell displacement, is, in effect, a representation of a physicochemical system which coordinates growth within the root apex. This may be equivalent to PI specified by hormonal fluxes. The organizing center, with its wandering property, is a point within the apex holding a particular positional value. B.

Cellular Patterns and Tissue Differentiation

One of the main functional aspects of the cellular patterns generated within the root apex relates to the wellknown, but little understood, process of tissue differentiation. All roots conform to a common histological pattern—stele to the interior, epidermis to the exterior, cortex in between, with a cap at the summit. In certain situations, the way in which differentiation proceeds can be modified during the course of root growth. For example, aerial roots of Ficus benghalensis show anatomical changes upon entering the soil, enabling them to be redesignated as terrestrial absorbing roots (Kapil and Rustagi, 1966). Although the same tissues are present in each type of root, there are alterations to the relative numbers of cells in the tissues (e.g., pith cells are plentiful in aerial roots of F. benghalensis but are absent in the terrestrial roots) and to the degree of tissue development (e.g., periderm is well developed in aerial roots whereas endodermis is poorly developed; the converse applies in terrestrial roots). Similar modifications occur within prop roots of Piper auritum as they enter the soil (Greig and Mauseth, 1991). By contrast, rather few quantitative differences were found between the tissues of aerial and subterranean roots of Monstera deliciosa (Hinchee, 1981). Whatever the changes, they occur when one set of external, abiotic environmental conditions is replaced by another during root growth. They are examples of the plastic developmental response of root apices. At another level of organization, different internal conditions impinging on root primordia at the time of their development bring into being dimorphic root systems (Barlow, 1986). Three species of Cyclanthaceae provide examples of dimorphic systems (Wilder and Johansen, 1992). The anatomy of their absorbing and anchoring roots differs in at least 15 quantifiable anatomical characteristics, as well as in more gross mor-

phological features. A second example is the system of air and water roots of Ludwigia peploides (Ellmore, 1981). Here, hormonal conditions at the time of root initiation, which also correlate with the position of the root primordia on the stem, influence which of the two types of root subsequently develops. Presumably, the determining event selects one of a number of alternative combinations of gene activity that accordingly leads to one particular pathway of rhizogenesis (see Barlow, 1994a). The stage of development at which a primordium becomes responsive to determining conditions is not known but, in roots of Convolvulus arvensis, it must occur quite early, as was shown by the experimentally induced switch to shoot development of primordia that would otherwise have formed new lateral roots (Bonnett and Torrey, 1966). As a result of the determining event, young primordia showed different patterns of cell divisions, one pattern being associated with the shoot-forming pathway, the other with root formation. The first group of cells to show differences in the division pattern was in the endodermis of the parent root. In the usual root-forming pathway, these cells divided and contributed to the emerging lateral root, whereas when shoot formation was induced, their contribution to the primordium was much more limited. There were also differences in the way in which cells of the growing primordium of each type made contact with the vascular system of the parent root. Another aspect of cellularity relates to its mechanical properties. The division of cells within a root elaborates a system of turgid protoplasts and semirigid walls which minimizes the force experienced by the root as it grows in mechanically resistant soil. Equally, the cellular pattern might maximize the root’s physiological response to an environment of this type should the walls be capable of transmitting mechanosensory information. When roots meet impenetrable soil, the cells of the cortex swell radially and those of the epidermis stretch tangentially. Together, these features enable the root apex to force apart the resistant soil structure and thereby maintain root elongation. The differential sensitivity of root cells to ethylene (Balus˘ ka et al., 1993) leads to the promotion of radial growth of the cortical cells, similar to the effects occurring in soil. The consequent swelling of the root amplifies their natural wedging ability due to the paraboloidal form of the apex. Whether different arrangements of cells within the cortex affect a root’s response to impedance is not known. For example, would roots with concentric rings of cortical cells arranged in radial rows (see Williams, 1947) be more

Cellular Patterning in Root Meristems

or less efficient in overcoming impedance than roots with less regular radial cortical cell patterns? The ever-present impedance exerted upon roots by soil, with the greatest pressure exerted upon their tips, is probably the reason for root meristems being apical rather than intercalary (as often accompanies shoot growth in monocots), because an apical location minimizes the chance of a root with circular cross section buckling as it grows. A further, mechanical attribute of the cortex is that, under anaerobic conditions, it contributes a system of air channels (aerenchyma), which helps maintain the growth of the root tip. Aerenchymas often develop lysigenously as a result of the selective death of certain cortical cells (Kawai et al., 1998). In extreme types of lysigenous aerenchymas, the cortical cells which remain alive comprise narrow, radial strips of tissue that provide mechanical struts linking the superficial dermal tissue (epidermis plus exodermis) to the vascular cylinder. These struts also channel water and solutes from the dermal tissues to the endodermis. The means by which some but not all cells are selected to die, and how this is regulated in a patterned way, is not known. It is theoretically possible for the pattern to be imprinted in the cortical tissue at the time of the latter’s construction in the meristem (i.e., it is prepatterned). Another way is through more local controls whereby already differentiated cells are protected from dying. For example, within both maize and willow root aerenchymas, living cortical cells are located on radii that extend from the protoxylem poles (Konings and Verschuren, 1980). Some cell-death-limiting influence may therefore emanate from the protoxylem, implying that cortical cells opposite the phloem may be more likely to die. Here, it is interesting to note that the oil reservoirs found in the inner cortex of roots of Solidago canadensis are also located opposite the phloem (Curtis and Lersten, 1990) and that these idioblasts have a schizogenous origin (which is perhaps a mild form of cellular lysis that results in a weakening of contact between cell walls). A third possibility (not incompatible with those previously mentioned) is that, within a region of cortex subjected to cellular lysis, the lytic process commences at random, but once started, it becomes self-regulating, stopping short of completely destroying all the competent cortical cells. One might speculate, therefore, that during the evolution of aerenchyma-bearing plants a fine balance has been achieved in the regulation of cortical cell death to ensure that enough autolysis occurs to provide the necessary number and volume of air channels but not

69

so much as to threaten complete cortical tissue collapse. Observations of Seago et al. (1999) on roots of the aquatic plant Hydrocharis morsus-ranae strongly suggest that the cell death necessary for lysigenous aerenchyma development in the root cortex is correlated with, and may even be regulated by, the cell division pattern (Fig. 16). Two divisions, one radial the other periclinal, produce a quartet of small cells which are destined to die and thereby form an aerenchymatous cavity in the cortex. Other interesting patterns of cell division were also observed by Seago et al. (2000) to be associated with the formation of a schizogenous cortical aerenchyma in roots of Nymphaea odorata (Fig. 17). Both in this last-mentioned species and in H. morsus-ranae, the geometric regularity with which large and small cells are distributed in the cortex is striking and begs for explanation not only in terms of the cell biological questions which they raise (relating to cell division control, positional information, cell separation), but also in relation to the advantages that such precise patterns of aerenchyma formation confer on these water-growing roots. This question of the geometry of cellular packing (square or hexagonal cell perimeters in root cross sections) in relation to waterlogging responses was examined in the root cortex of a number of species (Justin and Armstrong, 1987). A square packing arrangement (where the cells are arranged in radial rows) seems better for aeration of the cortical tissue on account of the relatively more voluminous intercellular space. Secondary tissues are often associated with specialized physical functions within the root. For example, the development of a protective periderm follows from the sloughing of their cortex. But notably, secondary vascular tissue develops from a cambium present within the stele of dicots species. The resulting secondary thickening, involving the differentiation of cambial derivatives as either secondary xylem or secondary phloem, occurs far behind the root apex in a zone where tissues have recently reached maturity (cf. Popham, 1955b). There are parallels (but also some striking differences) between the initiation of both the cambium and the lateral root primordia. One similarity is that, in both cases, the predominant direction of new growth is radial. Also, the first divisions are localized, rather than scattered, within the initiating tissue (Fig. 18). The cambium is derived initially from narrow bands of stelar parenchyma cells. Pericycle cells opposite the protoxylem poles are among the last to be recruited into the dividing cambial cell population. Interestingly, the first ray parenchyma cells to develop

70

Barlow

Figure 16 Cellular patterns associated with lysigenous aerenchyma formation in roots of Hydrocharis morsus-ranae. (A) Within a cross section of the root meristem (0.4 mm from tip of meristem), groups of one, two, or four cells are evident, the number of cells in the group depending on the number of previous periclinal and radial divisions. (B) Three millimeters further back along the meristem, the small cells comprising the groups of four have autolysed (stars). This has given rise to the air spaces. Groups of two larger cells remain (darts). These cells widen radially and form bridges across the aerenchymatous cortex. The first division (as seen in cross section) that led to the groups of two cells was periclinal; both these two cells survive. The first division on the pathway to the groups of four cells was radial and was then followed by a periclinal division; all four cells autolyse. (C) Radial bridges of cortical cells seen in longitudinal section (1.2 mm behind tip of the meristem). Mitoses are still present (darts). Scale bar ¼ 50
m; same magnification for all photos. (From Seago et al., 1999; with permission of the authors and Canadian Journal of Botany.)

are sited opposite the xylem poles (e.g., Loomis and Torrey, 1964; Torrey and Loomis, 1967). In woody species, rays are often initiated by the repeated anticlinal division of an elongated cambial cell. In this respect there is similarity with the early stages of lateral root primordium initiation (see Casero et al., 1993; Chapter 8 by Lloret and Casero in this volume). The new ray initials then divide periclinally. The ray parenchyma becomes conspicuous because the cells retain their cytoplasmic contents. Also, cambial ray cells do not divide frequently, and so they form files of cells greatly elongated in the radial plane (Fig. 19). A final aspect of longitudinal divisions in roots which deserves mention is the development of secondary, or supernumerary, cambia. This can lead to macroscopic features of root development which have both functional, structural, and sometimes eco-

nomic implications. In sugar beet (Beta vulgaris), for example, a number of secondary cambia develop simultaneously in the outer part of the stele (Artschwager, 1926). Their cellular descendants differentiate as xylem and phloem, but many cells remain parenchymatous. The parenchyma fills with sucrose, sometimes to a considerable degree (20% of sugar beet root fresh weight is sucrose), and the root as a whole increases in girth to give the familiar root storage organ. In this, and similar bulbous storage roots, the investment of resources in root growth is considerable and contributes to a strategy of ‘‘anchorage by root mass’’ for such soil-rooted plants (Ennos and Fitter, 1992). Often, a shallow groove runs from base to apex on the surface of roots of beet (Artschwager, 1926). This groove is due to a retardation of cambial growth and cell division opposite one of the two

Cellular Patterning in Root Meristems

71

Figure 17 Aerenchyma developing in the meristematic zone of the root cortex of Nymphaea odorata. Triangular cortical cells (S) divide unequally (open arrows) to produce a small daughter cell. The pattern of division is such that where the short walls of the small cells meet, the intercellular spaces open up by schizogeny. Scale bar ¼ 20
m; same magnification for all photos. (From Seago et al., 2000; with permission of the authors and the Annals of Botany Company.)

Figure 18 The first cell divisions that mark the initiation of vascular cambium in the tap root of Aesculus hippocastanum. Radially oriented groups of newly divided cambial cells, C, can be recognized by their thin cell walls. P is a primary phloem element. Scale bar ¼ 25
m. (Original. By courtesy of Dr. N. J. Chaffey.)

72

Barlow

Figure 19 Vascular cambium (C) and its derivatives (xylem, X, and phloem, P) in a tap root of Aesculus hippocastanum. Rays (R) are cells elongated in the radial plane. Scale bar ¼ 25
m. (Original. By courtesy of Dr. N. J. Chaffey.)

phloem poles. In economic terms, the groove is undesirable as it leads to significant amounts of unwanted soil being harvested along with the root. However, the groove does improve anchorage of the beet in the soil. Such a grooved root may be evolutionarily related to a type of root which, after a few seasons of growth, splits longitudinally, just as do (and using the same mechanism of differential cambial growth, but working over a longer period of time) the trunks of certain trees (e.g., Taxus baccata) and tap roots of Taraxacum koksaghyz (Bulgakov, 1944), and thus aids vegetative propagation.

V. ASYMMETRICAL CELL DIVISIONS Asymmetric formative divisions were discussed earlier (see Section III.A), especially in relation to embryonic root development. However, the production of unequally sized daughter cells by asymmetric division also occurs in the proliferative compartment of roots during the postgermination phase of their growth. In addition to the unequal partitioning of cytoplasm following mitosis, there can also be unequal repackaging of the nuclear proteins within the sister interphase

nuclei (Armstrong and Davidson, 1982). Just as may be the case in formative divisions, an asymmetric proliferative division may be evidence for some asymmetry of cytoplasm (polarization) in the mother cell. Unequal proliferative divisions are a feature of the root epidermal cells of some species (Clowes, 2000). An outcome of this is that one cell out of each pair of daughter cells subsequently gives rise to a root hair, whereas the other cell remains hairless. A similar inequality is also evident in the root cortex of Monstera deliciosa, where trichosclereids (internal hairs) develop from the smaller daughter cell of some divisions (Bloch, 1946; Barlow, 1984). In addition, asymmetric cortical cell division is a prelude to schizogenous aerenchyma formation in roots of Nymphaea odorata (Seago et al., 2000) (see Fig. 17). However, the chain of cause and effect that links inequality of cell size to different cell fates must remain circumstantial until more detail accrues on the consequences of unequal division for the partitioning of preexisting gene products within the two daughter cells and the subsequent patterns of gene expression. Nevertheless, the few experiments that have been performed are suggestive of such a link. They have been aimed, for example, either at suppressing hair development by means

Cellular Patterning in Root Meristems

of the promotion of equal, rather than unequal, division or, alternatively, at inducing all cells, irrespective of their size, to become hairs. One simple experiment involved the epidermis of maize roots. Equal divisions were induced by their growth in water: hair development, normally associated with unequal cell division, was suppressed (Fig. 20A) (Ivanov and Filippenko, 1979; Filippenko, 1980). Unequal divisions also naturally occur in the cortex of primary roots of Zea mays (cv. LG11), where, apart from their size, there is no other evident difference between the two sister cells (Barlow, 1987). However, this inequality was found to be restricted to the first few divisions immediately following germination. Later, as the roots lengthened, divisions became more equal. As in the cases mentioned above, the direction of unequal cell production in the maize cortex is not left to chance: the daughter cell in an apical position was usually longer than the basal daughter. One consequence of such a size differential is that the larger daughter cell divides before the smaller cell (Fig. 21). Similar findings have been reported for the cortex

73

of wheat roots (Demchenko, 1975; Demchenko and Ivanov, 1978). Persistence of this directed pattern of unequal cell productions means that, after a few cell generations, a particular distribution of cell lengths is noticeable in each cortical cell file. Such patterned division sequences have been analyzed in the young primary root of maize (Lu¨ck et al., 1994a,b; Barlow, 1987) but, as noted above, the initial pattern of inequalities ceded to one in which daughter cells were not so unequal in size (Lu¨ck et al., 1997). A possible contributory factor in the change in division pattern is that the cells in which unequal divisions were displayed were preformed embryonic cells and, later, they were displaced by descendents of postformed cells that originated close to the root tip. Interestingly, paired files of cells created in the outer cortex by an early longitudinal division subsequently showed a pattern of cell productions that was the inverse (i.e., divisions follow a basipetal sequence within the file) of cells in neighboring files which lack the longitudinal division (their cells show acropetal division sequences). The reason for this is obscure, but these inverted sequences

Figure 20 Evidence of unequal cell division and subsequent cell growth in the primary root of Zea mays. (A) Percentage of newly divided pairs of cells with varying ratios of apical to basal (a/b) cell lengths in the epidermis of roots grown either in air (where hair cells develop from the epidermis) or in water (where the epidermis is hairless). The a/b ratio shifts to a value closer to 1 (indicative of equal division) in the water-grown roots. The cells were measured in a zone of the meristem in advance of hair initiation, so it is not certain, in the case of air-grown roots, if all pairs of cells would have formed hairs or not. (B) Similar data to A, except that the values denoted by line 1 refer to newly divided pairs of cells in the most proximal tier of the root cap columella. Values denoted by line 2 refer to the a/b ratio for the same pair of cells, but measured just prior to division of the most proximal (b) cells. The shift in the percentages (with respect to the values denoted by line 1) suggests that the more distal (a) daughter cell of a division grows much less than the proximal cell. (A modified from Ivanov and Filippenko, 1979; B modified from Ivanov and Larina, 1976.)

74

Barlow

Figure 21 Unequal divisions in the cortex of a primary root of Zea mays. Usually, the more apical cell is the larger in a pair of daughter cells. As a consequence, it enters mitosis before the smaller basal daughter cell. This pattern of division persists for a number of divisions until these cells and their descendants are displaced from the meristem. The cells in this photograph are the products of the first two rounds of division following imbibition. The arrow points toward the root apex, emphasizing the polarity of the division sequence. Scale bar ¼ 10
m.

account for some of the low-frequency division pathways found in an analysis aimed to discover how the various cortical cell division patterns arose (Lu¨ck et al., 1994b). The presence of differently sized cells within a population of otherwise apparently homogeneous cells indicates that the division process cannot be taken for granted as being one from which equally sized cells will inevitably emerge. Little is known of how the size of each daughter could be regulated, so no ready answer is available to the important question of what factors regulate the siting of the new cell wall within the mother cell. Hofmeister’s division rule and a critical aspect ratio seem, at a first approximation, to be sufficient (see Barlow and Adam, 1989), but genetic control over divisional asymmetry has been discovered in many systems (Horvitz and Herskowitz, 1992; Way et al., 1994), and this type of control may override others which seem to be based on cell geometry. Genetic regulation of cell size may be related to subsequent cell differentiation programs which depend for their inception on the occurrence of quantal mitoses at precise times within a cell lineage (e.g., the scarecrow mutation of Arabidopsis). Unequal division presents a problem that is rather different from, but nevertheless

related to, that concerning whether a cell within a meristem is to divide longitudinally or transversely. The siting of the new division wall not only relates to local conditions within the dividing mother cell but, in turn, is also related to conditions within the tissue as whole, whereas the plane of division depends more on this broader, tissue-based set of conditions which includes growth polarities already established in the axial and radial planes (Lynch and Lintilhac, 1997). Both local and more global controls of division impinge on the siting of the preprophase band (PPB) of microtubules, a component of the cytoskeleton which seems invariably to predict where the developing cell plate will attach to the walls of a dividing mother cell (Wick, 1991). The siting of the PPB therefore has a direct influence on the overall pattern of mitosis within tissues (Gunning et al., 1978b; Lloyd, 1986). In the case of unequal transverse divisions in the epidermis, the PPB is, as expected, sited asymmetrically within the cell (Gunning et al., 1978b). This is presumably a response to asymmetry of cytoplasm and nuclear positioning during preprophase, if not beforehand. The unequal early transverse divisions in the maize cortex can be equalized by exposing the roots to methanol (Barlow, 1989b). This same chemical also equalizes

Cellular Patterning in Root Meristems

the early divisions of fern spores and, in so doing, changes the course of differentiation of one of the daughter cells (Miller and Greany, 1976; Vogelmann and Miller, 1981)—both daughters subsequently differentiate as rhizoids rather than as a rhizoid and protonema initial (which are the products when methanol is absent). It is not known how methanol affects either cell polarity or PPB positioning. In maize, the usual asymmetry of cytoplasmic RNA staining within stelar cells was also minimized by methanol (Barlow, 1989b), and this disturbance to cytoplasmic partitioning may have affected the location of the PPB. Transversely dividing cells in the cortex of maize roots lie within a growth field that has a more or less uniform rate of longitudinal extension. Hence, any pattern of differential cell sizes is a reflection of size at birth and is not a product of subsequent elongation. However, in the columella of the root cap (and the same may be true of cells on the basiscopic face of the QC), the meristematic cells lie upon a steep growth gradient. Those cells within the proximal, first tier of the short cap meristem have a faster elongation rate than cells in more distal tiers (Barlow, 1977). Thus, although divisions in the first tier may be equal, the two daughter cells quickly become unequal in length (Fig. 20B) (Ivanov and Larina, 1976; Ivanov, 1979). Such a gradient makes it difficult to judge, in microscope preparations of cap tissue, whether divisions are equal or not without examining cell sizes immediately after cytokinesis. Asymmetrical divisions in roots endow each of the two resulting daughter cells with different potentialities for their subsequent structure and function. The already mentioned root-hair-inducing division in epidermis is an example of this. At the time of their origin, the smaller hair cell initials (trichoblasts) are located in a region of root where elongation is ceasing. The hair initials, having acquired an RNA-rich cytoplasm following the unequal division, retain the ability for massive cytoplasmic and cell wall growth. The new cytoplasmic material bursts out from a small area of the external periclinal wall of the trichoblast, this being the only site where any further increase in wall area can take place (though some mutations can alter the location of this site, or even abolish it). Trichosclereids of Monstera roots likewise extend their side walls into the intercellular air spaces within the maturing zone of the cortex. That the division that creates hair cells is often the ultimate or penultimate one in a series of proliferative divisions means that the resulting hair cell initial will be endowed with numbers of plasmodesmata on each of its walls that are different from those asso-

75

ciated with nonhair cells. Hence, the hair and nonhair cells have different degrees of symplasmic connection with each of their neighboring cells, as occurs, for example, in the water plant Trianea bogotensis (Kurkova and Vakhmistrov, 1984). In part, the differential is due to the diminution of plasmodesmatal density brought about by the continuing expansion of the walls. A similar mechanism might explain the increasing symplasmic isolation of maturing epidermal cells on the Arabidopsis root and the hair cells to which they give rise (Duckett et al., 1994). Similar considerations also hold for roots that possess an exodermis (internal to the epidermis) consisting of short passage cells, a type of cell which seems to be developmentally less mature than the longer, nonpassage, sister cell. Although the anatomy of onion root exodermis suggests a function similar to that of endodermis, the short passage cells seem able to enhance the viability of overlying epidermal cells during stressful conditions (Barrowclough and Peterson, 1994). The patterning of the dimorphic exodermal cells can be variable with long and short cells arranged either irregularly or regularly. In the latter case, the long and short cells alternate within a given file (Peterson, 1989). Likewise, epidermal hair cells can show various patterns. In Azolla pinnata, the pattern of hairs results from a lineage-based differentiation program in which asymmetrical epidermal cell division is one of the final steps (Gunning et al., 1978a). In members of the Brassicaceae, the patterning of the differentiated epidermal cells is not only the outcome of asymmetrical transverse divisions but also results from a specific type of contact between their anticlinal walls and the underlying cortical cells (Bu¨nning, 1951; Barlow, 1984; Volkmann and Peters, 1995). Moreover, in Potomageton natans, there is a relationship between the asymmetrical division pattern in epidermis and exodermis. Here each tissue forms long and short cells late in the divisional history of the respective cell files. Exodermis cells underlying the short hair initials divide asymmetrically, but when exodermis cells overlie long, nonhair cells, they do not divide and, hence, remain long (Tschermak-Woess and Hasitschka, 1953). When considered in the light of the function that such morphologically specialized cells possess, the root epidermal hair system might justifiably be considered a paradigm for investigating the patterns of gene control and physiology within the context of a specific tissue structure–function relationship (Dolan, 1996). Not all the elements in this system of interacting cells are by any means certainly identi-

76

fied. Moreover, the function of root hairs is manifold: they participate in solute uptake and they stabilize the root in soil, to name but two evident functions. But to what extent are they dispensable in either of these processes? Whatever the emerging view on this question, it is clear that patterned cell division has a fundamental role not only in the specific case of the epidermis but also in the more general process of establishing root form during embryogenesis, as well as for the continued maintenance of form and function of roots in their natural environment.

REFERENCES Armstrong JE, Heimsch C. 1976. Ontogenetic reorganization of the root meristem in the Compositae. Am J Bot 63:212–219. Armstrong SW, Davidson D. 1982. Differences in protein content of sister nuclei: evidence from binucleate and mononucleate cells. Can J Biochem 60:371–378. Artschwager E. 1926. Anatomy of the vegetative organs of the sugar beet. J Agric Sci 33:143–176. Balus˘ ka F. 1990. Nuclear size, DNA content, and chromatin condensation are different in individual tissues of the maize root apex. Protoplasma 158:45–52. Balus˘ ka, F, Brailsford RW, Hauskrecht M, Jackson MB, Barlow PW. 1993. Cellular dimorphism in the maize root cortex: involvement of microtubules, ethylene and gibberellin in the differentiation of cellular behaviour in postmitotic growth zones. Bot Acta 106:394-403. Balus˘ ka F, Barlow PW, Lichtscheidl IK, Volkmann D. 1998. The plant cell body: a cytoskeletal tool for cellular development and morphogenesis. Protoplasma 202:1– 10. Balus˘ ka, F, Volkmann D, Barlow PW. 1999. Hormone– cytoskeleton interactions in plant cells. In: Hookyaas PJJ, Hall MA, Libbenga KR, eds. Biochemistry and Molecular Biology of Plant Hormones. Amsterdam: Elsevier, pp 363–390. Barlow PW. 1977. An experimental study of cell and nuclear growth and their relation to cell diversification within a plant tissue. Differentiation 8:153–157. Barlow PW. 1982. Root development. In: Smith H, Grierson D, eds. The Molecular Biology of Plant Development. Oxford, UK: Blackwell Scientific Publications, pp 185– 222. Barlow PW. 1984. Positional controls in root development. In: Barlow PW, Carr DJ, eds. Positional Controls in Plant Development. Cambridge, UK: Cambridge University Press, pp 281–318. Barlow PW. 1986. Adventitious roots of whole plants: Their forms, functions, and evolution. In: MB Jackson, ed. New Root Formation in Plants and Cuttings. Dordrecht, Netherlands: Nijhoff, pp 67–110.

Barlow Barlow PW. 1987. Cellular packets, cell division and morphogenesis in the primary root meristem of Zea mays L. New Phytol 105:27–56. Barlow PW. 1989a. Meristems, metamers and modules and the development of shoot and root systems. Bot J Linn Soc 100:255–279. Barlow PW. 1989b. Experimental modification of cell division patterns in the root meristem of Zea mays. Ann Bot 64:13-20. Barlow PW. 1992. The meristem and quiescent centre in cultured root apices of the gib-1 mutant tomato (Lycopersicon esculentum Mill.). Ann Bot 69:533– 543. Barlow PW. 1993. The cell division cycle in relation to root organogenesis. In: Ormrod JC, Francis D, eds. Molecular and Cell Biology of the Plant Cell Cycle. Dordrecht, Netherlands: Kluwer, pp 179–199. Barlow PW. 1994a. The origin, diversity and biology of shoot-borne roots. In: Davis TD, Haissig BE, eds. The Biology of Adventitious Root Formation. New York: Plenum Press, pp 1–23. Barlow PW. 1994b. Cell division in meristems and their contribution to organogenesis and plant form. In: Ingram DS, Hudson A, eds. Shape and Form in Plants and Fungi. London: Academic Press, pp 169–193. Barlow PW. 1994c. Structure and function at the root apex— phylogenetic and ontogenetic perspectives on apical cells and quiescent centres. Plant Soil 167:1–16. Barlow PW. 1994d. Evolution of structural initial cells in apical meristems of plants. J Theoret Biol 169:163–177. Barlow PW 1995a. Stem cells and founder zones in plants, particularly their roots. In: Potten CS, ed. Stem Cells. London: Academic Press, pp 29-57. Barlow PW. 1995b. Gravity perception in plants: a multiplicity of systems derived by evolution? Plant Cell Environ 18:951–962. Barlow PW. 1999. Living plant systems: how robust are they in the absence of gravity? Adv Space Res 23:1975– 1986. Barlow PW, Adam JS. 1988. The position and growth of lateral roots on cultured root axes of tomato, Lycopersicon esculentum (Solanaceae). Plant Syst Evol 158:141–154. Barlow PW, Adam JS. 1989. Experimental control of cellular patterns in the cortex of tomato roots. In: Loughman BC, Gasparı´ kova´ O, Kolek J, eds. Structural and Functional Aspects of Transport in Roots. Dordrecht, Netherlands: Kluwer, pp 21–24. Barlow PW, Rathfelder EL. 1984. Correlation between the dimensions of different zones of grass root apices, and their implications for morphogenesis and differentiation in roots. Ann Bot 53:249–260. Barlow PW, Lu¨ck HB, Lu¨ck J. 2001. Autoreproductive cells and plant meristem construction: the case of the tomato cap meristem. Protoplasma 215: 50–63.

Cellular Patterning in Root Meristems Barrowclough DE, Peterson CA. 1994. Effects of growing conditions and development of the underlying exodermis on the vitality of the onion root epidermis. Physiol Plant 92:343–349. Batygina TB. 1969. On the possibility of separation of a new type of embryogenesis in Angiospermae. Rev Cytol Biol Veg 32:335–341. Berleth T, Ju¨rgens G. 1993. The role of the monopteros gene in organizing the basal body region of the Arabidopsis embryo. Development 118:575–587. Binns AN, Chen RH, Wood HN, Lynn DG. 1987. Cell division promoting activity of naturally occurring dehydrodiconiferyl glucosides: do cell wall components control cell division? Proc Natl Acad Sci USA 84:980–984. Bloch R. 1946. Differentiation and pattern in Monstera deliciosa. The idioblastic development of the trichosclereids in the air root. Am J Bot 33:544–551. Bonnett HT Jr, Torrey JG. 1966. Comparative anatomy of endogenous bud and lateral root formation in Convolvulus arvensis roots cultured in vitro. Am J Bot 53:496–507. Bugnicourt M. 1983. Embryoge´nie des Hypericum de France. Bull Soc Bot Fr Lett Bot 1983:195–120. Bulgakov SW. 1944. Modified anatomical structure in koksaghyz roots and its biological import. C R Acad Sci URSS 45:35–38 Bu¨nning E. 1951. U¨ber die Differenzierungsvorgange in der Cruciferenwurzel. Planta 39:126–153. Burssens S, Engler JA, Beeckman T, Richard C, Shaul O, Ferreira P, Van Montagu M, Inze´ D. 2000. Developmental expression of the Arabidopsis thaliana CycA2;1 gene. Planta 211:623–631. Byrne J, Heimsch C. 1968. The root apex of Linum. Am J Bot 55:1011–1019. Byrne J, Heimsch C. 1970a. The root apex of Malva sylvestris. II. The quiescent center. Am J Bot 57:1179–1184. Byrne J, Heimsch C. 1970b. The root apex of Malva sylvestris. I. Structural development. Am J Bot 57:1170– 1178. Casero PJ, Casimiro I, Rodrı´ guez-Gallardo L, Martı´ nPartido G, Lloret PG. 1993. Lateral root initiation by asymmetrical transverse divisions of pericycle cells in adventitious roots of Allium cepa. Protoplasma 176:138–144. Cave MS, Arnott HJ, Cook SA. 1961. Embryogeny in the Californian peonies with reference to their taxonomic position. Am J Bot 48:397–404. Cedar H, Razin A. 1990. DNA methylation and development. Biochim Biophys Acta 1049:1–8. Cheng J-C, Seeley KA, Sung ZR. 1995. RML1 and RML2, Arabidopsis genes required for cell proliferation at the root tip. Plant Physiol 107:365–376. Chiang S-HT, Tsou A-P. 1974. The establishment of the anatomical zonation in the root apex of Phaseolus radiatus. Taiwania 19:96–105.

77 Clowes FAL. 1958. Development of quiescent centres in root meristems. New Phytol 57:85–88. Clowes FAL. 1959. Apical meristems of roots. Biol Rev 34:501–529. Clowes FAL. 1964. The quiescent center in meristems and its behaviour after irradiation. Brookhaven Symp Biol 16:46–57. Clowes FAL. 1978. Origin of the quiescent centre in Zea mays. New Phytol 80:409–419. Clowes FAL. 1981. The difference between open and closed meristems. Ann Bot 48:761–767. Clowes FAL. 2000. Pattern in root meristem development in angiosperms. New Phytol 146:83–94. Cooke TJ, Tilney MS, Tilney LG. 1996. Plasmodesmatal networks in apical meristems and mature structures: geometric evidence for both primary and secondary formation of plasmodesmata. In: Smallwood M, Knox JP, Bowles DJ, eds. Membranes: Specialized Functions in Plants. Oxford, UK: Bios, pp 471–488. Curtis JD, Lersten NR. 1990. Oil reservoirs in stem, rhizome, and root of Solidago canadensis (Asteraceae, tribe Astereae). Nord J Bot 10:443–449. Dawe RK, Freeling M. 1991. Cell lineage and its consequences in higher plants. Plant J 1:3–8. Deltour R, Antoine N, Bronchart R. 1989. Root ontogenesis during early germination of Zea mays. Ann Bot 64:107–116. Demchenko NP. 1975. The sequence of entering mitosis in wheat root sister cells and their difference in duration of mitotic cycles. Bot Z 60:88–198 (in Russian). Demchenko NP, Ivanov VB. 1978. Dependence of the duration of the mitotic cycle and G2 period in sister cells of wheat root meristem on the ratio of their lengths. Soviet J Dev Biol 9:233–241. Di Laurenzio L, Wysocka-Diller J, Malamy JE, Pysh L, Helariutta Y, Freshour G, Hahn MG, Feldmann KA, Benfey PN. 1996. The SCARECROW gene regulates an asymmetric cell division that is essential for generating the radial organization of the Arabidopsis root. Cell 86:423–433. Doerner P, Jørgensen J-E, You R, Steppuhn J, Lamb C. 1996. Control of growth and development by cyclin expression. Nature 380:520–523. Dolan L. 1996. Pattern in the root epidermis: an interplay of diffusible signals and cellular geometry. Ann Bot 77:547–553. Dolan L, Janmaat K, Willemsen V, Linstead P, Poethig S, Roberts K. 1993. Cellular organization of the Arabidopsis thaliana root. Development 119:71–84. Dubrovsky JG. 1997. Determinate primary-root growth in seedlings of Sonoran Desert Cactaceae; its organization, cellular basis, and ecological significance. Planta 203:85–92. Duckett CM, Oparka KJ, Prior DAM, Dolan L, Roberts K. 1994. Dye-coupling in the root epidermis of

78 Arabidopsis is progressively reduced during development. Development 120:3247–3255 Ellmore G. 1981. Root dimorphism in Ludwigia peploides (Onagraceae): development of two root types from similar primordia. Bot Gaz 142:525–533. Ennos AR, Fitter AH. 1992. Comparative functional morphology of the anchorage systems of annual dicots. Funct Ecol 6:71–78. Errera L. 1886. Sur une condition fondamentale d’e´quilibre des cellules vivantes. C R Acad Sci (Paris) 103:822– 824. Filippenko VN. 1980. Determination of the direction of rhizodermal cell differentiation in corn. Soviet J Dev Biol 11:189–196. Fitter AH. 1987. An architectural approach to the comparative ecology of plant root systems. New Phytol 106(suppl):61–77. Gallagher K, Smith LG. 1997. Asymmetric division and cell fate in plants. Curr Opin Cell Biol 9:842–848. Greig N, Mauseth JD. 1991. Structure and function of dimorphic prop roots in Piper auritum L. Bull Torrey Bot Club 118:176–183. Guignard JL. 1961. Recherches sur l’embryoge´nie des Gramine´es. Rapports des Gramine´es avec les autres Monocotyledones. Ann Sci Nat Ser 12 (2):491–611. Gunning BES, Hughes JE, Hardham AR. 1978a. Formative and proliferative cell divisions, cell differentiation, and developmental changes in the meristem of Azolla roots. Planta 143:121–144. Gunning BES, Hardham AR, Hughes JE. 1978b. Pre-prophase bands of microtubules in all categories of formative and proliferative cell division in Azolla roots. Planta 143:145–160. Haber AH. 1968. Ionizing radiations as research tools. Annu Rev Plant Physiol 19:463–489. Haccius B. 1978. Question of unicellular origin of non-zygotic embryos in callus cultures. Phytomorphol 28:74– 81. Halle´ F, Oldeman RAA Tomlinson PB. 1978. Tropical Trees and Forests. An Architectural Analysis. Berlin; Springer. Hayat MA. 1963. Apical organization in roots of the genus Cassia. Bull Torrey Bot Club 90:123–136. Hejnowicz Z. 1989. Differential growth resulting in the specification of different types of cellular architecture in root meristems. Environ Exp Bot 29:85–93. Hejnowicz Z, Hejnowicz K. 1991. Modeling the formation of root apices. Planta 184:1–7. Hill A. 1990. Entropy production as the selection rule between different growth morphologies. Nature 348:426–428. Hinchee MAW. 1981. Morphogenesis of aerial and subterranean roots of Monstera deliciosa. Bot Gaz 142:347– 359. Holliday R. 1987. The inheritance of epigenetic defects. Science 238:163–170.

Barlow Holtzer H, Rubinstein N, Fellini S, Yeoh G, Chi J, Birnbaum J, Okayama M. 1975. Lineages, quantal cell cycles, and the generation of cell diversity. Q Rev Biophys 8:523– 557. Horvitz HR, Herskowitz I. 1992. Mechanism of asymmetric cell division: two Bs or not two Bs, that is the question. Cell 68:237–255. Innocenti AM, Bitonti MB, Arrigoni O, Liso R. 1990. The size of the quiescent centre of Allium cepa L. growing with ascorbic acid. New Phytol 114:507–509. Irish VF. 1991. Cell lineages in plant development. Curr Opin Cell Biol 3:983–987. Irish VF. 1993. Cell fate determination in plant development. Semin Dev Biol 4:73–81. Ivanov VB. 1979. Determination of sequence of emergence of separate cells from the mitotic cycle and their transition to depositing starch in the sheath of maize roots. Dokl Bot Sci 245:766–768. Ivanov VB, Filippenko VN. 1979. Symmetry in mitotic division and differentiation of cells of the rhizodermis in maize. Dokl Bot Sci 245:763–765. Ivanov VB, Larina LP. 1976. The growth of root cap columella cells of maize seedlings. Tsitologiya 18:1360– 1363 (in Russian). Jacobs WP. 1994. Caulerpa. Sci Am 271(6):66–71. Jaffe MJ. 1985. Wind and other mechanical effects in the development and behavior of plants, with special emphasis on the role of hormones. In: Encyclopedia of Plant Physiology. New Series, Vol. 11. Heidelberg: Springer, pp 444–484. Jenı´ k J. 1978. Roots and root systems in tropical trees: morphologic and ecologic aspects. In: Tomlinson PB, Zimmermann MH, eds. Tropical Trees as Living Systems. Cambridge, UK: Cambridge University Press, pp 323–349. Jing H-C, Van Lammeren AAM, de Castro RD, Bino RJ, Hilhorst HWM, Groot SPC. 1999. -Tubulin accumulation and DNA synthesis are sequentially resumed in embryo organs of cucumber (Cucumis sativus L.) seeds during germination. Protoplasma 208:230–239. Johansen DA. 1945. A critical survey of the present status of plant embryology. Bot Rev 11:87–107. Jones PA. 1977. Development of the quiescent center in maturing embryonic radicles of pea (Pisum sativum L. cv. Alaska). Planta 135:233–240. Juniper BE, Barlow PW. 1969. The distribution of plasmodesmata in the root tip of maize. Planta 89:352–360. Ju¨rgens G, Mayer U. 1994. Arabidopsis. In: Bard J, ed. Embryos. Colour Atlas of Development. London: Wolfe, pp 7–21. Ju¨rgens G, Grebe M, Steinmann T. 1997. Establishment of cell polarity during early plant development. Curr Opin Cell Biol 9:849–852. Justin SHFW, Armstrong W. 1987. The anatomical characteristics of roots and plant response to soil flooding. New Phytol 106:465–495.

Cellular Patterning in Root Meristems Kapil RN, Rustagi PN. 1966. Anatomy of the aerial and terrestrial roots of Ficus benghalensis L. Phytomorphol 16:382–386. Kaplan DR, Hagemann W. 1991. The relationship of cell and organism in vascular plants. Are cells the building blocks of plant form? BioSci 41:693–703. Kaska DD, Myllyla¨ R, Cooper JB. 1999. Auxin transport inhibitors act through ethylene to regulate dichotomous branching of lateral root meristems in pine. New Phytol 142:49–58. Kawai M, Samarajeewa PK, Barrero RA, Nishiguchi M, Uchimiya H. 1998. Cellular dissection of the degradation pattern of cortical cell death during aerenchyma formation in rice roots. Planta 204:277–287. Keller B, Lamb CJ. 1989. Specific expression of a novel cell wall hydroxyproline-rich glycoprotein gene in lateral root initiation. Genes Dev 3:1639–1646. Kerk N, Feldman L. 1994. The quiescent center in roots of maize: initiation, maintenance and role in organization of the root apical meristem. Protoplasma 183:100–106. Kerk N, Feldman L. 1995. A biochemical model for the initiation and maintenance of the quiescent center: implications for organization of root meristems. Development 121:2825–2833. Knox JP, Day S, Roberts K. 1989. A set of cell surface glycoproteins forms an early marker of cell position, but not cell type, in the root apical meristem of Daucus carota L. Development 106:47–56. Konings H, Verschuren G. 1980. Formation of aerenchyma in roots of Zea mays in aerated solutions, and its relation to nutrient supply. Physiol Plant 49:265–270. Korn RW, Spalding RM. 1973. The geometry of plant epidermal cells. New Phytol 72:1357–1365. Kreuger M, Van Holst G-J. 1995. Arabinogalactan-protein epitopes in somatic embryogenesis of Daucus carota L. Planta 197:135–141. Kuras M. 1980. Activation of embryo during rape (Brassica napus L.) seed germination II. Transversal organization of radicle apical meristem. Acta Bot Soc Polon 49:387–395. Kurkova EB, Vakhmistrov DB. 1984. Distribution of plasmodesmata in the rhizodermis along the root axis in Trianea bogotensis. Soviet J Plant Physiol 31:115–119. Lachmann P. 1907. Origine et de´veloppement des racines et des radicelles du Ceratopteris thalictroides. Rev Gen Bot 19:523–556. Lee MM, Schiefelbein J. 1999. WEREWOLF, a MYBrelated protein in Arabidopsis, is a position-dependent regulator of epidermal cell patterning. Cell 99:473–483. Lintilhac PM. 1974. Differentiation, organogenesis, and the tectonics of cell wall orientation. II. Separation of stresses in a two-dimensional model. Am J Bot 61:135–140.

79 Litinskaya TK. 1993. Frequency of division in three direction in the apex of a growing root. Fiziol Biokhim Kult Rast 25:253–260 (in Russian). Liu C-M, Xu, Z-H, Chua N-H. 1993. Auxin polar transport is essential for the establishment of bilateral symmetry during early plant embryogenesis. Plant Cell 5:621– 630. Lloret PG, Casero PJ, Pulgarin A, Navascues J. 1989. The behaviour of two cell populations in the pericycle of Allium cepa, Pisum sativum, and Daucus carota during early lateral root development. Ann Bot 63:465–475. Lloyd CW. 1986. Microtubules and the cellular morphogenesis of plants. In: Bowder LW, ed. Developmental Biology. A Comprehensive Synthesis. Vol. 2. The Cellular Basis of Morphogenesis. New York: Plenum Press, pp 31–57. Loomis RS, Torrey JG. 1964. Chemical control of vascular cambium initiation in isolated radish roots. Proc Natl Acad Sci USA 52:3–11. LoSchiavo F, Pitto L, Giuliano G. 1989. DNA methylation of embryogenic carrot cell cultures and its variations as caused by mutation, differentiation, hormones and hypomethylating drugs. Theoret Appl Genet 77:325– 331. Lu P, Jernstedt JA. 1996. Rhizophore and root development in Selaginella martensii: meristem transitions and identity. Int J Plant Sci 157:180–194. Lu P, Porat R, Nadeau JA, O’Neill SD. 1996. Identification of a meristem L1 layer-specific gene in Arabidopsis that is expressed during embryonic pattern formation and defines a new class of homeobox genes. Plant Cell 8:2155–2168. Lu¨ck HB, Lu¨ck J. 1986. Unconventional leaves. (An application of map 0L-systems to biology.) In: Rozenberg G, Salomaa A, eds. The Book of L. Berlin: Springer, pp 275–289. Lu¨ck J, Lu¨ck HB. 1993. Parallel rewriting DW-cellwork systems for plant development. In: Demongeot J, Capasso V, eds. Mathematics Applied to Biology and Medicine. Winnipeg, Canada: Wuerz, pp 461–466. Lu¨ck J, Barlow PW, Lu¨ck HB. 1994a. Deterministic patterns of cellular growth and division within a meristem. Ann Bot 73:1–11. Lu¨ck J, Barlow PW, Lu¨ck HB. 1994b. Cell genealogies in a plant meristem deduced with the aid of a ‘bootstrap’ Lsystem. Cell Prolif 27:1–21. Lu¨ck J, Barlow PW, Lu¨ck HB. 1997. An automata-theoretical model of meristem development as applied to the primary root of Zea mays L. Ann Bot 79:375–389. Luxova´ M. 1975. Some aspects of the differentiation of primary root tissues. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. London: Academic Press, pp 73–90. Lyko F, Paro R. 1999. Chromosomal elements conferring epigenetic inheritance. BioEssays 21:824–831.

80 Lynch TM, Lintilhac PM. 1997. Mechanical signals in plant development: a new method for single cell studies. Dev Biol 181:246–256. MacLeod RD, McLachlan SM. 1974. The development of the quiescent centre in lateral roots of Vicia faba L. Ann Bot 38:535–544. Malamy JE, Benfey PN. 1997. Organization and cell differentiation in lateral roots of Arabidopsis thaliana. Development 124:33–44. Mansfield SG, Briarty LG. 1991. Early embryogenesis in Arabidopsis thaliana. II. The developing embryo. Can J Bot 69:461–476. Masand P. 1963. Embryology of Zygophyllum fabago Linn. Phytomorphol 13:293–302. Mayer U, Torres Ruiz RA, Berleth T, Mise´ra S, Ju¨rgens G. 1991. Mutations affecting body organization in the Arabidopsis embryo. Nature 353:402–407. Mayer U, Bu¨ttner G, Ju¨rgens G. 1993. Apical-basal pattern formation in the Arabidopsis embryo: studies on the role of the gnom gene. Development 117:149–162. Mayr E. 1961. Cause and effect in biology. Science 134:1501– 1506. McCabe PF, Valentine TA, Forsberg, LS, Pennell RI. 1997. Soluble signals from cells identified at the cell wall establish a developmental pathway in carrot. Plant Cell 9:2225–241. Meyer CF. 1958. Cell patterns in early embryogeny of the McIntosh apple. Am J Bot 45:341–349. Mezitt LA, Lucas WJ. 1996. Plasmodesmal cell-to-cell transport of proteins and nucleic acids. Plant Mol Biol 32:251–273. Miller JH, Greany RH. 1976. Rhizoid differentiation in fern spores: experimental manipulation. Science 193:687– 689. Mordhorst AP, Toonen MAJ, de Vries SC. 1997. Plant embryogenesis. Crit Rev Plant Sci 16:535–576. Munksgaard D, Mattsson O, Okkels FT. 1995. Somatic embryo development in carrot is associated with an increase in levels of S-adenosylmethionine, S-adenosylhomocysteine and DNA methylation. Physiol Plant 93:5–10. Nakielski J. 1992. Regeneration of the root apex: Modelling study by means of the growth tensor. In: Karalis TK, ed. Mechanics of Swelling (NATO ASI series H64.) Berlin: Springer, pp 179–191. Nakielski J, Barlow PW. 1995. Principal directions of growth and the generation of cell patterns in wild-type and gib1 mutant roots of tomato (Lycopersicon esculentum Mill.) grown in vitro. Planta 196:30–39. Natesh S, Rau MA. 1984. The embryo. In: Johri BM, ed. Embryology of Angiosperms. Berlin: Springer, pp 377– 443. Neuteboom LW, Veth-Tello LM, Clijdesdale OR, Hooykaas PJJ, Van der Zaal BJ. 1999. A novel subtilisin-like

Barlow protease gene from Arabidopsis thaliana is expressed at sites of lateral root emergence. DNA Res 6:13–19. Niklas KJ. 1989. The cellular mechanics of plants. Am Sci 77:344–349. Obroucheva NV. 1999. Seed Germination. A Guide to the Early Stages. Leiden, Netherlands: Backhuys. Olsen O-A, Brown RC, Lemmon BE. 1995. Pattern and process of wall formation in developing endosperm. BioEssays 17:803–812. Owens JN, Catalano GL, Morris SJ, Aitken-Christie J. 1995. The reproductive biology of kauri (Agathis australis). III. Proembryogeny and early embryogeny. Int J Plant Sci 156:793–806. Payne PI, Dobrzanska M, Barlow PW, Gordon ME. 1978. The synthesis of RNA in imbibing seeds of rape (Brassica napus) prior to the onset of germination: a biochemical and cytological study. J Exp Bot 29:77–88. Pennell RI, Janniche L, Kjellbom P, Scofield GN, Peart JM, Roberts K. 1991. Developmental regulation of a plasma membrane arabinogalactan protein epitope in oilseed rape flowers. Plant Cell 3:1317–1326. Peterson CA. 1989. Significance of the endodermis in root function. In: Loughman BC, Gasparı´ kova´ O, Kolek J, eds. Structural and Functional Aspects of Transport in Roots. Dordrecht, Netherlands: Kluwer, pp 35-40. Peterson RL. 1992. Adaptations of root structure in relation to biotic and abiotic factors. Can J Bot 70:661–675. Poddar K, Vishnoi RK, Kothari SL. 1997. Plant regeneration from embryogenic callus of finger millet Eleusine coracana (L.) Gaertn. on higher concentration of NH4NO3 as a replacement of NAA in the medium. Plant Sci 129:101–106. Pollock EG, Jensen WA. 1964. Cell development during early embryogenesis in Capsella and Gossypium. Am J Bot 51:915–921. Popham RA. 1955a. Zonation of primary and lateral root apices of Pisum sativum. Am J Bot 42:267–273. Popham RA 1955b. Levels of tissue differentiation in primary roots of Pisum sativum. Am J Bot 42:529–540. Pukhal’skaya EC. 1949. On the localization of the first mitoses in the root meristems of higher plants. Dokl Akad Nauk USSR 68:915–917 (in Russian). Racusen RH, Schiavone FM. 1988. Detection of spatiallyand stage-specific proteins in extracts from single embryos of the domesticated carrot. Development 103:665–674. Raff JW, Kellum R, Alberts B. 1994. The Drosophila GAGA transcription factor is associated with specific regions of heterochromatin throughout the cell cycle. EMBO J 13:5977–5983. Reeves RG, Beasley JO. 1935. The development of the cotton embryo. J Agric Res 53:935–941. Reihman MA, Rost TL. 1990. Regeneration responses in pea roots after tip excision at different levels. Am J Bot 77:1159–1167.

Cellular Patterning in Root Meristems Roy Chowdhury C. 1962. The embryogeny of conifers: a review. Phytomorphology 12:313–338. Sachs J von. 1887. Lecture XXVII. Relations between growth and cell-division in the embryonic tissues. In: Lectures in Plant Physiology (translated by HM Ward). Oxford, UK: Clarendon Press, pp 431–459. Schatt E, Landre´ P, Nougare`de A. 1985. E´tat nucle´aire des me´riste`mes du pois dans la graine se`che; imbibition et reprise du cycle cellulaire. Can J Bot 63:2200–2208. Scheres B, Wolkenfelt H, Willemsen V, Terlouw M, Lawson E, Dean C, Weisbeek P. 1994. Embryonic origin of the Arabidopsis primary root and root meristem initials. Development 120:2475–2487. Scheres B, McKhann H, Van den Berg C, Willemsen V, Wolkenfelt H, de Vrieze G, Weisbeek P. 1996. Experimental and genetical analysis of root development in Arabidopsis. Plant Soil 187:97–105. Schnell R. 1967. E´tudes sur l’anatomie et la morphologie des Podostemace´es. Candollea 22:157–225. Schoof H, Lenhard M, Haecker A, Mayer KFX, Ju¨rgens G, Laux T. 2000. The stem cell population of Arabidopsis shoot meristems is maintained by a regulatory loop between the CLAVATA and WUSCHEL genes. Cell 100:635–644. Seago JL, Peterson CA, Enstone DE. 1999. Cortical ontogeny in roots of the aquatic plant, Hydrocharis morsusranae L. Can J Bot 77:113-121. Seago JL, Peterson CA, Kinsley LJ, Broderick J. 2000. Development and structure of the root cortex in Caltha palustris L. and Nymphaea odorata Ait. Ann Bot 86:631–640. Seagull RW. 1983. Differences in the frequency and disposition of plasmodesmata resulting from root cell elongation. Planta 159:497–504 Singh H. 1978. Embryology of gymnosperms. In: Zimmermann W, Carlquist S, Ozenda P, Wulff HD, eds. Handbuch der Pflanzenanatomie, Vol 10, part 2. Berlin; Borntraeger. Sinnott EW. 1960. Plant Morphogenesis. New York; McGraw-Hill. Sitte P. 1992. A modern concept of the ‘‘cell theory.’’ A perspective on competing hypotheses of structure. Int J Plant Sci 153:S1–S6. Soue`ges R. 1924. Embryoge´nie des Gramines. Developpement de l’embryon chez le Poa annua L. C R Acad Sci (Paris) 178:860–862. Sporne K. 1974. The Morphology of Angiosperms. London; Hutchinson. Stein OL, Quastler H. 1963. The use of tritiated thymidine in the study of tissue activation during germination in Zea mays. Am J Bot 50:1006–1011. Steward FC. 1958. Growth and organized development of cultured cells. III. Interpretations of the growth from free cell to carrot plant. Am J Bot 45:709–713. Steward FC, Mapes MO, Smith J. 1958. Growth and organized development of cultured cells. I. Growth and

81 division of freely suspended cells. Am J Bot 45:693– 703. Swamy BGL, Padmanabhan D. 1961. Embryogenesis in Sphenoclea zeylanica. Proc Indian Acad Sci 54B:169– 187. Thomson RB, Hull KL. 1934. Physical laws and the cellular organization of plants. Bot Gaz 95:511–514. Toonen MAJ, Schmidt EDL, Van Kammen A, De Vries SC. 1997. Promotive and inhibitory effects of diverse arabinogalactan proteins on Daucus carota L. somatic embryogenesis. Planta 203:188–195. Torres-Ruiz RA, Ju¨rgens G. 1994. Mutations in the FASS gene uncouple pattern formation and morphogenesis in Arabidopsis development. Development 120:2967– 2978. Torrey JG. 1986. Endogenous and exogenous influences on the regulation of lateral root formation. In: Jackson MB, ed. New Root Formation in Plants and Cuttings. Dordrecht, Netherlands: Nijhoff, pp 31–66. Torrey JG, Callaham D. 1978. Determinate development of nodule roots in actinomycete-induced root nodules of Myrica gale. Can J Bot 56:1357–1364. Torrey JG, Loomis RS. 1967. Auxin-cytokinin control of secondary vascular tissue formation in isolated roots of Raphanus. Am J Bot 54:1098–1106. Tschermak-Woess E, Hasitschka G. 1953. U¨ber Musterbildung in der Rhizodermis and Exodermis bei einigen Angiospermen und einer Polypodiacee. O¨sterr Bot Z 100:646–651. Tykarska T. 1976. Rape embryogenesis I. The proembryo development. Acta Soc Bot Polon 45:3–16. Tykarska T. 1979. Rape embryogenesis. II. Development of embryo proper. Acta Soc Bot Polon 48:391–421. Tylicki A, Burza W, Kuras M, Dziadczyk E, Malepszy S. 2000. Structural and ultrastructural analysis of root primordia in vitro cultures (RPC) of Solanum lycopersicoides Dun. Plant Sci 156:73–83. Vallade J. 1970. De´veloppement embryonnaire chez un Petunia hybrida hort. C R Acad Sci Ser D 270:1893– 1896. Vallade J. 1972. Structure et fonctionnement du me´riste`me lors de la formation de la jeune racine primaire chez un Petunia hybrida hort. C R Acad Sci Ser D 274:1027– 1030. Vallade J. 1989. Embryogenesis and research of fundamental morphogenetic processes. In: Some Aspects and Actual Orientations in Plant Embryology. Dedicatory Volume to Professor A. Lebe`gue. Amiens, France: University of Picardie, pp 171–187. Van den Berg C, Willemsen V. Hage W, Weisbeek P, Scheres B. 1995. Cell fate in Arabidopsis root meristems determined by directional signalling. Nature 378:62–65. Vernon D, Meinke D. 1994. Embrogenic transformation of the suspensor in twin, a polyembryonic mutant of Arabidopsis. Dev Biol 165:566–573.

82 Vladesco MA. 1935. Recherches morphologiques et expe´rimentales sur l’embryoge´nie et organoge´nie des fouge`res leptosporangie´es. Rev Gen Bot 47:422–434, 513–528, 564–588. Vogelmann TC, Miller JH. 1981. The effect of methanol on spore germination and rhizoid differentiation in Onoclea sensibilis. Am J Bot 68:1177–1183. Volkmann D, Peters P. 1995. Structural basis of root hair formation: early development of trichoblasts and atrichoblasts. In: Balus˘ ka F, C˘iamporova´ M, Gasparı´ kova´ O, Barlow PW, eds. Structure and Function of Roots. Dordrecht, Netherlands: Kluwer, pp 61–67. Voronin NS. 1969. Root apical meristems in gymnosperms and the principles of their graphical interpretation. Bot Z (Leningrad) 54:67–76 (in Russian). Vroemen C, De Vries S, Quatrano R. 1999. Signalling in plant embryos during the establishment of the polar axis. Semin Cell Dev Biol 10:157–164. Vuylsteker C, Dewaele E, Rambour S. 1998. Auxin induced lateral root formation in chicory. Ann Bot 81:449–454. Ward M, Wetmore RH. 1954. Experimental control of development in the embryo of the fern, Phlebodium aureum. Am J Bot 41:428–434. Way JC, Wang L, Run J-Q, Hung M-S. 1994. Cell polarity and the mechanism of asymmetric cell division. BioEssays 16:925–931. Wick SM. 1991. The preprophase band. In: Lloyd CW, ed. The Cytoskeletal Basis of Plant Growth and Form. London; Academic Press, pp 231–244.

Barlow Wilder GJ, Johansen JR. 1992. Comparative anatomy of absorbing roots and anchoring roots in three species of Cyclanthaceae (Monocotyledoneae). Can J Bot 70:2384–2404. Willemsen V, Wolkenfelt H, de Vrieze G, Weisbeek P, Scheres B. 1998. The HOBBIT gene is required for formation of the root meristem in the Arabidopsis embryo. Development 125:521–531. Williams BC. 1947. The structure of the meristematic root tip and origin of primary tissues in the roots of vascular plants. Am J Bot 34:455–462. Williams EG, Maheswaran G. 1986. Somatic embryogenesis: factors influencing coordinated behaviour of cells as an embryogenic group. Ann Bot 57:443–462. Ye Z-H, Varner JE. 1991. Tissue specific expression of cell wall proteins in developing soybean tissues. Plant Cell 3:23–37. Yeung EC, Rahman MH, Thorpe TA. 1996. Comparative development of zygotic and microspore-derived embryos in Brasica napus L. cv. Topas. I. Histodifferentiation. Int J Plant Sci 157:27–39. Yeung EC, Stasolla C, Kong L. 1998. Apical meristem formation during zygotic embryo development of white spruce. Can J Bot 76:751–761. Zhang H, Jennings A, Barlow PW, Forde BG. 1999. Dual pathways for regulation of root branching by nitrate. Proc Natl Acad Sci USA 96:6529–6534. Zhang N, Hasenstein KH. 1999. Initiation and elongation of lateral roots in Lactuca sativa. Int J Plant Res 160:511– 519.

5 Root Hairs: Hormones and Tip Molecules Robert W. Ridge and Masayuki Katsumi International Christian University, Tokyo, Japan

I.

INTRODUCTION

supply them to root hairs. In addition, under natural conditions, root hairs are exposed to microorganisms that can also produce plant hormones (Wang et al., 1982; Pegg, 1985). (2) Stimulation of the normal process of development by exogenous supply of the hormone and/or suppression of the process by application of inhibitors. (3) Suppression of the process in mutants related to the hormone such as those deficient in the hormone or with a defective hormone response. Recent findings of many mutants of Arabidopsis, associated with root hair formation, morphology, and growth, implicate plant hormones, especially auxin and ethylene, are involved in these processes (Schiefelbein and Somerville, 1990; Baskin et al., 1992; Su and Howell, 1992; Schiefelbein et al., 1993; Galway et al., 1994; Masucci and Schiefelbein, 1994; 1996; Cernac et al., 1997; Grierson et al., 1997; Schneider et al., 1998; see also Aeschbacher et al., 1994).

The past 5 years have seen a significant increase in interest in root hair biology, especially in their cell and molecular biology. Current major areas in root hair research can be loosely grouped into four parts: cell biology, physiology, genetics, and plant–microbe interactions, though the last group encompasses aspects of the previous three (Ridge and Emons, 2000). Attempting to summarize these areas in a short chapter will give none of them justice and only repeat much of what is available in that monograph. Thus, we have decided to discuss two topics not previously covered in root hair chapters in Plant Roots (cf. Ridge, 1996), the role of hormones and the phenomenon of root hair tip molecules.

II.

ROOT HAIRS AND PLANT HORMONES

A.

The pattern of root hair growth and development is under the influence of various factors including genetic, physiological, and environmental factors (see Ridge, 1995b; Peterson and Farquhar, 1996). There are at least three criteria to be considered in whether or not a plant hormone is required for normal progress of the developmental process of root hairs: (1) Production of the hormone by root hairs or supply of the hormone from other sources to root hairs. For this, though it is not easy to demonstrate that root hairs per se produce plant hormones, we know that other root tissues can

Hormones in Root Hair Formation

Auxin and ethylene affect root hair formation. Stimulation of root hair formation by ethylene was probably first described by Borgstro¨m (1939) in Pisum sativum, Vicia faba, and Lupinus luteus (cf. Abeles et al., 1992; see also Chapter 27 by Hussain and Roberts in this volume). Masucci and Schiefelbein (1994) have demonstrated that an Arabidopsis mutant, rhd6, shows three distinct defects in root hair formation—a reduction in the 83

84

number, an overall basal shift in the site of formation, and a relatively high frequency of epidermal cells with multiple root hairs. These phenotypes can be rescued by application of exogenous indole-3-acetic acid (IAA) or 1-aminocyclopropane-1-carboxylic acid (ACC), an ethylene precursor. Ethylene-induced root hair formation in Arabidopsis has also been reported (Baskin and Williamson, 1992). Dark-grown wild-type Arabidopsis seedlings, which are largely root hairless and produce little ethylene, can develop root hairs by exogenous treatment with ethylene and ACC (Cao et al., 1999). Treatment of wild-type Arabidopsis with aminovinylglycine (AVG), an inhibitor of ACC synthase in the pathway of ethylene biosynthesis, can phenocopy the rhd6 phenotype. Abolishment of root hairs by treatments with AVG (Tanimoto et al., 1995; Masucci and Schiefelbein, 1996) or with silver nitrate, an ethylene action inhibitor (Baskin and Williamson, 1992) has also been reported for Arabidopsis. Additional supportive evidence for the involvement of ethylene/ACC in root hair production is ACCinduced ectopic root hair formation in wild-type seedlings (Tanimoto et al., 1995; Masucci and Schiefelbein, 1996). Ectopic root hairs are also present in ctr1, an ethylene-response mutant (constitutive triple responses) (Dolan et al., 1994; Tanimoto et al., 1995). Arabidopsis root hairs develop preferentially on trichoblasts overlying the cortical anticlinal cell walls, whereas nonhair cells develop on atrichoblasts overlying the cortical periclinal cell walls. Since in etiolated Arabidopsis, ethylene and ACC can induce root hairs on trichoblasts, and ethylene-overproducing mutants (eto1, eto2, eto3, abd, eto4) preferentially develop root hairs on trichoblasts, these cells should be more sensitive to ethylene than atrichoblasts (Cao et al., 1999). Exogenous auxins such as IAA and 2,4dichlorophenoxyacetic acid (2,4-D), can induce ethylene production but do not cause ectopic root hair formation. In addition, the fact that the number of root hairs in the ethylene response mutants etr1 and ein2 are normal, suggests that ethylene may not be necessary for absolute numbers of root hairs (Masucci and Schiefelbein, 1996). Lack of ectopic root hair formation due to ACC treatment has also been reported in wild-type and etr1 Arabidopsis seedlings (Pitts et al., 1998). Masucci and Schiefelbein (1996) have suggested that ethylene acts independently of TTG and GL2, positive regulators of trichome development (Hu¨lskamp et al., 1994), and possibly downstream of those genes. Cao et al. (1999) have also come to the same conclusion by a study with CTR1 gene, a negative regulator of the

Ridge and Katsumi

ethylene response (Kieber et al., 1993; Roman et al., 1995). An increase in root hair number by kinetin treatment has also been reported in Raphanus sativus (Bittner and Buschmann, 1983). B.

Hormones in Root Hair Elongation

1.

Auxins/Ethylene

Elongation of root hairs of Agrostis alba is stimulated by a low concentration of 0.1 pM IAA that inhibited at high concentrations (0.1–1 mM) (Jackson, 1960). Bates and Lynch (1996) reported that root hair elongation of the axr2 mutant of Arabidopsis is stimulated by low phosphorus availability, but an auxin antagonist, 2-(p-chlorophenoxy)-2-methylpropionic acid inhibits root hair elongation in low-phosphorus plants. Concomitantly, application of exogenous IAA stimulated root hair elongation in the wild-type high-phosphorus plants. If root hairs require auxin for normal growth, the level of endogenous auxin must be extremely low. This might be confusing since Pitts et al. (1998) obtained promotion of root hair growth of Arabidopsis with a relatively high concentration (10 nM to 10 mM) of 2,4-D. However, since they measured the effect 7 days after treatment, the root hairs might have adapted to the high concentration. If the effect would have been measured soon after the treatment, the effect could have been inhibitory. One of the major effects of auxin at subcellular levels, which leads to elongation of ordinary cells, is cell wall loosening due to the acidification of the cell wall matrix via proton secretion from the cytosol into the apoplast (Cleland, 1987; Cosgrove, 1986, 1993; see also Chapter 30 by Pilet in this volume). Though root hairs elongate by tip growth (see Ridge, 1993), it is possible that auxin controls root hair elongation through a similar mechanism. It has been demonstrated that tip-growing cells like pollen tubes and root hairs require a continuous influx of Ca2+ at the growing tip and an internal Ca2+ gradient with higher levels at the tip for their continued elongation (Reiss and Herth, 1979; Clarkson et al., 1988; Schiefelbein et al., 1992; de Ruijter et al., 1998; see also Chapter 31 by Poovaiah et al. in this volume). Auxin is known to affect cytoplasmic streaming in various types of plant cells including root hairs (see Ayling et al., 1994). Cytoplasmic streaming in Avena root hairs, is stimulated by low concentrations of auxin, though it is inhibited at high concentrations (Sweeney, 1944). A similar effect of IAA has also

Root Hairs

been shown by Ayling et al. (1994). IAA (10–100 nM) stimulates cytosolic streaming in tomato root hairs within 1–2 min after application, whereas concentrations of IAA > 1
M are inhibitory. Inhibition of streaming by a high concentration (10 mM) of the synthetic auxinNAA, has also been reported for root hairs of Hydrocharis dubia (Tominaga et al., 1998). Such inhibition was shown to be a result of acidification of the cytoplasm that disturbed the orientation of actin filaments and disrupted cortical microtubules. On the other hand, Tretyn et al. (1991) reported that auxin regulates cytoplasmic Ca2+ levels and pH in Sinapis alba root hairs through changes in membrane potential. They suggested that a high concentration (100 nM) of IAA inhibits root hair growth by depolarizing root hairs to diminish Ca2+ gradient. But, it is argued that auxin-induced changes in Ca2+ levels and in pH are too fast to correlate with auxin-induced stimulation of cytoplasmic streaming (Ayling et al., 1994). Cytoplasmic streaming is not indispensable for ordinary cell elongation, because cytochalasin B, which inhibited cytoplasmic streaming completely, did not entirely inhibit auxin-induced growth of Avena or maize coleoptiles (Cande et al., 1973). Recent studies on the molecular biology of Arabidopsis have revealed some auxin/ethylene related genes associated with root hair growth and development. Many auxin mutants of Arabidopsis have shorter root hairs than those of the wild-type. The aux1 (Okada and Shimura, 1995; Pitts et al., 1998), axr1 (Lincoln et al., 1990; Pitts et al., 1998), and axr2 and axr3 (Leyser et al., 1996) genes all affect root hair growth. The axr1 mutants of Arabidopsis have reduced numbers of root hairs (Cernac et al., 1997). On the other hand, a normal number of root hairs was found for the weak axr1-3 allele (Masucci and Schiefelbein, 1996). Pitts et al. (1998) examined this in detail and found that the discrepancy is due to the methods of measurement they employed; the lengths of the root hairs of axr1-12 and axr1-7 were clearly shorter than wild-type hairs, while axr1 mutations had little effect on root hair initiation. In the double mutants of sar1 and axr1-12 the length of root hairs was the same as that of the wild-type (Cernac et al., 1997; Pitts et al., 1998). SAR1 has been shown to act after AXR1 in the pathway of auxin response (Cernac et al., 1997). Pitts et al. (1998) have found that both 2,4-D, and ACC can stimulate root hair elongation of the wildtype, and of mutants deficient in auxin response (aux1) or in ethylene response (etr1) of Arabidopsis thaliana. Nevertheless, these mutations have no significant effect

85

on root hair initiation (Masucci and Schiefelbein, 1996). Since the AUX1 gene product was suggested to be an auxin flux carrier in Arabidopsis roots (Bennett et al., 1996), shorter root hairs of axr1 mutant might be attributed to a defect in auxin supply to trichoblasts which is indispensable for root hair elongation. On the other hand, the AXR1 gene product is similar to the ubiquitin-activating enzyme, E1, which is necessary for the activation of Rub1 protein in Arabidopsis (del Pozo et al., 1998). Modification of nuclear proteins in the root hair by Rub is necessary for root hair growth, and auxin probably stimulates root hair elongation via this AXR1 pathway. Whatever the mechanism is, the presence of auxinbinding protein in root hairs of Zea mays (Radermacher and Kla¨mbt, 1993) might indicate that auxin has some controlling role in root hair growth and development. Indeed, Baskin and Williamson (1992) have reported that ethylene increases root hair growth in Arabidopsis seedlings. 2.

Gibberellin

Application of GA3 had practically no growth stimulating effect on clover root hairs, in a concentration range from 1 nM to 10
M. However, uniconazole (10
M), an inhibitor of gibberellin biosynthesis, markedly inhibited root hair growth, suggesting that endogenous gibberellin was involved in the normal growth of clover root hairs (Izumo et al., 1995). Exogenous GA3 (0:1
M) could completely overcome the uniconazole-induced growth inhibition (Izumo et al., 1995). These results suggest that normal root hairs are dose saturated with endogenous gibberellin, and that the effective gibberellin concentration is very low. The results are also consistent with the finding of Tanimoto (1987, 1991) obtained by similar experiments using another gibberellin biosynthesis inhibitor, ancymidol. The conclusion is that roots require a very low level of endogenous gibberellin for normal growth (see Chapter 24 by Tanimoto in this volume). Gibberellininduced stimulation of root hair elongation has also been reported for Pisum sativum (Pecket, 1960) and Agrostis alba (Devlin and Brown, 1969). It is very likely that gibberellin is required for normal growth of root hairs, but in very small quantities. Gibberellin controls cortical microtubule orientation, thus affecting the arrangement of cellulose microfibrils (Ishida and Katsumi, 1991, 1992; Shibaoka, 1994). This effect correlates with GA-induced cell elongation (Ishida and Katsumi, 1991). Since in root hairs

86

microtubules are associated with tip growth (Emons and Wolters-Arts, 1983; Emons et al., 1990; Ridge, 1995a; Bibikova et al., 1999; Ridge and Emons, 2000), GA-involved control of root hair growth might be related to this cytoskeleton protein. 3. Cytokinin/Abscisic Acid/Brassinosteroid Roots of Arabidopsis seedlings are very sensitive to cytokinins. By treatment with 5 nM to 10
M benzyladenine (BA), the growth of primary roots was reduced, while that of root hairs was stimulated (Su and Howell, 1992). Su and Howell (1992) have isolated cytokinin response mutants (ckr1) of Arabidopsis which have longer primary roots and shorter root hairs than the wild-type in the absence of BA. Growth of root hairs of ckr1 was slightly stimulated by BA and reached to the same length as that of the control wild-type in the presence of 50–500 nM BA. In normal Arabidopsis seedlings, primary roots, and root hairs may counterresponse the endogenous cytokinins (Su and Howell, 1992). It is likely that cytokinin is involved in root hair growth. Since cytokinin can induce ethylene production (Abeles et al., 1992; see also Chapter 25 by Emery and Atkins in this volume), some of the cytokinin effects observed might be ascribed to ethylene. Endogenous ABA is known to mediate plant water stresses (Davies and Mansfield, 1983; Hartung and Davies, 1992; see also Chapter 26 by Hose et al. in this volume). The effect is expressed in a significant reduction in overall growth. Worrall and Roughley (1976) and Vartanian (1981) reported that drought stress caused existing root hairs to become fragile and thin walled, and produced short and swollen new root hairs in Trifolium subterraneum and Sinapis alba, respectively. This suggests that ABA is involved in the formation of deformed root hairs. Schnall and Quatrano (1992) have demonstrated that treatment of Arabidopsis with ABA resulted in the formation of similar deformed root hairs. But ABA had no significant effect on the total number of root hairs. However, the ABA-insensitive mutants abi1 and abi2 did not respond to exogenous ABA (no deformed root hairs). Nevertheless, an ABA-sensitive mutant, abi3, displayed a response similar to that of the wild-type (Schnall and Quatrano, 1992). It is unlikely that ABA is required for normal growth and development of root hairs. 4. Concluding Remarks About Hormones Cumulative data in the literature provide strong evidence that plant hormones are involved in the process

Ridge and Katsumi

of root hair development and elongation. Findings of various auxin/ethylene-associated mutants in Arabidopsis and their molecular biological studies have demonstrated that auxin and ethylene, in particular, are important endogenous factors which control not only the elongation but also the development of root hairs. Cytokinin, gibberellin, and brassinolide may also endogenously influence these processes, though evidence is still circumstantial. The mechanism of the hormonal control of root hair elongation is not clearly understood. Whatever it is, it must be different from the mechanism of ordinary cell elongation which depends on cell wall loosening, because root hair elongation proceeds by tip growth.

III.

MOLECULES AT THE ROOT HAIR TIP SURFACE

The first reports of distinct molecules at the tips of root hairs were in the middle to late 1980s (Werner and Kuhlmann, 1985; Diaz et al., 1986; Ridge and Rolfe, 1986; Diaz, 1989) (Fig. 1). Since then, there have been only a few publications directly addressing the role of known tip molecules or announcing the discovery of new ones. Indeed, the few molecules known to be specific for the tips of root hairs are lectins, sugars, glycoproteins, and a spectrinlike molecule. Extensin is probably present, though not visualized at the tip (Arsenijevic-Maksimovic et al., 1997). A.

Lectins at the Tips of Root Hairs

The role of lectins in legume nodulation was recently reviewed (Hirsch, 1999). Lectin on the surface of pea root hairs was detected, using an antibody against pea seed lectin (Diaz et al., 1986). Fluorescent-labeled lectins of root hair tips were found to be distributed in dense small patches rather than uniformly over the tip. Diaz et al. (1986) also showed that lectins are not present on all hairs, and gave evidence that lectin labeling of tips was detected predominantly opposite protoxylem poles. The significance of this finding is that early development of symbiotic nodules is usually opposite the protoxylem poles of the young root, and that infectivity of root hairs (as shown by the presence of lectin) is enhanced by unknown (hormonal?) factors transported from within the root. In the zone of mature hairs, where lectins are rarely found in hairs, inoculation with rhizobia produced few nodules compared to the immature root hair zone. The latter had high levels of lectin labeling and produced many nodules. This

Root Hairs

87

Figure 1 Lectin binding to root hair tips demonstrating the presence of different sugar moieties at the tip surface. (A) Macroptilium atropurpureum (Leguminaceae) labeled with MPA-FITC (from Maclura pomifera, haptens: N-acetyl galactosamine, -D-galactose). (B) Rafanus sp. (Brassicaceae) labeled with RCA-1-FITC (from Ricinus communis, hapten: -D-galactose). (From Ridge et al., 1998.)

proves that root lectins are likely to play a role in the early stages of nodulation. Further work in this area showed that lectins were involved in host specificity of nodulation (Diaz et al., 1989a,b, 1995). The pea lectin gene was transformed into white clover plants, which were then able to be nodulated by Rhizobium leguminosarum bv. viciae. These investigators also altered the carbohydrate-binding domain of the pea lectin by sitedirected mutagenesis and when this mutant gene was introduced into white clover plants, the Vicia rhizobia failed to nodulate them. More recently, Van Rhijn et al. (1998) showed similar results by transforming Lotus with soybean lectin gene. Soybean lectin was detected at the tips of young growing root hairs of transgenic Lotus plants. Inoculation with Bradyrhizobium japonicum resulted in nodulation of Lotus, which is normally not nodulated by B. japonicum. An interesting additional result of their work showed that exopolysaccharide of the bacteria is likely to be involved in early interactions with the root hair (see also Hirsch, 1999). This suggests that lectins in the root hair tip may directly bind the

bacteria to the root hair tip at the onset of the earliest interactions between plant and rhizobia. Certainly it is known that binding is an essential prerequisite for the colonization and infection of many kinds of tissue (Beuth et al., 1996), and also that lectins mediate many biological recognition processes (Sharon and Lis, 1995). A lectin detected on root hair surfaces carrying enzyme activity similar to apyrase (ATP diphosphohydrolase) and that binds Nod factors has also been reported, although it was not detected solely at root hair tips (Etzler et al., 1999). The lectin was called LNP by the authors (lectin nucleotide phosphohydrolase). Etzler et al. (1999) reported that this lectin binds Nod factor and that an antibody to LNP inhibits the ability of rhizobia to deform root hairs and to form nodules. However, their results are limited by single determination points in their Nod factor-binding experiments, difficult-to-interpret micrographs, and lack of explanation of some of their techniques. This is a potentially exciting result, but we must await confirmation from other laboratories.

88

B.

Ridge and Katsumi

Sugar Molecules at the Tips of Root Hairs

Lectins are widely used as tools for detecting specific sugar molecules, particularly on cell surfaces. Ridge and Rolfe (1986) found that RCA-I lectin (from Ricinus communis, specific for -D-galactose) bound to the root hair tips of siratro (Macroptilium atropurpureum), a cowpea-type legume known to be nodulated by several groups of rhizobia. Preincubation of roots, using a spot inoculation technique in the most susceptible zone (smallest emerging root hair zone) of the root for infection, showed that lectin bound to root hair tips prevented infection and nodulation by rhizobia. The authors concluded that carbohydrate molecules at the surface of the root hair tips could be directly involved in the initiation of the symbiosis. More recent work (Ridge et al., 1998) has shown that the broad host range of siratro has a multiplicity of sugars at its root hair tips, whereas narrow host range plants such as clover and alfalfa have usually one (RCA-I binding), sometimes two kinds of sugars at their tips. The authors also showed that some nonlegumes have sugars at their tips, though their study was limited to the Brassicaceae. Such a defined area of sugar molecules at the tips of root hairs suggests that the molecules are probably the sugar moieties of membrane-bound glycoproteins, Apparently they are recycled at the tip, where there is a large presence of clathrin at the base of the apical dome. Otherwise they would be detectable below the apical dome. Indeed, in the case of one lectin, VFA, which is specific for -Dmannose, this is true. In addition the results show that there may be a conserved sugar expression (or default expression) of one sugar, -D-galactose, present on narrow host range legumes and on one member of the Brassicaceae. More recently, Samaj et al. (1999) have shown specific localization of arabinogalactan-protein epitopes at the surface of maize root hairs, where strongest expression was at the bulge of root hair initiation (cf. Ridge and Rolfe, 1986) and at the tips of growing hairs. Samaj et al. (1999) also agreed with previous suggestions that such glycoproteins may be involved in plant-microbe interactions (by signaling events). What is the significance of sugar molecules at the tips of root hairs? In the case of root hairs that apparently have several kinds of sugars, it is asked whether they are the same glycoprotein molecule, or is each on different molecules, or variations thereof? Only recently it was begun to be understood how plants

use the structural diversity of oligosaccharides to regulate important cellular processes such as growth, development, and defense (John et al., 1997). These oligosaccharides are known to be biologically active at extremely low concentrations. In the case of sugar molecules at the tips of root hairs, especially those in legumes, it is imperative to understand if they are involved in early interactions with microorganisms in the rhizosphere. Indeed, the earlier work of Ridge and Rolfe (1986) showed that by blocking the sugar sites of siratro root hair tips, nodulation was prevented. Etzler et al. (1999) have recently reported a similar result whereby an antiserum to a lectin at the Dolichos root hair surface prevented nodulation by rhizobia. Despite the limited amount of information available, this is an important area of research for root hairs, because distinct molecules at the tips suggest many possibilities: control of polarity, communication to the rhizosphere through receptor activity, passive and active plant pathogen interactions. Indeed, the very nature of the growing tip of this cell, where the cell wall is expanding in direct exposure to the rhizosphere, makes it a likely site for many kinds of interactions. In this connection, it has been shown that in a nodulation-minus soybean, the stimulation of cell division in the cortex of the root by the synthetic auxin 2,4-D and the consequent cell wall expansion, allowed rhizobia-induced infection thread development directly in these cells rather than in root hairs (Akao et al., 1991, 1999) with a subsequent successful symbiosis. This means that for the wildtype soybean, invasion of root hairs by rhizobia requires an expanding cell wall rather than tip growth per se, and that symbiotic-specific communication molecules or receptors, if they exist, can also be expressed in other cells.

REFERENCES Abeles FB, Morgan PW, Salveit ME. 1992. Ethylene in Plant Biology. New York; Academic Press. Aeschbacher RA, Schiefelbein JW, Benfey PN. 1994. The genetic and molecular basis of root development. Annu Rev Plant Physiol Plant Mol Biol 45:25–45. Akao S, Nakata S, Yoneyama T. 1991. Formation of nodules on non-nodulating soybean T201 with treatment of 2,4-D. Plant Soil 138:207–212. Akao S, Minakawa Y, Taki H, Khan MK, Yuhashi K-I, Nakayama Y, Asis CA Jr, Chebotar V, Kang U-G, Minamisawa K, Ridge RW. 1999. Use of lacZ and gusA reporter genes to trace the infection process of nitrogen-fixing bacteria. Jpn Agric Res Q 33:77–84.

Root Hairs Arsenijevic-Maksimovic I, Broughton WJ, Krause A. 1997. Rhizobia modulate root-hair-specific expression of extensin genes. MPMI 10:95–101. Ayling SM, Brownlee, C, Clarkson, DT. 1994. The cytoplasmic streaming response of tomato root hairs to auxin; observations of cytosolic calcium levels. J Plant Physiol 143:184–188. Baskin TI, Williamson R. 1992. Ethylene, microtubules and root morphology in wild-type and mutant Arabidopsis seedlings. Curr Top Plant Biochem Physiol 11:118– 130. Baskin TI, Betzner AS, Hoggart R, Cork A, Williamson RE. 1992. Root morphology mutants in Arabidopsis thaliana. Aust J Plant Phsyiol 19:427–437. Bates JR, Lynch JP. 1996. Stimulation of root hair elongation in Arabidopisis thaliana by low phosphorous availability. Plant Cell Environ 19:525–538. Bennett MJ, Marchant A, Green HG, May ST, Ward SP, Millner PA, Walker AR, Schulz B, Feldmann KA. 1996. Arabidopsis AUX1 gene: a permease-like regulator of root gravitropism. Science 273:948–950. Beuth J, Stoffel B, Pulverer G. 1996. Inhibition of bacterial adhesion and infections by lectin blocking. Adv Exp Med Biol 408:51. Bibikova TN, Blancaflor FB, Gilroy S. 1999. Microtubules regulate tip growth and orientation in root hairs of Arabidopsis thaliana. Plant J 17:657–665. Bittner A, Buschmann C. 1983. Uptake and translocation of K+, Ca2+ and Mg2+ by seedlings of Raphanus sativus L. treated with kinetin. Z Pflanzenphysiol 109:181– 189. Borgstro¨m G. 1939. Theoretical suggestions regarding the ethylene responses of plants and observations on the influence of apple-emanations. Kungl Fisiografiska Sallskapets I Lund Forhandlingar 9:135–174. Cande WZ, Goldsmith MHM, Ray PM. 1973. Polar transport and auxin-induced elongation in the absence of cytoplasmic streaming. Planta 111:279–296. Cao XF, Linstead P, Berger F, Kieber J, Dolan L. 1999. Differential ethylene sensitivity of epidermal cells is involved in the establishment of cell pattern in the Arabidopsis root. Physiol Plant 106:311–317. Cernac A, Lincoln C, Lammer D, Estelle M. 1997. The SAR1 gene of Arabidopsis acts downstream of the AXR1 gene in auxin response. Development 124:1583–1591. Clarkson DT, Brownlee C, Ayling SM. 1988. Cytoplasmic calcium measurments in intact higher plant cells: results from fluorescence imaging of fura-2. J Cell Sci 91:71–80. Cleland R. 1987. The mechanism of wall loosening and cell extension. In: Cosgrove DJ, Knievel DP, eds. Physiology of Cell Expansion During Plant Growth. Madison, WI: American Society of Plant Physiology, pp 18–27. Cosgrove DJ. 1986. Biophysical control of plant cell growth. Annu Rev Plant Physiol 37:377–405.

89 Cosgrove DJ. 1993. How do plant cell walls extend? Plant Physiol 102:1–6. Davies WJ, Mansfield TA. 1983. The role of ABA in drought avoidance. In: Addicott, FT, ed. Abscisic Acid. New York: Praeger, pp 589–614. de Ruijter NCA, Rook MB, Bisseling T, Emons AMC. 1998. Lipochito-oligosaccharides re-initiate root hair tip growth in Vicia sativa with high calcium and spectrin-like antigen at the tip. Plant J 13:341–350. del Pozo JC, Timte C, Tan S, Callis J, Estelle M. 1998. The ubiquitin-related protein RUB1 and auxin response in Arabidopsis. Science 280:1760–1763. Devlin RM, Brown DP. 1969. Effect of gibberellic acid on the elongation rate of Agrostis alba root hairs. Physiol Plant 22:759–763. Diaz C. 1989. Root lectin as a determinant of host-plant specificity in the Rhizobium–legume symbiosis. PhD thesis, University of Leiden, Netherlands. Dı´ az CL, Van Spronsen PC, Bakhuizen R, Logman GJJ, Lugtenberg BJJ, Kijne JW. 1986. Correlation between infection by Rhizobium leguminosarum and lectin on the surface of Pisum sativum L. roots. Planta 168:350–359. Diaz C, Melchers L, Hooykaas P, Lugtenberg B, Kijne J. 1989a. Root lectin as a determinant of host-plant specificity in the Rhizobium-legume symbiosis. Nature 338:579–581. Diaz C, Van Spronsen P, Bakhuizen R, Longman G, Lugtenberg B, Kijne J. 1989b. Correlation between infection by Rhizobium leguminosarum and lectin on the surface of Pisum sativum L. roots. Planta 168:530–539. Diaz CL, Logman TJJ, Stam HC, Kijne JW. 1995. Sugarbinding activity of pea lectin expressed in white clover hairy roots. Plant Physiol 109:1167–1177. Dolan L, Duckett C, Grierson C, Linstead P, Schneider K, Lawson E, Dean C, Poethig S, Roberts K. 1994. Clonal relationships and cell patterning in the root epidermis of Arabidopsis. Development 120:2465– 2474. Emons AMC, Wolters-Arts A. 1983. Cortical microtubules and microfibril deposition in the cell wall of root hairs of Equisetum hyemale. Protoplasma 117:68–81. Emons AMC, Wolters-Arts A, Traas J, Derksen J. 1990. The effect of colchicine on microtubules and microfibrils in root hairs. Acta Bot Neerl 39:19–27. Etzler ME, Kalsi G, Ewing NN, Roberts NJ, Day RB, Murphy JB. 1999. A nod factor binding lectin with apyrase activity from legume roots. Proc Natl Acad Sci USA 96:5856–5861. Galway M, Masucci J, Lloyd A, Walbot V, Davis R, Schiefelbein J. 1994. The TTG gene is required to specify epidermal cell fate and cell patterning in the Arabidopsis root. Dev Biol 166:740–754. Grierson GS, Roberts K, Feldmann KA, Dolan L. 1997. The COW1 locus of Arabidopsis acts after RHD2, and in

90 parallel with RHD3 and TIP1, to determine the shape, rate of elongation and number of root hairs produced from each site of hair-formation. Plant Physiol 115:981–990. Hartung W, Davies WJ. 1992. Drought-induced changes in physiology and ABA. In: Davies WJ, Jones HG, eds. Abscisic Acid Physiology and Biochemistry. Oxford, UK: BIOS Sci Publ, pp 63–79. Hirsch AM. 1999. Role of lectins (and rhizobial exopolysaccharides) in legume nodulation. Curr Opin Plant Biol 2:320–326. Hu¨lskamp M, Misera S, Ju¨rgens G. 1994. Genetic dissection of trichome cell development in Arabidopsis. Cell 76:555–566. Ishida K, Katsumi M. 1991. Immunofluorescence microscopical observation of cortical microtubule arrangement as affected by gibberellin in d5 mutant of Zea mays L. Plant Cell Physiol 32:409–417. Ishida K, Katsumi M. 1992. Effect of gibberellin and abscisic acid on the cortical microtubule orientation in hypocotyl cells of light-grown cucumber seedlings. Int J Plant Sci 153:155–163. Izumo M, Ridge RW, Katsumi M. 1995. Hormonal control of root hair growth in clover. In: Abstracts of the 15th International Conference on Plant Growth Substances, Minneapolis, p 452. Jackson WT. 1960. Effect of indoleacetic acid on rate of elongation of root hairs of Agrostis alba L. Physiol Plant 13:36–45. John M, Rohrig H, Schmidt J, Walden R, Schell J. 1997. Cell signalling by oligosaccharides. Trends Plant Sci. 2:111– 115. Kauss H. 1987. Some aspects of calcium-dependent regulation in plant metabolism. Annu Rev Plant Physiol 38:47–72. Kieber JJ, Rothenberg M, Roman G, Feldman KA, Ecker JR. 1993. CTR1, a negative regulator of the ethylene response pathway in Arabidopsis, encodes a member of the Raf family of protein kinases. Cell 72:427–441. Leyser HMO, Pickett FB, Dharmaseri S, Estelle M. 1996. Mutations in the AXR3 gene of Arabidopsis result in altered auxin responses including ectopic expression of the SAUR-ACI promoter. Plant J 10:403–414. Lincoln C, Britton JH, Estelle M. 1990. Growth and development of the AXR1 mutants of Arabidopsis. Plant Cell 2:1071–1080. Masucci JD, Schieflbein JW. 1994. The rhd6 mutation of Arabidopsis thaliana alters root hair initiation through an auxin-and ethylene-associated process. Plant Physiol 106:1335–1346. Masucci JD, Schieflbein JW. 1996. Hormones act downstream of TTG and GL2 to promote root hair outgrowth during epidermis development in the Arabidopsis root. Plant Cell 8:1505–1517. Okada K, Shimura Y. 1995. Modulation of root growth by physical stimuli. In: Cold Spring Harbor Laboratories,

Ridge and Katsumi eds. Arabidopsis. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, pp 665–684. Pecket RC. 1960. Effects of gibberellic acid on excised pea roots. Nature 185:114–115. Pegg GF. 1985. Pathogenic and nonpathogenic microorganisms and insects. In: Pharis RP, Reid DM, eds. Hormonal Regulation of Development III. Berlin; Springer-Verlag, pp 599–624. Peterson L, Farquhar ML. 1996. Root hairs: Specialized tubular cells extending root surfaces. Bot Rev 62:1–40. Pitts RJ, Cernac A, Estelle M. 1998. Auxin and ethylene promote root hair elongation in Arabidopsis. Plant J 16:553–560. Radermacher E, Kla¨mbt D. 1993. Auxin-dependent growth and auxin-binding proteins in primary roots and root hairs of corn (Zea mays L.). Plant Physiol 141:698– 703. Reiss H, Herth W. 1979. Calcium gradients in tip growing plant cells visualized by chlorotetracycline fluorescence. Planta 146:615–621. Ridge RW. 1993. A model of legume root hair growth and Rhizobium infection. Symbiosis 14:359–373. Ridge RW. 1995a. Microvesicles, pyriform vesicles and macro-vesicles associated with the plasma membrane in the root hair of Vicia hirsuta after freeze-substitution. J Plant Res 108:363–368. Ridge RW. 1995b. Recent developments in the cell and molecular biology of root hairs. J Plant Res 108:399–405. Ridge RW. 1996. Root hairs: cell biology and development. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. 2nd ed. New York; Marcel Dekker, pp 127–147. Ridge RW, Emons AMC. 2000. Root Hairs: Cell and Molecular Biology. Tokyo; Springer-Verlag. Ridge RW, Rolfe BG. 1986. Lectin binding to the root and root hair tips of the tropical legume Macroptilium atropurpureum Urb. Appl Environ Microbiol 51:328–332. Ridge RW, Kim R, Yoshida F. 1998. The diversity of lectindetectable sugar residues on root hair tips of selected legumes correlates with the diversity of their host ranges for rhizobia. Protoplasma 202:84–90. Roman G, Lubarsky B, Kieber JJ, Rothenberg M, Ecker JR. 1995. Genetic analysis of ethylene signal transduction in Arabidopsis thaliana: five novel mutant loci integrated into a stress response pathway. Genetics 139:1393–1409. Samaj J, Braun M, Balus˘ ka F, Ensikat H-J, Tsumuraya Y, Volkmann D. 1999. Specific localisation of arabinogalactan-protein epitopes at the surface of maize root hairs. Plant Cell Physiol 40:874–883. Schiefelbein J, Galway M, Masucci J, Ford S. 1993. Pollen tube and root-hair tip growth is disrupted in a mutant of Arabidopsis thaliana. Plant Physiol 103:979–985. Schiefelbein JW, Somerveille C. 1990. Genetic control of root hair development in Arabidopsis thaliana. Plant Cell 2:235–243.

Root Hairs Schiefelbein JW, Shipley A, Rowse P. 1992. Calcium influx at the tip of growing root-hair cells of Arabidopsis thaliana. Planta 187:455–459. Schnall JA, Quatrano RS. 1992. Abscisic acid elicits the water-stress response in root hairs of Arabidopsis thaliana. Plant Physiol 100:216–218. Schneider K, Mathur J, Boudonck K, Wells B, Dolan L, Roberts K. 1998. The ROOT HAIRLESS 1 gene encodes a nuclear protein required for root hair initiation in Arabidopsis. Genes Dev 12:2013–2021. Sharon N, Lis H. 1995. Lectins—proteins with a sweet tooth: function in cell recognition. Essays Biochem 30:59. Shibaoka H. 1994. Plant hormone-induced changes in the orientation of cortical microtubules: alterations in the cross-linking between microtubules and the plasma membrane. Annu Rev Plant Physiol Plant Mol Biol 45:527–544. Su W, Howell S. 1992. A single genetic locus, Ckr1, defines Arabidopsis mutants in which root growth is resistant to low concentrations of cytokinin. Plant Physiol 99:1569–1574. Sweeney BM. 1944. The effect of auxin on protoplasmic streaming in root hairs of Avena. Am J Bot 31:78–80. Tanimoto E. 1987. Gibberellin-dependent root elongation in Lactuca sativa: recovery from growth retardant-suppressed elongation with thickening by low concentration of GA3. Plant Cell Physiol 28:963–973. Tanimoto E. 1991. Gibberellin requirement for the normal growth of roots. In: Takahashi N, Phinney BO, MacMillan J, eds. Gibberellins. New York; SpringerVerlag, pp 229–240.

91 Tanimoto M, Roberts K, Dolan L. 1995. Ethylene is a positive regulator of root hair development in Arabidopsis thaliana. Plant J 8:943–948. Tominaga M, Sonobe S, Shimmen T. 1998. Mechanism of inhibition of cytoplasmic streaming by auxin in root hair cells of Hydrocharis. Plant Cell Physiol 39:1342– 1349. Tretyn A, Gottfried W, Felle HH. 1991. Signal transduction in Sinapis alba root hairs: auxins as external messengers. J Plant Physiol 139:187–193. Van Rhijn P, Goldberg RB, Hirsch AM. 1998. Lotus corniculatus nodulation specificity is changes by the presence of a soybean lectin gene. Plant Cell 10:1233– 1249. Vartanian N. 1981. Some aspects of structural and functional modifications induced by drought in root systems. In: Brouwer R, Gasparikova O, Kolek J, Loughman BC, eds. Structure and Function of Plant Roots. The Hague; Maritinus/Nijhoff/Dr. W Junk, pp 309–318. Wang TL, Wood EA, Brewin NJ. 1982. Growth regulators, Rhizobium and nodulation in pea. Indole-3-acetic acid from the cultured medium of nodulating and nonnodulating strains of R. leguminosarum. Planta 155:345–349. Werner D, Kuhlmann KP. 1985. Legume root hair proteins. In: Magnien E, de Nettancourt D, eds. Genetic Engineering of Plants and Microorganisms Important for Agriculture. Dordrecht, Netherlands: Nijhoff and Junk Publ, pp 66–67. Worrall VS, Roughley RJ. 1976. The effect of moisture stress on infection of Trifolium subterraneum L. by Rhizobium trifolii Dong. J Exp Bot 27:1233–1241.

6 Secondary Growth of Roots: A Cell Biological Perspective Nigel Chaffey IACR-Long Ashton Research Station, University of Bristol, Long Ashton, Bristol, England

I.

INTRODUCTION

body, which in the past 10–20 years has benefited greatly from the development of new techniques in cell and molecular biology. The application of those techniques, largely concentrated upon the ‘‘model plant’’ Arabidopsis thaliana (e.g., Somerville, 2000), has generated a tremendous amount of information concerning plant development. Because of problems—real or perceived—inherent in studying the cambium of trees (see Chaffey, 2001c), there has been a marked reluctance to pursue such investigations, and it is only recently that the techniques of modern cell biology have begun to be applied to this secondary meristem (e.g., Lachaud et al., 1999).

The primary apical meristems of the root and shoot produce the cells and tissue systems of the primary plant body (Cutter, 1978), such as the epidermis, cortex, and primary vascular tissues, and give rise to elongation growth. Secondary growth (synonymous with secondary thickening) is the province of the two laterally situated secondary meristems, the vascular cambium (herein simply referred to as the cambium) and the cork cambium, and results in increase in girth of the organ. Products of these two cambia together constitute the secondary plant body—secondary xylem and secondary phloem, and periderm, respectively. The terms primary and secondary indicate the order in which these two forms of growth occur; they should not be seen as an indication of their relative importance. Indeed, both are necessary for the successful growth and establishment of the majority of vascular plants, and secondary growth is essential to the development of the tree habit. Furthermore, although separated in space, both types of growth can occur at the same time. At the heart of secondary growth is the cambium. Notwithstanding the numerous reviews of this meristem (almost solely of the shoot) over the past 40 years (cf. Larson, 1994; Catesson, 1994; Lachaud et al., 1999), there has been little new cambial cell biology to review since the heyday of cambial ultrastructural studies in the 1960s and early 1970s. That contrasts markedly with the situation for the primary plant

II.

SCOPE OF THIS REVIEW

Aspects of primary growth and the biology of the primary body of roots are considered in other chapters in this volume. Background information on the anatomy and secondary growth of roots can be found in most standard plant anatomy texts. General aspects of the secondary growth of roots have previously been reviewed, most notably by Fayle (1968), Wilson (1975), and Woods (1991). As is proper, those reviews reflected the prevailing knowledge and interests of the time, and generally stood outside of the root ‘‘looking in.’’ However, since the last of those reviews was published, new aspects of the cell biology of secondary growth in roots have been studied, and it is therefore appropriate now to consider the inner workings of the 93

94

Chaffey

system. Since most is known about the ultrastructure of, and the dynamics of the cytoskeleton during, the cambial seasonal cycle and secondary vascular differentiation of the angiosperm trees, Aesculus hippocastanum L. (horse-chestnut), and Populus tremula L:  P. tremuloides Michx. (hybrid aspen), this review will concentrate upon those topics in those taxa. Accordingly, this chapter summarizes the published work of Chaffey et al. (1996, 1997a–d, 1998a, 1999, 2000a,b) and Chaffey and Barlow (2000, 2001), and the work in progress of Chaffey, Barlow, and Sundberg, and Chaffey and Barlow.

III.

identify a region of similar-looking cells—the ‘‘cambial zone’’—wherein the true cambium lies. However, it is confidently predicted that suitable markers will soon be forthcoming as a result of the current interest in the molecular biology of cambial growth and activity (e.g., Hertzberg and Olsson, 1998; Newman and Campbell, 2000). The established cambium consists of two morphologically distinct cell types: fusiform initials (which are greatly elongated in the axis of the root) and ray initials (which are cuboid) (Fig. 3). Both are thin-walled, highly vacuolate cells (Fig. 4) with the usual complement of organelles one would expect for a cell in the root (see Table 1).

SECONDARY GROWTH: A RECAP

Even the most cursory glance at a transverse section of a tree root (Fig. 1; see also Fig. 18 in Chapter 4 by Barlow in this volume) shows the essential features of its secondary vascular system (SVS), concentric tissuezones—the phloem toward the outside, the cambial zone, and the xylem—and radial files of cells that traverse those zones. Both the radial and concentric features are products of periclinal (synonyms in the literature: tangential, additive) cell division of the cambial initials (Fig. 2), and of their daughter cells. Subsequently these cambial derivatives grow and differentiate as phloem cells or xylem cells in a positiondependent manner. Several subdivisions of the tissue zones are recognized in arboreal plants. For example, Bailey (1952) identified six zones related to wood formation: the cambial zone; the zones of cell enlargement, cell maturation, mature sapwood, and sapwood-heartwood transition; and the inner core of heartwood. That those same zones are still recognized today is both a testament to the insight of those earlier workers, and an acknowledgement of how little our understanding of the SVS has advanced. As our methods of analysis and inquiry become more refined, we will probably have cause, and need, to reevaluate those zones, possibly defining them more specifically and precisely by use of ‘‘molecular markers.’’ The powerhouse that drives all this growth activity within the SVS is the cambium. Unfortunately, although the cambium undeniably exists, there are currently no known markers that allow us unambiguously to identify true cambial initial cells (e.g., Catesson, 1994). Indeed, it is still a matter of debate as to whether the cambium proper is comprized of one or more tiers of cells (cf. Catesson, 1974; Larson, 1994; Iqbal, 1995). For that reason, the best we can presently do is to

Figure 1 Light micrograph of a transverse section of the root of hybrid aspen showing the arrangement of tissues of the secondary vascular system. Key: CZ, cambial zone; f, fusiform cambial cell; F, fiber; r, ray cambial cell; SE, sieve elements; SPh, secondary phloem; SXy, secondary xylem; VE, vessel element. Scale bar ¼ 20
m.

Secondary Growth

Figure 2 Transmission electron micrograph of a transverse section of the cambial zone of the taproot of horse chestnut showing a recently periclinally divided ray cambial cell. Key: p, plastid; rw, radial wall; tw, tangential wall; v, vacuole; star, nascent division wall. Scale bar ¼ 2:5
m.

Fusiform initials give rise to the cells of the vertical, or axial, component of the SVS, which, in horse chestnut and hybrid aspen roots, are: axial parenchyma, fibers, sieve elements, companion cells, and vessel elements (Table 2; Chaffey et al., 1999). Note, however, that transverse cell divisions of axial cambial derivatives within the cambial zone are necessary to produce axial parenchyma cells in both the xylem and phloem. The ray initials produce the ray cells (Fig. 1), which, in the xylem, can be further distinguished as two subtypes (Braun, 1984). Contact ray cells, where they lie next to vessel elements, have large-diameter pits, which are mirrored by the adjacent vessel element. Isolation ray cells, which may also lie next to vessel elements, have only small, elliptical simple pits (as for fibers and xylem axial parenchyma). All phloem ray cells appear to be of the same type. Although the rays are predominantly uniseriate, biseriate rays are occasionally found in the xylem and phloem of hybrid aspen (particularly in the root). Additionally, phloem rays are subject to dilatation growth, which effectively makes the rays multiseriate in this tissue.

95

Figure 3 Light micrograph of a tangential longitudinal section through the cambial zone of the taproot of horse chestnut showing the ray (r) and fusiform (f) cambial cells; the encircled region indicates a relatively recent pseudotransverse division wall. Scale bar ¼ 20
m.

At maturity, and relative to the precursor cambial cell, all secondary vascular cells have different shapes, their cell walls are thicker, variously elaborated, and, in the case of all xylem cells and phloem fibers, also lignified (Table 2; Chaffey et al., 1999). Many ultrastructural changes also accompany secondary vascular differentiation (Iqbal, 1995).

IV.

ROOT AND SHOOT CAMBIA ARE ALIKE

Superficially, secondary growth in roots appears the same as it does in shoots, and for many years it has been almost an article of faith that root and shoot cambia were the same. However, in the absence of definitive information regarding the cell biology of root cambia, and in light of the differences in wood anatomy between roots and shoots of angiosperm trees (Patel, 1965), this view was difficult to sustain.

96

Chaffey

Figure 4 Transmission electron micrograph of a transverse section of the taproot of horse chestnut showing a file of phloem ray cells. Key: n, nucleus; r, ray cambial cell; R 0 and R 00 , successive stages in differentiation of phloem ray cells; arrow indicates direction of differentiation. Scale bar ¼ 5
m.

To my knowledge, the only work on the cell biology of tree root cambia is the series of articles by Chaffey et al. (1996, 1997a–d, 1998a, 1999, 2000a,b), which examined the ultrastructure of, and cytoskeletal dynamics within taproots of Aesculus hippocastanum. Very little difference between the cambia of roots and shoots of horse chestnut was found in the ultrastructure of ray and cambial cells, in their seasonal cycle (Table 1), and in the behavior of their microtubule and microfilament cytoskeleton, both during the cambial seasonal cycle and during differentiation of secondary xylem and phloem. The major benefit of this revelation is that shoot and root cambia do indeed appear to be similar, and information derived from one can be integrated with that derived from the other. However, lest we be tempted to ignore root cambia completely, it should be stressed that so few taxa have been studied that to conclude that all shoot and root cambia are the same may be premature. Certainly, much more work is needed to cover the range of tree and wood types, which embraces angiosperms—gymnosperms, temperate—tropical, ring-porous—diffuseporous, storied—nonstoried cambia.

Table 1 Seasonal Changes in Ultrastructure of Fusiform and Ray Cambial Cells Within Taproots of Aesculus hippocastanum L. Feature Cell walls Nucleus1 Mitochondria Plastids Microtubules2 Microfilaments2 Vacuome Endoplasmic reticulum Dictyosomes Secretory vesicles Coated vesicles PCR/TGN6 Plasmalemma Oleosomes Microbodies Ribosomes7

Active cambium Thin, amorphous Single þþ þþ (little starch) Random3 Axial4 Single, large vacuole þþ (rough only) þþ (active) þþ þþ þ Undulating þ þ þþ

Dormant cambium Thick, lamellate Single þþ þþ (more starch) Oblique and helical3 Axial4 Dispersed, small vacuoles þþ (rough mainly, some smooth5 ) þþ (less active) þþ þðþÞ þ Undulating þþ þðþÞ þþ

Key: 1 Occasionally containing microtubule-like structures; 2 cortical elements; 3 endoplasmic microtubules particularly prominent in ray cells; 4 random, but rare in ray cells; 5 smooth endoplasmic reticulum only seen in fusiform cells; 6 partially coated reticulum/trans Golgi networklike structure (Chaffey et al., 1997d); 7 frequently in polysomal configurations; þ, þðþÞ, þþ, increasing abundance.

Secondary Growth

97

Table 2 Comparison of Some of the Characteristics of Cells of the Secondary Vascular System in the Taproot of Aesculus hippocastanum Tissue/cell type Phloem Companion cell Sieve element Fiber Axial parenchyma Ray cell Cambium Ray cell Fusiform cell Xylem Fiber Axial parenchyma Vessel element Contact ray cell Isolation ray cell

Walla ;  lignified; type of elaboration

Widtha

Lengtha

Wider Wider Wider Wider Equald

Shorter Equal Longer Shorter Longere

Thicker, Thicker, Thicker, Thicker, Thicker,

N/A N/A

N/A N/A

Thin, nonlignified; pit fields only Thin, nonlignified; pit fields only

Wider Wider Wider

Longer Shorter Equal

Equald Equald

Longere Longere

Thicker, lignified; simple pits, obliquely angledc Thicker, lignified; simple pits, transversely angledc Thicker, lignified; bordered and contact pits, tertiary thickenings, simple perforations Thicker, lignified; contact pits, interray cell pits Thicker, lignified; simple pits, interray cell pits

nonlignified; pitted nonlignified; sieve areasb lignified; simple pits, obliquely angledc nonlignified; pitted nonlignified; pitted

Key: a Relative to precursor cambial cell; b at sieve plate and within lateral cell walls; c orientation of long axis of elliptical pit relative to cell’s long axis; d measured axially; e measured radially.

V.

CAMBIAL CELL DIVISION AND GROWTH

A.

The Arithmetic Cambium: Division and Multiplication

As a result of the combination of meristematic activity of the cambium and the growth of the daughter cells, the cambial cylinder moves outward. If radial growth were to continue unchecked, it is likely that the pressures so generated within the cambium and the phloem would result in these tissues being torn apart. Clearly, this does not happen; the radial expansion is somehow accommodated. The integrity of the SVS is maintained by at least three sites of cell division/growth in the radial plane, which together lead to an increase in girth. The main ‘‘pressure relief valve’’ is anticlinal (synonyms in the literature: radial, multiplicative) division of cambial initials to produce new cambial initials circumferentially, which subsequently grow and develop their own radial files of secondary vascular cells. In nonstoried cambia, such as those of horse chestnut and hybrid aspen, where the ends of adjacent cambial cells are at different levels, these anticlinal divisions vary from nearly transverse to nearly longitudinal, and are termed ‘‘pseudotransverse.’’ Because the division wall is offset from the vertical, both daugh-

ter cells are shorter than the mother cell. Thus, to increase the circumference of the cambial cylinder, not only do the daughters have to increase in diameter, but they also have to elongate to regain the original parental length. This contrasts with periclinal cell division, which is essentially vertical, longitudinal division that gives rise to two daughter cells of approximately equal length. In storied cambia, such as that of Robinia, where all fusiform cambial cells are the same length, anticlinal cell division is truly longitudinal and two equal-length daughters are produced, which only have to increase in diameter to achieve increase in cambial circumference. A second line of defense against unchecked radial expansion, but which is more pronounced in the stem, which increases in girth much more than the root, is so-called dilatation growth (e.g., Larson, 1994) of phloem rays. In hybrid aspen the typically uniseriate rays undergo anticlinal radial divisions and cell growth to broaden into multiseriate rays. Thirdly, the combination of anticlinal cell division in the cork cambium (Y. Waisel, personal communication, 2000) and cell growth also serve to relieve any buildup of pressure. The conditions that promote and sustain these anticlinal cell divisions are unknown, but it seems plausible that some sort of pressure/stretch-detection mechanism exists at the cellular level, which subsequently sets off an appropriate number of rounds of cell divi-

98

Chaffey

sion/enlargement sufficient to dissipate the immediate danger. However, a further consideration is that in dilatation growth, parenchymatous phloem ray cells, which are differentiated cambial derivatives, can subsequently be encouraged to undergo cell division. How similar is this cell division to that which takes place within the cambium proper? Do the differentiated ray cells have to dedifferentiate to a meristematic state before they can undergo division? Is this cell division also influenced by auxin flow, as considered to be the case for the cambium (e.g., Aloni et al., 2000)? If so, how is the appropriate level/flow of auxin generated in this specialized region of the phloem? The presence of a radial gradient of auxin across the tissues of the SVS in stems of angiosperm and gymnosperm trees has recently been established (Sundberg et al., 2000). Extension of this study to a consideration of root cambia, dilatation growth, and the much understudied cork cambium, should prove quite rewarding. However, mere production of more initial cells by anticlinal cell division in the cambium is not enough. The right sort of cambial initial has to be produced. A certain ratio of fusiform:ray cells needs to be maintained for proper functioning of the SVS (Larson, 1994). To maintain that ratio it will be necessary to produce more ray cells. Ray cells can generally be produced in one, or all, of three ways. Transformative cell divisions can occur in which either a part, or the whole, of a fusiform cambial cell becomes subdivided, by anticlinal, transverse cell divisions, to produce ray initial(s) (e.g., Phillips, 1976). Thirdly, a most unequal lateral cell division of a fusiform cambial cell can occur whereby a ray initial is cut off from the side of the fusiform cell (e.g., Phillips, 1976). Understanding the conditions that promote the production of ray initials represents a considerable challenge. B.

Breaking the Rules

Of the two types of cambial initials, understanding cell division of the fusiform cells is the most problematic. They are long cells (165–329
m in horse chestnut root—Chaffey et al., 1999) but can be up to several millimeters long in conifers (Larson, 1994), whose periclinal or pseudotransverse divisions ‘‘do not obey the usual laws of cell division’’ (Cutter, 1975). For example, they violate Errera’s Law (Cutter, 1978), which states that cells divide by a wall of minimal surface area. As if that were not intriguing enough, they also appear to lack a preprophase band (PPB). The PPB is a circumferential aggregation of cortical microtubules (MTs) that appears to define both the plane of subse-

quent cell division and the site of attachment on the parental cell wall of the new division wall (e.g., Wick, 1991). However, ultrastructural study of cell division in stem cambia of Robinia (Farrar and Evert, 1997) has failed to identify a PPB in fusiform initials, as has my extensive search by immunofluorescent techniques for MTs in roots of horse chestnut and hybrid aspen. If the PPB performs the role ascribed to it, how do fusiform initials manage without it? What guides the advancing cell plate to the correct place on the parental cell wall? Indeed, is the cell plate ‘‘guided’’ in the usual sense? Ultrastructural study of cell division in fusiform cambial cells of Fraxinus (Goosen-de Roo et al., 1984) identified a previously unsuspected class of MTs, within the phragmosome, which advances before the phragmoplast. They suggested that these longitudinally oriented (i.e., parallel to the developing cell wall) MTs guide the growing cell plate to the ‘‘parental cell wall site previously marked by the preprophase band of microtubules’’ (Goosen-de Roo et al., 1984). However, my own immunofluorescent work, admittedly on different species and using a different visualization technique, has not demonstrated any MTs in this location in dividing fusiform cells. Perhaps the best we can say at present is that it is uncertain whether a PPB is present in such cells, or whether such phragmosomal MTs can perform the role suggested by Goosen-de Roo et al. Of course, it might be that a PPB is unnecessary and a wide variety of daughter cell lengths can be tolerated—ranging from those that result from near-transverse to those from near-longitudinal variants of pseudotransverse cell division. The system then would be reliant upon elongation growth of both daughter cells to reconstitute the original cell length, rather than the accurate placement of the division wall at a specific site on the parental wall. On this basis, it is proposed that a PPB is only essential to cells in situations where it is important to preserve a particular cell shape or to maintain specific cell-cell geometry. An example of this in the SVS could be the retention of the regular, radial symplasmic transport route, which is provided by the xylem and phloem rays. Consistent with this suggestion, we have identified a PPB in ray cambial cells by immunolocalization techniques. Nevertheless, it is necessary to establish in a more rigorous way whether a PPB really is absent from fusiform cells. What do we currently know about the cell biology of cambial cell division in the root? To date, cytoskeletal dynamics have only been examined during periclinal cell division and the transverse cell divisions that

Secondary Growth

generate axial phloem parenchyma cells in roots. Those studies, which have so far considered immunofluorescent localization of tubulin (for MTs), actin (for microfilaments—MFs), unconventional myosin VIII (‘‘myosin’’) (characterized by Reichelt et al., 1999) and callose, have revealed that karyokinesis is accompanied by development of a mitotic spindle, which contains tubulin, but in which neither myosin nor actin have yet been detected. During subsequent cytokinesis, both tubulin and actin are immunolocalized either side of the new division wall (Fig. 5a,b, see color insert), within the phragmoplast, whereas myosin and callose are found within the wall (Fig. 5c,d). Interestingly, cell walls at this very early stage of development are apparently unstained by Calcofluor (Chaffey, 2001d), indicating that they are almost devoid of cellulose, and in agreement with earlier reports of a callosic phase in cell wall development (Waterkeyn, 1967). Whether callose is made in situ, as is assumed for subsequent cellulose synthesis, or is transported there in secretory vesicles, is not yet known. As for cell division in other cell systems, in longitudinal section it is evident that the tubulin and actin are concentrated within the phragmosomes at the two ends of the centrifugally advancing cell plate. The MTs here, which lie at right angles to the plane of the developing cell wall (Fig. 6), can be viewed as providing guides for transport of secretory vesicles to the developing cell wall, as can the associated MFs. Once vesicles arrive at the phragmoplast, the dense MT-MF array could help to trap such vesicles there, ensuring that their contents are released at the correct place. Immunolocalization of actin, and possibly tubulin, is also evident along the nascent cell wall between these two phragmoplasts, as is myosin and callose within the wall. The actin and myosin can be envisaged as acting in concert here as a motile cytoskeletal system (e.g., Reichelt and Kendrick-Jones, 2000), which could translocate those dictyosome-derived secretory vesicles that arrive at the developed cell plate to the phragmoplasts, where they can contribute to the development of the nascent division wall. Soon after this stage, the callose-staining throughout the wall (Chaffey, 2001e) becomes lost and the cell walls stain strongly with Calcofluor, indicating the presence of cellulose. Whether callose stainability disappears because the callose is replaced, or is altered in some way so that its antigenicity is masked or lost, is not known. In line with current interest in the chemistry of cambial cell walls, and the enzymes that modify it (e.g., Follet-Gueye et al., 2000; Micheli et al., 2000),

99

Figure 6 Transmission electron micrograph of a transverse section of the cambial zone of the taproot of horse chestnut showing a recently periclinally divided fusiform cambial cell undergoing cell wall formation. Key: rw, radial wall; tw, tangential wall; star, nascent division wall; arrows, microtubules. Scale bar ¼ 2:5
m.

further study of this phenomenon is clearly warranted. However, neither callose nor myosin is completely absent from the mature cell wall; both are retained at the pit fields, presumably in connection with the plasmodesmata at those sites (e.g., Radford et al., 1998; Radford and White, 1998). The association of callose with early stages of development of otherwise cellulosic structures, such as cambial division walls and the peripheries of bordered and contact pits, suggests that the deposition of this polysaccharide may be a necessary precursor to cellulose deposition. Perhaps the colocalization of callose and myosin at these sites is necessary to attract and/or retain and/or anchor the MTs or MFs that are related to further wall elaboration at those sites.

VI.

CYTOSKELETAL CONSIDERATIONS OF THE SECONDARY VASCULAR SYSTEM

Although many changes take place as cambial derivatives differentiate as secondary vascular cells (Table 2) (Iqbal, 1995; Chaffey, 2001b), alterations of their size,

100

Chaffey

shape, and walls are among the most obvious. In view of the roles ascribed to cellulose microfibrils in influencing the direction of cell expansion (cf. Green, 1969), and to the cytoplasmic MTs in influencing the arrangement of those microfibrils within the cell wall (reviewed by Giddings and Staehelin, 1991), to date the cell walls and the cytoskeleton are the most studied aspects of root cambial cell biology. Using a combination of conventional transmission electron microscopy and immunofluorescence localization of the cytoskeletal proteins tubulin (for MTs) and actin (for MFs), the changes taking place to these cell components during secondary vascular differentiation have been well characterized in taproots of A. hippocastanum (Chaffey et al., 1996, 1997a–d, 1998a, 1999, 2000a,b; summarized for wood formation in this tree in Table 3). Our immunofluorescent work in progress has now extended that study to roots of hybrid aspen. A recent review (Barlow and Balus˘ ka, 2000) which concentrates on the primary plant body also deals in detail with the MT cytoskeleton in roots.

A.

Cambium

In active cambia, the cortical MTs of both ray and fusiform cambial cells are randomly organized, and frequently overlap. Although cambial cell wall structure has not specifically been examined in horse chestnut or in hybrid aspen, the literature shows that the primary walls of secondary xylem cells—which are derived from the cambial cell wall—contain disorganized cellulose microfibrils (e.g., Awano et al., 2000). Thus, the similarly randomly arranged MTs can be seen as circumstantial evidence for the view that MTs influence the orientation of microfibrils. What is not obvious, however, is the significance of the overlapping MTs. Do they accord with corresponding production of similarly overlapping cellulose microfibrils? Or do such images record the changeover from one MT/ microfibril orientation to another? Perhaps of more immediate significance is that the random arrangement of MTs and microfibrils in both cambial cell types potentially permits growth in any direction, i.e., isotropically. However, cambial derivatives grow anisotropically, but this takes place principally along one or both of two axes—longitudinal and radial—and the combination of the two varies quite

Table 3 Behavior of the Cortical Microtubule and Microfilament Components of the Cytoskeleton During the Cambial Seasonal Cycle and Xylogenesis in the Taproot of Aesculus hippocastanum Cell type Ray cambial (active) Ray cambial (dormant) Fusiform cambial (active) Fusiform cambial (dormant) Fiber (early stages—phase I) Fiber (middle stages—phase II) Fiber (maturity) Axial parenchyma (early stages) Axial parenchyma (maturity) Vessel element (early stages) Vessel element (middle stages) Vessel element (maturity) Contact ray (early stages) Contact ray (maturity) Isolation ray (early stages) Isolation ray (maturity)

Microtubules

Microfilaments

Random Parallel-helical Random Parallel-helical Parallel-helical, wavy Parallel-helical, dense Absent Parallel-helical Axial-helical Random Various3 Absent Random, rings5 Axial-helical Random Axial-helical

Random Random Axial Axial Axial1 Axial2 Absent Axial1 Axial2 Axial Axial2 , various3;4 Absent Axial, rings5 Axial2 Axial1 Axial2

Key: 1 Ellipse associated with developing simple pits; 2 branched, but overall net-axial orientation maintained; 3 rings at peripheries of bordered and contact pits, and perforation plate, and at aperture of developing bordered pit, parallel-aligned arrays associated with developing tertiary thickenings; 4 meshwork over perforation plate; 5 rings at periphery of developing contact pit.

Secondary Growth

dramatically depending upon the cell-type. Thus, fibers grow considerably along both axes, sieve and vessel elements grow more radially than longitudinally, and ray cells grow almost solely radially. Further complications arise from the facts that ray cambial cells can ‘‘transform’’ into fusiform cambial cells (e.g., Larson, 1994), fusiform cambial cells can give rise to ray cells, daughters of pseudotransverse division of a fusiform initial must grow in both axes (although considerably more longitudinally) to regain the parental cell volume, and fusiform cambial cells can undergo transverse divisions to produce axial parenchyma cells. In other words, the randomness of the MTs and microfibrils within the cambial cells can be interpreted as an indication that these cells are as yet uncommitted to a particular cell fate. Potentially, each cambial derivative could follow one of several pathways of differentiation. Nevertheless, it is unusual to find random MTs in a cell whose shape is other than that of a sphere (cf. cultured tobacco protoplasts—Vissenberg et al., 2000b). However, as for the protoplast, which will subsequently develop new MT arrangements as it differentiates (Vissenberg et al., 2000b), the most likely explanation for the random MTs is the same in each case. By contrast, the prominent MFs within fusiform cambial cells are found in axially oriented bundles throughout the cambial seasonal cycle. In ray cambial cells, however, the MF bundles are much less common and apparently randomly oriented, although they too persist in these cells throughout cambial activity and dormancy. Nowhere in cambial cells have MFs been observed co-oriented with MTs. In both cambial cell types the MF bundles are implicated in cytoplasmic streaming during the active part of the cambial seasonal cycle (Chaffey et al., 2000a).

B.

Early Stages of Secondary Vascular Differentiation

As cambial derivatives differentiate, their MTs generally adopt new arrangements, and it is tempting to suggest that this may indicate commitment of a cambial derivative to a particular cell fate (e.g., Chaffey et al., 1997b). However, since the earliest stages of differentiation are more easily identified by the increase in cell diameter, and relocation of the nucleus to a tangential wall (e.g., Fig. 4; Chaffey et al., 1999), it is not yet possible to decide whether changes to the MT cytoskeleton are more cause or effect of the differentiation process.

101

1.

Fibers

a.

Biphasic Fiber Growth

Although both normal wood and gelatinous fibers are found within roots and shoots of hybrid aspen, only the cytoskeletal dynamics of normal wood fibers have so far been studied in such roots. Accordingly, only normal wood fibers are considered further. Two phases of development of wood fibers are apparent. Phase I commences with the establishment of a parallel alignment of MTs, but although the individual MTs themselves appear wavy and may overlap, crosslinks between MTs and the plasmalemma have not been observed. It is suggested that this phase is concerned with cell growth, hence with production of primary cell wall, and takes place whilst the cells are within the enlargement zone of the secondary xylem. The succeeding phase II shows an increase in the number of MTs relative to phase I, but they are still present as a layer that is almost exclusively one MT deep. The MTs are all at the same angle within a cell (Fig. 7a, see color insert), and in places observed to be crosslinked to the plasmalemma, but not to each other. It is suggested that this phase occurs in the maturation zone when fibers have ceased enlargement, and are undergoing the main stage of cell wall thickening. Although crosslinks between MTs and the plasmalemma have so far been seen only in ultrastructural study of hybrid aspen stem material, a similar association between MTs and plasmalemma in horse chestnut root material is concluded from study of plasmolysed fibers. In such material, the parallel alignment of MTs is still seen, despite the fact that the plasmalemma has retracted from the cell wall. Presumably, these crosslinks act to hold the MTs in a specific orientation to each other, which thereby promotes a more ordered orientation to the cellulose microfibrils (or to other components of the cell wall). It is widely acknowledged that fibers, and other wood cells, have highly orderd microfibrils in the various S(econdary) wall layers. Presence of plasmalemmal crosslinked MTs in fibers during phase II of growth, but not during phase I or in fusiform cambial cells, can be related to this apparent need for a much more ordered secondary cell wall, the better to perform its strengthening role. By contrast, for a primary cell wall, which is laid down in a cell that is still growing, the emphasis is more on flexibility and the ability of the wall to accommodate further increase in cell volume. In such a case, a less ordered microfibril arrangement will perhaps be of more use, and may indeed be favoured by the presence of non-plasmalemmal-crosslinked MTs.

102

If we accept that MT orientation has a role in determining the arrangement of cell wall microfibrils, then our observed variation in MT orientation from fiber to fiber, which is inferred to relate to successive stages of fiber differentiation, could be related to the deposition of different layers of the secondary cell wall (S1, S2, or S3), or to successive development of different lamellae within a single wall layer (Barnett et al., 1998). Clearly, further work is required to decide between these possibilities. It is noteworthy that development of phloem fibers appears to be similar to that of their xylem counterparts. Certainly, they undergo a growth phase that is the same as the phase II of xylem fibers (Fig. 7b, see color insert). However, what we don’t yet know is the full sequence of phloem fiber differentiation. This largely relates to the fact that more effort has been concentrated upon understanding xylogenesis in the SVS, but partly to the apparent inability to identify phloem fibers at earlier stages of their differentiation (e.g., Chaffey et al., 2000b). Parallels are also evident between development of xylem fibers and cotton ‘‘fibers’’ (which, strictly speaking, are seed coat trichomes). Detailed ultrastructural work by Seagull (1992) with in vitro grown cotton fibers has established that MTs are generally randomly oriented during fiber initiation and early elongation, reorienting to shallow pitched (near-transverse) helices as elongation and primary wall deposition continue. Finally they give way to steeply pitched (near longitudinal) MT helices during secondary wall deposition. Accompanying these changes were increases in MT length, number, proximity to the plasmalemma, and a decrease in variability of MT orientation. With the exception of a shallow helical phase, and in the absence of data regarding MT length, the similarity of cotton fiber and wood fiber MT dynamics and the associated phases of growth is remarkable. Although the economic importance of cotton means that research on this crop is more highly funded than that on wood fibers, in view of the similarity of the cell biology of growth between the two cell types, a bonus of the research on cotton is likely to be its relevance to a greater understanding of wood fiber biology. b. Bipolar Fiber Growth Over many years it has been established that wood fibers grow primarily, if not solely, at their tips (Larson, 1994). Although the mechanism of this tip growth is not yet known, it has been suggested that it may be facilitated by secretion of wall-degrading enzymes at the tips (Wenham and Cusick, 1975). The

Chaffey

MFs retain their net axial orientation (Figs 7c,d) throughout all phases of fiber development. This is consistent with a role in supplying precursors or secretory vesicles, either via cyclosis or by directed transport (Boevink et al., 1998) of enzymes, to the growing tips of these cells. Support for this view has been provided by the polar distribution of expansin mRNAs found in putative xylem fiber cells in stems of Zinnia elegans (Im et al., 2000). Along with xyloglucan endotransglycosylases (Vissenberg et al., 2000a), expansins are a class of enzymes that are implicated in ‘‘wall-loosening’’ events (Cosgrove, 1998), such as are necessary to facilitate tip growth of xylem fibers. Citing work of Roberts and Uhnak (1998) using the in vitro Zinnia system, in which leaf mesophyll cells are induced to differentiate as tracheary elements, Im et al. (2000) proposed that longitudinally oriented MTs might play a role in the translocation of the mRNAs to the fiber tips. In planta, in cells of the SVS of angiosperm trees, longitudinally oriented MTs have only so far been seen in gelatinous fibers of hybrid aspen (see also ultrastructural studies in Populus (Mia, 1968; Fujita et al., 1974) and Salix (Robards and Kidwai, 1972), and the immunofluorescent studies of Fraxinus (Prodhan et al., 1995). It therefore seeems more likely that the axially (i.e., longitudinally) oriented MFs are a better candidate for any polar distribution of mRNAs that may be found in angiosperm trees. This interpretation also emphasizes the need for caution in making inferences about the natural system from an in vitro one. Is the growth of fibers really ‘‘tip growth?’’ How does it compare with that in the more usual tip-growing systems of fungal hyphae, pollen tubes, root hairs, and fern protonemata (Geitman and Emons, 2000; see also Chapter 5 by Ridge and Katsumi in this volume)? The straightforward answer is that wood fiber growth has not yet been studied in sufficient detail to permit a proper comparison with other tip-growing systems. However, some observations are pertinent. Wood fibers are single cells, both ends of which grow (i.e., they undergo bipolar growth—Chaffey et al., 1999), and whose MTs (Fig. 7a) and MFs appear to extend to the tips of the cell. Other such systems are unipolar and their apical regions usually have different MT and MF distributions from the rest of the cell. The growing tip of a wood fiber is long tapered and more varied in shape than the ‘‘hemisphere’’ of, e.g., a pollen tube, and may bifurcate (Chaffey et al., 1999). Wood fibers bear wall elaborations—pits—which are absent from fungal hyphae or root hairs. During early development, wood fibers are in contact with other wood

Secondary Growth

cells via plasmodesmata; all the other tip-growing systems are essentially isolated from other cells. The wall of fibers is probably much more structurally elaborate than that of most of the other systems, and is lignified. Taken together, wood fibers have a number of features that serve to distinguish them from the unipolar tip-growing systems. However, fibers are such an important economic cell type, e.g., for wood pulp, that further study of their mode of growth is warranted, and may even contribute to a greater understanding and appreciation of more conventional tip growth. c.

Pits in Fibers

The only deviation from the axial MF array in wood and phloem fibers is seen in the association of MFs with sites of simple pit formation in these cells (Fig. 7d), where ellipses of antiactin reactive material are found. This is similar to the situation in axial and isolation ray parenchyma in both Aesculus and hybrid aspen. It may be significant that these ellipses are not seen in isolation, but are usually attached to MF bundles. Such an association might indicate that the MF bundles act as guides for the transport of enzymes of cell wall modification, or precursors for cell wall synthesis, to the site of pit development. That no specific array of MTs has been seen at these sites implies that this cytoskeletal component is not directly involved with either the siting or the development of the pits. Furthermore, the absence of MTs from the pit membrane of simple pits—and all other pit types in the secondary xylem—which remain as regions of unthickened primary wall middle lamella, is consistent with the view that presence of MTs is required for secondary wall thickening to take place (Chaffey et al., 1999). d.

Waste Not, Want Not

Notwithstanding 50 years of ultrastructural study of the SVS, new discoveries continue to be made. A good example of this is the role of plasmatubules in wood cells. Plasmatubules (PTs) (Harris et al., 1982) are small-diameter, tubular evaginations of the plasmalemma found at sites where there is an inferred, but relatively short-term, high flux between the apoplasm and the symplasm (Harris and Chaffey, 1985, 1986). During ultrastructural studies of the SVS of Aesculus root, PTs were only seen in differentiating xylem cells (Chaffey et al., 1999; Chaffey, 2001e). The plasmalemma-bound symplasmic compartment of an individual differentiating xylem cell is surrounded by a large apoplasmic space, which consists of its own cell wall, the walls of other cells, and the lumina of fully

103

differentiated fibers and vessel elements. The observed distribution of PTs can thus be related to their inferred role in symplasm–apoplasm exchange, facilitating the resorption of the products of xylem cell lysis, which might otherwise be lost, from the extracellular medium. In this way PTs act as an adjunct to the symplasmic intercellular fluxes within and between the longlived ray and axial parenchyma cells. This notion of PT-facilitated recycling receives experimental support from a study of the uptake of radioactively labeled lysine by dwarf mistletoe from its host (Coetzee and Fineran, 1989), in which PTs were implicated. Furthermore, absence of PTs from differentiating phloem cells is consistent with this suggestion for xylem, since here symplasmic routes between adjacent cells are maintained via plasmodesmata. However, even in situations within the phloem where such symplasmic continuity is lost, e.g., during the differentiation of bast fibers, PTs have been recorded (Sal’nikov et al., 1993). Elsewhere, PTs have been illustrated and discussed in the study of tracheid differentiation in Cryptomeria sp. (Takabe and Harada, 1986), and differentiating tracheary elements of Salix dasyclados (Sennerby-Forsse and Von Fircks, 1987). PTs are also evident in xylem fibers of Acer (Cronshaw, 1965) and Aesculus (Barnett, 1981). 2.

Vessel Elements

Vessel elements possess several distinct arrays of MTs and MFs in association with the numerous wall elaborations found in this cell type (Table 3); however, I shall only consider the development of the perforation. The perforation plate is a highly specialized region of the vessel element wall. It is an MT-free zone (Chaffey et al., 1999) that is devoid of callose (unpublished observations), and almost devoid of cellulose (Benayoun et al., 1981), and which neither secondarily thickens nor becomes lignified (Chaffey et al., 1999). Since removal of the perforation plate facilitates longdistance water transport throughout the tree, it has been much studied (Meylan and Butterfield, 1981; Butterfield, 1995), although the details of its formation and subsequent loss are not fully known. From a cytoskeletal point of view, we know that in roots (and shoots) of Aesculus, the periphery of the perforation plate is marked by a ring of both MTs and MFs, although in hybrid aspen, the MF ring appears to be absent, and a dense meshwork of MFs lies over the perforation plate itself. In the context of formation of the perforation, most significance is attached to the MF meshwork. In accordance with

104

the view that enzymatic dissolution may play a role in the removal of the perforation plate (Benayoun, 1983), it is suggested that transport of the secretory vesicles containing the necessary enzymes to the perforation plates is facilitated by the net axial MF bundles within these cells. However, once there, the MF meshwork effectively traps them so that their lytic cargoes are released only at the perforation plate. Related to the development of the perforation, and to the more general consideration of how the various types of pits in vessel elements and other wood cells are sited, is the question of establishment of the ‘‘perforation domain.’’ Chemically, the perforation domain is characterized by the absence of cellulose, which elsewhere is a common component of the cell wall and which was apparently present within the precursor cambial cell wall. How is the cellulose removed from this specific location? How does the subsequent organization of MFs and/or MTs at the periphery of the domain come about? Is it due to the presence of particular MT- and/or MF-associated proteins at this site? If so, how do they get there? If one considers the tip of an axial cambial derivative, it can be likened to the two sides that form the apex of a steeply angled triangle. The two sides look the same, yet only one will go on to develop a perforation. What is different between these two sides of the cell tip? One way of approaching this question is to examine these cells with a wide range of antibodies against cell surface epitopes such as the John Innes monoclonal antibodies (JIMs) (Willats et al., 2000), looking for differences between and within the walls. Both cell wall and plasmalemmal epitopes should be considered, since it is not clear if the domain is a product of one or other or both of these cell compartments. 3. Multifaceted Microfilaments A net axial orientation of MFs is retained in the fusiform cambial cells and the axial components of the secondary xylem. However, there is a pronounced increase in the degree of branching of the MF network in axial xylem cells as cells differentiate from fusiform cambial cells into fibers or into vessel elements. Thus, there appears to be a correlation between the degree of radial growth of the cell and the extent of branching of the MF network. In keeping with the notion that axial MFs are related to tip-directed cell wall growth/elaboration/modification, it is suggested that the branched MF elements are involved with delivery of enzymes (such as expansins or xyloglucan endotransglycosylases) and/or wall precursors to specific sites on

Chaffey

the side walls. There it may be necessary to break intermolecular bonds, such as xyloglucan crosslinks (Fujino et al., 2000), so facilitating wall growth, or to permit wall elaboration such as development of simple pits in fiber cells. The degree of branching can be viewed as being at a maximum for the axial cell components in the case of transversely oriented MFs that are seen in vessel elements. Although MFs appear not to have a role in the change of orientation of MTs (Chaffey et al., 1997b), their putative involvement in formation of all pit types and perforations, and in directional cell expansion suggests that they have important roles in cell differentiation generally, and secondary vascular differentiation specifically. Generally, work on MFs lags considerably behind that on MTs, particularly in plant cells. It is hoped that studies such as the one reviewed here, and a recent book devoted to MFs in plants (Staiger et al., 2000), will improve awareness of this remarkable cytoskeletal component. In particular, given the responsiveness of MTs to plant hormones (Shibaoka, 1994; Balus˘ ka et al., 1999) and the role of both MTs (Chaffey, 2000; Funada, 2000) and plant hormones (Aloni et al., 2000) in wood formation, further studies of the effects of plant hormones on MFs is desirable. C.

Seasonal Cycle of Cambial Activity and Dormancy: A Cytoskeletal Angle

It is well known that cambial cells of the stems of temperate tree species undergo a seasonal cycle of activity and dormancy, which is paralleled by changes in cell wall thickness (Catesson, 1994). A similar seasonal cycle was found within root cambia of horse chestnut (Chaffey et al., 1998a). However, associated with the changes in cambial wall thickness was a previously unexpected cycle of MT reorientation. During cambial dormancy, when the cell walls are thickened, the MTs are arranged in a parallel-aligned helical orientation. When growth resumes and the walls become thin again, the random array of MTs reappears. These observations raise a number of questions. First, how do the MTs, which appear to persist throughout the life of the cambial cells, survive dormancy? Although it is widely reported in the literature that MTs are depolymerized by low temperatures, their presence in cambia of roots from frozen soil (Chaffey et al., 1998a) contradicts the view that MTs are absent from dormant cambia (Savidge, 1993). Are the MTs in both active and dormant cambial cells naturally resistant to cold depolymerization? Or are the MTs of dormant cambial cells comprized of a cold-resistant type

Secondary Growth

of tubulin, possibly an isoform of - or -tubulin, heterodimers of which constitute the MT, or a posttranslationally modified form of tubulin? Although our preliminary immunofluorescent examination shows presence of acetylated and tyrosinated tubulins in stem SVS material of horse chestnut and hybrid aspen, no similar studies have been undertaken with roots. What is the significance of the helical orientation of the MTs during dormancy? Certainly, there appears to be an association of parallel, helical MTs with wall thickening throughout the SVS. It is seen during differentiation of fibers, vessel elements, and phloem cells. Therefore, by comparison with secondary wall formation in those other cells of the SVS, and in keeping with the definition of a secondary wall as that which is laid down over an existing primary wall in a nongrowing cell, this ‘‘dormancy wall’’ of the cambium is considered a secondary wall (Chaffey et al., 1998a). Although it seems likely that the change of orientation of MTs from random to helical is related to the wall thickening during dormancy, whether this orientation of cytoplasmic MTs is mirrored by the orientation of cellulose microfibrils within the cell wall remains to be determined. It is tempting to suggest (Chaffey et al., 1998a) that the dormancy-associated wall differs somehow from the primary wall which surrounds active cambial cells, and that it is this difference that allows this dormancy wall to be lysed during spring cambial reactivation (Funada and Catesson, 1991), but prevents the ‘‘active cambial primary wall’’ from suffering a similar fate. However, once wall thickening has taken place, the MTs persist. Why? Do they serve other functions unrelated to wall-elaboration? For example, do they act as a protein store that can be subsequently degraded and reutilized? This suggestion is superficially attractive for horse chestnut, which, unlike hybrid aspen, appears not to store protein in the SVS. However, we do not know if the MTs are used in that way. What seems less questionable is that the partial wall lysis in springtime will have several effects: it will reduce a major physical constraint to cell expansion, and it will also produce oligosaccharides. These sugars could serve as substrates for energy production and synthesis of cell components, and/or act as solutes to raise cell turgor facilitating cell growth (Boldingh et al., 2000). It is also possible that oligosaccharins may be produced which can act as signaling molecules, which can have a number of roles in plant growth and development (Dumville and Fry, 2000). Indeed, the simple sugars produced generally as part of wall digestion

105

could have profound roles in aspects of growth and differentiation far beyond their use as substrates (Sheen et al., 1999; Smeekens, 2000), and, potentially, far beyond the immediate vicinity of the cambium. In this regard it is worth recalling the role attributed to a ‘‘secreted factor,’’ possibly an oligosaccharide, in coordinating cell expansion and differentiation in the Zinnia mesophyll system (Roberts et al., 1997). Superficially, the enzymatic digestion that appears to accompany cambial reactivation has similarities with some aspects of the breaking of seed dormancy and fruit ripening. A more detailed comparison of all three phenomena is likely to provide further insights into cambial dormancy and spring activation. Another important question arises in the context of cambial dormancy. Should dormant cambial cells be considered cambial cells? By definition, the cambium is a meristematic tissue. Clearly, dormant cambium is not meristematic, nor does it look like active cambium. This consideration is more than a semantic nicety, since it raises the further question of whether a dormant cambial cells undergoes a more profound ‘‘transdifferentiation’’ to become meristematic again, rather than simply wall thinning and vacuolation. D.

Keeping It All Going

To maintain the meristematic activity of the cambium, and the growth and differentiation of its derivatives, input of materials is needed, both as respirable substrates to support the high energy demand of cell division and growth, and as structural raw materials. Although hydrolysis of reserve materials and cambial wall-lysis in situ may contribute to the cambium’s demand for substrates, it is assumed that the majority of the required material is ultimately derived from photosynthesis within the leaves. For stem cambia it is also likely that local bark—or even pith—photosynthesis can supplement this leaf supply of photosynthate. Of course, that is a luxury not available to roots. However, although it is well established that the axial pathway of photosynthate transport is within the sieve tubes of the phloem, and that the rays are implicated in phloem–xylem exchange (Van Bel, 1990), the pathway from sieve tube to cambium is less clear (Sauter, 2000; Van Bel and Ehlers, 2000). Recent examination of the cytoskeleton within the long-lived cells of the SVS, particularly the ray cells, has helped to shed some light on the problem of photosynthate transfer to the cambial sink (Chaffey and Barlow, 2001). At the final stages of xylem ray cell differentiation, when wall thickening and elaboration

106

are complete, MFs and MTs are both bundled and netaxially oriented. At maturity, the MFs and MTs of axial xylem parenchyma cells are similarly oriented, parallel to the long cell axis. However, since the long axes of axial parenchyma and ray cells are at right angles to each other, so are the axes of orientation of the MT and MF cytoskeletons. Similarly, in mature phloem ray and axial parenchyma cells, although the MTs are more helically oriented (Fig. 8a,b, see color insert), a net-axial orientation of both MTs and MFs (Fig. 8c) is attained. Since cell differentiation has finished, and given that such cells live for many months (up to several years in the xylem), what is the role of MTs and MFs here? In view of the established importance of the long-lived ray cells as the symplasmic transport pathway between xylem and phloem (Van Bel, 1990) and their role in storage of reserve products (e.g., Sauter and Van Cleve, 1990; Ho¨ll, 2000), we have proposed involvement of both cytoskeletal components in intracellular transport (Chaffey and Barlow, 2001). Additionally, unconventional myosin VIII and callose (a marker for plasmodesmata) have been immunolocalized to the pit fields between adjacent cells within the rays of the phloem and xylem, and between ray cells and adjacent axial parenchyma. This apparent plasmodesmatal coupling of living parenchyma cells thus provides an extensive 3-dimensional symplasmic network, which extends radially from the central pith to outer bark, circumferentially within the axial parenchyma of the xylem and phloem, and axially via the sieve tubes of the phloem, which ramifies throughout the tree. The apparent colocalization of actin and myosin to the plasmodesmata of the membranes of the pits has led us to speculate further that intercellular transport, between ray and axial parenchyma cells, may be mediated by plasmodesmata whose diameter can be increased/reduced via relaxation/contraction of an actomyosin complex. It is thus envisaged that this network is responsible for not only transport of photosynthate and other materials around the tree, but also, by appropriate degrees of opening/closing of plasmodesmata, the establishment of symplasmically isolated domains (Lucas et al., 1993), which may have importance for coordination of developmental events throughout the tree. This view of the radial-axial symplasmic pathway also provides an additional role for the dilatation growth that can occur within phloem rays. Seen in a transverse section of the root (or shoot), dilatation growth appears like a triangle, with its base toward

Chaffey

the outside and its apex pointing toward the center of the organ. In that way it can be considered to act like a funnel, receiving photosynthate and other materials at its base and sides, from the sieve elements and associated cells, and channeling it toward the bottom of the funnel to enter the uniseriate ray that then extends from the phloem into the cambial zone. The cytoskeleton-facilitated solute transport is envisaged to break down as the helical/axial arrays of MTs and MFs are lost within the cells of the cambial zone, i.e., in the cells where the transported material is required to fuel respiration and cell growth. One of the attractions of this idea is that transport can be bidirectional, from either the phloem side or the xylem side. Potentially, cambial reactivation in spring can be ‘‘kick-started’’ by release of reserves from the axial and ray parenchyma cells of the xylem; thereafter, leaf photosynthate can be channeled to the cambium via the phloem route.

VII.

QUO VADIMUS?

Where do we go from here? Since ‘‘here’’ is a state of comparative ignorance about the cell—and molecular—biology of tree root cambia, further study in almost any direction will generate useful information. In nature, roots grow in the soil. That makes them messy and more difficult to access than aerial portions of the plant. Trees are rather large, and working with their roots is even more problematical than with smaller plants. It often takes a long time to dig up a tree, or even a sapling, clean the roots, and subsequently fix them, etc. It is quite likely that cell changes will have taken place during that time, compromising the interpretation of results. It is thus no wonder that aerial portions of plants are a more attractive alternative. Can we make root work a little easier? Probably, yes (see also Chapter 19 by Waisel in this volume). However, as with so many other aspects of modern plant biology, it is likely that we will have to turn to Arabidopsis to help us. Amongst the many virtues of this crucifer are that its roots can be grown in agar or another, similar nonsoil medium, permitting more ready harvest, etc. Nevertheless, how can a small, ephemeral weed help us to understand secondary growth? Although most work with Arabidopsis is currently directed at its primary plant body, under appropriate conditions it can undergo quite substantial secondary thickening and wood formation, particularly in its hypocotyl and root (e.g., Dolan and Roberts, 1995; Regan et al., 1999; Zhao et al., 2000).

Secondary Growth

Ultrastructurally, the cambial cells of Arabidopsis appear like those of angiosperm tree roots and shoots, and the fibers and vessel elements of its wood are morphologically similar to those of hybrid aspen and horse chestnut. Undeniably, there are limits to the relevance of wood formation in Arabidopsis to the process in trees. However, there is a lot of fundamental work that can be done, and Arabidopsis will have an important role to play in future development of hypotheses, which can subsequently be tested in the tree (Chaffey, 1999a). Ultimately, however, one would like techniques of investigation that will enable us to study the living processes in situ, in vivo and in planta. Unfortunately, although recent advances have been made in applying the techniques of modern cell and molecular biology to the cambium (Savidge et al., 2000; Chaffey, 2001a), until such work can be carried out with living cambia it will be necessary to use alternative models. In that regard, using intact roots of Arabidopsis to understand secondary growth of trees has obvious attractions over the other main model system for studying tracheary element formation, isolated Zinnia mesophyll cells (McCann et al., 2000). In this regard, we should recall that xylogenesis has long been viewed as a model system for the understanding of plant cell differentiation, and that one of the earliest model systems proposed was roots (Torrey et al., 1971). What tree species should be used as the ‘‘model species’’? Most of the fundamental groundwork on the ultrastructure and cytoskeletal dynamism of the root SVS has been carried out in Aesculus hippocastanum (Chaffey, 2000). However, although it is a majestic tree when in flower, it is more sensible to concentrate upon poplar/aspen and other members of the genus Populus, which by general consensus are the model angiosperm trees (Chaffey, 1999a, 2001c). As reviewed here, hybrid aspen and horse chestnut roots appear similar, certainly in respect of their cytoskeletal dynamics during wood formation. It is to be hoped that concentration on a single tree in this way will generate the critical mass of interest that has been so successful with Arabidopsis.

VIII.

THE NEXT 10 YEARS

These are exciting times for biology. The near-daily advent of technical advances means that previously intractable systems, such as the SVS, are beginning to give up their secrets, and we can now dare to ask

107

questions we couldn’t even think of 20 or 10 years ago. What advances might we expect in cambial cell biology in the next 10 years? Just as any developmental phenomenon can be studied at many different levels, consistent with the questions being asked and the availability of the necessary techniques, for the SVS we can identify anatomical, biochemical, and molecular-genetic levels (Chaffey, 2001b). We have barely begun to make inroads into the basics of the anatomical-biochemical (i.e., ‘‘cellular’’) changes that accompany secondary vascular differentiation, and there is much still to do in this area. Further documentation of that will help to establish hypotheses that can be tested, no doubt in tandem with Arabidopsis (e.g., Chaffey et al., 1998b; Chaffey, 1999b). However, the most dramatic developments are expected at the molecular-genetic level of inquiry, particularly with regard to differentiation of individual cell types and to the overall control of the SVS. Several groups are interested in identifying cambialspecific genes or genes that are involved in differentiation of secondary vascular cells. For example, Sterky et al. (1998) have recently generated a library of nearly 5700 expressed sequence tags (ESTs) from the woodforming tissues of hybrid aspen. Using a subtractive technique, Bossinger and Leitch (2000) have produced a library that is enriched in gene fragments that are differentially expressed in woody tissues, from in vitro–grown stem explants of Eucalyptus globulus. Over the next few years, we shall see many more such DNA libraries created, with a corresponding increase in the exploitation of the opportunities that they will provide for investigating the SVS. Understanding the ‘‘master genes’’ that control pathways of differentiation, such as xylogenesis in angiosperm trees, has already begun (Hertzberg and Olsson, 1998). All developmental processes depend for their success on the regulated and coordinated expression of many genes. In turn, those genes are ‘‘controlled’’ at the gene transcription level, which is mediated via transcription factors. Newman and Campbell (2000) recently presented evidence from loblolly pine that MYB proteins could act as transcription factors regulating some aspects of xylogenesis. Rather satisfyingly, they have identified orthologs of their pine genes in Arabidopsis, paving the way for use of both systems in tandem to understand this aspect of cambial activity. Homeobox genes (genes coding for transcription factors, which are involved in ‘‘switching on’’ other genes that control particular pathways of development), particularly in regard to vascular differentia-

108

Chaffey

tion, are being investigated in Arabidopsis (Baima et al., 2000). One such gene, ATHB-2, is of considerable interest because elevated levels interfere with auxin response pathways and appear to affect secondary thickening by acting as a negative regulator of gene expression (Steindler et al., 1999). This work highlights another area where our understanding is likely to increase dramatically in the near future, the molecular basis for the involvement of plant hormones in influencing the activity of the SVS. IX.

CONCLUDING COMMENTS

Although it is probably more satisfying to have a complete story, the action-packed thriller, which ends with a ‘‘to be continued . . .’’ is arguably the more exciting. I believe that The Story of Cambium falls into that latter category. We currently know just enough to make the subject interesting, but the questions that it raises, and their relevance to the really big issues of modern biology—cell differentiation and determination—should make us all keen to tackle the next instalment. ACKNOWLEDGMENTS I thank John Barnett and Peter Barlow for their patience in trying to teach me ‘‘trees and cambium’’ and ‘‘roots and cytoskeleton,’’ respectively, and both for lively discussions of those subjects over the past few years. I also thank Peter and the editors for their helpful comments on an earlier version of this chapter. The IACR receives grant-aided support from the Biotechnology and Biological Sciences Research Council of the U.K. REFERENCES Aloni R, Feigenbaum P, Kalev N, Rozovsky S. 2000. Hormonal control of vascular differentiation in plants: the physiological basis of cambium ontogeny and xylem evolution. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 223–236. Awano T, Takabe K, Fujita M, Daniel G. 2000. Deposition of glucuronoxylans on the secondary cell wall of Japanese beech as observed by immuno-scanning electron microscopy. Protoplasma 212:72–79. Bailey IW. 1952. Biological processes in the formation of wood. Science 115:255–259. Baima S, Tomassi M, Matteucci A, Altamura MM, Ruberti I, Morelli G, 2000. Role of the ATHB-8 gene in xylem

formation. In: Savidge R. Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 445–455. Balus˘ ka F, Volkmann D, Barlow PW. 1999. Hormone-cytoskeleton interactions in plant cells. In: Hooykaas PJJ, Hall MA, Libenga KR, eds. Biochemistry and Molecular Biology of Plant Hormones. Amsterdam; Elsevier, pp 363–390. Barlow PW, Balus˘ ka F. 2000. Cytoskeletal perspectives on root growth and morphogenesis. Annu Rev Plant Physiol Plant Mol Biol 51:289–322. Barnett JR. 1981. Secondary xylem cell development. In: Barnett JR, ed. Xylem Cell Development. Tunbridge Wells, UK: Castle House Publications, pp 47–95. Barnett JR., Chaffey NJ, Barlow PW. 1998. Cortical microtubules and microfibril angle. In: Butterfield BG, ed. Microfibril Angle in Wood. Christchurch, New Zealand: IAWA/IUFRO, pp 253–271. Benayoun J. 1983. A cytochemical study of cell wall hydrolysis in the secondary xylem of poplar (Populus italica Moench). Ann Bot 52:189–200. Benayoun J, Catesson AM, Czaninski Y. 1981. A cytochemical study of differentiation and breakdown of vessel end walls. Ann Bot 47:687–698. Boldingh H, Smith GS, Klages K. 2000. Seasonal concentration so non-structural carbohydrates of five Actinidia species in fruit, leaf and fine root tissue. Ann Bot 85:469–476. Boevink P, Oparka K, Cruz SS, Martin B, Betteridge A, Hawes C. 1998. Stacks on tracks: the plant Golgi apparatus traffics on an actin/ER network. Plant J 15:441–447. Bossinger G, Leitch MA. 2000. Isolation of cambium-specific genes from Eucalyptus globulus Labill. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 203–207. Braun HJ. 1984. The significance of the accessory tissues of the hydrosystem for osmotic water shifting as the second principle of water ascent, with some thoughts concerning the evolution of trees. IAWA Bull ns, 5:274– 294. Butterfield BG. 1995. Vessel element differentiation. In: Iqbal M, ed. The Cambial Derivatives. Berlin; Gebru¨der Borntraeger, pp 93–106. Catesson A-M. 1974. Cambial cells. In: Robards AW, ed. Dynamic Aspects of Plant Ultrastructure. London; McGraw-Hill, pp 358–390. Catesson A-M. 1994. Cambial ultrastructure and biochemistry: changes in relation to vascular tissue differentiation and the seasonal cycle. Int J Plant Sci 155:251– 261. Chaffey NJ. 1999a. Cambium: old challenges—new opportunities. Trees 13:138–151. Chaffey NJ. 1999b. Wood formation in forest trees: from Arabidopsis to Zinnia. Trends Plant Sci 4:203–204.

Secondary Growth Chaffey NJ. 2000. Cytoskeleton, cell walls and cambium: new insights into secondary xylem differentiation. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 31–42. Chaffey NJ, ed. 2001a. Wood Formation in Trees: Cell and Molecular Biology Techniques. Amsterdam; Harwood Academic Publishers (in press). Chaffey NJ. 2001b. Introduction. In: Chaffey NJ, ed. Wood Formation in Trees: Cell and Molecular Biology Techniques. Amsterdam; Harwood Academic Publishers (in press). Chaffey NJ. 2001c. An introduction to the problems of working with trees. In: Chaffey NJ, ed. Wood Formation in Trees: Cell and Molecular Biology Techniques. Amsterdam; Harwood Academic Publishers (in press). Chaffey NJ. 2001d. Wood microscopical techniques. In: Chaffey NJ, ed. Wood Formation in Trees: Cell and Molecular Biology Techniques. Amsterdam; Harwood Academic Publishers (in press). Chaffey NJ. 2001e. Conventional (chemical-fixation) transmission electron microscopy and cytochemistry of angiosperm trees. In: Chaffey NJ, ed. Wood Formation in Trees: Cell and Molecular Biology Techniques. Amsterdam; Harwood Academic Publishers (in press). Chaffey NJ, Barlow PW. 2000. Actin in the secondary vascular system of woody plants. In: Staiger C, Balus˘ ka F, Volkmann D, Barlow PW, eds. Actin: A Dynamic Framework for Multiple Plant Cell Functions. Dordrecht, Netherlands: Kluwer, pp 587–600. Chaffey N, Barlow PW. 2001. The cytoskeleton facilitates a three-dimensional symplasmic continuum in the longlived ray and axial parenchyma cells of angiosperm trees. Planta (in press). Chaffey NJ, Barlow PW, Barnett JR. 1996. Microtubular cytoskeleton of vascular cambium and its derivatives in roots of Aesculus hippocastanum L. (Hippocastanaceae). In: Donaldson LA, Butterfield BG, Singh PA, Whitehouse LJ, eds. Recent Advances in Wood Anatomy. Rotorua, NZ: New Zealand Forest Research Institute, pp 171–183. Chaffey NJ, Barlow PW, Barnett JR. 1997a. Formation of bordered pits in secondary xylem vessel elements of Aesculus hippocastanum L.: an electron and immunofluorescent microscope study. Protoplasma 197:64–75. Chaffey NJ, Barlow PW, Barnett JR. 1997b. Microtubules rearrange during differentiation of vascular cambial derivatives, microfilaments do not. Trees 11:333-341. Chaffey NJ, Barnett JR, Barlow PW. 1997c. Visualization of the cytoskeleton within the secondary vascular system of hardwood species. J Microsc 187:77–84. Chaffey NJ, Barnett JR, Barlow PW. 1997d. Endomembranes, cell walls and cytoskeleton: aspects of the biology of the vascular cambium of Aesculus hippocastanum L. Int J Plant Sci 158:97–109.

109 Chaffey NJ, Barlow PW, Barnett JR. 1998a. A seasonal cycle of cell wall structure is accompanied by a cyclical rearrangement of cortical microtubules in fusiform cambial cells within taproots of Aesculus hippocastanum L. (Hippocastanaceae). New Phytol 139:623–635. Chaffey NJ, Regan S, Sundberg B. 1998b. Arabidopsis: the new model tree? 8th International Cell Wall Meeting, John Innes Centre, Norwich, UK, 1–5 September, 1998. Chaffey NJ, Barnett JR, Barlow PW. 1999. A cytoskeletal basis for wood formation in angiosperm trees: the involvement of cortical microtubules. Planta 208:19– 30. Chaffey NJ, Barlow PW, Barnett JR. 2000a. A cytoskeletal basis for wood formation in angiosperm trees: the involvement of microfilaments. Planta 210:890–896. Chaffey NJ, Barlow PW, Barnett JR. 2000b. Structure-function relationships during secondary phloem development in Aesculus hippocastanum: microtubules and cell walls. Tree Physiol 20:777–786. Coetzee J, Fineran BA. 1989. Translocation of lysine from the host Melicope simplex to the parasitic dwarf mistletoe Korthalsella lindsayi (Viscaceae). New Phytol 112:377–381. Cosgrove DJ. 1998. Cell wall loosening by enzymes. Plant Physiol 118:333–339. Cronshaw J. 1965. The organization of cytoplasmic components during the phase of cell wall thickening in differentiating cambial derivatives of Acer rubrum. Can J Bot 43:1401–1407. Cutter EG. 1975. Plant Anatomy: Experiment and Interpretation, Part 2, Organs. London; Edward Arnold. Cutter EG. 1978. Plant Anatomy, Part 1, Cells and Tissues. 2nd ed. London; Edward Arnold. Dolan L, Roberts K. 1995. Secondary thickening in roots of Arabidopsis thaliana. New Phytol 131:121-128. Dumville JC, Fry SC. 2000. Uronic acid-containing oligosaccharins: their biosynthesis, degradation and signalling roles in non-diseased plant tissues. Plant Physiol Biochem 38:125–140. Farrar JJ, Evert RF. 1997. Seasonal changes in the ultrastructure of the vascular cambium of Robinia pseudoacacia. Trees 11:191–202. Fayle DCF. 1968. Radial growth of tree roots. Fac For Univ Toronto Tech Rep 9:1–183. Follet-Gueye ML, Ermel FF, Vian B, Catesson AM, Goldberg R. 2000. Pectin remodelling during cambial derivative differentiation. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 289–294. Fujino T, Sone Y, Mitsuishi Y, Itoh T. 2000. Characterization of cross-links between cellulose microfibrils, and their occurrence during elongation growth in pea epicotyl. Plant Cell Physiol 41:486–494.

110 Fujita M, Saiki H, Harada H. 1974. Electron microscopy of microtubules and cellulose microfibrils in secondary wall formation of poplar tension wood fibers. Mokuzai Gakkaishi 20:147–156. Funada R. 2000. Control of wood structure. In: Nick P, ed. Plant Microtubules: Potentials for Biotechnology. Berlin; Springer-Verlag, pp 51–81. Funada R, Catesson A-M. 1991. Partial cell wall lysis and the resumption of meristematic activity in Fraxinus excelsior cambium. IAWA Bull, ns 12:439–444. Geitman A, Emons AMC. 2000. The cytoskeleton in plant and fungal cell tip-growth. J Microsc 198:218–245. Giddings TH Jr, Staehelin LA. 1991. Microtubule-mediated control of microfibril deposition: a re-examination of the hypothesis. In: Lloyd CW, ed. The Cytoskeletal Basis of Plant Growth and Form. London; Academic Press, pp 85-99. Goosen–de Roo L, Bakhuizen R, Van Spronsen PC, Libbenga KR. 1984. The presence of extended phragmosomes containing cytoskeletal elements in fusiform cambial cells of Fraxinus excelsior L. Protoplasma 12:145–152. Green PB. 1969. Cell morphogenesis. Annu Rev Plant Physiol 20:365–394. Harris N, Chaffey NJ. 1985. Plasmatubules in transfer cells of pea (Pisum sativum L.). Planta 165:191–196. Harris N, Chaffey NJ. 1986. Plasmatubules—real modifications of the plasmalemma. Nord J Bot 6:599–607. Harris N, Oparka KJ, Walker-Smith DW. 1982. Plasmatubules: an alternative to transfer cells? Planta 156:461–465. Hertzberg M, Olsson O. 1998. Molecular characterisation of a novel plant homeobox gene expressed in the maturing xylem zone of Populus tremula  tremuloides. Plant J 16:285–295. Ho¨ll W. 2000. Distribution, fluctuation and metabolism of food reserves in the wood of trees. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 347–362. Im K-H, Cosgrove DJ, Jones AM. 2000. Subcellular localization of expansin mRNA in xylem cells. Plant Physiol 123:463–470. Iqbal M, ed. 1995. The Cambial Derivatives. Berlin; Gebru¨der Borntraeger. Lachaud S, Catesson A-M, Bonnemain J-L. 1999. Structure and function of the vascular cambium. C R Acad Sci, Paris, Ser III, Sci de la Vie 322:633–650. Larson PR. 1994. The Vascular Cambium: Development and Structure. Berlin; Springer-Verlag. Lev-Yadun S, Aloni R. 1995. Differentiation of the ray system in woody plants. Bot Rev 61:45–84. Lucas WJ, Ding B, Van der Schoot C. 1993. Plasmodesmata and the supracellular nature of plants. New Phytol 125:435–476.

Chaffey McCann MC, Domingo C, Stacey NJ, Milioni D, Roberts K. 2000. Tracheary element formation in an in vitro system. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 457–470. Meylan BA, Butterfield BG. 1981. Perforation plate differentiation in the vessels of hardwoods. In: Barnett JR, ed. Xylem Cell Development. Tunbridge Wells, UK: Castle House Publications, pp 96–114. Mia AJ. 1968. Organization of tension wood-fibers with special reference to the gelatinous layer in Populus Tremuloides Michx. Wood Sci 1:105–115. Micheli F, Bordenave M, Richard L. 2000. Pectin methylesterases: possible markers for cambial derivative differentiation? In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 295–304. Newman LJ, Campbell MM. 2000. MYB proteins and xylem differentiation. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 437–444. Patel RN. 1965. A comparison of the anatomy of the secondary xylem in roots and stems. Holzforschung 19:72–79. Phillips IDJ. 1976. The cambium. In: Yeoman MM, ed. Cell Division in Higher Plants. London; Academic Press, pp 347–390. Prodhan AKMA, Funada R, Ohtani J, Abe H, Fukazawa K. 1995. Orientation of microfibrils and microtubules in developing tension-wood-fibers of Japanese ash (Fraxinus mandshurica var. japonica). Planta 196:577– 585. Radford JE, Vesk M, Overall RL. 1998. Callose deposition at plasmodesmata. Protoplasma 201:30–37. Radford JE, White RG. 1998. Localization of a myosin-like protein to plasmodesmata. Plant J 14:743–750. Regan S, Chaffey NJ, Sundberg B. 1999. Exploring cambial growth with Arabidopsis and Populus. J Exp Bot 50(suppl):33. Reichelt S, Kendrick-Jones J. 2000. Myosins. In: Staiger C, Balus˘ ka F, Volkmann D, Barlow PW, eds. Actin: A Dynamic Framework for Multiple Plant Cell Functions. Dordrecht, Netherlands: Kluwer, pp 29–44. Reichelt S, Knight AE, Hodge TP, Balus˘ ka F, S˘amaj J, Volkmann D, Kendrick-Jones J. 1999. Characterization of the unconventional myosin VIII in plant cells and its localization at the post-cytokinetic cell wall. Plant J 19:555–567. Robards AW, Kidwai P. 1972. Microtubules and microfibrils in xylem fibers during secondary cell wall formation. Cytobiologie 6:1–21. Roberts A, Uhnak KS. 1998. Tip growth in xylogenic suspension cultures of Zinnia elegans: implicatinos for the relationship between cell shape and secondary cell wall

Secondary Growth pattern in tracheary elements. Protoplasma 204:103– 113. Roberts AW, Donovan SG, Haigler CH. 1997. A secreted factor induces cell expansion and formation of metaxylem-like tracheary elements in xylogenetic suspension cultures of Zinnia. Plant Physiol 115:683–692. Sal’nikov VV, Ageeva MV, Yumashev VN, Lozovaya VV. 1993. Ultrastructural analysis of bast fibers. Russian Plant Physiol 40:416–421. Sauter JJ. 2000. Photosynthate allocation to the vascular cambium: facts and problems. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 71–83. Sauter JJ, Van Cleve B. 1990. Biochemical, immunochemical, and ultrastructural studies of protein storage in poplar (Populus  canadensis ‘‘robusta’’) wood. Planta 183:92–100. Savidge AR. 1993. Formation of annual rings in trees. In: Rensing L, ed. Oscillations and Morphogenesis. New York; Marcel Dekker, pp 434–463. Savidge, R. Barnett JR, Napier R, eds. 2000. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers. Seagull RW. 1992. A quantitative electron microscopic study of changes in microtubule arrays and wall microfibril orientation during in vitro cotton fiber development. J Cell Sci 101:561–577. Sennerby-Forsse L, Von Fircks HA. 1987. Ultrastructure of cells in the cambial region during winter hardening and spring dehardening in Salix dasyclados Wim. grown at two nutrient levels. Trees 1:151–163. Sheen J, Zhou L, Jang J-C. 1999. Sugars as signalling molecules. Curr Opin Plant Biol 2:410–418. Shibaoka H. 1994. Plant hormone-induced changes in the orientation of cortical microtubules: alterations in the cross-linking between microtubules and the plasma membrane. Annu Rev Plant Physiol Plant Mol Biol 45:527–544. Smeekens S. 2000. Sugar-induced signal transduction in plants. Annu Rev Plant Physiol Plant Mol Biol 51:49–81. Somerville C. 2000. The twentieth century trajectory of plant biology. Cell 100:13–25. Staiger C, Balus˘ ka F, Volkmann D, Barlow PW, eds. 2000. Actin: A Dynamic Framework for Multiple Plant Cell Functions. Dordrecht, Netherlands: Kluwer. Steindler C, Matteucci A, Sessa G, Weimar T, Ohgishi M, Aoyama T, Morelli G, Rubertis I. 1999. Shade avoidance responses are mediated by the ATHB-2 HD-Zip protein, a negative regulator of gene expression. Development 126:4235–4245. Sterky F, Regan S, Karlsson J, Rohde A, Holmberg A, Amini B, Bhalerao R, Larsson M, Villarroel R, Van Montagu M, Sandberg G, Olsson O, Teeri TT,

111 Boerjan W, Gustafsson P, Uhle´n M, Sundberg B, Lundeberg J. 1998. Gene discovery in the wood-forming tissues of poplar: analysis of 5,692 expressed sequence tags. Proc Natl Acad Sci USA 95:13330– 13335. Sundberg B, Uggla C, Tuominen H. 2000. Cambial growth and auxin gradients. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers, pp 169–188. Takabe K, Harada H. 1986. Polysaccharide deposition during tracheid wall formation in Cryptomeria. Mokuzai Gakkaishi 32:763–769. Torrey JG, Fosket DE, Hepler PK. 1971. Xylem formation: a paradigm of cytodifferentiation in higher plants. Am Sci 59:338–352. Van Bel AJE. 1990. Xylem-phloem exchange via the rays: the undervalued route of transport. J Exp Bot 41:631–644. Van Bel AJE, Ehlers K. 2000. Symplasmic organization of the transport phloem and the implications for photosynthate transfer to the cambium. In: Savidge R, Barnett JR, Napier R, eds. Cell and Molecular Biology of Wood Formation. Oxford, UK: Bios Scientific Publishers Ltd, pp 85–99. Vissenberg K, Martinez-Vilchez IM, Verbelen J-P, Miller JG, Fry SC. 2000a. In vivo colocalization of xyloglucan endotransglycosylase activity and its donor substrate in the elongation zone of roots. Plant Cell 12:1229– 1238. Vissenberg K, Quwlo A-H, Van Gestel K, Olyslaegers G, Verbelen J-P. 2000b. From hormone signal, via the cytoskeleton, to cell growth in single cells of tobacco. Cell Biol Int 24:343–349. Waterkeyn L. 1967. Sur l’e´xistence d’un ‘‘stade callosique,’’ pre´sente´ par la paroi cellulaire, au cours de la cytocine`se. C R Acad Sci Paris 265:1792–1794. Wenham MW, Cusick F. 1975. The growth of secondary wood-fibers. New Phytol 74:247–261. Wick SM. 1991. Spatial aspects of cytokinesis in plant cells. Curr Opin Cell Biol 3:253–260. Willats WGT, Steele-King CG, McCartney L, Orfila C, Marcus SE, Knox JP. 2000. Making and using antibody probes to study plant cell walls. Plant Physiol Biochem 38:27–36. Wilson BF. 1975. Distribution of secondary thickening in tree root systems. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. New York; Academic Press, pp 197–219. Woods FW. 1991. Cambial activity of roots. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. New York; Marcel Dekker, pp 149–160. Zhao C, Johnson BJ, Kositsup B, Beers EP. 2000. Exploiting secondary growth in Arabidopsis. Construction of xylem and bark cDNA libraries and cloning of three xylem endopeptidases. Plant Physiol 123:1185–1196.

7 The Kinematics of Primary Growth Wendy Kuhn Silk University of California, Davis, California

I.

KINEMATIC ANALYSES ILLUMINATE ROOT FUNCTION

shape change in fluids and continuua, are particularly relevant. (A distinction is made between kinematics and ‘‘dynamics,’’ the study of the forces and energies that produce the observed motions and shape changes.) Applications of kinematic analyses include consideration of the ‘‘material’’ aspects of root development to elucidate spatial–temporal relationships in growing tissue. Growth itself can be characterized in terms of relative elemental rates, analogous to strain rates used in continuum mechanics. In structural studies, kinematic expressions can be used to find basic physical relationships among growth rates, cell division rates, anatomy, and morphology. In studies of mineral nutrition, kinematic expressions allow us to find local nutrient deposition rates in expanding tissue from spatial concentration patterns. In ecological studies, the growth strain rate field provides a quantitative characterization of effects of environmental variation on growth. And some basic ideas from growth kinematics, including the use of a moving reference frame attached to the growing root tip, allow us to understand the relationship between the growth zone and its rhizosphere.

Symbols used to represent kinematic variables and parameters: x z r t X xðX,tÞ Z zðZ,tÞ v vx and vr u L g @V=V@t vf ; k; n

Longitudinal distance from root tip Longitudinal distance from soil surface Radial distance from root center Time Particle found x mm from root tip at initial time Location, in co-moving reference frame, at time t of particle X Particle found z mm from soil surface at initial time Location, in stationary reference frame, at time t of particle Z Rate of displacement from root tip Components of growth velocity in axial and radial directions Rate of displacement from soil surface Length of root Relative elemental growth rate (REG rate or growth strain rate) Local relative rate of increase in volume Parameters to fit velocity field v(x) to logistic function

II.

The concepts and numerical methods from fluid dynamics and continuum mechanics are powerful tools for solving problems in root physiology. The methods of ‘‘kinematics,’’ i.e., the study of motion or

ROOT DEVELOPMENT HAS SPATIAL AND MATERIAL ASPECTS

As one looks back from the root cap and root apex one sees a zone of cell division (with slow growth), then a zone of rapid expansion that overlaps a zone of tissue 113

114

Silk

differentiation. Differentiation may be further characterized by the locations of specialized cells. For example, under constant environmental conditions in a particular species the formation of phloem, root hairs, and xylem may occur at identifiable distances from the root apex. In botany textbooks the pattern along the root axis is shown to represent a developmental gradient, and the root tip is described in terms of anatomically and functionally distinct zones. However, to understand the root we must also recognize that the developmental zones are formed from a procession of tissue elements each of which first grows slowly while producing new cell walls, then elongates rapidly, and then further differentiates specialized cells. When the root is viewed over time, a marked cohort of cells appears to flow from the root tip through the standing developmental pattern. The existence of a form that is unchanging (or slowly changing) in time, and that is composed of elements that are experiencing rapid change, is reminiscent of fluid structures. And indeed, allusions to rivers, fountains, and the wakes of boats appear often in classical botanical literature. Root growth zones, like boat wakes, are composed of changing ‘‘material elements’’—cells, analogous to water droplets in the boat wake. To understand root structure it is important to appreciate both spatial patterns, as observable in photographs, and the ‘‘material’’ aspects or real elements, i.e., the properties of the cells that comprise the spatial pattern as they move through it. This distinction between spatial (or site-specific) and material (or cell particle-specific) descriptions of developmental variables is central to the understanding of plant form and to clarifying the relationships among spatial and temporal patterns in developing roots (Green, 1976; Silk and Erickson, 1979; Gandar, 1980, 1983; Silk, 1989, 1992).

III.

ROOT GROWTH CAN BE VISUALIZED IN STATIONARY AND MOVING REFERENCE FRAMES

In the discussion of material specifications the emphasis was on the continuing flow of cells away from the root tip. From another point of view, it is the pattern that moves through a constantly extending sequence of cells. If we consider spatial coordinates attached to the soil surface, each cell is displaced only a small distance from where it is formed. But the various zones move from one place to another down the growing root, keeping a constant distance behind the moving tip. This view of root organization arises when we use a

stationary reference frame and is useful, for instance, for understanding the relationships between the developing root surface and the stationary soil particles that surround it. In contrast, the use of a ‘‘comoving’’ reference frame attached to the root tip allows us to find the many steady (time-invariant) or quasi-steady patterns that characterize root development. Historically, botanists have observed that primary extension growth is confined to apical regions of roots and shoots. Marks placed far from the apex do not separate from each other, although they are found progressively farther from the growing apex. One way of specifying growth is to plot the positions of cellular particles, or applied marks, versus time (Fig. 1). The resulting ‘‘pathline’’ (Gandar, 1983) or ‘‘growth trajectory’’ (Silk and Wagner, 1980; Hejnowicz, 1984) is a material specification of growth, because material particles are followed. If distance is measured from the apex, the slope of the growth trajectory increases monotonically from small values near the apex to a constant value at the base of the growth zone. To characterize the growth pattern, a family of curves can be obtained to show growth trajectories of particles at different initial positions (Fig. 1). The relationships between the growth trajectories in the stationary and comoving reference frames can be understood with the use of some formalisms from continuum mechanics (Silk, 1989). To connect the stationary and comoving reference frames, we can use two different variables for distance: Let x denote the distance from the root tip, and z denote the distance from the soil surface (or any other stationary reference origin). To distinguish properties associated with moving cellular particles from properties associated with spatial locations (instantaneously occupied by a particular particle), we can use capital letters for the material properties and lowercase letters for spatial properties. In particular, let X denote the material or real cellular particle located initially at location x0 ¼ X. Then xðX,tÞ represents at time t the spatial location occupied by particle found initially at the location X. This is the growth trajectory or particle pathline followed over time by the cellular particle, specified by its initial distance from the root tip. The expression zðX,tÞ represents the position of the same particle in the stationary reference frame. We can see that xðX,tÞ þ zðX,tÞ equals the length of the root at time t. Formally, for the material particle X it is easily seen that zðX,tÞ þ xðX,tÞ ¼ LðtÞ where LðtÞ represents root length at time t.

ð1Þ

Kinematics of Primary Growth

115

Figure 1 Growth trajectories of cellular particles in the maize root, shown in two reference frames. The root has a growth zone 12 mm long and extends initially (at time 0) 12 mm below the soil surface. The stationary reference frame is indicated with dashed lines showing zðZ,t) where z is distance from the soil surface of cellular particles initially located, respectively, at 10, 10.5, 11, and 11.5 mm from the soil surface. These particles are also initially located, respectively, at 2.0, 1.5, 1.0, and 0.5 mm from the root tip. On the graph the material particle located initially z ¼ 10 and x ¼ 2 is displaced from the soil surface as shown by the lowest dashed line in the figure and also displaced from the root tip as shown by the highest solid line on the figure. With time each particle decelerates away from the soil surface to a constant depth in the soil and also accelerates to a constant rate of displacement from the root tip.

A similar relationship can be found to connect the growth velocities in the two reference frames. Let v represent the velocity of displacement from the tip, and u represent the velocity of displacement from the soil surface. Then vðX,tÞ þ uðX,tÞ ¼ @L=@t

ð2Þ

where the capital letters are used to denote velocities associated with the material particle. Equation (2) says the rate at which a particle moves away from the soil

surface plus the rate at which it moves away from the apex equals the elongation rate of the root, i.e., the time rate of change of root length. If instantaneous point-particle interchangeability is invoked, equation (2) also implies that at time t the approximately steady field vðxÞ can be used to compute uðx,tÞ as the difference between the overall elongation rate and velocity in the moving reference frame. These uðx,tÞ can then be assigned to the appropriate z values via equation (1). The characteristics of the velocity field (Erickson and Sax, 1956) are apparent in Fig. 2. The velocity

116

Silk

Figure 2 The growth velocity field, shown in two reference frames. When plotted as a function of x, distance from the root tip, displacement velocities are time invariant (upper figure). That is, rate of displacement, from either the tip or the base of the root, is constant at a particular location. When plotted in a stationary reference frame, the velocity field is shifted to the right over time (lower figure).

of displacement from the root tip increases monotonically with position, because progressively more expanding tissue is found between the origin and progressively more basal locations (Fig. 2, solid line). At the base of the growth zone, where expansion is no longer occurring, the growth velocity acquires a uniform value equal to the elongation rate of the organ. In

our experiments, in the maize primary root growing at 28 C, individual particles accelerate to a final growth velocity (rate of displacement from the apex) of 3:2 mm h1 . Velocity of displacement from the soil surface is minimum at the base of the root (Fig. 2, dashed lines). Thus the uðxÞ and uðzÞ fields give us a physically intuitive picture of the growth velocity, since we know

Kinematics of Primary Growth

the root tip moves most rapidly away from the soil surface while the basal regions of the root are not being displaced. With time the curves of vðzÞ and uðzÞ are shifted to the right (Fig. 2, bottom), while uðxÞ and vðxÞ remain unchanged (Fig. 2, top). It is the time invariance of vðxÞ that makes it so useful. Graphs of growth trajectories can be constructed in the moving and stationary reference frames by integrating particle velocity over time, as shown in Fig. 1. In the moving reference frame we see that a cellular particle accelerates through the growth zone to achieve a constant velocity of displacement from the tip. As long as the root continues to grow, the particle will increase its distance from the tip with time. In the stationary reference frame, cellular particles decelerate to zero velocity of displacement. The final location of the particle depends on its initial location; particles initially close to the root tip move deeper into the soil than particles initially farther from the tip. The growth trajectories of the comoving reference frame show quantitatively how the ‘‘stationary pattern’’ on the root axis must be occupied by a changing population of cells that flows away from the tip. And the growth trajectories in the stationary reference frame show how the ‘‘moving pattern’’ on the root axis is maintained a constant distance from the root tip. IV.

A.

ROOT GROWTH IS QUANTITATIVELY DESCRIBED IN TERMS OF PARTICLE TRAJECTORIES, VELOCITIES, AND RELATIVE ELEMENTAL GROWTH RATES (STRAIN RATES) Specification of Axial Growth in One Dimension

Contemporary growth analysis emphasizes the relationships among growth trajectories, growth velocities, and relative elemental growth rates. We have seen that the growth trajectory provides a material specification of growth; a family of curves represents the displacements of the many cellular particles during the tissue expansion. The growth trajectory can be used as a space–time map. Tabulated values of position and time can be used to infer the time course for any developmental variable with a known spatial distribution along the root axis (Silk and Erickson, 1979). The velocity field is a spatial, or Eulerian, specification of growth because it involves the movements of many particles found instantaneously at different locations. The velocity field can be calculated from shortterm marking experiments. It is used in many impor-

117

tant physiological applications, including calculations of biosynthesis rates, cell production rates, and material derivatives in growing tissue (Silk and Erickson, 1979; Gandar, 1980). For the physiologist the most useful growth descriptor is probably the relative elemental growth (REG) rate (Erickson and Sax, 1956), called the growth strain rate by engineers. In one dimension the REG rate, g, is the velocity gradient; mathematically, g ¼ @v=@x ¼ @v=@z. The REG rate is also approximated by the local relative growth rate of a segment of initial length L, g
L=L
t. The local relative growth rate approaches the REG rate as
L and
t become small. The REG rate field gives a quantitative description of the location and magnitude of growth within the root. In our experiments in a maize root elongating at 3:2 mm h1 , for example, growth occurs within the region extending 12 mm from the tip (Fig. 3, top curves). The REG rate maximum, 0:45 h1 , is found 4–6 mm from the apex. The REG rate profile gives a quantitative characterization of growth and is thus ideal to show the effect of environmental variation on growth. The REG rate also appears in theoretical studies, including the basis for axial curvature (Silk and Erickson, 1978) and growth sustaining water potential (Silk and Wagner, 1980). It provides the spatial pattern for determining the ultrastructural control of growth (Sugimoto et al., 2000). B.

Radial Growth Rates

A comprehensive analysis of growth including bending and twisting may involve the use of natural coordinate systems and be quite complex (Hejnowicz, 1984; Silk, 1989; Jirasek et al., 2000). Simpler analyses can be made for the many cases in which the root is radially symmetrical, has steady morphology and cell size distribution, and grows without bending or twisting. In such cases a two-dimensional analysis in cylindrical coordinates can provide a comprehensive description of the growth. We can use the coordinate system (x; r), where x is longitudinal distance from the root tip, and r is radial distance from the root center. The components of growth velocity can be given as vx and vr . As described in the section on axial growth, the pattern of vx can be determined with an axial marking experiment. If root morphology is steady, the maximum radial growth velocity (at the root surface) can be expressed in terms of the spatial pattern of root radius and longitudinal growth velocity. That is, at the root surface vr ¼ vx @r=@x. However, to find the spatial pattern of radial expansion, we need to know the spatial

118

Silk

increase in cell widths at two longitudinal positions and dividing by the time required for displacement between the two positions. The longitudinal component of the growth strain rate @vx =@x can be obtained as usual with an axial marking experiment, and a circumferential relative growth rate is given by vr =r. Note that even in the absence of twist the local, relative growth rate in volume, V, includes the circumferential as well as the radial strain rates: @V=V@t ¼ @vx =@x þ @vr =@r þ vr =r

Figure 3 REG rate fields fit to an asymmetric logistic function for maize roots grown at different temperatures (upper figure) and soil water potentials (lower figure). (From Pahlavanian and Silk, 1988; Sharp et al., 1988; and Morris and Silk, 1992.)

distribution of vr along the root radius. This can be difficult to obtain noninvasively. For the case of steady cell size distributions we can infer the radial growth velocity distribution by comparing cell widths in the same file at different longitudinal positions. For this simple case we see that a longitudinal marking experiment can be combined with a determination of the profile of cell widths at different distances from the tip to obtain the growth strain rates in three dimensions. Now vr ¼ vx @r=@x, where r is the radial location of the cell file at position x. A radial growth rate @vr =@r can be approximated by measuring the relative

ð3Þ

Relatively few two-dimensional analyses of root growth can be found in the literature. Hejnowicz and colleagues (1984, 1993) pioneered the use of curvilinear coordinates to show relationships among growth trajectories and isovelocity curves within root apices. Barlow et al. (1991) characterized the effects of a mutation on components of the growth rate in tomato roots. Water stress has been shown to affect radial growth rates differently from longitudinal growth rates in the root meristem (Sharp et al., 1988; Van der Weele et al., 2000). The osmotic adjustment under water stress involves a decrease in radial growth rates, related to an overall decrease in osmoticum deposition rates (Sharp et al., 1990; but see Ober and Sharp, 1994). These studies directed attention to the control of growth anisotropy. Baskin and Bivens (1995) found that different toxins affect longitudinal and radial growth rates in Arabidopsis roots in different ways. And studying the ultrastructural basis for the different effects of water stress on the radial and longitudinal components of the growth rates, Baskin and Sharp and colleagues (Liang et al., 1997; Baskin et al., 1999) showed that the degree of growth anisotropy in maize roots was not correlated with the degree of alignment of either microtubules or microfibrils. The growth analyses combined with spatial patterns of ultrastructural detail have suggested the need for a new paradigm to replace the multinet theory of cell expansion.

V.

FLAWS IN EXPERIMENTAL DESIGN AND INTERPRETATION CAN OCCUR IN THE STUDY OF ROOT GROWTH ZONES

Physiological research on growing roots is tricky. Physiologists are trained in chemistry where the most commonly held assumptions are what Paul Green called ‘‘the assumptions of the well-stirred beaker.’’ In the famous Michaelis-Menton model for enzyme kinetics, for example, spatial homogeneity and tem-

Kinematics of Primary Growth

poral variation are assumed. In the root growth zone, by contrast, the simplest possible set of assumptions includes spatial variation (the presence of the developmental gradient) and temporal constancy (if the apex is chosen as origin of the coordinate system). Experiments on growing tissue must be designed to take account of cellular displacements during growth. A.

Assessing Environmental Effects

A naı¨ ve, but not uncommon, experimental design is to subject the root tip to an environmental perturbation, such as immersion in a toxic chemical. After a day or two the treated and control roots are compared. To make physiologically meaningful comparisons, the fates of treated cells should be mapped so that the properties of the tissue generated during the treatment can be compared to the tissue generated in the absence of the treatment. Thus, good experimental design requires that growth trajectories as well as spatial patterns in developmental variables should be determined in studies of growing parts of roots. Assessment of the effect of environmental variation on root properties must include identification of the tissue produced during the environmental perturbation. B.

Designing a Meaningful Growth Analysis

Marking experiments to obtain growth velocities and REG rates also need some attention to find meaningful spatial and temporal scales. A single marking experiment, with proper spatial and temporal resolution, can be used to infer the growth trajectory (by following identifiable particles on the root surface) and the REG rate profile (by differentiation of particle velocity with respect to position). To determine the REG rate profile, the marking experiment must be designed with spatial resolution sufficient to evaluate the REG rate at perhaps 10 locations within the growth zone. Adequate temporal resolution requires that marks not move too far from their initial positions during the observation period (Erickson and Silk, 1980; Silk, 1984). As longer time intervals are used, then the calculated REG rates become progressively overestimated, and the length of the growth zone becomes badly underestimated. If the calculated local relative growth rate is assigned to the final midpoint of a segment rather than the initial apical location, then the observation period can be somewhat longer, but the magnitude of the REG rate becomes progressively underestimated (Peters and Bernstein, 1997). A good rule of thumb is that

119

< 20% of the length of the growth zone should be displaced past the boundary of the growth zone during the observation period. Then, whether assigned to initial, midpoint, or final locations during the growth analysis, the local relative growth rates will approximate the instantaneous values of the REG rate. To determine whether growth is steady, growth trajectories or velocity profiles should be evaluated at different times. If growth is steady, then many developmental regularities occur: successively formed growth trajectories can be superimposed, and velocity and REG rate profiles coincide for observations made at different times. If growth is known to be steady, then a single determination of the growth velocity field can be integrated (in a special way, following a particle as it moves through the growth zone) to obtain the growth trajectory. However, a growth trajectory obtained by numerical integration of the growth velocity will be subject to some uncertainty. Experimental error in measured growth velocities near the root tip becomes amplified by the integration process through the growth zone. This is a reflection of a physical variability inherent in the growth process. In fact, small differences in growth velocity near the root tip do result in large differences—for instance, in final segment length as a root segment is displaced through the growth zone. This has led to some controversies in interpretation of effects of genetic and hormonal variation on growth patterns (Peters et al., 1999). C.

Measurement of Local Biosynthesis and Uptake Rates in Growing Tissue

A related artifact involves the design of radioactive labeling experiments to measure biosynthesis or incorporation rates. In designing and interpreting radioactive labeling experiments, one confronts the difficulty that the experiment is not instantaneous. Some time is required for penetration of label into the tissue, and in this time period the cells initially at one location will be displaced. The problem can be illustrated by considering the deposition rate profile calculated using continuity equations for uronide incorporation into cell walls (Fig. 4 from Silk et al., 1984). In the roots of this study, a cell located initially at 4 mm from the tip will be located 1 h later at 5.2 mm, where the incorporation rate is quite different. In a labeling experiment, as the incubation period increases, the amount of label found at a location is less likely to be an accurate representation of the local deposition rate; the cells at a given location, having arrived from

120

Silk

growing roots. Thus, in labeling experiments, as in growth analysis, it is important that the cells at a particular location remain near the attributed sites during the time interval for the observation of change.

VI.

Figure 4 The apparent incorporation rate of uronide into the growth zone of the maize root for different times of incubation in radioactive label. Top curve (asterisks) shows rates calculated using a continuity equation with data on uronide content and growth velocity. If this curve is assumed to be the true, instantaneous rate of uronide incorporation, then incubation in label for 15 min, 2 h, and 5 h would produce the set of progressively flatter curves. To use labeling experiments to localize uptake patterns within the growth zone, the investigator must choose incubation times short enough to minimize the growth displacements. (From Silk et al., 1984.)

progressively more apical sites, will present an integration of the incorporation rates associated with the different positions traversed en route. A computer simulation of the deposition process reveals that for these fast-growing roots, only incubation periods < 30 min will reveal an accurate profile of the local deposition rates. Before 2 h of incubation, the important characteristics of the deposition rate profile are lost: the peak at 3 mm is not readily observable, and the low values at the basal end of the growth zone are shown erroneously large. After 5 h of incubation in radioactive label, the deposition rate profile would appear almost flat. In light of published observations that 3–6 h is required for good penetration of polysaccharide precursor into the corn root, it can be concluded that it is not possible to obtain incorporation rate profiles with radioactive label supplied to rapidly

EFFECTS OF ENVIRONMENTAL AND GENETIC VARIATION ON GROWTH HAVE BEEN QUANTIFIED WITH REG RATES

It has long been recognized that the REG rate field can be used to describe quantitatively the spatial pattern of root growth within the root (Goodwin and Stepka, 1945; Erickson and Sax, 1956). Thus the REG rate field provides the basic phenomenology for the physiological understanding of growth. By now, many kinematic studies have established the effects on the growth rate distribution of environmental factors, including temperature (Pahlavanian and Silk, 1988), water stress (Sharp et al., 1988; Van der Weele et al., 2000), irradiance (Muller et al., 1998), salinity (Zhong and La¨uchli, 1993), and ambient pH (Peters and Felle, 1999; Walter et al., 2000). REG rate profiles have also been used to quantify growth patterns at different developmental stages (Beemster and Baskin, 1998). Most recently, with the explosion of work on mutant plants, the REG rate field has been used to infer the mechanisms for genetic control of growth (Sugimoto et al., 2000; Beemster and Baskin, 2000). A number of different spatial patterns and corresponding physiological mechanisms have been found to underlie environmentally induced variation in root elongation rates. An interesting contrast was observed between the effects of water availability and temperature. Pahlavanian and Silk (1988) concluded that within a certain range, temperature affected the rates of developmental processes, including REG rates, without much affecting the coordination of other processes. In maize roots growing at particular temperatures between 19 C and 29 C, there were similar patterns of cell length. Furthermore, the length of the growth zone appeared invariant although root elongation rate varied threefold over this range of incubation temperatures. In contrast, Sharp et al. (1988) found that as water stress increased, the length of the growth zone became shorter, while the magnitude of the REG rates was conserved in the apical parts of the growth zone. These effects were summarized by Morris and Silk (1992) who fitted a flexible, asymmetric logistic function to the primary data sets. The velocity fields were well fit in the spatial domain using the function

Kinematics of Primary Growth

vðxÞ ¼ 

vf 1 þ ekðxx0 Þ

121

leagues (e.g., Ober and Sharp, 1994;) revealed that the maintenance of the REG rate during water stress depended on increased levels of abscisic acid, promoting increased deposition rates of proline as an osmoticum. The abscisic acid acts by decreasing ethylene production under water stress (Spollen et al., 2000).

ð4Þ

1=n

where vðxÞ represents the velocity at position x, vf is the velocity at the base of the growth zone, and x0 is chosen as that spatial position satisfying ½vðx0 Þn ¼

vnf 2

ð5Þ

VII.

The effects of temperature and water stress were displayed by plotting the longitudinal REG rate field, @v=@x against x (Fig. 3). Then the environmental effects were summarized by tabulating values of the root elongation rate, vf ; the parameter k, a measure of the spread of the velocity curve along the x-axis; the parameter n, related to the position of the inflection point of the REG rate curve; and the maximum REG rate for the different environmental conditions (Table 1). For the chosen range of temperatures and soil water potential, environmental variation produced similar changes in overall root elongation rate for the two factors. However, water stress and temperature affected the spatial growth patterns in different ways. Confirming the conclusions of the empirical studies, the fitting parameters indicate that temperature does not much affect the spread of the velocity or longitudinal REG rate curves, while increasing water stress causes a progressive shortening of the growth zone and a basal displacement of the point of maximum REG rate. The maximum longitudinal REG rate changed in proportion to incubation temperature but was not much affected by water stress. Thus, the physiology of the responses to the two environmental factors must be different. Subsequent work by Sharp and col-

KINEMATIC ANALYSES REVEAL INTERACTIONS BETWEEN GROWTH ZONES AND THEIR RHIZOSPHERES*

Symbols used to represent variables and parameters in model for pH of rhizosphere: F a pH1 bHS DHS D r t 0; t r 0; r z0 z V L R

Radial flux of H þ (from the root surface) Root radius Initial soil pH Soil buffering capacity Soil acidity diffusion coefficient Diffusivity of H þ Gradient operator Time Radial distance from root center Longitudinal distance from soil surface Longitudinal distance from root tip Growth velocity Distance parameter Rhizosphere or Pe´clet number

The ability of the root to change the pH of the soil in its immediate vicinity affects the uptake of both beneficial nutrients and phytotoxic metals, as described in Chapter 33 by Gerendas and Ratcliffe in this volume. Thus, an understanding of root-induced pH patterns in the rhizosphere is central to the study of plant nutri-

Table 1 Effects of Incubation Temperature and Soil Water Potential on the Growth Velocity of the Primary Maize Root Temperature ( C) 29 24 19 16 29 29 29

Water potential (MPa)

Elongation rate (mm h1 ) vf

k

n

Maximum REG rate ðh1 Þ

0:03 0:03 0:03 0:03 0:20 0:81 1:60

3.14 2.12 1.62 1.18 2.04 1.58 1.13

0.55 0.53 0.58 0.55 1.06 1.25 1.52

0.83 0.80 0.75 0.68 1.66 1.66 2.19

0.46 0.30 0.25 0.18 0.45 0.41 0.32

The parameters gave the best fit of the velocity field to an asymmetric, logistic function. Source: Morris and Silk (1988). *Section VII is adapted from Kim et al., 1999, and Nichol and Silk, in press, with permission from Blackwell Science Ltd.

122

Silk

tion and to the physiology of root growth. This is recognized in many studies of genetic variation in capacity to modify the pH of the rhizosphere (e.g., Gollany and Schumacher, 1993; Yan et al., 1996; Degenhardt et al., 1998). Most of the models of root–rhizosphere interactions consider the mature part of the root, since the largest proportion of root tissue is no longer growing. However, recent work recognizes that many of the physiologically important uptake processes occur in the soil surrounding the meristem and growth zone (e.g., Henriksen et al., 1992; Jaillard et al., 1996; Felle, 1998). Many of the ideas from growth kinematics, including the importance of a moving reference frame attached to the root tip and the recognition of spatial variation within the growth zone, can be used to extend our understanding of the rhizosphere of the growth zone. Recently, we have used growth kinematics in a mathematical model to predict pH in the rhizosphere of a growth zone (Kim et al., 1999).

A.

Classical Theory and Modifications Needed to Understand Growth Zones

Classically, a diffusion theory has been used to predict the plant-induced pH fields in the rhizosphere (Nye, 1981, and references therein). Pointing out that a charge balance is axiomatic, Nye hypothesized that the root would release protons if more cations than anions are absorbed from the soil. Assuming the hydrogen ions would diffuse according to a modified Fick’s Law, Nye used an analytic solution for the diffusion equation with flux over the surface of a cylinder to find rhizosphere pH as a function of distance, r (from the root surface), and time, t:     aF 2:25DHS t pH ¼ pH1  ln ð6Þ 2bHS DHS r2 where F is the flux of Hþ (from the root surface); a is root radius; pH1 is the initial soil pH; bHS is soil buffering capacity; and DHS is the soil acidity diffusion coefficient. The existing theory is powerful in explaining how hydrogen ion fluxes associated with root metabolism affect soil properties. However, it is one-dimensional and assumes constant, spatially uniform Hþ flux. Thus, while it works well for the rhizosphere around mature root tissue, it needs modification to be used for growth zones. This is because growing tissue is characterized by spatial patterns that correspond to changes in developmental stage. Primary growth is

localized in small (cm-scale) zones at the tips of roots. The pH at the surface of growth zones has often been observed to be as much as 2 pH units higher or lower than neighboring mature root tissue (e.g., Weisenseel et al., 1979; Ha¨ussling et al., 1985; Gollany and Schumacher, 1993; Taylor and Bloom, 1998). Thus axial gradients in ion deposition rate and proton flux should be resolved on the scale of millimeters within the growth zone. In Kim et al. (1999), several conceptual modifications are made to adapt the Nye theory to growing tissue. The Hþ flux is plotted as a function of distance from the root tip. In this reference frame the pattern of Hþ flux can be taken to be steady (independent of time) if environmental conditions are held constant. (The soil next to the root above the growth zone is strongly affected by transpiration. However, since there is no functional xylem in the growth zone, we hypothesize transpiration has only indirect effects on the rhizosphere of the growth zone.) Thus, in the plantbased reference frame the Hþ flux is steady (like the classical theory) but nonuniform (in contrast to the classical theory). The root tip is propelled at a constant velocity of, say, 2 mm h1 through bulk soil. From the point of view of a soil particle next to the root, the Hþ fluxes corresponding to the different root locations will be encountered in a predictable sequence. The fixed soil particle will absorb the flux associated with a neighboring root element. Since the root tip is moving 2 mm h–1 downward, every half-hour the flux into the soil particle will be from a root element that is located 1 mm farther from the root tip. Thus, we can simplify the problem by imagining a stationary root, surrounded by a moving soil medium that is flowing 2 mm h1 upward. Solving for the pH field around the growth zone involves following a slice of soil as it ‘‘moves’’ away from the tip upward, keeping track of the history of the radial pH profile and updating the Hþ flux over time as the soil encounters the older tissue elements. B.

The Convection Diffusion Model

To derive our model, we began with the diffusion model (as described by Nye, 1981).      @ Hþ ð7Þ ¼ r: Dr Hþ 0 @t In this equation t 0 is time, D is the diffusion coefficient, and r is the gradient operator. (In this general formulation, in terms of proton flux, the diffusion coefficient,

Kinematics of Primary Growth

123

D, is determined empirically from the measured diffusion of protons in soil or other growth substrates of interest and includes the soil buffering capacity.) Assuming that the solution is radially symmetric, equation (7) can be reduced to the following axisymmetric diffusion equation, written in cylindrical coordinates;    þ     þ  @ Hþ @H 1 @ @ 0@ H þ 0 D ð8Þ ¼ 0 0 Dr r @r @z @t 0 @r 0 @z 0 where r 0 is the radial distance from the center line of the root, and z 0 is the vertical distance from the soil surface. If the diffusion coefficient, D, is constant then equation (8) can be rewritten as "        # þ @ Hþ @2 Hþ @2 Hþ 1@ H ¼D 0 þ þ ð9Þ r @r 0 @t 0 @z 02 @r 0 2 To take into account the root growth rate, we pick our reference frame to be stationary relative to the tip of the root. In other words, our reference frame moves with the root growth zone at the growth rate of the tip of the root. In this reference frame, the soil would appear to move around the root just as a person on a moving glass bottom boat would observe, in looking at the water through the glass, that the water is moving around the boat. As described in Section III, this transformation is familiar to those who do root growth analysis. Mathematically, in this reference frame we have the following coordinate transformation: t ¼ t 0;

r ¼ r 0;

z ¼ z 0  Vt

ð10Þ

where V is the growth rate (and direction) of the tip of the root. Here we assume that the root only grows in the axial direction, z 0 . So our model, taking into account the motion of the growth of the root, is given by; @½Hþ  @½Hþ  ¼V @t @z " # 1 @½Hþ  @2 ½Hþ  @2 ½Hþ  þ þ þD r @r @r2 @z2

ð11Þ

The solution to equation (11) can be characterized by a dimensionless number, R ¼ VL D , called the rhizosphere number, analogous to the Pe´clet number of transport theory (Tennekes and Lumley, 1994). The variable L represents a characteristic length scale; for most roots L can be taken to be 1 mm. Physically, large R implies convection (i.e., root elongation) is more important than diffusion in dispersing the protons; small R implies that diffusion is relatively more important. Dimensional analysis reveals that the rhizosphere

number can be used to predict the extent of the rhizosphere of the growth zone and the time required for it to become steady. C.

Radial pH Patterns in the Rhizosphere of the Growth Zone

Our convection diffusion model predicts that after several hours of growth in soil a steady pH pattern will surround the moving root tip, and that acidification of soil pH will be detectable only within 1 mm from the root surface (Kim et al., 1999). A second prediction is that after just 50 min of incubation an agar medium, the root will acidify the surrounding substrate to a distance of > 5 mm. The wider zone of root influence in agar is due to the 2000-fold increase in proton diffusivity in agar relative to soil. The time to achieve a steady-state pH pattern around the tip of a root growing in agar or aqueous solution is on the order of 1 day. These predictions were obtained by solving equation (11) using our empirical data for diffusivities in sandy soil (Nichol and Silk, in press) and Felle’s data (Felle, 1998) for proton flux across roots growing in solution culture. Recently we made some empirical tests of the published predictions of the convection diffusion model (Nichol and Silk, in press). We could not make a precise comparison between our measured pH fields and the predictions of our model, because our boundary conditions (initial substrate pH and proton flux across the root surface) were different from those in Felle’s experiments. However, we were able to test the robust conclusions for rhizosphere dimensions. We measured pH at different radial distances from the root surface to test these predictions of the theory (Table 2). The sandy soil was found to be acidified at r ¼ 0:2 mm but not at r ¼ 1, 3, or 5 mm from the root surface at the locations z = 4 and 10 mm behind the root tip. In contrast, the agar was acidified at all measured radial locations, r = 0.2, 1, 3, and 5 mm from the root surface. Thus, at least qualitatively, the empirically determined dimensions of the rhizosphere in the different substrates support the theoretical predictions of the convection diffusion model. In summary, to study pH in the rhizosphere of growth zones we needed to modify the classical model to include the growth rate and a spatially varying flux of Hþ ions from the root surface. Use of a moving reference frame attached to the root tip simplified the problem. We retained the idea that diffusion is the main process for the distribution of Hþ in the rhizosphere. Our major conclusion from preliminary

124

Silk

Table 2 Empirical Verification of Rhizosphere Dimensions in Different Substrates (initial bulk substrate pH
6:2) Soil pH measurements after 18 h Radial distance from surface (mm) Axial distance from tip (mm)

4 10

0:2a 5.43b 5.74b

(SD) (0.23) (0.14)

0–2 6.17 6.20

(SD) (0.12) (0.10)

2–4 6.31 6.11

(SD) (0.19) (0.31)

4–6 6.10 6.13

(SD) (0.14) (0.16)

0 10

5.44b 5.64b

(0.15) (0.09)

5.62c 5.75

(0.14) (0.17)

5.79c 5.84c

(0.11) (0.08)

5.82c 5.92c

(0.11) (0.09)

Agar pH measurements after 50 min

a

pH measurements in substrate directly behind the growing root. Sign test shows rhizosphere is more acidic than the bulk soil (P < :01). c Friedmann test shows an influence of the root on the pH of the substrate (P < :05). Source: Nichol and Silk, in press. b

microprobe measurements, as well as from the numerical simulations, is that the inclusion of the root growth rate leads to a steady (time invariant) pH pattern on the root surface and in the rhizosphere of the root growth zone. In contrast, the stationary soil particles change in pH with time, as the root tissue next to a particular soil element is located progressively farther from the moving root tip. D.

Implications of the Convection Diffusion Model for Plant Growth and Nutrition

For the soil, the effect of a buffered growth zone is transient. This is because the growth zone moves rapidly—on the time scale of hours to a day—through a soil layer. However, for the growth zone the pH can be constant during the time required for many populations of cells to be displaced through the structure. This conservation of surface pH in the growth zone is probably important for the physiology of the root extension. It may also have an important effect on nutrition of the growing tissue, as suggested in Chapter 36 by Neumann and Ro¨mheld on rhizosphere chemistry. A branched rooting system extends, manufacturing a collection of locally acidified (or alkalinized) microenvironments that surround the tips as they penetrate to new locations. The conceptual framework for our recently published convection diffusion model draws attention to the inadequacy of the classical ideas for predicting properties of the rhizosphere around the growth zone. The soil adjacent to the mature part of the root experiences continuing proton flux, so that, according to the classical model, pH of the rhizosphere will change logarithmically with time (Nye, 1981). However, unlike the nongrowing part of the root, the growth zone, located at the moving root tip, continu-

ally changes its position with respect to the stationary soil particles. Furthermore, the expansion zone is extremely active as a sink for nutrient deposition, so that fluxes of cations and anions are an order of magnitude higher in the growth zone than in the mature part of the plant (compare Mengel and Barber, 1974, to Silk et al., 1986). Thus even on a time scale of hours, the growth zone can cause changes in the chemistry of the adjacent soil. The convection diffusion model is of general interest. For any substance that is secreted or taken up by growing tissue and that moves down a concentration gradient in the soil, the model can be used to characterize the spatial and temporal distribution patterns within the rhizosphere. Thus, from an ecological perspective, our model gives insight into many interactions between moving growth zones and the soil. Solution of the model equations revealed the possibility for steady patterns to exist in the rhizosphere of the growth zone, if the soil is homogeneous (Kim et al., 1999). This leads to the insight that as it moves through successive soil layers the root tip can surround itself with a constant pH field. However, this paradigm needs to be tempered by the realization that proton and bicarbonate fluxes across the root surface may be strongly affected by environmental conditions, especially soil physical and chemical properties. Recent work has already shown that the axial pH pattern on the root surface depends on external pH if the root is grown in solution culture (Peters and Felle, 1999). Preliminary work in our laboratory suggests that proton fluxes across the surface of roots grown in solution culture may differ in magnitude from those for roots grown in soil. Dimensional analysis indicates that the results of the model are robust in predicting a larger rhizosphere and longer time dependence in agar (or solution culture) relative to soil (Kim et al., 1999),

Kinematics of Primary Growth

and these predictions have been empirically verified (Nichol and Silk, in press). However, it may not be possible to solve the convection diffusion model for one growth medium, and scale the results according to the proton diffusivity of other media to obtain an accurate description of the pH field in particular soils. Quantitative analysis of the important problem of rhizosphere chemistry around the root tip will require both an understanding of the model relating flux to pH field and detailed empirical studies to characterize the proton flux patterns in different soils. The technology for measurement of pH gradients on the millimeter scale is now available, but work remains to show the interactions between rhizosphere pH and the nutrition of the growth zone and to find the effects of important environmental variables such as soil moisture content, mechanical impedance, and nutrient availability.

VIII.

TECHNOLOGY IS AVAILABLE TO MAKE KINEMATIC ANALYSES FASTER AND SIMPLER

Kinematic analysis of plant growth, involving analysis of particle trajectories, velocities and relative elemental growth rates, is facilitated by the technology of automated image analysis. This decade has seen the commercial availability of digital cameras that can be interfaced with high-speed computers with gigabytes of memory. Flexible software packages can now be developed to replace the laborious digitizing techniques of the last century (Schmundt et al., 1998). This should streamline kinematic studies and lead to an extensive database quantifying genetic, developmental, and environmental effects on growth. A comprehensive database of the growth patterns will provide a sound basis for the understanding of the physiology of root growth.

REFERENCES Barlow PW, Brain P, Parker JS. 1991. Cellular growth in roots of a gibberellin-deficient mutant of tomato (Lycopersicon esculentum Mill.) and its wild-type. J Exp Bot 42:339–351 . Baskin TI, Bivens NJ. 1995. Stimulation of radial expansion in Arabidopsis roots by inhibitors of actomyosin and vesicle secretion but not by various inhibitors of metabolism. Planta 197:514–521. Baskin TI, Meekes HTHM, Liang BM, Sharp RE. 1999. Regulation of growth anisotropy in well-watered and water-stressed maize roots. II. Role of cortical micro-

125 tubules and cellulose microfibrils. Plant Physiol 119:681–692. Beemster GTS, Baskin TI. 1998. Analysis of cell division and elongation underlying the developmental acceleration of root growth in Arabidopsis thaliana. Plant Physiol 116:1515–1526. Beemster GTS, Baskin TI. 2000. STUNTED PLANT1 mediates effects of cytokinin, but not of auxin, on cell division and expansion in the root of Arabidopsis. Plant Physiol 124:1718–1727. Degenhardt J, Larsen PB, Howell SH, Kochian LV. 1998. Aluminum resistance in the Arabidopsis mutant alr-104 is caused by an aluminum-induced increase in rhizosphere pH. Plant Physiol 117:19–27. Erickson RO, Sax KB. 1956. Elemental growth rate of the primary root of Zea mays. Proc Am Phil Soc 100:487– 498. Erickson RO, Silk WK. 1980. The kinematics of plant growth. Sci Am 242:134–151. Felle HH. 1998. The apoplastic pH of the Zea mays root cortex as measured with pH-sensitive microelectrodes: aspects of regulation. J Exp Bot 49:987–995. Gandar P. 1980. The analysis of growth and cell production in root apices. Bot Gaz 141:131–138. Gandar P. 1983. Growth in root apices. I . The kinematic description of growth. Bot Gaz 144:1–10. Gijsman AJ. 1990. Rhizosphere pH along different root zones of Douglas-fir (Pseudotsuga menziesii), as affected by source of nitrogen. Plant Soil 124:161–167. Gollany HT, Schumacher TE. 1993. Combined use of colorimetric and microelectrode methods for evaluating rhizosphere pH. Plant Soil 154:151–159. Goodwin RH, Stepka W. 1945. Growth and differentiation in the root tip of Phleum pratense. Am J Bot 43:36–46. Green P. 1976. Growth and cell pattern formation on an axis. Critique of concepts, terminology and modes of study. Bot Gaz 137:187–202. Ha¨ussling M, Leisen E, Marschner H, Ro¨mheld V. 1985. An improved method for nondestructive measurements of the pH at the root–soil interface (rhizosphere). J Plant Physiol 117:371–375. Hejnowicz Z. 1984. Trajectories of principal directions of growth, natural coordinate system in growing plant organ. Acta Soc Bot Polon 53:29–42. Hejnowicz Z, Karczewski J. 1993. Modeling of meristematic growth of root apices in a natural coordinate system. Am J Bot 80:309–315. Henriksen GH, Raman DR, Walker LP, Spanswick RM. 1992. Measurement of net fluxes of ammonium and nitrate at the surface of barley roots using ion-selective microelectrodes. 2. Patterns of uptake along the root axis and evaluation of the microelectrode flux estimation technique. Plant Physiol 99:734–747. Jaillard B, Ruiz L, Arvieu J-C. 1996. pH mapping in transparent gel using color indicator videodensitometry. Plant Soil 183:85.

126 Jirasek C, Prusinkiewicz P, Moulia B. 2000. Integrating biomechanics into developmental plant models expressed using L-systems. In: Spatz H-C, Speck T, eds. Plant Biomechanics. Stuttgart, Germany: Georg Thieme Verlag, pp 615–624. Kim TK, Silk WK, Cheer AY. 1999. A mathematical model for pH patterns in the rhizospheres of growth zones. Plant Cell Environ 22:1527–1538. Liang BM, Sharp RE, Baskin TI. 1997. Regulation of growth anisotropy in well-watered and water-stressed maize roots. I. Spatial distribution of longitudinal, radial, and tangential expansion rates. Plant Physiol 115:101–111. Morris AK, Silk WK. 1992. Use of a flexible logistic function to describe axial growth of plants. Bull Math Biol 54:1069–1081. Mengel DB, Barber, SA. 1974. Rate of nutrient uptake per unit of corn under field conditions. Agron J 70:695– 698. Muller B, Stosser M, Tardieu F. 1998. Spatial distributions of tissue expansion and cell division rates are related to irradiance and to sugar content in the growing zone of maize roots. Plant Cell Environ 21:149–158. Nichol SA, Silk WK. Empirical evidence for a convectiondiffusion model for pH patterns in the rhizospheres of root tips. Plant Cell Environ (in press). Nye P. 1981. Changes of pH across the rhizosphere induced by roots. Plant Soil 61:7–26. Pahlavanian A, Silk WK. 1988. Effect of temperature on spatial and temporal aspects of growth in the primary maize root. Plant Physiol 87:529–532. Ober ES, Sharp RE. 1994. Proline accumulation in maize (Zea mays L.) primary roots at low water potentials. I. Requirement for increased levels of abscisic acid. Plant Physiol 105:981–987. Peters WS, Bernstein N. 1997. The determination of relative elemental growth rate profiles from segmental growth rates: a methodological evaluation. Plant Physiol 113:1395–1404. Peters WS, Felle HH. 1999. The correlation of profiles of surface pH and elongation growth in maize roots. Plant Physiol 121:905–912. Peters WS, Fricke W, Chandler PM. 1999. XET-related genes and growth kinematics in barley leaves. Plant Cell Environ 22:331–332. Sacks MM, Silk WK, Burman P. 1997. Effect of water stress on cortical cell division rates within the apical meristem of primary roots of maize. Plant Physiol 114:519– 527. Schmundt D, Stitt M, Jahne B, Schurr U. 1998. Quantitative analysis of the local rates of growth of dicot leaves at a high temporal and spatial resolution, using image sequence analysis. Plant Journal 16:505–514. Sharp RE, Silk WK, Hsiao TC. 1988. Growth of the maize primary root at low water potentials. Plant Physiol 87:50–57.

Silk Sharp RE, Hsiao TC, Silk WK. 1990. Growth of the maize primary root at low water potentials. II. The role of growth and deposition of hexose and potassium in osmotic adjustment. Plant Physiol 93:1337–1346. Silk WK. 1984. Quantitative descriptions of development. Annu Rev Plant Physiol 35:479–518. Silk WK. 1989. Growth rate patterns which maintain a helical tissue tube. J Theor Biol 138:311–327. Silk WK. 1992. Steady form from changing cells. Int J Plant Sci 153:S49–S58. Silk WK, Erickson RO. 1978. Kinematics of hypocotyl curvature. Am J Bot 65:310–319. Silk WK, Erickson RO. 1979. Kinematics of plant growth. J Theor Biol 76:481–501. Silk WK, Wagner KK. 1980. Growth sustaining water potential distributions in the primary corn root. Plant Physiol 66:859–863. Silk WK, Walker RC, Labavitch J. 1984. Uronide deposition rates in the primary root of Zea mays. Plant Physiol 74:721–726. Silk WK, Hsiao TC, Diedenhofen U, Matson C. 1986. Spatial distributions of potassium, solutes, and their deposition rates in the growth zone of the primary corn root. Plant Physiol 82:853–858. Spollen WG, LeNoble ME, Samuels TD, Bernstein N, Sharp RE. 2000. Abscisic acid accumulation maintains maize primary root elongation at low water potentials by restricting ethylene production. Plant Physiol 122:967–976. Sugimoto K, Williamson RE, Wasteneys GO. 2000. New techniques enable comparative analysis of microtubule orientation, wall texture, and growth rate in intact roots of Arabidopsis. Plant Physiol 124:1493–1506. Taylor AR, Bloom AJ. 1998. Ammonium, nitrate, and proton fluxes along the maize root. Plant Cell Environ 21:1255–1263. Tennekes H, Lumley JL. 1994. A First Course in Turbulence. Cambridge, MA: MIT Press. Van der Weele CM, Spollen WG, Sharp RE, Baskin TI. 2000. Growth of Arabidopsis thaliana seedlings under water deficit studied by control of water potential in nutrient-agar media. J Exp Bot 51:1555–1562. Walter A, Silk WK, Schurr U. 2000. Effect of soil pH on growth and cation deposition in the root tip of Zea mays L. J Plant Growth Regul 19:65–76. Weisenseel MH, Dorn A, Jaffe LF. 1979. Natural H+ currents traverse growing roots and root hairs of barley (Hordeum vulgare L.). Plant Physiol 64:512–518. Yan X, Lynch JP, Beebe SE. 1996. Utilization of phosphorus substrates by contrasting common bean genotypes. Crop Sci 36:936–941. Zhong H, La¨uchli A. 1993. Spatial and temporal aspects of growth in the primary root of cotton seedlings: effects of NaCl and CaCl2. J Exp Bot 44:763–771.

8 Lateral Root Initiation Pedro G. Lloret and Pedro J. Casero Universidad de Extremadura, Badajoz, Spain

I.

INTRODUCTION

1. The first step is a division of the founder cells. This has initiated an interesting controversy. Do founder cells cease to divide as they leave the apical meristem but later recover the capacity to proliferate far from the apical meristem? Alternatively, does the proliferation of the founder cells not really stop? 2. Founder cells undergo transversal, periclinal, anticlinal, and oblique divisions giving rise to a group of cell derivatives in which it is difficult to recognize the future tissues of the LRs. Such cell divisions produce the cells in the LR primordium. This event includes induction of important changes in the cell walls, cytoskeleton, cytoplasm, nuclear displacement, genetic expression, etc. 3. Derivative cells differentiate acquiring the morphological features of the LR tissues that show the same features of the parent roots. 4. LRs originate endogenously from tissues lying inside the parent root. Therefore, they must grow through the parents root tissues in order to emerge. 5. LRs may originate either from a single parent tissue or from several tissues. 6. The parent tissues adjacent to the lateral roots can undergo major structural and functional changes to facilitate the growth of the lateral root inside the parent root and its emergence. 7. Regulation by hormones such as auxins and cytokinins. 8. The founder cells are similar to the neighbor cells but the lateral root primordia are delimited structures separated from the parent tissues. This constitu-

Lateral roots (LRs) as defined here are those that originate from some other root. This means that the lateral roots can be derived either from a seminal root, an adventitious root, or another lateral root. The lateral roots, therefore, constitute almost the whole root system. They influence many aspects that are so important for the plant’s development as its adaptability to the environment and its absorption capacity. Lateral roots affect plant size, plant production, vitality, etc. All these aspects justify the scientific interest that the study of lateral roots has always earned. Moreover, in many plants the lateral roots originate far from the apex. Therefore, cell proliferation during lateral root development does not overlap the cell proliferation of the apical meristem. The two processes are temporally and spatially separated. In these plants, the analysis of lateral roots is more relevant because these roots originate from a very few parent cells, called founder cells, which are very long and highly vacuolated mature cells. These cells proliferate and give rise to very short derivative cells, which are nearly isodiametric, with thinner cell walls, a marked increase in cytoplasmic basophilia and volume, a pronounced nucleolar enlargement, and numerous small vacuoles. These short cells show the typical morphological features of meristematic cells, suggesting that the founder cells undergo a dedifferentiation process. Lateral root initiation and development involve the following: 127

128

Lloret and Casero

tes a very attractive developmental model because the founder cells must be targets for the hormones, diffusing molecules, etc. 9. The LRs are located in a specific position in relation to the vascular structures, forming longitudinal ranks along the parent roots. 10. The distribution of the lateral roots along the parent roots is not at random and is related to the environmental conditions, allowing optimal utilization of soil resources. Excellent reviews have been published on LR development (Van Tieghem and Douliot, 1888; Torrey, 1961, 1986; Von Guttenberg, 1968; McCully; 1975; Peterson and Peterson, 1986; Charlton, 1991, 1996). After Zobel’s (1975, 1991) general reviews on the genetics of root development, a more specific analysis of the genetics of LR development is lacking. Probably the reason has been that this research field has been in continuous growth during recent years. Now, we are witnessing the maturing of the great array of molecular, genetic and cellular techniques that constitute a very reliable and consistent research system which will ensure significant results in the near future leading to a better understanding of the control mechanisms of lateral root initiation and development. II.

EVENTS DURING LATERAL ROOT FORMATION

A.

Sequence of Formation

Roots elongate continuously due to the activity of their apical meristem. Since LRs are formed from relatively young parent root tissues and usually form in an acropetal sequence (Abadı´ a-Fenoll et al., 1986), they are initiated between the previously formed LRs and the apical meristem (Fig. 1). Nevertheless, the formation of LRs sometimes proceeds in an order that is not strictly acropetal (MacLeod, 1990; Pellerin and Tabourel, 1995). The acropetal origin of LRs should thus be taken as a trend rather than a rule. One possible cause for altering the basic scheme of acropetal formation might be that the activity of the different cell ranks is not synchronized. Charlton (1975) described in Pontederia cordata the existence of retarded ranks producing their distalmost new LR primordia at a greater distance from the apex than more advanced ranks. This results in small LR primordia intercalated between others at later stages of development. In some extreme cases, the competence of the parental root pericycle or endodermis to initiate LR primordia lasts for so long that in broad regions of

the parent root new and old LR primordia can coexist. This has been described in root systems of Zea mays. Initiation of LR primordia of Zea mays is not confined to the subapical region but new anlages grow up to as far as 5–8 cm from the apex (MacLeod, 1990). A similar situation occurs in Triticum aestivum: from the root apex to the base the frequency of LR primordia increases to a maximum located 40–50 mm from the apex (Bingham et al., 1997). New primordia of wheat also appear in tissues that already contain visible primordia. The appearance of small primordia between emerged growing LRs (Fig. 2) does not imply a nonacropetal sequence of development. It may happen that during the development of LR growth of some primordia is delayed. A slower growth of late-formed primordia would also yield images such as that of Fig. 2. Nevertheless, it appears that commonly all primordia eventually emerge even if initiated out of sequence (Lloret and Pulgarı´ n, 1992; MacLeod and Thompson, 1979).

Figure 1 Segment of primary root of Zea mays showing the development of LRs in an apparent acropetal order, with longer LRs at the basal most region and progressively shorter LRs toward more apical regions. Bar: 20 mm.

Lateral Root Initiation

129

B.

Figure 2 Intact maize root with young LRs showing short LRs between longer ones. This suggests that the sequence of formation is not strictly acropetal. Note that there is a clear LR dimorphism with narrow LRs (arrow) which are longer than broad LRs (arrowhead). Bar: 1 mm.

The existence of dormant or late-formed LR primordia presents a problem for studies of the LR distribution pattern. Indeed, if these minute anlages are not detected under the microscope, the pattern of externally apparent LR distribution in a given species may be obscured (Mallory et al., 1970). The possibility that LRs emerge out of sequence is a normal situation for roots developing under experimental conditions (Charlton, 1996). Nevertheless, the formation of LRs in an inverse order to the natural acropetal sequence is unusual. To the best of our knowledge, there has only been one report of such finding (Dyanat-Nejad and Neville, 1972). This basipetal sequence of LR development occured on the preformed root region in the germinative radicle of Theobroma cacao. This region has a poorly defined transition to the hypocotyl, and consequently these roots can really be adventitious roots originated from the hypocotyl rather than true LRs.

General Aspects of Lateral Root Development

In some plants, the LRs originate so close to the apical meristem that either none of the vascular tissues of the main root is differentiated or only the protophloem appears to be mature (Mallory et al., 1970; McCully, 1975). However, it is more common to find plants in which LRs are initiated from region including well differentiated tissues, away from the apical meristem (Bell and McCully, 1970; McCully, 1975; Blakely et al., 1982; Charlton, 1983b; Lloret et al., 1988, 1989; Pulgarı´ n et al., 1988; Casero et al., 1989, 1993, 1995, 1996; Seago and Marsh, 1990). It is generally accepted that the distance from the apex of the main root to the point of appearance of the LRs is fairly constant and is characteristic for each species (Torrey, 1961; Charlton, 1991). Plants in which the LRs originate far from the apex of the mother root are interesting for studies of LR initiation, because the cell proliferation involved in LR initiation does not overlap the cell proliferation in the apical meristem. The two phenomena are separated in time and space. However, new findings regarding LR initiation raises the question whether cell proliferation really ceases in the apical meristem and is then reactivated to give rise to the LRs. Root anatomy is relatively uncomplicated. Roots of primary growth are cylindrical with a diameter that is practically constant except at the apex. At the center of such roots is the vascular cylinder, which presents a characteristic alternating arrangement of the conducting tissues, xylem and phloem. In either case, the outermost cells of the two tissues, xylem and phloem, are in contact with the pericycle, that usually forms a continuum and delimits the outside of the vascular cylinder. LRs originate endogenously from the endodermis in ferns (Clowes, 1961; McCully, 1975; Lin and Raghavan, 1991; Charlton, 1996) and from the pericycle in gymnosperms and angiosperms (McCully, 1975; Peterson and Peterson, 1986; Fahn, 1990; Charlton, 1996). In some angiosperms, both the pericycle and endodermis contribute to the tissues of the LR, although in many cases the derivatives of the endodermis are short-lived (reviewed by McCully, 1975). Therefore, the early LR development occurs inside the parent roots (Fig. 3). The LRs are located in defined relation to the vascular pattern (Fig. 4). In some species, they appear close to the xylem poles whereas in other species they appear next to phloem poles (Van Thieghem and Douliot, 1888; Esau, 1977;

130

Figure 3 A lateral root primordia ( ) can be observed attached to an xylem rank inside an intact adventitious root. Bar: 200
m.

Nishimura and Maeda, 1982; Lloret et al., 1989; Fahn, 1990; Casero et al., 1995). Thus, it can be assumed that in the pericycle, cells located opposite either xylem or phloem poles (depending on the species) and capable of initiating LR primordia alternate with cells which are apparently unable to initiate LR primordia. What

Lloret and Casero

is the factor responsible of this difference is still far from being answered. In carrot, LRs originate from pericycle cells located next to the phloem poles where the pericycle is unilayered while pericycle in front of the xylem is bistratified (Esau, 1977; Lloret et al., 1989; Knox et al., 1989; Casero et al., 1995, 1998). However, in Pisum sativum, LRs initiate from a multilayered pericycle in front of the xylem while the pericycle is single-layered in front of the phloem (Lloret et al., 1989; Casero et al., 1998). In many plants, including Allium cepa, Raphanus sativus, Helianthus annuus, and Arabidopsis thaliana, in which the pericycle consists of only one layer of apparently uniform cells, initiation of LR primordia only occurs in the pericycle in front of the xylem (Blakely et al., 1982; Lloret et al., 1989; Casero et al., 1995; Laskowski et al., 1995). Therefore, the number of layers in the pericycle has its effects but does not determine its capacity to form LRs. In onion, carrot, and pea, species which show very different pericycle patterns, Lloret et al. (1989) and Casero et al. (1989b) have describe two types of pericycle cells distinguishable by their different length and position with respect to the vascular pattern. In onion and pea, the shorter cells were found opposite xylem poles, where LR initiation occurs, whereas the longer cells were opposite the phloem poles. This is true also

Figure 4 Transverse section of an onion adventitious root. After successive periclinal divisions, the pericycle (P) becomes multiseriate between three xylem poles (X). Bar: 30
m.

Lateral Root Initiation

for A. thaliana (Laskowski et al., 1995). In carrot, the shorter cells were found adjacent to the phloem poles, where LR initiation occurs. Thus, apparently the shorter pericycle cells seem to be better suited for LR initiation. Because roots normally exhibit symplastic growth, differences in mean length between pericycle cells located opposite xylem and phloem poles at any given distance from the tip reflect differences in rates of transverse division between these two cell types in the meristem (Luxova´, 1975; Webster and MacLeod, 1980; Rost et al., 1988; Casero et al., 1989b). Therefore, cell length, transverse proliferation in the meristem, and LR initiation could be related. While this is a very attractive hypothesis, it is necessary to remember that LR development is a spatially located phenomenon in which only a very limited number of grouped cells of the mother root are involved. This number varies considerably depending on the species (MacLeod and Thompson, 1979). Thus, in radish, a group of 30 pericycle cells originate a LR (Blakely et al., 1982), while in A. thaliana only 11 adjacent pericycle cells are involved in the formation of an LR (Laskowski et al., 1995). Like the rest of the root tissues, pericycle cells form vertical columns of cells that can be traced to initial cells in the root apical meristem. Many columns of such cells are involved in the process of LR initiation in the root. Therefore, it is difficult to

131

explain why at any transverse level only one LR is formed in front of one of the xylem poles, and why LRs are not initiated in front of the remaining xylem poles at the same distance from the apex although the corresponding opposite pericycle cells are just as short (Casero, unpublished data). Moreover, the longitudinal extension of the LR primordia suggests that very few pericycle cells in the same column can be involved in the initiation. Adjacent pericycle cells located both above and below those involved in the initiation are similar in length to the pericycle cells from which the LR is initiated, but they are unable to initiate a LR under normal conditions. All this suggests that several interacting factors regulate the initiation of the LRs by discriminating between adjacent pericycle cells. In angiosperms, periclinal division of a few pericycle cells is one of the most common morphological criteria used to define LR initiation (Esau, 1940, 1977; Clowes, 1961; Foard et al., 1965; Bonnett and Torrey, 1966; Bell and McCully, 1970; Charlton, 1975; McCully, 1975; Blakely et al., 1982; Lloret et al., 1989; Fahn, 1990; Seago and Marsh, 1990). Periclinal divisions are very noticeable in cross section in plants with an uniseriate pericycle because the two radially arranged derivatives can be easily recognized in contrast with the rest of the pericycle (Blakely et al., 1982; Lloret et al., 1989; Casero et al., 1993) (Fig. 5). Just before the first periclinal divisions, pericycle cells undergo a noticeable

Figure 5 Transverse section of an onion adventitious root showing one of the pericycle cells in front of the xylem undergoing a periclinal division (arrowhead). C, cortex; E, endodermis; P, pericycle; X, xylem. Bar: 30
m.

132

radial enlargement (Popham, 1955; Clowes, 1961; Bonnett and Torrey, 1966; Charlton, 1975; Casero et al., 1993), a marked increase in cytoplasmic basophilia and volume (Bell and McCully, 1970; Karas and McCully, 1973), and a pronounced nucleolus enlargement (Seago, 1973). It is particularly interesting that pericycle cells undergoing periclinal divisions are very short (Van Tieghen and Douliot, 1888; Bell and McCully, 1970; McCully, 1975; Blakely et al., 1982; Seago and Marsh, 1990; Laskowski et al., 1995; Malamy and

Lloret and Casero

Benfey, 1997). In onion, for instance, they are only 30–50
m long (Fig. 6) (Lloret et al., 1989; Casero et al., 1993). Since LR initiation occurs after cell elongation has terminated, elongated and highly vacuolated pericycle cells must divide transversely to become short enough before dividing periclinally (Van Tieghem and Douliot, 1888; Bell and McCully, 1970; Blakely et al., 1982; Lloret et al., 1989; Malamy and Benfey, 1997). In the apical meristem, cells divide and elongate, but beyond the meristem zone cells elongate only. At a certain distance from the tip elongation ceases. In onion, no mitotic cells can be found beyond 1600–1800
m from the apex. At this level, the mean length of the pericycle cells is < 60
m (Casero et al., 1989a). Then, pericycle cells increase in length. Between 6 and 7 mm from the apex, all pericycle cells are highly vacuolated, containing a central nucleus and measuring over > 200
m in length. Transverse divisions are a necessary stage prior to the periclinal divisions. This led Casero et al. (1993, 1995, 1996) to focus their attention to the precise sequence of cellular events from which such short pericycle cells are derived in different plants. The resulting proliferation pattern is original and very attractive and allows clear identification of the site of LR initiation closer to the apex than previously documented.

C.

Figure 6 Longitudinal section of an onion adventitious roots showing a pericycle column (P) in front of one of the xylem poles (X). One short pericycle cell within a group of short cells has undergone periclinal division, forming two cells located at the same transversal level (arrowhead). Bar: 30
m.

Transverse Division in the Pericycle Subsequent to Elongation

The occurrence of transversal divisions in pericycle cells proximal to the elongation zone has been noted in onion and in other species in which LRs are initiated far from the apex (Van Tieghem and Douliot, 1888; Bell and McCully, 1970; Lloret et al., 1989). In roots of Allium cepa and Pisum sativum, Lloret et al. (1989) describe pericycle cells undergoing symmetric (Fig. 7) and asymmetric transversal divisions (Fig. 8) closer to the apex than where the periclinal divisions occur. Symmetric transverse divisions can occur in pericycle cells situated either opposite the xylem or opposite the phloem, but do not give rise to cells as short as those which can divide periclinally (Casero et al., 1993). Asymmetric transverse divisions occur only in pericycle cells situated in front of xylem, giving rise to short cells which can divide periclinally (Casero et al., 1993). The results strongly suggest that these asymmetric divisions are related to LR initiation.

Lateral Root Initiation

133

Figure 8 Longitudinal section of an onion adventitious root showing a pericycle cell undergoing an asymmetrical transverse division in front of one of the xylem poles. C, cortex; E, endodermis; P, pericycle; X, xylem. Bars: 20
m.

Figure 7 Longitudinal section of an onion adventitious root showing a pericycle cell undergoing a symmetrical transverse division in front of one of the xylem poles. C, cortex; E, endodermis; P, pericycle; X, xylem. Bars: 20
m.

D.

Asymmetric Transverse Divisions During the Earliest LR Development in Vascular Plants

The pericycle cells that are involved in the initiation and development of the LRs can be named ‘‘founder pericycle cells’’ (FPC) following the definition given by Laskowski et al. (1995), because they are activated and divide giving rise to derivative pericycle cells which differentiate into the apical meristem and the tissues of the lateral roots. Although Casero et al. (1993, 1995, 1996) had previously used the term ‘‘mother pericycle cells’’ to define them, FPC is now more generally used.

The first morphological event related to the LR initiation can be observed in just one pericycle column. Approximately when elongation ceases, two elongated and highly vacuolated adjacent pericycle cells (FPC) located in the same column undergo nearly synchronous asymmetric transverse divisions. This occurs in front of one of the xylem poles in onion, radish, and sunflower and near one of the phloem poles in carrot and corn (Casero et al., 1993, 1995, 1996). Two short pericycle derivatives are then produced, lying end to end in the same column flanked above and below by the two longer pericycle derivatives (Fig. 9). The short cells tend to have a central nucleus, while the longer cells have nuclei, which tend to be displaced toward the short cells. After these transverse divisions, it is easy to distinguish the original extension of the FPCs because the new transverse cell walls are noticeably thinner than the older ones. Moreover, a small intercellular space can be seen between the two FPCs and the cells of the adjacent tissues. No intercellular space can be distinguished at the level of the new cell walls (Casero et al., 1993, 1995, 1996). Observations by

134

Figure 9 Longitudinal section of an onion adventitious root showing two short pericycle cells derivatives, lying end to end, which are flanked by two longer cells in front of one of the xylem poles. Arrowheads mark the original length of the founder pericycle cells. C, cortex; E, endodermis; P, pericycle; X, xylem. Bars: 30
m.

Demchenko and Demchenko (1996) in Triticum aestivum also strongly suggest that LR initiation occurs when two FPCs in front of the phloem undergo asymmetric transverse divisions like those described above. More recently, similar phenomena were observed in two adjacent FPC pericycle cells in front of the xylem poles in Lactuca sativa (Zhang and Hasenstein, 1999) and A. thaliana (Casimiro et al., unpublished results). These results allow us to generalize and assume that in angiosperms the polarized asymmetric transverse division of two adjacent FPCs is a general feature and can be considered the first recognizable stage of LR development. This objective morphological criterion allows clear identification of the site of LR initiation. In onion,

Lloret and Casero

radish, sunflower, corn, and carrot the pair of short pericycle cells appears
8–10 mm, 10–13 mm, 12–15 mm, 12–15 mm, and 12–15 mm, respectively, behind the tip, whereas the first periclinal division takes place at
21–23 mm, 15–18 mm, 19–21 mm, 21–24 mm, and 21–24 mm behind the tip, respectively (Casero et al., 1995). The distance behind the root apex where LRs are initiated is important when the proliferation of the pericycle cells from the apical meristem is considered. It is well known that in species in which LR initiation occurs opposite the xylem, the pericycle cells in front of the xylem divide further from the tip than the pericycle cells located in front of the phloem (Luxova´, 1975; Rost et al., 1988; Casero et al., 1989a). Therefore, the pericycle cells near the xylem are shorter than those next to the phloem. By contrast, in those species in which LRs appear next to the phloem, the pericycle cells located near the phloem are shorter than those next to the xylem (Lloret et al., 1989). This implies that, in these species, the pericycle cells close to the phloem divide further from the tip than those in front of the xylem. The question is whether or not the FPC cease to divide before LR initiation. The results presented by Jensen and Kavaljian (1958), Balodis (1968), Luxova´ (1975), Rost et al. (1988), and Casero et al. (1989a) regarding the occurrence of mitosis in the pericycle near the root apex, suggest that proliferation of the pericycle cells ceases, because no mitoses could be seen beyond a certain distance behind the tip. John et al. (1993) also suggested that proliferation ceases in the pericycle cells but they point out that the development of some pericycle cells is arrested in the G2 phase. These cells can, therefore, be induced to divide very quickly with auxin. Blakely et al. (1982) also speculate that the cycle of the pericycle cells stops at G2, except for that of the FPC. Our results indicating that LR initiation occurs nearer the tip support the hypothesis that pericycle cells do not leave the division cycle when they enter the elongation zone. Thus, in onion roots, the first periclinal divisions of the pericycle cells occur
21– 23 mm behind the tip, but the first asymmetric transverse divisions occur at
7 mm behind the tip (Casero et al., 1993). Assuming that an onion root increases its length by
15 mm/day (unpublished data) and that the cycle time lasts 13.5 h at 25 C (Gime´nez-Martı´ n et al., 1977), it is possible that a cell that undergoes its last division in the apex could divide again further behind the tip, for example,
7–8 mm behind the tip, where transversal divisions have indeed been seen. This observation strongly suggests that the pericycle cells involved

Lateral Root Initiation

in LR initiation continue dividing once they leave the meristem (Casero et al., 1993). Furthermore, the PFC show a very important characteristic, which distinguishes them from the rest of the pericycle cells: a regular pattern of cell divisions which is initiated with the coordinated asymmetric division of two neighboring cells (Casero et al., 1993, 1995). Asymmetric transverse division during the LR initiation is not exclusive to angiosperms. During LR initiation in ferns the nucleus of endodermal cells in front of xylem poles moves toward a position close to the wall and then divides asymmetrically giving rise to two cells of different sizes (Clowes, 1961; Lin and Raghavan, 1991). However, two main differences must be emphasized. In ferns, unlike higher plants, just one cell in front of the xylem divides asymmetrically and, furthermore, this is an endodermal cell. Studying polarized asymmetric transverse divisions during lateral root intiation in angiosperms opens up very attractive possibilities in plant development and cell biology. Recently, it has also been suggested, in comparing mutant and wild-type embryogenesis in A. thaliana, that GNOM gene activity is required for the asymmetric division of the zygote, and in its absence a symmetric division occurs which leads to other embryonic defects (Aeschbacher et al., 1994). The SCR (scarecrow) gene is required for the two distinct asymmetric cell divisons necessary for establishing the two ground tissue layers in the embryo and postembryonic root meristem (Dolan, 1997). Therefore, in root meristem of scr seedlings, the cortical/endodermal initial cells fail to undergo their asymmetric division which would produce the cortical and endodermal cell layers. These mutant roots have a single layer which shows attributes of both cortex and endodermis (Scheres et al., 1995; Di Laurenzio et al., 1996). The short-root (SHR) gene is required for the asymmetric cell division responsible for endodermis and cortex formation (Helariutta et al., 2000). These findings suggest that such asymmetric divisions depend on the expression of these genes. During LR initiation two pericycle cells, that appear identical to their neighbors, well differentiated and highly vacuolated, undergo asymmetric divisions almost simultaneously. A number of interesting questions arise from these observations: Are these asymmetrical divisions also conditioned by a gene expression? Why does it occur in just two cells out of so many? Does some signal induce the gene expression? Does the signal pass through both pericycle cells to synchronize them? Why do the two nuclei move in opposite directions?

135

The hypothesis of intercellular communication during LR initiation seems more significant after observing that a wave of proliferation extends from the first FPCs and as neighbor pericycle cells become successively FPCs they also undergo asymmetric transverse divisions. This has been studied in detail in onion roots between 10 and 15 mm behind the tip (Casero et al., 1996). The adjacent FPCs were observed to undergo polarized asymmetrical division in relation to the first pair of short cells. This occurs in FPCs located in the same column above and below the first two FPCs. Thus, alternating sequences of short and long pericycle derivatives can be found in the same column. At that time, proliferation extends laterally to the adjacent columns. Therefore, pairs of short pericycle cell derivatives located at the same transverse level can be seen in longitudinal tangential sections (Fig. 10). Asymmetric proliferation also extends upward and downward in these other columns. This means that new adjacent FPCs are successively involved in the LR initiation reproducing the division pattern of the first two FPCs. Moreover, a clear polarization is shown in all the new FPCs in relation to the first pair of short cell derivatives. The longer pericycle cell derivatives now undergo asymmetric transversal divisions, giving rise to new short cells close to those formed previously. These follow the same sequence as the first asymmetric transverse divisions. This explains the formation of the groups of short cells observed in onion between 15 and 20 mm behind the tip. These groups of short cells show a gradual increase in their radial diameter towards the center. The radial diameter of the longer cells gradually increases toward the end nearer the short cells (Casero et al., 1993, 1996). In onion roots, the first periclinal divisions of LR initiation were seen 20–25 mm behind the tip. The centermost cells of a group of short cells, which show a noticeable radial expansion, are generally the first cells to undergo periclinal divisions, giving rise to two radially arranged cells. This means that the first periclinal divisions occur in the older short pericycle derivatives. Periclinal proliferation then extends to the neighboring short cells following approximately their formation sequence (Casero et al., 1993, 1996). Subsequently, successive waves of cell divisions emulating that in the two pairs of FPCs initially engaged in LR formation (i.e., asymmetric transverse cell division, formation of groups of short cells, and periclinal divisions) extend centrifugally as a wave from the center of the LR primordia.

136

Lloret and Casero

serve as a signal that moves between cells. One could speculate that pericycle cells, which are involved in LR initiation, would form a symplastically isolated group of cells. In such a case, the signal could not move farther above or below, and would be restricted to the limits of the LR anlage. Experiments carried out by Duckett et al. (1994) strongly suggest that differentiating epidermal cells can become symplastically isolated from other epidermal cells and from the underlying cortex cells, forming a ‘‘symplastic unit.’’ Indeed, the symplastic isolation of cells during the development of a tissue or an organ has been widely observed (Carr, 1976; Kwiatkowska, 1988; Duckett et al., 1994). If the group of founder cells indeed form a symplastic unit which becomes isolated from its neighbors during pericycle differentiation, LR initiation could be an excellent model with which to study cellcell interactions in plants. However, the pericycle cells of onion located near the xylem poles are activated before the intervening pericycle cells in front of the phloem poles (Casero et al., 1996). It is difficult to explain this observation on the basis of the assumption that the signal moves symplastically, via plasmodesmata, to neighboring cells. But it cannot be ruled out that the cell cycle of the cells near the phloem is much longer than that of the cells near the xylem. Another possibility is that each group of neighboring pericycle cells is independently activated.

Figure 10 Tangential longitudinal section of onion adventitious root showing pairs of short pericycle cell derivatives located approximately at the same transversal level in adjacent pericycle columns. C, cortex; E, endodermis; P, pericycle; X, xylem. Bars: 30
m.

The above results suggest that the FPC could be symplastically coupled, allowing small molecules to pass from cell to cell, as has been observed in root epidermal cells of Arabidopsis by using Lucifer Yellow and carboxyfluorescein (Duckett et al., 1994). After injection into an epidermal cell, the dye was seen in surrounding epidermal cells, having moved preferentially to cells within the same column. In onion roots, pericycle cells in the same column also tend to divide before cells in adjacent columns. Jackson et al. (1994) suggested that changes in gene expression patterns precede, or at least are concomitant with, the determination of the position of primordia in the shoot meristem. The proteins, which are products of the expression of these genes, e.g., KN1 protein, could

E.

Development of Lateral Roots After Initiation

Malamy and Benfey (1997) have described in detail the lateral root development after the occurrence of the periclinal divisions in the relatively simple roots of A. thaliana. The structure generated has a radial organization similar to that of the mature root tip. Fig. 11 is an attempt to summarize the main development stages following the results of Malamy and Benfey (1997). Stage I: The pericycle file contains many short cell derivatives (Fig. 11, St I). It is now known that they are produced after asymmetric transverse divisions of two adjacent founder cells as in other plants (Casimiro et al., unpublished results). Stage II: Lateral root primordia contain two layers of short-cell pericycle derivatives due to the periclinal proliferation of the previously formed short derivatives (Fig. 11, St II). The inner layer was termed IL, and the outer one, OL. Arabidopsis thaliana, Allium cepa and Raphanus

Lateral Root Initiation

137

Figure 11 Schematic representation to summarize the main development stages following the results by Malamy and Benfey (1997). St I: The pericycle file contains many short cell derivatives. St II: Lateral root primordia contain two layers of short-cell pericycle derivatives due to the periclinal proliferation of the previously formed short derivatives. IL, inner layer; OL, outer layer. St III: The lateral primordia contain three layers because the OL layer divides periclinally again giving rise to OL1 and OL2 layers. St IV: The lateral primordia contain four layers because the IL layer undergo periclinal divisions giving rise to IL1 and IL2 layers. St V: The centermost cell derivatives of OL1 and OL2 divide transversely. St VI: The centermost cells of OL1 undergo periclinal divisions. These cells will give rise to the future cap. The rest of the OL1 cells do not undergo periclinal divisions and will produce the future epidermis. The centermost cells of OL2 will produce the future quiescent center. The rest of the cells of OL2 undergo a periclinal division, forming the layers OL2a and OL2b that will form the future cortex and endodermis, respectively. IL1 will form the future pericycle. The centermost cells of IL2 undergo a noticeable radial expansion and periclinal divisions. They will give rise to the rest of the vascular cylinder. C, future cortex; E, future endodermis; Ep, future epidermis; St, stage; VC, future vascular cylinder.  Future cap cells.

sativus are examples of plants in which the periclinal divisions are symmetric (Blakely et al., 1982; Casero et al., 1993, 1995, 1996; Malamy and Benfey, 1997). However, in other plants, such as Zea mays, periclinal divisions are asymmetric. The daughters closer to the endodermis are narrower than the others (Bell and McCully, 1970; Ashford and McCully, 1973; McCully, 1975; Casero et al., 1995). In Malva silvestris, Byrne (1973) showed that cells in the two layers had very different behavior although they were formed after a symmetrical periclinal division. More recently, Malamy and Benfey (1997) have elegantly demonstrated in A. thaliana seminal roots that, although IL and OL are morphologically similar, they also present distinct identities as is shown by the differential gene expression after a promoterless GUS gene is inserted approximately 1 kb upstream of the start site of the SCR gene. Cells may also have different identities within a single layer (Malamy and Benfey, 1997). Results by Karas and McCully (1973) show major changes in the cell walls preceding the periclinal division indicating that in corn enzymatic hydrolysis must be involved in these changes.

Stage III: The OL layer divides periclinally again. The two new layers, OL1 and OL2, replace the OL layer, with OL1 being the outer one (Fig. 11, St III). Stage IV: The lateral primordia contain four layers because the IL undergo periclinal divisions giving rise to two layers (IL1 and IL2), with IL2 being the closer to the seminal vascular cylinder (Fig. 11, St IV). The pericycle probably arises from IL1 while the other stele cells arise from proliferation of IL2 (Malamy and Benfey, 1997). In Ipomoea purpurea, Seago (1973) also described the formation of four layers of pericycle derivatives with the inner layer dividing before the outer one. The author suggested that the initial cells were formed from these layers. The two inner layers gave rise to the vascular initial cells, the next one to the cortex, and the outermost to the initial cells of the capepidermis. Stage V: The centermost cell derivatives of OL1 and OL2 divide transversely. Proliferation then extends to the adjacent cell derivatives in these layers. The most centered IL2 cells undergo expansion and division distorting the shape of IL1 and OL2. (Fig. 11, St V).

138

Stage VI: This development stage is very interesting because the main cellular layers of the future mature root are established (Fig. 11, St VI). Malamy and Benfey (1997) showed that these layers can be recognized by means of cell line markers. There are no marker lines that specifically stain the initial cells. However, in the stage VI primordium there are cells that can be tentatively identified as initial cells, based on their position at the ends of differentiated cell files and comparison with the positions of initial cells in the mature root tip. The centermost cells of OL1 undergo periclinal divisions. These cells will give rise to the future cap as is demonstrated by means of the cell marker LRC244, which is also expressed in the cap cell of the seminal root. The rest of the OL1 cells do not undergo periclinal divisions and will produce the future epidermis. The cell marker EpiGL2 expressed in these OL1 cells and in the epidermis of the seminal root the epidermal cells that will not produce hairs. The centermost cells of OL2 will produce the future quiescent center. The rest of the cells of OL2 undergo a periclinal division forming the layers OL2a and OL2b that will form the future cortex and endodermis, respectively. End195 marks the OL2b in the LR primordia and endodermis in the seminal roots. End195 also marks the cortex/endodermal initial cell which undergoes a periclinal division to generate cortex and endodermal cell files. The marker corAx92 identifies the OL2a cells in the LR primordia and the cortical cells in the seminal roots. IL1 will form the future pericycle. The centermost cells of IL2 undergo a noticeable radial expansion and periclinal divisions. They will give rise to the rest of the vascular cylinder. The end of development stage VI is reminiscent of the primary root tip. Other studies have also tried to identify the future mature root tissues in the lateral root primordia before emergence based on the cell location. In LR primordia of I. purpurea, Seago (1973) observed that the apical cells of the third layer of pericycle derivatives, counted outward, do not divide but the remaining third layer cells undergo periclinal divisions giving rise to the ground meristem. These results are interesting to compare with those of Malamy and Benfey (1997) because this third layer in I. purpurea would be equivalent to the OL2 layer in A. thaliana. Therefore, the endodermis and cortical cells of the lateral roots are formed from an

Lloret and Casero

equivalent layer in the two plants. Periclinal divisions were also observed in the flanks of the outermost pericyclic-derived layer of Typha glauca to produce the future ground meristem and protodermis (Seago and Marsh, 1990). The remaining pericycle derivative cells of this layer do not divide periclinally. In this species, pericycle derivatives are covered by a layer of endodermis derivatives. The endodermis derivative cells at the tip quickly undergo periclinal divisions like the outermost pericycle derivative layer in Arabidopsis. However, in T. glauca, cells formed from endodermis derivatives do not produce lateral root tissues. Cap cells will be formed from the ground meristem protodermis just after emergence (Seago and Marsh, 1990). Thus, the mature lateral root tissues of A. thaliana are exclusively formed from the pericycle cell derivatives. This was previously reported by Van Tieghem and Douliot (1888) for Convolvulus siculus and C. tricolor, by Dittmer and Spensley (1947) for Descurainia, by Davidson (1965) for Vicia faba, by Riopel (1966) for Musa acuminata, by Byrne (1973) for Malva sylvestris, by Seago (1973) for I. purpurea, by Charlton (1975) for Pontederia cordata, by McCully (1975) for Daucus, and by Seago and Marsh (1990) for Typha glauca. However, in many other plants, Zea mays for example (Bell and McCully, 1970; Karas and McCully, 1973), pericycle, endodermis, cortex, and stelar parenchyma divide during the lateral root initiation. All these cells undergo changes in the structure of cytoplasm and cell walls which are reviewed by McCully (1975). The endodermis outside the lateral root primordia of A. cepa, also undergo asymmetric transverse divisions like the pericycle. A similar situation occurs in Helianthus annuus (Fig. 1a of Charlton, 1996). Transverse proliferation would explain how endodermis cells incorporate 3H-thymidine prior to radial expansion and anticlinal divisions in Malva sylvestris (Byrne, 1973). Many authors have found that the endodermis shows similar morphological features to the pericycle during lateral root initiation: increasing cytoplasmic content (Bonnett and Torrey, 1966; Seago, 1973), increasing cytoplasmic basophilia (Bell and McCully, 1970), and enlargement of the nucleoli (Seago, 1973). The cell walls show major changes (Karas and McCully, 1973). Endodermis cells outside the lateral primordia undergo tangential expansions and anticlinal divisions to accommodate the growth of the pericycle derivatives (McCully, 1975; Clowes,

Lateral Root Initiation

1978). The new cells do not form Casparian strip (Peterson and Peterson, 1986). The original Casparian strip regions do not change and remnants of them can occasionally be seen on some of the anticlinal walls of the cells derived from the endodermis which cover the young primordium (McCully, 1975). The endodermis also undergoes periclinal divisions giving rise to several radially arranged layers (Popham, 1955; McCully, 1975; Charlton, 1996). Early authors (Tschermak-Woess and Dolezal, 1953; Gramberg, 1971) described proliferation of the cortical cells outside the lateral primordia. More recently, Casero et al. (1996) observed that the cortex outside the lateral root primordium of A. cepa undergoes asymmetric transverse divisions like the pericycle. In Glycine max, the cortical cells divide periclinally (Byrne et al., 1977). Several layers of cortical cells dedifferentiate becoming incorporated into the primordium in Cucurbita pepo (Charlton, 1996). Endodermis and cortex cells enter successively into proliferation after those of the pericycle (Casero et al., 1996). This allows us to extend the hypothesis of symplastic communication to the endodermis and the cortex. McCully (1975) suggested that a substance diffusing from the young primordia can influence cortical cells. Cells other than the pericycle also proliferate during LR initiation and can also contribute to the formation of the lateral root. The contribution of the parent root tissues to lateral root formation depends on the plant species. In maize, for example, the endodermis of the parent root gives rise to the epidermis of lateral roots. This is based on the fact that the epidermis of the parent root and endodermis derivatives of the lateral primordia show a similar morphology and secrete a thick coat of mucilage to the outer surface of their outer tangential wall. Moreover, a few endodermal derivatives at its tip divide periclinally to produce the root cap initial cells (cf. McCully, 1975). The contribution of the parent endodermis of pea roots is more extensive than in maize, producing the cortex, the epidermis and the cap of the lateral root (Popham, 1955). In Cucurbita maxima, the endodermis forms the cortex, the epidermis, the cap and the vascular tissue of the lateral root (Mallory et al., 1970). In maize (Bell and McCully, 1970; Ashford and McCully, 1973) and in Ipomoea purpurea (Seago, 1973), the parenchyma cells of the parent root contribute to the basal tissue of the lateral primordia, playing an important role in the vascular connection. Cell proliferation does not necessarily mean that the cell derivatives will form part of the future LR. In many examples, endodermis and cortical derivatives

139

form a covering over the young primordium that is finally stretched, broken, and shed (McCully, 1975). The German term Tasche—pocket—was used by Von Guttenberg (1968) to refer to the structures derived from the endodermis. In some cases the tasche persists and forms a substitute root cap; in others the epidermis of the lateral root is derived from the inner layer of the endodermal derivatives and persists, while the remainder of the tasche is shed (Schade and Von Guttenberg, 1951). The review by Charlton (1996) also refers to the poche digestive of Van Tieghem and Douliot (1888) because it is also a structure formed of endodermis derivatives although it can also contain overlying cortical derivatives. The tasche and the poche digestive could have similar roles secreting enzymes to facilitate the passage of lateral primordia through the parent root tissues. Coated vesicles that fused to the plasmalemma of the outermost cells of the tasche of Convolvulus arvensis were observed, suggesting that they contained hydrolytic enzymes (Bonnett, 1969). In Z. mays, the outermost derivatives of the parent root endodermis show high levels of acid phosphatase and -glycosidase activity (Ashford and McCully, 1973). Peretto et al. (1992) and Bonfante and Peretto (1993) found polygalacturonases in the outer part of the primordium and tasche and in the pectic material between the LR and parent cortical tissues of Allium porrum. Enzymatic hydrolysis facilitates LR emergence. Pectinolytic enzymes are secreted by the lateral primordium to separate the parent cells and collapse them. However, Bell and McCully (1970) reported that the walls of hypodermis and epidermis of Z. mays are lignified and suberized, and are not easily penetrated. Some authors claim that a simple mechanical process is enough to explain the passage of the lateral primordium through the parent root tissues. Sutcliffe and Sexton (1968) suggested that the high levels of glycerophosphatase activity detected in the cortical cells adjacent to lateral root primordia during emergence are due to breakage of these parent root cortical cells induced by mechanical pressure. Peterson and Peterson (1986), Charlton (1991), Lin and Raghavan (1991), and Charlton (1996) observed that the cortical cells adjacent to primordia collapse as the primordium emerges. Charlton (1996) showed that in Cicer arietinum cortical cells outside the primordia separate. Therefore, it is reasonable to assume that both physical and enzymatic processes are involved in lateral root emergence (Bonnett, 1969) and that they can occur either simultaneously or separate in time.

140

During early developmental stages the LR primordia are made up of similar proliferative cells, and it is sometimes difficult to distinguish the derivative cells from the parent pericycle, endodermis, and cortex. However, populations of cells can be distinguished in small primordia by their different mean cycle time (McCully, 1975). Early growth of primordia was mostly due to proliferation of cells occupying a central core (Friedberg and Davidson, 1971), while later, growth occurred preferentialy through proliferation of peripheral cells. Cells of Vicia faba undergo major changes 24 h before emergence, which depend on their location. The most proliferative central cells become almost completely quiescent while the peripheral cells increase their rate. Mitotic activity then resumes in the central cells and falls off in the peripheral cells (McCully, 1975). Laskowski et al. (1995) suggested that the formation of a meristem in A. thaliana is a two-step process involving the establishment of a population of rapidly dividing cells followed by the process of meristem organization, during which initial cells are established. Although in A. thaliana LR primordia are sufficiently organized before emergence to proceed with development of LRs even if isolated from the parent root (Laskowski et al., 1995; Malamy and Benfey, 1997), an active apical meristem must be established later (Cheng et al., 1995). These authors show that rlm mutants are unable to activate a root meristem. Therefore, growth of the LR of these mutants is normal until emergence, but is then arrested. Emergence appears to be largely due to enlargement of the basal cells rather than to cell division. Before emergence the basal cells elongate, being less proliferative than the more apical and superficial cells (Malamy and Benfey, 1997). The central basal cells contain small vacuoles in contrast with the highly vacuolated cells which surround them. The quiescent center is normally detectable in the apical meristem of emerged lateral roots (Clowes, 1958). The vascular connection between LR and the parent root is then formed. Phloem connections are formed from the pericycle and vascular parenchyma derivatives while the xylem connection is established through the parenchyma derivatives neighboring the parent xylem (McCully, 1975; Peterson and Peterson, 1986; Luxova´, 1990; Vidal-Bernabe´ et al., 1998). It is generally accepted that maturation of the vascular cell connectors into vessels proceeds acropetally into the lateral roots, and occurs either at the time of lateral root emergence or later. The vascular tissues produced by the apical meristem of the lateral roots mature basipetally towards the parent roots (Byrne et al., 1982).

Lloret and Casero

Oparka et al. (1995) have demonstrated the necessary symplastic connection between the primary root and the lateral ones in roots of A. thaliana. This was done by means of carboxyfluorescein and confocal laser scanning microscopy. Oparka et al. (1995) observed that the establishment of intercellular transport between the parent root phloem and the lateral root primordia was enabled by structural differentiation of adjacent phloem connector elements. This occurred very early, when the lateral root primordia were visible as small bulges protruding from the cortex. Phloem connectors were detectable so early because they showed intense point fluorescence on their walls. The sieve elements were structurally recognizable immediately after the emergence of the LR.

III.

TYPES OF LATERAL ROOTS

LRs are structures that originated by cell division in the parent root pericycle. When they emerge they have an apical meristem similar to that of the root axis that originated it (Torrey, 1986). This definition is perhaps too broad as the LR population in a specific axis is not always as homogeneous as expected. The first source of variation comes from the sequence of development that is not always perfectly acropetal. Frequently, poorly developed abortive or dormant LRs are interspersed between more-developed LRs, apparently breaking the acropetal sequence. They may be acropetal LRs whose development has stopped. Nevertheless, true new LR primordia can develop between older formed LRs (Charlton, 1991). Hence, in a first approach, the LRs can be classified into acropetal laterals and adventive late-forming laterals. The basal roots, often considered to be LRs, are a special type of roots. Genetically and physiologically it appears that they may really be different from LRs (Zobel, 1975). Basal roots were initially demonstrated in tomato and have later also been observed in other species (Stoffella et al., 1992; Bonser et al., 1996). In tomato, they are perfectly discriminated in homozygotic double tomato mutants dgt ro that in theory should not form LRs (dgt mutation) or adventitious roots (ro mutation), but do form roots probably from the hypocotyl (Zobel, 1975). Hence in strict terms, they cannot be regarded as LRs because of their shoot origin. Cluster rootlets (Peterson, 1992; Skene et al., 1996), root primordial masses and fasciated roots (Hinchee and Rost, 1992a), and hairy roots (Biondi et al., 1997) have a structure and origin similar to those of normal LRs. However, they still differ by being much

Lateral Root Initiation

closer to each other than normal LRs. Probably, all these types of LR result from response of the parent root to altered physiological conditions and overcoming possible inhibitions between adjacent LR primordia. The type of alteration hairy roots undergo is believed to be the result of auxin overproduction promoted by the insertion into their DNA of bacterial aux and/or rolB gene sequences plus possible auxin hypersensitivity (Biondi et al., 1997). Ectomycorrhizal roots are present in many gymnosperms and angiosperms (see Chapters 28 by Bacon et al. and 50 by Kottke in this volume). A detailed description of this association between LRs and symbiotic fungi in Pinus resinosa was given (Wilcox, 1968b). This gymnosperm shows a clear LR dimorphism with long and short laterals (Wilcox, 1968a). Long LRs have a greater diameter and faster growth than short LRs. It is these short roots that establish the symbiosis (Peterson and Peterson, 1986). As the name indicates, in an ectomycorrhizal association a mantle of hyphae encloses the root without penetrating into its cells. Nevertheless, the presence of such a mantle significantly modifies the structure and function of the affected LR (Peterson, 1992). The main structural differences between them and regular LRs are rounding of the root tip, reduced growth, meristem vacuolisation, inhibition of root hair formation, and in several cases also dichotomous branching (Peterson, 1992; Peterson and Peterson, 1986). Other particular LR types that were described in the literature are the narrow and wide LRs of maize, which differ not only in diameter (Fig. 2) but also in growth (MacLeod, 1990). Are all these patterns of development really very different? Probably such root types develop in response to particular physiological and/or environmental conditions, but at the cellular level they do not represent substantial deviation from the basic aspects of LR development. So many LR types rather support the concept of plasticity of development, which allows the LR, for example, to stay dormant when environmental conditions are adverse and resume growth when they become to favorable (Dubrovsky et al., 1998). LRs could also transform into clustered rootlets when grown under low P concentration (Johnson et al., 1996). The most impressive example of plasticity is the transformation of endogenous root primordia into endogenous buds. Such a phenomenon occurs in a small group of species (Bonnett and Torrey, 1966). The initial endogenous primordia originate from the pericycle at a specific location relative to the parent root vascular system which would usually give rise to

141

LRs. Excised Convulvulus roots formed either lateral roots or endogenous buds. The initiation of an endogenous bud had a higher probability on the proximal part of the root. At early stages of development, root primordia were identical to endogenous buds. Later, the pathways of development differed between these two structures, mainly in their contribution of parent root tissues to the body of the primordia, in the growth rate, and in the orientation of cell divisions (Bonnett and Torrey, 1966). Despite the apparent initial similarity between root and shoot primordia, conclusive evidence of an undetermined primordium, or of an undetermined primordium site, was not obtained in this study. Hence, this situation deserves further analysis now that we have more powerful research tools. IV.

LATERAL ROOT INITIATION AND REGULATION OF THEIR DEVELOPMENT

A.

Lateral Root Development Appears to Be a Multiphasic Process

It has been estimated that a red oak tree has
500 million living root tips, most of which are LRs (Lyford, 1975). This impressive number of LRs shows how important it is for the plant to explore the soil and to acquire enough water and nutrients to stay alive. Evidently, given the importance of having a root system extending over large soil volume, the development of LRs must be under fine regulatory controls. The formation of LRs has a high adaptive value and is therefore under a strong selective pressure. The formation of LRs is a multiphase process consisting of two or three main stages under different development controls. Some authors have considered two stages: initiation and meristem formation (MacIsaac et al., 1989; Laskowski et al., 1995). The moment that marks the transition from initiation to the following stages of growth has been placed early, when the LR primordium is considered autonomous— i.e., grows in vitro without supplementary growth regulators (Laskowski et al., 1995). It could also happen later, when the primordia has emerged as a LR (MacIsaac et al., 1989). Others have proposed three stages: initiation, development of primordia into roots, and emergence coupled with subsequent growth (Van Staden and Ntingane, 1996; Baum et al., 1998; Zhang and Hasenstein, 1999). As the process of LR development can stop at initiation (Celenza et al., 1995) shortly before emergence (Mallory et al., 1970) or after it (Baum et al., 1998), perhaps this view reflects

142

Lloret and Casero

most accurately the fine physiological controls that determine the overall processes involved in the formation of LRs. The initiation phase is represented by the active proliferation of a few FPCs. During this stage reactivated pericycle cells continue to their proliferative activities and organize an anlage formed by a few tiers of cells. Frequently the pericycle-derived cells are partially surrounded by an endodermal covering originating from the parent root endodermis (Esau, 1940). During LR initiation pericycle cells proliferate actively and duplication times as short as 3–4 h have been described (MacLeod and Thompson, 1979). The second stage of development occurs later, when the LR primordium acquires sufficient physiological autonomy to grow in a culture medium (Laskowski et al., 1995). There are different populations of cells according to the length of the cell cycle (Friedberg and Davidson, 1971). As a culmination of the processes the LR primordium emerges in this phase. This occurs rather by the rapid elongation of cells at the base of the LR primordium than by cell divisions in the apical meristem (Bell and McCully, 1970). The growth of the emerged LR is considered the third stage of development. At this moment the LR apical meristem is very similar to the apical meristem of the parent root (Dolan et al., 1993). In most species the LR now acquires a quiescent center (MacLeod and McLachlan, 1974; McCully, 1975) and reaches its vascular connection with the parent root (MacLeod and Francis, 1976; Byrne et al., 1977). For sustained growth, the young LR requires auxin supply (Celenza et al., 1995). It has a root cap derived from the pericycle, which replaces the ephemeral endodermal covering (Clowes, 1978; Seago, 1973), and as a consequence it gradually develops a gravitropic response (Moore and Pasieniuk, 1984; see also Chapter 30 by Pilet and Chapter 3 by Sievers et al., in this volume). B.

Endogenous and Exogenous Factors Regulate Lateral Root Formation

As a result of their evolution, the higher plants show the capacity to regulate the development of LRs to fulfill efficiently all the functions assigned to root systems. The LRs have a decisive influence on root system capacity to anchor the plant to the soil and to acquire water and nutrients. Consequently, practically all vascular plants can develop LRs. There are a few exceptions of species with primary roots that cannot develop laterals. Noteworthy cases are the aquatic fern Azolla pinnata (Barlow, 1984) and the angiosperm Lemna

(Clowes, 1985). Several mutants with a more or less impeded capacity of LR formation, such as axr1, aux1, axr4, Dwf, alf4, tir3, and dgt, were described in Arabidopsis thaliana, Zea mays, and Lycopersicon esculentum, species that normally develop LRs (Mirza et al., 1984; Schiefelbein and Benfey, 1991; Hobbie, 1998). Some aerial roots do not develop LRs until they reach the soil (Zimmerman and Hitchcock, 1935). Finally, there are lettuce cultivars that show delayed LR formation (Zhang and Hasenstein, 1999). All these examples suggest that the capacity of LR formation can be modulated by exogenous and endogenous factors (see also Chapter 14 Feix et al., in this volume). Changes in the physical or chemical environment that modify the physiology of roots can strongly influence on LR development. Such changes probably interact with metabolic processes in the aerial parts of the plant or in the root itself (see also Chapter 9 by Waisel and Eshel in this volume), which in turn are necessary for root system development. By means of these mechanisms, physical factors such as temperature (Sattelmacher et al., 1990; Gladish and Rost, 1993; McMichael and Burke, 1998), mechanical impedance (Goss, 1977; Feldman, 1984; Singh and Sainju, 1998), or light (Furuya and Torrey, 1964; Reinhardt and Rost, 1995) can modulate LR formation. Heavy metals, NaCl, and other chemicals in the root environment may be toxic for LR development (Malone et al., 1978; Waisel and Breckle, 1987; Reinhardt and Rost, 1995; Baligar et al., 1998). A deficit of certain nutrients also has a strong effect on LR formation (Baligar et al., 1998). Most of these factors probably regulate LR development through systemic influences on the metabolism at the whole-plant level. It is well known that nitrates favor branching of root systems (Hackett, 1972; Drew, 1975; Granato and Raper, 1989; Robinson, 1994). Similar effects have been reported for other inorganic nutrients such as ammonia and phosphates (Drew, 1975). As phosphates promote first-order LR elongation rather than branching, they can bring about the same results as nitrate, having different developmental effects (Robinson, 1994). Frequently, LRs that find favorable soil patches grow and branch profusely, whereas other roots, initiated in less favorable sites, suffer reduced growth. Such a local response to inorganic nutrients operates by different mechanisms to stimulate branching by enhancing growth at the whole-plant level. It has been proposed that tissues that receive adequate nitrogen supply are preferred sinks for photosynthates compared to tissues receiving limited nitrogen

Lateral Root Initiation

(Sattelmacher and Thoms, 1991). Another possibility is that nutrients, particularly nitrates, regulate hormone metabolism or transport in localized regions or even act themselves as developmental signals (Cao et al., 1993). Because recent research has offered evidence for this type of control mechanism, the effects of nitrate on LR initiation are of special interest (Zhang et al., 1999). Those authors have demonstrated that in A. thaliana roots nitrate ions control the formation of LRs by means of a dual mechanism; primarily they constitute a signal which promotes LR development from pericycle cells. This effect is localized in the root regions directly affected by nitrate, and seems to affect the development of primordia into LRs rather than their initiation. At the same time, the assimilated nitrates affect LR development by controls traceable to the whole-plant level. Curiously, this systemic effect is inhibitory rather than stimulatory (Zhang et al., 1999; Zhang and Forde, 2000). Experiments with auxinresistant mutants of A. thaliana have provided evidence for an overlap between the localized stimulatory effect of nitrate on LR elongation and the mechanism of response to auxin (Zhang et al., 1999; Zhang and Forde, 2000). Increase of branching frequency in response to nutrient availability may occur not only on a localized scale of a single-root axis, but also on a general scale of whole-root system (Granato and Raper, 1989; Gersani and Sachs, 1992; Bingham et al., 1997). A split-root hydroponics system was designed to test the effects of a heterogeneous supply of nitrate to only one of two axes of maize (Granato and Raper, 1989). Total allocation of dry matter between shoot and root was not altered in this experiment, but partitioning between the two axes was strongly affected. The þN axis receiving increased dry matter and producing more LRs comparatively to N axis. Pea plants grown in hydroponic culture with one half of their root system bathed in nutrient-rich solution and the other half in a less rich solution, also showed that LR development was favored in the former and inhibited in the latter (Gersani and Sachs, 1992). Hence, a compensatory growth seems to control branching in deprived versus well-fed roots. Endogenous factors clearly condition LR initiation. Excised roots growing in culture medium require a wide range of substances for LR formation. For example, pea roots required, in addition to auxin and sucrose, thiamin, nicotinic acid, adenine, and one or more micronutrients (Torrey, 1956). For plants growing in more natural conditions, LR formation is an energy-consuming process that can only occur at the

143

expense of available metabolic substrates, particularly of assimilates originating in the shoot (Philipson, 1988). Recent investigations have confirmed the importance of sugars in LR formation. Feeding halves of wheat root systems with glucose resulted in an increase of branching in the fed half (Bingham et al., 1997). Such results have suggested that one of the main limiting endogenous factors regulating LR formation is sugars. As they sustain root growth, sugars are a necessary component of growth media when excised roots are grown in vitro. Sugars also accumulate in cavities just outside the LR primordia and contribute to their growth (Friedberg and Davidson, 1971; MacLeod and Francis, 1976). As with inorganic nutrients, it is difficult to know whether sugars act by increasing metabolism, as true signals (cf. Bingham et al., 1998), or by both mechanisms. Assimilate partitioning is mainly controlled by the sink strength of the different organs. The main root apex and the apex of developing LRs clearly compete for assimilates. Hence, when the root apex is removed the assimilates are redirected to LRs. The mode of dry matter partitioning between the LR in Pinus pinea depends on the time of decapitation. Removing the root tip of young seedlings without emerged laterals results in a homogeneous redistribution of assimilates along the main root axis. When an older taproot is decapitated after it has begun to form LRs, then the assimilates accumulate in basal regions of the root system where the laterals are developing. These results strongly suggest that there is competition for assimilates between the taproot apex and developing laterals. They are also consistent with the hypothesis that assimilates are more necessary for the elongation of LRs than for their initiation (Atzmon et al., 1994). The plant hormone auxin plays various key regulatory roles during LR initiation, organization of the apical meristem of the LR primordia, and finally the emergence and subsequent growth of the LR. In most plant species, exogenous auxins promote the initiation of LRs (Torrey, 1962; Blakely et al., 1972, 1982; Webster and Radin, 1972; Wightman et al., 1980; Zeadan and MacLeod, 1984; Hinchee and Rost, 1986; Hurren et al., 1988; MacIsaac et al., 1989; Lloret and Pulgarı´ n, 1992; Vuylsteker et al., 1997, 1998; Baum et al., 1998; Zhang and Hasenstein, 1999; see also Chapter 23 Gaspar et al., in this volume). The typical response to exogenous auxin treatment is the formation of new LR primordia out of the acropetal sequence. These out-of-sequence primordia have been called adventive LRs to distinguish them from adventitious or shootborne roots (Barlow,

144

1986). Concerning the acropetal series of LRs, some species appear to be insensitive to exogenous auxins (Charlton, 1983a, 1991). Furthermore, it has not been possible to correlate the LR frequency and endogenous auxin contents of maize excised root fragments (Golaz and Pilet, 1987). Hence, the evidence that links auxins to the formation of LRs has been considered to be circumstantial (Charlton, 1996). Nevertheless, using transgenic plants and developmental mutants we now have more clues that link auxins and LR initiation (Celenza et al., 1995; Larkin et al., 1996). Studies of LR development based on classical experiments involving exogenous auxin treatments certainly have limited value. Indeed, most studies performed with exogenous auxin treatments have suffered from three main handicaps: (1) they have frequently not distinguished between acropetal and adventive LRs; (2) they have not demonstrated the final auxin concentration that affect the pericycle cells; and (3) no distinction between the effects of exogenously applied versus endogenous auxin was made. Nevertheless, the evidence obtained suggests that auxins are involved in the regulation of LR formation. Several explanations were given to those particular cases where auxins apparently did not promote LR. The apparent insensitivity to exogenous auxin in some species could result from the permeation barriers impeding the exogenous auxin from reaching the target tissue—i.e., the pericycle (Blakely et al., 1986). An alternative explanation for the negative results in exogenous auxin treatments takes into account the limited time in which the root pericycle can initiate LRs (Abadı´ a-Fenoll et al., 1986; Lloret and Pulgarı´ n, 1992). Hence, the current view is that endogenous auxins are the main regulators of LR formation (Hobbie, 1998). Moreover, not only the auxin concentration but also the sensitivity of the pericycle and the efficiency of the auxin transport is relevant for the development of LRs. Perhaps the most direct evidence that links auxin to LR development comes from the demonstration that plants that overproduce auxins have also increased numbers of LRs. The sur1 mutants of A. thaliana have high auxin contents in root tissues and are considered to be auxin overproducers (Boerjan et al., 1995). The sur1 mutants showed increased LR formation (Boerjan et al., 1995). Transgenic plants that overexpress bacterial iia genes and consequently have high auxin levels, also tend to form many LRs (Klee et al., 1987). On the other hand, the dgt mutant of Lycopersicon esculentum, which shows reduced auxin sensitivity, has also impeded capacity of LR develop-

Lloret and Casero

ment (Zobel, 1973; Kelly and Bradford, 1986; Muday et al., 1995). Similarly, in A. thaliana, the combination of two auxin resistance mutations (axr4 and axr1) results also in few LRs compared to the parental lines or single mutants (Hobbie and Estelle, 1995). These mutants have multiple phenotype alterations compared to wild plants other than their modified capacity for LR formation. Nevertheless, there are three mutants which show specific alterations in LR formation in A. thaliana (Celenza et al., 1995). These have been named alf mutants (alf1, alf4 and alf3) and show defects at different stages of LR development. The first, alf1, is probably defective in auxin catabolism in the root, leading to high auxin concentrations in root tissues, which in turn promotes increased LR formation. This mutant is thought to be the same as sur1, hls3, and rty (Scheres et al., 1996). The second, alf4, seems to have a defect in primary perception or response to auxin as its pericycle cells are unable to form LRs. The third, alf3, initiates LR primordia, but they abort later. The mutant alf3 is rescued by exogenous auxin treatments, but alf4 is not. All three mutants provide strong support for the hypothesis that endogenous auxin is the major regulator of LR initiation and subsequent growth of the primordia (Celenza et al., 1995). Auxin transport in higher plants is a directional and regulated process. It has been known for a long time that auxin applied to the shoot eventually reaches the root (Morris et al., 1969; McDavid et al., 1972). Once in the root, auxin is transported acropetally in the central cylinder (Mitchel and Davies, 1975; Tsurumi and Ohwaki, 1978; Kerk and Feldman, 1995). It can accumulate in developing LR primordia (Rowntree and Morris, 1979) or at the root apex (Kerk and Feldman, 1994), where it induces a zone with organized pattern and polarity properties (Sabatini et al., 1999). When the auxin reaches the apex the movement is reversed, and auxin moves basipetally between the root tip and the elongation zone probably through the root epidermis and/or the outer cortex (Tsurumi and Ohwaki, 1978; Yang et al., 1990; Balus˘ ka et al., 1994; Estelle, 1998). The acropetal transport of auxins in the root is dependent on metabolic energy (Torrey, 1976). Given its pathway inside the root, the acropetal flux of auxin might influence the formation of LR primordia in the pericycle. There are three lines of evidence which relate auxin transport to LR initiation. In Pisum sativum seedlings, exogenous auxin applied to the plant to substitute the excised cotyledons mimics the role of these organs in the regulation of LR initiation and emergence

Lateral Root Initiation

(Hinchee and Rost, 1986). The second line of evidence for this relationship comes from the study of the aux1 and tir3 mutants of A. thaliana that show reduced numbers of LRs owing to a probable defect in the auxin transport mechanism (Ruegger et al., 1997; Bennett et al., 1996, 1998). Nevertheless, these mutants did not show dramatically altered LR formation. This fact probably occurs because the genetic redundancy in the multigene family encoding auxin transport proteins attenuate the effect of a single mutant locus. Treatments with auxin transport inhibitors result in much stronger effects. These compounds, e.g., naphtylphtalamic acid or semicarbazone derivative 1, are effective in inhibiting the formation of LRs in tomato (Muday and Haworth, 1994) and A. thaliana roots (Reed et al., 1998). We do not know whether primordia are initiated under such treatments in these species but either they subsequently do not emerge or their initiation has also been inhibited. In Pisum sativum roots, it appears that under auxin inhibitor treatment, structures similar to laterals called ‘‘root primordial masses’’ are initiated but fail to grow out because they are not capable of organizing an apical meristem. Such structures are distributed in regions along the parent root axis which normally produce laterals. When treatment is suppressed or exogenous auxin is supplied, they can grow as clusters of LRs (Hinchee and Rost, 1992a). Some of the alterations promoted by auxin transport inhibitors seem to be similar to the case of alf3 mutants (Celenza et al., 1995). The results in P. sativum roots are also consistent with the hypothesis that auxin transport inhibitors cause auxin redistribution and apparent morphogenetic effects in the root (Sabatini et al., 1999). Nitrogen-fixing nodules of leguminous plants are structures a related to LR development. Characterization of the genetic and physiological control of nodule formation has demonstrated that mutants with altered nodulation also show altered LR formation (Kneen et al., 1994). All the available evidence points to an increased auxin sensitivity as a cause for the relationship between nodule and LR formation. Furthermore, in the model species for determinate nodule formation, Lotus japonica, it was shown that the induction of nodule organogenesis and the initiation of LRs share common regulatory elements (de Bruijn et al., 1998). Auxin is not the only growth regulator that modulates LR formation. Exogenous kinetin at low concentrations stimulates auxin-induced LR production in excised pea roots, but at higher concentrations (higher than 4:6  105 ) it is a powerful inhibitor (Torrey,

145

1962). The cytotokinin concentrations that inhibited LR formation have been considered as physiological, or at least nontoxical (Torrey, 1962). On the other hand, the spontaneous formation of LRs also seems to be inhibited by exogenous cytokinin in lettuce roots (MacIsaac et al., 1989). Furthermore, the studies of Wightman et al. (1980) and Van Staden and Ntingane (1996) have demonstrated that exogenous cytokinins inhibited not only initiation but also the emergence of LRs. As cytokinins are if not produced at least accumulated at the primary root apex and the apices of developing LRs (Feldman, 1975; Van Staden and Davey, 1979; Forsyth and Van Staden, 1981; but see Chapter 25 by Emery and Atkins in this volume), and the decapitation of primary roots transitorily stimulates LR formation (Wightman and Thimann, 1980; Forsyth and Van Staden, 1981; Lloret et al., 1988), it has been propossed that these hormones are the inhibitors that explain why LRs are not initiated near the root apex (Van Staden and Ntigane, 1996). The concept of apical dominance applied to roots has recently been revisited by Zhang and Hasenstein (1999) in lettuce roots (Lactuca sativa, cv. Baijianye). The seedlings of this cultivar do not normally produce LRs, but removal of the primary root tip triggers a rapid response and LR primordia are initiated in
9 h. As LR primordia are newly initiated in response to root decapitation and growth regulators that usually promoted the growth of shoot buds inhibited LR primordia development, it appears that root apical dominance is not a quite similar phenomenon to its counterpart in shoot development, although the same growth regulators, auxin and cytokinin, are involved (Cline, 1997). In fact, auxin and cytokinin concentrations along the parent root are thought to directly control LR initiation and emergence (Hinche and Rost, 1986). Webster and Radin (1972) reported interaction between auxin and cytokinins in excised pea roots growing in culture medium. In such roots the pericycle can initiate LRs or contribute to the formation of a cambium. The behavior of the pericycle depends on the ratio of exogenous auxin to cytokinins. Auxins alone promote LR initiation. Cytokinins alone bring about formation of a multiseriate pericycle. The combination of both growth regulators organizes a functional cambium. The control of LR formation would be better understood if it were analyzed in terms of interactions between hormones rather than by considering the effects of an isolated growth regulator. The exact nature of such interactions is for the moment unknown, and this is an aspect that deserves further research.

146

Several Arabidopsis mutants with altered response to cytokinins have been isolated. These pas mutants show abnormal root system development. The mutants pas1 and pas3 develop short primary roots with no or with very few LRs, whereas pas2 shows long primary roots and increased numbers of LRs. Physiological and biochemical analyses show that cytokinins are involved in pas phenotypes. The pas mutants are not cytokinin overproducers but are probably altered in their sensitivity to cytokinin (Faure et al., 1998). If cytokinins are supposed to inhibit LR initiation, one important question remains unresolved: What happens in plant species that initiate their LR primordia close to the root apex. This occurs, for example, in Ceratopteris thalictroides and Cucurbita maxima (Mallory et al. 1970) or Pontederia cordata, Pistia stratiotes, and Musa acuminata (Charlton, 1987). We still have no clear explanation for this. One possibility is that LR formation is a constitutive capacity of the root apex and proceeds in these species even under high cytokinin concentrations at the site of initiation. Another is that the sensitivity to or the actual concentration of cytokinins differs from one species to another. Other phytohormones have a less clear influence and often give rise to inconsistent or even contradictory reports about their effects on LR development. Perhaps the reason is that they regulate this process no directly but through interaction with auxins and cytokinins. One phytohormone that is produced at the root apex and that inhibits LR initiation and emergence when applied exogenously is abcisic acid (ABA) (Hooker and Thorpe, 1998). It is interesting to note that the treatment of excised tomato roots cultured in vitro with fluoridone, an inhibitor of ABA biosynthesis, stimulates the initiation of LRs but not their emergence and subsequent growth (Hooker and Thorpe, 1998). The overall evidence suggests that gibberellic acid plays at most a minor role in regulating LR initiation (Torrey, 1976). The same is true for the ethylene, although this hormone might affect LR initiation through its well-known effect on auxin lateral transport (Lee et al., 1990). It has recently been reported that silver ions, known inhibitors of ethylene action, promote the elongation of LRs but have no effect on LR initiation in Lactuca sativa roots (Zhang and Hasenstein, 1999). From a cellular point of view, the initiation of a LR primordium at a specific site implies that pericycle and/ or endodermis cells can perceive signals that are responsible for organized LR development. Furthermore, they are able to integrate such signals

Lloret and Casero

with promotive or inhibitory physiological inputs. All these interactions between cells are far from being understood. Nevertheless, the concept is arising that root cells react to stimulus according to their location on a positional framework, which is used as a reference for patterning. One factor that contributes to this framework is polar auxin transport (Sabatini et al., 1999; Doerner, 2000). Hence, auxin is beginning to be viewed not as a classical morphogen as defined for animal organisms, but as somehow important in letting the cells know where they are located. V.

LATERAL ROOT PATTERNING

A.

Relationship of Lateral Root Initiation Sites with the Vascular Pattern

The most evident aspect of LR arrangement is that the LR primordia are initiated at specific positions related to the vascular pattern of the parent root. Hence they form ranks along the parent root. In some instances these ranks run along the protoxylem poles in others along the protophloem poles (McCully, 1975). B.

Arrangement of Laterals Along the Parent Root

When roots grow under controlled environmental conditions, they tend to form LR primordia at a regular rate. Consequently, the total number of LRs is positively correlated with the root length (Lloret et al., 1988), and similar numbers of roots per centimeter are observed along wide LR-bearing regions (Lloret and Pulgarı´ n, 1992). In some instances, the number of LRs formed decreases toward distal regions of intact roots (Hummon, 1962; Granato and Raper, 1989). Near the apex and in some instances also near the root base, there are two zones that lack LRs (Mallory et al., 1970; Lloret and Pulgarı´ n, 1992). The reason for the absence of LRs near the apex has been explained in general to the age of pericycle cells located in this region (Abadı´ a-Fenoll et al., 1986) or to the accumulation of cytokinins (Atzmon and Van Staden, 1994). What causes the basal LR-free region is unknown. Under experimental conditions we can promote or inhibit LRs in specific regions of the parent root. Exogenous auxin treatments result in accumulation of LRs at the distal end of the root (Lloret et al., 1992; Vuylsteker et al., 1998). Tritiated thymidine treatments suppress the initiation of LRs at regions of active formation, whereas they do not affect older

Lateral Root Initiation

root regions where LR formation has finished (Hummon, 1962). This means that not all cells along the root are equally reactive to the treatment, being more sensitive the younger they are. There is a gradient of reaction to experimental treatments, which suggests that the capacity for LR formation is limited in time and space (Abadı´ a-Fenoll et al., 1986).

C.

Pattern of Distribution Within Ranks

Some studies have analyzed the LR ranks individually (Mallory et al., 1970; Charlton, 1975, 1977, 1982, 1987; Barlow and Adam, 1988; Lloret et al. 1988, 1998; Pulgarı´ n et al., 1988; Hinchee and Rost, 1992b; Newson et al., 1993). In general, it is accepted that along each rank the LRs are rather regularly arranged (Charlton, 1991, 1996). Nevertheless, the distance between LRs is highly variable in most species. Perhaps the data that fit best a regular mathematical model are those obtained in Ceratopteris thalictroides. In this fern the root is diarch. As growth retracts from the root apex, the root is twisted and the LR primordia that develop opposite the xylem poles form two helical series. Within each series the LR primordia are regularly spaced (Mallory et al., 1970). Excised tomato roots show an interesting feature. The histogram of interlateral spacing of ranks belonging to individual roots of the same rank seems to be multimodal (Barlow and Adam, 1988). The peaks of frequency occur at multiples of a fundamental distance, or ‘‘quantum.’’ This opens up the possibility that segments or metamers could be at the basis of root development (architecture). When the available data were tested statistically they did not always show a significant trend (Newson et al., 1993). Nevertheless, this view of roots with metamers marked by LR positions should be further tested (see also Chapter 4 by Barlow in this volume). When LR primordia are initiated near the apex the spacing within ranks seems to follow a logarithmic scale. This is a consequence of the nearly exponential elongation of roots behind the apex. Hence, in Pontederia cordata and Pistia stratiotes, Charlton (1975, 1983b) analyzed root spacing in the form of log (p2/p1), where p2 and p1 are the positions of two successive LRs behind the root tip and p2 is the smaller value. This parameter was defined as the rhizotron ratio and has been used for the study of LR patterning in Musa acuminata (Charlton, 1982; Draye et al., 1999). All these species also showed a regular spacing of successive LRs along each rank.

147

Between different ranks, an apparent interaction could arise from a persistent synchrony of LR initiation in different ranks (Fig. 12). The simplest example of such a synchrony is the root of C. thalictroides (Mallory et al., 1970). In this fern, two protoxylembased ranks of LRs initiate LR primordia sequentially, thus forming pairs of LRs. These LR primordia are slightly displaced with respect to each other in the longitudinal plane but away from other pairs. Hence groups of two LRs can be discerned in this species. The regularity in LR spacing seems to be a consequence of the regular segmentation of the root apical cell and consequently could be due to a mechanism of synchrony between ranks similar to that shown in Fig. 12. The LR positioning in another fern with diarch roots has also been related to the pattern of segmentation of the apical cell (Charlton, 1983a). Higher-plant roots, of Musa acuminata, Pistia stratiotes, and Pontederia cordata, have shown apparent interactions between ranks, which finally have been demonstrated to be spurious (Charlton, 1987). It appears that in tomato roots the two main ranks of laterals also form primordia independently (Barlow and Adam, 1988). On the other hand, there were some references to clumping of LRs (Yorke and Sagar, 1970; Lloret et al., 1988), which suggest possible interaction between ranks in P. sativum and A. cepa. The tendency to clump may lead to the formation of LR clumps at the same transversal level of the parent root. The number of groups and the number of members of each group of onion roots widely exceed what one would expect based on random coincidence (Pulgarı´ n et al., 1988; Lloret et al., 1988, 1998). The correlation between ranks is also statistically significant for roots of Potentilla palustris (Charlton, 1987). The possibility of interaction between ranks, is intriguing but difficult to prove. The fact that in some species LR primordia inhibit the formation of new primordia in the longitudinal plane, whereas they tend to be formed grouped at the same transversal level, supports the hypothesis of the control by a growth regulator that is transported in a polar pathway. The pattern of LR distribution may be set by the root apex as part of the general process of cell production and differentiation. Alternatively, the pattern of LRs can be determined later, by stimulating mother cells by hormones or by other mobile morphogens. In reality, the two models are not mutually exclusive. Regular spacing within ranks may arise from differences in the longitudinal distribution of nutrients or

148

Lloret and Casero

Figure 12 The possible theoretical relationships between two ranks forming LR primordia (arrowheads) in a physiological background of stimulating or inhibiting factors. When LR primordia coincide at the same transverse level (left center), it can occur because of a simple coincidence without interaction between LR primordia belonging to both ranks (upper left drawing), because of a local stimulus originated in surrounding tissues (left middle) or because of a stimulating effect originated within the proper primordia which tend to be formed grouped at the same transverse level (left bottom). When the LR primordia do not coincide (right center), then it also can occur randomly (upper right) or by an inhibiting influence originated in initiating LR primordia which would move through the vascular cylinder impeding the formation of LR primordia at the same transverse level (bottom right).

growth regulators within the pericycle and/or endodermis cells (Riopel, 1966, 1969; Pulgarı´ n et al., 1988). Initially, such chemical differences do not have to be large as they can undergo self-amplification by a sort of autocatalytic process (Meinhardt, 1984). This would break the uniformity of pericycle and/or endodermis cells in the longitudinal direction. Once generated, the more or less regularly spaced periodic chemical patterns of primordial initiation would proceed at predictable distances from each other and with apparent mutual repulsion. This is similar to the classical view of the order of leaf arrangement on the shoot (Chapman and Perry, 1987). Another possibility, not necessarily incompatible with the former, is that the pattern of LR distribution would be set by the root apex by the formation of a series of mother cells at regular intervals (Barlow and Adam, 1988). These intervals may be considered on the basis of distances between successive LR primordia or better as intervening cells between them. The FPCs initially commited to develop LR primordia could eventually do it or not, depending on the physiological conditions at the moment of LR initiation (Charlton, 1983a; Barlow and Adam, 1988). Hence, more sites for LRs may be determined than actually develop. In fact, an investigation of the relationships between pericycle cell length and the number of LR primordia in decapitated onion

roots has shown that decapitation stimulated LR formation in distal regions to a greater extent than it inhibited pericycle cell elongation (Lloret et al., 1985). D.

Asymmetry of Lateral Root Development

Not all ranks of a given root necessarily form LRs at the same rate. The LRs of adventitious roots of onion are initiated near two neighboring ranks (Lloret et al., 1988; Pulgarı´ n et al., 1988). This pattern is altered if the roots are treated with auxins (Lloret et al., 1998). Asymmetric distribution of laterals between ranks has also been described for the development of early laterals (probably basal roots) on the pregerminative root of Theobroma cacao L. (Dyanat-Nejad and Neville, 1972). The ranks with the greatest number of LRs of such roots are connected with the central xylem poles of the cotyledons. A similar relation was found for the LR primordia of P. sativum preformed on the embrionary radicle (Hinchee and Rost, 1992b). In this case, the primordia were most abundant on the vascular strand connected with one of the cotyledons Apparently, the capacity for LR formation of each individual rank is related to the liberation of some growth regulator by the underlying vascular strands. Sometimes the asymmetry occurs in more distant locations along the root. For example, in curved root

Lateral Root Initiation

segments, more LR primordia arise on the convex side and none on the concave side (McCully, 1975; Fortin et al., 1989). For the moment we have no clear explanation for the reason of all these asymmetries in LR development.

VI.

CONCLUDING REMARKS

Noticeable results have recently been obtained by using new techniques. However, a lot of very important questions about LR formation yet remain unresolved and deserve particular attention in future research. Now, it is known that in angiosperms the first morphological event associated to LR initiation is the polarized asymmetrical transverse division of a pair of pericycle cells. Nevertheless, other molecular and cellular phenomena should precede the activation of FPCs to initiate LR development. These events are under study and it is certain that new cues about them will emerge in the next few years. The transition from the initial phase of development to the organization of the apical meristem of the LR has been shown to involve changes in the expresion of several cell markers and in the physiology of the cells of LR primordia. The interactions between relevant cells for LR development are partially understood. Concerning the analysis of the factors controlling LR initiation, the situation is no better. Although we are aware that this event is related to the effect of various growth regulators, the mode of action of phytohormones on LR development is to be understood at the cellular level. Moreover, it is not known if LR spacing within ranks is due to the distribution of nutrients and growth regulators within pericycle cells, or if it is traceable to the divisional history at the apex producing FPCs at regular intervals, or if it is due to a combination of these two possibilities. It is also unknown why in some species LRs arise opposite xylem parent root poles, whereas in others they do it opposite phloem poles. Partial answers to some of these problems are available. However, a full understanding is still far away. To optimize resources it is particularly important to concentrate research effort on a few well-known models—i.e., Allium, Arabidopsis, Lycopersicon, Pisum, and Zea. As more genes regulating LR development become available, specific steps of LR development will be targets for genetic modification to dissect accurately the whole process.

149

REFERENCES Abadı´ a-Fenoll F, Casero PJ, Lloret PG, Vidal MR. 1986. Development of lateral primordia in decapitated adventitious roots of Allium cepa. Ann Bot Lond 58:103–107. Aeschbacher RA, Schiefelbein JW, Benfey PN. 1994. The genetic and molecular basis of root development. Annu Rev Plant Phys Plant Mol Biol 45:25–45. Ashford AE, McCully ME. 1973. Histochemical localisation of -glycosidases in roots of Zea mays. III. Glucosidase activity in the meristems of lateral roots. Protoplasma 77:411–425. Atzmon N, Van Staden J. 1994. Metabolism of [8-14C] transzeatin by intact and decapitated tap roots of Pinus pinea. S Afr J Bot 60:27–30. Atzmon N, Reuveni O, Riov J. 1994. Lateral root formation in pine seedlings. II. The role of assimilates. Trees Struct Funct 8:273–277. Baligar VC, Fageria NK, Elrashidi MA. 1998. Toxicity and nutrient constraints on root growth. Hortscience 33:960–965. Balodis VA. 1968. Some aspects of mitosis distribution in the root tip. Tsitologiya 10:1374–1383. Balus˘ ka F, Barlow PW, Kubica S. 1994. Importance of the post-mitotic isodiametric growth (PIG) region for growth and development of roots. Plant Soil 167:31– 41. Barlow PW. 1984. Positional controls in root development. In: Barlow PW, Carr DJ, eds. Positional Controls in Plant Development. Cambridge, UK: Cambridge University Press, pp 281–318. Barlow PW. 1986. Adventitious roots of whole plants: their forms, functions and evolution. In: Jackson MB, ed. New Root Formation in Plants and Cuttings. Dordrecht, Netherlands: Martinus Nijhoff, pp 67–110. Barlow PW, Adam JS. 1988. The position and growth of lateral roots on cultured root axes of tomato, Lycopersicon esculentum (Solanaceae). Plant Syst Evol 158:141–154. Baum SF, Karanastasis L, Rost TL. 1998. Morphogenetic effect of the herbicide cinch on Arabidopsis thaliana root development. J Plant Growth Regul 17:107–114. Bell JK, McCully ME. 1970. A histological study of lateral root initiation and development in Zea mays. Protoplasma 70:179–205. Bennett MJ, Marchant A, Green HG, May ST, Ward SP, Millner PA, Walker AR, Schulz B, Feldmann KA. 1996. Arabidopsis AUX1 gene: A permease-like regulator of root gravitropism. Science 273:948–950. Bennett MJ, Marchant A, May ST, Swarup R. 1998. Going the distance with auxin: unravelling the molecular basis of auxin transport. Phil Trans R Soc B 353:1511–1515.

150 Bingham IJ, Blackwood JM, Stevenson EA. 1997. Site, scale and time-course for adjustments in lateral root initiation in wheat following changes in C and N supply. Ann Bot Lond 80:97–106. Bingham IJ, Blackwood JM, Stevenson EA. 1998. Relationship between tissue sugar content, phloem import and lateral root initiation in wheat. Physiol Plant 103:107–113. Biondi S, Lenzi C, Baraldi R, Bagni N. 1997. Hormonal effects on growth and morphology of normal and hairy roots of Hyoscyamus muticus. J Plant Growth Regul 16:159–167. Blakely LM, Rodaway SJ, Hollen LB, Crocker SG. 1972. Control and kinetics of branch root formation in cultured root segments of Haplopappus ravenii. Plant Physiol 50:35–42. Blakely LM, Durham M, Evans TA, Blakely RM. 1982. Experimental studies on lateral root formation in radish seedling roots. I. General methods, developmental stages, and spontaneous formation of laterals. Bot Gaz 143:341–352. Blakely LM, Blakely RM, Galloway CM. 1986. Effects of dimethyl sulfoxide and pH on indoleacetic acid– induced lateral root formation in the radish seedling root. Plant Physiol 80:790–791. Boerjan W, Cervera MT, Delarue M, Beeckman T, Desitte W, Bellini C, Caboche M, Van Onckelen H, Van Montagu M, Inze D. 1995. superroot, a recessive mutation in Arabidopsis, confers auxin overproduction. Plant Cell 7:1405–1419. Bonfante P, Peretto R. 1993. Cell-wall separation during the outgrowth of lateral roots in Allium porrum L. Acta Bot Neerl 42:187–197. Bonnett HT Jr. 1969. Cortical cell death during lateral root formation. J Cell Biol 40:144–159. Bonnett HT Jr, Torrey JG. 1966. Comparative anatomy of endogenous bud and lateral root formation in Convolvulus arvensis roots cultured in vitro. Am J Bot 53:496–507. Bonser AM, Lynch J, Snapp S. 1996. Effect of phosphorus deficiency on growth angle of basal roots in Phaseolus vulgaris. New Phytol 132:281–288. Byrne JM. 1973. The root apex of Malva sylvestris. III. Lateral root development and the quiescent center. Am J Bot 60:657–662. Byrne JM, Pesacreta TC, Fox JA. 1977. Development and structure of the vascular connection between the primary and secondary root of Glcine max (L.) Merr. Am J Bot 64:946–949. Byrne JM, Byrne, JM, Emmitt DP. 1982. Development and structure of the vascular connection between the primary and lateral root of Lycopersicon esculentum. Am J Bot 69:287–297. Cao YW, Glass ADM, Crawford NM. 1993. Ammonium inhibition of Arabidopsis root growth can be reversed

Lloret and Casero by potassium and by auxin resistance mutations aux1, axr1, and axr2. Plant Physiol 102:983–989. Carr DJ. 1976. Plasmodesmata in growth and development. In: Gunning BES, Robards AW, eds. Intercellular Communication in Plants: Studies on Plasmodesmata. Berlin; Springer-Verlag, pp 246–289. Casero PJ, Garcı´ a-Sa´nchez C, Lloret PG, Navascue´s J. 1989a. Changes in cell length and mitotic index in vascular pattern-related pericycle cell types along the apical meristem and elongation zone of the onion root. Protoplasma 153:85–90. Casero PJ, Garcı´ a-Sa´nchez C, Lloret PG, Navascue´s J. 1989b. Morphological features of pericycle cells in relation to their topographical location in onion adventitious roots. New Phytol 112:527–532. Casero PJ, Casimiro I, Rodrı´ guez-Gallardo L, Martı´ nPartido G, Lloret PG. 1993. Lateral root initiation by asymmetrical transverse divisions of pericycle cells in adventitious roots of Allium cepa. Protoplasma 176:138–144. Casero PJ, Casimiro I, Lloret PG. 1995. Lateral root initiation by asymmetrical transverse divisions of pericycle cells in four plant species: Raphanus sativus, Helianthus annuus, Zea mays, and Daucus carota. Protoplasma 188:49–58. Casero PJ, Casimiro I, Lloret PG. 1996. Pericycle proliferation pattern during the lateral root initiation in adventitious roots of Allium cepa. Protoplasma 191:136–147. Casero PJ, Casimiro I, Knox JP. 1998. Occurrence of cell surface arabinogalactan-protein and extensin epitopes in relation to pericycle and vascular tissue development in the root apex of four species. Planta 204:252– 259. Celenza JL, Grisafi PL, Fink GR. 1995. A pathway for lateral root formation in Arabidopsis thaliana. Gene Dev 9:2131–2142. Chapman JM, Perry R. 1987. A diffusion model of phyllotaxis. Ann Bot Lond 60:377–389. Charlton WA. 1975. Distribution of lateral roots and pattern of lateral initiation in Pontederia cordata L. Bot Gaz 136:225–235. Charlton WA. 1977. Evaluation of sequence and rate of lateral-root initiation in Pontederia cordata L. by means of colchicine inhibition of cell division. Bot Gaz 138:71–79. Charlton WA. 1982. Distribution of lateral root primordia in root tips of Musa acuminata Colla. Ann Bot Lond 49:509–520. Charlton WA. 1983a. Patterns and control of lateral root initiation. In: Jackson MB, Stead AD, eds. Growth Regulators in Root Development. London; British Plant Regulator Group, pp 1–14. Charlton WA. 1983b. Patterns of distribution of lateral root primordia. Ann Bot Lond 51:417–427.

Lateral Root Initiation Charlton WA. 1987. Relationship between lateral root primordia in different ranks. Ann Bot Lond 60:455– 458. Charlton WA. 1991. Lateral root initiation. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. New York; Marcel Dekker, pp 103–128. Charlton WA. 1996. Lateral root initiation. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. 2nd ed. New York; Marcel Dekker, pp 149–173. Cheng J, Seeley KA, Sung ZR. 1995. RML1 and RML2, Arabidopsis genes required for cell proliferation at the root tip. Plant Physiol 107:365–376. Cline MG. 1997. Concepts and terminology of apical dominance. Am J Bot 84:1064–1069. Clowes FAL. 1958. Development of quiescent centres in root meristems. New Phytol 57:85–88. Clowes FAL. 1961. Origin of root apices. In: James WO, ed. Apical Meristems. Oxford, UK: Blackwell Scientific Publications, pp 161–167. Clowes FAL. 1978. Chimeras and the origin of lateral root primordia in Zea mays. Ann Bot Lond 42:801–807. Clowes FAL. 1985. Origin of epidermis and development of root primordia in Pistia, Hydrocharis and Eichhornia. Ann Bot Lond 55:849–857. Davidson D. 1965. Cytological chimeras in roots of Vicia faba. Bot Gaz 126:149–154. De Bruijn FJ, Wopereis J, Kapranov P, Dazzo FB, Szczyglowski K. 1998. The use of the model legume Lotus japonicus to study nodulation and lateral root formation. In: Flores HE, Lynch JP, Eissenstat D, eds. Radical Biology: Advances and Perspectives on the Function of Plant Roots. Rockville, MD: American Society of Plant Physiologists, pp 153–163. Demchenko KN, Demchenko NP. 1996. Early stages of lateral root development in Triticum aestivum L. Acta Phytogeogr Suec 81:71–75. Di Laurenzio L, Wysocka-Diller J, Malamy JE, Pysh L, Helariutta Y, Freshour G, Hahn MG, Feldmann KA, Benfey PN. 1996. The SCARECROW gene regulates an asymmetric cell division that is essential for generating the radial organization of the Arabidopsis root. Cell 86:423–433. Dittmer HJ, Spensley RD. 1947. The developmental anatomy of Descurainia pinnataochrolenca (Woot.) Detling. New Mexico Univ Publ Biol 3:1–47. Doerner P. 2000. Root patterning: does auxin provide positional cues? Curr Biol 10:R201–R203. Dolan L. 1997. SCARECROW: specifying asymmetric cell divisions throughout development. Trends Plant Sci 2:1–2. Dolan L, Janmaat K, Willemsen V, Linstead P, Poethig S, Roberts K, Scheres B. 1993. Cellular organisation of the Arabidopsis thaliana root. Development 119:71–84. Draye X, Delvaux B, Swennen R. 1999. Distribution of lateral root primordia in root tips of Musa. Ann Bot Lond 84:393–400.

151 Drew MC. 1975. Comparison of the effects of a localized supply of phosphate, nitrate, ammonium and potassium on the growth of the seminal root system, and the shoot, in barley. New Phytol 75:479–490. Dubrovsky JG, North GB, Nobel PS. 1998. Root growth, developmental changes in the apex, and hydraulic conductivity for Opuntia ficus-indica during drought. New Phytol 138:75–82. Duckett CM, Oparka KJ, Denton AM, Prior DAM, Dolan L, Roberts K. 1994. Dye-coupling in the root epidermis of Arabidopsis is progressively reduced during development. Development 120:3247–3255. Dyanat-Nejad H, Neville P. 1972. E´tude expe´rimentale de l’initiation et de la croissance des racines late´rales pre´coces du cacaoyer (Theobroma cacao L.). Ann Sci Nat Bot, Paris, 12me se´rie 13:211–246. Esau K. 1940. Developmental anatomy of the fleshy storage organ of Daucus carota. Hilgardia 13:175–226. Esau K. 1977. Plant Anatomy. 2nd ed. New York: Wiley. Estelle M. 1998. Polar auxin transport: new support for an old model. Plant Cell 10:1775–1778. Fahn A. 1990. Plant Anatomy. 4th ed. Oxford, UK: Pergamon Press. Faure JD, Vittorioso P, Santoni V, Fraisier V, Prinsen E, Barlier I, Van Onckelen H, Caboche M, Bellini C. 1998. The PASTICCINO genes of Arabidopsis thaliana are involved in the control of cell division and differentiation. Development 125:909–918. Feldman LJ. 1975. Cytokinins and quiescent center activity in roots of Zea. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. London: Academic Press, pp 55–72. Feldman LJ. 1984. Regulation of root development. Annu Rev Plant Physiol 35:223–242. Foard DE, Haber AH, Fishman TN. 1965. Initiation of lateral root primordia without completion of mitosis and without cytokinesis in uniseriate pericycle. Am J Bot 52:580–590. Forsyth C, Van Staden J. 1981. The effects of root decapitation on lateral root formation and cytokinin production in Pisum sativum. Physiol Plant 51:375–379. Fortin MC, Pierce FJ, Poff KL. 1989. The pattern of secondary root formation in curving roots of Arabidopsis thaliana (L.) Heynh. Plant Cell Environ 12:337–339. Friedberg SH, Davidson D. 1971. Cell population studies in developing root primordia. Ann Bot Lond 35:523– 533. Furuya M, Torrey JG. 1964. The reversible inhibition by red and far-red light of auxin-induced lateral root initiation in isolated pea roots. Plant Physiol 39:987–991. Gersani M, Sachs T. 1992. Development correlations between roots in heterogeneous environments. Plant Cell Environ 15:463–469. Gime´nez-Martı´ n G, de la Torre C, Lo´pez-Sa´ez JF. 1977. Cell division in higher plants. In: Rost TL, Gifford EM Jr, eds. Mechanisms and Control of Cell Division.

152 Stroudsburg, PA: Dowden, Hutchinson and Ross, pp 261–307. Gladish DK, Rost TL. 1993. The effects of temperature on primary root growth dynamics and lateral root distribution in garden pea (Pisum sativum L., cv. Alaska). Environ Exp Bot 33:243–258. Golaz F, Pilet PE. 1987. Root primordia and endogenous auxin in maize roots cultured in vitro. Physiol Plant 70:389–393. Goss MJ. 1977. Effects of mechanical impedance on root growth in barley (Hordeum vulgare). I. Effects on elongation and branching of seminal root axis. J Exp Bot 28:96–111. Granato TC, Raper CD. 1989. Proliferation of maize (Zea mays L.) roots in response to localized supply of nitrate. J Exp Bot 40:263–275. Gramberg JJ. 1971. The first stages of the formation of adventitious roots in petioles of Phaleoulus vulgaris. Koninkl Nederl Akad Wetenshappen Amsterdam 74:42–45. Guttenberg HV von. 1968. Der prima¨re bau der angiospermenwurzel. In: Linsbauer K, ed. Handbuch der Pflanzenanatomie. Berlin; Gebru¨der Borntraerger. Helariutta Y, Fukaki H, Wysocka-Diller J, Nakajima K, Jung J, Sena G, Hauser MT, Benfey PN. 2000. The SHORT-ROOT gene controls radial patterning of the Arabidopsis root through radial signaling. Cell 101:555–567. Hackett C. 1972. A method for applying nutrients locally to roots under controlled conditions and some morphological effects of locally applied nitrate on the branching of wheat roots. Aust J Biol Sci 25:1169–1180. Hinchee MAW, Rost TL. 1986. The control of lateral root development in cultured pea seedlings. I. The role of seedling organs and plant growth regulators. Bot Gaz 147:137–147. Hinchee MAW, Rost TL. 1992a. The control of lateral root development in cultured pea seedlings. II. Root fasciation induced by auxin inhibitors. Bot Acta 105:121– 126. Hinchee MAW, Rost TL. 1992b. The control of lateral root development in cultured pea seedlings. III. Spacing intervals. Bot Acta 105:127-131. Hobbie LJ. 1998. Auxin: molecular genetic approaches in Arabidopsis. Plant Physiol Biochem 36:91–102. Hobbie L, Estelle M. 1995. The axr4 auxin-resistant mutants of Arabidopsis thaliana define a gene important for root gravitropism and lateral root initiation. Plant J 7:211–220. Hooker TS, Thorpe TA. 1998. Effects of fluridone and abscisic acid on lateral root initiation and root elongation of excised tomato roots cultured in vitro. Plant Cell Tiss Org 52:199–203. Hummon MR. 1962. The effects of tritiated thymidine incorporation on secondary root production by Pisum sativum. Am J Bot 49:1038–1046.

Lloret and Casero Hurren NG, Zeadan SM, MacLeod RD. 1988. Lateral root development in attached and excised roots of Zea mays L. cultivated in the presence or absence of indol-3-yl acetic acid. Ann Bot Lond 61:573–579. Jackson D, Veit B, Hake S. 1994. Expression of maize KNOTTED1 related homeobox genes in the shoot apical meristem predicts patterns of morphogenesis in the vegetative shoot. Development 120:405–413. Jensen WA, Kavaljian LG. 1958. An analysis of cell morphology and the periodicity of division in the root tip of Allium cepa. Am J Bot 45:365–372. John PCL, Zhang K, Dong C, Diederich L, Wightman F. 1993. p34cdc2 related proteins in control of cell cycle progression, the switch between division and differentiation in tissue development, and stimulation of division by auxin and cytokinin. Aust J Plant Physiol 20:503–526. Johnson JF, Vance CP, Allan DL. 1996. Phosphorus deficiency in Lupinus albus. Altered lateral root development and enhanced expression of phosphoenolpyruvate carboxylase. Plant Physiol 112:31–41. Karas I, McCully ME. 1973. Futher studies of the histology of lateral root development in Zea mays. Protoplasma 77:243–269. Kelly MO, Bradford KJ. 1986. Insensitivity of the diageotropica tomato mutant to auxin. Plant Physiol 82:713– 717. Kerk N, Feldman L. 1994. The quiescent center in roots of maize: initiation, maintenance and role in organization of the root apical meristem. Protoplasma 183:100–106. Kerk NM, Feldman LJ. 1995. A biochemical model for the initiation and maintenance of the quiescent center: implications for organization of root meristems. Development 121:2825–2833. Klee HJ, Horsch RB, Hinchee MA, Hein MB, Hoffmann NL. 1987. The effects of overproduction of two Agrobacterium tumefaciens T-DNA auxin biosynthetic gene products in transgenic petunia plants. Gene Dev 1:86–96. Kneen BE, Weeden NF, LaRue TA. 1994. Non-nodulating mutants of Pisum sativum (L.) cv. Sparkle. Heredity J 85:129–133. Knox JP, Day S, Roberts K. 1989. A set of cell surface glycoproteins forms an early marker of cell position, but not cell type, in the root apical meristem of Daucus carota L. Development 106:47–56. Kwiatkowska M. 1988. Symplasmic isolation of Chara vulgaris antheridium and mechanisms regulating the process of spermatogenesis. Protoplasma 142:137–146. Larkin PJ, Gibson JM, Mathesius U, Weinman JJ, Gartner E, Hall E, Tanner GJ, Rolfe BG, Djordjevic MA. 1996. Transgenic white clover. Studies with the auxin-responsive promoter, GH3, in root gravitropism and lateral root development. Transgenic Res 5:325– 335.

Lateral Root Initiation Laskowski MJ, Williams ME, Nusbaum HC, Sussex IM. 1995. Formation of lateral root meristems is a twostage process. Development 121:3303–3310. Lee JS, Chang WK, Evans ML. 1990. Effects of ethylene on the kinetics of curvature and auxin redistribution in gravistimulated roots of Zea mays. Plant Physiol 94:1770–1775. Lin B-L, Raghavan V. 1991. Lateral root initiation in Marsilea quadrifolia. I. Origin and histogenesis of lateral roots. Can J Bot 69:123–135. Lloret PG, Pulgarı´ n A. 1992. Effect of naphthaleneacetic acid on the formation of lateral roots in the adventitious root of Allium cepa: number and arrangement of laterals along the parent root. Can J Bot 70:1891– 1896. Lloret PG, Vidal MR, Casero PJ, Navascue´s J. 1985. The relationship between lateral-root distribution and endodermis and pericycle cell length in Allium cepa L. adventitious roots. Ann Bot Lond 56:189– 195. Lloret PG, Casero PJ, Navascue´s J, Pulgarı´ n A. 1988. The effects of removal of the root tip on lateral root distribution in adventitious roots of onion. New Phytol 110:143–149. Lloret PG, Casero PJ, Pulgarı´ n A, Navascue´s J. 1989. The behaviour of two cell populations in the pericycle of Allium cepa, Pisum sativum, and Daucus carota during early lateral root development. Ann Bot Lond 63:465– 475. Lloret PG, Pulgarı´ n A, Casimiro I, Molina M, Casero PJ. 1998. Effect of a treatment with naphthalene-acetic acid on the arrangement of lateral roots in onion (Allium cepa L.): general pattern and coordination between ranks. Bot Acta 111:55–61. Luxova´ M. 1975. Some aspects of the differentiation of primary root tissues. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. London; Academic Press, pp 73–90. Luxova´ M. 1990. Effect of lateral root formation on the vascular pattern of barley roots. Bot Acta 103:305– 310. Lyford WH. 1975. Rhizography of non-woody roots of trees in the forest floor. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. London; Academic Press, pp 179–196. MacIsaac SA, Sawhney VK, Pohorecky Y. 1989. Regulation of lateral root formation in lettuce (Lactuca sativa) seedling roots: Interacting effects of -naphthaleneacetic acid and kinetin. Physiol Plant 77:287–293. MacLeod RD. 1990. Lateral root primordium inception in Zea mays L. Environ Exp Bot 30:225–234. MacLeod RD, Francis D. 1976. Cortical cell breakdown and lateral root primordium development in Vicia faba. J Exp Bot 27:922–932.

153 MacLeod RD, McLachlan SM. 1974. The development of a quiescent centre in lateral roots of Vicia faba L. Ann Bot Lon 38:535. MacLeod RD, Thompson A. 1979. Development of lateral root primordia in Vicia faba, Pisum sativum, Zea mays and Phaseolus vulgaris. Rates of primordium formation and cell doubling times. Ann Bot Lond 44:435– 449. Malamy JE, Benfey PN. 1997. Organization and cell differentiation in lateral roots of Arabidopsis thaliana. Development 124:33–44. Mallory TE, Chiang SH, Cutter EG, Gifford EM Jr. 1970. Sequence and pattern of lateral root formation in five selected species. Am J Bot 57:800–809. Malone CP, Miller RJ, Koeppe DE. 1978. Root growth in corn and soybeans: effects of cadmium and lead on lateral root initiation. Can J Bot 56:277–281. McCully ME. 1975. The development of lateral roots. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. London; Academic Press, pp 105– 124. McDavid CR, Sagar GR, Marshall C. 1972. The effect of auxin from the shoot on root development in Pisum sativum L. New Phytol 71:1027–1032. McMichael BL, Burke JJ. 1998. Soil temperature and root growth. Hortscience 33:947–951. Meinhardt H. 1984. Models of pattern formation and their application to plant development. In: Barlow PW, Carr DJ, eds. Positional Controls in Plant Development. Cambridge, UK: Cambridge University Press, pp 1–32. Mitchell EK, Davies PJ. 1975. Evidence for three different systems of movement of indoleacetic acid in intact roots of Phaseolus coccineus. Physiol Plant 33:290–294. Mirza JI, Olsen GM, Iversen TH, Maher EP. 1984. The growth and gravitropic responses of wild-type and auxin-resistant mutants of Arabidopsis thaliana. Physiol Plant 60:516–522. Moore R, Pasieniuk J. 1984. Graviresponsiveness and the development of columella tissue in primary and lateral roots of Ricinus comunis. Plant Physiol 74:529–533. Morris DA, Briant RE, Thompson PG. 1969. The transport and metabolism of 14C-labelled indoleacetic acid in intact pea sedlings. Planta 89:178–197. Muday GK, Haworth P. 1994. Tomato root growth, gravitropism, and lateral development: correlation with auxin transport. Plant Physiol Biochem 32:193–203. Muday GK, Lomax TL, Rayle DL. 1995. Characterization of the growth and auxin physiology of roots of the tomato mutant, diageotropica. Planta 195:548–553. Newson RB, Parker JS, Barlow PW. 1993. Are lateral roots of tomato spaced by multiples of a fundamental distance? Ann Bot Lond 71:549–557. Nishimura S, Maeda E. 1982. Cytological studies on differentiation and dedifferentiation in pericycle cells of excised rice roots. Jpn J Crop Sci 51:553–560.

154 Oparka KJ, Prior DAM, Wright KM. 1995. Symplastic communication between primary and developing lateral roots of Arabidopsis thaliana. J Exp Bot 46:187–197. Pellerin S, Tabourel F. 1995. Length of the apical unbranched zone of maize axile roots. Its relationship to root elongation rate. Environ Exp Bot 35:193–200. Peretto R, Favaron F, Bettini V, De Lorenzo G, Marini S, Alghisi P, Cervone F, Bonfante P. 1992. Expression and localization of polygalacturonase during the outgrowth of lateral roots in Allium porrum L. Planta 188:164–172. Peterson RL. 1992. Adaptations of root structure in relation to biotic and abiotic factors. Can J Bot 70:661–675. Peterson RL, Peterson CA. 1986. Ontogeny and anatomy of lateral roots. In: Jackson MB, ed. New Root Formation in Plants and Cuttings. Dordrecht, Netherlands: Martinus Nijhoff, pp 1–30. Philipson JJ. 1988. Root growth in Sitka spruce and Douglas-fir transplants: dependence on the shoot and stored carbohydrates. Tree Physiol 4:101–108. Popham RA. 1955. Zonation of primary and lateral root apices of Pisum sativum. Am J Bot 42:267–273. Pulgarı´ n A, Navascue´s J, Casero PJ, Lloret PG. 1988. Branching pattern in onion adventitious roots. Am J Bot 75:425–432. Reed RC, Brady SR, Muday GK. 1998. Inhibition of auxin movement from the shoot into the root inhibits lateral root development in Arabidopsis. Plant Physiol 118:1369–1378. Reinhardt DH, Rost TL. 1995. Primary and lateral root development of dark- and light-grown cotton seedlings under salinity stress. Bot Acta 108:457–465. Riopel JL. 1966. The distribution of lateral roots in Musa acuminata ‘‘Gros Michel.’’ Am J Bot 53:403–407. Riopel JL. 1969. Regulation of lateral root positions. Bot Gaz 130:80–83. Robinson D. 1994. The responses of plants to non-uniform supplies of nutrients. New Phytol 127:635–674. Rost TL, Jones TJ, Falk RH. 1988. Distribution and relationship of cell division and maturation events in Pisum sativum (Fabaceae) seedling roots. Am J Bot 75:1571–1583. Rowntree RA, Morris DA. 1979. Acumulation of 14C from exogenous labelled auxin in lateral root primordia of intact pea seedlings (Pisum sativum L.). Planta 144:463–466. Ruegger M, Dewey E, Hobbie L, Brown D, Bernasconi P, Turner J, Muday G, Estelle M. 1997. Reduced naphthylphthalamic acid binding in the tir3 mutant of Arabidopsis is associated with a reduction in polar auxin transport and diverse morphological defects. Plant Cell 9:745–757. Sabatini S, Beis D, Wolkenfelt H, Murfett J, Guilfoyle T, Malamy J, Benfey P, Leyser O, Bechtold N, Weisbeek P, Scheres B. 1999. An auxin-dependent dis-

Lloret and Casero tal organizer of pattern and polarity in the Arabidopsis root. Cell 99:463–472. Sattelmacher B, Thoms K. 1991. Morphology of maize root systems influenced by local supply of nitrate or ammonia. In: McMichael BL, Persson H, eds. Plant Roots and Their Environment. Amsterdam; Elsevier, pp 149– 156. Sattelmacher B, Marschner H, Ku¨hne R. 1990. Effects of the temperature of the rooting zone on the growth and development of roots of potato (Solanum tuberosum). Ann Bot Lond 65:27–36. Schade C, Von Guttenberg H. 1951. U¨ber die Entwicklung des Wurzelvegetations Punktes der Monokotyledonen. Planta 31:170–198. Scheres B, Di Laurenzio L, Willemsen V, Hauser MT, Janmaat K, Weisbeek P, Benfey PN. 1995. Mutations affecting the radial organisation of the Arabidopsis root display specific defects throughout the embryonic axis. Development 121:53–62. Scheres B, McKhann HI, Van den Berg C. 1996. Roots redefined: anatomical and genetic analysis of root development. Plant Physiol 111:959–964. Schiefelbein JW, Benfey PN. 1991. The development of plant roots. New approaches to underground problems. Plant Cell 3:1147–1154. Seago JL. 1973. Developmental anatomy in roots of Ipomoea purpurea. II. Initiation and development of secondary roots. Am J Bot 60:607–618. Seago JL, Marsh LC. 1990. Origin and development of lateral roots in Tipha glauca. Am J Bot 77:713–721. Singh BP, Sainju UM. 1998. Soil physical and morphological properties and root growth. Hortscience 33:966–971. Skene KR, Kierans M, Sprent JI, Raven JA. 1996. Structural aspects of cluster root development and their possible significance for nutrient acquisition in Grevillea robusta (Proteaceae). Ann Bot Lond 77:443–451. Stoffella PJ, Lipucci Di Paola M, Pardossi A, Tognoni F. 1992. Seedling root morphology and shoot growth after seed priming or pregermination of bell pepper. Hortscience 27:214–215. Sutcliffe JF, Sexton R. 1968. -Glycerophosphatase and lateral root development. Nature 217:1285. Torrey JG. 1956. Chemical factors limiting lateral root formation in isolated pea roots. Physiol Plant 9:370–388. Torrey JG. 1961. The initiation of lateral roots. In: Recent Advances in Botany. Toronto; University of Toronto Press, pp 808–812. Torrey JG. 1962. Auxin and purine interactions in lateral root initiation in isolated pea root segments. Physiol Plant 15:177–185. Torrey JG. 1976. Root hormones and plant growth. Annu Rev Plant Physiol 27:435–459. Torrey JG. 1986. Endogenous and exogenous influences on the regulation of lateral root formation. In: Jackson MB, ed. New Root Formation in Plants and

Lateral Root Initiation Cuttings. Dordrecht, Netherlands: Martinus Nijhoff, pp 31–66. Tschermak-Woess E, Dolezal R. 1953. Durch Seitenwurzelbildung induzierte und spontane Mitosen in den Dauergeweben der Wurzel. Ost Bot Zeit 100:358–402. Tsurumi S, Ohwaki Y. 1978. Transport of 14C-labeled indoleacetic acid in Vicia faba root segments. Plant Cell Physiol 19:1195–1206. Van Staden J, Davey JE. 1979. The synthesis, transport and metabolism of endogenous cytokinins. Plant Cell Environ 2:93–106. Van Staden J, Ntingane BM. 1996. The effect of a combination of decapitation treatments, zeatin and benzyladenine on the initiation and emergence of lateral roots in Pisum sativum. S Afr J Bot 62:11–16. Van Tieghem P, Douliot H. 1888. Recherches comparatives sur l’origine des membres endoge`nes dans les plantes vasculaires. Ann Sci Nat (Bot), 7ie`me se´rie 8:1–660. Vidal-Bernabe´ MR, Garcı´ a de la Puerta-Lo´pez P, Lo´pezGarrido E. 1998. Origin, formation and maturation of vascular tracheary elements between the main root and the lateral root of Allium cepa. Phyton Int J Exp Bot 63: 211–220. Vuylsteker C, Leleu O, Rambour S. 1997. Influence of BAP and NAA on the expression of nitrate reductase in excised chicory roots. J Exp Bot 48:1079–1085. Vuylsteker C, Dewaele E, Rambour S. 1998. Auxin induced lateral root formation in chicory. Ann Bot Lond 81:449–454. Waisel Y, Breckle SW. 1987. Differences in responses of radish roots to salinity. Plant Soil 104:191–194. Webster PL, MacLeod RD. 1980. Characteristics of root apical meristem cell population kinetics: a review of analyses and concepts. Environ Exp Bot 20:335–358. Webster BD, Radin JW. 1972. Growth and development of cultured radish roots. Am J Bot 59:744–751. Wightman F, Thimann KV. 1980. Hormonal factors controlling the initiation and development of lateral roots. I. Sources of primordia-inducing substances in the primary root of pea seedlings. Physiol Plant 49:13–20. Wightman F, Schneider EA, Thimann KV. 1980. Hormonal factors controlling the initiation and development of

155 lateral roots. II. Effects of exogenous growth factors on lateral root formation in pea roots. Physiol Plant 49:304–314. Wilcox HE. 1968a. Morphological studies of the roots of red pine, Pinus resinosa. I. Growth characteristics and patterns of branching. Am J Bot 55:247–254. Wilcox HE. 1968b. Morphological studies on the root of red pine, Pinus resinosa. II. Fungal colonization of roots and the development of mycorrhizae. Am J Bot 55:686–700. Yang RL, Evans ML, Moore R. 1990. Microsurgical removal of epidermal and cortical cells evidence that the gravitropic signal moves through the outer cell layers in primary roots of maize. Planta 180:530–536. Yorke JS, Sagar GR. 1970. Distribution of secondary root growth potential in the root system of Pisum sativum. Can J Bot 48:699–704. Zeadan SM, MacLeod RD. 1984. Some effects of indol-3-ylacetic acid on lateral root development in attached and excised roots of Pisum sativum L. Ann Bot Lond 54:759–766. Zhang HM, Forde BG. 2000. Regulation of Arabidopsis root development by nitrate availability. J Exp Bot 51:51– 59. Zhang NG, Hasenstein KH. 1999. Initiation and elongation of lateral roots in Lactuca sativa. Int J Plant Sci 160:511–519. Zhang HM, Jennings A, Barlow PW, Forde BG. 1999. Dual pathways for regulation of root branching by nitrate. Proc Natl Acad Sci USA 96:6529–6534. Zimmerman PW, Hitchcock PW. 1935. The response of roots to ‘‘root forming’’ substances. Contrib Boyce Thompson Inst 7:439–445. Zobel RW. 1973. Some physiological characteristics of the ethylene-requiring tomato mutant diageotropica. Plant Physiol 52:385–389. Zobel RW. 1975. The genetics of root development. In: Torrey JG, Clarkson DT, eds. The Development and Function of Roots. London; Academic Press, pp 261– 275. Zobel RW. 1991. Genetic control of root systems. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. New York; Marcel Dekker, pp 27–38.

9 Functional Diversity of Various Constituents of a Single Root System Yoav Waisel and Amram Eshel Tel Aviv University, Tel Aviv, Israel

environments (Caldwell, 1994; Fitter, 1996; see also Chapter 53 by Nobel in this volume.) Very few of the numerous roots that constitute one root system are exposed to the same conditions. Usually, different roots within each root system are exposed to different physical or chemical conditions that prevail in various microsites within the rooting volume. Moreover, the roots themselves increase the environmental variability by depleting certain zones of the soil of minerals and water (Grime, 1994; Stark, 1994), by secreting organic compounds that are utilized by microorganisms and by changing the ionic composition of their immediate rhizosphere (see Chapter 36 by Neumann and Roemheld in this volume). One source of variation is the gradual maturation and aging of root tissues. As the root system develops, younger parts are added while others mature and eventually senesce and die. In most cases the meristem of the root apex can continue to generate young undifferentiated cells for a long period of time. Therefore, at any instance in time, root segments of different ages and degrees of development can be found along the same root axis (see Chapter 7 by Silk in this volume). These stages of maturation correlate with differences in associated microbial and mycorrhizal activities (Stark, 1994). Such differences in microbial associations between tap and lateral roots can be found even in roots of the

Because the Faculty or Power of a Body, lieth not in any of its Principles apart; but is a Resultance from them all; or from their being, in such peculiar sort and manner, United and Combined together. So the several parts of a Clock, it is their Form, by which they are, what they are; yet it is the setting together of such Parts, and in such a way only, that makes them a Clock. Nehemia Grew (1672)

I.

INTRODUCTION

Root systems are congregates of several individual components that together constitute the functional ‘‘hidden half’’ of plants (cf. Bohm, 1979; Feldman, 1984; Eshel and Waisel, 1996). This attribute is of special importance for roots of terrestrial plants that occupy heterogeneous environments that vary spatially and temporally (Caldwell and Pearcy, 1994). The ability of a plant to produce different types of roots is an inherent aspect of its plasticity which has an important adaptive characteristic (Barlow, 1993; Bell and Lechowicz, 1994). Variation in traits among the various components of plant root systems affects the capability of these plants to cope with their complex 157

158

same plant that are grown in aeroponics, i.e., in an absolutely uniform environment (Waisel, 1996). As the individual parts of a root system develop at different microsites, under different internal and environmental conditions, and are of a variety of ages, variations in growth and physiological characteristics among them should be expected. However, though the structure and function of ‘‘typical’’ roots have been amply investigated, little attention has been given to the fine distinction in physiological traits that exist among them (Waisel and Breckle, 1987). In certain plants, several types of roots originate from different tissues and organs. Root systems of grasses are composed of a few basic types of roots: embryonic roots (including the primary root and the other seminal roots that develop adventitiously from the embryonic nodes) and roots that develop later from the lower stem nodes (adventitious, nodal, or crown roots). Roots of both systems remain active for long periods, and some of them may support the plant during the entire course of its life (Klepper et al., 1984). Nevertheless, during different stages of development, each of the two groups of roots supports different allotments of the shoots: the seminal roots support mainly the primary shoot, although some support is also given to the tillers. On the other hand, adventitious roots are basically connected only with one, or with very few of them. The two groups also differ in their physiological performance, and the contribution of seminal roots to the whole plant exceeds what would have been implied from their fractal mass. However, because of their ramified connections with various parts of the shoot, seminal roots are more important for the survival of whole plants than the adventitious roots. Many prostrate plants grow adventitious nodal roots from aboveground stems, from stolon, and from rhizomes. This additional variation of functional roots allows such plants to expand the soil area which they exploit. This may be an important trait for adaptation to habitats where the soil is shallow or waterlogged. The development of different root classes determines eventually the shape of the root system of mature plants. Root systems of various forage grasses (e.g., Lolium multiflorum, Phleum pratense, Festuca arundinacea, or Dactylis glomerata) develop bilayered root systems. Such plants have one group of surface roots and a second group of roots that penetrate deep into the soil. The roots of the first group are relatively uniform in length and diameter and exhibit high tensile strength (Kobayashi, 1977; Stokes and Mattheck,

Waisel and Eshel

1996; Nicoll, 2000). The deep roots are more heterogeneous but generally have a lower tensile strength. A similar phenomenon is found in trees (e.g., Ficus sp., Tamarix aphylla) and other woody spieces (e.g., Retama roetam). There are also differences among roots of dicotyledonous plants, composed of main roots and of several types of laterals (see Chapter 2 by Fitter in this volume). Annual crop plants like beans (Phaseolus vulgaris) (Lynch and van Beem, 1994), sunflowers (Helianthus annuus) (Tamir, Eshel, and Waisel, unpublished data), tomato (Lycopersicon esculentum), and others have in addition to their taproot and the laterals that emerge from it, a group of basal roots which originate in the hypocotyl immediately beneath the soil surface (Zobel , 1986). Such basal roots are less sensitive to gravity and extend the root system horizontally. They branch as much as the taproot or even more, and increase the specific root density at the upper soil layers. This root mass exploits the most fertile portions in agricultural soils and can perform its physiological roles efficiently. Surface roots of wheat (Triticum aestivum) and rape (Brassica campestris) are outstanding in their capability to absorb nutrients. Most of the nutrient requirements of the mature plants of these species were satisfied by the upper roots. Only a fraction of the nutrients were absorbed from deeper soil layers (Bole, 1977). Such behavior was also observed for water uptake by wheat plants. The differences in uptake between the surface roots and the deeper ones are typical of the species; they were found for plants grown in root boxes, despite the uniform water content, soil composition, bulk density, and temperature of the medium (Bole, 1977). The adaptation of Lupinus albus plants to heterogeneous soils depends, in part, on their capability to produce short root clusters in response to acidic soil conditions. The development of cluster roots in acid soil horizons enables the plants to exploit the whole depth of the soil profile (Karley, 2000). Explanation of root functions at the level of the physiological mechanism requires knowledge of the composition and characteristic of the population of roots (Aguirrezabal et al., 1993). From an agricultural point of view, knowledge of the differences in behavior that exist among various root types, and of the magnitude of their variations, is most important. It determines the functional efficiency of the whole root system. However, not enough is known about these characteristics to serve as a basis for breeding or for genetic engineering of new cultivars for more efficient root systems. Among a list of some 1400 officially

Functional Diversity

released mutant varieties only two were for root architecture (Maluszynski et al., 2000). In this chapter we will address the following questions: What are the inherited differences among roots? What characteristics are involved? How do different types of roots respond to their environment?

II.

STRUCTURAL DIFFERENCES

A.

Architecture

Every root system comprises several components. In dicots the taproot is the first leader, which bears laterals. In most cases, some of the first-order laterals and the basals become leaders that elongate fast, persist for a long time, and thicken with time. These leaders contain most of the biomass of the root system and form the long-distance transport pathways that conduct water and nutrients. The higher-order laterals are usually fine roots that make up most of the surface area of the root system. These fine roots have a limited elongation period and they are usually short-lived (see Chapter 13 by Eissenstadt and Yanai in this volume for discussion of root lifespan). Differences in the thickness of the cortex and the number of xylem vessels exist between the main axis and the laterals of tomato plants (Eshel et al., 2000). The most obvious differences among the various groups of roots are the structural differences. It was noted that even in the simple and conservative root system of Arabidopsis thaliana the anatomical structure of lateral roots is different from that of the taproot (Aeschbacher et al., 1994). Differences in structure among individual roots can be traced to the site of initiation, to the size of the primordia, and to the age of the roots. The dimensions of the proximal meristem of primary roots of maize vary within a relatively wide range. The size of the meristems has an important impact on rates of their elongation and on their performance. The size of primordia of adventitious roots was shown to have a greater impact on subsequent development than the time of their initiation (cf. Luxova and Lux, 1981). The size of the primordia affects not only the subsequent growth rates of roots but also their longevity. Certain plant genera have an additional type of roots: they have dense bottle-brush-like clusters of short lateral roots covered with a dense mat of root hairs. The archetype of this so-called Proteid roots is found in the Proteaceae (see Chapter 55 by Pate and Watt in this volume). Lupinus albus plants growing on

159

P deficient soil formed such proteoid roots as well (Skene and James, 2000). Short-root clusters are also common in pines (Wilcox 1967). Short roots of Pinus resinosa plants seem to be rapidly aborted, presumably because of the small size of their meristems. Primordia of adventitious roots of various water plants are also dimorphic: The ventral primordia of Nymphaea candida and of Nuphar lutea are capable of developing into full-size roots, whereas the upper ones either remain undeveloped (Kadej and Kadej, 1983), or develop into short roots or into aerenchymatous ones (Ellmore, 1981; Waisel and Agami, 1996). B.

Anatomy

Various roots differ in their capability for lateral transport of nutrients. To a certain degree, this has a structural basis and may be caused by variations in the conductance of the plasmodesmata. Indeed, the pore size of the plasmodesmata was found to differ among different cell types of one organ as well as among different organs (Erwee and Goodwin, 1985). Roots also differ in the structure of various mature cells. Casparian strips (i.e., the endodermal structures that control ion transport into the stele) are formed several millimeters beyond the tips of fast-growing leader roots but appear much closer to the tips in fine laterals. Thus, transport of ions into fast-growing roots would be less selective because solutes would be drawn into the xylem via the apoplast of the gap between the mature endodermis and the root tip. Ion selectivity of fine roots is much better, because of the smaller gap between the mature endodermis and the tips. Plant strategy may follow one of two alternatives for root architecture: production of long, strong, fastgrowing roots, thus sacrificing some of the selective capability; or production of fine, slowly growing roots, with the gain of better control of ion movement into the tops. In salt-treated plants, the second option seems to be preferred (cf. Hajibagheri et al., 1985). Thus, ion content of the plant seems to be determined not only by the physiological traits of the roots but also by the ratio of long roots to short ones. A series of studies McCully and colleagues (McCully and Canny, 1988; Wenzel et al., 1989; Varney and McCully, 1991; Varney et al., 1991; McCully and Mallett, 1993) described in detail the structure of maize root systems and of their distinct components. The younger parts of the nodal roots, and of their laterals, were shown to have closed immature metaxylem vessels. These root portions were

160

covered by soil sheaths. As the roots mature, their metaxylem elements open and lose their sheaths. McCully and Canny (1988) attributed the shedding of this sheath to the drying of the soil surrounding the root surface. Most of the lateral roots of maize are short and lose their active meristems as they mature. Such lateral roots have been classified into four different types on the basis of their thickness and anatomy (Varney et al., 1991). The frequency distribution of diameters of roots of maize was bimodal with a minimum diameter of 0.6 mm; 97% of axile roots were larger than this value and 98% of laterals were smaller (Chan et al., 1989). There was a correlation between root diameter and elongation rate, with laterals having a steeper slope. The data have also indicated that laterals elongated during 2.5 days only, while axile roots continued to grow throughout the plant’s life. Jordan et al. (1993) recognized 14 different models of xylem organization in roots of maize plants grown in soil or in aeroponic culture. The number of differentiated metaxylem vessels increased in nodal roots and in long laterals with internode rank, with the nodal roots having higher proportion of xylem and higher calculated hydraulic conductivity. The short laterals stopped their development at the time of late metaxylem differentiation. Thus, 92% of the short roots of maize had only one differentiated vessel. Short laterals also had smaller-diameter vessels than long ones. Significant variation in diameter was also found among long laterals of nodal roots that were initiated from various nodes. Variations among individual roots in the numbers of xylem and phloem strands of the vascular bundles were also reported for certain species of trees. Variations between three and six strands per root were reported for seedlings of Abies alba, between four and six strands for individual roots of Quercus rubra, and between six and eight strands for seedlings of Quercus robur (Rypacek et al., 1976). No doubt that roots with the larger number of vascular strands would have a better capability for long-distance transport of water and of minerals than would roots with a lower number of strands. Moreover, the larger number of phloem strands seems also to affect the subsequent development of such roots. This would have a selfenhancement effect, first increasing the translocation of assimilates to the phloem-rich roots, and consequently increasing their size and accelerate their growth and requirement for additional assimilates. An increase in the mechanical resistance of the soil reduced very much the elongation rate of main root

Waisel and Eshel

axes of pea (Pisum sativum) but did not affect elongation of the laterals (Tsegaye and Mullins, 1994). It also brought about a larger reduction in the diameter of main axes and first-order laterals than in the diameter of the second-order laterals. Distribution of lateral roots along the axes was not uniform with some segments of the root axes that were devoid of laterals and did not have soil particles clinging to them. Those were interpreted as root portions which grew across air gaps within the soil. It can be concluded that the structural differences among the various root types are tightly connected with their functional differences. The ability of a root to absorb and transport nutrients and water is governed by the morphological and anatomical characteristics of this organ, by its position within the root system, by the vascular connections with the other parts of the plant, and by its other physiological traits. The effects of structural differences among roots should be given more emphasis in future research, and attempts to quantitize such effects to explain the operation of entire root systems should be made. III.

FUNCTIONAL DIFFERENCES

A.

Growth

Elongation of main roots, initiation of lateral roots, and extension growth of laterals of various orders were frequently summed up under the general terms of root mass or root growth. This was done despite of the fact that each of the processes mentioned above reacts differently to various environmental and physiological determinants and consequently affects differently the characteristics of the entire plant (Eshel et al., 2001). One of the topics that attracted attention was the possible effect of the global increase in CO2 level on plant growth. Soybean plants grown under elevated CO2 level had larger root mass than those grown under current atmospheric conditions (Del Castillo et al., 1989). However, their treatment did not affect the rate of elongation of individual root axes. Instead, there was a significant increase in the number of actively growing roots. Thus, the root length density increased but the volume of the soil explored by the roots did not. Different growth rates for several components of the root systems of hydroponically grown sunflowers were reported (Aguirrezabal et al., 1993). Their results have indicated that the control of carbon partitioning among various components of a single root system is

Functional Diversity

determined by a combination of the distance of each sink from the source, and by its level of branching. Different portions of the root system had different rates of energy consumption, with the highest values exhibited by the growing tissue fractions (Bingham and Stevenson, 1993). The time course of initiation of lateral roots may follow different patterns: it can be linear in some roots but exponential or logistic in others. Averaging both processes for an entire root system would conceal the actual nature of development of some types of roots and mask that interesting source of variability (Buwalda et al., 1984; Waisel and Breckle, 1987; Eshel and Waisel, 1996). The subsequent development of lateral roots may also differ considerably in different species. For example, the lateral roots of spruce seedlings (Picea glauca) differ so much in their growth pattern that they actually form three different classes (Johnson-Flanagan and Owens, 1985). The growth patterns of roots may vary with their parent material. Roots of Ipomoea batatas that have developed from stored cut sprouts were longer, their cambium became active earlier, and the rate of their xylem lignification was faster than roots that have developed from fresh sprouts (Nakatani et al., 1987). In the developing system of young pea seedlings, the main root is supported by current photosynthates of the growing shoots, whereas the laterals are supplied more by the cotyledons. Changes in the structure of the root caps of tea plants cause changes in root gravitropism of main and of lateral roots and by that affect the architecture of the whole root system (Yamashita et al., 1997). Mineral status of the soil is one of the important factors affecting root growth. Heterogeneous distribution of nitrogen within the root zone of black birch (Betula lenta) brought about changes in whole root system architecture, increased link length, and induced simpler branching patterns (Crabtree and Berntson, 1994). It was concluded that in this case plant integration of the environmental responses may override local control of root growth. The proteid roots of Lupinus albus secreted acid which increased P availability in the immediate vicinity of these root clusters (Dinkelaker et al., 1989). Exposure of Lolium perenne to a low-P medium has reduced the number of adventitious roots but increased the length of each individual root (Troughton, 1977). Nitrogen nutrition and the mode of its application had specific effects on the number of lateral roots (Marriott and Dale, 1977). Abundant production of laterals is highly important for root growth in heterogeneous media. It affects the nutrient supply

161

of the plants, the allocation of assimilates, and the production and supply of growth substances (Waisel and Shapira, 1971; Marriott and Dale, 1977). Lynch and coworkers have shown that low P affects several characteristics of bean and Arabidopsis roots. The basal roots were less gravitropic under low P leading to more efficient utilization of the upper soil horizon (Bonser et al., 1996; Liao et al., 2001; Lynch and Brown, 2001). Density of root hairs was higher and their elongation was faster under low P, allowing a lager soil volume exploited by each root axis (Bates and Lynch, 1996, 2000a,b; Zhong, 2001). Growth of roots of a divided root system depends on the supply of assimilates from the tops as well as on the spatial nutrient supply from the soil. In this respect, various roots of seedlings of Picea sitchensis have developed competitive relationships; growth stimulation of some of them caused an equivalent growth inhibition of the others (Philipson and Coutts, 1977). Subjection of single roots of Suaeda monoica (Waisel and Wilcox, unpublished data) and of Rhodes grass (Chloris gayana) (Waisel, 1985) to high concentrations of NaCl or of PEG has accelerated their extension growth. Such an acceleration of growth involved a larger allocation of assimilates to the treated roots and increased supply of construction material as well as of osmotically active organic substances. Indeed, tracing the movement of 14C-labeled assimilates into single roots of Pinus contorta, or to the water-stressed half of a Picea root system (Wolswinkel, 1985; Lang and Thorpe, 1986), has proven that more assimilates have moved into such stressed roots (Reid and Mexal, 1977). This seems to present an adaptation of the plant which helps it to cope with local environmental water stresses. However, in some cases, such adaptive changes persist for a short time only. When roots were given sufficient time for adjustment the direction of flow was reversed, and more assimilates were preferentially transported to the unstressed fraction (Farrar and Williams, 1988). B.

Water Transport

The soil–root interface is highly variable with respect to water and ion availability. However, because water conductivity of the moist bulk soil can be several orders of magnitude higher than that of roots, the main bottleneck for water uptake by plants lies in the water flux across the root tissues up to the xylem. The resistance to such a flux is affected by the age of the roots, by the degree of their development, by the magnitude of their suberization, and by the resistance of

162

their endodermis (cf. Moreshet et al., 1996; Chapter 38 by Sperry et al. and Chapter 39 by Nardini et al. in this volume). Seminal roots of maize were shown to be much more efficient water suppliers to the shoots than are nodal roots (Navara, 1987). Thus, the contribution of individual roots to the water flux into a plant depends very much on the type and structure of roots and on the quantitative contribution of each of the developmental classes to the total performance of the root system. A most important characteristic that varies as root tissue mature is the degree of differentiation of the xylem vessels. In the youngest parts of the root, the xylem vessels are full of cytoplasm and other cell components. At that stage they do not conduct water as efficiently as when they are fully differentiated. The hydraulic conductivity of each root segment depends on the degree of differentiation of various xylem vessels and their respective diameters (Wenzel et al., 1989; Nobel, 1991; Wang et al., 1994; Oertli, 1996; Eshel et al., 2000). The ability of maize roots to take up water, as measured by movement of a dye into the tissue, also varied in accordance with root maturation and size (Varney and Canny, 1993; Canny and Huang, 1994). The vascular connections between the branch roots and the nodal roots is thought to have a significant role in connecting phloem and xylem conduits enabling the recycling of nutrients and organic compounds in the plant. These connections also contain tracheid barriers that are thought to reduce the chance of air bubbles from blocking the main vessels thus maintaining high hydraulic conductivity even under low water potentials (McCully and Mallett, 1993). Wan et al. (1994) compared shallow and deep lateral roots of Gutierrezia sarothrae. The deep roots were nonsuberized and had 40% more large-diameter xylem vessels than the shallow suberized laterals. As a result of these anatomical differences, the hydraulic conductivity of the deep roots was severalfold higher. Magnetic resonance imaging (MRI) was used to measure water uptake by individual roots of loblolly pine (Pinus taeda) seedlings. On the basis of weight and surface area, the tertiary fine roots were more efficient than the lateral or taproots in water uptake (MacFall et al., 1991). The so called Hydraulic lift phenomenon takes place in desert plants whenever water is taken up by the deep-reaching part of the root system, and supplied to the shallow roots which are surrounded by dry soil. The water is exuded to the upper soil layers through the night and is taken up again during the day (Richards and Caldwell, 1987). It was recently shown

Waisel and Eshel

that it may also play a role in dryland agriculture (Wan et al., 2000). Different roots differ also in the direction of water flux. Reverse flow of water was reported for vertically descending roots, while water uptake by lateral roots continued. The continuation of water uptake by lateral roots enabled a downward siphoning of water induced when the surface soil is wetted (Smith et al., 1999). Efficiency of water uptake and transport is highly dependent on individual root structure. Water uptake is directly proportional to root surface area. Water transport, on the other hand, is proportional to diameter, and degree of differentiation of the xylem vessels. When compared on the basis of carbon cost, each root is capable of performing only one of the two functions at the most efficient way. These considerations are of prime importance, when plant adaptation to arid environment is considered (see Chapter 53 by Nobel in this volume). Individual roots of grapevines supply water and nutrients to different twigs along the same branch. There was no compensation by other roots, for the loss of supply capability of any root subjected to local adverse edaphic conditions (Shani et al., 1993). This might be characteristic also of other vine-like plants as well. C.

Ion Metabolism

One aspect of the heterogeneity of root environment is the spatial and temporal variation of mineral supply. The reaction of plants to localized higher concentration of nutrients have been the subject of many studies (cf. Robinson, 1994). Those which have employed the split-root system technique either in solution or soil media, imposed a constant uneven mineral distribution conditions. Others, which have used banded fertilization or decomposing organic matter sources, had a temporary concentrated source which might have either been used up or dispersed by water movement and diffusion. The initial response of the plants in both cases is thought to be the same, but the overall ecological implication is of course different. The general trends that have been observed are localized enhancements of root proliferation at the site of higher mineral concentration, and concomitant reduction in root growth elsewhere in the rooting medium. This is, then, not a mere localized response but a change in the pattern of development of the whole root system resulting in a clustered root distribution. Such a change might have a large effect on the functional capability of the root system. The reaction of the

Functional Diversity

roots to a localized increase in mineral availability is very rapid and was shown to result from changes in lateral initiation (Gersani and Sachs, 1992), change in size of first-order laterals, change in the size of the second-order laterals, or overall change in root relative growth rate (Robinson, 1994). Different species exhibited different types of responses, and also varied in their response to the different nutrients. It was demonstrated that several grass species responded by proliferation of roots in microsites fertilized with a full nutrient solution (Larigauderie and Richards, 1994). In general, ammonium, nitrate, and phosphate increased root growth at the higher supply zone, but potassium failed to induce such a reaction (Brouder and Cassman, 1994). It was therefore concluded that banded fertilization of potassium will not be of benefit for cotton, unless it is accompanied by banded application of one of the other nutrients. From measurements of uptake and accumulation of nitrate by various parts of the maize root system, Lazof et al. (1992) concluded that in addition to nitrate uptake by the root tip it also receives nitrogen from more mature parts of the root (cf. Brady et al., 1993). Over the whole root system, the lateral roots accounted for 60% of nitrate influx. A rapidly exchanging translocation pool was located within the laterals, from which nitrate was transported to the shoot. A comparison of the uptake kinetics of nitrate by seminal and crown roots of barley, revealed that Vmax of the seminal roots decreased while that of the crown roots did not. Thus, the contribution by the seminal roots to total nitrate uptake decreased from > 50% in vegetative plants to
20% just after main shoot anthesis, and to < 5% during grain filling stage (Mattsson et al., 1993). Pate and Jeschke (1993) compared the ion content of xylem sap from various components of the root system of Banksia prionotes. In this proteaceous plant they distinguished among the taproot, which grows straight down to the underground watertable, lateral roots, and clusters of proteoid—short roots characteristic to this family. Cluster roots were found to be principal xylem donors of malate, phosphate, chloride, sodium, potassium, and amino acids. Other parts of the root system served as major sources of other cations, nitrate, and sulfate (see also Chapter 55 by Pate and Watt in this volume). Despite the ostensible coordination among the various components of root systems, the independent behavior of individual roots is obvious. Chaillou et al. (1994) studied uptake of ammonium and nitrate

163

by different root parts of hydroponically grown soybeans. They concluded that proportionality of uptake by the different parts was not governed by the shoot, which only limited the total uptake of nitrogen by supply of carbohydrates. Translocation of phosphorus from various roots to stolons of white clover (Trifolium repens) depended on the relative position of each root with respect to the stolon and on its age (Chapman and Hay, 1993). A stolon that originated from the same node as the root was the most significant sink, and its importance as a sink increased the older and larger it became (Kemball and Marshall, 1994). Nutrition of the shoots can also be affected by the storage capabilities of the roots. For example, cortex cells of maize roots delay the nutrient transport to the shoots by intercepting the transported P and by storing it. As roots age, the accumulated P is released out of their cells and eventually reaches the shoots. However, this does not occur simultaneously for all roots. Senescence of the cortex of lateral roots of wheat was shown to be faster than that of seminal roots (Fusseder, 1987). Therefore, such roots release P out of their cortex before the main axes and constitute a better P source for the shoot under conditions of marginal P supply. Behavior of shoots may also depend on the function of various root types. Salt-treated grasses contain higher levels of NaCl in the primary shoots than in their tillers. The increase in salt accumulation by the primary shoots is presumably a result of the decrease in selectivity of their older roots. On the other hand, roots of the younger tillers discriminate more efficiently against undesirable ions and therefore keep a lower salt content in their tillers. Salt content of the tillers remains low and the shoots remain unharmed until their own roots also age. The capability for ion uptake by seminal roots of maize differs from that of the adventitious ones (Fig. 1). Rates of uptake of N out of a mixture of NO3 and NH4 by the nodal roots of wheat were two to six times higher than by seminal roots (Kuhlmann and Barraclough, 1987). The uptake rates of K by taproots and by laterals of faba beans (Lev, 2000) also differed but to a smaller extent (Fig. 2). In both cases the question of whether uptake is affected by the type of the roots or by their age and developmental stage calls for additional research. Roots induce changes in the pH of their rhizosphere. Differences in pH were shown to occur along the root axis (see Chapter 33 by Gerandas and Ratcliffe in this volume) and among roots within the same root system. Such changes are determined by the

164

Waisel and Eshel

Figure 2 Uptake of RbK by taproots and by lateral roots of Vicia faba plants in the presence and absence of NaCl.

Figure 1 (a) Inflow and (b) specific absorption rates of N and K by seminal and by nodal roots of maize. (From Kuhlmann and Barraclough, 1987.)

species characteristics as well as by the metabolic responses of the individual roots to environmental influences, for example, to the ionic form of supplied N (cf. Chapter 36 by Neumann and Ro¨mheld in this volume). However, beyond the taxonomic and the environmental effects, various roots of one plant grown in a uniform soil differ in a typical way. Three examples have been presented (cf. Marschner et al., 1986): 1. Main roots of maize have alkalized their rhizosphere, raising the pH from 6.0 to 7.5. However, lateral roots of the same axis have acidified the same soil, lowering the pH from 6.0 to pH 4.5. 2. Roots of Picea abies vary with regard to their acidification capability: some of them have lowered the pH of their rhizosphere, whereas others have remained unaffected. 3. Normal roots of Lupinus albus have raised the pH of the medium from 6.3 to 7.5. However, proteoid roots of the same root system, which had been formed

under P deficiency, have acidified the medium to below pH 4.5. Such changes in pH are, of course, crucial for the acquisition of nutrients from the soil and affect the N, P, Fe, and Mn economy of the entire plant. Thus, the use of average values of rhizosphere pH, for studies of nutrient dynamics in a plant–soil system, could be inaccurate and misleading. A similar case was demonstrated for kinetic analyses of ion uptake. Such analyses have been calculated for years using average values of heterogeneous batches of roots, despite the fact that such averages may have led to erroneous and misleading conclusions. Robinson et al. (1991) estimated that only a small fraction of the roots take part in nutrient uptake. Consequently, plant nutrient uptake rates based on total root length grossly overestimate the true values. Yanai (1994) presented a modification of the BarberCushman nutrient uptake model (Barber and Cushman, 1981; see also Chapter 37 by Silberbush in this volume) by introducing a separate term to account for nutrient uptake by newly grown roots. Further developments in that direction are hindered by the lack of quantitative data regarding the variation of ion uptake kinetic parameters among roots of different ages and different positions in the root system. Proofs that the rates of ion uptake, the modes of ion interactions, the electrical properties of root cells (Ishikawa et al., 1984), and the nature of the uptake mechanisms differ along roots of maize, barley, and beans have been available for a long time (Waisel and Eshel, 1992). Variability among the individual roots did not end with differences in uptake patterns. Such roots have also shown different magnitudes of

Functional Diversity

165

uptake and a different balance between the metabolic uptake mechanism and the nonmetabolic ones. Variations along roots are not unique for ion transport and a similar phenomenon was also shown for the uptake of tritiated water. Such data provide additional arguments against the use of unspecified root batches without a detailed knowledge of their individual characteristics. Developments in microelectrode techniques allow accurate measurements of ion fluxes to be made at different locations among and along the roots with high resolution. Henriksen et al. (1992) used such a technique to measure ammonium and nitrate fluxes at various locations along seminal roots of barley. Their measurements revealed that uptake rates varied among the roots, spatially along the root, and temporally for every root. Therefore, different flux patterns were recorded during successive scans along the same portion of the root. Highest fluxes were found at the tip or within the apical two centimeters. They suggested that metabolic state of the cells at various portions of the apical region of the root may regulate these fluxes. Kochian et al. (1992) described the application of vibrating microelectrode technique to studies of ion fluxes at various sites along the root. Such a vibrating probe together with its associated computerized control and software allows for measurement of ion fluxes across the boundary layer of the root surface. Its high sensitivity enables measurements of ion fluxes at the surface of a single cell to be performed accurately (see Chapter 20 by Porterfield in this volume).

D.

Responses to Stress

Differences among individual roots, or among various root types, are usually smallest under optimal condi-

tions; that is, the conditions under which many root investigations are conducted. However, differences among roots become much more pronounced under the suboptimal conditions of salinity (Table 1), when roots are stressed (Waisel and Breckle, 1987). This is also true for O2 stresses. Oxygen-stressed roots of sunflower plants are short but have a larger density of laterals than do nonstressed roots (Wample and Reid, 1975). Eucalyptus camaldulensis plants develop fine lateral roots ( 50 240–280 21 4 4

15.5 124.9 16.9 49.8 49 26.2 46.3 54.2 58.2 49.9 10.8 2.8 10.3 5.0

Pinus Pinus Pinus Pinus Pinus Pinus

sylvestris sylvestris sylvestris sylvestris sylvestris sylvestris

(30)a (23) (22) (24) (13) (17) (17) (21) (29) (17) (34) (15) (14) (14)

Stem

Total

27.7 356.1 48.0 134.5 221 131.1 232.7 168.0 165.5 202.2 18.1 7.9 63.9b 29.6b

50.9 548.9 76.7 211.3 375 157.3 279.0 258.4 202.2 296.5 32.1 19.3 74.2 34.6

References Grier et al., 1981 Griet et al., 1981 Ovington and Madgwick, 1959a,b Ovington and Madewick, 1959a,b Nihlga˚rd, 1972 Mo¨ller et al., 1954 Mo¨ller et al., 1954 Tadaki et al., 1969 Tadaki et al., 1969 Tadaki et al., 1969 Kajimoto et al., 1999 Satoo, 1960 Parrotta, 1999 Parrotta, 1999

4

7.6

(11)

63.3b

70.9

Parrotta, 1999

4 4

18.4 15.7

(19) (18)

77.9b 69.3b

96.3 85.0

Parrotta, 1999

4

12.1

(18)

53.4b

65.5

Parrotta, 1999

4 4

11.3 4.6

(20) (16)

45.6b 23.8b

56.9 28.4

Parrotta, 1999 Parrotta, 1999

4

4.4

(18)

20.4b

24.8

Parrotta, 1999

24 38 55 60 93 30–100c 30–100c 30–100c 30–100c 30–100c 30–100c 15 7 9 12 14 14 23

20.2 38.1 58 64.7 65.4 15.5 17.0 16.8 15.5 13.9 12.3 10.8 6.6 1.3 0.6 1.6 2.2 28.1

(23) (25) (16) (25) (21) (18) (20) (19) (18) (16) (15) (17) (7) (4) (4) (5) (6) (48)

69.1b 113.1b 262 195.6b 249.5 40.4 42.2 43.0 43.3 43.5 43.7 41.9 67.0 9.6 7.3 15.5 21.7 44.3

89.3 151.2 367 260.3 314.9 87.9 86.2 86.3 85.6 84.4 83.9 64.0 94.9 30.2 16.5 32.3 38.3 58.1

28 45 47 27 26 28

7.0 19.3 11.0 6.1 6.6 7.5

(28) (20) (21) (6) (5) (7)

11.6 60.9 30.4 74.1 84.3 75.8

17.9 75.9 41.9 104.7 121.0 112.6

Sonn, 1960 Sonn, 1960 Nihlga˚rd, 1972 Sonn, 1960 Sonn, 1960 Chibisov, 1995 Chibisov, 1995 Chibisov, 1995 Chibisov, 1995 Chibisov, 1995 Chibisov, 1995 Satoo, 1960 Albrektson, 1980 Albrektson, 1980 Albrektson, 1980 Albrektson, 1980 Albrektson, 1980 Ovington, 1957; 1959; Ovington and Madgwick, 1959a,b Ma¨lko¨nen, 1974 Ma¨lko¨nen, 1974 Ma¨lko¨nen, 1974 Albrektson, 1980 Albrektson, 1980 Albrektson, 1980 (continued)

198

Persson

Table 4 (continued) Species

Age

Pinus sylvestris Pinus sylvestris Pinus sylvestris Pinus sylvestris Pinbus sylvestris

29 33 34 50 55

Pinus sylvestris Pinus sylvestris Pinus sylvestris Pinus sylvestris Pinus sylvestris Populus deltoides Populus deltoides Populus deltoides Populus deltoides Quercus borealis Pseudosuga menziesii Pseudotsuga menziesii Equatorial forest Equatorial forest Intermediate savanna Open shrub savanna Dense shrub savanna Savanna woodland Mixed savanna Tropical evergreen Tropical deciduous Tropical deciduous Temperatate deciduous Northern hardwoods Northern hardwoods Pygmy cypress Subalpine spruce-fir

84 85 88 100 120–150 1d 2d 3d 4d 49 35–50 Uneven Uneven Uneven Uneven Uneven Uneven Uneven Uneven 18 Uneven 50 Uneven Uneven 79–106 Uneven 200–500

Woody roots

Stem

Total

5.7 32.7 8.6 7.6 34.1

(6) (18) (9) (8) (31)

63.2 118.8 66.7 64.6 96.7

94.3 185.9 98.5 90.8 109.0

10.0 17.9 21.1 12.5 14.2 13.8 16.0 16.5 16.7 15.0 67.4 141.3 2.8 2.6 3.6 9.2 14.3 26.6 18.6 31.3 34.3 24.7 39 28.3 20.6 7.1 18

(7) (34) (34) (11) (19) (19) (19) (19) (19) (9) (23) (16) (9) (11) (32) (29) (29) (31) (26) (21) (15) (9) (17) (18) (18) (32) (14)

108.6 41.0 48.2 81.3 45.2 48.2 49.4 53.6 55.1 111.9 263.282 704.5b 19.2 14.7 7.4b 21.9 32.6 54.3b 53.4b 116.4b 141.1 261.6b 231b 132.8 51.0 20.4 94

146.8 53.4 62.4 117.0 74.0 72.0 85.8 88.1 88.7 176.4 293.8 857.1 31.5 24.3 11.3 31.8 48.6 84.7 72.0 147.7 234.8 286.3 231 161.1 111.8 21.8 125

References Albrektson, 1980 Ovington and Madgwick, 1959a,b Albrektson, 1980 Albrektson, 1980 Ovington, 1957; 1959; Ovington and Madgwick, 1959a,b Albrektson, 1980 Fine´r, 1992 Fine´r, 1992 Albrektson, 1980 Bringmark, 1977 Lodihyal & Lodihyal, 1997 Lodihyal & Lodihyal, 1997 Lodihyal & Lodihyal, 1997 Lodihyal & Lodihyal, 1997 Ovington et al., 1962 Fogel and Hunt, 1992 Sollins et al., 1980 Bartholomev et al., 1953 Nye, 1961 Menat and Cesar, 1979 Menat and Cesar, 1979 Menat and Cesar, 1979 Menat and Cesar, 1979 Ogawa et al., 1961 Greenland and Kowal, 1960 Bandhu, 1970 Greenland and Kowal, 1960 Anderson, 1970 Gosz et al., 1976 Whittaker et al., 1974 Westman, 1978 Arthur and Fahey, 1991

a

Percentage share in woody root systems is indicated within parenthesis. Data are available only for stem þ branches. c Picea abies stands for natural origin that have undergone thinning of various intensity during the past 30 years (1000, 2000, 3000, 4000, 5000, and 6000 stems ha1 , respectively). d Plantations. b

influence of nutrient availability, light intensity and CO2 on growth and shoot/root ratio in structural components of young Betula pendula plants. He concluded that (1) root growth was favored when N, P, or S were major constraints for short roots; (2) shoot growth was favored when K, Mg, and Mn restricted shoot growth; (3) that shortage of Ca, Fe, and Zn regime had almost no effect on shoot:root ratio—that the light regime had no effect on dry matter allocation, and (4) that shortage of CO2 strongly decreased root development. The root biomass of forests is not static; it represents a varying proportion of the total tree biomass

during stand development (Table 4). Biomass accumulation in the woody roots does not occur at the same time as accumulation in aboveground structures. During stand development, root growth investments may even be higher than investments in aboveground structures (Lyr and Hoffman, 1967). Nevertheless growth of the above- and belowground structures are mutually interdependent. The ultimate rooting depth or the total lateral root spread in forest stands is usually not correlated with the aboveground size of the trees. While ‘‘living’’ fine roots persist more frequently in the upper part of the

Root Systems of Arboreal Plants

soil horizon, structural roots are generally spread more deeply in the soil profile (Persson, 1980b). The growth of the root tip determines the position, length, and orientation of the structural woody roots. The distribution of woody roots in forest stands, in relation to distance from trees, is dependent on the stand (Persson, 2000). Coarse roots are usually found, in dense forest stands, more frequently in the soil near the trees, while fine roots are distributed more regularly over the total stand (Persson, 1980a,b). Root biomass in the structural root system is usually minimal at maximum tree density (stem ha1 ) but increases proportionally to thinning intensity (Chibisov, 1995). However, the relationships between forest productivity and thinning operations are still not well understood. Tree roots are often spread as far as, or beyond, the projected crown (Fig. 2; Kutschera et al., 1997). In boreal forests, small and poorly developed trees frequently developed extensive root systems extending far beyond the tree crowns. Conservation of essential nutrients and their effective utilization is required to sustain forest production, especially in nutrient-poor sites. Trees on poor sandy soils allocate considerably more biomass to the structural roots than trees growing on fertile clay soil (Nielson and Dencker, 1998). At high elevations or in permafrost areas, the root/shoot ratio of forest trees is higher than in neighboring, climatically more favourable areas (Van Cleve et al., 1981). Roots of woody plants are stiffened and thickened by secondary xylem/wood production. Variations in growth rates around the root circumference are a general feature, especially for horizontal roots. Therefore root cross sections tend to be very eccentric, though their structure on the wind side differs from that on the lee side. Young roots are generally circular in shape. However, later they change toward greater xylem production on the upper side near the root base or in the strained side than on the stressed side. Thickening of

199

roots started first in more distal parts of the roots (Head, 1968). During some years the roots do not thicken at all. Thus, the number of rings of the secondary wood cannot be safely used for estimation of the age of roots. Moreover, radial growth tends to differ in root sections of different-diameter classes. The largest contribution to root biomass and the largest relative radial growth was found in thinner-diameter classes (Deans, 1981). The uprooting of trees by wind (i.e., windthrow) occurs very irregularly in forests, mainly on sites with shallow structural roots. The increased root/stem ratio with increasing thinning intensity of forest stands is presumably caused by the increase in water and nutrient availability, but also by enhanced mechanical stress in stems and structural roots (Chibisov, 1995; Nielson and Dencker, 1998; see also Chapter 10 by Stokes in this volume). The tree root systems subjected to wind movements lose a high proportion of fine roots, which must continuously be replaced. The impression of stability and permanence of the trees mask the essentially dynamic nature of forest ecosystems. In forest ecosystems, large-diameter tree roots survive and function for many seasons. Many large diameter roots remain alive as long as the trees stay alive, and in temperate regions may extend to > 100 years. As the forest stand ages, a considerable weight fraction accumulates in large-diameter roots (Ovington, 1962; Ovington and Madgwick, 1959a,b; Persson et al., 1995). However, carbon budgets suggest that the annual investment in the maintenance of the fine roots may even be higher (A˚gren et al., 1980; Axelsson and Axelsson, 1986). There are no good methods for determination of mortality and decomposition of large roots (Fogel, 1985, 1990). Many large-diameter roots of forest trees survive for a long time, though generally they exhibit a varying rate of mortality and high masses of dead root tissue have been reported for various forest ecosystems (Puhe, 1994).

Figure 2 Distribution of the structural root system in a 13-year-old Pinus sylvestris stand on a sandy podsol in Central Sweden. The soil water table was at 10 m. No taproot was found under those circumstances. (From Kutschera et al., 1997.)

200

V. CONCLUSIONS Structural roots of arboreal trees play important roles in the mechanical support of the trees, and protecting the soil against erosion. Fine roots derive their greatest functional importance in supplying the needs of the aboveground parts of trees with water and nutrients. Roots play an important role as producers of growthregulating compounds that affect the shoots. The amount of fine roots is for functional reasons in constant flux; fine roots repeatedly penetrate the soil horizons, often at the expense of high rates of death and renewal. Production and replacement of fine roots accounts for the high rate of primary production of forests. Extensive growth of fine root requires an adequate supply of carbohydrates and hormones and sufficient soil water and nutrient supply, adequate temperature and oxygen, low soil impedance, and limited competition with neighboring roots. Water and some nutrients move through the soil to the roots, but this is not enough. Tree roots extend into new areas by means of fast-growing lateral roots originally extending from the taproot. By penetrating the soil substrate, the roots improve soil structure, and the nutrient and soil water holding capacity. Roots compete with each other for carbohydrates, and those occupying favorable soil sites tend to grow at the expense of other components of the same root system. Most roots of the common tree species of the world grow within the upper 1–2 m of the soil. The most active fine roots are concentrated in the top 5 cm. Roots of most trees are usually spread far beyond the tree crowns. Biomass studies of tree roots at the ecosystem level indicate that the belowground investment is high and that root systems are expensive to maintain. Compared to our knowledge of the aboveground organs of trees, our knowledge about root function and behavior still needs boosting.

REFERENCES Aaltonen VT. 1920. U¨ber die Ausbreitung und den Reichtum der Baumwurzeln in den Heidewa¨ldern Lapplands. Acta Forest Fenn 14:1–75. Abuzinadah RA, Read DJ. 1986. The role of proteins in the nitrogen nutrition of ectomycorrhizal plants. III. Protein utilization of Betula, Picea and Pinus in mycorrhizal association with Hebeloma crustuliniforme. New Phytol 103:507–314. Abuzinadah RA, Read DJ. 1989. Carbon transfer associated with assimilation of organic nitrogen sources by silver birch (Betula pendula Roth.). Trees 3:17–23.

Persson A˚gren G, Axelsson B, Flower-Ellis JGK, Linder S, Persson H, Staaf H, Troeng E. 1980. Annual carbon budget for a young Scots pine. Ecol Bull (Stockh) 32:307–313. Ahlstro¨m K, Persson H, Bo¨rjesson I. 1988. Fertilization in a mature Scots pine (Pinus sylvestris L.) stand—effects on fine roots. Plant Soil 106:179–190. Albrektson A. 1980. Relationships between tree biomass fractions and conventional silvicultural measurements. Ecol Bull (Stockh) 32:315–327. Andersson F. 1970. An ecosystem approach to vegetation, environment and organic matter of a mixed woodland and meadow area. Summary of doctoral Thesis, Univ. Lund, Sweden. Arthur MA, Fahey TJ. 1991. Biomass and nutrients in an Engelmann spruce—subalpine fir forest in north central Colorado: pools, annual production, and internal cycling. Can J For Res 22:315–325. Atkinson D. 1980. The growth and activity of fruit tree root systems under simulated orchard conditions. In: Sen DN, ed. Environment and Root Behaviour. Jodhpur, India: Geobios International, pp 171–185. Axelsson E, Axelsson B. 1986. Changes in carbon allocation patterns in spruce and pine trees following irrigation and fertilization. Tree Physiol 2:189–204. Bandhu D. 1970. A study of the productive structure of northern tropical dry deciduous forests near Varanasi. I. Stand structure and non-photosynthetic biomass. Trop Ecol 11:92–106. Bartholomew WY, Meyer J, Laudeleaut H. 1953. Mineral nutrient immobilization under forest and grass fallow in the Jangambi (Congos region) with some preliminary results. Publ Inst Natn Etude Agron Congo Belge, Ser Sci 57:1–40. Bartsch N. 1987. Response of root systems of young Pinus sylvestris and Picea abies plants to water deficits and soil acidity. Can J For Res 17:805–812. Bormann FH. 1966. The structure, function, and ecological significance of root grafts in Pinus strobus L. Ecol Monogr 36:2–26. Bowen GD. 1985. Roots as a component of tree productivity. In: Cannel MGR, Jackson JE, eds. Attributes of Trees as Crop Plants. Edinburgh; Institute of Terrestrial Ecology, Natural Environment Research Council, pp 303–315. Bringmark L. 1977. A bioelement budget of an old Scots pine forest in Central Sweden. Silva Fenn 11:201–257. Caldwell MM. 1987. Competition between roots in natural communities. In: Gregory PJ, Lake JV, Rose DA, eds. Root Development and Function. New York; Cambridge University Press, pp 167–185. Cannell MGR, Dewar RC. 1994. Carbon allocation in trees: a review of concepts for modelling. Adv Ecol Res 25:59–104. Chibisov GA. 1995. Bioproductivity of spruce stands in northern European Russia. Water Air Soil Pollut 82:87–96.

Root Systems of Arboreal Plants Chung H-H, Kramer PJ. 1975. Absorption of water and 32P through suberized and unsuberized roots of loblolly pine. Can J For Res 5:229–235. Cossmann KF. 1940. Citrus roots: their anatomy, osmotic pressure and periodicity of growth. Palestine J Bot 3:65–106. Cropper WP, Gholz HL. 1991. In situ needle and fine root respiration in mature slash pine (Pinus elliottii) trees. Can J For Res 21:1589–1595. Deans JD. 1981. Dynamics of coarse root production in a young plantation of Picea sitchensis. Forestry 54:25– 41. Dickmann DL, Pregitzer KS. 1992. The structure and dynamics of woody plant root systems. In: Michell JB, Ford-Robertson T, Hincley TM, Sennerby-Forsse L, eds. Ecophysiology of Short Rotation Forest Crops. New York; Elsevier Science, pp 95–123. Drexhage M, Chauvie`re M, Colin F, Nielsen CNN. 1999. Development of structural root architecture and allometry of Quercus petraea. Can J For Res 29:600–608. Eissenstat DM, Yanai RD. 1997. The ecology of root lifespan. Adv Ecol Res 27:1–60. Ericsson A, Persson H. 1980. Seasonal changes in starch reserves and growth of fine roots of 20-year-old Scots pines. Ecol Bull (Stockh) 32:239–250. Ericsson T. 1995. Growth and shoot:root ratio of seedlings in relation to nutrient availability. Plant Soil 168– 169:205–214. Eshel A, Waisel Y. 1996. Multiform and multifunction of various constituents of one root system. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. 2nd ed. New York; Marcel Dekker, pp 175–191. Farrell EP, Leaf AL. 1974. Effects of fertilization and irrigation on root numbers in a red pine plantation. Can J For Res 4:366–371. Fine´r L. 1992. Biomass and nutrient dynamics of Scots pine on a drained ombrotrophic bog. Fin For Res Inst Res Pap 420:1–122. Fitter AH. 1982. Morphometric analysis of root systems: application of the technique and influence of soil fertility on root system development in two herbaceous species. Plant Cell Environ 5:313–322. Fogel R. 1985. Roots as primary producers in below-ground ecosystems. In: Fitter AH, Atkinson D, Read DJ, Usher M, eds. Ecological Interactions in Soils. Oxford, UK: Blackwell Scientific, British Ecol Soc, Spec Publ 4:23–36. Fogel R. 1990. Root turnover and production in forest trees. HortScience 25:270–273. Fogel R, Hunt G. 1982. Contribution of mycorrhizae and soil fungi to nutrient cycling in a Douglas-fir ecosystem. Can J For Res 13:219–232. Ford ED, Deans JD. 1977. Growth of a Sitka spruce plantation: spatial distribution and seasonal fluctuation of

201 lengths, weights and carbohydrate concentrations of fine roots. Plant Soil 47:463–485. Gill RA, Jackson RB. 2000. Global patterns of root turnover for terrestrial ecosystems. New Phytol 147:13–31. Gosz JR, Likens GE, Bormann FH. 1976. Organic matter and nutrient dynamics of the forest and forest floor in the Hubbard Brook Forest. Oecologia 22:305–320. Greenland DJ, Kowal JML. 1960. Nutrient content of the moist tropical forest of Ghana. Plant Soil 12:154–174. Grier CC, Vogt KA, Keyes MR, Edmonds RL. 1981. Biomass distribution and below and above-ground production in young and mature Abies amabilis zone ecosystems of the Washington Cascades. Can J For Res 11:155–167. Harley JL. 1959. The Biology of Mycorrhiza. London; Leonard Hill. Hatch AB, Doak KD. 1933. Mycorrhizal and other features of the root systems of pine. J Arnold Arbor 14:85–99. Head GC. 1968. Seasonal changes in the diameter of secondarily thickened roots of fruit trees in relation to growth of other parts of the tree. J Hort Sci 43:275– 282. Heikurainen L. 1955. U¨ber Vera¨nderungen in den Wurzelverha¨ltmissen der Kiefernbesta¨nde auf Moorbo¨den im Laufe des Jahres. Acta For Fenn 65:1–70. Hermann RK. 1977. Growth and production of tree roots. In: Marshall JK, ed. The Belowground Ecosystem: A Synthesis of Plant-Associated Processes. Fort Collins, CO: Colorado State University, Range Science Department Science Series 26:7–28. Holstener-Joergensen H. 1958–59. Investigations of root systems of oak, beech and Norway spruce on groundwater-affected moraine soils with a contribution to elucidation of evapotranspiration of stands. Forstl Forsoeksv Danm 15:227–289. Jackson RB, Canadell J, Ehleringer JR, Mooney HA, Sala OE, Schulze ED. 1996. A global analysis of root distributions for terrestrial biomes. Oecologia 108:389– 411. Jenik J. 1971. Root structure and underground biomass in equatorial forests. In: Duvigneau P, ed. Productivity of Forest Ecosystems. Paris; UNESCO, pp 323–331. Kalela E. 1954. U¨ber die Wurzelverha¨ltnisse der Kiefernsamenba¨ume und -Baumbesta¨nde. Acta For Fenn 61:1–17. Kajimoto T, Matsuura Y, Sofronov MA, Volokitina AV, Mori S, Osawa A, Abaimov AP. 1999. Above and belowground biomass and net primary productivity of a Larix gmelinii stand near Tura, Central Sibiria. Tree Physiol 19:815–822. Kolesnikov VA. 1968. Cyclic renewal of roots in fruit plants. In: Ghilarov MS, Kovda A, Novichkova-Ivanova LN, Rodin LE, Sveshnikova VM, eds. Methods of Productivity Studies in Root Systems and

202 Rhizosphere Organisms. St. Petersburg, Russia: Publishing House NAUKA, pp 102–106. Kozlowski TT. 1971a. Growth and Development of Trees. I. Seed Germination, Ontogeny, and Shoot Growth. New York; Academic Press. Kozlowski TT. 1971b. Growth and Development of Trees. II. Cambial Growth, Root Growth, and Reproductive Growth. New York; Academic Press. Kozlowski TT, Kramer PJ, Pallardy SG. 1991. The Physiological Ecology of Woody Plants. New York; Academic Press. Ko¨stler JN, Bru¨ckner E, Bibelriether H. 1968. Die Wurzeln der Waldba¨ume. Berlin; Verlag Paul Parey. Kramer PJ, Bullock HC. 1966. Seasonal variations in the proportions of suberized and unsuberized roots of trees in relation to the absorption of water. Am J Bot 53:200–204. Kuhns MR, Garret HE, Teskey RO, Hinckley TM. 1985. Root growth of black walnut trees related to soil temperature, soil water potential, and leaf water potential. Forest Sci 31:617–629. Kutschera L, Lichtenegger E, Sobotik M, Haas D. 1997. Bewurzelung von Pflanzen in der verschiedenen Lebensra¨umen. 5. Band der Wurzelatlas-Reihe. Stapfia 49:1–331. Ku¨lla T, Lo˜hmus K. 1999. Influence of cultivation method on root grafting in Norway spruce (Picea abies [L.] Karst.). Plant Soil 217:91–100. Laing EV. 1932. Studies in Tree Roots. Forestry Com Bull 13, 81 pp. Lambers H. 1987. Growth, respiration, exudation and symbiotic associations: the fate of carbon translocated to roots. In: Gregory PJ, Lake JV, Rose DA, eds. Root Development and Function. London; Cambridge University Press, pp 24–35. Lambers H, Scheurwater I, Atkin OK. 1996. Respiratory pattern in roots in relation to their functioning. In: Waisel Y, Eshel, A, Kafkafi U, eds. Plant Roots. The Hidden Half. 2nd ed. New York; Marcel Dekker, pp 323–362. Ledig FT. 1983. The influence of genotype and environment on dry matter distribution in plants. In: Huxley PA, ed. Plant Research and Agroforestry. Nairobi, Kenya: Int. Council Res. Agroforestry, pp 427–454. Ledig FT, Bormabb FH, Wenger KF. 1970. The distribution of dry matter between shoot and roots in loblolly pine. Bot Gaz 131:349–359. Leshem B. 1965. The annual activity of intermediary roots of the Aleppo pine. Forest Sci 11:291–298. Leshem B. 1970. Resting roots of Pinus halepensis: structure, function, and reaction to water stress. Bot Gaz 131:99– 104. Loescher WH, McCamant T, Keller JD. 1990. Carbohydrate reserves, translocation, and storage in woody plants. HortScience 25:274–281. Lodhiyal LS, Lodhiyal N. 1997. Variation in biomass and net primary productivity in short rotation high density

Persson central Himalayan popular plantations. For Ecol Manag 98:167–179. Lyford WH, Wilson BF. 1966. Controlled growth of forest tree roots: techniques and application. Harvard For Pap 16:1–12. Lyr H, Hoffmann G. 1967. Growth rates and growth periodicity of tree roots. Int Rev For Res 2:181–236. Ma¨lko¨nen E. 1974. Annual primary production and nutrient cycle in some Scots pine stands. Comm Inst For Fenn 84:1–87. Marschner H. 1988. Mineral Nutrients of Higher Plants. London: Academic Press. Marshall JD, Waring RH. 1985. Predicting fine root production and turnover by monitoring root starch and soil temperature. Can J For Res 15:791–800. Menaut JC, Cesar J. 1979. Structure and primary productivity of Lamto savannas, Ivory Coast. Ecology 60:1197– 1210. Mo¨ller CM, Mu¨ller D, Nielsen J. 1954. Respiration in stem and branches of beech. Forstl Sorsoeksv Danm 21:273–301. Morrow RR. 1950. Periodicity and growth of sugar maple surface layer roots. J For 48:875–881. Nambiar EK. 1987. Do nutrient retranslocate from fine roots? Can J For Res 17:913–918. Newman EI. 1988. Mycorrhizal links between plants: their functioning and ecological significance. Adv Ecol Res 18:243–270. Nielson CN, Dencker I. 1998. Root architecture and root/ shoot-ratios of Norway spruce as affected by thinning intensity and soil type in Denmark. In: Box JE, ed. Root Demographics and Their Efficiencies in Sustainable Agriculture, Grasslands and Forest Ecosystems. Dev Plant Soil Sci 82:721–735. Nihlga˚rd B. 1972. Plant biomass, primary production and distribution of chemical elements in a beech and a planted spruce forest in South Sweden. Oikos 23:69– 81. Noelle W. 1910. Studien zur vergleichenden Anatomie und Morphologie der Koniferenwurzeln mit Ru¨cksicht auf Systematik. Bot Zeit 68:169–266. Noland TL, Mohammed GH, Scott M. 1997. The dependence of root growth potential on light level, photosynthetic rate, and root starch content in jack pine seedlings. New For 13:105–119. Nye PH. 1961. Organic material and nutrient cycles under moist tropical forest. Plant Soil 13:250–263. Ogawa H, Yoda K, Kira T. 1961. A preliminary survey on the vegetation of Thailand. Nat Life SE Asia 1:21–157. Olsthoorn AFM, Titak A. 1991. Fine root density and root biomass of two Douglas-fir stands on sandy soils in the Netherlands. Neth J Agric Sci 39:67–77. Orlov AY. 1980. Cyclic development of roots of conifer and their relation to environmental factors. In: Sen DN, ed. Environment and Root Behaviour. Jodhpur, India; Geobio International, pp 43–61.

Root Systems of Arboreal Plants Ovington JD. 1957. Dry-matter production by plantations of Pinus sylvestris L. Ann Bot NS 21:287–314. Ovington JD. 1962. Quantitative ecology and the woodland ecosystem concept. Adv Ecol Res 1:103–192. Ovington JD, Madgwick HAI. 1959a. The growth and composition of natural stands of birch. 1. Dry-matter production. Plant Soil 10:271–283. Ovington JD, Madgwick HAI. 1959b. The growth and composition of natural stands of birch. 2. The uptake of mineral nutrients. Plant Soil 10:389–400. Parrotta JA. 1999. Productivity, nutrient cycling, and succession in single- and mixed-species plantations of Casuarina equisetifolia, Eucalyptus robusta, and Leucaena leucocephala in Pourto Rico. For Ecol Manag 124:45–77. Persson H. 1978. Root dynamics in a young Scots pine stand in Central Sweden. Oikos 30:508–519. Persson H. 1980a. Spatial distribution of fine root growth, mortality and decomposition in a young Scots pine stand in Central Sweden. Oikos 34:77–87. Persson H. 1980b. Death and replacement of fine roots in a mature Scots pine stand. Ecol Bull (Stockh) 32:251– 260. Persson H. 1980c. Fine-root dynamics in a Scots pine stand, with and without near optimum nutrient and water regimes. Acta Phytogeogr Suec 68:101–110. Persson H. 1982. Changes in the tree and dwarf shrub fineroots after clear cutting in a mature Scots pine stand. Swed Con For Proj, Techn Rep 31, 18 pp. Persson H. 1995. The role of roots in carbon cycling in forests. In: Helmisaari H-S, Smolander A, Suokas A, eds. The Role of Roots, Mycorrhizas and Rhizosphere Microbes in Carbon Cycling in Forest Soil. Fin For Res Inst Res Pap 537:119–126. Persson H. 1996. Fine-root dynamics in forest trees. Acta Phytogeogr Suec 81:17–23. Persson H. 2000. Adaptive tactics and characteristics of tree fine roots. Dev Plant Soil Sci 33:337–346. Persson H, Majdi H, Clemensson-Lindell A. 1995. Effects of acid deposition on tree roots. Ecol Bull (Stockh) 44:158–167. Pettersson R, McDonald AJS, Stadenberg I. 1993. Response of small birch plants (Betula pendula Roth.) to elevated CO2 and nitrogen supply. Plant Cell Environ 16:1115– 1121. Puhe J. 1994. Die Wurzelentwicklung der Fichte (Picea abies [L.] Karst.) bei unterschiedlichen chemischen Bodenbedingungen. Ber. Forschungszentrums Waldo¨kosysteme Univ Go¨ttingen 198:1–129. Raich JW, Nadelhoffer KJ. 1989. Belowground carbon allocation in forest ecosystems: global trends. Ecology 70:1346–1354. Rickman RW, Letey J, Stolzy LH. 1966. Plant response to oxygen supply and physical resistance in the root environment. Proc Soil Sci Soc Am 30:304–307.

203 Roberts J. 1976. A study of root distribution and growth in a Pinus sylvestris L. (Scots pine) plantation in East Anglia. Plant Soil 44:607–621. Rodin LE, Bazilevich NI. 1967. Production and Mineral Cycling in Terrestrial Vegetation. London; Oliver and Boyd. Santantonio D, Hermann RK. 1985. Standing crop, production, and turnover of fine roots on dry, moderate, and wet sites of mature Douglas-fir in western Oregon. Ann Sci For 42:113–142. Santantonio D, Santantonio E. 1987. Effect of thinning on production and mortality of fine roots in a Pinus radiata plantation on a fertile site in New Zealand. Can J For Res 17:919–928. Santantonio D, Hermann RK, Overton WS. 1977. Root biomass studies in forest ecosystems. Pedobiologia 17:1– 31. Satoo T. 1966. Production and distribution of dry matter in forest ecosystems. Miscell Inf Tokyo Univ For 16:1– 15. Sen DN. 1980. Root system and root ecology. In: Sen DN, ed. Environment and Root Behaviour. Jodhpur, India: Geobios, pp 1–24. Scarratt JB, Glerum G, Plexman CA, eds. 1981. Proceedings of the Containerized Tree Seedling Symposium. Toronto; Canadian Forestry Service. Schoettle AW, Fahey TD. 1994. Foliage and fine root longevity of pines. Ecol Bull (Stockh) 43:136–153. Sollins P, Grier CC, McCorison FM, Cromack K, Fogel R, Fredriksen RL. 1980. The internal element cycles of an old-growth Douglas-fir ecosystem in Western Oregon. Ecol Monogr 50:261–285. Soon SW. 1960. Der Einfluss des Waldes auf die Bo¨den. Jena, Germany: Gustav Fischer Verlag. Stattin E. 1999. Root freezing tolerance and stability of Scots pine and Norway spruce seedlings. Doctoral thesis, Swedish University of Agricultural Science. Stevens CL. 1931. Root growth of white pine (Pinus strobus L.). Yale Univ School For Bull 32:1–64. Stone EL, Kalisz PJ. 1991. On maximum extent of tree roots. For Ecol Manage 46:59–102. Strober C, Eckart GA, Persson H. 2000. Root growth and response to nitrogen. In: Schulze E-D, ed. Carbon and Nitrogen Cycling in European Forest Ecosystems, Ecol Stud 142. Berlin; Springer, pp 99–121. Sutton RF. 1969. Form and development of conifer root systems. Comm For Bur Oxford, Techn Commun 7:1–131. Sutton RF. 1980. Root system morphogenesis. NZ J For Sci 10:264–292. Sutton RF. 1983. Root growth capacity; relationship with field root growth and performance in outplanted jack pine and black spruce. Plant Soil 71:111–122. Sutton RF, Tinus RW. 1983. Roots and root system terminology. For Sci Monogr 24:1–125.

204 Tadaki Y, Hatiya K, Tochiaki K. 1969. Contribution from JIBP-PT 63, Gov For Exp Sta. Meguro (Tokyo) 51:331–339. Torrey JG. 1976. Root hormones and plant growth. Annu Rev Plant Physiol 27:433–459. Tschirch A. 1905. U¨ber die Heterorhizie bei Dikotylen. Flora 94:68–79. Van Cleve K, Barney R, Schlentner R. 1981. Evidence of temperature control of production and nutrient cycling in two inferior Alaska black spruce ecosystems. Can J For Res 11:258–273. Van Rees KCJ, Comerford NB. 1990. The role of woody roots of slash pine seedlings in water and potassium absorption. Can J For Res 20:1183–1191. Vogt KA, Persson H. 1991. Measuring growth and development of roots. In: Lassoie JP, Hincley TM, eds. Techniques and Approaches in Forest Tree Ecophysiology. Boca Raton, FL: CRC Press, pp 477–501.

Persson Vogt K, Grier CC, Vogt DJ. 1986. Production, turnover, and nutrient dynamics of above- and belowground detritus of world forests. Adv Ecol Res 15:303–377. Wargo PM. 1979. Starch storage and radial growth in woody roots of sugar maple. Can J For Res 9:49–56. Westman WE. 1978. Patterns of nutrient flow in the pygmy forest region of Northern California. Vegetatio 36:1– 15. Whittaker RH, Bormann FH, Likens GE, Siccama TG. 1974. The Hubbard Brook ecosystem study: forest biomass and production. Ecol Monogr 44:233–252. Wilcox H. 1954. Primary organization of active and dormant roots of noble fir, Abies procera. Am J Bot 41:812–821. Wilson BF. 1964. Structure and growth of woody roots of Acer rubrum L. Harv For Pap 1:1–14. Woods FH, Brock K. 1964. Interspecific transfer of Ca-45 and P-32 by root systems. Ecology 45:886–889. Yin X, Perry JA, Dixon RK.1989. Fine-root dynamics and biomass distribution in a Quercus ecosystem following harvesting. For Ecol Manag 27:159–177.

12 Root–Shoot Relations: Optimality in Acclimation and Adaptation or the ‘‘Emperor’s New Clothes’’? Peter B. Reich University of Minnesota, St. Paul, Minnesota

I.

INTRODUCTION

also focused on the range of components of root and shoot systems (including differences in morphology, chemistry, phenology, metabolism, and longevity), although data on root properties under contrasting environmental conditions are still scarce. In this chapter I revisit the issues raised above, focusing on the existing literature. I will not consider the actual internal communication devices used by plants to signal differential rates of growth or senescence of aboveground versus belowground tissues. Instead I will ask whether root/shoot relationships (1) maintain a so-called functional balance over time; (2) differ phenotypically in relation to differences in supply of light, nutrients, CO2, or water, or merely differ due to variation in plant size resulting from variation in these factors; (3) differ among plant species differing in habitat affinity; and (4) of aggregations of plants in stands are similar in their behavior to individual plants.

Differences in the relative sizes and/or function of root and shoot systems can arise due to differences in allocation of biomass, in morphology and chemistry of absorptive root and shoot tissues, and in the turnover rates of those tissues. For as long as the topic has been considered by plant biologists, a consistent theme has prevailed—that, given the relatively opposite roles of shoots and roots in uptake and use of key resources, there should be some kind of balance in size (e.g., biomass) and surface area (or related metric) between the root and shoot systems of individual plants (and perhaps stands). Such a balance should vary in relation to factors such as resource supply and species habitat affinity. For example, this would lead to a proportionally greater root than shoot system when nutrients were in short supply. Although such ideas have been taken as near-paradigm by several generations of physiologists and ecologists, we need to ask: How firmly does the evidence support these ideas? Historically, the majority of discussion of root– shoot relationships has focused on the relative biomass fractions in roots versus shoots. Scientists have

A.

Conceptual Considerations

The close coordination of growth of root and shoot systems has long intrigued plant biologists. It led to the functional equilibrium model of coordinated shoot and root growth (Brouwer, 1962a, 1983). It also led to the optimality theory (Wilson, 1988; Bloom et al., 1985; Thornley, 1972, 1998) that argues for preferential partitioning to the part of the plant

Abbreviations: LAR, leaf area ratio; LMF, leaf mass fraction; NAR, net assimilation rate (mol m2 h1); RGR, relative growth rate (d1); RMF, root mass fraction; SLA, specific leaf area (cm2 g1); SRL, specific root length (cm g1).

205

206

Reich

which acquires the currently most limiting resource. Such partitioning hypothetically increases the ability of the plant to acquire that most limiting resource. Consistent with the optimality theory is the idea that species will show marked differences in biomass distribution in a fashion related to their differences in habitat affinities. According to such hypotheses, for example, plants adapted to growth in deep shade should tend to display high biomass allocation to leaves (Givnish , 1988), and plants adapted to growth in conditions with limited soil resources (nutrients or water) should tend to display high biomass allocation to roots (Chapin, 1980). Thus, theory suggests that there should be functional balance among root and shoot systems, that this balance should vary with resource supply, and that this balance should vary among species in relation to variation in life history traits and/or habitat affinities. As demonstrated below, the first point has been shown to be true so far as it has been tested. In fact, the evidence suggests that a tendency toward functional balance may be so pronounced that the second and third points, often taken as paradigm and nearparadigm, respectively, are in fact weak trends compared to the tendency toward individual functional balance. They have been overemphasized in studies of young plants due to failure to account for ontogenetic drift (e.g., Rice and Bazzaz, 1989; Coleman et al., 1994; Reich et al., 1998a). Moreover, differences in morphology, chemistry, and metabolism of roots and leaves are so profound in terms of impact on uptake and use of resources (e.g., Fitter, 1985; Poorter and Remkes, 1990; Poorter et al., 1990; Garnier, 1992, 1998; Lambers and Poorter, 1992; Ryser and Lambers, 1995; Reich et al., 1998b; Poorter and Nagel, 2000) that they exert greater influence on growth than differences in root–shoot biomass fraction in young plants. B.

Terminology

1. Allocation, Partitioning, and Distribution Although root/shoot ratio has often been criticized for being a static ratio, at least it is defined similarly by most authors, as the ratio of dry mass of belowground [i.e., roots] vs. aboveground plant parts at a given point in time. In contrast, the terms biomass allocation and partitioning (often used similarly) have had diverse uses. These range from narrow definitions in which allocation or partitioning refers only to the shortterm transfer and placement of new increments of biomass or specific substances (including carbohydrates,

proteins, lipids, etc.) to broad definitions in which allocation or partitioning refers specifically to the relative distribution of biomass pools among plant parts at a point in time (e.g., Poorter and Nagel, 2000). For purposes of this chapter, I will follow the definitions used by Lutze and Gifford (1998) and refer to allocation and partitioning as synonyms for the transfer and placement of new increments of biomass. I will refer to the amount of biomass present in any organ system at any comparable time or plant size, relative to the total plant biomass, as the biomass distribution. Hereafter I will use the term carbon roughly synonymously with biomass, since carbon represents roughly half of plant biomass. Biomass distribution then is a function of biomass allocation and biomass turnover rate, as shown in this example for root mass fraction (RMF): RMF = (Biomass allocated to roots – root biomass turnover)/(Biomass allocated to roots – root biomass turnover) + (Biomass allocated to stems – stem biomass turnover) + (Biomass allocated to foliage – foliage biomass turnover) In such a definition, turnover would include biomass lost via respiration as well as by tissue senescence or herbivory. It is clear that biomass distribution (which is frequently measured) is equally a result of turnover as of allocation (both of which are rarely measured), although this important and simple point is often overlooked. Hence, since turnover rates of roots are still very poorly quantified for the vast majority of species and situations (Eissenstat and Yanai, 1997), we have little reliable data for biomass allocation for whole plants, except during very early rapid growth phases when fine-root turnover is likely to be very low. This lack of empirical data on allocation is further highlighted when we consider our failure to account for carbon allocated to root symbionts such as mycorrhiza in most species and to root nodule bacteria in others, and carbon exuded from roots. Thus, the vast majority of usable, interpretable data are for biomass distribution. Again, for young seedlings, this might largely reflect biomass allocation since turnover has not yet begun in a major fashion. However, for older plants (including older ‘‘seedlings’’), any conclusions reached in the literature about biomass allocation per se (usually based on biomass distribution) are likely suspect, since the turnover and allocations parts of the equation are generally unknown.

Root–Shoot Relations

2.

Root–Shoot Ratios, Biomass Fractions, and Allometric Analyses

Given historic use of the root–shoot ratio I will repeatedly refer to it in this chapter, although I prefer the term biomass fraction to refer to the fraction of total plant mass in a given plant component. The term biomass fraction is preferred for statistical reasons over root/shoot ratio, although they convey the same information. In addition to evaluating patterns of biomass fractions, I shall also discuss the use and value of allometric relationships in which the log-transformed biomass of the shoot is plotted against the logtransformed biomass of the roots or of the whole plant (Pearsall, 1927; Evans, 1972). This enables the separation of differences in biomass distribution due to differences in size from those due to true shifts in partitioning. Although allometry is not perfect, it is ‘‘the only routine method of showing an effect of treatment on net partitioning’’ (Farrar and Gunn, 1998). When allometry is performed to relate roots to shoots, the allometric slope constant k is the ratio of the relative growth rate (RGR) of the shoot versus the RGR of the root, and hence a precise measure of relative net change in biomass for the shoot vs. the root. It is useful to evaluate whether k is less than or greater than unity, since this directly defines the direction of the ontogenetic drift in partitioning. Although k is useful as typically defined, it does not distinguish among stems, leaves, and roots, which may limit its usefulness (Poorter and Nagel, 2000). However, one can determine a k for any specific pair of tissues. Additionally, nonlinearities in the allometric relationships are problematic and the slopes of allometric relationships do not fully account for differences in the overall elevation of the allometric lines. Apparently, plants might allocate proportionally less to one component (roots, for example) as they grow larger under a given treatment A as compared to another treatment B (i.e., have a lower slope under treatment A if roots are the Y-axis variable). Yet they still might have a greater proportion of total biomass in that tissue component at any plant size because of differences in intercept. Hence, allometric relationships should consider the slopes and intercepts to adequately interpret potential differences in allocation and distribution. Finally, allometry provides some information on the outcome of partitioning, not the process itself. Despite these limitations it is far more appropriate as a means of addressing questions of partitioning than comparing plants in different treatments at the same time (Evans, 1972; Coleman et al., 1994; Lutze

207

and Gifford, 1998). A number of authors have also compared leaf mass fraction (LMF) or RMF to plant size (Peace and Grubb, 1982; McConnaughay and Coleman, 1999; Poorter and Nagel, 2000), which accomplishes roughly the same purpose as the direct allometric approach. C.

Evidence

In evaluating the published data in support of or against the idea of optimality in acclimation and adaptation of biomass allocation or distribution, I used all my reprints plus 675 articles gleaned from an electronic search for ‘‘root’’ and ‘‘shoot’’ using the database Agricola. Roughly 200 citations met the criteria needed to be relevant to the topic at hand. Although not an exhaustive search, it is likely relatively comprehensive. Moreover, when I refer to papers that did not account for ontogenetic drift and yet arrived at conclusions regarding shifts in biomass allocation, there is no attempt to denigrate these authors. To demonstrate this, I point out here that my own work (e.g., Reich et al., 1987; Walters and Reich, 1989) has suffered from the failure to account for ontogenetic drift. II.

FUNCTIONAL BALANCE AND RESPONSES OF ROOT VERSUS SHOOT SYSTEMS TO VARIATION IN RESOURCES

A.

Functional Balance, Optimality Theories, and Ontogenetic Drift

The relative differences in response of root versus shoot systems to variation in resources such as light, water, nutrients, or CO2 has long been a central question in plant biology. Brouwer (1962a,b) performed several pioneering investigations that showed that leaf- or root-pruned seedlings rapidly regained their original root–shoot biomass balance and that the pruned plants were then on the same line in the allometric analysis as the control plants; i.e., when compared to a (younger) control plot with a similar dry mass, they had the same proportion of that mass in roots and shoots. Surprisingly few studies were ever made to validate this finding, although work by Caloin et al. (1991), Farrar and Gunn (1998), and Poorter and Nagel (2000) demonstrate the same result for herbaceous plants. Indirect evidence in this same vein has been seen for woody plants by Eissenstat and Duncan (1992), Reich et al. (1993), Kruger and Reich (1993), and Vanderklein and Reich (1999) for lightly to

208

moderately shoot-pruned or defoliated woody plants. Hence I will presume that this is a general process until proven otherwise. If true, this is remarkable, in that the control and conservation of biomass distribution occurs only due to plastic variation in the allocation or partitioning of new carbon (Farrar and Gunn, 1998). The recovery of root–shoot balance in partially defoliated plants comes at a cost, such as of fruit production in citrus (Eissenstat and Duncan, 1992). Even if we accept that all plants behave in such a fashion, many possible mechanisms may be involved, and understanding of this process is surprisingly limited to this date (Farrar and Gunn, 1998; Poorter and Nagel, 2000). Building conceptually on earlier studies that showed greater shoot fraction in low light or root fraction under low soil resource supply (which one should note were typically based on root–shoot ratios of plants of different sizes), Brouwer (1983) synthesized his own work to develop the ‘‘functional equilibrium’’ theory. In a simplified framework this can be stated as (Poorter and Nagel, 2000): plants shift their allocation toward shoots if the carbon gain of the shoot is limited by a low level of aboveground resources, such as light or CO2. Allocation is shifted toward roots if belowground resources, such as nutrients or water, are at low levels. Optimal partitioning theory further suggests that plants respond to variation in the environment by partitioning biomass among various plant organs to optimize the acquisition of nutrients, light, water, and carbon in a manner that maximizes plant growth (Thornley, 1969; Bloom et al., 1985). For example, plants experiencing low levels of nutrient supply would be predicted to shift resources toward root growth and to processes associated with nutrient capture rather than those associated with carbon uptake. In this sense we can think of the optimal partitioning theory as an extension of the ‘‘functional equilibrium’’ model. Both functional equilibrium and optimal partitioning models explain the recovery of a stable root–shoot biomass distribution following pruning or nutrient starvation. They also explain shifts toward higher root biomass fraction under low soil resources and higher shoot biomass fraction under high soil resources believed to be a common whole-plant behavior. Moreover, these shifts could be considered adaptive, since they might lead toward enhanced capture of the resources most limiting to plant growth. However, optimality may not be a good explanation. Instead, one could argue that the plant is reestablishing a functional equilibrium selected for over time for reasons unrelated to optimality and determined genetically.

Reich

Since the mechanisms controlling functional equilibrium are still at best partially known, and other aspects of the ‘‘optimal allocation vis-a`-vis resource supply’’ theory are only weakly supported, I propose that functional equilibrium may be as much or more a case of plants maintaining homogeneity rather than responding optimally to variation in resource supply. Optimal partitioning theory is generally accepted— in fact, it is loosely used as a paradigm in plant ecology, based on concept and empirical evidence (Bloom et al., 1985; Chapin, 1991; Reynolds and D’Antonio, 1996), although there is no consensus about the mechanisms involved. This idea still merits partial support, but a variety of studies have indicated that some significant fraction of the evidence taken as support for such theory is actually due to failure to account for ontogenetic drift (Evans, 1972; Rice and Bazzaaz, 1989; Walters et al., 1993a,b; Coleman et al., 1994; Reich et al., 1998a; Farrar and Gunn, 1998; McConnaughay and Coleman, 1999). Such allometric evidence suggests that the plasticity of allocation in response to resource variation is often relatively muted (and differentially dependent on the resource involved). The large differences reported in the literature and attributed to plasticity sometimes reflect large differences in plant size, rather than true differences in allocation. In fact, failure to account for ontogenetic drift can cause authors to misjudge the magnitude, duration, and even direction of many of the adjustments in biomass allocation patterns found for many species (McConnaughay and Coleman, 1999). Those investigators noted that a surprisingly small number of studies have explicitly distinguished between biomass distribution differences that result as a natural consequence of ontogenetic drift (i.e., plant growth and development) and ‘‘true’’ adjustments in biomass distribution (i.e., those that require an adjustment in biomass allocation). The numbers of such publications are small as compared to the much larger number of papers on plant responses to nutrients, light, water, or CO2 published in the past several decades without considering ontogenetic drift. Poorter and Nagel (2000) have argued that root– shoot differences that occur among treatments at a point in time due only to differences in prior growth rate, and hence in plant size at that time, are nonetheless meaningful. This is because the plants actually co-occur in time and differences in biomass fractions may influence their response to current conditions. Although this latter possibility may be true, the hypothetical plants are also likely to differ in many other ways (leaf or root chemistry, metabolism, plus

Root–Shoot Relations

overall plant size) that may influence their functions. Hence, Evans (1972), Coleman et al. (1994), and Lutze and Gifford (1998) have argued, and I agree, that the comparison of biomass distribution in plants, that are of widely differing sizes because of microenvironmental differences, sheds no light on the actual process of allocation, or of plasticity in allocation. If plants of a given species under low light conditions actually shifted their proportional allocation of new biomass toward leaves (which I define as a plastic response), they would have a higher k and likely a higher LMF at the same size as high light plants. If the plants make absolutely no change in allocation, high light will cause faster overall growth and hence over time the plants diverge along the same developmental trajectory and with exactly the same partitioning patterns, and yet have divergent LMF at the same time. B.

Responses to Resource Gradients

1.

Light

Optimal partitioning theory hypothesizes increased allocation to leaf production should be a phenotypic response of plants to lower light environments (e.g., Givnish, 1988). The vast majority (in fact, every test I could find) of studies that compared plants under varying light conditions reported higher leaf mass fraction in low-light-grown plants when plants were harvested at a common time (cf. Olff et al., 1990; Callaway, 1992; Latham, 1992; Lei and Lechowicz, 1998) and a meta-analysis of such studies also reports a significant shift toward leaves and stems and away from roots under low light (Poorter and Nagel, 2000). This evidence of differences in biomass distribution of plants of widely different size that grew under differing light conditions led to the common paradigm that plants shift allocation toward leaves when grown at low light. However, most of the studies that addressed this issue examined plants only at a common time. In contrast, the vast majority of the smaller set of studies that explicitly tested for allocation by accounting for ontogenetic drift, have not found increasing allocation to leaf mass in low light. This is the case for 23 species (of 26 tested) as disparate as herbs (Evans and Hughes, 1961; Hughes and Evans, 1962; Peace and Grubb, 1982; Rice and Bazzaz, 1989; Philippot et al., 1991; Casper et al., 1998; McConnaughay and Coleman, 1999), crop plants (Terry, 1968; Corre, 1983), and woody plant seedlings (Ledig et al., 1970; Steinbrenner and Rediske, 1964; Walters et al., 1993a; Stoneman and Bell, 1993; Chen, 1997; Chen and Klinka, 1998; Reich et al., 1998a). Thus, most

209

studies that accounted for ontogenetic drift found no evidence of allocation shift toward leaf production at low light. Using an alternative approach, Ingestad and Agren (1991) also concluded that there is no effect of irradiance on allocation if a plant is at steady-state nutritional conditions. Thus the preponderance of evidence suggests that plastic allocational shifts in biomass in response to light gradients predicted by optimality theory are the exception rather than the rule. As an example of the evidence regarding biomass allocation patterns under contrasting light environments, I present data for seedlings of nine woody species that were summarized in a condensed fashion in Reich et al. (1998a). It is evident (Fig. 1; Table 1) that there is no consistent difference in the allometric coefficient k (here defined as the change in leaf RGR versus root RGR, ignoring stem mass) in contrasting light treatments. Only one of the nine species had a significant shift in k, and many species even had lower k in lower light. The intercepts were generally not significantly different either. Only Larix had a significantly different intercept, showing for this species that even for a given root size, low-light plants had greater leaf mass despite a similar k. Using these data, we also see no significant differences between light treatments if LMF is regressed against plant size. One alternative way of examining allocational shifts bears mention. Some studies which reported an effect of light on biomass allocation and attempted to adjust for plant size, did so using analysis of covariance (ANCOVA), wherein the relationship between leaf fraction and light environment was statistically adjusted for plant size (cf. Veneklass and Poorter, 1998; Poorter et al., 1999). For our data set (Fig. 1; Table 1), ANCOVA results were often substantially different from those from a strictly allometric analysis, and the range of plant size occupied by each treatment was quite distinct, so that the statistical adjustment using ANCOVA may not adequately account for ontogenetic drift. Therefore, until such data are examined using both allometry and ANCOVA, studies that use only ANCOVA should be considered weak evidence as a test of plasticity. When analyzed allometrically, for woody plants there was no shift toward higher LMF at lower light for 12 species (Walters et al., 1993a; Reich et al., 1998a), but there was a consistent shift toward stems and away from roots. This was also seen earlier for etiolated peas (Pearsall, 1927) and for the woodland herb Impatiens in studies by Evans and Hughes (1961), Huges and Evans (1962), and Peace and

210

Reich

Figure 1 Variation in leaf mass versus root mass, shown on a logarithmic scale, for young seedlings of nine woody plant species grown at low and high light (5% and 25% of full sunlight, respectively) (n ¼ 59 plants on average per light level per species). Statistical summaries of slope and intercept differences across light treatments are provided in Table 1. Individual data points are individual plants harvested at nine different times over the course of the growing season. (Expanded from data summarized in Reich et al., 1998a.)

Table 1 Summary of Values of the Allometric Constant (k) for the Relationship Between Leaf Mass and Root Mass for Each of Nine Woody Species in Two Contrasting Light Environments—Low (5% of full sunlight) and High (25% of full sunlight). Species Populus tremuloides Betula papyrifera Larix laricina Pinus banksiana Betula allegheniensis Pinus strobus Picea glauca Picea mariana Thuja occidentalis

Shade tolerance classification Intolerant Intolerant Intolerant Intolerant Intermediate Intermediate Tolerant Tolerant Tolerant

Low light 0.92 1.07 0.91 0.76 0.96 0.73 0.89 0.90 0.89

        

0.05 0.03 0.04 0.07 0.07 0.04 0.08 0.08 0.08

High light

P ðkÞ

P (intercept)

        

ns * ns ns ns ns ns ns ns

ns — * ns ns ns ns ns ns

0.84 0.89 0.93 0.87 0.97 0.73 0.94 0.96 0.81

0.02 0.02 0.02 0.03 0.02 0.02 0.03 0.03 0.03

Mean is followed by 1 SE. Species are arrayed from least to most shade tolerant. Probabilities (P) contrast the two light levels and are those for the slope (the allometric constant k) and for the intercept of the allometric relationship (not testable for Betula papyrifera since there were significant slope differences). Source: Expanded from data summarized in Reich et al. (1998a).

Root–Shoot Relations

Grubb (1982). These patterns could be construed as indirect support for the optimal allocation model, since stems and leaves collectively are needed to harvest light and fix carbon, while roots acquire water and nutrients. If both leaf fraction and stem fraction increase under low-light conditions then one could argue that these changes supported the optimality theory, since immediate resource gain would occur. However, if only stem fraction increases, as commonly observed, it suggests a strategy to improve the likelihood of future potential resource gain and perhaps to increase the likelihood of survival. Moreover, since the plants in these experiments were growing as isolated individuals, with shade coming from high above the plant, increased carbon gain would have come directly from more leaves or leaf area but not from more stem mass or stem length. Hence, the observed shift toward stems rather than leaves must represent an acclimation that has been selected for because of its value in natural environments, rather than a plastic response to increase carbon gain in the near term. In contrast to the weak evidence for optimal biomass partitioning with respect to light—or, in fact, any consistent shift in LMF—dramatic shifts in leaf morphology were routinely shown, e.g., with leaves in low light having lower density, lower thickness, or both. This was demonstrated by higher SLA values, even when accounting for ontogenetic drift in SLA with plant size (Hughes, 1965; Peace and Grubb, 1982; Walters et al., 1993a; Reich et al., 1998a). There are few data available examining shifts in root properties (such as specific root length) under differing light conditions, but the data are consistent with leaf responses. Plants of nine species grown under low-light conditions had lower specific root lengths and higher SLA than plants of the same size grown under higher light conditions (Fig. 2; Reich et al., 1998a). Although all nine species showed shifts in SRL with light environment, these are not easily explained simply by differences in the slope of the change in root length with root dry mass or whole plant mass. The initial differences between light levels (in the smallest, youngest plants measured) were sometimes so large that SRL was higher in high-light than in low-light plants across their common range of plant sizes. This is despite a steeper slope of root length versus root mass for low light plants. Acclimation of tissue morphology appears to represent a greater means of plasticity than shifts in biomass allocation in response to variation in light (Evans, 1972; Walters et al., 1993a; Reich et al., 1998a). Moreover, greater SLA and lesser SRL at low light

211

Figure 2 Specific leaf area (cm2 foliage per g foliage) and specific root length (m root length per g root) for seedlings of nine species grown at low and high light (5% and 25% of full sunlight, respectively). See Fig. 1 or Table 1 for species names. (From Reich et al., 1998a.)

are consistent with the idea of improving the uptake of the most limiting resource (Givnish, 1988), assuming that uptake of resources is related to high surface area or length of the tissue in question (Reich et al., 1998b). Hence, optimality theory might be better supported if it were considered using models that incorporate the full constellation of plant traits and responses, including morphological ones that result in changes in the relative area or length of absorptive tissues. 2.

Nutrients and Water

The evidence for phenotypic shifts in partitioning is stronger for nutrients than any other resource, but still is not universal. The predominant response noted in studies that did not account for ontogenetic drift was a large shift toward roots and away from shoots at low nutrient supply (Poorter and Nagel, 2000). Most studies that adjusted for plant size also found shifts in allocation across varied nitrogen supply, as reported by Ryser and Lambers (1995), Volin and Reich (1996), Gedroc et al. (1996), Lutze and Gifford (1998), Baxter et al. (1997) and McConnaughay and Coleman (1999) (Fig. 3). The relative magnitude of the shift was typically smaller when examined ontogenetically rather than at a single

212

Figure 3 Variation in root mass versus shoot mass, shown on a logarithmic scale, for young seedlings of the annual herbaceous plant species Abutilon theophrasti (left) and Chenopodium album (right) grown at low (open circles) and high (closed circles) nutrient regimes (n=22 plants per nutrient regime per species). Individual data points are individual plants harvested at different times over the course of the growing season. (From Gedroc et al., 1996.)

point in time. Evidence in response to supply of P and other nutrients is less clear. Eissenstat et al. (1993) found shifts toward decreasing RMF with increasing P supply in sour orange, even when comparing plants at a common size. However, using data from a study of common bean plants (Nielsen et al., 1998), there is no evidence of decreasing RMF with increasing P supply, and in fact, the evidence suggests the opposite (Fig. 4). Ryser and Lambers (1995) found declining RMF with increasing P supply in one grass species (Brachypodium pinnatum) but not in another (Dactylis glomerata). Using a different experimental approach, several authors concluded that there is a systematic shift in allocation for variable P and N supply, but not for K (Ingestad and Agren, 1991; see also Van der Werf et al., 1993a,b). Similar to the limited database on root traits across light gradients, there are few data for root traits across nutrient gradients. Ryser and Lambers (1995) found that in the grass species Dactylis glomerata, SRL decreased with increasing N supply at various levels of P supply, but in the grass species Brachypodium pinnatum there was no decline in SRL with increasing N supply. In neither species was there a consistent shift in SRL with variation in P supply. In contrast, Eissenstat et al. (1993) found a decrease in SRL with increasing P supply in sour orange. Only scant data on plastic allocation responses to water availability are available in experiments where ontogenetic drift was accounted for in the design. Of those data, increased allocation toward roots, when water was in low supply, have been reported for some woody plants (Ledig et al., 1970; Tomlinson and Anderson, 1998) and during certain phenotypical phases in different woody seedlings (McMillan and

Reich

Wagner, 1995). However, it was not observed for annual herbs (McConnaughay and Coleman, 1999). This database is too scant and inconsistent to provide a basis for reaching firm conclusions. Moreover, with regard to water acquisition, optimality theory may not apply in cases of uncertainty (Lerdau, 1992; Lerdau and Gershenzon, 1997). If plants were to develop roots after water shortage occurred, it would be too late. Therefore they may have been selected to invest in excessive amounts of roots in habitats that may be subjected to periodic and unpredictable drought.

Figure 4 Variation in root mass versus total plant mass, shown on a logarithmic scale, for young common bean plant seedling grown (lower panel) at low and high phosphorus supply rates and (upper panel) with or without mycorrhizal infection. Individual data points are average values for plants harvested at four dates over the time course of the experiment, pooled across the other treatment. (From Nielsen et al., 1998.)

Root–Shoot Relations

3.

213

Elevated Atmospheric CO2

C.

Consistent with optimality theory, it was hypothesized (e.g., Reynolds and Thornley, 1982) and often empirically observed (e.g., Eamus and Jarvis, 1989; Ceulemanns and Mousseau, 1994; among many such publications) that growth in elevated CO2 results in increased dry-mass partitioning to roots. However, few if any of these reports accounted for ontogenetic drift in biomass fraction with plant size. In contrast, opposite results were reported by researchers who made explicit tests with woody and herbaceous species of the role of ontogenetic drift vis-a`-vis biomass allocation under elevated CO2 (Tjoelker et al., 1998; Gunn et al., 1999; Lutze and Gifford 1998). They found that the strong size dependence of partitioning of dry mass among plant parts resulted in ‘‘apparent’’ differences in root and shoot dry mass partitioning, that in fact were the result of differences in plant size and not a functional adjustment to CO2 environment (e.g., Table 2, Fig. 5). Recent meta-analyses that did not account for plant size (Curtis and Wang, 1998; Poorter and Nagel, 2000) also indicated that these earlier suggestions of optimality in root–shoot partitioning in response to elevated CO2 are not supported by critical evidence.

Conclusions Regarding Experimental Tests of Allocation Across Resource Gradients

A better understanding of acclimation of biomass distribution under variable environmental conditions is critical to our ability to comprehend plant growth, or the responses of plants to multiple resource scenarios (McConnaughay and Coleman, 1999). Optimal partitioning theory has often been the basis for models which attempt to predict plant responses to global environmental changes. If the assumption that plants shift biomass partitioning in response to environmental cues is untrue, or much more limited in degree than previously believed, the predictions of such models must be questioned. Some studies show pronounced shifts in allocation as hypothesized, but others do not. What can we therefore conclude? If optimality allocation theory holds for all plant resources, we should see predicted shifts in response to variation in light, CO2, water, and nutrients. There is strong evidence against such claims for light and CO2, strong evidence in favor of these claims for N, and variable evidence for P or water.

Table 2 Allocation of Dry Mass to Roots Among Five Boreal Species (Populus tremuloides, Betula papyrifera, Larix laricina, Pinus banksiana, and Picea mariana) Grown as Seedlings Under Ambient (370
mol mol1) and Elevated (580
mol mol1) Concentrations of CO2 in Combination with Five (day/night) Growth Temperatures Growth temperature (8C) Genus

CO2

18/12

21/15

24/18

27/21

30/24

Populus

370 580 P > Fa 370 580 P>F 370 580 P>F 370 580 P>F 370 580 P>F

1.19  0.04 1.18  0.05 ns 1.07  0.02 1.07  0.01 ns 0.93  0.01 0.96  0.01 0.09 1.05  0.02 1.07  0.02 ns 1.01  0.01 1.01  0.01 ns

1.10  0.02 1.08  0.02 ns 1.05  0.01 1.05  0.01 ns 1.03  0.02 0.99  0.02 ns 1.01  0.02 1.05  0.02 ns 0.97  0.02 0.94  0.01 ns

1.11  0.02 1.21  0.02 .001 1.15  0.03 1.09  0.02 ns 1.02  0.01 1.01  0.01 ns 0.98  0.02 1.02  0.01 0.05 1.00  0.01 1.00  0.01 ns

1.04  0.02 1.08  0.02 ns 1.05  0.01 1.05  0.01 ns 0.99  0.01 1.01  0.01 ns 0.99  0.01 0.99  0.01 ns 0.97  0.01 0.98  0.01 ns

1.11  0.02 1.12  0.01 ns 1.08  0.02 1.04  0.01 ns 1.04  0.01 1.05  0.01 ns 1.00  0.02 0.99  0.01 ns 0.99  0.02 1.04  0.02 0.08

Betula

Larix

Pinus

Picea

Mean (SE) allometric coefficients relating change in root mass to change in plant mass are shown. Slope values determined from individual plant dry mass data using the linear regression equation: ln (root) = slope  ln (plant) + b, all R2  0.97. Values >1.0 indicate an increased partitioning of dry mass to roots for a given change in plant mass. a P values for test of separate slopes for CO2 treatment in analysis of covariance; ns indicates P > :1. Source: Tjoelker et al. (1998).

214

Reich

Figure 5 The effect of atmospheric CO2 concentration on (A) shoot–root ratio at a number of harvest dates, and (B) the allometric relationship between shoot and root, for Bellis perennis and Trifolim repens. Plants were grown hydroponically in controlled environments of either 350 or 700
mol CO2 mmol1. Open bars of each pair (above) represent elevated CO2 levels. Open squares (below) represent elevated CO2 levels. (From Gunn et al., 1999.)

Collectively, the body of evidence suggests that the optimal allocation theory is not a general theory. Why is this so? An idea central to this theory is that plants plastically shift biomass allocation to have a higher resource gain under a new situation than would otherwise occur without such a shift. However, several important issues make this idea untenable: 1. A variety of traits influence plant resource balance, including morphology, metabolism, and chemistry, that are often more biologically significant and influential than biomass allocation (Evans, 1972; Lambers and Poorter, 1992; Reich et al., 1998a,b). For example, a plant can change carbon gain by either changing LAR or by changing the photosynthetic rate

of that leaf area (NAR). A plant can change leaf area by changing either SLA or LMF. In short, there are numerous ways to enhance C gain and RGR, and only few of them require shifts in biomass allocation. Thus, even if plants did routinely phenotypically change to maximize resource gain, they would not necessarily need to shift biomass allocation to do so. Apparently it is a less effective means of doing so as compared to other alternatives. 2. Shifts in allocation, morphology, architecture, chemistry, and metabolism have likely been selected for to simultaneously improve C and N balance and survival, not just to improve C and N balance. For instance, responses to an increase from a very low to a low level of a limiting resource are often inconsistent with optimality arguments. Almost all shifts, such as

Root–Shoot Relations

decreased SLA (and hence decreased LAR) in higher but still limiting light or CO2 conditions, should in theory reduce C gain and RGR (Lambers and Poorter, 1992) when compared to a plant that had not shifted SLA or LAR. 3. Traits that enhance resource gain do not necessarily enhance survival and vice versa, especially in resource-poor and/or disturbed habitats (Walters and Reich, 1996, 2000a,b). Thus, shifts that have been selected for to increase survival may in fact work against a shift toward enhanced resource status. Only a whole-plant model that simultaneously examines shifts in allocation, morphology, chemistry, metabolism, turnover, and architecture, and their interaction, can characterize such changes. Unfortunately, it is easier to propose such a model than to develop it. III.

VARIATION IN ROOT–SHOOT RELATIONSHIPS AMONG SPECIES

Species of differing resource environments have been hypothesized to differ in their root–shoot relations (Grime, 1979; Chapin, 1980; Tilman, 1988; Gleeson and Tilman, 1990), especially for resources that either are soil based, such as nutrients and water, or aboveground based, such as light and CO2. In essence, if it is advantageous to be rootier as an individual when soil resources are poor, all species should phenotypically act this way, and species adapted to poor soils should be intrinsically rootier than those adapted to richer soils. But what do the data show? Species found in soil resource-poor habitats have been hypothesized to be inherently ‘‘rootier’’ than species from richer habitats (Chapin, 1980). A test of this idea requires comparison of individuals in comparable habitats. Several studies designed to address this question have had conflicting results (Poorter and Remkes, 1990; Garnier, 1991; Fichtner and Schultze, 1992; Van der Werf et al., 1993a,b), and no conclusion can be made in support of the hypothesis. Moreover, the best objects for such test would be of species in natural field settings. However, there are few if any such published data available. Moreover, no clear distinction in root morphological or chemical traits was apparent among species varying in fertility of habitat (cf. Ryser, 1996; Craine et al., 2001). Optimality theory also suggests that if it is advantageous to allocate proportionally more biomass to leaves when an individual plant is shaded, shadeadapted species should tend to be leafier than shade-

215

intolerant ones (Givnish, 1988). However, in comparing shade-tolerant versus intolerant species, recent results show opposite patterns vis-a`-vis biomass distribution, allocation and tissue traits as long-held hypotheses. Shade-tolerant species actually appear to have equal or higher RMF than intolerant species. This was true in a comparison of tropical woody species considered shade tolerant versus intolerant, (Veneklass and Poorter, 1998) and for winter deciduous temperate broadleaf species (Walters and Reich, 1999). A neutral pattern was observed for evergreen broad-leaved species (Walters and Reich 1999) and for our study species (Table 1; note that k is not different for tolerant vs. intolerant species). Hence, comparisons among species whose distribution is related to fertility or light gradients do not lend strong support for the optimal allocation or distribution theory. In contrast to the weak evidence for differences in biomass distribution among species in relation to habitat affinity, there is much stronger evidence of differences in tissue morphology, chemistry, metabolism, and turnover rates, increasingly of roots as well as shoots (Walters et al., 1993a,b; Ryser, 1996; Reich et al., 1998a,b; Walters and Reich, 2000b). Plants adapted to low fertility and/or low light may in fact be selected for low turnover of leaf and root tissues, rather than to high allocation to roots or shoots (Aerts, 1990; Reich et al., 1992, 1998a). They are also selected for traits associated with slow tissue turnover that minimize resource losses, such as dense tissues with low respiration rate. These include extended nutrient residence time via long tissue life span, for plants adapted to infertile habitats (Aerts, 1990). They also maintain low carbon losses via the combination of low respiration rates of all tissue types, and low tissue turnover rates, for plants adapted to deep shade (Walters and Reich, 1999, 2000a,b). Generally, species from resource-rich microhabitats have higher concentrations of nutrients, faster photosynthetic and respiration rates, higher SRL and SLA, and faster turnover of leaves than those from resource-poor environments (Chapin, 1980; Reich et al., 1992, 1998a,b, 1999). However, the comparative multispecies database for roots is still weak. Thus, it is uncertain whether root turnover rates are also higher for species from resource-rich microhabitats (Eissenstat and Yanai, 1997). The importance of leaf and root morphology and metabolism, rather than biomass allocation, was shown in contrasts of diverse species (Poorter and Remkes, 1990; Walters et al., 1993a; Reich et al., 1998a; Cornellissen et al., 1996) and among populations within a species (Ryser and Aeschlimann, 1999).

216

Reich

V. PLANT COMMUNITIES AND PLANTS IN COMMUNITIES Field studies that include whole-plant biomass data are rare. The root–shoot distributions of five conifer species planted in the field did not differ in trenched plots with higher nitrogen and water supply than in untrenched plots (Machado et al., 2001), even for plants of common size. Similarly, naturally grown sugar maple seedlings growing across a soil fertility and soil moisture gradient did not differ in biomass fraction distributed to roots, even when accounting for plant size and light environment (Walters and Reich, 1997), but did differ in other important plant traits such at tissue %N. It is possible that modest natural gradients in soil resource supply have different, or inconsequential, effects, compared to the large gradients observed in experimental studies with very young plants. It is also likely that biomass fractions in older seedlings are markedly influenced by turnover rates of leaves and roots. Do stands or patches of plants behave similarly vis-a`-vis root–shoot relationships as individual plants? Answering this question is difficult since most accounts of individual plants involve small chamber-grown plants, whereas most studies of plant stands involve larger, older, field-grown plants. For forests, both observational and experimental studies show that plants growing under lower nutrient and water availability tend to have a greater RMF and a greater shift toward fine-root than toward foliage production, as compared with plants of high nutrient and water availability (Linder and Axelsson, 1982; Gower et al., 1992). These patterns could result either from shifts toward greater allocation to roots at low resource supply, from lower fineroot turnover rates, or from both. The relative contribution of each is practically impossible to calculate, given the paucity of root turnover data. Across a N gradient that resulted from 32 years of variable experimental fire frequency regimes, a strong gradient in root:foliage fraction was found (Fig. 6; Reich et al., 2001). Oak savanna stands under chronically low N supply have a greater fraction of fine biomass in fine roots rather than in foliage. This pattern is consistent with optimality theory, but it is impossible to know whether it results from variation in dominance of one vegetation type over another, since the low-fertility, frequently burned end of the continuum was dominated more by grasses than trees. Moreover, again we cannot separate allocation from turnover rates in this instance.

Figure 6 The fraction of total fine root plus leaf mass distributed to fine roots in relation to net nitrogen mineralization rate for 14 oak savanna stands in a fire frequency experiment in Minnesota. Along with the variation in nitrogen mineralization rate are parallel gradients in fire frequency and relative tree versus grass dominance, with frequently burned stands being dominated more by grasses than trees and having low mineralization rates, and vice versa. (From Reich et al., 2001.)

At a very coarse, but very large scale, Jackson et al. (1996) summarized existing data regarding root and shoot biomass for different biomes of the world. Even segregating woody from nonwoody systems (because these two classes vary in root–shoot ratio primarily due to the differences in aboveground woody biomass), there were substantial differences among biomes. For forests, there were differences among boreal (0.32), temperate coniferous (0.18), temperate deciduous (0.23), tropical deciduous (0.34), and tropical evergreen (0.19) forests, but these do not provide any discernible pattern. This could be a result of the confounding of the myriad of differing environmental factors across such broad gradients. One way to separate out genetic from environmental effects would be to compare different species or populations in common gardens, or the same genotype in plantations across environmental gradients. Some such data are available for Scots pine populations, wherein those populations from colder environments show greater biomass fractions in roots, both as stands of trees (Oleksyn et al., 1999) and as individual seedlings. Across this climate of origin gradient, there are strong gradients in temperature, nutrient supply, length of growing season, and day length, so discerning the most responsible factors for such patterns is difficult.

Root–Shoot Relations

VI.

SUMMARY

It is not clear why young plants phenotypically shift biomass allocation in response to gradients in N supply in a manner consistent with optimality theory, fail to do so consistently for light or CO2, and show mixed patterns for P and water. This may not even be an important question to consider in isolation. It may well be that plants benefit more from shifting morphology, phenology, metabolism, chemistry, tissue longevity, or architecture in response to such factors than from shifting allocation. Only holistic consideration at the whole-plant level can answer such questions. Knowledge of ontogenetic drift in root–shoot biomass distribution with plant size and age is not at all new. There are now a considerable number of publications calling attention to the possible misinterpretation of root–shoot ‘‘snapshots’’ in time. Nonetheless, despite systematic tests of the role of ontogenetic drift, these have failed to alert plant scientists sufficiently to these patterns. Instead, many authors continue to refer to differences in root–shoot relations as differences in plasticity of allocation that may well be largely the result of ontogenetic drift (e.g., Walters and Reich, 1989; Conroy et al., 1992; Latham, 1992; Graves, 1994; Lusk et al., 1997; Lei and Lechowicz, 1998; Messier et al., 1999; Valladares et al., 2000). For instance, root–shoot ratio was reported to be significantly higher in bean plants under low than moderate P supply and higher in nonmycorrhizal than mycorrhizal plants, based on comparisons of plants at common harvest times (Nielsen et al., 1998). It is difficult to reconcile results of such ‘‘snapshots’’ with those obtained by allometric relationships (Fig. 4) showing that at a common plant size, low P plants had fewer roots, not more. The point of such a specific example for P (Fig. 4), plus an example for CO2 (Fig. 5), is to provide graphic examples of a persistent problem of inadequate evaluation of whether shifts in biomass allocation are real, or whether we often ‘‘see’’ what we have been trained to ‘‘see.’’ Similarly, many elegant studies have used sophisticated techniques to study the mechanisms involved in the control of allocation, as influenced by resource supply. However, their results may teach us about how plants developmentally adjust biomass allocation as they grow larger and older, rather than teaching us about how resource supply influences allocational processes. Only careful scrutiny of all such studies can help us decide when the right interpretation was made. Changes in allocation patterns are relatively strong when nutrient supply is varied, but changes in other

217

aspects of plant morphology or physiology more generally dominate the responses of plants to variation in a wide set of resources, including light, nutrients, water, and CO2. In essence, rather than adjust allocation of biomass, plants change tissue morphology, chemistry, metabolism, and turnover to alter their capability for resource interception, and to vary the rate of resource capture or resource loss. To the best of our knowledge, evidence for optimality in biomass allocation is even scarcer when examining differences among species that vary in habitat association, along either fertility or light gradients. Again, shifts in other attributes that may positively or negatively affect rates of resource capture or resource conservation appear to be far more important. Hence, it seems timely to suggest the removal of the significance of biomass allocation from its historically premier position, when discussing both how plants respond phenotypically to their environment and how plants have arrived through natural selection at their intrinsic genotypic traits. REFERENCES Aerts R. 1990. Nutrient use efficiency in evergreen and deciduous species from heathlands. Oecologia 84:391–397. Baxter R, Ashenden TA, Farrar JF. 1997. Effect of elevated CO2 and nutrient status on growth, dry matter partitioning and nutrient content of Poa alpina var. vivipara L. J Exp Bot 48:1477–1486. Bloom AJ, Chapin FS, Mooney HA. 1985. Resource limitations in plants—an economic analogy. Annu Rev Ecol Syst 16:363–392. Brouwer R. 1962a. Distribution of dry matter in the plant. Neth J Agric Sci 10:361–376. Brouwer R. 1962b. Nutritive influences on the distribution of dry matter in the plant. Neth J Agric Sci 10:399–408. Brouwer R. 1983. Functional equilibrium: sense or nonsense? Neth J Agric Sci 31:335–348. Callaway RM. 1992. Morphological and physiological responses of three California oak species to shade. Int J Plant Sci 153:434–441. Caloin M, Cle´ment B, Herrmann S. 1991. Regrowth kinetics of Dactylis glomerata following root excision. Ann Bot 68:435–440. Casper BB, Cahill JF Jr, Hyatt LA. 1998. Above-ground competition does not alter biomass allocated to roots in Abutilon theophrasti. New Phytol 140:231–238. Ceulemans R, Mousseau M. 1994. Effects of elevated atmospheric CO2 on woody plants. New Phytol 127:425– 446. Chapin FS III. 1980. The mineral nutrition of wild plants. Annu Rev Ecol Syst 11:233–260. Chapin FS III. 1991. Effects of multiple stresses on nutrient availability and use. In: Mooney HA, Winner WE, Pell

218 EJ, eds. Response of Plans to Multiple Stresses. San Diego, CA: Academic Press, pp 67–88. Chen HYH. 1997. Interspecific responses of planted seedlings to light availability in interior British Columbia: survival, growth, allometric, and specific leaf area. Can J For Res 27:1383–1393. Chen HYH, Klinka K. 1998. Survival, growth, and allometry of planted Larix occidentalis seedlings in relation to light availability. For Ecol Manag 106:169–179. Coleman JS, McConnaughay KDM, Ackerly DD. 1994. Interpreting phenotypic variation in plants. Trends Ecol Evol 9:187–191. Conroy JP, Milham PJ, Barlow EWR. 1992. Effect of nitrogen and phosphorus availability on the growth of Eucalyptus gradis to high CO2. Plant Cell Environ 15:843–847. Cornelissen JHC, Castro-Diez P, Hunt R. 1996. Seedling growth, allocation and leaf attributes in a wide range of woody plant species and types. J Ecol 84:755–765. Corre WJ. 1983. Growth and morphogenesis of sun and shade plants. I. The influence of light intensity. Acta Bot Neerl 32:49–62. Craine JM, Tilman DG, Wedin DA, Reich PB, Tjoelker MJ, Knops JMH. 2001. The relationship between plant functional strategies and growth in low-nutrient habitats. (Submitted to J Veg Sci.) Curtis PS, Wang X. 1998. A meta-analysis of elevated CO2 effects on woody plant mass, form, and physiology. Oecologia 113:299–313. Eamus D, Jarvis PG. 1989. Direct effects of CO2 increases on trees and forests (natural and commercial) in the UK. Adv Ecol Res 19:1–55. Eissenstat DM, Duncan LW. 1992. Root growth and carbohydrate responses in bearing citrus trees following partial canopy removal. Tree Physiol 10:245–257. Eissenstat DM, Yanai RD. 1997. The ecology of root lifespan. Adv Ecol Res 27:1–60. Eissenstat DM, Graham JH, Syvertsen JP, Drouillard DI. 1993. Carbon economy of sour orange in relation to mycorrhizal colonization and phosphorus status. Ann Bot 71:1–10. Evans GC. 1972. The Quantitative Analysis of Plant Growth. Berkeley, CA: University of California Press. Evans GC, Hughes AP. 1961. Plant growth and the aerial environment. I. Effect of artificial shading on Impatiens parviflora. New Phytol 60:150–180. Farrar JF, Gunn S. 1998. Allocation: allometry acclimation—and alchemy? In: Lambers H, Poorter H, Van Vuuren MMI, eds. Inherent Variation in Plant Growth. Physiological Mechanisms and Ecological Consequences. Leiden, Netherlands: Backhuys Publishers, pp 183–198. Fichtner K, Schulze E-D. 1992. The effect of nitrogen nutrition on growth and biomass partitioning of annual plants originating from habitats of different nitrogen availability. Oecologia 92:236–241.

Reich Fitter AH. 1985. Functional significance of root morphology and root system architecture. In: Fitter AH, Atkinson D, Read DJ, eds. Ecological Interactions in Soil, Plants, Microbes and Animals. Oxford, UK: Blackwell, pp 87–106. Garnier E. 1991. Resource capture, biomass allocation and growth in herbaceous plants. Trends Ecol Evol 6:126– 131. Garnier E. 1992. Growth analysis of congeneric annual and perennial grass species. J Ecol 80:665–675. Garnier E. 1998. Resource capture, biomass allocation and growth in herbaceous plants. Tree 6:126–131. Gedroc JJ, McConnaughay KDM, Coleman JS. 1996. Plasticity in root/shoot partitioning: optimal, ontogenetic, or both? Funct Ecol 10:44–50. Givnish TJ. 1988. Adaptation to sun and shade: a whole plant perspective. Aust J Plant Physiol 15:63–92. Gleeson SK, Tilman D. 1990. Allocation and the transient dynamics of succession on poor soils. Ecology 71:1144–1155. Gower ST, Vogt KA, Grier CC. 1992. Carbon dynamics of Rocky Mountain Douglas-fir: influence of water and nutrient availability. Ecol Monogr 62:43–65. Graves WR. 1994. Seedling development of sugar maple and black maple irrigated at various frequencies. HortScience 29:1292–1294. Grime JP. 1979. Plant Strategies and Vegetation Processes. Chichester, UK: Wiley. Gunn S, Bailey SJ, Farrar JF. 1999. Partitioning of dry mass and leaf area within plants of three species grown at elevated CO2. Funct Ecol 13 (suppl 1):3–11. Hughes AP. 1965. Plant growth and the aerial environment. IX. A synopsis of the autecology of Impatiens parviflora. New Phytol 64:399–413. Hughes AP, Evans GC. 1962. Plant growth and the aerial environment. II. Effects of light intensity on Impatiens parviflora. New Phytol 61:154–174. Ingestad T, Agren GI. 1991. The influence of plant nutrition on biomass allocation. Ecol Appl 1:168–174. Jackson RB, Canadell J, Ehleringer JR, Mooney HA, Sala OE, Schulze E-D. 1996. A global analysis of root distributions for terrestrial biomes. Oecologia 108:389– 411. Kruger EL, Reich PB. 1993. Coppicing affects growth, root:shoot relations and ecophysiology of potted Quercus rubra seedlings. Physiol Plant 89:751–760. Lambers H, Poorter H. 1992. Inherent variation in growth rate between higher plants: a search for physiological causes and ecological consequences. Adv Ecol Res 22:187–261. Latham RE. 1992. Co-occurring tree species change rank in seedling performance with resources varied experimentally. Ecology 73:2129–2144. Ledig, FT, Bormann FH, Wenger KF. 1970. The distribution of dry matter growth between shoot and roots in loblolly pines. Bot Gaz 13:349–359.

Root–Shoot Relations Lei TT, Lechowicz MJ. 1998. Diverse responses of maple saplings to forest light regimes. Ann Bot 82:9–19. Lerdau M. 1992. Future discounts and resource allocation in plants. Funct Ecol 6:371–375. Lerdau M, Gershenzon J. 1997. Allocation theory and the costs of chemical defenses in plants. In: Bazzaz F, Grace J, eds. Resource Allocation in Plants and Animals. San Diego, CA: Academic Press, pp 265– 277. Linder S, Axelsson B. 1982. Changes in carbon uptake and allocation patterns as a result of irrigation and fertilization in a young Pinus sylvestris stand. In: Waring RH, ed. Carbon Uptake and Allocations: Key to Management of Subalpine Forest Ecosystems, International Union Forest Research Organization (IUFRO) Workshop. Corvallis, OR: Forest Research Laboratory, Oregon State University. Lusk CH, Contreras O, Figueroa J. 1997. Growth, biomass allocation and plant nitrogen concentration in Chilean temperate rainforest tree seedlings: effects of nutrient availability. Oecologia 109:49–58. Lutze JL, Gifford RM. 1998. Acquisition and allocation of carbon and nitrogen by Danthonia richardsonii in response to restricted nitrogen supply and CO2 enrichment. Plant Cell Environ 21:1133–1141. Machado J-L M, Walters MB, Reich PB. 2001. In deeply shaded forest understories, belowground resources limit seedling growth but do not alter biomass distribution patterns or survival. (Submitted to Forest Ecology and Management.) McConnaughay KDM, Coleman JS. 1999. Biomass allocation in plants: ontogeny or optimality? A test along three resource gradients. Ecology 80:2581–2593. McMillan JD, Wagner MR. 1995. Effects of water stress on biomass partitioning of ponderosa pine seedlings during primary root growth and shoot growth periods. For Sci 41:594–610. Messier C, Doucet R, Ruel J-C, Claveau Y, Kelly C, Lechowicz MJ. 1999. Functional ecology of advance regeneration in relation to light in boreal forests. Can J For Res 29:812–823. Nielsen KL, Bouma TJ, Lynch JP, Eissenstat DM. 1998. Effects of phosphorus availability and vesicular-arbuscular mycorrhizas on the carbon budget of common bean (Phseolus vulgaris). New Phytol 139:647–656. Oleksyn J, Reich PB, Chalupka W, Tjoelker MG. 1999. Differential above- and belowground biomass accumulation of European Pinus sylvestris populations in a 12year-old provenance experiment. Scand J For Res 14:7–17. Olff H, Van Andel J, Bakker JP. 1990. Biomass and shoot/ root allocation of five species from a grassland succession series at different combinations of light and nutrient supply. Funct Ecol 4:193–200.

219 Peace WJH, Grubb PJ. 1982. Interaction of light and mineral nutrient supply in the growth of Impatiens parviflora. New Phytol 90:127–150. Pearsall WH. 1927. Growth studies VI. On the relative sizes of growing plant organs. Am J Bot 41:549–556. Philippot S, Allirand JM, Chartier M, Gosse G. 1991. The role of different daily irradiations on shoot growth and root/shoot ratio in Lucerne (Medicago sativa L.). Ann Bot 68:329–335. Poorter L. 1999. Growth responses of 15 rain-forest tree species to a light gradient: the relative importance of morphological and physiological traits. Funct Ecol 13:396–410. Poorter H, Nagel O. 2000. The role of biomass allocation in the growth response to plants to different levels of light, CO2, nutrients and water: a quantitative review. Aust J Plant Physiol 27:595–607. Poorter H, Remkes C. 1990. Leaf area ratio and net assimilation rate of 24 wild species differing in relative growth rate. Oecologia 83:553–559. Poorter H, Remkes C, Lambers H. 1990. Carbon and nitrogen economy of 24 wild species differing in relative growth rate. Plant Physiol 94:621–627. Reich, PB, Schoettle AW, Stroo HF, Troiano J, Amundson RG. 1987. Influence of O3 and acid rain on white pine seedlings grown in five soils. I. Net photosynthesis and growth. Can J Bot 65:977–987. Reich PB, Walters MB, Ellsworth DS. 1992. Leaf lifespan in relation to leaf, plant and stand characteristics among diverse ecosystems. Ecol Monogr 62:365–392. Reich PB, Walters MB, Krause SC, Vanderklein D, Raffa KF. 1993. Growth, nutrition and gas exchange of Pinus resinosa following artificial defoliation. Trees 7:67–77. Reich PB, Tjoelker MG, Walters MB, Vanderklein D, Buschena C. 1998a. Close association of RGR, leaf and root morphology, seed mass and shade tolerance in seedlings of nine boreal tree species grown in high and low light. Funct Ecol 12:327–338. Reich PB, Walters MB, Tjoelker MG, Vanderklein D, Buschena C. 1998b. Photosynthesis and respiration rates depend on leaf and root morphology and nitrogen concentration in nine boreal tree species differing in relative growth rate. Funct Ecol 12:395–405. Reich PB, Ellsworth DS, Walters MB, Vose J, Gresham C, Volin J, Bowman W. 1999. Generality of leaf traits relationships: a test across six biomes. Ecology 80:1955–1969. Reich PB, Peterson DA, Wrage K, Wedin D. 2001. Fire and vegetation effects on productivity and nitrogen cycling across a forest-grassland continuum. Ecology (in press). Reynolds HL, D’Antonio C. 1996. The ecological significance of plasticity in root weight ratio in response to nitrogen: opinion. Plant Soil 185:75–97.

220 Reynolds JF, Thornley JHM. 1982. A shoot:root partitioning model. Ann Bot 49:585–597. Rice SA, Bazzaz FA. 1989. Quantification of plasticity of plant traits in response to light intensity: comparing phenotypes at a common weight. Oecologia 78:502– 507. Robinson D. 1986. Compensatory changes in the partitioning of dry matter in relation to nitrogen uptake and optimal variations in growth. Ann Bot 86:841–848. Robinson D, Rorison IH. 1988. Plasticity in grass species in relation to nitrogen supply. Funct Ecol 2:249–257. Ryser P. 1996. The importance of tissue density for growth and life span of leaves and roots: a comparison of five ecologically contrasting grasses. Funct Ecol 10:717– 723. Ryser P, Aeschlimann U. 1999. Proportional dry-mass content as an underlying trait for the variation in relative growth rate among 22 Eurasian populations of Dactylis glomerata s.l. Funct Ecol 13:473–482. Ryser P, Lambers H. 1995. Root and leaf attributes accounting for the performance of fast- and slow-growing grasses at different nutrient supply. Plant Soil 170:251–265. Steinbrenner EC, Rediske JH. 1964. Growth of ponderosa pine and Douglas-fir in a controlled environment. Weyerhauser Forest Paper No. 1. Centralia, WA: Weyerhauser Forestry Research Center. Stoneman GL, Dell B. 1993. Growth of Eucalyptus marginata (Jarrah) seedlings in a greenhouse in response to shade and soil temperature. Tree Physiol 13:239–252. Terry N. 1968. Developmental physiology of sugar beet. I. The influence of light and temperature on growth. J Exp Bot 61:795–811. Thornley JM. 1969. A model to describe the partitioning of photosynthate during vegetative plant growth. Ann Bot 33:419–430. Thornley JM. 1972. A balanced quantitative model for root:shoot ratios in vegetative plants. Ann Bot 36:431–441. Thornley JHM. 1998. Modelling shoot:root relations: the only way forward? Ann Bot 81:165–171. Tilman D. 1988. Plant Strategies and the Dynamics and Structure of Plant Communities. Princeton, NJ: Princeton University Press. Tjoelker MG, Oleksyn J, Reich PB. 1998. Temperature and ontogeny mediate growth response to elevated CO2 in seedlings of five boreal tree species. New Phytol 140:197–210. Tomlinson PT, Anderson PD. 1998. Ontogeny affects response of northern red oak seedlings to elevated CO2 and water stress. New Phytol 140:493–504. Valladares F, Wright SJ, Lasso E, Kitajima K, Pearcy RW. 2000. Plastic phenotypic response to light of 16 congeneric shrubs from a Panamanian rainforest. Ecology 81:1925–1936.

Reich Van der Werf A, Van Nuenen M, Visser AJ, Lambers H. 1993a. Contribution of physiological and morphological plant traits to species’ competitive ability at high and low nitrogen supply. Oecologia 94:434–440. Van der Werf A, Visser AJ, Schieving F, Lambers H. 1993b. Evidence for optimal partitioning of biomass and nitrogen at a range of nitrogen availabilities for a fast- and slow-growing species. Funct Ecol 7:63–74. Vanderklein DW, Reich PB. 1999. The effect of defoliation intensity and history on photosynthesis, growth and carbon reserves of two conifers with contrasting leaf lifespans and growth habits. New Phytol 144:121–132. Veneklaas EJ, Poorter L. 1998. Growth and carbon partitioning of tropical tree seedlings in contrasting light environments. In: Lambers H, Poorter H, Van Vuuren MMI, eds. Inherent Variation in Plant Growth. Leiden, Netherlands: Backhuys Publishers, pp 337–361. Volin JC, Reich PB. 1996. Interaction of carbon dioxide and ozone on C3 and C4 grasses and trees under contrasting nutrient supply. Physiol Plant 97:674–684. Walters MB, Reich PB. 1989. Response of Ulmus americana seedlings to varying nitrogen and water status. I. Photosynthesis and growth. Tree Physiol 5:159–172. Walters MB, Reich PB. 1996. Are shade tolerance, survival, and growth linked? Low light and nitrogen effects on hardwood seedlings. Ecology 77:841–853. Walters MB, Reich PB. 1997. Growth of Acer saccharum seedlings in deeply shaded understories of northern Wisconsin: effects of nitrogen and water availability. Can J For Res 27:237–247. Walters MB, Reich PB. 1999. Low light carbon balance and shade tolerance in the seedlings of woody plants: do winter deciduous and broad-leaved evergreen species differ? New Phytol 143:143–154. Walters MB, Reich PB. 2000a. Seed size, nitrogen supply and growth rate affect tree seedling survival in deep shade. Ecology 81:1887–1901. Walters MB, Reich PB. 2000b.Trade-offs in low-light CO2 exchange: a component of variation in shade tolerance among cold temperate tree seedlings. Funct Ecol 14:155–165. Walters MB, Kruger EL, Reich PB. 1993a. Growth, biomass distribution and CO2 exchange of northern hardwood seedlings in high and low light: relationships with successional status and shade tolerance. Oecologia 94:7– 16. Walters MB, Kruger EL, Reich PB. 1993b. Relative growth rate in relation to physiological and morphological traits for northern hardwood seedlings: species, light environment and ontogenetic considerations. Oecologia 96:219–231. Wilson JB. 1988. A review of evidence on the control of shoot:root ratio, in relation to models. Ann Bot 61:433–449.

13 Root Life Span, Efficiency, and Turnover David M. Eissenstat The Pennsylvania State University, University Park, Pennsylvania

Ruth D. Yanai State University of New York, Syracuse, New York

I.

INTRODUCTION

than shoot competition (Wilson, 1988). Just as perennial structures aboveground can give plants a competitive advantage for light capture, there may be advantages to long-lived roots in the capture of limited soil resources. Resource preemption can be an important component of competitive success. For example, in climates with winter precipitation, perennial grasses with an established root system are much more effective than seedlings of perennial grasses at competing with annual grass species during the spring and summer (Harris, 1967). Clearly, root demography can have important consequences on species distribution and abundance. The demography of roots also influences ecosystem processes associated with material and energy flows. Approximately 33% of global net primary production is used for fine-root production, based on fine-root biomass in 253 field studies in a wide range of ecosystems and assuming roots have a life span of 1 year, possibly a conservative estimate (Jackson et al., 1997). In other studies, belowground net primary productivity (BNPP) has been estimated to be at least as great as aboveground net primary productivity (Vogt et al., 1986; Caldwell, 1987). Clearly, the production and death of fine roots can have a substantial influence on ecosystem carbon and mineral nutrient cycling. Many ecologists have been concerned with understanding how BNPP varies among ecosystems and pre-

Like other plant organs, roots have a life history in which they pass from birth to death. The size and population structure of the root system is determined by the birth rate and death rate of the individual roots. The study of root demography is of interest to many disciplines, including crop science, physiology, ecology, and soil science. For example, a better understanding of root demography could enable agronomists and horticulturalists to increase yields while reducing agrochemical inputs. Severe root losses, such as those caused by drought or pathogens, clearly are not conducive to crop production. Growing too many roots, however, may also be undesirable, since large amounts of carbohydrates and mineral nutrients are needed for root growth and maintenance that otherwise might be allocated to photosynthetic organs or harvested parts. An optimization approach suggests that, other things being equal, total plant growth should be greatest when a root system maximizes water and nutrient acquisition per unit resource supplied from the shoot (e.g., Thornley, 1998). If roots are produced in the most favorable soil patches and shed when they are no longer efficient in water and nutrient absorption, then production, theoretically, should be maximized. The birth and death of roots also influence plant competition. Root competition can be more intense 221

222

Eissenstat and Yanai

dicting how it may change in response to tropospheric ozone concentrations (Coleman et al., 1996), nitrogen deposition (Nadelhoffer, 2000), temperature (Gill and Jackson, 2000; Pregitzer et al., 2000a), drought (Joslin et al., 2000) and elevated CO2 (Arnone et al., 2000; Tingey et al., 2000). We need a more mechanistic understanding of factors controlling root longevity if BNPP is to be incorporated into models of ecosystem response to global climate change (Norby and Jackson, 2000; Jackson et al., 2000). The first roots of plants developing from seed are indeterminate, typically extending greatly in length as the taproot or other seminal roots develop. The major laterals that first emerge from these primary roots and the adventitious or nodal roots that emerge from the stem base are also typically indeterminate, often extending decimeters or more in length. These indeterminate roots form the basic framework of the root system and may live as long as the plant lives. This chapter focuses on the more ephemeral portion of the root system. Ephemeral roots are the fine laterals that may be replaced several times during a growing season and may have only a few orders of branching. In at least some woody species, these roots never undergo secondary development of the stelar tissue or the development of a periderm (Brundrett and Kendrick, 1988; Eissenstat and Achor, 1999). In this chapter, we examine variations in root life span and causes for this variation. We discuss different methods of assessing root life span and root turnover. We describe a cost-benefit model of root deployment, which defines the root life span that maximizes the efficiency of resource acquisition. We review studies that have examined biotic and abiotic factors that influence root life span in the context of our hypothesis that plants modulate root life span to maximize root efficiency. Finally, we extend the model of individual root efficiency to describe a cohort of roots with a median life span, and we include allocation to defense in defining the optimal root life span. II.

VARIATION IN ROOT LIFE SPAN

A.

Sources of Variation

Estimates of root life span vary widely. The median life span of the finest roots can range from 1 year in slow-growing forest trees, according to studies using transparent windows in the soil (Eissenstat and Yanai, 1997). In a data set containing 190 studies in nonagricultural ecosystems, based mainly on changes

in biomass from sampling soil monoliths, soil cores or ingrowth cores, average root life spans ranged from
290 d in tropical ecosystems to
3 years in highlatitude ecosystems, with considerable variation within each ecosystem type (Gill and Jackson, 2000). Recent studies based on tracer approaches have indicated that fine roots may live considerably longer—averaging 4–8 yr in some temperate forests (Matamala et al., 2000; Gaudinski et al., 2000). Although differences in methods contribute substantially to differences in estimates of life span, as we will discuss, undoubtedly much of the variation in reported root life span is caused by differences in environmental conditions and plant species. B.

Patterns of Variation Among Species

It is difficult to assess the relative importance of genetic and environmental variation on root life span. Few studies have tracked individual roots of more than one species under the same environmental conditions. In a greenhouse study of seedlings of four tree species, root life span varied from 26 d in Prunus avium to 86 d in Picea sitchensis (Black et al., 1998). In a Valencia orange citrus rootstock trial in central Florida, we measured a median root life span of 90 d in Poncirus trifoliata and 152 d in Citrus volkameriana. Weaver and Zink (1946) banded individual nodal roots of perennial range and pasture grasses. After 3 years, root survival ranged from 45% in Bouteloua gracilis to 10% in Stipa spartea. The fine laterals of the nodal roots presumably had shorter life spans, but they could not be followed with this approach. The same theories that attempt to explain variation in leaf life span have been applied to roots (Grime, 1977; Chapin, 1980; Aerts, 1995). Plants that have slow growth rates and are adapted to chronically low-nutrient sites, for example, should have long life span of the absorptive organs compared to more fertile sites. Tissue retention in nutrient-poor sites allows nutrients to be retained as well, which is important if root and shoot growth rates are restricted by nutrient limitations. There is considerable evidence that leaf longevity is consistent with this hypothesis (Reich et al., 1997), but roots have been less well studied. In the pot study of Black et al. (1998), the species with the shortest root life span, Prunus avium, had considerably faster growing root and shoot systems than the species with longest root life span, Picea sitchensis. In a study comparing grasses from nutrient-poor and nutrient-rich habitats in pots in the field, the grasses from the nutrient-rich habitat had lost a greater

Life Span, Efficiency, and Turnover

percentage of their leaves and roots by the end of the second growing season (Ryser, 1996). Roots also lived longer in species adapted to more infertile soils among trees in mixed hardwood forests in Wisconsin (estimated by nitrogen budgeting; Aber et al., 1985; Nadelhoffer et al., 1985) and among heathland shrubs and grasses (estimated by minirhizotron and soil core sampling; Aerts et al., 1989, 1992). These results suggest that roots and leaves do have similar adaptations of longevity to resource availability. There are notable exceptions to this generalization, however. Desert succulents have long-lived leaves but short-lived ‘‘rain’’ roots (Huang and Nobel, 1992; North et al., 1993; see also Chapter 53 by Nobel in this volume). In seasonal dry climates, cluster roots of evergreen woody plants (Lamont, 1995) and ericoid mycorrhizal root hairs of plants in the Epacridaceae (Smith and Read, 1997) are shed during extended dry periods. Generalizations about the relationship of tissue longevity to resource availability may apply better to nutrients than water. Long leaf life span has been associated with other leaf traits, including low specific leaf area (area/mass ratio), N concentration, maximum assimilation rate, high leaf thickness, toughness, lignin content, and tissue density (Reich et al., 1997). Similar suites of correlated traits may also occur in roots (Eissenstat, 1992; Reich et al., 1998), but the scarcity of observations makes patterns more difficult to detect. One such study by Ryser (1996) found higher tissue density in grasses with longer-lived roots. Similarly, in a comparison of apple and citrus, long root life span was associated with coarse root diameter and high tissue density (Eissenstat et al., 2000), low maintenance respiration, and a low P uptake capacity (Bouma et al., 2001). III.

METHODOLOGICAL CONSIDERATIONS

A.

Difficulties and Definitions

The single greatest impediment to the study of root life span is the difficulty of studying roots in their natural environment. Many approaches have been taken with varying degrees of success. Often studies are not long enough to establish clear year-to-year variation or to have allowed the plants to fully adjust to installation of root measuring devices (e.g., minirhizotrons) or treatments. For example, fertilization studies are often conducted for only a few years, so they may not characterize steady-state responses to a new level of fertility. The interpretation of estimates of root life span is hindered by inconsistencies in methods of reporting

223

root dynamics. In many ecosystem studies, the main objective is to estimate BNPP (kg ha1 yr1). The term ‘‘root turnover’’ has often been used synonymously with annual root production or annual root mortality and thus has units such as kg ha1 yr1. Alternatively, root turnover may be used to describe the specific rate of root mortality, in units of yr1. One way to report the specific rate of root mortality is the rate constant in exponential decay (described in Section VII, below). Root turnover rate is also commonly reported as annual root production or annual root mortality divided by root standing crop. Studies differ in whether minimum (Hendrick and Pregitzer, 1993), average (Aber et al., 1985; Aerts et al., 1992), or maximum (Dahlman and Kucera, 1965; Gill and Jackson, 2000) root biomass are used to estimate standing crop. Gill and Jackson (2000) found that about onethird of root turnover studies report only the mean standing crop. An important disadvantage of using minimum or maximum standing crop is that the minimum or maximum value in any distribution is dependent on the number of samples collected and the sampling error associated with sample measurement. Nonetheless, Gill and Jackson (2000) found that maximum standing crop could be accurately estimated by mean standing crop by a regression approach ðr2 ¼ :90Þ, based on 20 data sets that included both maximum and mean root biomass. Root life span is inversely proportional to root turnover rate, with the constant of proportionality dependent on the definitions of turnover rate and life span. Many recent studies that follow the fate of individual roots with minirhizotrons report only median life span (or similarly half-life of the cohort), partly because many of the roots in the study have not died by the end of the study and partly because the median is a better estimator of the central location of a highly skewed distribution—a condition common to survivorship curves. Clearly, average life span may be considerably longer than median life span if an appreciable fraction of the population lives a very long time. Studies that follow individual roots typically report median life spans of specific cohorts (roots born at the same point in time), because different cohorts may exhibit very different median life spans (Kosola et al., 1995). B.

Methods of Estimating Root Life Span

Early techniques estimated root turnover at ecosystem scales by measuring average standing crop and seasonal root production using sequential coring, root

224

ingrowth cores, and elemental budgeting (see reviews by Caldwell and Eissenstat, 1987; Vogt and Persson, 1991; Fahey et al., 1999). More recently, studies have used minirhizotrons (Cheng et al., 1990; Hendrick and Pregitzer, 1993) and other direct observational techniques (Fahey and Hughes, 1994), which focus on the fate of individual roots. In addition, tracer techniques hold considerable promise as an independent estimator of root longevity (Gaudinski et al., 2000; Matamala et al., 2000). No method of estimating root turnover has emerged as the best for all conditions. In the 1970s and ’80s, the most popular approach to estimating ecosystem BNPP was sequential coring. This approach involves collecting soil cores over the growing season (often monthly) and estimating BNPP based on changes in the mass of live and dead roots (e.g., Vogt et al., 1981). The advantages of sequential coring are that the roots being measured have not been altered in any way prior to coring, estimates can be scaled up to the ecosystem, and equipment costs are low. The labor required to separate roots from the soil core and to separate live from dead roots, however, is considerable (Bloomfield et al., 1996). The very finest roots, which may be very fragile, are probably never completely separated from the soil. Another limitation of this method is a lack of information on turnover of deeper roots; cores are commonly collected only to 20 cm depth. There are also several sources of error in the calculations, which involve the differences between cores collected over time. Simultaneous birth and death of roots during a single sampling interval is not detected (Rytter, 1999). The very finest roots probably die within weeks, not months (Wells and Eissenstat, 2001). It is also difficult to separate spatial and sampling variation in root mass from the parameter of interest, temporal variation (Singh et al., 1984; Sala et al., 1988). Typically, soil-coring or soil monolith methods are used to estimate annual root production, and a steady-state assumption is required to equate annual root mortality with annual root production. Root turnover (yr1) is obtained by dividing production (kg ha1 yr1) by some estimate of standing crop (kg ha1), which can introduce further errors. Various approaches have been used to improve biomass-based estimates using compartment-flow models (Santantonio and Grace, 1987; Ma¨kela¨ and Vanninen, 2000), but these methods require accurate information on fine-root decomposition, which is difficult to acquire (Fahey et al., 1999), especially for the very finest roots (Comas et al., 2000; Wells and Eissenstat, 2001).

Eissenstat and Yanai

The ingrowth-core technique for estimating root production and turnover assumes that root production in a soil volume initially devoid of roots reflects root production in the undisturbed soil (Fabia˜o et al., 1985; Fine´r et al., 1997). It also assumes that no root mortality (or if dead roots are followed, no root decomposition) has occurred during a sampling interval. Like sequential coring, this method of estimating root mortality assumes steady-state conditions and requires an estimate of root standing crop. Ingrowth-core approaches are relatively inexpensive, requiring considerably less labor than either minirhizotron or sequential coring methods, because they do not involve distinguishing live roots from dead ones or removing them from soil organic matter. The biggest drawback is their artificiality. Soil disturbance can increase water and nutrient availability by increasing decomposition and reducing root competition (Eissenstat, 1991). Soil disturbance may also favor root growth by decreasing soil bulk density and impedance. For these reasons, root growth may be considerably higher in small volumes of disturbed soil than in surrounding undisturbed soil, biasing estimates of production and turnover. Various elemental budgeting techniques have been used to assess root turnover (Nadelhoffer, 2000). The nitrogen budgeting approach estimates fine-root production as the difference between annual net mineralization and net N uptake into aboveground production (Aber et al., 1985; Nadelhoffer et al., 1985). This method depends on the accuracy of the estimate of net N mineralization, which is usually based on in situ soil incubations, and assumes that there is no change in ecosystem storage of mineralized N. To estimate root turnover or root longevity further requires estimating the standing crop of roots and assuming a steady state, as for the other methods described above. The budgeting techniques are also subject to errors in other fluxes such as N deposition, denitrification, and leaching. Another budgeting approach uses soil carbon fluxes to estimate root turnover. Soil respiration less root respiration and aboveground litterfall should equal fine root production (Raich and Nadelhoffer, 1989; Nadelhoffer and Raich, 1992; Haynes and Gower, 1995). This method depends on the accuracy of the estimates of soil and root respiration. It also assumes that soil organic matter is at steady state, unless the rate of change can be estimated (see Chapter 40 by Cramer in this volume). Tracers of C and N have been used to estimate root turnover, typically by calculating the dilution of the

Life Span, Efficiency, and Turnover

tracer in the structural tissues at various intervals after labeling (Caldwell and Camp, 1974; Milchunas and Lauenroth, 1992; Hendricks et al., 1997). The chief problems with this approach have been achieving uniform labeling of the structural tissue of the fine roots, estimating turnover rates of the very finest roots, which may be more rapid than the sampling intervals, and, for C, labeling whole trees. Recently, 13C in free-air CO2 exposure (FACE) experiments (Matamala et al., 2000) and the spike in atmospheric 14C caused by bomb-testing in the 1950s (Gaudinski et al., 2000) have been used to provide estimates of root longevity for large trees. There are some additional techniques that allow for the examination of factors influencing root demography. Tagging roots (Weaver and Zink, 1946) and following tillers of known age and root number (Shaver and Billings, 1975; Brundrett and Kendrick, 1988) permit estimation of the life span only of the major nodal roots, not the fine laterals. Root screens (Fahey and Hughes, 1994) can be useful for estimating the longevity of fine roots that form a readily accessible root mat. The most versatile technique for the direct observation of root demography is to track roots growing against transparent windows. Large root observation windows, referred to as rhizotrons, were initially used to study root phenology (seasonal patterns of root growth), including root mortality, in relation to shoot phenology (Head, 1973). Rapid progress in our understanding of root demography has occurred with the development of minirhizotrons (transparent tubes typically 2–6 cm in diameter), which allow roots to be observed in diverse ecosystems with minimal disturbance and a reasonable degree of replication (Taylor, 1987; Fahey et al., 1999; see Chapter 18 by Polomski and Kuhn in this volume). This technique suffered in its early years from limitations in the quality of images and the amount of labor required to process thousands of root images. In the late 1990s, improvements in miniature cameras or borescopes, direct digital capture of images, fast low-cost computers with greater storage capacity, and more sophisticated statistical approaches have made this technique more powerful and accessible. Minirhizotron studies have provided detailed information about root life span, such as age-specific mortality rates, mortality rates of roots born at different times of year or at different depths in the soil, mortality rates among roots of different orders or diameters, and effects of localized soil conditions on root mortality (Hendrick and Pregitzer, 1993; Ruess et al., 1998; Arnone et al., 2000; Wells and Eissenstat, 2001).

225

Despite the widespread use of transparent wall techniques, they, too, have disadvantages. Transparent walls create an unnatural environment that may affect root production and longevity (Samson and Sinclair, 1994; Joslin and Wolfe, 1999). They can be difficult to use in rocky soils, shrink-swell soils, and clays that smear the tube surface, although various modifications have been devised (Gijsman et al., 1991; Meyer and Barrs, 1991; Lopez et al., 1996; Phillips et al., 2000). The biggest limitations are the cost of the camera equipment and the still considerable labor required to process the large numbers of root images. C.

Influence of Root Diameter and Root Order

The reported variation in root life span is partly due to the imprecise definition of the classes of roots under study. ‘‘Fine’’ roots are typically defined by an arbitrary diameter limit. The diameter limits for tree roots are generally large (1–5 mm) relative to the very finest roots. These finest roots can have much shorter life spans than the larger-diameter roots, which are still considered part of the fine-root system. For example, the median life span of apple roots 0.1–0.2 mm in diameter was only
40 d, while the median life span of roots 0.5–1.1 mm in diameter was longer than the observation period (211–240 d, depending on the year; Wells and Eissenstat, 2001). In peach, roots

0.25 mm in diameter had a median life span of 77 d while not a single root in the 0.5–1.7-mm class (n = 45) had died by the end of the study (369 d; Wells et al., submitted). Root order, which describes the position of a root in the branching pattern, is also important to root life span. In sugar maple, among roots 23 mmol

Figure 5 The efficiency of a cohort of roots as a function of C expended for defense (CDef). The efficiency of the cohort is based on respiration and uptake rates of individual roots (see Fig. 1) and exponential decay at rates determined by the defensive C investment. The inset shows the assumed relationship between C expended for defense of the root and the resulting half-life of the cohort of roots, for scenarios of higher (High) and lower (Low) pressure from herbivores and pathogens.

234

P/mol C (Fig. 4). Under low pressure, where we assumed the median life span would be reduced to 35 d without an additional C investment, the optimal median life span was increased to 42 d by an additional C investment that resulted in 94% of the ideal maximum efficiency. With higher pressure, where we assumed the median life span would be only 14 d without the expenditure of C for defense, the optimal C investment was 25 mmol/g root, compared to only 5 mmol C/g root in the low-pressure case. This investment, according to our guess at the return on investment shown in Fig. 5 (inset), resulted in a median life span of 50 days, and a 10% loss in cohort efficiency, but a vast improvement over the efficiency achieved by the unprotected cohort. We do not expect the quantitative relationships illustrated here to apply to any real situations in nature. The relationship between defensive C investment and median life span (Fig. 5, inset), upon which the optimal investment in defense depends (Fig. 5), was not based on data from any specific case. This illustration, however, serves to highlight some useful concepts and some needs for future research, as follows. Applying a cost-benefit analysis to a cohort of roots with a distribution of life spans reveals that the optimal individual life span is not the same as the optimal median life span of the population. The relationship between the individual and population optima depends on the form of the survivorship curve or on its hazard function. Plant control over root life span may be indirect, as the direct causes of mortality, such as herbivory and disease, are largely external to the root. One indirect influence of the plant over root life span is investment in defense of the root. The optimal investment in defense of the root will depend on the magnitude of the imposed risk of mortality. The optimal investment in defense will be larger when the herbivore or pathogen pressure is high. Inducible defenses may help finetune the relationship between allocation to defense and herbivore or pathogen pressure. The maximal efficiency of the cohort will be reduced by the C expenditure for defense, compared to a situation without herbivore or pathogen pressure. Beyond the optimal expenditure, allocating additional C to root defense would increase life spans but reduce efficiencies, such that nutrient acquisition would be better served by constructing new roots than by defending old ones. The absence of data required to better quantify or test these relationships is due to the difficulty of collecting information on root demography, C allocation, and age-dependent C expenditures

Eissenstat and Yanai

and uptake rates of ephemeral roots under field conditions

VIII.

SUMMARY

The ability of a root system to forage efficiently for water and nutrients depends on the production and loss of individual roots in soil of spatially and temporally heterogeneous moisture and fertility and on the physiological activity of these roots, which changes with age. There is enormous variation in root life span, and sources for the variation are not well understood. We hypothesize that plant variation in root life span often relates to plant potential growth rate and the nutrient availability where the plant has evolved. More and wider species comparisons under common garden conditions are needed to test this hypothesis. One of the difficulties in generalizing about factors controlling root life span is the lack of agreement among methods. No method has emerged as best for all conditions, although the minirhizotron approach seems to hold the most promise for developing a better understanding of root demography under a wide range of conditions. Both abiotic and biotic factors affect root life span, and often these factors interact. Higher temperature, for example, may diminish root life span more by allowing for more root herbivores and pathogens than by directly affecting root maintenance costs. Reductions in available photosynthate for root maintenance, such as caused by grazing or pruning of the shoot or by high fruit production, often leads to greater root mortality. We previously approached the cost-benefit analysis of root life span with the implicit assumption that plants controlled the life span of roots. Our current approach acknowledges the role of exogenous factors in root mortality, with the plant having indirect control through allocation to defense. A cohort analysis of root efficiency allows a distribution of life spans to be optimized. Additional studies examining factors influencing root life span combined with optimization modeling are needed to unravel the numerous controls and constraints on the life span of plant roots.

REFERENCES Aber JD, Melillo JM, Nadelhoffer KJ, McClaugherty CA, Pastor J. 1985. Fine root turnover in forest ecosystems in relation to quantity and form of nitrogen availabil-

Life Span, Efficiency, and Turnover ity: a comparison of two methods. Oecologia 66:317– 321. Aerts R. 1995. The advantages of being evergreen. Trends Ecol Evol 10:402–407. Aerts R, Berendse F, Klerk NM, Bakker C. 1989. Root production and rot turnover in two dominant species of wet heathlands. Oecologia 81:374–378. Aerts R, Bakker C, De Caluwe H. 1992. Root turnover as determinant of the cycling of C, N and P in a dry heathland ecosystem. Biogeochemistry 15:175–190. Alexander IJ, Fairley RI. 1983. Effects of N fertilization on populations of fine roots and mycorrhizas in spruce humus. Plant Soil 71:49–53. Arnone III JA, Zaller JG, Spehn EM, Niklaus PA, Wells CE, Ko¨rner C. 2000. Dynamics of root systems in native grasslands: effects of elevated CO2. New Phytol 147:73–85. Benhamou N, Fortin JA, Hamel C, St-Arnaud M, Shatilla A. 1994. Resistance responses of mycorrhizal Ri T-DNAtransformed carrot roots to infection by Fusarium oxysporum f. sp. Chrysanthemi. Phytopathology 84:958– 968. Black KE, Harbron CG, Franklin M, Atkinson D, Hooker JE. 1998. Differences in root longevity of some tree species. Tree Physiol 18:259–293. Bloomfield J, Vogt K, Wargo PM. 1996. Tree root turnover and senescence. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. 2nd ed. New York; Marcel Dekker, pp 383–381. Bouma TJ, Yanai RD, Elkin A, Hartmond U, Flores DE, Eissenstat DM. 2001. Estimating age-dependent costs and benefits of roots with contrasting life span: comparing apples and oranges. New Phytol (in press). Brundrett MC, Kendrick B. 1988. The mycorrhizal status, root anatomy, and phenology of plants in a sugar maple forest. Can J Bot 66:1153–1173. Bryla DR, Bouma TJ, Eissenstat DM. 1997. Root respiration in citrus acclimates to temperature and slows during drought. Plant Cell Environ 20:1411–1420. Burton AJ, Pregitzer KS, Hendrick RL. 2000. Relationships between fine root dynamics and nitrogen availability in Michigan northern hardwood forests. Oecologia (in press). Caldwell MM. 1979. Root structure: The considerable cost of belowground function. In: Solbrig OT, Jain S, Johnson GB, Raven PH, eds. Topics in Plant Population Biology. New York; Columbia University Press, pp 408–427. Caldwell MM. 1987. Competition between roots in natural communities. In: Gregory PJ, Lake JV, Rose DA, eds. Root Development and Function. New York; Cambridge University Press, pp 167–185. Caldwell MM, Camp LB. 1974. Belowground productivity of two cool desert communities. Oecologia 17:123–130. Caldwell MM, Eissenstat DM. 1987. Coping with variability: examples of tracer use in root function studies. In:

235 Tenhunen JD, Catarino FM, Lange OL, Oechel WC, eds. Plant Response to Stress—Functional Analysis in Mediterranean Ecosystems. NATO Adv Sci Inst Ser. New York; Springer-Verlag, pp 95–106. Chapin FS III. 1980. The mineral nutrition of wild plants. Annu Rev Ecol Syst 11:233–260. Chandler WH. 1923. Results of Some Experiments in Pruning Fruit Trees. Ithaca, NY; New York Agricultural Exp Sta Bull No. 415. Cheng W, Coleman DC, Box, JE Jr. 1990. Root dynamics, production and distribution in agroecosystems on the Georgia Piedmont using minirhizotrons. J Appl Ecol 27:592–604. Clarkson DT. 1991. Root structure and sites of ion uptake. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. New York; Marcel Dekker, pp 417– 453. Coleman MD, Dickson RE, Isebrands JG, Karnosky DF. 1996. Root growth and physiology of potted and field-grown trembling aspen exposed to tropospheric ozone. Tree Physiol 16:145–152. Comas LH, Eissenstat DM, Lakso AN. 2000. Assessing root death and root system dynamics in a study of grape canopy pruning. New Phytol 147:171–178. Dahlman RC, Kucera RL. 1965. Root productivity and turnover in native prairie. Ecology 46:84–89. Eissenstat DM. 1992. The costs and benefits of constructing roots of small diameter. J Plant Nutr 15:763–782. Eissenstat DM. 1991. On the relationship of specific root length and the rate of root proliferation: a field study using citrus rootstocks. New Phytol 69:870–873. Eissenstat DM, Achor DS. 1999. Anatomical characteristics of roots of citrus rootstocks that vary in specific root length. New Phytol 141:309–321. Eissenstat DM, Duncan LW. 1992. Root growth and carbohydrate responses in bearing citrus trees following partial canopy removal. Tree Physiol 10:245–257. Eissenstat DM, Yanai RD. 1997. The ecology of root life span. Adv Ecol Res 27:1–62. Eissenstat DM, Graham JH, Syvertsen JP, Drouillard DL. 1993. Carbon economy of sour orange in relation to mycorrhizal colonization and phosphorus stress. Ann Bot 71:1–10. Eissenstat DM, Whaley EL, Volder A, Wells CE. 1999. Recovery of citrus surface roots following prolonged exposure to dry soil. J Exp Bot 50:1845–1854. Eissenstat DM, Wells CE, Yanai RD, Whitbeck JL. 2000. Building roots in a changing environment: implications for root longevity. New Phytol 147:33–42. Eissenstat DM, Wells CE, Wang L. 2002. Root efficiency and mineral nutrition in apple. Acta Hort (in press). Espeleta JF, Eissenstat DM. 1998. Responses of citrus fine roots to localized soil drying: a comparison of seedlings with adult fruiting trees. Tree Physiol 18:113–119. Espeleta JF, Eissenstat DM, Graham JH. 1999. Citrus root responses to localized drying soil: a new approach to

236 studying mycorrhizal effects on the roots of mature trees. Plant Soil 206:1–10. Fabia˜o A, Persson HA, Steen E. 1985. Growth dynamics of superficial roots in Portuguese plantations of Eucalyptus globulus Labill studied with a mesh bag technique. Plant Soil 83:233–242. Fahey TJ, Hughes JW. 1994. Fine root dynamics in a northern hardwood forest ecosystem, Hubbard Brook Experimental Forest, NH. J Ecol 82:533–548. Fahey TJ, Bledsoe CS, Day FP, Ruess R, Smucker A. 1999. Root production and demography. In: Robertson GP, Bledsoe CS, Coleman DC, Sollins P, eds. Standard Soil Methods for Long-Term Ecological Research. New York; Oxford University Press, pp 437–455. Fine´r L, Messier C, De Grandpr, L. 1997. Fine-root dynamics in mixed boreal conifer-broad-leafed forest stands at different successional stages after fire. Can J For Res 27:304–314. Fitter AH, Graves JD, Self GK, Brown TK, Bogie DS, Taylor K. 1998. Root production, turnover and respiration under two grassland types along an altitudinal gradient: influence of temperature and solar radiation. Oecologia 114:20–30. Fitter AH, Self GK, Brown TK, Bogie DS, Graves JD, Benham D, Ineson P. 1999. Root production and turnover in an upland grassland subjected to artificial soil warming respond to radiation flux and nutrients, not temperature. Oecologia 120:575–581. Forbes PJ, Black KE, Hooker JE. 1997. Temperatureinduced alteration to root longevity in Lolium perenne. Plant Soil 190:87–90. Gange AC, Brown VK, Sinclair GS. 1994. Reducion of black vine weevil larval growth by vesicular-arbuscular mycorrhizal colonization. Entom Exp Appl 70:115– 119. Gaudinski JB, Trumbore SE, Davidson EA, Zheng S. 2000. Soil carbon cycling in a temperate forest: radiocarbonbased estimates of residence times, sequestration rates and partitioning of fluxes. Biogeochemistry 51:33–69. Gijsman AJ, Floris J, Van Noordwijk M, Brouwer G. 1991. An inflatable minirhizotron system for root observation with improved soil-tube contact. Plant Soil 134:261–270. Gill RA, Jackson RB. 2000. Global patterns of root turnover for terrestrial ecosystems. New Phytol 147:13–31. Graham JH. 1988. Interactions of mycorrhizal fungi with soilborne plant pathogens and other organisms: an introduction. Phytopathology 78:365–366. Graham JH, Brylansky RH, Timmer LW, Lee RF. 1985. Comparison of citrus tree declines with necrosis of major roots and their association with Fusarium solani. Plant Dis 69:1055–1058. Grime JP. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. Am Nat 111:1169–1194.

Eissenstat and Yanai Gross K, Peters A, Pregitzer KS. 1994. Fine root growth and demographic responses to nutrient patches in four oldfield plant species. Oecologia 95:61–64. Harris GA. 1967. Some competitive relationships between Agropyron spicatum and Bromus tectorum. Ecol Monogr 37:89–111. Hayes DC, Seastedt TR. 1987. Root dynamics of tallgrass prairie in wet and dry years. Can J Bot 65:787–791. Haynes BE, Gower ST. 1995. Belowground carbon allocation in unfertilized and fertilized red pine plantations in northern Wisconsin. Tree Physiol 15:317–325. Head GC. 1966. Estimating seasonal changes in the quantity of white unsuberized root on fruit trees. J Hort Sci 41:197–206. Head GC. 1969. The effects of fruiting and defoliation on seasonal trends in new root production on apple trees. J Hort Sci 44:175–181. Head GC. 1973. Shedding of roots. In: Kozlowski TT, ed. Shedding of Plant Parts. New York; Academic Press, pp 237–293. Hendrick RL, Pregitzer KS. 1992. The demography of fine roots in a northern hardwood forest. Ecology 73:1094– 1104. Hendrick RL, Pregitzer KS. 1993. Patterns of fine root mortality in two sugar maple forests. Nature 361:59–61. Hendricks JJ, Nadelhoffer KJ, Aber JD. 1997. A 15N tracer technique for assessing fine root production and mortality. Oecologia 112:300–304. Hodge A, Robinson D, Griffiths BS, Fitter AH. 1999a. Nitrogen capture by plants grown in N-rich organic patches of contrasting size and strength. J Exp Bot 50:1243–1252. Hodge A, Stewart J, Robinson D, Griffiths BS, Fitter AH. 1999b. Plant, soil fauna and microbial responses to Nrich organic patches of contrasting temporal availability. Soil Biol Biochem 31:1517–1530 Hooker JE, Black KE, Perry RL, Atkinson D. 1995. Arbuscular mycorrhizal fungi induced alteration to root longevity of poplar. Plant Soil 172:327–329. Huang B, Eissenstat DM. 2000. Root plasticity in exploiting water and nutrient heterogeneity. In: Wilkinson RE, ed. Plant-Environment Interactions. 2nd ed. New York; Marcel Dekker, pp 111–132. Huang B, Fry JD. 1998. Root anatomical, physiological, and morphological responses to drought stress for tall fescue cultivars. Crop Sci 38:1017–1022. Huang B, Nobel PS. 1992. Hydraulic conductivity and anatomy for lateral roots of Agave deserti during growth and drought-induced abscission. J Exp Bot 43:1441– 1449. Huang B, Duncan RR, Carrow RN. 1997. Drought-resistance mechanisms of seven warm-season turfgrasses under surface soil drying. II. Root aspects. Crop Sci 37:1863–1869.

Life Span, Efficiency, and Turnover Huck MG, Hoogenboom G, Peterson CM. 1987. Soybean root senescence under drought stress. In: Taylor HM, ed. Minirhizotron Observation Tubes: Methods and Applications for Measuring Rhizosphere Dynamics. ASA Special Publ. No. 50. Madison, WI: Agronomy Society of America, pp 168–174. Jackson RB, Mooney HA, Schulze ED. 1997. A global budget of fine root biomass, surface area, and nutrient contents. Proc Natl Acad Sci USA 94:7362–7366. Jackson RB, Schenk HJ, Jobbagy EG, Canadell J, Colello GD, Dickinson RE, Field CB, Friedlingstein P, Heimann M, Hibbard K, Kicklighter DW, Kleidon A, Neilson RP, Parton WJ, Sala OE, Sykes MT. 2000. Belowground consequences of vegetation change and their treatment in models. Ecol Appl 9:470–483. Johnson MG, Phillips DL, Tingey DT, Storm MJ. 2000. Effects of elevated CO2, N-fertilization, and season on survival of ponderosa pine fine roots. Can J For Res 30:220–228. Joslin JD, Wolfe MH. 1999. Disturbances during minirhizotron installation can affect root observation data. Soil Sci Soc Am J 63:218–221. Joslin JD, Wolfe MH, Hanson PJ. 2000. Effects of altered water regimes on forest root systems. New Phytol 147:117–129. Keyes MR, Grier CC. 1981. Above- and below-ground net production in 40-year-old Douglas-fir stands on low and high productivity sites. Can J For Res 11:599– 605. King JS, Pregitzer KS, Zak DR. 1999. Clonal variation in above- and below-ground growth responses of Populus tremuloides Michaux: influence of soil warming and nutrient availability. Plant Soil 217:119–130. Kosola, KR, Eissenstat DM. 1994. The fate of surface roots of citrus seedlings in dry soil. J Exp Bot 45:1639–1645. Lamont BB. 1995. Mineral nutrient relations in Mediterranean regions of California, Chile and Australia. In: Arroyo MTK, Zedler PH, Fox MD, eds. Ecology and Biogeography of Mediterranean Ecosystems of Chile, California, and Australia. New York; Springer-Verlag, pp 211–238. Lopez B, Sabate S, Garcia C. 1996. An inflatable minirhizotron system for stony soils. Plant Soil 179:255–260. Majdi H, Kangas P. 1997. Demography of fine roots in response to nutrient applications in a Norway spruce stand in southwestern Sweden. Ecoscience 4:199–205. Ma¨kela¨ A, Vanninen P. 2000. Estimation of fine root mortality and growth from simple measurements: a method based on system dynamics. Trees 14:316–323. Marshall JD. 1986. Drought and shade interact to cause fineroot mortality in Douglas-fir seedlings. Plant Soil 91:51–60. Matamala R, Gonzalez-Meler MA, Schlesinger WH. 2000. How long do roots live? A 13C tracer technique for assessing fine root longevity in a North Carolina pine

237 forest. Abstracts of Ecological Society of America Annual Meeting, Snowbird, UT, p 154. Meyer WS, Barrs HD. 1991. Roots in irrigated clay soils: measurement techniques and responses to root zone condition. Irrig Sci 12:125–134. Milchunas DG, Lauenroth WK. 1992. Carbon dynamics and estimates of primary production by harvest, 14C dilution, and 14C turnover. Ecology 73:593–607. Nadelhoffer KJ. 2000. The potential effects of nitrogen deposition on fine-root production in forest ecosystems. New Phytol 147:-131-139. Nadelhoffer KJ, Raich JW. 1992. Fine root production estimates and belowground carbon allocation in forest ecosystems. Ecology 73:1139–1147. Nadelhoffer KJ, Aber JD, Melillo JM. 1985. Fine roots, net primary production and nutrient availability: a new hypothesis. Ecology 66:1377–1390. Newsham KK, Fitter AH, Watkinson AR. 1995. Arbuscular mycorrhiza protect an annual grass from root pathogenic fungi in the field. J Ecol 83:991–1000. Nobel PS, Alm DM, Cavelier J. 1992. Growth respiration, maintenance respiration and structural-carbon costs of roots of three desert succulents. Func Ecol 6:79–85. Norby RJ, Jackson RB. 2000. Root dynamics and global change: seeking an ecosystem perspective. New Phytol 147:3–12. North GB, Huang B, Nobel PS. 1993. Changes in the structure and hydraulic conductivity of desert succulents as soil water status varies. Bot Acta 106:126–135. Peng S, Eissenstat DM, Graham JH, Williams K, Hodge NC. 1993. Growth depression of mycorrhizal citrus at high phosphorus supply: analysis of carbon costs. Plant Physiol 101:1063–1071. Phillips DL, Johnson MG, Tingey DT, Biggart C, Nowak RS, Newsom JC. 2000. Minirhizotron installation in sandy, rocky soils with minimal soil disturbance. Soil Sci Soc Am J 64:761–764. Pregitzer KS, Hendrick RL, Fogel R. 1993. The demography of fine roots in response to patches of water and nitrogen. New Phytol 125:575–580. Pregitzer KS, Zak DR, Crutis PS, Kubiske ME, Teeri JA, Vogel CS. 1995. Atmospheric CO2, soil nitrogen and fine root turnover. New Phytol 129:579–585. Pregitzer KS, Kubiske ME, Yu CK, Hendrick RL. 1997. Relationships among root branch order, carbon and nitrogen in four temperate species. Oecologia 111:302–308. Pregitzer KS, Laskowski MJ, Burton AJ, Lessard VC, Zak DR. 1998. Variation in sugar maple root respiration with root diameter and soil depth. Tree Physiol 18:665–670. Pregitzer KS, King JS, Burton AJ, Brown SE. 2000a. Responses of tree fine roots to temperature. New Phytol 147:105–115. Pregitzer KS, Zak DR, Maziasz J, DeForest J, Curtis PS, Lussenhop J. 2000b. Interactive effects of atmospheric

238 CO2 and soil-N availability on fine roots of Populus tremuloides. Ecol Appl 10:18–33. Raich JW, Nadelhoffer KJ. 1989. Belowground carbon allocation in forest ecosystems: global trends. Ecology 70:1346–1354. Reich PB, Walters MB, Ellsworth DS. 1997. From tropics to tundra: global convergence in plant functioning. Proc Natl Acad Sci USA 94:13730–13734. Reich PB, Walters MB, Tjoelker MG, Vanderklein D, Buschena C. 1998. Photosynthesis and respiration rates depend on leaf and root morphology and nitrogen concentration in nine boreal tree species differing in relative growth rate. Func Ecol 12:395–405. Ruess RW, Hendrick RL, Bryan JP. 1998. Regulation of fine root dynamics by mammalian browsers in early successional Alaskan taiga forests. Ecology 79:2706–2720. Ryser P. 1996. The importance of tissue density for growth and life span of leaves and roots: a comparison of five ecologically contrasting grasses. Func Ecol 10:717– 723. Rytter, R-M. 1999. Fine-root production and turnover in a willow plantation estimated by different calculation methods. Scand J For Res 14:526–537. Sala OE, Biondi ME, Lauenroth WK. 1988. Bias in estimates of primary production: an analytical solution. Ecol Model 44:43–55. Samson BK, Sinclair TR. 1994. Soil core and minirhizotron comparison for the determination of root length density. Plant Soil 161:225–232. Santantonio D, Grace JC. 1987. Estimating fine root production and turnover from biomass and decomposition data: a compartment flow model. Can J For Res 17:900–908. Shaver GR, Billings WD. 1975. Root production and root turnover in a wet tundra system. Ecology 56:401–409. Singh JS, Lauenroth WK, Hunt HW, Swift DM. 1984. Bias and random errors in estimators of net root production: a simulation approach. Ecology 65:1760–1764. Smith PF. 1976. Collapse of ‘Murcott’ tangerine trees. J Am Soc Hort Sci 101:23–25. Smith SE, Read DJ. 1997. Mycorrhizal Symbiosis. 2nd ed. San Diego, CA: Academic Press. Smucker AJM, Aiken RM. 1992. Dynamic root responses to soil water deficits. Soil Sci 154:281–289. Steele SJ, Gower ST, Vogel JG, Norman JM. 1997. Root mass, net primary production and turnover in aspen, jack pine and black spruce forests in Saskatchewan and Manitoba, Canada. Tree Physiol 17:577–587. Taylor HM, ed. 1987. Minirhizotron Observation Tubes: Methods and Applications for Measuring Rhizosphere Dynamics. ASA Special Publ. No. 50. Madison, WI: American Society of Agronomy. Tingey DT, Phillips DL, Johnson MG. 2000. Elevated CO2 and conifer roots: effects on growth life span and turnover. New Phytol 147:87–103.

Eissenstat and Yanai Thornley JHM. 1998. Modelling shoot:root relations: the only way forward? Ann Bot 81:165–171. Van Rees KCJ, Comerford NB. 1990. The role of woody roots of slash pine seedlings in water and potassium absorption. Can J For Res 20:1183–1191. Vogt KA, Edmonds RL, Grier CC. 1981. Seasonal changes in biomass and vertical distribution of mycorrhizal and fibrous-textured conifer fine roots in 23- and 180-yearold subalpine Abies amabilis stands. Can J For Res 11:223–229. Vogt KA, Grier CC, Vogt DJ. 1986. Production, turnover, and nutrient dynamics of above- and belowground detritus of world forests. Adv Ecol Res 15:303–377. Vogt KA, Persson H. 1991. Root methods. In: Lassoie JP, Hinckley TM. Ecophysiology of Forest Trees. Vol. 1, Techniques and Methodologies. Boca Raton, FL: CRC Press, pp 477–501. Watson CA, Ross JM, Bagnaresi U, Minotta GF, Roffi F, Atkinson D, Black KE, Hooker JE. 2000. Environment-induced modifications to root longevity in Lolium perenne and Trifolium repens. Ann Bot 85:397–401. Weaver JE, Zink E. 1946. Length of life of roots of ten species of perennial range and pasture grasses. Plant Physiol 21:201–217. Wells CE. 1999. Advances in the fine root demography of woody species. PhD thesis. Pennsylvania State University, University Park, PA. Wells CE, Eissenstat DM. 2001. Marked differences in survivorship among apple roots of different diameters. Ecology 82:882–892. Wells CE, Eissenstat DM, Glenn DM. Submitted. Soil insects alter fine root demography in peach (Prunus persica). Submitted to Plant Cell Environ. Wells CE, Eissenstat DM, Glenn DM. 2002 Submitted 2 Changes in the risk of fine root mortality with age: a case study in peach (Prunus persica). Am J Bot. Weste G. 1986. Vegetation changes associated with invasion by Phytophthora cinnamomi of defined plots in the Brisbane Ranges, Victoria, 1975–1985. Aust J Bot 34:633–648. Whaley EL. 1995. Uptake of phosphorus by citrus roots in dry surface soil. MSc thesis, University of Florida, Gainesville, FL. Williams M, Yanai RD. 1996. Multi-dimensional sensitivity analysis and ecological implications of a nutrient uptake model. Plant Soil 180:311–324. Wilson JB. 1988. Shoot competition and root competition. J Appl Ecol 25:279–296. Yanai RD. 1994. A steady-state model of nutrient uptake improved to account for newly-grown roots. Soil Sci Soc Am J 58:1562–1571. Yanai RD, Fahey TJ, Miller SL. 1995. Efficiency of nutrient acquisition by fine roots and mycorrhizae. In: Smith WK, Hinckley TM, eds. Resource Physiology of Conifers. New York; Academic Press, pp 75–103.

14 Maize Root System and Genetic Analysis of Its Formation Gu¨nter Feix, Frank Hochholdinger, and Woong June Park University of Freiburg, Freiburg, Germany

I.

INTRODUCTION

tion and morphogenesis of various root types and root hairs. Identification and correct phenotypic assessment of such mutants were essential for detailed analysis of the structure and development of the root system (cf. Feldman, 1994; Kisselbach, 1999). As this aspect remains important for further mutant work, a short structural account of the root system of maize is given below.

The genetic analysis of root formation in maize has long been neglected, although maize has been a favored plant object for genetic analysis (cf. Coe et al., 1988; Sheridan and Clark, 1988). The lack of genetic research of root structure and its formation in monocotyledonous plants was largely influenced by the difficulty of observing the complex underground root systems. Furthermore, the large environmental influence on root formation makes it difficult to identify mutants in large plant populations. Thus, only one maize root mutant had been characterized until the mid-1990s, rt1 (Jenkins, 1930), a mutant that shows a reduced lodging resistance caused by defects in shootborne root formation. The deficiency of available root mutants was recently overcome in the model plant Arabidopsis by applying new methods and concepts, and this led to the isolation of an impressive number of mutants allowing new insights into the general mechanisms of root formation and morphogenesis (cf. Scheres and Wolkenfelt, 1998). However, as the root system of maize is more complex (Feldman, 1994) than that of Arabidopsis, and because a detailed knowledge of root structure and formation of food plants like maize is still needed, a concerted effort was recently undertaken in search for root mutants in maize (Wen and Schnable, 1994; Hetz et al., 1996; Hochholdinger and Feix, 1998a). This effort has concentrated so far on the isolation and characterization of monogenic mutants with an influence on the forma-

II.

ROOT SYSTEM OF MAIZE

A.

Various Root Types and Their Formation During Development

The primary root system of maize consists of the embryonic primary and seminal roots, the postembryonic shootborne crown and brace roots, and lateral roots emerging from all root types. The term adventitious root has often been associated with the crown and brace roots (Esau, 1977; Fahn, 1990; Kisselbach, 1999), but adventitious root formation occurs, strictly spoken, only after an external influence like injury or hormone treatment and is not developmentally programmed (Feldman, 1994). The primary root is an endogenous part of the developing embryo (Yamashita, 1991). It grows out of the basal meristem and appears to be essential for the growing at the seedling stage only. At later stages of development the growth of the primary root usually stops (Larson and Hanway, 1977; Feldman, 1994). However, it may resume growth and support the 239

240

Feix et al.

plant to maturity if necessary (Kausch, 1967; McCully and Canny, 1988; Hetz et al., 1996; Kisselbach, 1999). The seminal roots are also of embryonic origin and their primordia emerge from the scutellar node during the later phase of embryogenesis. In more recent work they have also been called lateral-seminal roots (Hetz et al., 1996). The number of the seminal roots is a polygenic trait and highly variable (0–20) between different maize lines (Sass, 1977; Feldman, 1994; Kisselbach, 1999). The ontogenic origin of the seminal roots is still unclear especially when relating to the issue of allorhizie and homorhizie (Tillich, 1977). It is still unclear which role the seminal roots play for a good performance of the seedling. The shootborne crown roots are formed in variable numbers at consecutive underground stem nodes and initiate at the inner cell layer of the nodes (Martin and Harris, 1976; Varney and McCully, 1991; Varney et al., 1991; Wang et al., 1991, 1994). Numbers and sizes of the crown roots increase on higher nodes of the stem. When formed from the aboveground nodes, such roots are designated as brace roots (Feldman, 1994). Crown roots represent the backbone of the overall root architecture of maize endowing the plant its lodging resistance and are essential for the stability of the growing plant (McCully and Canny, 1988). Lateral roots initiate in the pericycle cell layer (Esau, 1977) or sometimes from some endodermal cells of the root differentiation zone (Bell and McCully, 1970). The primordia of such roots are formed by a multistep mechanism (Laskowski et al., 1995; Malamy and Benfey, 1997). They are much thinner than crown

roots (Peterson, 1991) and can lead to the formation of secondary and higher-order lateral roots which have a great influence on the architecture of the root system (Lynch, 1995). Lateral roots increase the absorbing surface of the root system and are essential for water and nutrient uptake. Root hairs are epidermal extensions that are formed on both embryonic and postembryonic roots and have an essential function for nutrient and water uptake by further increasing the absorbing surface of the root (Gilroy and Jones, 2000; see also Chapter 5 by Ridge and Katsumi in this volume). However, the importance of these root hair functions in maize is unclear, because some mutants lacking root hairs showed normal growth and development (Wen and Schnable, 1994). A summary of some important properties of the various root types is given in Table 1. B.

Structure of Individual Roots

The basic anatomical structure of maize roots (Avery, 1930; Sass, 1977; Feldman, 1994; Ishikawa and Evans, 1995) is comparable to that of Arabidopsis roots (Dolan et al., 1994; Benfey and Schiefelbein, 1994; Scheres and Wolkenfelt, 1998), but maize roots display some additional features. The quiescent center is much larger (it contains >1000 cells; Feldman, 1994), the cortical cell files are more numerous (up to 10–15; Feldman, 1998) and border cells are continuously secreted at the tip region (Vermeer and McCully, 1982). The mesocotyl connecting the scutellar and coleoptilar node has structural features similar to

Table 1 The Root System of Maize Root type Primary root Nodal roots Seminal roots Crown roots Brace roots Lateral roots Adventitious roots

Site and time of root outgrowth

Time of primordia formation

Number of roots

Influence on lodging resistance

From root apical meristem in germinating embryo

Early in embryogenesis

1

—

From scutellar node in first days of seedlings growth From underground nodes at developing stem From aboveground nodes at stem (later growing stage) From pericycle of differentiated roots From stem at random sites after exogenous induction

Late in embryogenesis

—

Late in plant development

0–12(variety dependent) Major root stock of plant Several

+

Not defined

Many

(+)

After dedifferentiation and reprogramming of stem regions

Stimulus dependent

(+)

During growth of plant

+++

Maize Root System

roots. This is demonstrated by the extensive growth of fine roots from the mesocotyl in some maize lines. Field-grown roots normally have a well-developed rhizosphere that influences the structural features of the root; e.g., hypoxic or anaerobic soil condition can lead to arenchyma formation (He et al., 1996).

241

tance was undertaken with the option for a later isolation of the genes impaired by the mutation. An account of these attempts will be given in the following paragraphs.

A. C.

Transition from Early to Late Root Architectures

The overall architecture of the root system is determined by the transition from the temporary embryonic to the persisting postembryonic root structures. The early root system consists of the primary and other seminal roots together with their lateral roots. It is only essential for the development of the young seedling. The shape of this structure is comparable to the root system of the dicot plants (e.g., Arabidopsis) except for the additional presence of the seminal roots. At about 2 weeks after germination, the postembryonic shootborne roots start to develop and gradually determine the form of the root system architecture. The final shape of the root system is governed by the number and growth direction of the postembryonic crown roots together with their branching laterals (Lynch, 1995). The development of the root system is strongly influenced by environmental factors like water and nutrient availability. Those can determine the rate of root initiation and growth and are the major causes for the plasticity of the root architecture (Aiken and Smucker, 1996; McCully, 1999).

III.

GENETIC ANALYSIS OF ROOT FORMATION

The formation of the complex root structure of maize is controlled by many interacting genes, leading to polygenic inheritance patterns of many root-related traits. Breeders succeeded in manipulating and modifying the root system in line with agronomic requirements leading to the isolation of varieties with a more shallow or deep root system responsive to changes of soil conditions. More recently, quantitative trait loci (QTL) analyses were performed with rootconnected traits associating genetic determinants of root formation to chromosome regions (Lebreton et al., 1995; Zheng et al., 2000). The dissection, however, of such QTLs into their monogenic components has not been achieved. In a first step to identify genes involved in root formation and morphogenesis, a search for mutants with a strictly monogenic inheri-

Variability of Root Formation and Induced Mutagenesis

The variability of root architecture of maize varieties is mostly of polygenic nature and could only rarely be used as starting material for the isolation of monogenic mutants (Schiefelbein and Benfey, 1991). Depending on the crosses performed with lines displaying particular structural features of their root system, a monogenic trait of interest may be identified by segregation analysis and be used for the isolation of a mutant. One such rare case was apparently present in the progenitor line used for the isolation of the rtcs mutant (described in Section III.C). However, such an approach is very unpredictable in its outcome and not suitable for a more systematic search for root mutants. Systematic searches for monogenic recessive mutants were successfully performed with maize lines mutagenized by an ethylmethanesulfonate (EMS) treatment of pollen (Neuffer, 1994). The search for mutants was then undertaken in segregating F2 families generated from the mutagenized line by selfings. The identification of monogenic recessive mutants required the availability of large numbers of F2 families. For example, in the case of the screening for the lateral root mutant lrt1 (described in Section III.C), 9811 segregating F2 families were used. The second mutagenesis procedure used depends on the insertion of transposons by working with lines containing active transposons in large amounts. After the initial use of the Ac-Ds and the En-Spm transposon systems, the Mutator Mu system is now preferred because of its high mutation rate and the apparent random insertion of Mutator into chromosomal sites (Gierl and Saedler, 1992). It should be remembered that the Mu system does generate germinal reversions only at a very low rate that have been proven advantageous for the identification of mutants (Bennetzen et al., 1993; Bennetzen, 1996). For example, the transposon mutagenesis was applied in the case of the mutants slr1 and slr2 (1969 segregating families were used in this screening; described in Section III.D) or in the screening of root hair mutants (2892 segregating families revealed the mutants rth2 and rth3; described in Section III.E).

242

B.

Feix et al.

Isolation of Mutants

Two major test systems were applied for the screening of large numbers of segregating F2 families for recessive monogenic mutants. The search for mutants with aberrations in the formation of the early root system was performed in a paper roll test (Hetz et al., 1996). This is a system which allows the screening of large numbers of seedlings under controlled conditions, in limited space. In brief, surface sterilized kernels were placed on properly arranged filter paper sheets, which were rolled up and put with their lower part into beakers containing water only. The kernels were then left to germinate for up to 4 weeks. The developing root system of the seedlings was visually analyzed after opening the rolled-up sheets. Such examinations could be repeated several times if needed. This test system is very reliable and could, of course, be extended by supplying further screening criteria to the germinating kernels. The search for mutants with defects at later stages of root formation was performed in field tests of the segregating F2 population. This test is more demanding because only in rare cases a direct correlation of the observed changes to the root system can be obtained. This was the case with the laterdescribed root lodging mutant rtcs. Isolates, with changes in root phenotypes and with wild-type siblings showing the expected Mendelian segregation pattern, should in the beginning be considered as tentative mutants. After further selfings and outcrossings to other genetic backgrounds, the mutant candidates may then prove to be monogenic recessive, stable, and true mutants. The screening conditions,

under which the isolation of the mutants was done, need also to be strictly reproducible to exclude environmental influences on the observed phenotypes. The major classes of ‘‘root’’ mutants isolated so far showed defects in the formation or morphogenesis of the primary or of lateral roots. They displayed a different architecture or were agravitropic. The mutant classes described below in more detail were selected for further analysis because of their highly root-specific phenotype with relevance to the mechanism of root formation and development. C.

Root Initiation Mutants

The two mutations rtcs (deficient in the formation of nodal roots; Hetz et al., 1996) and lrt1 (deficient in the formation of lateral roots; Hochholdinger and Feix, 1998a) are of particular interest. In both mutants primordia cannot be detected by microscopic examination, indicating that the defects of these loss-offunction mutants act at an early stage of root initiation. The monogenic recessive mutant rtcs (rootless for crown and seminal roots; Hetz et al., 1996) was isolated from a cross of a dent line with the local variety DK 105 from Germany and is apparently not caused by any of the known active transposon lines. A graphic illustration of a 14-day-old seedling of rtcs is depicted in part B of Fig. 1. Although the aboveground phenotype (complete loss of root lodging resistance) is very dramatic, no mutant of this phenotype has been reported before. The mutation affects all crown roots and the embryonic seminal roots by inhibiting primor-

Figure 1 Graphic illustrations of the mutants described. The root systems of 14-day-old seedlings of wild-type (A), rtcs (B), lrt1 (C) and slr1 (D) are shown. Drawings by Miwa Kojima, Agronomy Department, Iowa State University, Ames, IA.

Maize Root System

dia formation. The effect of the mutation seems to be highly specific, since no pleiotropic effects on other parts of the plant have been observed. Normal seed set can be achieved if the impaired lodging mutant plants are supported by a stick. The absence of cyclines as shown by in situ hybridization experiments (Hochholdinger and Feix, 1998b) indicates that this mutation prevents all cell divisions in the region of primordia formation. The deficiency of the mutant is confined to the root-initiating part of the nodes with no influence on their tiller-forming part tested by introgressing the rtcs locus into the heavily tillered gaspe flint line (Hochholdinger and Feix, 1998c). The RTCS locus was mapped with the help of a cosegregating RAPD marker (Hetz et al., 1996) and by a B-A translocation test (Hochholdinger and Feix, unpublished) to the small arm of chromosome 1. The monogenic recessive mutant lrt1 (lateral rootless 1; Hochholdinger and Feix, 1998a) was isolated from EMS-mutagenized B73 seeds. It lacks postembryonic root formation at the early seedling stage; i.e., no lateral roots emerge from the primary root or from the crown roots that grow out of the coleoptilar node (see part C of Fig. 1). Root growth of these plants resumes later leading to normally developed root systems of the mature plants. The defect of lrt1 operates very early in root initiation before the primordia can be seen. Furthermore, the wild-type phenotype cannot be rescued from mutants by the application of auxin to germinating kernels. The mutant does not form an elongated mesocotyl in the dark, although its photomorphogenic response appears to be normal in the mutant. Double mutants prepared from lrt1 and rtcs show a strict additive behavior. The LRT1 locus was mapped to chromosome 2S by B-A translocation analysis (Hochholdinger and Feix, 1998a). D.

Root Elongation Mutants

In the mutants slr1 and slr2 (short lateral root), the mechanism of cell elongation of lateral roots is disturbed leading to lateral roots with impaired morphogenesis (Hochholdinger et al., 2001). The two recessive and nonallelic mutants isolated from selfed Mutator stocks both displayed short lateral roots of about one-fourth the length of wild-type roots (see part D of Fig. 1) and very limited root hair formation because of their incomplete cell elongation zone. In the case of slr1, the mutation was mapped by B-A translocation analysis to the short arm of chromosome 3. A particular feature of these mutants is the strict root type specificity, since only lateral roots of the embryonic

243

primary and of other seminal roots are affected. Furthermore, the deficiency of the mutants acts transiently during the first 4 weeks, and the wild-type phenotype is restored afterward. Confocal laser scanning microscopy demonstrated that the short lateral roots are caused by the presence of smaller cells compared to wild-type lateral roots. In the double mutant slr1 slr2, the defect does not only influence lateral root-specific cell elongation, but also leads to disarranged longitudinal cellular patterns in the primary and in the other seminal roots. The transient nature of the single mutants is, however, retained in the double mutant, indicating that the two loci affected in slr1 and slr2 are cooperating in the establishment of the lateral root specificity during early root development. E.

Root Maturation Mutants

Root hair formation is a typical trait of the root maturation zone. So far, the three monogenic recessive mutants rth1-rth3 (root hairless), which are defective in the morphology of root hairs of the primary root and the seminal roots, have been isolated from selfed Mutator stocks (Wen and Schnable, 1994). These mutants are of particular interest, since root hairs play an important role in plant nutrition by facilitating the uptake of water and nutrients (Gilroy and Jones, 2000). The rth1 and rth2 mutants form normal root hair primordia. However, the root hairs of the mutant rth1 hardly elongate and reach only 1/20 to 1/10 of the length of the root hairs of the wild-type plants. The root hairs of the mutant rth2 show elongation to about one-fifth to one-fourth of those of the wild type. In contrast to rth1 and rth2, the mutant rth3 displays abnormal root hair primordia already at a very early stage of development, as can be seen by scanning electron microscopy. Such an early expression eventually leads to very short root hairs similar to the case of rth1. This implies that different genes regulate root hair elongation at different stages of development. The mutant rth1 shows pleiotropic nutritional defects leading to stunted plants with purplish leaves that never form ears and only rarely produce tassels. In contrast to rth1, the plants of the mutants rth2 and rth3 show a normal growth performance during their entire development despite their drastically impaired root hair morphology. This might suggest that under some conditions root hairs are less important for plant growth than had been previously thought. rth1, rth2, and rth3 were mapped by B-A translocation tests to chromosome arms 1L, 5L, and 1S, respectively (Wen and Schnable, 1994).

244

Feix et al.

F. Conclusions from Mutants Analysis

phenotype that persists until about the third week after germination cannot be rescued by auxin and is not under the influence of a phytochrome signal transduction pathway. Similarly, a development-specific signal can also be implied in the case of the slr mutants which are affected only in the lateral root formation of the embryonic but not of the postembryonic shootborne roots. Another aspect that can be learned from the chromosomal map positions of the mutant loci indicates that a larger root locus combining root relevant genes is not evident from the available data. A summary of these findings of the mutant analysis is given in Table 3. Although root hairs immensely increase the absorbing surface and are known to play an important role in nutrient and water uptake (Gilroy and Jones, 2000), they should, depending on growth conditions, be considered as less important in maize growth than previously thought. This is indicated by the root hair mutants rth2 and rth3 which are drastically impaired in root hair elongation but do not show any other defects during the development of the whole plant (Wen and Schnable, 1994).

The use of mutants revealed several properties of the mechanism of root formation not easily accessible otherwise. For example, the analysis of rtcs demonstrated that the initiation of embryonic and postembryonic nodal roots depends on an intact RTCS gene, while the initiation of crown roots from the scutellar node depends also on the LRT1 gene. The two genes act independently of each other as shown by the additive nature of the double mutant from rtcs and lrt1. This implies that at least two independent mechanisms are effective in the formation of root primordia. The LRT1 gene is also essential for the initiation of lateral roots emerging from the primary root showing that the formation of roots initiated either from preformed roots or from nodal tissue shares expression features of LRT1. Interestingly, gene functions with an influence on several root types had been seen by most of the genes impaired in the available mutants (see Table 2). A case of extreme root type specificity is demonstrated by the transient deficiency in lateral root elongation effective only on lateral roots emerging from the primary root. This root type, tissue, and stage specificity is governed by at least two independent genes (SLR1 and SLR2). The transient occurrence of the deficiency of several mutants together with further findings indicates that various steps of root formation depend on sets of signals operating for defined periods of time in the development of the plant. In the case of lrt1, the mutant

IV.

OUTLOOK

The mutant work, although in its early stage and still very fragmentary, has already proven very valuable for the advancement of the knowledge on root formation and morphogenesis. A schematic summary of the mutants isolated so far, with indications of where in

Table 2 Root Typesa Affected in the Mutants Are Indicated by Shaded Areas Mutant

PR

SR

CR

rtcs rt1 lrt1 rth1-3 slr1 slr2 brt1 des21 agt Asr1 a

PR, primary root; SR, seminal root; CR, crown root; BR, brace root. See references for des21 (Gavazzi et al., 1993), agt (Pilet, 1983), and Asr1 (De Miranda et al., 1980).

BR

Maize Root System

245

Table 3 Characteristics of the Mutants

Name

Gene

Chromosomal map position

rootless for crown and seminal root lateral rootless

rtcs

1S

lrt1

2S

short lateral root 1

slr1

3S

short lateral root 2

slr2

N/A

root hairless 1

rth1

1L

root hairless 2

rth2

5L

root hairless 3

rth3

1S

Root type affected All crown and brace roots Lateral roots and crown roots from coleoptilar node Lateral roots from primary and seminal roots Lateral roots from primary root and seminal roots Root hairs from primary and seminal roots (others N/A) Root hairs from primary and seminal roots (others N/A) Root hairs from primary and seminal roots (others N/A)

the pathways of root formation they can be placed, is shown in Fig. 2. It is evident that mutants for only a few steps in the mechanism of root formation have been found, and it is hoped that great efforts will be made to fill in the many gaps of lacking knowledge. So far, only visual inspection has been used for mutant isolation using the paper roll test. Screening procedures with better information are expected to be used in the future. For example, the monogenic recessive mutant brt1 (for brown root), displaying a brown-looking central cylinder of the primary root starting 7 days after germination and leading to a drying out of the seedling, was identified under water stress conditions in a paper roll screening experiment (Hochholdinger and Feix, 1998d). It was found that this mutant shows deficiencies in a function relevant to survival under anaerobic conditions (Chang et al., 2000). The search for mutants in late-appearing traits of the root system of maize will continue to depend on their detection in field tests. This is because of the difficulties to test thousands of segregating families under hydroponic or aeroponic conditions or in root boxes. The need for additional efforts for the identification of mutants in field experiments is evident from

Observed defects Primordia formation Primordia formation Cell elongation

Plant growth period affected after germination

References

Permanent

Hetz et al. (1996) First 4 weeks Hochholdinger and Feix (1998a) First 4 weeks Hochholdinger et al. (2001)

Cell elongation

First 4 weeks Hochholdinger et al. (2001)

Root hair elongation

Permanent

Root hair elongation

Not affected

Abnormal primordia, root hair elongation

Not affected

Wen and Schnable (1994) Wen and Schnable (1994) Wen and Schnable (1994)

the isolation and investigation of the recessive mutant rtd (root degradation). This mutant is characterized by the drying of lower leaves of approximately 4-week-old plants (Krebs and Feix, unpublished). It was then found, in an experimental setup allowing the direct observation of the roots, that the drying of the leaves was the consequence of a premature degradation of the root system. The screening for specific mutants with defects at later stages of development is important for the recognition of genes acting at later phases of postembryonic development. It is hoped that aboveground traits cosegregating with root deficiencies and relevant markers will become increasingly available for the detection of such mutants in field experiments. Further insights into the genetic network operating in the formation and function of roots can be expected from double-mutant analyses and the generation of second-site mutations. Such studies have proven powerful in many gene systems for the analysis of gene interactions. Double mutants will not only be informative if generated from different single-root mutants as shown in the case of slr1 and slr2, but also in combinations with single mutants with defects in other parts of the plants. For example, it will be interesting to check, by using the appropriate double

246

Feix et al.

Figure 2 Genetic control of root formation in maize. All mutants are described in the text.

mutants, whether the root apical meristem shares important gene functions and characteristics with the other meristems of the maize plant. Furthermore, double-mutant analyses may yield important information on common features between developmentally controlled root initiation and root formation induced by external signals. Regarding the identification of target genes of the mutated gene structures and of the influence of such genes on cell functions, the use of newly established profiling techniques like proteomics (Touzet et al., 1996; Thiellement et al., 1999; Chang et al., 2000) or

EST microarray technology (Richmond and Sommerville, 2000; Van Hal et al., 2000) will be very rewarding. Assistance in the identification of gene activities, with relevance to root formation and function, can also be expected from the use of reverse genetic methods by identifying cDNAs and ESTs with root-specific expression. Using this approach it was possible to isolate a cDNA coding for a glycine rich protein (Goddemeier et al., 1998) which was shown by in situ hybridization to be highly specifically expressed in the exodermis of the root tip of several root types

Maize Root System

(Hochholdinger and Feix, unpublished). Furthermore, the isolated DNA elements can be used in the Trait Utility System of Corn (TUSC) with the aim to search for mutants caused by genes coding for the cDNA of interest. Isolation of the mutated genes, which is in progress in the case of some of the mutants, will allow the understanding of the genes involved in root development. Sequence elements identified that way might reveal cases of root-specific genes as part of larger gene families. They will also allow scientists to search for homologous genes in other plants, like rice. Furthermore, sequence elements will be very helpful in the identification of more alleles of the genes affected by the mutation as well as the phenotypes caused by these alleles. This could be achieved by using the TUSC system. It is of particular interest to identify the specific promoters and potential root-specific regulatory elements of the isolated genes. Root-specific promoters are very much in demand for the construction of transgenes, as for instance in expression cassettes of transgenes intended for the tissue- and stage-specific in planta synthesis of Bt-type insecticides (de Maagd et al., 1999). The information gained by the analysis of the cloned genes will be very useful for selecting the bestsuited alleles for the construction of transgenic plants for application purposes.

REFERENCES Aiken RM, Smucker AJM. 1996. Root system regulation of whole plant growth. Annu Rev Phytopathol 34:325– 346. Avery GS. 1930. Comperative anatomy and morphology of embryos and seedlings of maize, oats and wheat. Bot Gaz 89:1–39. Bell JK, McCully ME. 1970. A histological study of lateral root initiation and development in Zea mays. Protoplasma 70:179–205. Benfey PN, Schiefelbein JW. 1994. Getting to the root of plant development: the genetics of Arabidopsis root formation. TIG 10:84–88. Bennetzen JL. 1996. The mutator transposable element system of maize. Curr Top Microbiol Immunol 204:195–229. Bennetzen JL, Springer PS, Cresse AD, Hemdrickx M. 1993. Specificity and regulation of the mutator transposable element system of maize. Crit Rev Plant Sci 12:57–95. Chang WWP, Huang L, Shen M, Webster C, Burlingame AL, Roberts JKM. 2000. Patterns of protein synthesis and tolerance of anoxia in root tips of maize seedlings acclimated to a low-oxygen environment, and identification of proteins by mass spectrometry. Plant Physiol 122:295–317.

247 Coe EH Jr, Neuffer MG, Hoisington DA. 1988. The genetics of corn. In: Sprague GF, ed. Corn and Corn Improvement. Madison, WI: American Society of Agronomy. de Maagd RA, Bosch D, Stiekema W. 1999. Bacillus thuringiensis toxin-mediated insect resistance in plants. TIPS 4:913. De Miranda LT. 1980. Inheritance and linkage of root characteristic from Puebla maize. Maize Genet Newslett 54:18–19. Dolan L, Duckett CM, Grierson C, Linstead P, Schneider K, Lawson E, Dean C, Poethig S, Roberts K. 1994. Clonal relationships and cell patterning in the root epidermis of Arabidopsis. Development 120:2465– 2474. Esau K. 1977. Plant Anatomy. 2nd ed. New York; John Wiley and Sons. Fahn A. 1990. Plant Anatomy. 4th ed. Oxford, UK: Pergamon Press. Feldman L. 1994. The maize root. In: Freeling M, Walbot V, eds. The Maize Handbook. New York; SpringerVerlag, pp 29–37. Feldman LJ. 1998. Not so quiet quiescent centers. TIPS 3:80–81. Gierl A, Saedler H. 1992. Plant transposable elements and gene tagging. Plant Mol Biol 19:39–49. Gavazzi G, Dolfini M, Galbiati M, Helentjari T, Landon M, Pelucci N, Todesco G. 1993. Mutants affecting germination and early seedling development in maize. Maydica 38:265–274. Gilroy S, Jones DL. 2000. Through form to function: root hair development and nutrient uptake. TIPS 5:56–60. Goddemeier M, Wulff D, Feix G. 1998. Root-specific expression of a Zea mays gene encoding a novel glycine-rich protein, zmGRP3. Plant Mol Biol 36:799–802. He C-J, Morgan PW, Drew MC. 1996. Transduction of an Ethylene signal is required for cell death and lysis in the root cortex of maize during aerenchyma formation induced by hypoxia. Plant Physiol 112:463–472. Hetz W, Hochholdinger F, Schwall M, Feix G. 1996. Isolation and characterisation of rtcs, a mutant deficient in the formation of nodal roots. Plant J 10:845– 857. Hochholdinger F, Feix G. 1998a. Early post-embryonic root formation is specifically affected in the maize mutant lrt1. Plant J 16:247–255. Hochholdinger F, Feix G. 1998b. Cyclin expression is completely suppressed at the site of crown root formation in the nodal region of the maize root mutant rtcs. J Plant Physiol 153:425–429. Hochholdinger F, Feix G. 1998c. Tiller formation in Gaspe Flint is not affected by the rtcs mutation. Maize Genet Newslett 72:30–31. Hochholdinger F, Feix G. 1998d. Isolation of the new necrotic mutant brt1. Maize Genet Newslett 72:29–30.

248 Hochholdinger F, Park WJ, Feix G. 2001. Cooperative action of SLR1 and SLR2 is required for lateral root-specific cell elongation in Maize. Plant Physiology 125:1529–1539. Ishikawa H, Evans ML. 1995. Specilized zones of development in roots. Plant Physiol 109:725–727. Jenkins MT. 1930. Heritable characters of maize XXXIVrootless. J Hered 21:79–80. Kausch W. 1967. Lebensdauer der Primaerwurzel von Monokotylen. Naturwissenschaften 54:475. Kisselbach TA. 1999. The Structure and Reproduction of Corn. 50th Anniversary ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Laskowski MJ, Williams ME, Nusbaum C, Sussex IA. 1995. Formation of lateral root meristems is a two-stage process. Development 121:3303–3310. Larson WE, Hanway JJ. 1977. Corn production. In: Spargue GF, ed. Corn and Corn Improvement. Madison, WI: American Society of Agronomy, pp 625–669. Lebreton C, Lazic-Jancic V, Steed A, Pekic S, Quarrie SA. 1995. Identification of QTL for drought responses in maize and their use in testing causal relationships between traits. J Exp Bot 46:853–865. Lynch J. 1995. Root architecture and plant productivity. Plant Physiol 109:7–13. Malamy JE, Benfey PN. 1997. Organization and cell differentiation in lateral roots of Arabidopsis thaliana. Development 124:33–44. Martin EM, Harris WM. 1976. Adventitious root development from the coleoptilar node in Zea mays L. Am J Bot 63:890–897. McCully ME. 1999. Roots in soil: unearthing the complexities of roots and their rhizospheres. Annu Rev Plant Physiol Plant Mol Biol 50:695–718. McCully ME, Canny MJ. 1988. Pathways and processes of water and nutrient movements in roots. Plant Soil 111:159–170. Neuffer MG. 1994. Mutagenesis. In: Freeling M, Walbot V, eds. The Maize Handbook. New York; SpringerVerlag, pp 212–219. Peterson RL. 1991. Adaptations of root structure in relation to biotic and abiotic factors. Can J Bot 70:661–675. Pilet PE. 1983. Elongation and gravireactivity of roots from an agravitropic maize mutant—implications of growth inhibitors. Plant Cell Physiol. 24:333–336. Richmond T, Sommerville S. 2000. Chasing the dream: plant EST microarrays. Curr Opin Plant Biol 3:108–116. Sass JE. 1977. Morphology. In: Spargue GF, ed. Corn and Corn Improvement. Madison, WI: American Society of Agronomy, pp 89–110.

Feix et al. Scheres B, Wolkenfelt H. 1998. The Arabidopsis root as a model to study plant development. Plant Physiol Biochem 36:21–32. Schiefelbein JW, Benfey PN. 1991. The development of plant roots: new approaches to underground problems. Plant Cell 3:1147–1154. Sheridan WF, Clark JK. 1988. Maize developmental genetics: genes of morphogenesis. Annu Rev Genet 22:353– 358. Thiellement H, Bahrman N, Damerval C, Plomion C, Rossignol M, Santoni V, de Vienne D, Zivy M. 1999. Proteomics for genetic and physiological studies in plants. Electrophoresis 20:2013–2026. Tillich HJ. 1977. Vergleichend morphologische Untersuchungen zur Identitaet der GramineenPrimaerwurzel. Flora 166:415–421. Touzet P, Riccardi F, Morin C, Damerval C, Huet J-C, Pernollet J-C, Zivy M, de Vienne D. 1996. The maize two-dimensional gel protein database: towards an integrated genome analysis program. Theor Appl Genet 93:997–1005. Van Hal NL, Vorst O, Van Houwelingen AM, Kok EJ, Peinenburg A, Aharoni A, Van Tunen AJ, Keijer J. 2000. The application of DNA microarrays in gene expression analysis. J Biotechnol 78:271–280. Varney GT, McCully ME. 1991. The branch roots of Zea. II. Developmental loss of the apical meristem in fieldgrown roots. New Phytol 118:535–546. Varney GT, Canny MJ, Wang XL, McCully ME. 1991. The branch roots of Zea. First order branches, their number, sizes and division into classes. Ann Bot 67:357– 364. Vermeer J, McCully ME. 1982. The rhizosphere in Zea: new insight into its structure and development. Planta 156:45–61. Wang XL, Canny MJ, McCully ME. 1991. The water status of the roots of soil-grown maize in relation to the maturity of their xylem. Physiol Plant 82:157–162. Wang XL, McCully ME, Canny MJ. 1994. The branch roots of Zea. IV. The maturation and openness of xylem conduits in first-order branches of soil-grown roots. New Phytol 126:21–29. Wen T-J, Schnable PS. 1994. Analyses of mutants of three genes that influence root hair development in Zea mays (Gramineae) suggest that root hairs are dispensable. Am J Bot 81:833–843. Yamashita T. 1991. Ist die Primaerwurzel bei Samenpflanzen exogen oder endogen? Beitr Bio Pflanzen 66:371–391. Zheng HG, Babu RC, Pathan MS, Ali L, Huang N, Courtois B, Nguyen HT. 2000. Quantitative trait loci for rootpenetration ability and root thickness in rice: comparison of genetic backgrounds. Genomics 43:53–61.

15 Root Architecture—Wheat as a Model Plant Gu¨nther G. B. Manske and Paul L. G. Vlek Center for Development Research, University of Bonn, Bonn, Germany

I.

INTRODUCTION

However, not all primordia always develop. In wheat, three to six seminal roots normally emerge from the seed, but genetic variations exist among the various cultivars (Robertson et al., 1979). Seed size and seminal root number are positively correlated, although in some genotypes this is not the case (O’Toole and Bland, 1987). The seminal roots constitute 1–14% of the entire root system. They grow and function throughout the whole vegetative period, and penetrate the soil earlier and deeper than the adventitious roots. The wheat crown root system develops on the lower shoot internodes, typically 1–2 cm beneath the soil surface. The seminal root system is very important for the establishment of wheat seedlings. Crown roots begin to develop only at the first foliar node, when the fourth mainstem leaf appears. Root axes (generally, two to four per node) then emerge from successive higher nodes (Klepper et al., 1984). The unique relationships between stem nodes and crown roots permits the establishment of correlations between shoot and root development (Klepper, 1991). The stage of development of the crown root system can be calculated from the phyllochron, which is the unit of time between equivalent growth stages of successive leaves. The phyllochron depends on the air temperature. In wheat, each phyllochron requires
100 growing degree days (Rickman et al., 1984). Adventitious roots mostly occupy the upper soil layers and their number depends mainly on the tillering ability of the plant. Root number and tillering thus are

Root characteristics play an important role in the development of new wheat germplasm with improved drought tolerance, nutrient and water uptake efficiency, lodging resistance, and tolerance to mineral toxicity. These traits are especially relevant for the adaptation of wheat to marginal environments, with increasing demand for wheat production at a time when water and phosphorus are becoming scarce commodities worldwide. This chapter describes the morphology, physiology, ecology, and function of wheat roots, and reviews the current status of knowledge regarding heritability and genetic diversity of root traits of wheat.

II.

MORPHOLOGY AND PHYSIOLOGY OF WHEAT ROOTS

Two root types are distinguished in cereals—the seminal roots (also called primary roots), which develop at the scutellar and epiblast nodes of the embryonic hypocotyl of the germinating caryopsis, and adventitious roots (also called shootborne, nodal, secondary, or crown roots), which subsequently emerge from the coleoptilar nodes at the base of the apical culm and tillers. These two categories of roots function in a complementary manner, and thus the root system must be considered as a whole. The number of seminal roots in cereals is usually five to seven but may reach 10. 249

250

positively correlated (Hockett, 1986). The ratio of seminal to adventitious roots is also altered by the degree of tillering, and consequently by interplant competition. A high-input ideotype of wheat is characterized by a low number of tillers, by a high harvest index, and by seminal root dependence. In contrast, low-input genotypes develop a larger root system, essentially based on adventitious roots exploring a greater soil volume. Indeed, research results at CIMMYT in Mexico showed that the number of tillers was positively correlated with root length density and grain yield of semidwarf bread wheat cultivars (BWs) grown under P deficiency in acid soils (Manske et al., 2000). Tiller numbers did not affect grain yield and root length density when P was amply available. The average root radius of wheat plants ranges between 0.07 and 0.15 mm. The root length density (RLD) varies between 2 and 10 cm cm3 soil depending on the stage of plant development, soil depth, and environmental factors. The horizontal spread of the roots of a wheat plant is usually between 30 and 60 cm. However, under deficient moisture conditions, they invade lower horizons and have a greater vertical distribution (Mishra et al., 1999). The roots can be abundant at soil depths >100 cm with some roots even reaching 200 wheat genotypes grown in different soils and under different water and fertilizer regimes, the roots were always infected by AMF (authors’ unpublished information). In one trial in India, where sewage water was used for irrigation, AMF was absent (Lu¨ttger, 1996). AMF spores are sensitive to the accumulation of heavy metals.

III.

ECOLOGY AND FUNCTION OF WHEAT ROOTS

Extensive root growth is often required for improved growth of shoots and for higher yields in marginal environments. Complex feedback systems between roots and soil take place in the rhizosphere, which commonly comprises 20–30% of the topsoil volume. In compacted soils, roots are shorter, thicker, and more irregularly shaped than the thinner, fibrous roots that develop under noncompacted conditions (Dexter, 1987; Chapter 45 by Masle in this volume). Wheat produces finer roots per unit soil volume under conditions of low nutrient and water supply. Moreover, it exhibits a remarkable plasticity in root growth, adjusting to the soil nutrient and water status (Cholick et al., 1977; Vlek et al., 1996; Chapter 34 by Glass in this volume). Wheat plants can respond to nutrient and water stress via various mechanisms— alternations of root branching and root extension rates (Horst et al., 1996), increased rate of uptake per unit root length or weight (Egle et al., 1999), higher root:shoot ratios (Manske, 1989; Hamblin et al., 1990), more and longer root hairs (Foehse et al., 1991), enhancement of root exudates (Neumann and Romheld, 1999), and AMF infection (Manske et al., 1995)—and by lowering demand for nutrients and water for growth. Each of these parameters can be altered by selection and breeding. Roots tend to proliferate in the soil zones of more favorable nutrient and water supply, but avoid sites of hypoxia or toxic levels of minerals. As a result, wheat roots are more abundant in the upper soil layers which are better aerated and richer in nutrients than deeper horizons. Usually, 70% of the total root length are found in the 0–30 cm soil layer, where the nutrients are concentrated in most agricultural soils. The five

Root Architecture

principal functions of roots are listed below and discussed in relation to wheat. A.

Nutrient Uptake Efficiency

Wheat roots in the upper soil layers have an outstanding capability to absorb nutrients, and most of the nutrient requirements can be satisfied by these roots. Only a fraction of the nutrients are absorbed from the deeper soil (Bole, 1977). However, under P stress, roots will proliferate in the deeper soil profile, allowing access to residual, native P in the deeper soil layer in an Andisol in Mexico (cf. Manske et al., 2000). The geometry of the root system is a key to improvement of nutrient uptake as it may be manipulated to maximize RLD in those places where nutrients and water are more readily available. Wheat plants can only extract immobile nutrients such as P from soil close to the root surface. The quantities extracted are limited by the concentration at the root–soil interface. The soil solution concentrations of phosphate are small (10 years. We are aware of two large aeroponic research facilities that were constructed recently and are now in operation: 1. A relatively large aeroponic facility was constructed at the University of Bielefeld, Germany, by Prof. S.W. Breckle. It constitutes 20 individual planting sites, with a depth of the root chamber of 2.5 m. The plants are watered and nourished through four fogging systems, with the nutrient solution of each pumped and recycled through a pressurized container. A pH-stat controls the pH of the recycled solution, and the temperature of the root chamber is controlled by a water-heating system. 2. A compact aeroponic facility, designed for large-scale mutant screening experiments, was constructed at the Universite´ Katholique de Louvain in Belgium by Dr. Xavier Draye. Plants are placed on top of a 1-m-long, large-diameter plastic tube, with their roots allowed to grow into the tube. A nutrient solution is sprayed through a fine atomizer by a pulse of compressed air. The fogger is located near the top of the tube, and the runoff solution is collected at the base the tube. A small chamber under the collector, installed with two opposite one-way valves, serves as a volumetric pump. A single magnetic valve and a single pressure source are used for simultaneous driving of hundreds of such experimental units. The measurements and analysis of the collected solution allows the calculation of the water and nutrient uptake by the tested plants. An aeroponic facility for commercial growing vegetables was recently established in the Philippines (see http://members.tripod.com/aerogreen/).

V.

Figure 2 Cross section of the Sarah Racine Root Research Laboratory at Tel Aviv University. From top to bottom, 12 m.

SYNOPSIS

The development of each of the roots within a root system is determined by its individual genetic potential as well as by several environmental determinants. As a rule, development can be restricted by the availability of water and nutrients, by the presence of a hostile ionic composition, by the pH, by the gas composition (pO2 and CO2), by the environmental temperature, by the neighboring biota, and by the pressure against soil particles and the friction with them. All those determinants can be controlled in an aeroponic system, thus

Aeroponics

enabling the study of the potential development of root systems in an environment with minimal restrictions. Aeroponics overcomes many of the difficulties that have hampered root research in other growth systems. It is used for observation of root growth and spatial organization without the mechanical, physical, and electrochemical limitations which are imposed by the soil or by an aqueous environment. Aeroponic rhizotrons have several unique advantages: 1.

2.

3. 4.

5.

6.

7.

8.

9.

They enable continuous observation of each of the roots and the measurement of their development and function under a variety of environments. They enable manipulation, observation, and measurement of the growing roots of perennial plants through several annual cycles. Aeroponics enables the distinction, in real time, between various growth patterns of various roots. It enables comparison between the behavior of different roots of a similar/different status, and/or location of one plant. Aeroponic enables sampling of selected roots without causing damage to any of the others. Aeroponics provides information regarding the growth potential of roots, rather than mere measurements of growth under the limitations imposed by the mechanical impedance of the soil or by the aeration stresses of the environment. Aeroponics is the only research system where roots are subjected to a uniform environment without a limiting supply of minerals and without the formation of depletion zones. Roots are grown under what seem to be optimal conditions (high pO2, lack of mechanical impedance, optimal water conditions, good and constant mineral nutrition, etc.). Roots of aeroponically grown plants are free of slimes of fungi and bacteria, so common for roots grown in hydroponics. Root measurement in an aeroponic system is nondestructive; it enables frequent observation of each of the roots of the whole root system. The use of aeroponics enables repetition of measurements of the same roots, thus reducing the sampling errors. Aeroponics give a uniform environment to all the roots of the system, with the possibility to alter the nutrient milieu, and the gas composition of the atmosphere around the roots, very rapidly. It subjects all the roots to a uniform

329

10.

environment but concomitantly enables the induction of spatial changes in the nutrient milieu or the gas composition of the ambient atmosphere. Aeroponic chambers can be constructed in various sizes and can fit the specific requirement and budget limitations of any experimental project.

REFERENCES Barker BTP. 1922. Studies of root development. Long Ashton Res Sta Rep 1921:9–20. Biddinger EJ, Liu CM, Joly RJ, Raghothama KG. 1998. Physiological and molecular responses of aeroponically grown tomato plants to phosphorus deficiency. J Am Soc Hort Sci 123:330–333. Bohm W. 1979. Methods of Studying Root Systems. Berlin; Springer-Verlag. Buer CS, Correll MJ, Smith TC, Towler MJ, Weathers PJ, Nadler M, Seaman J, Walcerz D. 1996. Development of a nontoxic acoustic window nutrient-mist bioreactor and relevant growth data. In Vitro Cell Dev Biol Plant 32:299–304. Burgess T, McComb J, Hardy G, Colquhoun I. 1998. Influence of low oxygen levels in aeroponics chambers on eucalypt roots infected with Phytophthora cinnamomi. J Plant Dis 82:368–373. Burgess T, Hardy GES, McComb JA, Colquhoun I. 1999. Effects of hypoxia on root morphology and lesion development in Eucalyptus marginata infected with Phytophthora cinnamomi. J Plant Pathol 48:786–796. Caba JM, Recalde L, Ligero F. 1998. Nitrate-induced ethylene biosynthesis and the control of nodulation in alfalfa. Plant Cell Environ 21:87–93. Carter WA. 1942. A method of growing plants in water vapor to facilitate examination of roots. Phytopathology 32:623–625. Chung SJ, Chi SH, Shinohara Y, Ikeda H, Suzuki Y. 1933. Effect of misting intervals of nutrient solution on the growth and fruit yield of tomato. J Korean Soc Hort Sci 34:91–98. Clayton MF, Lamberton JA. 1964. A study of root exudates by the fog-box technique. Aust J Biol Sci 17:855–866. Correll MJ, Weathers PJ. 1998. Studies on hyperhydration of Dianthus caryophyllus in an acoustic window mist reactor. In Vitro Cell Dev Biol Anim 34:3. Dodd IC, He J, Turnbull CGN, Lee SK, Critchley C. 2000. The influence of supra-optimal root-zone temperatures on growth and stomatal conductance in Capsicum annuum L. J Exp Bot 51:239–248. du Toit LJ, Kirby HW, Pedersen WL. 1997. Evaluation of an aeroponics system to screen maize genotypes for resistance to Fusarium graminearum seedling blight. Plant Dis 81:175–179.

330 Engenhart M. 1987. Der Einfluss von Bleiionen die Produktivitat und den Mineralstoffhaushalt von Phaseolus vulgaris L. in Hydroponik und Aeroponik. Flora 175:273–282. Eshel A, Zilberstaine M, Waisel Y. 1992. Characterization of various root types of avocado. In: Kutschera L, Huebel E, Lichtenegger E, Persson H, Sobotik M, eds. Root Ecology and its Practical Application. Proc 3rd ISRR Symposium, Vienna, pp 691–694. Finlayson SA, Reid DM. 1996. The effect of CO2 On ethylene evolution and elongation rate in roots of sunflower (Helianthus annuus) seedlings. Physiol Plantarum 98:875–881. Freundl E, Steudle E, Hartung W. 2000. Apoplastic transport of abscisic acid through roots of maize: effect of the exodermis. Planta 210:222–231. Fuchs EE, Livingston NJ, Rose PA. 1999. Structure–activity relationships of ABA analogs based on their effects on the gas exchange of clonal white spruce (Picea glauca) emblings. Physiol Plantarum 105:246–256. Garrido I, Espinosa F, Paredes MA, Alvarez-Tinaut MC. 1998a. Effect of some electron donors and acceptors on redox capacity and simultaneous net H+/K+ fluxes by aeroponic sunflower seedling roots: evidence for a CN-resistant redox chain accessible to nonpermeative redox compounds. Protoplasma 206:141–155. Garrido I, Espinosa F, Paredes MA, Alvarez-Tinaut MC. 1998b. Net simultaneous hydrogen and potassium ion flux kinetics in sterile aeroponic sunflower seedling roots: effects of potassium ion supply, valinomycin, and dicyclohexylcarbodiimide. J Plant Nutr 21:115– 137. Henzler T, Waterhouse RN, Smyth AJ, Carvajal M, Cooke DT, Schaffner AR, Steudle E, Clarkson DT. 1999. Diurnal variations in hydraulic conductivity and root pressure can be correlated with the expression of putative aquaporins in the roots of Lotus japonicus. Planta 210:50–60. Hessel MI Jr, Reichert GE Jr, Nevill GE Jr. 1933. Airflowcontained aeroponic nutrient delivery for a microgravity plant growth unit. Biotronics 21:33–38. Hoffman TK, Kolb FL. 1997. Effects of barley yellow dwarf virus on root and shoot growth of winter wheat seedlings grown in aeroponic culture. Plant Dis 81:497– 500. Hung LLL, Sylvia DM. 1988. Production of vesicular-arbuscular mycorrhizal fungus innoculum in aeroponic culture. App Environ Micribiol 54:353–357. Ingestad T, Lund A-B. 1979. Nitrogen stress in birch seedlings. I. Growth technique and growth. Physiol Plantarum 45:137–148. Jackson TJ, Burgess T, Colquhoun I, Hardy GES, Jackson TJ. 2000. Action of the fungicide phosphite on Eucalyptus marginata inoculated with Phytophthora cinnamomi. Plant Pathol 49:147–154.

Waisel Jarstfer AG, Farmer-Koppenol P, Sylvia DM. 1998. Tissue magnesium and calcium affect arbuscular mycorrhiza development and fungal reproduction. Mycorrhiza 7:237–242. Jie H, Kong LS. 1998a. Growth and photosynthetic characteristics of lettuce (Lactuca sativa L.) under fluctuating hot ambient temperatures with the manipulation of cool root-zone temperature. J Plant Physiol 152:387– 391. Jie H, Kong LS. 1998b. Growth and photosynthetic responses of three aeroponically grown lettuce cultivars (Lactuca sativa L.) to different rootzone temperatures and growth irradiances under tropical aerial conditions. J Hort Sci Biotechnol 73:173–180. Klotz LGA. 1944. A simplified method of growing plants with roots in nutrient vapors. Phytopathology 34:507–508. Lee SK, Cheong SC. 1996. Inducing head formation of iceberg lettuce (Lactuca sativa L.) in the tropics through root-zone temperature control. Trop Agric 73:34–42. Martin NE, Hendrix JW. 1967. Comparison of root systems produced by healthy and stripe rust-inoculated wheat in mist-, water-, and sand-culture. Plant Dis Rep 51:1074–1076. Martin-Laurent F, Lee SK, Tham FY, He J, Diem HG, Durand P. 1997. A new approach to enhance growth and nodulation of Acacia mangium through aeroponic culture. Biol Fertile Soil 25:7–12. Martin-Laurent F, Lee SK, Tham FY, Jie H, Diem HG. 1999. Aeroponic production of Acacia mangium saplings inoculated with AM fungi for reforestation in the tropics. For Ecol Manage 122:199–207. Martin-Laurent F, Tham FY, Lee SK, He J, Diem HG. 2000. Field assessment of aeroponically grown and nodulated Acacia mangium. Aust J Bot 48:109–114. Mohammad A, Khan AG, Kuek C. 2000. Improved aeroponic culture of inocula of arbuscular mycorrhizal fungi. Mycorrhiza 9:337–339. Nir I. 1982. Growing plants in aeroponics growth system. Acta Hort 126:435–445. Padgett PE, Leonard RT. 1993. Contamination of ammonium-based nutrient solutions by nitrifying organisms and the conversion of ammonium to nitrate. Plant Physiol 101:141–146. Peterson LA, Krueger AR. 1988. An intermittent aeroponics system. Crop Sci 28:712–713. Quintana JM, Harrison HC, Palta JP, Nienhuis J, Kmiecik K, Miglioranza E. 1999. Xylem flow rate differences are associated with genetic variation in snap bean pod calcium concentration. Studies of Ca nutrition. J Am Soc Hort Sci 124:488–491. Rao A, Gritton ET, Grau CR, Peterson LA. 1995. Aeroponic chambers for evaluating resistance to Aphanomyces root rot of peas (Pisum sativum). Plant Dis 79:128–132.

Aeroponics Sachs J. 1874. Ueber das Wachstum der Haupt- und Nebenwurzlen. Arbeiten Bot Inst Wurzburg 4:586– 589. Shimura K. 1970. Root research phytotron. Jpn Agric Res Q 5:54–57. Smucker AJM, Erickson AE. 1976. An aseptic mist chamber system: a method for measuring root processes of peas. Agron J 68:59–62. Sylvia DM, Hubbell DH. 1986. Growth and sporulation of vesicular–arbuscular mycorrhizal fungi in aeroponic and membrane systems. Symbiosis 1:259–267. Shtrausberg DV, Rakitima EG. 1970. On the aeration and gas regime of roots in aeroponics and water culture. Agrokhimiya 4:101–110. Takahashi H, Mizuno H, Kamada M, Fujii N, Higashitani A, Kamigaichi S, Aizawa S, Mukai C, Shimazu T, Fukui K, Yamashita M. 1999. A spaceflight experiment for the study of gravimorphogenesis and hydrotropism in cucumber seedlings. J Plant Res 112:497– 505. Varney GT, Canny MJ. 1993. Rates of water uptake into the mature root system of maize plants. New Phytol 123:775–786. Vincenzoni A. 1979. Aeroponics: method of soilless culture (hydroponics). Coltura Protette 8:51–59. Vyvyan MC, Travell GF. 1953. A method of growing trees with their roots in a nutrient mist. Annu Rep East Malling Res Station, pp 95–98. Wagner RE, Wilkinson HT. 1992. An aeroponics system for investigating disease development on soybean taproots

331 infected with Phytophthora sojae. Plant Dis 76:610– 614. Wagner RE, Carmer SG, Wilkinson HT. 1993. Evaluation and modelling of rate-reducing resistance of soybean seedlings to Phytophthora sojae. Phytopathology 83:187–192. Waisel Y. 1988. Aeroponic observatories: useful tools for teaching and for long term studies of root behavior. Proc 4th ISSR Meeting, Uppsala, Sweden, 4:27. Waisel Y, Breckle SW. 1987. Differences in responses of various radish roots to salinity. Plant Soil 104:191–194. Ycas JW, Zobel RW. 1983. The response of maize radicle orientation to soil solution and soil atmosphere. Plant Soil 70:27–35. Zimmermann HM, Hartmann K, Schreiber L, Steudle E. 2000. Chemical composition of apoplastic transport barriers in relation to radial hydraulic conductivity of corn roots (Zea mays L.). Planta 210:302–311. Zobel RW. 1989. Steady state control and investigation of root system morphology. In: Torrey JG, Winship LJ, eds. Applications of Continuous and Steady State Methods to Root Biology. Dordrecht, Netherlands: Kluwer, pp 165–182. Zobel RW, Del Tredici P, Torrey JG. 1976. Methods for growing plants aeroponically. Plant Physiol 57:344– 346. Zsoldos F, Vashegyi A, Erdei L. 1987. Lack of active K+ uptake in aeroponically grown wheat seedlings. Physiol Plantarum 71:359–364.

20 Use of Microsensors for Studying the Physiological Activity of Plant Roots D. Marshall Porterfield University of Missouri-Rolla, Rolla, Missouri

I.

INTRODUCTION

To fully understand the physiological activity of plant roots, one must understand the physical and chemical properties of the rhizosphere immediately near the root surface. For the root this is the direct environment for physiological interaction with the bulk soil. The attributes of the rhizosphere are determined not only by the soil, but also by the biochemical and physiological activity of the root. The properties of the rhizosphere can vary in size, shape, and composition, even from root to root on the same plant. Root biologists have sought to explore the rhizosphere and understand plant-mediated rhizosphere activities by utilizing microsensors to probe this environment. These sensors have included various types of ion-selective microelectrodes, as well as polarographic electrochemical oxygen sensors to determine the concentration of important molecules within the rhizosphere. Advanced microsensor techniques now make it possible to measure dynamic flux of these molecules in real time with a relatively high degree of spatial and temporal resolution. This technique has been referred to as the microelectrode ion flux estimation (MIFE) technique, the vibrating probe (VP), and more recently the self-referencing microsensor (SRM) technique. The use of such microsensors and microsensor techniques to study the plant root and rhizosphere processes is discussed in this chapter.

II.

THE ION-SELECTIVE MICROELECTRODE (ISM)

A.

ISM Basics

For this general class of sensors the term ‘‘selective’’ is preferred because the sensors tend to favor one ion species over others. In some cases the sensors are specific, but that is rare and care must always be taken to consider the sensor’s capabilities and limitations. Ionselective electrodes are constructed based on the ion selective properties of various inorganic soluble salts, organic ion exchangers (ionophores), and glass materials. These ion-selective components are used to create an effective barrier or membrane to separate the electrode’s internal electrolyte solution from the solution being measured. This ion-selective membrane is electrically capacitive, and the voltage potential that develops across the membrane is based on the ion concentration differential across the membrane. This voltage potential is measured using a pair of Ag/ AgCl electrodes on each side of the electrode membrane, where one of the electrodes is contained within a defined solution separated by the membrane from an undefined medium into which the second electrode (reference electrode) is placed. Ion-selective microelectrodes are constructed utilizing ionophore-based membranes. The ionophores are typically incorporated into lipophylic ion exchange (LIX) cocktails that are in most cases commercially 333

334

Porterfield

available. The details of how to build these types of microelectrodes have been described (Amman, 1986; Miller, 1995; Smith et al., 1999), and a brief description of the protocols is reviewed in Fig. 1. For intracellular probes the liquid membranes of the electrodes can be plasticized using polyvinyl chloride (PVC), but this tends to slow the response time of the sensor. For non-invasive measurements of ion activities around plant roots this is not necessary. Since the selectivity of the sensor is mediated by the properties of the LIX, the basic electrode construction (Fig. 1) can be utilized to build any number of ion-selective microelectrodes simple by changing the electrode electrolyte backfill solution and the LIX (Table 1). The response of these electrodes is commonly termed to be ‘‘Nernstian’’ as the relationship between ion concentration and the resulting electrode potential conforms to the Nernst equation (28 mV or 56 mV change in electrode potential per 10-fold change in divalent or monovalent ion concentration, respectively). B.

Use of ISM Techniques to Study Roots and Root Cells

There has been fairly broad application of ISMs in measurements of ion activities in, or at the surface

of, plant cells and tissues. This is in contrast to the use of ISMs as noninvasive self-referencing flux sensors as discussed in subsequent sections. Discussion here will be limited to the use of ISMs as invasive sensors to probe the activities of ions in individual root cells and in root tissues. That is not to say that the measurement of ion activities at the root surface using a static probe does not provide valuable information. In fact a recent study has shown a strong relationship between surface pH and profiles of elongation in maize roots (Peters and Felle, 1999). This study does, however, illustrate some limitations in the use of static sensors to characterize the surface of a root. Static measurements do not provide any direct signal relating to the polarity of ion movements. Thus, pH scale changes may not accurately describe the magnitude of physiological ion flux activity from the root and significant differences in ion transport activity may be overlooked (Table 2). Despite the limitations of the use of static ISMs, they definitely are valuable tools. While surface concentration and transport activity can be best measured using self-referencing ion-selective (SRIS) microelectrodes, static ISMs are effectively used for determination of intracellular ion concentrations, and in some cases offer a good alternative to the use of ratiometric

Figure 1 Diagram depicting the construction and calibration of a calcium-selective microelectrode. The specificity of this microelectrode is based on the characteristics of the commercially available lipophyllic ion exchange cocktail (Ca2+ Ionophore I Cocktail A) that forms an organic phase liquid membrane barrier in the tip of a silanized glass microelectrode. This membrane separates the internal electrolyte (100 mM CaCl2) from the solution being measured. The selectivity of an ionophore is defined by the Nikolsky-Eisenman equation (Ammann, 1986). Other types of ISMs are constructed filling the glass micropipette with the appropriate backfill solution and tipped with a membrane made from a particular ion-selective cocktail (Table 1). All of these types of microelectrodes have Nernstian calibration responses; a 10-fold change in concentration generates a constant voltage difference.

Microsensors

335

Table 1 Ionophore Cocktails and Appropriate Backfill Solutions for Several Ion-Selective Membranes Used to Construct Ion-Selective Microelectrodes for Use as a Self-Referencing Sensor Ion-selective Membrane

Column length (mm)

Backfill solution

200 30 150 30 80 200 200 50 100

500mM NH4C1 100 mM CaCl2 100 mM NaCl 100 mM KCl 500 mM MgCl2 500 mM KNO3 500 mM KNO2 100 mM KCl 100 mM NaCl

Ammonium Fluka cat. #09879 Calcium Fluka cat. #21048 Chloride Fluka cat. #24902 Hydrogen Fluka cat. #95293 Magnesium Fluka cat. #63048 Nitrate Fluka cat. #72549 Nitrite Fluka cat. #72549 Potassium Fluka cat. #60398 Sodium Fluka cat. #I7397

dyes. This would be the case when microinjection is required to get the dye into the cell or when there is not a dye available for the particular ion of interest. Developmentally important gradients in the concentration of calcium in root hairs of Sinapis alba were measured using a combination of intracellular calcium ISM and ratio imaging. Generally, the two techniques showed good agreement. Using ISMs’ intracellular calcium levels have also been measured in individual root hairs and in the stele of maize roots (Felle, 1998) and in root hairs of alfalfa and soybean during the transduction of the Nod factor signal (Felle et al., 1999a,b). The use of ISMs really has an advantage over ratio metric dyes is in the ability to measure more than one ion at a time using separate and multi barreled microelectrodes. Both H+ and Ca2+ and H+ and Cl concentrations were measured in a study of tip growth in root hairs of Sinapis alba (Felle, 1994; Herrmann and Felle, 1995) using multibarreled microelectrodes that combine measures of the separate ions with measures of membrane potential. Triple-barreled electrodes have

also been used to combine measurements of multiple ions and membrane potential into a single unit (Walker et al., 1995). In such a unit H+, membrane potential, and either K+ or NO3 were measured simultaneously in barley root epidermal cells. III.

POLAROGRAPHIC ELECTROCHEMICAL SENSORS

A.

Theory and Operation of Electroanalytical Sensors

Electrochemical sensors mediate analyte detection by reducing or oxidizing an analyte. The redox changes in the analyte are driven by a polarized potential on the electrode’s sensing surface that facilitates electron transfers between the surface and the molecule being detected. The flow of electrons is therefore dependent on the concentration of the analyte, and these sensors typically show a linear relationship between analyte concentration and the resulting current (Bard and

Table 2 Relationship Between pH and H+ in the Bath and at the Root Tip as Measured Using a H+ Ion-Selective Microelectrode Bath pH 7.8 6.5 4.2

Root tip surface pH

Bath H+ (mM)

Root tip surface H+ (mM)


pH (root tip–bath)


(H+) (root tip–bath) (mM)

6.7 5.8 4.45

0.015848932 0.316227766 63.09573445

0.199526231 1.584893192 35.48133892

1.1 0.7 0.25

0.18368 1.26867 27.6144

Notice that the largest changes in H+ are associated with relatively small pH changes in the range of 4.2 and 4.45 pH. It is also important to note that the decrease in pH occurring in the bath at pH 7.8 and 6.5 relates to an increase in the H+ ion concentration, presumably due to H+ efflux and H+/ATPase activity. The pH increases (decrease in H+) associated with the pH 4.2 bath is conversely presumed to be associated with a complete reversal of the direction of H+ ion flux back into the root. Source: Peters and Felle (1999).

336

Faulkner, 1980). Since the primary electronic signal that relates to the analyte concentration is a current, these sensors are sometimes referred to as amperometric sensors. This is in contrast to ion-selective electrodes where the primary electronic signal is a voltage potential and the term voltametric sensor is often used. Specificity of analyte detection by electroanalytical sensors is mediated by the redox properties of the analyte, the polarization potential of the electrode, and membrane coatings on the electrode surface that may facilitate phase or size exclusion of potential interferants. Electroanalytical sensors have been widely applied in animal and biomedical research. These techniques have been used to measure compounds ranging from simple gases (oxygen and nitric oxide [NO]) to complex hormones like insulin (Huang et al., 1995). In

Porterfield

plant sciences, electroanalytical techniques have been applied to measure oxygen in relation to photosynthesis, respiration, and soil oxygenation (Armstrong, 1994). There are basically only two types of electroanalytical oxygen electrodes—the Whalen- and the Clark-style microelectrodes (Fig. 2). It is important to note that all polarographic oxygen sensors are variations of these two basic designs. The Whalen-style electrode is a platinum or gold cathode that is directly membrane coated and that utilizes a separate reference electrode, whereas the Clark electrode has the noble metal cathode (Au or Pt) and the reference electrode contained within a KCl solution behind a common membrane. The membrane is important in both of these designs as it acts not only to protect the noble metal surface from fowling by compounds in the measuring media (most commonly proteins and

Figure 2 Construction and calibration of Clark (A) and Whalen (B) style oxygen-selective microelectrodes. The Clark oxygen electrode is constructed by building a noble metal (Au or Pt) cathode and sealing this inside of a membrane-tipped electrolyte containing micropipette body along with an Ag/AgCl reference electrode. The Whalen style electrode is simply a noble metal cathode with a recessed tip that contains an oxygen-permeable membrane. For either style electrodes, the cathode can be constructed by filling a micropipette with a low melting point alloy such as Wood’s metal. This metal-filled electrode is then etched and plated with gold in order to create an electrode with the desired tip geometry. In the case of the Whalen electrode, the recessed tip is filled with Pt or gold but gold is reported to be less susceptible to redox changes itself when reducing oxygen to water. Finally, the electrode tip recess is filled with a gas-permeable membrane. The electrode is calibrated (C) against solutions of known oxygen concentrations made by bubbling with gases of known oxygen partial pressures.

Microsensors

sulfides), but it alters the oxygen consumption (resistance) of the electrode in such a way as to render the electrode stir insensitive. This quality of stir insensitivity is very important because the electrode, in the process of reducing oxygen to water, is consuming oxygen. If the consumption of oxygen by the electrode were not modified in this way, the electrode could potentially alter the concentration of oxygen in the local area where it is actually measuring. The membrane effectively solves this problem by increasing the resistance to oxygen transport through the membrane. A properly functioning electrode should have 99% of the electrode oxygen depletion gradient contained entirely within the electrode’s membrane (Sneiderman and Goldstick, 1978). Some Whalenstyle electrodes utilize a deep recess in the electrode tip that is not membrane coated to contain the depletion gradient. While this does effectively deal with the problem of stir sensitivity, it does not adequately protect the electrode surface from fowling.

B.

Use of O2-Selective Microelectrodes in Root Research

Roots are invariably composed of heterogeneous cell types characterized by different levels of metabolic activity. Understanding cellular and tissue metabolism in roots turns out to be the key to understanding normal physiology and various types of stress responses. This can be accomplished using spatial oxygen measurements that will provide a general indicator of many aspects of cellular metabolism and physiological status. Davies and Brink (1942) first described the use of oxygen microelectrodes for direct measurement of oxygenation biological tissues. Later the electrodes were miniaturized with the goal of measuring intracellular O2 concentrations (Whalen et al., 1967). Electrode designs have been subject to further development (Forstner and Gnaiger, 1983) and widespread use, including the study of metabolism and aeration in plant roots (Armstrong, 1994) and in plant aerial tissues (Porterfield et al., 1999). The first notable use of an oxygen electrode to study root tissue oxygenation was by Bowling (1973), who studied the oxygen concentration profiles across excised sunflower roots. He used a commercial electrode (tip diameter 1 mm) and revealed a shallow radial oxygen gradient across the root cortex. Tjepkema and Yocum (1974) studied oxygenation in soybean root nodules using glass-insulated Pt microelectrodes and documented that the major resistance to oxygen trans-

337

port in root nodules lies in the cortex. This was followed by studies of oxygen distribution in the root nodules of pea and French bean (Witty et al., 1987) and of oxygen profiles in the barley rhizosphere (Hojberg and Sorensen, 1993). Much of what we know about root oxygenation and oxygen transport during hypoxia and anoxia come from the use of in vivo oxygen microelectrodes. Studies of radial oxygen distribution in maize roots have showed that during low oxygen stress, oxygen was supplied from the shoot (Armstrong et al., 1993). Steep radial diffusion gradients were characteristic of the nonporous epidermal/hypodermal shell and stele, while shallow profiles were generally found in the cortex (Armstrong et al., 1990). Anoxia in the stele was inducible by manipulating the oxygen concentrations around the shoot (Armstrong et al., 1990). Measurement of oxygen profiles obtained by slowly advancing a microelectrode through the meristemic zone and the root cap of roots grown in stagnant anaerobic medium, suggests that the bulk of the root tip tissue is likely to be anoxic. (See also Chapter 42 by Armstrong and Drew in this volume). Root oxygenation in solution and vermiculite grown maize roots were studied using a Clark-style microelectrode (Ober and Sharpe, 1996). Treatments used to alter water potential in solution grown maize roots by application of polyethylene glycol reduced the availability of oxygen in the medium by altering the solubility and diffusion of oxygen in the system. This reduction in oxygen availability produced substantial decreases in root tissue oxygenation (Versules et al., 1998). This suggests that experimental artifact associated with simulated water stress could occur, and indeed it was noted that the hypoxic enzyme alcohol dehydrogenase (EC 1.1.1.1) is induced under simulated water stress situation.

IV.

THE SELF-REFERENCING MICROELECTRODE (SRM) TECHNIQUE

A.

Background and History

A very important consideration in the use of any microsensor for studying biological transport activity is the mode of operation of the sensor. The use of static measurements to characterize a component of the roots dynamic physiology are thus subject to limitations and sources of artifacts as previously discussed. In many cases the signal-to-noise characteristics of the recording instrumentation and electrodes hinder the

338

ability to detect the intricate biological patterns that are present. The use of invasive microelectrodes can introduce experimental error (Silver, 1967; Whalen, 1974; Schneiderman and Goldstick, 1978), the most significant of which in plants is the induction of an oxidative burst in the tissue being probed (Lamb and Dixon, 1997). Such limitations in the use of microelectrodes are in many ways overcome by the self-referencing microelectrode (SRM) technique. This technique allows noninvasive monitoring of physiological transport activity (flux) from single cells or whole tissues in real time. Apparently this approach has been reinvented several times, and with each new rediscovery a new name has been coined. The basic approach has been referred to as the vibrating probe, the self-referencing microelectrode, the microelectrode ion flux estimation (MIFE), and the microelectrode flux estimation (MFET) techniques. What these sometimes subtle variations all have in common is that a single microelectrode is used to sample two distinct positions within a biologically derived gradient. The self-referencing approach (Zisman, 1932) used a metal electrode as a probe to measure low-magnitude currents by determining voltage differences between two distinct points in an electric field using a single electrode. This is what is specifically referred to as the vibrating voltage probe and is used to measure bulk ion fluxes. Despite the nonspecificity of the technique this approach has provided a practical method to show total electrical currents in a number of biological systems. The first biological use of the technique was for plant research (Bluh and Scott, 1950). Later it was used to study gravitropic responses of corn coleoptiles (Hertz, 1960; Grahm, and Hertz 1962, 1964; Grahm, 1964). It was even used to measure net ion fluxes on skeletal muscle fibers (Davies, 1966). A notable rediscovery of the vibrating probe (Jaffe and Nucitelli, 1974) resulted in significant improvements of the technique based on the use of a commercial lock-in amplifier. The vibrating voltage probe was used to investigate the presence of ionic currents during barley and clover root development (Weisenseel et al., 1979; Miller et al., 1986), gravitropism in cress roots (Weisenseel et al., 1992), and wounding responses in Nicotiana (Miller et al., 1988) and pea roots (Hush et al., 1992). The limitation of the self-referencing vibrating voltage probe is the inability of the sensor to determine the actual ionic makeup of the measured biological currents. The ability to measure the active flux of specific ions came with the development of the self-refer-

Porterfield

encing ion-selective (SRIS) microelectrode. This approach differs from the vibrating probe in that it is based on the use of ion-selective microelectrodes instead of the original metal electrodes, and the frequency of probe movement is substantially slower (0.1–0.3 Hz vs 300–1000 Hz). Initially, investigators drove the probe movements manually to sample the biological gradients (Newman et al., 1987). Later, Khu¨trieber and Jaffe (1990) used computer-driven electrode movement and data acquisition to automate the measurement of calcium ion flux. The automated SRIS approach was subsequently diversified to measure other ions like H+ and K+ (Kochian et al., 1992) and even heavy metals like Cd2+ (Pineros et al., 1998). The self-referencing technique has also recently been used to measure the biological flux of nonionic compounds through the development of the self-referencing electrochemical microelectrode (SREM) technique (Porterfield et al., 1998; Land et al., 1999; Porterfield and Smith, 2000).

B.

SRM Theory and Operation

As previously stated, the goal of the use of SRM is to actively measure compounds fluxing near the surface of single cells or whole tissues based on the diffusionary movement of molecules within the gradient (Newman et al., 1987; Khu¨trieber and Jaffe, 1990; Shabala et al., 1997; Porterfield et al., 1998; Land et al., 1999; Porterfield and Smith, 2000). This is accomplished by controlling the translational movement of a selective microelectrode in a gradient. A recording system is constructed using a microscope and a head-stage electrode amplifier driven by a translational motion control system. Such an assembly is mounted on an antivibration table and housed within a Faraday cage. The amplifier and the electrical components of the motion control system are commercially available, whereas in the past automated systems were custom built and required custom software development. The motion control system allows the electrode tip to be moved through the gradient at a known frequency and between known points (commonly 10–50 mm apart). The automated electrode movement induces a phasedependent waveform on the electrode output signal, where the amplitude of said waveform is proportional to the differential analyte concentration within the gradient (Fig. 3). This effectively turns a static concentration sensor into a dynamic flux sensor, all while minimizing the impact of random noise and drift on the differential electrode output (Fig. 4).

Microsensors

339

Figure 3 Diagrams showing the basic procedure for conducting a self-referencing microelectrode flux measurement. In panel (A) a schematic of an experiment is depicted where a generated diffusional gradient is established outside of a micropipette. The relationship between probe positions (A) and electrode output (B) during the probe movement cycle is plotted to illustrate how a waveform is induced on top of the electrode output signal. For automated systems a PC typically controls the translational frequency allowing signal analysis software to extract voltage changes in phase with the probe moving through the gradient. The relationship between the concentration differential (C1–C2) is directly proportional to the amplitude of the waveform of the differential electrode output value (
fA for amperometric electrochemical sensors and
mV for potentiometric ion selective sensors).

There are two main approaches to acquiring and analyzing amplitude of the waveform or the electrode differential signal. Some of the first automated systems utilized an approach sometimes referred to as AC coupling. Here the electrode signal coming from the preamplifier (mV) is passed through a capacitor

that nondiscretely subtracts off the baseline output signal so that the differential waveform signal (mV) can be amplified enough to digitally analyze these minute signals. The problem with this approach is that the capacitor discharges exactly when the differential signal is being analyzed, invariably resulting in

Figure 4 Data showing how the self-referencing technique effectively filters out noise based on the basic principles of phasesensitive detection. For this experiment a SRIS-H+ sensor was operated at a frequency of 0.3 Hz and over an excursion distance of 10 mm. The artificial gradient was created outside of a micropipettes with a 10-mm tip diameter that was filled with a pH 4 buffer immobilized in an agar gel matrix (1.0%). Data were collected only during the time when the probe was stopped at the two positions and the data that were collected are plotted here as a scatter plot. This clearly shows that the data segregate into two distinct groups representing the two measuring positions. Note how electrode drift and noise patterns are common to both groups of data while the differential (
V) between the two populations stays relatively constant.

340

degradation of the differential signal (Khu¨trieber and Jaffe, 1990), and in an underestimation of the true flux values (Smith et al., 1999). Despite the fact that the signal degradation problem was described and quantified a decade ago (Khu¨trieber and Jaffe, 1990), the approach is still used today (Smith et al., 1999). The problem can be overcome by subtraction of the baseline signals before amplification. This is referred to as DC coupling, and it allows for baseline signal subtraction without the associated problems of differential signal degradation due to capacitive discharge (Shipley and Fiejo, 1999). Once the measurements of the differential electrode signals within the gradient have been acquired, they must be converted into a measure of differential concentration before flux can be calculated. The formulas for doing the conversion from differential electrode output to differential concentration are another area where the AC-coupling and DC coupling approaches diverge. Because of the signal degradation and the way that the AC-coupled method logs data, it is impossible to directly calculate these values (Smith et al., 1999). Because the DC-coupling discretely subtracts off the baseline signal, this allows for logging of the actual electrode output values at both measuring positions. It thereby allows calculation of these concentration differentials without having to make any types of assumptions regarding background electrode signals and measurement efficiency. Flux calculations are all based on the Fick equation J ¼ Dð
C=
XÞ where J ¼ flux, D is the diffusion coefficient,
C is the differential concentration, and
X is the distance between the two electrode measuring positions. This version of the formula assumes flat planar geometry of the surface being analyzed. Some researchers, using a variation of the Fick equation, have gone to great lengths to calculate flux using cylindrical diffusion geometry for the plant roots. Obviously a geometric cylinder is an oversimplification of a plant root, especially when considering a high-resolution scan of the heterogeneous root tip. This approach was compared to the planar flux model using actual root recording data and quantitatively shown to be unnecessary (Kochian et al., 1992). The planar model yielded almost identical results to the cylindrical model, as would be expected given the small electrode tip size and excursion distance, and the fact that any surface can be described mathematically as a series of flat planar surfaces. When possible, the less complex planar approach is preferred as estimations of some of

Porterfield

the values required for the cylindrical formula may actually introduce analytical error.

C.

Self-Referencing Ion-Selective (SRIS) Microelectrode Root Studies

The SRIS technique has been used in investigations of various plant systems, including studies of ion transport mechanisms, in plant roots. All plant cells have sophisticated ion transport mechanisms, but the structural characteristics of the cell wall and the presence of a large osmotically active central vacuole limit the application of standard intracellular electrodes and electrophysiological studies of these processes. Given the noninvasive nature of the SRM technique, membrane ion transport properties can be measured in real time without disrupting the cell or tissue. Roots and root hairs have been the subject of many important studies. Calcium and hydrogen ion fluxes have been measured in the root hairs of Sinapis alba and shown to correlate with intracellular events and other properties of growth (Hermann and Felle, 1995; Felle and Hepler, 1997). Root hair calcium fluxes have been investigated in relation to normal development (Schiefelbein et al., 1992) in Arabidopsis, and under the influence of the Nod factor in leguminous roots (Allen et al., 1994). Calcium fluxes in the root apex and in root hairs of wheat and in Limnobium stonoloniferum were studied under the influence of aluminum in toxic concentrations (Huang et al., 1992; Jones et al., 1995). Proton and potassium flux in maize roots and maize suspension cells (Kochian et al., 1992) has been characterized. Ammonium and nitrate ionophores have been used to measure and map the uptake of these ions along the axis of a barley root (Henricksen et al., 1992), and ammonium, nitrate, and proton fluxes have also been measured along maize roots (Taylor and Bloom, 1998). Ionophores have recently been developed for the measure of heavy-metal ions, and a cadmium-selective SRIS was used for plant toxicity studies (Pineros et al., 1998) by measuring the flux of cadmium into the roots of Thlaspi and wheat. Using advanced hardware and software that includes digital electrode position tracking with video feedback, high-resolution scans of ion flux activities are now possible (Fig. 5). A root tip of Typhya latifolia was scanned using an H+ ion-selective microelectrode. The electrode positioning data was used to reconstruct the

Microsensors

341

Figure 5 H+ flux around an actively growing root tip of Typha latifolia. Flux measurements were made using a SRIS-H+ sensor (10 mm excursion/0.3 Hz). At closest approach the probe was within 1 mm of the root surface and the excursion distance of 10 mm was entirely within the rhizosphere gradient as shown in a step-back experiment (data not shown). The experiment was done using an IPA-2 DC-coupled amplifier (Applicable Electronics, Forestdale, MA) and a PC running ASET software (ScienceWares, Falmouth, MA). The 3D profile of the root tip was reconstructed using the microelectrode positioning data that were collected along with the electrode output by the computer data acquisition system.

3D root surface and to plot a contour map of the flux from the surface. D.

Self-Referencing Electrochemical Microelectrode (SREM) Techniques

The SREM sensors are a recent development (Porterfield et al., 1998; Land et al., 1999; Porterfield and Smith, 2000) that was originally based on the use of a Whalen-type oxygen microelectrode used as a selfreferencing sensor (SREM-O2). Given that this is a recent development, it has seen only limited application in plant science research but already has demon-

strated very high sensitivity and special resolution. Oxygen and proton fluxes have been studied in single algal cells of Spirogyra and Acetabularia using a combination of probes (Porterfield and Smith, 2000; Serikawa et al., 2000, 2001) and demonstrated the subcellular resolution of the technique. The SREM-O2 sensor has been used in a study of metabolic oxygen consumption patterns along the root of Elodea (Mancuso et al., 2000). However, care must be taken in interpreting these results because a stir-sensitive bare cathode was used as an oxygen sensor. The SREM-O2 oxygen microsensor has great promise as an important tool for root physiology research.

342

Like other SRM sensors it is noninvasive and allows for direct measurement of rhizosphere oxygen transport into and even out of roots. Perhaps the most significant applications of the sensor will come from

Porterfield

simultaneous SREM-O2/SRIS recordings. An example of such application can be seen in Fig. 6. High-resolution scans of the surface of a maize root were done using a combination of oxygen- and proton-

Figure 6 Metabolic oxygen flux patterns and H+ flux around the roots of Zea mays (cv Kandy Korn) as measured using a dual sensor arrangement. Here the plants were exposed to different forms of nitrogen (A=200 mm (NH4)2SO4, 100 mM Ca2SO4; B=200 mm Ca(NO3)2, 100 mM Ca2SO4), and the resulting differences in flux patterns were scanned using two separate sensors mounted together (tips within 2 mm of one another) on a single computer-driven motion control system. The two separate preamplifiers were operated using a dual-channel DC-coupled amplifier (model IPA-2, Applicable Electronics, Forestdale, MA) and a PC running ASET software (ScienceWares, Falmouth, MA). When the roots are exposed to nitrogen in the form of NH4, the H+ efflux values are typically higher, along the entire root, than the root exposed to NO3. Although the flux patterns were virtually identical in the two treatments, the H+ efflux values ranged between 10 to 20 pmoles cm2 sec1 along the NH4 roots whereas the NO3-exposed roots had flux values that ranged between –4 and 14 pmoles cm2 sec1. Note that the negative values associated with the positions in the range of 600–2400 mm from the root tip indicate that this was a zone of net H+ influx. Oxygen consumption patterns were again very similar in the two treatments; however, the flux values tended to be almost three times higher in the NO3-exposed roots. The differences in root metabolic oxygen influx and H+ flux are thought to be related to the additional metabolic energy and H+ required to reduce NO3 to NH4.

Microsensors

selective microelectrodes. The sensors were mounted together and driven using a single motion control system. The flux patterns were mapped around roots when exposed to either ammonium or nitrate. Results reveal significant differences in hydrogen ion flux and metabolic oxygen consumption resulting from the differences in transport and assimilation of these different nitrogen forms. While the SREM-O2 sensor will be an important tool, perhaps the most significant use of the SREM sensor technology will result from application of other types of electroanalytical sensors. Already the SREM technique has been expanded to include sensors for NO, ascorbic acid, and hydrogen peroxide (Heck et al., 1999; Kumar et al., 1999; Pepperell et al., 1999a,b; Billack et al., 2001; Kumar et al., 2001; Porterfield et al., in press). While these sensors have not been used in plant research, they do have great potential as tools for root research, possibly advancing our understanding in many fundamental areas. NO has been suggested to be involved in plant signaling (Pfeiffer et al., 1994), stress physiology (Leshem et al., 1997), root nodule biochemistry in Lupinus albus (Cueto et al., 1996), soil biology (Vermoesen et al., 1996), and pathological responses in potato tubers (Noritake et al., 1996). There are also studies that suggest NO might regulate growth and cell elongation in the root of pea (Sen and Cheema, 1995) and maize (Gouvea et al., 1997). In the root system, ascorbate plays a significant role in protecting the root system from free radicals produced as a result of exposure to metals like aluminum (Lukaszewski and Blevins, 1996), copper (Gupta et al., 1999), lead (Renata et al., 1999; Mishra and Choudhuri, 1998), and mercury (Mishra and Choudhuri, 1998). Ascorbate-mediated antioxidant protection has also been implicated in root tolerance and recovery from salinity (Meneguzzo et al., 1999), chilling (Queiroz et al., 1998), and hypoxic stresses (Biemelt et al., 1998). Ascorbate also has been shown to be involved in mineral nutrition as iron deficiency induces changes in both ascorbate concentrations and enzyme activities related to ascorbate metabolism (Zaharieva et al., 1999). There is now evidence that inhibition of root growth associated with boron deficiency is a result of impaired ascorbate metabolism (Lukaszewski and Blevins, 1996). Depending on the concentration H2O2 can induce cell protective responses, programmed cell death (apoptosis), or necrosis in plant systems. Low concentrations of H2O2 can induce antioxidant responses, whereas a higher level of H2O2 triggers apoptosis (Levine et al.,

343

1994). One of the best-studied and best-understood roles of H2O2 in plants is the hypersensitive response (HR), which is induced in plant tissues during pathogen infection (Lamb and Dixon, 1997). The SREM probes that are now being developed will be invaluable tools in these areas. E.

Developing Self-Referencing Biosensor Technologies

Development of enzyme-coupled electrochemical sensors (biosensors) for use as self-referencing microelectrodes has a great potential for application in plant science. The basic concept is to utilize an enzyme to convert the presence of an undetectable substance into a detectable signal via the production of a measurable reporter. For the sake of brevity this discussion will focus on electrochemical detection of the reporter species and will not deal with other types of reporter detection such as the use of ion electrode, colorimetric, photometric, luminometric, or enthalmetric coupled techniques. There are two classes of electrochemical enzyme electrodes and these are based on the uses of either oxidase or dehydrogenase enzymes. The most common type of these biosensors is the oxidase-based biosensor. Oxidase activity can be coupled to the electrochemical detection of hydrogen peroxide. During the enzymatic step the analyte reacts with oxygen producing the enzymatic product, which itself is electrochemically inactive, and H2O2. The hydrogen peroxide, however, is detected electrochemically by oxidation at 0.65 V with a platinum electrode. This approach has been used to develop biosensors capable of detecting compounds such as bilirubin, creatinine, dextrin, glucose, glucose-6-phosphate, glutamate, maltose, and sucrose (Danielson and Mosbach, 1988). Oxidase base biosensors have also been built upon oxygen electrodes, as was done in the development of the first enzyme electrode (Updike and Hicks, 1967) for the measurement of glucose. Here, increases in analyte concentration are correlated with decreases in measured oxygen availability. Sensors capable of detecting biotin, creatinine, d-lactate, l-lactate, lactose, l-lysine, phenol, and sucrose have been built using this approach (Danielson and Mosbach, 1988). The second type of enzyme-based biosensors is those that utilize NAD+- or NADP+-dependent dehydrogenase enzymes. In the enzymatic step, the analyte is converted to the product along with the requisite reduction of NAD+ to NADH. NADH is subsequently oxidized to produce an electrochemical signal that relates to the concentration of the analyte. While

344

direct electrochemical measurement of NADH oxidation is possible, this requires such high overpotentials that the electrode surface will become fouled. Using a mediator to facilitate the transfer of electrons from NADH to the electrode or the direct transfer of electrons from the enzyme to the electrode can solve this problem. Numerous mediator-based strategies for modifying electrodes for the detection of NADH have been reported (Albery et al., 1987; Hin and Lowe, 1987; Nowall and Kuhr, 1995; Nakamura et al., 1996; Silber et al., 1996; Alvarez-Crespo et al., 1997; Curulli et al., 1997; Pandey et al., 1998; Lorenzo et al., 1998). Many of these strategies involve the use of both conducting and nonconducting polymers that are electrodeposited on the active surface of the electrode. The potential of biosensors is in allowing direct measurement, and it will be of value for identification of compounds that have not previously been measurable because they are nonionic and electrochemically inactive. Biosensors should allow for direct measurement of any number of low-molecular-weight organic compounds in and around plant roots. These include glucose, sucrose, and ethanol, as well as low-molecular-weight organic acid exudates, like malate and citrate, that are believed to mediate root tolerance to soil metals like aluminum.

V. SUMMARY Microelectrode sensor technologies have proven to be important tools in root biology. The importance of the application of microelectrodes to root biology will increase as application grows from the use of ISM sensors to include electroanalytical and biosensorbased microelectrodes. For root surface transport studies, the SRM technique is an extremely powerful technique as it allows for noninvasive mapping of dynamic flux from individual plant roots in real time. Furthermore, these fluxes can be scanned and mapped along the entire surface of an active tissue, such as a root tip or root hair, and correlated spatially with subsequent analysis of the morphology and anatomy. The utility of the SRM technique will also be enhanced by the application of various types of analytical sensor techniques that include new types of ISMs, electroanalytical sensors, and biosensors. In addition to the application of existing analytical technologies, there is much to be gained in actually exploring the development of new classes of electroanalytical and biosensors that are specific for com-

Porterfield

pounds that are unique to and important in plant systems. These new techniques could then be used to develop new classes of SRMs that could provide detailed information for the flux of a diverse array of compound that may even include plant hormones like auxin and ethylene. REFERENCES Albery WJ, Bartlett PN, Cass AE. 1987. Amperometric enzyme electrodes. Phil Trans R Soc Lond B Biol Sci 316:107–119. Allen NS, Bennet MN, Cox DN, Shipley A, Ehrhardt DW, Long SR. 1994. Effects of nod factors on alfalfa root hair Ca++ and H+ currents and on cytoskeletal behavior. In: Daniels MJ, Donnie JA, Osbourn AE, eds. Advances in Molecular Genetics of Plant–Microbe Interactions, Vol. 3. Dordrecht, Netherlands: Kluwer, pp 107–113. Alvarez-Crespo SL, Lobo-Castanon MJ, Miranda-Ordieres AJ, Tunon-Blanco P. 1997. Amperometric glutamate biosensor based on poly(o-phenylenediamine) film electrogenerated onto modified carbon paste electrodes. Biosens Bioelectron 12:739–747. Ammann D. 1986. Ion Selective Micro-electrodes. New York; Springer-Verlag. Armstrong W. 1994. Polarographic oxygen electrodes and their use in plant aeration studies. Proc R Soc Edinb 102B:511–527. Armstrong W, Beckett PM, Justin SHFW, Lythe S. 1990. Modelling and other aspects of root aeration. In: Jackson MB, Davies DD, Lambers H, eds. Plant Life Under Oxygen Stress. The Hague; SPB Academic Publishing, pp 267–282. Armstrong W, Cringle S, Brown M, Greenway H. 1993. A microelectrode study of oxygen distribution in the roots of intact maize seedlings. In: Jackson MB, Black CR, eds. Interacting Stresses on Plants in a Changing Climate. NATO ASI Series I; Global Change, Vol 16. Berlin; Springer-Verlag, pp 287– 304. Bard AJ, Faulkner LR. 1980. Electrochemical methods: fundamentals and applications. New York; John Wiley and Sons, pp 142–145. Billack B, Heck DE, Porterfield DM, Malchow RP, Smith PJS, Gardner CR, Laskin DL, Laskin JD. 2001. Minimal amidine structure for inhibition of nitric oxide biosynthesis. Biochem Pharmacol 61:1581–1586. Biemelt S, Keetman U, Albrecht G. 1998. Re-aeration following hypoxia or anoxia leads to activation of the antioxidative defense system in roots of wheat seedlings. Plant Physiol 116:651–658. Bluh O, Scott BIH. 1950. Vibrating probe electrometer for the measurement of bioelectric potentials. Rev Sci Inst 10:867–868.

Microsensors Bowling DJF. 1973. Measurement of a gradient of oxygen partial pressure across the intact root. Planta 111:323– 328. Cueto M, Hernandez-Perera O, Martin R, Bentura ML, Rodrigo J, Lamas S, Golvano MP. 1996. Presence of nitric oxide synthase activity in roots and nodules of Lupinus albus. FEBS Lett 398:159–164. Curulli A, Carelli I, Trischitta O, Palleschi G. 1997. Assembling and evaluation of new dehydrogenase enzyme electrode probes obtained by electropolymerization of aminobenzine isomers and PQQ on gold, platinum and carbon electrodes. Biosens Bioelectron 12:1043–1055. Davies WP. 1966. Membrane potential and resistance of perfused skeletal muscle fibers with control of membrane current. Fed Proc 25:332. Davies PW, Brink F. 1942. Microelectrodes for measuring local oxygen tension in animal tissues. Rev Sci Instr 13:524–533. Danielson B, Mosbach K. 1988. Analytical applications with emphasis on biosensors. Introduction. In: Mosbach K, ed. Methods in Enzymology, Vol. 137. New York; Academic Press, pp 3–14. Felle HH. 1994. The H+/Cl symporter in root-hair cells of Sinapis alba. An electrophysiological study using ionselective microelectrodes. Plant Physiol 106:1131– 1136. Felle HH. 1998. The apoplastic pH of the Zea mays root cortex as measured with pH-sensitive microelectrodes: aspects of regulation. J Exp Bot 49:987–995. Felle HH, Hepler PK. 1997. The cytosolic Ca2+ concentration gradient of Sinapis alba root hairs as revealed by Ca2+-selective microelectrode tests and fura-dextran ratio imaging. Plant Physiol 114:39–45. Felle HH, Kondorosi E, Kondorosi A, Schultze M. 1999a. Elevation of the cytosolic free (Ca2+) is indispensable for the transduction of the Nod factor signal in alfalfa. Plant Physiol 121:273–279. Felle HH, Kondorosi E, Kondorosi A, Schultze M. 1999b. Nod factors modulate the concentration of cytosolic free calcium differently in growing and non-growing root hairs of Medicago sativa L. Planta 209:207–212. Forstner H, Gnaiger E. 1983. Calculation of equilibrated oxygen concentration. In: Gnaiger E, Forstner H, eds. Polarographic Oxygen Sensors. Heidelberg, Germany: Springer-Verlag, pp 334–336. Gouvea CMCP, Souza JF, Magalhaes ACN, Martins IS. 1997. NO releasing substances that induce growth elongation in maize root segments. Plant Growth Regul 21:183–187. Grahm L. 1964. Measurements of geoelectric and auxininduced potentials in coleoptiles with a refined vibrating electrode technique. Physiol Plant 17:231–261. Grahm L, Hertz CA. 1962. Measurement of the geoelectric effect in coleoptiles by a new technique. Physiol Plant 15:96–114.

345 Grahm L, Hertz CA. 1964. Measurements of the geoelectric effect in coleoptiles. Physiol Plant 17:186–201. Gupta M, Cuypers A, Vangronsveld J, Clijsters H. 1999. Copper affects the enzymes of the ascorbate-glutathione cycle and its related metabolites in the roots of Phaseolus vulgaris. Physiol Plant 106:262–267. Heck DE, Billack B, Porterfield DM, Smith PJS, Malchow RP, Laskin JD 1999. Lipids and serumderived proteins sequester nitric oxide. Toxicologist 48(1–S):158. Herrmann A, Felle HH. 1995. Tip growth in root hair cells of Sinapis alba L.: Significance of internal and external Ca2+ and pH. New Phytol 129:523–533. Henriksen GH, Raman DR, Walker LP, Spanswick RM. 1992. Measurement of net fluxes of ammonium and nitrate at the surface of barley roots using ion-selective microelectrodes. II. Patterns of uptake along the root axis and evaluation of the microelectrode flux estimation technique. Plant Physiol 99:734–747. Hertz CA. 1960. Electrostatic measurement of the geoelectric effect in coleoptiles. Nature 187:320–321. Hin BF, Lowe CR. 1987. Catalytic oxidation of reduced nicotinamide adenine dinucleotide at hexacyanoferrate-modified nickel electrodes. Anal Chem 59:2111– 2115. Huang JW, Grunes DL, Kochian LV. 1992. Aluminium effects on the kinetics of calcium uptake into cells of the wheat root apex. Quantification of calcium fluxes using a calcium-selective vibrating microelectrode. Planta 188:414–421. Huang L, Shen H, Atkinson MA, Kennedy RT. 1995. Detection of exocytosis at individual pancreatic beta cells by amperometry at a chemically modified microelectrode. Proc Natl Acad Sci USA 92:9608–9612. Hush JM, Newman IA, Overall RL. 1992. Utilization of the vibrating probe and ion-selective microelectrode techniques to investigate electrophysiological responses to wounding in pea roots. J Exp Bot 43:1251–1257. Hojberg O, Sorensen J. 1993. Microgradients of microbial oxygen consumption in a barley rhizosphere model system. Appl Environ Microbiol 59:431–437. Jaffe LF, Nucitelli R. 1974. An ultrasensitive vibrating probe for measuring steady extracellular current. J Cell Biol 63:614–628. Jones DL, Shaff JE, Kochian LV. 1995. Role of calcium and other ions in directing root hair tip growth in Limnobium stonoloniferum. Planta 197:672–680. Kumar SM, Porterfield DM, Muller KJ, Sahley CL, Smith PJS. 1999. In situ measurement of injury-induced nitric oxide (NO). Soc Neurosci Abstr 25:313. Kumar SM, Porterfield DM, Muller KJ, Smith PJS, Sahley CL. 2001. Nerve injury causes a rapid efflux of nitric oxide (NO). J Neurosci 21:215–220. Kochian LV, Shaff JE, Khu¨trieber WM, Jaffe LF, Lucas WJ. 1992. Use of an extracellular, ion-selective, vibrating microelectrode system for the quantification of K+,

346 H+, and Ca2+ fluxes in maize roots and maize suspension cells. Planta 188:601–610. Khu¨trieber WM, Jaffe LF. 1990. Detection of extracellular calcium gradients with a calcium-specific vibrating electrode. J Cell Biol 110:1565–1573. Lamb C, Dixon RA. 1997. The oxidative burst in plant disease resistance. Annu Rev Plant Physiol Plant Mol Biol 48:251–257. Land SC, Porterfield DM, Sanger RH, Smith PJS. 1999. The self-referencing oxygen-selective microelectrode: detection of trans-membrane oxygen flux from single cells. J Exp Biol 202:211–218. Leshem YY, Haramaty E, Iluz D, Malik Z, Sofer Y, Roitman L, Leshem Y. 1997. Effect of stress nitric oxide (NO): interaction between chlorophyll fluorescence, galactolipid fluidity and lipoxygenase activity. Plant Physiol Biochem 35:573–579. Levine A, Tenhaken R, Dixon R, Lamb C. 1994. H2O2 from the oxidative burst orchestrates the plant hypersensitive disease resistance response. Cell 79:583–593. Lorenzo E, Pariente F, Hernandez L, Tobalina F, Darder M, Wu Q, Maskus M, Abruna HD. 1998. Analytical strategies for amperometric biosensors based on chemically modified electrodes. Biosens Bioelectron 13:319– 332. Lukaszewski KM, Blevins DG. 1996. Root growth inhibition in boron-deficient or aluminum stressed squash may be a result of impaired ascorbate metabolism. Plant Physiol 112:1135–1140. Mancuso S, Papeschi G, Marras AM. 2000. A polarographic, oxygen-selective, vibrating-microelectrode system for the spatial and temporal characterisation of transmembrane oxygen flux in plants. Planta 211:384–389. Meneguzzo S, Navari-Izzo F, Izzo R. 1999. Antioxidative responses of shoots and roots of wheat to increasing NaCl concentrations. J Plant Physiol 155:274–280. Miller AJ. 1995. Ion-selective microelectrodes for measurement of intracellular ion concentration. Methods Cell Biol 49:275–291. Miller AL, Raven JA, Sprent JI, Weisenseel MH. 1986. Endogenous ion currents traverse growing roots and root hairs of Trifolium repens. Plant Cell Environ 9:79– 83. Miller AL, Shand E, Gow NAR. 1988. Ion currents associated with root tips, emerging laterals and induced wound sites in Nicotiana tabacum: spatial relationship proposed between resulting electrical fields and phytophtoran zoospore infection. Plant Cell Environ 11:21–25. Mishra A, Choudhuri MA. 1998. Amelioration of lead and mercury effects on germination and rice seedling growth by antioxidants. Biol Plant 41:469–473. Nakamura Y, Suye S, Kira J, Tera H, Tabata I, Senda M. 1996. Electron-transfer function of NAD+-immobilized alginic acid. Biochim Biophys Acta 1289:221–225.

Porterfield Newman IA, Kochian LV, Grusak MA, Lucas WJ. 1987. Fluxes of H+ and K+ in corn roots. Plant Physiol 84:1177–1184. Noritake T, Kawakita K, Doke N. 1996. Nitric oxide induces phytoalexin accumulation in potato tuber tissues. Plant Cell Physiol 37:113–116. Nowall W, Kuhr B. 1995. Electrocatalytic surface for the oxidation of NADH and other anionic molecules of biological significance. Anal Chem 67:3583–3588. Ober ES, Sharpe RE. 1996. A microelectrode for direct measurement of O2 partial pressure within plant tissues. J Exp Bot 47:447–454. Pandey PC, Upadhyay S, Upadhyay BC, Pathak HC. 1998. Ethanol biosensors and electrochemical oxidation of NADH. Anal Biochem 260:195–203. Pepperell J, Porterfield DM, Liu L, Smith PJS, Keefe DL. 1999a. Ascorbic-acid regulation in eggs and zygotes of the hamster. Biol Reprod 60:418. Pepperell J, Porterfield DM, Liu L, Smith PJS, Keefe DL. 1999b. Noninvasive detection of ascorbate fluxes in individual, oxidatively-stressed early embryos. Human Reprod 14:157. Peters W, Felle HH. 1999. The correlation of profiles of surface pH and elongation growth in maize roots. Plant Physiol 121:905–912. Pfeiffer S, Janistyn B, Jessner G, Pichorner H, Ebermann R. 1994. Gaseous nitric oxide stimulates guanosine-3 0 ,5 0 cyclic monophosphate (cGMP) formation in spruce needles. Phytochemistry 36:259–262. Pineros MA, Shaff JE, Kochian LV. 1998. Development, characterization and application of a cadmium-selective microelectrode for the measurement of cadmium fluxes in roots of Thlaspi species and wheat. Plant Physiol 116:1393–1401. Porterfield DM, Smith PJS. 2000. Characterization of transcellular oxygen and proton fluxes from Spirogyra grevilleana using self-referencing microelectrodes. Protoplasma 212:80–88. Porterfield DM, Trimarchi J, Keefe DL, Smith PJS. 1998. Metabolism and calcium homeostasis during development of the mouse embryo to the blastocyst stage in M2 culture medium. Biol Bull 195:208–209. Porterfield DM, Kuang A, Smith PJS, Crispi ML, Musgrave ME. 1999. Oxygen-depleted zones inside reproductive structures of Brassicaceae: implications for oxygen control of seed development. Can J Bot 77:1439–1446. Porterfield DM, Laskin J, Smith PJS, Malchow RP, Billack B, Heck D. Direct measurement of nitric oxide fluxes from single cells using a novel self-referencing microsensor. Am J Physiol (in press). Queiroz CGS, Alonso A, Mares-Guia M, Magalhaes AC. 1998. Chilling-induced changes in membrane fluidity and antioxidant enzyme activities in Coffea arabica L. roots. Biol Plant 41:403–413.

Microsensors Renata R, Waplak S, Gwozdz E. 1999. Free radical formation and activity of antioxidant enzymes in lupin roots exposed to lead. Plant Physiol Biochem 37:187–194. Schneiderman G, Goldstick TK. 1978. Oxygen electrode design criteria and performance characteristics: recessed cathode. J Appl Physiol 45:145–154. Schiefelbein JW, Shipley A, Rowse P. 1992. Calcium influx at the tip of growing root-hair cells of Arabidopsis thaliana. Planta 187:455–459. Sen S, Cheema IR. 1995. Nitric oxide synthase and calmodulin immunoreactivity in plant embryonic tissue. Biochem Arch 11:221–227. Serikawa KA, Porterfield DM, Mandoli DF. 2001. Asymmetrical subcellular mRNA distribution correlates with enzymatic activity for carbonic anhydrases from Acetabularia acetabulum. Plant Physiol 125: 900– 911. Serikawa KA, Porterfield DM, Smith PJS, Mandoli DF. 2000. Calcification and measurements of proton and oxygen flux reveal subcellular domains in Acetabularia acetabulum. Planta 211:474–483. Shabala SN, Newman IA, Morris J. 1997. Oscillations in H+ and Ca2+ ion fluxes around the elongation region of corn roots and effects of external pH. Plant Physiol 113:111–118. Shipley AM, Feijo´ JA. 1999. The use of the vibrating probe technique to study steady extra-cellular currents during pollen germination and tube growth. In: Nagel W, Scheffey C, Kunkel J, eds. Fertilization in Higher Plants—Molecular and Cytological Aspects. Berlin; Springer-Verlag, pp 235–252. Silber A, Hampp N, Schuhmann W. 1996. Poly(methylene blue)-modified thick-film gold electrodes for the electrocatalytic oxidation of NADH and their application in glucose biosensors. Biosens Bioelectron 11:215–223. Silver IA. 1967. Problems in the investigation of tissue oxygen microenvironment. In: Reneau D, ed. Chemical Engineering in Medicine. Washington; Am Chem Soc Press, pp 343–351. Smith PJS, Hammar K, Porterfield DM, Sanger RH. 1999. Self-referencing, non-invasive, ion selective electrode

347 for single cell detection of trans-plasma membrane calcium flux. Microsc Res Tech 46:398–417. Taylor AR, Bloom AJ. 1998. Ammonium, nitrate, and proton fluxes along the maize root. Plant Cell Environ 21:1255–1263. Tjepkema JD, Yocum CS. 1974. Measurement of oxygen partial pressure within soybean nodules by oxygen microelectrodes. Planta 119:351–360. Updike SJ, Hicks GP. 1967. The enzyme electrode. Nature 214:986–988. Vermoesen CJ, De Groot L. Nollet P, Boeckx P, Van Cleemput O. 1996. Effect of ammonium and nitrate application on the NO and N2O emission out of different soils. Plant Soil 181:153–162. Verslues PE, Ober ES, Sharp RE. 1998. Root growth and oxygen relations at low water potentials: impact of oxygen availability in polyethylene glycol solutions. Plant Physiol 116:1403–1412. Walker DJ, Smith SJ, Miller AJ. 1995. Simultaneous measurement of intracellular pH and K+ or NO3 in barley root cells using triple-barrelled, ion-selective microelectrodes. Plant Physiol 108:743–751. Weisenseel MH, Dorn A, Jaffe LF. 1979. Natural H+ currents traverse growing roots and root hairs of barley (Hordeum vulgare L.). Plant Physiol 64:512–518. Weisenseel MH, Becker HF, Ehlgotz JG. 1992. Growth, gravitropism, and endogenous ion currents of cress roots (Lepidium sativum L.). Plant Physiol 100:16–25. Whalen WJ. 1974. Some problems with an intracellular pO2 electrode. Adv Exp Med Biol 50:39–41. Whalen, WJ, Riley J, Nair P. 1967. A microelectrode for measuring intracellular PO2. J Appl Physiol 23:798– 801. Witty JF, Skot L, Revsbech NP. 1987. Direct evidence for changes in the resistance of legume nodules to oxygen diffusion. J Exp Bot 38:1129–1140. Zaharieva T, Yamashita K, Matsumoto H. 1999. Iron deficiency induced changes in ascorbate content and enzyme activities related to ascorbate metabolism in cucumber roots. Plant Cell Physiol 40:273–280. Zisman WA. 1932. A new method of measuring contact potential differences in metals. Rev Sci Inst 7:367–370.

21 Rooting of Micropropagules Geert-Jan de Klerk Centre for Plant Tissue Culture Research, Lisse, The Netherlands

I.

INTRODUCTION

Micropropagation, i.e., vegetative propagation of plants in tissue culture, has been broadly applied starting since the early 1980s. This chapter deals first briefly with general principles of micropropagation, and then focuses on adventitious root formation from cuttings produced in tissue culture (microcuttings).

The term tissue culture is used to describe culture of plant tissues under sterile conditions on top of or submerged in an artificial medium. Such media are composed of an aqueous solution of organic nutrients (usually sucrose), inorganic nutrients (often based on the formulation of macro- and micronutrients published by Murashige and Skoog [1962]), plant growth regulators (mostly a cytokinin and/or an auxin), and vitamins. The medium may be either liquid, or by addition of agar semisolid. In tissue culture, excised plant organs (shoots, roots, embryos, meristems), callus, cells and even protoplasts can be cultured and the direction of their development can be controlled relatively easily. Tissue culture has become an important tool for studies of physiological and developmental processes because of two major advantages: the conditions during experiments can be strictly controlled, and research can be carried out in simplified systems. As an example of the latter, for research of adventitious root formation a system was developed consisting of 1-mm slices excised from stems of apple microcuttings (Van der Krieken et al., 1993). In this system, the interferences by other organs of the plants, in particular by leaves and by apical and axillary buds, are avoided. Because the explant is very small, the factors under examination have an almost direct access to the founder cells from which the adventitious roots develop.

II.

METHODS IN MICROPROPAGATION

First practical applications of tissue culture concerned freeing plants from diseases. The potential of propagating plants (orchids) through tissue culture was recognized by Morel in 1960. Major breakthroughs were made when a generally applicable nutrient medium was devised by Murashige and Skoog in 1962, and when the use of benzylaminopurine (BAP) was introduced in order to force outgrowth of axillary buds (Sachs and Thimann, 1964). Micropropagation occurs in three steps: establishment (in which plant tissue is transferred to the in vitro environment and growth is initiated), multiplication, and reestablishment ex vitro. In the establishment phase, tissues from plants growing ex vitro in a glasshouse or in the field are surface-sterilized and transferred to a tissue culture tube with nutrient medium and plant growth regulators. Either buds or tissues without preexisting meristems may be taken as starting material. Buds are allowed to grow out to shoots, and tissues without preexisting meristems are induced to produce callus or to generate adventitious shoots. Two major prob349

350

lems are met. First, the chlorine that is used to kill microorganisms on the surface of explants cannot reach endophytic bacteria and fungi. This problem of endogenous contamination can be met by keeping the mother plants under special conditions—for example, by growing them in a greenhouse and by avoiding wetting the shoots. Endogenous contamination is also reduced when meristems are excised and cultured instead of complete buds. In addition, antibiotics can be added to the initiation medium. In tissues that are capable of resisting environmental stresses, i.e., bulbs, tubers, corms, and dormant buds of perennial plants, a warm-water treatment may kill endogenous contaminants (Langens-Gerrits et al., 1998). The second major problem is that the explants may not resume growth. This can be solved by mother plant pretreatments, by starting the cultures in the proper season or by medium adaptations. Multiplication may be carried out in three ways: 1. Via axillary branching from shoots. Apical dominance is broken in blocked axillary buds by addition of cytokinin to the medium or by removal of the apical bud. The released buds develop into side shoots. For the next cycle, the newly formed side shoots are excised and apical dominance is broken again, resulting in the formation of secondary side shoots. The secondary side shoots are excised and the process is repeated until sufficient shoots have been produced. These shoots are reestablished ex vitro. Axillary branching can be applied in many crops and is, because of its ease and reliability, the most frequently applied method in tissue culture companies. A major disadvantage, though, is that it is a very labor-intensive method. Problems that are met include a low rate of propagation, hyperhydricity, decline in vigor, necrosis, loss of the chimeric structure, and, after reestablishment, bushiness. 2. Via adventitious shoot formation. Shoot meristems are regenerated from tissues without preexisting meristems (e.g., from stem, leaf, or scale fragments). The shoots may develop directly from cells of the explants or after an intermediate phase of callus. This method is applied commercially for a few crops, among others, for lilies and African violets. In many crops, however, regeneration of shoots at large numbers is difficult to achieve. It should be noted that plants originating from adventitious meristems may suffer from a high incidence of mutations, in particular after an intermediate callus phase. This phenomenon is denoted as somaclonal variation (reviewed in De Klerk, 1990).

De Klerk

3. Via somatic embryogenesis in liquid medium. By this method, new meristems are also being formed from tissues without preexisting meristems. A major difference from the previous method is a practical one: the embryos are singulated in liquid medium, allowing automation. Since major problems are met, this method is being used on a commercial scale for a few crops only. The three types of propagules that are produced in tissue culture—viz., microcuttings, microstorage organs (bulblets, tubers, corms), or somatic embryos—have to be reestablished in soil. This is problematic for somatic embryos because they are often tiny and vulnerable, and for microcuttings. Microcuttings have a phenotype different from that of normal cuttings: They have a diminished stature, a cuticle with reduced wax, poorly functioning stomata, poorly developed palisade tissue, ample spongy parenchyma tissue with large intercellular air spaces, and hypolignified stems (Ziv, 1995). These features are related to the in vitro environment that is very different from the normal environment. In the headspace of tissue culture containers, the atmosphere has a very high humidity, high diurnal fluctuations of CO2 and O2, and high levels of organic gases, among others of ethylene. Other notable characteristics of the in vitro environment are the medium that contains high concentrations of sucrose and plant growth regulators, low light intensity, and the incompleteness of the cultured plants. With respect to the latter, in shoot cultures the absence of roots may have serious consequences: roots are uptake organs for water and nutrients, and they may produce compounds that are necessary for the shoots. After transfer from tissue culture, microcuttings require a transitional environment for acclimatization to the ex vitro conditions. In the transitional environment in which the plantlets are cultured for a few days to several weeks, relative humidity is at first close to saturation and then gradually brought down to ambient values. The light intensity is kept low at first but is raised later. The transitional period allows the microcuttings to develop an adequate root system, to improve the water retention capacity of the persistent leaves, to form new leaves, and to become autotrophic. For example, microcuttings of a Juglans hybrid remain heterotrophic for the first 7 days after transfer to soil (Chenevard et al., 1997). Just after transfer from the in vitro environment, microcuttings are often very vulnerable to mechanical damage and to attacks by pathogens.

Micropropagules

351

Like normal cuttings, microcuttings require a rooting treatment. They may be either rooted like normal cuttings, e.g., by giving them a dip in rooting powder (an auxin, usually IBA, with talc as carrier), or they may be rooted in vitro after which microcuttings with roots are transferred to soil. Microstorage organs usually do not receive a rooting treatment, but such treatment may be crucial, e.g., in Narcissus bulblets (Langens-Gerrits and Nashimoto, 1997). III.

THE ROOTING PROCESS

A.

Steps in the Rooting Process

Differentiated somatic cells may reinitiate the developmental program and give rise adventitious shoots, adventitious roots, or somatic embryos. Just like the formation of new organs from somatic cells of animals, this phenomenon is referred to as regeneration. The formation of shoots (caulogenesis) or roots (rhizogenesis) are also specified as ‘‘adventitious organogenesis’’ whereas ‘‘somatic embryogenesis’’ describes the formation of adventitious (somatic) embryos. Regeneration occurs frequently during a normal plant life, but its occurrence is strongly enhanced in tissue culture, among others because plant hormones may be easily applied via the sterile tissue culture medium. Addition of cytokinin, often together with a low concentration of auxin, promotes adventitious shoot formation. Addition of auxin promotes adventitious root formation. Addition of certain auxins, in particular 2,4-D, increases the incidence of somatic embryogenesis. The most important achievement in the study of adventitious root regeneration during the last decade is a conceptual one (Kevers et al., 1997): rooting is now envisaged as a process composed of well-defined, successive phases each with its own requirements. The timing of these phases has been examined in histological and biochemical studies. However, physiological experiments using differential sensitivities to pulses with specific plant growth regulators are more instructive (De Klerk, 1996). We have proposed a time schedule for three successive phases of the rooting process in apple microcuttings (Fig. 1) based on experiments in which 24-h pulses with auxin (indolebutyric acid; IBA), cytokinin (BAP), or a genuine antiauxin (p-chlorophenoxyisobutyric acid; PCIB) were given (De Klerk, 1995; De Klerk et al., 1995): 1. In an initial dedifferentiation phase that lasts up to 24 h after excision, certain cells in the base of the stem develop competence to respond to the rhizogenic signal.

Figure 1 Dissection of the rooting process in apple Jork 9 microcuttings. For the timing of these phases, their differential sensitivities to 24-h pulses with auxin, cytokinin, and antiauxin were used.

2. Subsequently during an induction phase (between 24 and 96 h), these founder cells become determined to form roots by the rhizogenic action of auxin. Histological studies have shown that during this period the root meristems are being formed. 3. Thereafter, a phase of morphological differentiation occurs during which the roots develop. During this phase the rhizogenic signal is no longer required, and is—at the concentration used to induce root meristems—even strongly inhibitory. B.

Auxin

Auxin is known to enhance adventitious root formation from cuttings (Thimann and Went, 1934). In the practice of vegetative propagation, application of auxin is still the only method used to achieve rooting. When auxin is supplied to normal cuttings, uptake occurs almost exclusively via the cut surface (Kenney et al., 1969). The cuticle of microcuttings is often in a very poor condition (Ziv, 1995). Nevertheless, auxin uptake by microcuttings also occurs predominantly via the cut surface (Guan and De Klerk, 2000). Uptake of auxin into the cells occurs very rapidly either by diffusive permeation of the membrane of the relatively lipophilic undissociated molecule or by mediated uptake of the anion (Lomax et al., 1995). To my knowledge, the rate of auxin uptake from rooting powder has never been examined. In tissue culture, IAA uptake from solidified medium is very rapid,

352

De Klerk

depleting the medium close to the cut surface (Guan et al., 1997). Therefore it is likely that auxin is taken up very rapidly from the rooting powder, probably within a few hours. It should be remembered that auxin is rapidly deactivated after uptake, leaving 2% or less in the active, free form (De Klerk et al., 1999). There are two major pathways of auxin deactivation: by conjugation to sugars or amino acids, or by oxidation. Conjugation is reversible but oxidation is not (Smulders et al., 1990). As noted above, auxin exerts its rhizogenic action from
24 h to 72 h after excision of apple microcutting cuttings (Fig. 1). It is important to stress again that after the induction phase, once the cells have been determined to root formation, a high auxin level is no longer required. Actually, the high concentration of auxin becomes deleterious since it blocks the outgrowth of root primordia, the growth of roots, and the development of the shoots (De Klerk et al., 1990, 1997; see also Chapter 23 by Gaspar et al. in this volume). By enhancing ethylene production, auxin also promotes senescence of shoots. C.

Ethylene

All other hormones, viz., cytokinins, gibberellins, ethylene, and abscisic acid, and the ‘‘new’’ hormones (e.g., brassinosteroids, jasmonate), influence adventitious root formation, sometimes having a vast effect. However, with the exception of ethylene their mechanism of action is not well understood and research is still too fragmentary. Before discussing the effect of ethylene, two essential features of this hormone should be stressed. First, ethylene is a gas. This has various implications. Whereas plants deactivate other hormones by enzymatic conversion (oxidation or conjugation), such metabolic deactivation systems do not exist for ethylene, except for a low rate of oxidation. Nevertheless, plants do not accumulate ethylene since it simply diffuses away from the tissue into the atmosphere. However, since the rate of ethylene diffusion in water is approximately 10,000 times less than the rate in air (Jackson, 1985), ethylene may accumulate in submerged tissues. This has an important consequence for in vitro rooting: Microcuttings are submerged with the basal portion of the stem in solidified medium, and therefore ethylene may accumulate in the tissue where the adventitious roots are mostly formed. How much ethylene accumulates depends on how far the stem is submerged, and on the rate of ethylene transport in the plant. Secondly, synthesis of ethylene is enhanced by auxin, by wounding, the orientation of

the cuttings as well as by many other factors (De Wit et al., 1990). The type of effect of ethylene depends on the phase of the rooting process (compare Fig. 1). 1. During the dedifferentiation phase, just after the cuttings have been excised from the mother plant, ethylene is stimulatory. This became apparent from 24h pulse treatments with the ethylene antagonist siverthiosulfate (STS) or with the ethylene precursor 1-aminocyclopropane-1-carboxylic acid (ACC). It has been found that elicitors enhance rooting (De Klerk et al., 1999). Thus, it is tempting to speculate that ethylene acts as an elicitor making cells capable of responding to the rhizogenic signal of auxin. 2. During the induction period, when the meristems are formed and auxin has its rhizogenic effect, ethylene becomes inhibitory. This was also shown by pulse treatments with ACC and STS. The reason why ethylene is inhibitory at this stage is unclear. Possibly, ethylene disrupts the establishment of polarity during meristem formation (cf. Kalev and Aloni, 1999). 3. During the differentiation period, ethylene is required for the formation of root hairs (Tanimoto et al., 1995). Thus, in the presence of STS, no root hairs are formed (Fig. 2). Furthermore, ethylene is required for the formation of aerenchyma in the roots of waterlogged plants (see Chapter 42 by Armstrong and Drew in this volume). The same may apply for aerenchyma of normal roots. Ethylene may also inhibit xylem formation (Zobel and Roberts, 1978). D.

Improvement of Rooting

In spite of much research, the standard rooting treatment for normal cuttings is still the one that was developed in 1936, using the auxin IBA with talc as carrier. Apparently auxin is still the only growth regulator that has a consistent, enhancing effect. Recently developed methods, viz., application of slow-release type of auxins or elicitors, may improve rooting of many crop plants (De Klerk et al., 1999). Rooting may also be promoted by enhancement of the rooting capability of (micro)cuttings. A major factor is the use of juvenile cuttings. In normal plants, rejuvenation is obtained by repeated pruning of the apex so that successively primary, secondary, tertiary, etc., branches are formed. As noted above, micropropagation is mostly based on repeated outgrowth of axillary buds. This may be the reason why during micropropagation rejuvenation occurs (e.g., Webster and Owen, 1989). For normal cuttings it is known that rooting capability of etiolated stems is improved

Micropropagules

353

Figure 2 Microcuttings of apple Jork 9 were rooted with 1
M IBA without (A) or with (B) addition of 10
M STS. Note that the roots formed in the presence of STS have hardly any root hairs and that they are much thinner. The bar is 0.5 mm.

(Kawase, 1965). In microcuttings of apple, long internodes root better than short ones (Fig. 3B) and elongation obtained by a dark treatment at the end of propagation or by using double layer, i.e., a layer of liquid medium on top of the semisolid medium, strongly enhances rooting (Fig. 3A). IV.

ROOTING AND PERFORMANCE

A.

Ex Vitro or In Vitro Rooting

As noted above, leaves of microcuttings have a very poor water retention capacity because of poorly functional stomata (Santamaria and Kerstiens, 1994). Thus, it is essential that the new root system of microcuttings functions as soon as possible after transfer to ex vitro conditions so that the extensive water loss from the leaves can be compensated for by water taken up by the roots. Often, microcuttings do not require application of auxin to achieve rooting, but treatment with auxin increases the rooting percentage, the number of roots, and the speed and synchronicity of rooting (Marks and Simpson, 2000). Microcuttings of some plants can be rooted like normal cuttings by dipping them in rooting powder and planting them in a suitable ex vitro medium, for example in a peat–perlite mixture. Microcuttings may also be rooted in vitro on a semisolid nutrient medium containing auxin and, after roots have been formed, transferred to soil. It is surprising that no critical comparisons have been made between in vitro and ex vitro rooting. Many published reports

involve major pitfalls because they monitor only survival and not growth performance, and/or because the in vitro rooting treatment was not optimized. In particular, investigators did not search for the appropriate auxin treatment and did not consider the detrimental effect of accumulation of ethylene in the headspace during in vitro rooting. In vitro rooting is a controversial issue. It was stated that in vitro roots die after planting in soil (Debergh and Maene, 1981; McClelland et al., 1990; Ziv, 1995) or that they do not function properly (Grout and Ashton, 1977). On the other hand, it was shown that the compensation for water loss by Picea microcuttings correlates with the number of in vitro formed roots (Mohammed and Vidaver, 1991). Van Telgen et al. (1993) describe that performance after transfer to soil is correlated with the number of roots formed during in vitro rooting. Almost all in vitro formed roots of the apple rootstock Jork 9 resumed growth after transfer to soil, and root growth resumed almost immediately after the transfer (Fig. 4). Performance ex vitro was related to the number of roots (albeit only weakly; De Klerk, 2000). Ex vitro rooting of apple microcuttings ensured survival but did not result in rapid resumption of growth. Microcuttings rooted in vitro showed much better growth, in particular when ethylene had been removed from the headspace (Fig. 5). The situation was different for roses where ex vitro rooting gave good results. It should be noted here that the differences between rooting treatments diminish when the conditions in the greenhouse have been optimized (Anthonis, 2001).

354

De Klerk

Figure 4 Length of the longest root of apple Jork 9 microcuttings after transfer to soil. The microcuttings had been rooted for 3 weeks in vitro.

Figure 3 Effect of shoot elongation on rooting of apple. (A) Effect of stem elongation in a ‘‘double-layer’’ system. Microcuttings of the apple rootstock M26 were examined for their rooting capability after a large number of subculture cycles (data are from C. Denissen, Lisse). (B) Effect of stem internode length on rooting of Jork 9 microcuttings. Onemillimeter stem slices were cut from these internodes and rooted with 3 mM IBA.

96 h in apple microcuttings. Thus, when auxin is supplied as a brief, early pulse, it is important to supply the cutting with a stable auxin that remains present for a sufficiently long period of time. Indeed, when auxin was supplied as a brief 1-h pulse, the concentration of applied IAA (unstable) to achieve sufficient rooting was so high (Fig. 6) that the shoots suffered. In such a case, IBA (stable) was the preferable auxin. During in vitro rooting, though, uptake of auxin occurs during an extended period of time of at least 3 days (Guan et

In vitro rooting is also advantageous because microcuttings are often tiny and vulnerable at the end of the propagation cycle, and considerably increase in size during the rooting treatment. Dry weight of the shoots of apple microcuttings increases about fourfold during the 3-week rooting treatment (De Klerk, 2000). The obvious disadvantages of in vitro rooting are that, because of the additional labor, the price of rooted microcuttings is approximately twice as high and the rooted microcuttings are more difficult to plant in soil. B.

Choice of the Type of Auxin

As noted above, rooting requires presence of a high level of auxin for a protracted period, e.g., for 72–

Figure 5 Weight of Jork 9 shoots that had been rooted either ex vitro (1 h exposure to 1 mM IBA) or 3 weeks in vitro with 10 mM IAA. When the shoots were rooted in vitro, grains coated with KMnO4 were also added. Weight was measured after 4–7 weeks in soil.

Micropropagules

355

Figure 6 Concentrations of auxin that gave optimal rooting after exposure of Jork 9 cuttings to auxin for 1 h or 1 d, or after continuous exposure. Note that for a 1-h exposure a very high concentration of IAA had to be supplied.

al., 1997). As auxin is taken up over such a long period, in in vitro rooting an unstable auxin may result in high rooting. Moreover, because the high concentration of auxin, necessary for the induction of rooting, inhibits the outgrowth of root primordia and is detrimental for both root and shoot growth, the use of an unstable auxin for in vitro rooting may be even preferable. Indeed, IAA is the preferential choice for in vitro rooting of apple microcuttings in comparison with the stable auxins IBA and NAA: The maximum number of roots is the same as with IBA but it is reached over a much wider range of concentrations, callus formation was little, the roots are longer, and the shoots hardly show signs of aging (De Klerk et al., 1997). C.

Figure 7 Microcuttings of Jork 9 that had been rooted with 3 mM NAA. To reverse the effects of ethylene, either 10 mM STS had been added or ethylene had been captured from the headspace by grains coated with KMnO4.

Removal of Ethylene

Shoots exposed to a high concentration of auxin show leaf senescence, indicating an effect of ethylene (De Klerk et al., 1997). When the ethylene inhibitor STS was applied together with auxin, the shoots recovered almost completely (Fig. 7). However, upon planting out, many of the shoots died, possibly because of a carryover of STS that inhibited root hair formation (Fig. 2). Thus, STS does not seem to be appropriate to counteract ethylene. Significant improvement of performance was observed when ethylene was removed from the headspace of the Petri dish by KMnO4. STS has also a tremendous positive effect on shoot quality of roses but at the same time blocks rooting (De Klerk, 2000). KMnO4 treatment strongly increases both survival and growth of rose microcuttings (Fig. 8).

Figure 8 Survival and growth of surviving shoots of rose microcuttings (Madelon) rooted in vitro with 10 mM IAA. Ethylene was allowed to accumulate in the headspace or was removed by grains coated with KMnO4.

356

V. PROSPECTS The central theme of this chapter is that there are two main differences between rooting in vitro and ex vitro. First, for ex vitro rooting a brief pulse with a high concentration of a stable auxin is supplied. Rooting of microcuttings in vitro involves a protracted uptake of auxin. Therefore, unstable auxins may perform well in in vitro rooting. Actually, they may be preferable because stable auxins remain present for a too long time. Thereby, they inhibit outgrowth of root primordia and affect the quality of shoots. Secondly, because of the wounding effect and of the auxin medium, ethylene may accumulate in the tissue culture containers during the rooting treatment and affect the microcuttings. This can be overcome by addition of KMnO4. The interface between in vitro and ex vitro has hardly been explored. In addition to adapting the rooting conditions, there are many other possibilities that may greatly improve performance after establishment. Acclimatization may be carried out to some extent in vitro, the shoots may be bacterized before transfer to soil, additional nutrients may be added before transfer, and hormonal or dormancy breaking treatments can be used. Such research can greatly enhance growth performance of microcuttings and is therefore of high importance for horticulture and the micropropagation industry. The treatment used for rooting of normal cuttings was developed some 70 years ago. New molecular techniques, e.g., DNA microarrays and studies on mutants, and the rapid developments in molecular research on lateral root formation and auxin action, will have significant impact on the understanding of rooting. This should result in new rooting treatments. Tissue culture systems, e.g., root formation in vitro from excised stem segments, constitute an excellent and indispensable tool for such research.

REFERENCES Anthonis B. 2001. The commercial practice of rooting in a nursery. In: De Klerk GJ, Van der Krieken W, eds. Root Formation. Proceedings of the Third International Congress on Adventitious Root Formation. Dordrecht, Netherlands: Kluwer (in press). Chenevard D, Frossard JS, Jay-Allemand C. 1977. Carbohydrate reserves and CO2 balance of hybrid walnut (Juglans nigra No. 23  Juglans regia) plantlets during acclimatisation. Sci Hort 68:207–217.

De Klerk Debergh PC, Maene LJ. 1981. A scheme for commercial propagation of ornamental plants by tissue culture. Sci Hort 14:335–345. De Klerk GJ. 1990. How to measure somaclonal variation: a review. Acta Bot Neerl 39:129–144. De Klerk GJ. 1995. Hormone requirements during the successive phases of rooting of Malus microcuttings. In: Terzi M, Cella R, Falavigna A. eds. Current Issues in Plant Cellular and Molecular Biology. Dordrecht, Netherlands: Kluwer, pp 111–116. De Klerk GJ. 1996. Markers of adventitious root formation. Agronomie 16:563–571. De Klerk GJ. 2000. Rooting treatment and the ex vitro performance of micropropagated plants. Acta Hort 530:277–288. De Klerk GJ, Ter Brugge J, Smulders R, Benschop M. 1990. Basic peroxidases and rooting in microcuttings of Malus. Acta Hort 280:29–36. De Klerk GJ, Keppel M, Ter Brugge J, Meekes H. 1995. Timing of the phases in adventitious root formation in apple microcuttings. J Exp Bot 46:965–972. De Klerk GJ, Ter Brugge J, Marinova S. 1997. Effectiveness of indoleacetic acid, indolebutyric acid and naphthaleneacetic acid during adventitious root formation in vitro in Malus ‘Jork 9’. Plant Cell Tissue Org Cult 49:39–44. De Klerk GJ, Van der Krieken W, De Jong J. 1999. The formation of adventitious roots: new concepts, new possibilities. In Vitro Cell Dev Biol 35:189–199. De Wit L, Liu JH, Reid DM. 1990. Production of ethylene by gravistimulation; a potential problem with the interpretation of data from some experimental techniques. Plant Cell Environ 13:237–242. Grout BWW, Aston H. 1977. Transplanting of cauliflower plants regenerated from meristem culture. 1. Water loss and transfer related to changes in leaf wax and to xylem regeneration. Hort Res 17:1–7. Guan H, Huisman P, De Klerk GJ. 1997. Rooting of apple stem slices in vitro is affected by rapid decline of indoleacetic acid in the medium. J Appl Bot 71:80–85. Guan H, De Klerk GJ. 2000. Stem segments of apple microcuttings take up auxin predominantly via the cut surface and not via the epidermis. Sci Hort 86:23–32. Jackson MB. 1985. Ethylene and responses of plants to soil waterlogging and submergence. Annu Rev Plant Physiol 36:146–174. Kalev N, Aloni R. 1999. Role of ethylene and auxin in regenerative differentiation and orientation of tracheids in Pinus pinea seedlings. New Phytol 142:307–313. Kawase M. 1965. Etiolation and rooting in cuttings. Physiol Plant 32:170–173. Kenney G, Sudi J, Blackman GE. 1969. The uptake of growth substances. XIII. Differential uptake of indole-3yl-acetic acid through the epidermal and cut surfaces of etiolated stem segments. J Exp Bot 20:820–840.

Micropropagules Kevers C, Hausman JF, Faivre-Rampant O, Evers D, Gaspar T. 1997. Hormonal control of adventitious rooting: progress and questions. J Appl Bot 71:71–79. Langens-Gerrits M, Nashimoto S. 1997. Improved protocol for the propagation of Narcissus in vitro. Acta Hort 430:311–313. Langens-Gerrits M, Albers M, De Klerk GJ. 1998. Hotwater treatment before tissue culture reduces initial contamination in Lilium and Acer. Plant Cell Tissue Org Cult 52:75–77. Lomax TL, Muday GK, Rubery PH. 1995. Auxin transport. In: Davies PJ, ed., Plant Hormones: Physiology, Biochemistry and Molecular Biology. Dordrecht, Netherlands: Kluwer, pp 509–530. Marks TR, Simpson SE. 2000. Manipulation of rooting competence in vitro in a range of difficult and easy-to-root woody plants. Plant Cell Tissue Org Cult 62:65–74. McClelland MT, Smith MAL, Carothers ZB. 1990. The effects of in vitro and ex vitro root initiation on subsequent microcutting root quality in three woody plants. Plant Cell Tissue Org Cult 23:115–123. Mohammed GH, Vidaver WE. 1991. Plantlet morphology and the regulation of water loss in tissue-cultured Douglas-fir. Physiol Plant 83:117–121. Morel G. 1960. Producing virus-free Cymbidium. Am Orchid Soc Bull 29:495–497. Murashige T, Skoog F. 1962. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol Plant 95:814–821. Smulders MJM, Van de Ven ETWM, Croes AF, Wullems GJ. 1990. Metabolism of 1-naphthaleneacetic acid in

357 explants of tobacco: evidence for release of free hormone from conjugates. J Plant Growth Regul 9:27–34. Santamaria JM, Kerstiens G. 1994. The lack of control of water loss in micropropagated plants is not related to poor cuticle development. Physiol Plant 91:191–195. Sachs T, Thimann KV. 1964. Release of apical buds from apical dominance. Nature 201:939–940. Tanimoto M, Roberts K, Dolan L. 1995. Ethylene is a positive regulator of root hair development in Arabidopsis thaliana. Plant J 8:943–948. Thimann KV, Went FW. 1934. On the chemical nature of the root forming hormone. Proc Kon Akad Wetensch 37:456–459. Van der Krieken WM, Breteler H, Visser MHM, Mavridou D. 1993. The role of the conversion of IBA into IAA on root regeneration in apple: introduction of a test system. Plant Cell Rep 12:203–206. Van Telgen HJ, Van Mil A, Kunneman B. 1992. Effect of propagation and rooting conditions on acclimatization of micropropagated plants. Acta Bot Neerl 41:453– 460. Webster CA, Owen OP. 1989. Micropropagation of the apple rootstock M. 9: effect of sustained subculture on apparent rejuvenation in vitro. J Hort Sci 64:421–428. Ziv M. 1995. In vitro acclimatization. In: Aitken-Christie J, Kozai T, Lila Smith M, eds. Automation and Environmental Control in Plant Tissue Culture. Dordrecht, Netherlands: Kluwer, pp 493–516. Zobel RW, Roberts LW. 1978. Effects of low concentrations of ethylene on cell division and cytodifferentiation in lettuce pith explants. Can J Bot 56:987–990.

22 Modeling Root System Architecture Loı¨c Page`s INRA, Centre d’Avignon, Avignon, France

I.

INTRODUCTION

Many root system models were developed as part of larger models which aimed at describing either the crop functioning in relation to its environment, including the soil (Ritchie et al., 1985; Whisler et al., 1986; Jones and Kiniry, 1986; Klepper and Rickman, 1990; Chapman et al., 1993; Adiku et al., 1995; Asseng et al., 1997) or some aspects of the soil–plant–atmosphere system (Hansen, 1975; Skiles et al., 1982). In such models, the root system was considered as an uptake system (water and minerals) and/or as a sink for photoassimilates (Reynolds and Thornley, 1982; Huck and Hillel, 1983; Cannell, 1985; Hoogenboom and Huck, 1986; Ho, 1988). Representing the uptake function of the root system requires a minimal representation of its morphology. For this purpose, the concept of root density profile has been used extensively (Huck and Hillel, 1983). Root density profiles describe the amount of roots in terms of biomass, length, surface area, etc., in horizontal layers of the soil. In relation to this very simple representation of the root system, the functional assumptions of the uptake models are also very simplified. They generally assume the spatial distribution of the roots to be homogeneous in the soil layer, and the uptake to be similar between roots. They formulate the uptake function as related to root length and to soil layer water potential (for instance) and do not take into consideration transport and resistance in pathways. (See Chapter 37 by Silberbush and Chapter 35 by Jungk in this volume for details.) The role of the root system as a sink for photoassimilates is sometimes

The main purpose of this chapter is to discuss the interest of using mathematical simulation models to study the development and the architecture of root systems. Methodological difficulties and the development of new techniques and methods for studying root systems have long been, and still are, a dominant subject in the scientific literature on roots (see review by Bo¨hm, 1979, and Chapter 18 by Polomski and Kuhn in this volume). These efforts have led to significant advances in the way of conducting experiments on roots, and obtaining more reliable data at various scales and in various conditions. Another major issue, often overshadowed until now because data acquisition appeared as the true bottleneck, is the parallel development of concepts and models to acquire, interpret, and use these new data. The set of roots, organized into a root system, can either be considered per se or as part of the larger soil–plant system (soil–plant system), as a complex system requiring specific tools and methods to be studied (Legay, 1997). The behavior of each component is the result of numerous interactions, many of which cannot be discarded without taking the risk of distorting the image of the system under study. Therefore, I would like to emphasize the methodological purpose of this chapter. The models will be mainly reviewed as tools corresponding to research steps, associated with various viewpoints on the investigated system. 359

360

considered in these global models, using very simple rules. The root system either receives a given proportion of the total amount of assimilates produced by the plant, is used as a spillway for excess of assimilates after consumption by the shoots, or is provided according to its overall calculated demand (e.g., as a function of its biomass and soil conditions). These oversimplifications in describing the root system are justified for these models which have predictive objectives, or aim at describing the functioning of a larger system, including roots, while taking several additional aspects of its functioning into account. Whenever the objective of a model is to investigate the development and function of the root system, the framework imposed by root depth or root density models is unsuitable. The major simplifications in these models are generally designed for calculation simplification and for matching data availability (root density profiles as outcome of a long tradition of soil coring) rather than for the achievement of a clear hierarchy between interacting mechanisms. Therefore, it is necessary to develop models that include details of the architecture and dynamics of root systems in order to gain new insight. In this respect, the word ‘‘architecture’’ needs to be clarified by specifying its two complementary meanings: shape and structure. Shape refers to the root system geometry, or to the spatial distribution of the roots. Shape, as well as its time-dependent variations, is most important during the first phase of soil resource acquisition because at that stage, uptake is limited by the soil transport an interception of the resources by the growing roots. The shape has sometimes been characterized by a functional criterion: distribution of distances between sample points in the soil and their nearest neighboring root (Tardieu, 1988). Root distribution often presents non uniform but clumped patterns (Tardieu, 1988; Logsdon and Allmaras, 1991; Pellerin and Page`s, 1996) which may have significant effect on the uptake processes (Van Noordwijk and Brouwer, 1991; Bruckler et al., 1991; Tardieu et al., 1992). The significance of the kinetics of root spatial distribution depends on the mobility of the resources considered (Baldwin et al., 1973). The less mobile the resources, the more crucial is this distribution pattern. The root system geometry is also a major component of the anchoring function (Coutts, 1983; Ennos and Fitter, 1992; Guingo and He´bert, 1997; see also Chapter 11 by Persson and Chapter 10 by Stokes in this volume). Root orientations, as well as the spatial distribution of their diameters and mechanical properties, are in this case the most important geometric characteristics.

Page`s

The term structure refers to the differentiation of components within the root system, and to their mutual relationships: differences in their ontogenetic characteristics which are organized along the axes (gradients) and differences between the connected roots (e.g., daughter roots are necessarily younger and generally finer than the mother roots). Connection relationships define the topology of the branching system (Fitter, 1982; Fitter, 1986; see also Chapter 2 by Fitter in this volume). The morphogenetic and functional differences between and along the roots significantly influence the root-foraging potential and resource acquisition (see Chapter 9 by Waisel and Eshel in this volume). The issues concerning transport pathways within the root system, organization, and effective conductance are still points of debate and new analyses (Fitter, 1987; Berntson, 1994a; Fitter, 1996). They cannot be dismissed just because of the calculation problems they raise. It is also evident that the structure of the root system plays a prime role in other functions of the root system, such as sink and storage functions, deposition and excretion of various biochemical compounds, sending of signals toward the plant or the soil, or association with symbiotic organisms. Thus, we must consider strong relationships between architecture and function and, from a dynamic point of view, among the time-dependent geometric, developmental, and functional properties. It shows the need for integration of the various aspects of the dynamics of the root system architecture into a consistent framework, such as a simulation model based on developmental processes. Such a model can be used to simulate these close links between the generation of the root system components, their architectural position, and their functions. In this chapter we present dynamic architectural models, i.e., models simulating both the shape and the structure of the root system and its progressive development. Some static models have been developed (Henderson et al., 1983a,b), which will not be discussed, because they rely on very different objectives and design. Fractals (Mandelbrot, 1983) can be used to simulate the dynamics of the root system architecture, but they will not be specifically considered here. Moreover, in respect to the purpose of this chapter, they can be considered as a particular case among the models that we are presenting, having more restrictive hypotheses regarding developmental processes in order to meet the autosimilarity assumption (Tatsumi et al., 1989; Fitter and Stickland, 1992; Berntson, 1994b; Shibusawa, 1994; Van Noordwijk et al., 1994; Eshel, 1998; Ozier-Lafontaine et al., 1999).

Modeling Root System Architecture

The chapter will obviously focus on the root system, with only few references to the shoot system. Still, it is worth noting that the approach and tools are not fundamentally different from those applied to shoots, and some concepts have only been translated. However, the main basic processes, as well as the constraints considered, are quite different and justify specific considerations. This review will complement and update other interesting reviews on the subject (e.g., Klepper and Rickman, 1990; Jones et al., 1991; Lynch, 1995; Lynch and Nielsen, 1996). We shall successively analyze the basic developmental hypotheses of these models and discuss how the interaction with various environmental characteristics has been considered. Then, we shall illustrate various aspects of architecture modeling using models which have been developed according to different viewpoints. This will lead us to look at the present and future links between simulation of the root system architecture and various aspects of the research conducted on root system functioning. II.

MODEL DEVELOPMENT

A.

Basic Principles

The basic principle of most root system architecture models from the first (Hackett and Rose 1972a,b; Lungley, 1973) to the most recent ones (Fitter et al., 1991; Clausnitzer and Hopmans, 1994; Somma et al., 1998) has been to mimic developmental phenomena of individual roots using very simple rules. This idea resulted from the observation that part of the global complexity is actually the result of a rather limited number of axis types and dynamic processes that are repeated many times in space and time, with some quantitative variations. Thus, the plasticity of the whole root system contrasts with the high organization of the basic processes and the relative invariance of structural properties. Therefore, the modeling approach has been (1) to identify and categorize within the root system a reduced number of components having a homogeneous developmental behavior, and (2) to characterize and formalize the dynamics of appearance, time variation, and sometimes disappearance of these components. This approach is somehow close to what Halle´ and Oldeman (1970) and Halle´ et al. (1978) have called l’analyse architecturale (architectural analysis), which they applied to the shoots of a large number of tropical tree species. The various combinations of axis types and developmental processes led them to define about 20 qualitatively different mode`les architec-

361

turaux (architectural models), which described all the observed species. It is noteworthy that the relevance of such an approach relies on the relative invariance of the components and basic processes during development. 1.

Components of the Root System

In many models the components that have been defined are the roots, or the root apices, belonging to different developmental branching orders (order 1 being directly connected to the shoot system, and order i giving rise to order i+1 by branching). The first papers, which have mainly worked on young plants (Hackett and Rose, 1972a,b; Lungley, 1973; Diggle, 1988; Page`s and Arie`s, 1988), have grouped roots according to their branching order, considering only the three or four first branching orders representing in this case almost the whole root system. This root categorization is an essential assumption, contrasting with what had been assumed before in root modeling. It resulted from the observation that roots of different branching orders have very different developmental characteristics (e.g., appearance, growth, and branching). For example, Diggle (1988) considered growth rates decreasing by a factor of 10 from order 1 to 3. Later on, several authors (Page`s et al., 1989; Le Roux and Page`s, 1994; Jourdan et al., 1995) extended the approach, defining general root types which are not strictly associated to branching orders. Apparently, root growth rates, as well as other morphogenetic properties, may be highly variable according to the origin of the root, and not only because of variations in the soil or substrate (see Chapter 9 by Waisel and Eshel in this volume). The classification of root types is based on several criteria that describe the morphogenetic properties of the roots: growth rate, growth duration, branching ability and density, and tropisms (cf. Wilcox, 1962, 1968; Charlton, 1967; Hackett, 1969; Varney et al., 1991; Atger and Edelin, 1992; Le Roux, 1994; Jourdan and Rey, 1997; Vercambre and Page`s, 1998). Coutts (1987), and then Page`s (1995) and Vercambre and Page`s (1998), showed that the apical diameter was a reliable synthetic criterion to distinguish roots of different types and to locate them on a common scale. Schematically, the big roots (those having a large apical diameter) have a high morphogenetic potential (axial and radial growth, branching, gravitropism), whereas fine roots have low developmental potential. In conclusion, it is possible to generalize the concept of root components by considering not only roots

362

grouped into types, by also types of apices and segments (which are pieces of axes). Thus, we define root system modules by analogy with the modules which make up the shoot system, as defined in the LSystem models (Prusinkiewicz and Lindenmayer, 1990). 2. Morphogenetic Rules and Their Formalization Identifying root system components and their connection relationships (topology) is not sufficient in itself to make dynamic simulations of its development. It is also necessary to investigate the processes by which the components change of state from their initiation onward. Three basic processes, involved in the development of root systems have first been identified and modeled: emission, resulting in the appearance of firstorder roots, at the base of the plant that is directly connected to the shoot; branching, by which new order i+1 roots are produced laterally on order i mother roots; and axial growth, leading to the elongation of existing roots. The emission process has been studied in cereals (Klepper et al., 1984; Picard et al., 1985; Nemoto and Yamazaki, 1986) and then included as submodels (modules) in the simulation of wheat (Diggle, 1988) and maize (Page`s et al., 1989). Cereals continuously produce nodal roots according to a highly organized sequence in space and time. As the phyllochron, which is generally tightly correlated to the cumulated sum of temperatures since sowing, nodal roots are emitted similarly from the phytomers at a regular rate, from the base upward. In these models, devoted to cereals, a simple linear function was used to link the rank of the phytomer where emergence occurs to cumulated thermal time. The branching process that has been studied the most is the acropetal branching, which results in the emergence of lateral roots in a limited zone along the bearing axis that shifts acropetally, following at a given distance the apex of the bearing axis (see Chapter 8 by Lloret and Casero in this volume). This acropetal branching process is of prime importance in the development of most root systems because it results generally in a large number of roots and it arranges lateral roots according to their age. Therefore, such branching leads to a certain spatial organization of root age. In order to formalize the spatial and temporal aspects of such a branching process, the simplest models (Hackett and Rose, 1972a,b; Lungley, 1973) were based on two rules: (1) successive lateral roots appear

Page`s

at fixed distance from one another (parameter interbranch distance), and (2) they cannot appear closer to the apex than a given threshold distance (parameter length of the apical nonbranching zone). In later models (Diggle, 1988; Page`s et al., 1992; Clausnitzer et Hopmans, 1994), the apical nonbranching zone was simulated using a temporal parameter, which is the time duration before emergence of the lateral primordia that are initiated close to the apex of the bearing axis. Using this formalization, which defines duration by a parameter, the length of the apical nonbranching zone varies according to the growth rate of the bearing axis, in agreement with observations made on several species (Page`s and Serra, 1994; Pellerin and Tabourel, 1995; Aguirrezabal and Tardieu, 1996). In the model of Jourdan and Rey (1997), as well as that of Vercambre and Page`s (1998), the type of lateral axis is not strictly determined by the type of bearing axes, but is simulated using a stochastic process. These models are more flexible regarding the topological structure of the branching system. The general idea is that the type of an axis is only partly determined by its bearing axis (e.g., a fine root cannot give rise to big lateral roots), but it also depends on random factors (e.g., among the branches of a big root, fine roots can also be observed). The initial branching direction is calculated in threedimensional models by considering an insertion angle and a radial angle. The radial orientation of laterals was sometimes simulated in relation to the internal structure of the roots (Berntson, 1994a), forcing lateral roots to emerge only in front of xylem poles along orthostichies (Esau, 1989). Axial growth, or axis elongation, was assumed to be primarily dependent on the type of axis, with strong differences reaching a factor of 10–20, showing the significance of an architectural factor on elongation. In the simplest versions (Hackett and Rose, 1972a; Lungley, 1973; Fitter, 1987; Diggle, 1988; Fitter et al., 1991), growth rate was constant for each root type and calculated from a single parameter per root type. In a slightly different way, Page`s et al. (1989), simulating maize root systems, considered that root types did not differ only in their growth rates, but also in their growth patterns (determinate or undeterminate depending on the root type, according to Cahn et al., 1989; Varney et al., 1991; Varney and McCully, 1991). For this purpose, they used a common growth function allowing the simulation of both growth patterns. In their model, Clausnitzer and Hopmans (1994) and then Somma et al. (1998) also used predetermined growth functions, and calculated a growth potential on

Modeling Root System Architecture

this basis. Environmental factors are used to calculate reduction factors (values between 0 and 1), which are then multiplied by the potential growth rate to obtain the actual growth rate. Other authors, such as Page`s et al. (1992) or Jourdan and Rey (1997), have used stochastic elongation models to account for growth variations, even for a given root type, which cannot be modeled on the basis of environmental variations. Lastly, the model of Thaler and Page`s (1998) did not use a predetermined growth function for each root type, but calculated a potential growth rate from the root apical diameter using a single function. In this model, the apical diameter, indicating the meristem volume of the root, varies with time according to carbohydrate availability. This growth model predicts large variations in growth patterns from one root to another. Growth directions were also specifically considered in several three-dimensional models, since the trajectory of the root is generally not straight and is considered as a major feature of the root system architecture. Growth direction, for the faster-growing roots especially, contributes along with growth rate to define the actual volume of soil explored. Growth direction is affected by soil constraints, which will be discussed in the next section, but also depends on the ability of the different root types to direct themselves according to specific mechanisms, the tropisms. For example, different gravitropic behaviors of roots have been described and classified by Riedacker et al. (1982). In addition to this well-known gravitropism, root orientation by various mechanisms is a very common and major feature in plants (see Chapter 29 by Porterfield in this volume). From an architectural point of view, it gives the main roots a globally organized trajectory and, simultaneously, an important local tortuosity (see Kutschera, 1960; Fig. 1). The gravitropic trends of roots were first modeled by Lungley (1973) and then formalized with more detail in several three-dimensional models (Page`s and Aries, 1988; Berntson, 1994a; Clausnitzer and Hopmans, 1994). The general principle is to give, at each elongation step, an additional vertical component to growth direction. Tardieu and Pellerin (1990) have shown the ability of this type of model to account for the general trajectory of the maize nodal roots under field conditions. Le Roux (1994) has also proposed a representation of the plagiogravitropism of some axes, i.e., their tendency to come back to a horizontal direction after deflection, by gravitropism (Riedacker et al., 1982; Le Roux and Page`s, 1996). In his model, a horizontal directional component is calculated and added to the growth

363

direction. Other submodels of tropism based on the same method could be included, in order to help in analyzing and quantifying other mechanisms which may concur to the growth direction of roots (chemiotropism, thermotropism, hydrotropism, etc.) (Coutts, 1989; see also Chapter 29 by Porterfield in this volume). In addition to these main developmental processes, which were the bases of root system architectural models, some authors (Clausnitzer and Hopmans, 1994; Page`s et al., 1995; Vercambre and Page`s, 1998) have suggested application of the same approach to other processes that affect root system dynamics: reiteration, decay and abscission, and radial growth. The reiteration concept was suggested by Oldeman (1974) to describe the dynamics of shoot architecture in tropical trees. He considered the architectural development as a sequence (the ‘‘architectural unit,’’ according to Barthe´le´my et al., 1989), in which axes of defined types appear in ordered steps at predictable locations both in time and space. Reiteration is the development of a new sequence within the first one. The new sequence may start from the beginning (in case of complete reiteration) making the initial (youngest) axes, or at a later stage (partial reiteration). This phenomenon occurs spontaneously, with aging of the tree crown, giving rise to a set of small treelike structures (complete reiterated structures) within the old tree (Halle´ et al., 1978). It may also be an adaptive response to various pruning events and changing conditions. The same phenomenon is observed in root systems. It has been described by Lyford and Wilson (1964) and by Kahn (1978), who interpreted for example the ‘‘sinkers’’ as new ‘‘architectural units’’ developing within the overall root system, starting from a typical taproot, which is the initial axis of the sequence (complete reiteration). Atger and Edelin (1992) have also used this concept to describe the dynamics of the root systems of several tree species. Vercambre and Page`s (1998) have formalized and quantified this phenomenon in a model applied to the peach tree (Fig. 2). They observed that roots not only branched by acropetal branching, but also by reiteration, creating forks in which the bearing axes were duplicated. The reiterations were partial, since the new sequences were not started from the beginning. These authors showed that the process occurred periodically, as a result of drastic variations in the growing conditions (e.g., winter or dry season). In their model, a set of new axes of the same type were generated reiteratively on the bearing axis, substituting it in the process. The number of

364

Page`s

Figure 1 Drawing of a root system, drawn from a field excavation of a Polygonum lapathifolium plant. The general organization of the trajectories of the main roots contrasts with their local tortuosity. (From Kutschera, 1960.)

reiterated axes was calculated using a predetermined distribution law. Decay and the subsequent abscission are common phenomena of the architecture genesis in long-term development of perennial plants. Clausnitzer and Hopmans (1994), Le Roux (1994), Page`s et al. (1995), and Vercambre and Page`s (1998) have developed modules to account for these decay phenomena. In their model, Vercambre and Page`s (1998) assumed that the root tip has a given life expectancy (parameter defined for each type of axis) after which it dies. Later, the axis generated by this apex disappears, either completely if it bears no living branch, or partly—i.e., its distal part—if it bears living branches. This decay module, which is very simple mainly owing to the lack of understanding of this complex phenomenon, allowed a first realistic representation of the spatial

and temporal organization of decay and abscission. This phenomenon is not randomly distributed within the root system, and its complex distribution can be described, at least roughly, by this algorithm in an efficient way. This point illustrates once more the relevance of linking local and basic developmental processes to the whole root system architecture using simulation models. Radial growth, and the production of secondary xylem and phloem, are also of major significance in the aged root system of dicotyledon species. Their functional consequences are obvious on several whole plant processes: sink for carbohydrates, axial water conductance, storage, and anchorage (see also Chapter 6 by Chaffey in this volume). In their architectural model, Vercambre and Page`s (1998) developed a radial growth submodel based on the strong allo-

Modeling Root System Architecture

365

Figure 2 Simulations of the peach tree root system in orchard conditions (after the model of Vercambre and Page`s, 1998). A. Main roots (macrorhizae) 2 (left), 3 (center), and 4 (right) years after plantation. B. Top view of these main roots, 4 years after plantation. Forks are the consequence of the reiteration process (see text), which is typical of several tree species. Radial growth, represented by the line width, was simulated using the principles of the pipe model (Shinozaki et al., 1964). Root decay and abscission concerns the fine roots (enlargement) and results in a decrease of the density near the base.

Page`s

366

metric relationships of root section areas observed at the branching points of lignified roots. The cross-sectional area of the mother root was proportional to the sum of the cross-sectional areas of the daughter roots, with a proportion coefficient close to unity. Moreover, this type of relationship is quite common to various branching systems, and is the basis of the pipe model proposed by Shinozaki et al. (1964), as well as of the fractal model of Van Noordwijk et al. (1994). More specifically, Vercambre and Page`s (1998) used this close relationship concerning root cross-sectional areas and simulated radial growth on any root segment as a function of the emergence of new roots at distal position relative to this segment. These new roots are assumed to require new pipes to connect to the shoot system (Eshel et al., 2000). Thus, radial growth is linked during the simulation to the branching processes (acropetal branching and reiteration in this case). Such a simulation renders the empirical allometric relationship at each branching point and contributes to realistic representations (Fig. 2). 3. Driving the Simulation and Coding the System Most models developed until now rely on almost the same principles and approximately the same coding system. Time is divided into time steps; at each time step, the model calculates the structure alterations induced by the application of the development modules, and the possible interactions with the environment and its variations during the considered time step (Porter et al., 1986; Clausnitzer et Hopmans, 1994; Bengough, 1997). The simulation operates typically on individual plants, clearly identified, and not at the crop scale. However, several authors produced outputs by subsequent calculations at the crop scale (Lungley, 1973; Porter et al., 1986; Bengough, 1997), considering that the simulated plant is an average plant. The characteristics of such plant could be converted to surface or volume per ground area units, knowing the plant density of the crop. Some models (Page`s et al., 1989) allow the simulation of a small crop, via the simulation of several plants located as they are in the field, in order to produce two- or three-dimensional outputs, which can be compared to plane maps (Pellerin and Page`s, 1996) or three-dimensional density maps (Grabarnik et al., 1998). The time step of the simulation varies from one model to another, from 12 h (Bengough, 1997) to several days (Vercambre and Page`s, 1998), depending on

the simulation duration, the considered processes and the expected accuracy. It can sometimes be adjusted by the user (Somma et al., 1998). In several models, some of the modules run using a shorter time step to increase numerical accuracy within these specific modules, without increasing by too much the duration of the program execution, or enlarging unnecessarily the size of the simulated structures. The root system is generally coded in the computer memory as a set of components, which represents the segments in most models, and sometimes also the generating apices. The segments code the parts of axes generated at each time step. They contain more or less detailed information on their location and development (spatial coordinates, age, diameter, type of the bearing axis) as well as pointers keeping the connections with the rest of the structure (originating segment, set of segments originating from it). Lynch and Nielsen (1996) and Lynch et al. (1997) called this structure the ‘‘extensible tree structure’’ because it actually represents a branching structure (a tree) during its development. Some authors (Vercambre and Page`s, 1998; Bidel et al., 2000a) have proposed a more elaborated structure in which they also defined an axis structure containing several segments and identified apices with specific characteristics (e.g., type, state, and diameter). When these simulated structures are stored in permanent files, they can later be represented by several different graphical or numerical outputs, using other software. When explicitly included in the system, the soil is also divided and coded into elements, either by horizontal soil layers (e.g., Porter et al., 1986; Bengough, 1997) or by cubic volume elements (Clausnitzer and Hopmans, 1994; Somma et al., 1998). It is then possible to use numerical methods to solve differential equations concerning water and mineral transport in the soil (Somma et al., 1998; see also Chapter 37 by Silberbush, Chapter 34 by Glass, and Chapter 35 by Jungk in this volume). The shoot system, when considered, was modeled as a global compartment (Clausnitzer and Hopmans, 1994) or as a simplified structure with several organs (Thaler and Page`s, 1998) providing carbohydrates and taking up water and nutrients. B.

Interactions with the Soil Environment

The first part of this chapter describes the means of simulating various aspects of the developmental program of the plant in architectural models. The common feature of these models is that they focus on the

Modeling Root System Architecture

plant root system, with rules attached to its components. It contrasts with other models predicting root density profiles, which use as basic elements the soil elements organized in space. This structuring approach is the result of a viewpoint which naturally accounts for the numerous features of root system architecture. However, the root system is also known for its plasticity, or even its opportunism, in relation to the soil spatial and temporal heterogeneity. Thus, the extension of the root is subject to numerous conditions and constraints which are non uniformly distributed, in space or time, and have several interacting effects on the whole architecture. Several authors have underlined this difficulty and have shown the relevance of architectural models to help integrating, at the root system level, the response of its components in relation to the diversity of situations they may experience. The general idea formulated in the models is that the root component that appears at a given time and in a given location is faced to specific environmental conditions which it responds to or interacts with. The environment can be described as a fixed map, or can be modeled itself as a dynamic system, with some of its properties changing with time, particularly under the root system influence. A large number of physical, chemical, or biotic characteristics may have a more or less significant impact on root development. Therefore, the first steps in modeling are to sort and select factors among all the possible influences, to give priority to the major ones (which are not always those for which the effects are best known), and to identify the most affected processes. Thus, architectural models are integrating frames for modules that relate soil variables to developmental variables, at the root scale. Up to now, the process of root growth has received the main attention. 1.

Temperature

Soil temperature is an extremely variable parameter, to which root systems are known to respond (Porter et al., 1986; Diggle, 1988; Clausnitzer and Hopmans, 1994). Given the time step of these models, generally about 1 day, seasonal variations only were considered. Diurnal variations, whose amplitude may be large in surface horizons, with maximal values typically higher than the optimal temperatures, were usually left out. In the Rootmap model, for example (Diggle, 1988), time is expressed in cumulated temperature for each apex, and growth rates are expressed in unit length per unit of thermal time (degreeday). Thus, tempera-

367

ture variations were taken into account according to a linear model, between a threshold temperature and an optimal temperature, provided that a dynamic map of temperatures is supplied. The model of Porter et al. (1986) relies on the same assumption of a linear influence of temperature and includes a module to calculate the variations of temperature with depth. In the model of Clausnitzer and Hopmans (1994), a maximal growth rate is defined at an optimal temperature. The actual growth rate is calculated by reducing this value with a multplicative factor (between 0 and 1) calculated from a sine function of the temperature, with a minimum and maximum located at the threshold and optimal temperatures. In these models, temperature also determines the progress of acropetal branching, since the time lag between initiation and emergence of a lateral root is expressed in thermal time. The emergence kinetics for the nodal roots of cereals was also linked to thermal time (e.g., Page`s et al., 1989). Even though temperature may have many other influences on root development, including growth directions and branching densities (Coutts, 1989; Chapter 41 by McMichael and Burke in this volume), such relationships have not yet been integrated into the models. 2.

Soil Strength

The mechanical impedence of the soil to root penetration is a major factor in many soils, which has to be considered to understand the root system architecture and especially the rooting depth (Bengough, 1997; Chapter 45 by Masle in this volume). Unfortunately, this mechanical resistance is not easy to characterize by a relevant variable. It depends on several other soil characteristics, among which some are complex or exhibit short-term variations. They include soil texture, structure, bulk density, and water content. However, several authors (e.g., Bengough and Mullins, 1990) have obtained clear experimental relationships between root growth rate and penetration resistance, as measured with a penetrometer. These empirical relationships, as well as relationships between soil resistance and other soil characteristics (e.g., bulk density and water content), were used as basic functions to integrate this aspect of root growth in architectural models (Clausnitzer and Hopmans, 1994; Bengough, 1997). Regarding elongation, the general trend was to reduce the potential growth rate by a multiplicative coefficient (beween 0 and 1), calculated by a predictive function of the soil penetration resistance (Clausnitzer and Hopmans, 1994; Bengough, 1997).

Page`s

368

Mechanical constraints also influence the direction of root growth, leading to tortuous trajectories and sometimes global directional trends (Fig. 3). Such effects have been modeled by combining several influences in the growth direction, each influence being represented by a vector contributing through a vectorial sum to the resulting growth direction (Diggle, 1988; Page`s et al., 1989; Clausnitzer and Hopmans, 1994). The general anisotropic mechanical constraint is represented by a random vector, generating noise in the tra-

jectory, whereas local resistance gradients are represented by a directed vector, giving an organized trend to the trajectory (Clausnitzer and Hopmans, 1994; Fig. 3). 3.

Water and Mineral Availability

Combining root system architecture modeling and water uptake modeling, Bengough (1997) and Clausnitzer and Hopmans (1994) studied the influence

Figure 3 Observed (top) and simulated (bottom) root systems of Juncus squarrosus in a heterogenous soil environment. The model includes functions of root growth rate and growth direction according to mechanical impedance. (From Clausnitzer and Hopmans, 1994.)

Modeling Root System Architecture

of soil water content via the mechanical impedance associated with soil drying. Thus, the simulated effect of water content variations is indirect. The assimilate supply to the root system is also indirectly dependent on water availability in the model of Clausnitzer and Hopmans (1994). Concerning mineral distribution in the soil, Somma et al. (1998) suggested a module simulating a direct influence of mineral concentration on root elongation. According to this module, growth is optimal in a given range of concentration. Below and beyond this opimal range, growth is linearly reduced, until two other limits, below and beyond which there is no more growth. It is worth mentioning that the influence of nitrate availability on root branching density, although shown by several authors (e.g., Drew, 1975; Drew and Saker, 1975; Granato and Raper, 1989), has not yet been included in such models. However, its role is probably important for simulating adaptation of the root system to heterogeneous distribution of the nitrogen resources (Robinson, 1994). C.

Interactions Within the Plant (Photoassimilate Availability)

In addition to interactions with the external environment, the root system components have numerous relationships, between each other and with the rest of the plant. These endogenous interactions have often been neglected or have not been specified in models, for obvious reasons of simplification, although their role in architecture genesis is fundamental. Considering independently local interactions between the development of the root system components and their surrounding environment may lead to lack of realism. Developmental correlations within the plant were described long ago (e.g., Dyanat-Nejad and Neville, 1972; Champagnat, 1974; Champagnat et al., 1974; Wightman and Thiman, 1980; Favre, 1985). They lead to various phenomena (growth compensation, synchronization, competition, inhibition, promotion), the determining factors of which are complex. Within the root system particularly, growth reallocation occurs when axis growth has stopped (Torrey, 1976; Lamond et al., 1983; Amin et al., 1987; Atzmon et al., 1994a,b) or when different parts of the root system are subjected to variable conditions of water and mineral availability (Drew and Saker, 1975; Coutts and Philipson, 1976; Granato and Raper, 1989). Interactions with the shoot system have also been observed. For example, the alternating growth of shoots and roots, in the rythmically growing

369

species Hevea brasiliensis, was described by Thaler and Page`s (1996a). It revealed tight interactions between the shoot and the root system, showing a clear control of the root system development and architecture by interactions within the plant (see also Chapter 12 by Reich in this volume). In order to describe such interactions, phenomenological models were designed to synchronize the development of the nodal roots through the development (Porter et al., 1986). Close relationships between the rate of leaf emergence and the rate of nodal root emergence have been shown in several cereals, such as wheat (Klepper et al., 1984), maize (Picard et al., 1985; Pellerin, 1993), and rice (Nemoto and Yamazaki, 1986). In such plants the phyllochron can be considered as a plant internal clock that determines nodal root initiation. Among the numerous factors likely to lead to interactions within the whole plant, photoassimilate availability has been considered by root system modelers to be the most significant (Clausnitzer and Hopmans, 1994; Nielsen et al., 1994; Thaler and Page`s, 1998; Bidel et al., 2000a). The dependence of the root system on carbohydrate supply from the shoot and the distance of the sinks from the carbohydrate sources have a profound effect on root system architecture (Aguirrezabal et al., 1993). In contrast with some crop models which formalized this interaction via a global sink function of the root system, architectural models have contributed to specify a more realistic sink function (distributed in time and space) and to assign it to certain structures within the root system. Using this approach, Nielsen et al. (1994) evaluated and located, using a modeling tool, the carbon cost of the root system contruction. The model was used to estimate the overall construction cost as it can be measured locally by respiration (Bidel et al., 2000b), biomass deposition, and exudation. This approach showed the spatial variations in carbon allocation, and the effect of architectural diversity on carbohydrate cost. In the model of Clausnitzer and Hopmans (1994), the interaction between the root and the shoot systems is considered in terms of water uptake (by the roots) and photoassimilate supply (by the shoots). Assimilate production is related to the amount of water absorbed and transpired, through a water use efficiency coefficient. The part of biomass allocated to the roots is shared between axes, proportionately to their demand. This is calculated from their individual potential growth rate after taking into account the possible reductions related to the soil conditions surrounding

Page`s

370

the root tips. This model makes it possible to represent time variations in terms of supply and demand, and growth compensations. Roots located in the most favorable environment will also have the largest demand, and subsequently the largest supply, at the expense of less favorably located roots. In the model of Thaler and Page`s (1998), carbohydrate allocation is also proportionate to the demand, but demand is not calculated from typical growth functions associated to each root type. It uses the apical diameter of each root at each time step. The basic assumption is that the larger the meristem volume (tightly related to apical diameter), the stronger the meristemic sink and the higher the potential elongation of the root. The meristem volume is not definite, but subjected to buffered variations at each time step: it may increase when carbohydrate supply is high and decrease when it is low. Moreover, the initial apical diameter of the root is the result of both its architectural location (size of the mother root) and carbohydrate availability during the primordium development. This model also represents growth compensation, because the assimilate resources are shared and may limit growth under specific conditions. It may also account for the large plasticity in root thickness and in their carbon demand according to the availability within the plant. In contrast to the previous two models, which consider an overall amount of carbon resource shared between the demanding sinks and give a common availability, the model of Bidel et al. (2000a) specifies the transport of carbohydrate resources, thus calculating a local availability for each sink. This approach accounts for local compensation phenomena, which operate for example between a mother root and its daughter roots because they are proximal sinks. D.

Tools for Translating and Simulating

As in other fields using modeling and simulation, advances in simulating root system architecture are clearly consistent with those made in computer science concerning hardware and software. The first model published by Hackett and Rose (1972a,b) was an algebraic model devoted to the analytical calculation of a number of state variables, such as root length, root number, and root surface or volume, classified by branching order. Though this approach did not include the spatial features of the architecture, the general idea of the model was based on an architectural vision of the root system. It is highly probable that the analytical formulation of the model, with strong simplifica-

tions such as the uniformity of growth rates, was imposed by the lack of computer technology available. The model proposed by Lungley (1973) took advantage of advances in this field, allowing numerical formulation, coding in a programming language (Fortran), and graphical representation of outputs. This model was in two dimensions and was used to calculate one-dimensional root density profiles. It is worth mentioning that Lungley’s work was consistent with this traditional representation of the root system, and therefore ensured the link between an architectural representation and a density profile representation. The first three-dimensional models of root system architecture were published independently by Diggle (1988) and by Page`s and Aries (1988). Both these models were written in programming language Pascal, which made this sort of programming easier, by using structured programming and user-adapted data structures. Since then, various significant advances have been made concerning modeling tools, which make developing and using elaborated models much easier. Regarding programming languages first, recent models make extended use of object oriented modeling methods (such as OMT, or UML—Rumbaugh et al., 1991; Muller, 1997) and associated languages, also object oriented (Smalltalk, C++, Java). These tools are particularly adapted to complex applications (Muller, 1997; Bidel et al., 2000a). The biological objects that we call the components (such as the apices, segments, roots) have a computer representation closer to the real (biological) world, being modeled as entities containing both attributes (object characteritics) and computer functions (permitting to account for the behavior). These autonomous objects (computer scientists say ‘‘encapsulated’’) can also interact with each other by sending messages. The recent development of computer graphics associated with an increased computing power allows a direct and realistic representation of the simulated structures. These software programs carry out image synthesis, by modeling the visual effects produced by the simulated volumes and surfaces, and therefore significantly contribute to the development of biological models. Such programs are now available in the public domain (e.g., Geomview—Phillips, 1996). In addition to these advances, a specific language for simulating biological structures was developed by Lindenmayer (1968, 1971), called the L-System. This language uses an axiom (i.e., an initial structure as starting point), modules (i.e., elementary components of the structure), and production rules operating on

Modeling Root System Architecture

the modules (i.e., formal translation of morphogenetic rules). It allows translation quite directly of the various concepts presented previously, even though it has not been applied that much to root systems (the hidden half!). Thanks to subsequent improvements, now included in the syntax, this language also permits to capture a number of interaction models (context-sensitive L-System), either between the system modules or with the environment. Moreover, this formal language, which was rather theoretical at the beginning, has been complemented with specific software (Pruzinkiewicz and Lindenmayer, 1990; Kurth, 1994). It enables the direct interpretation of the models and their simulation, and the production of graphical outputs (tool Grogra, by Kurth, 1994; tools L-Studio and VLab, by Pruzinkiewicz, 1998).

III.

EXAMPLES OF MODELING APPROACHES

After this analytical presentation of the knowledge and hypotheses which were put together in architectural models, we will now illustrate the application of the approach, by concrete examples of models or pieces of modeling approach, showing more specifically some research topics in which models may be helpful. A.

Estimation of Parameters (Calibration)

One of the most recurrent criticisms against architectural models is the large number of parameters included (Passioura, 1996). This number is actually highly variable, depending on the level of detail and the number of interactions taken into account. More specifically, it first depends on the number of different types of axes which have been identified, because each type is generally associated with a set of parameters. It also depends on the number of developmental processes described: the simplest models considered only two or three basic processes (emission, elongation, acropetal branching), whereas more complete models integrated other processes, described previously (e.g. reiteration, decay-abscission, radial growth). Each new process is formalized using a module, which requires new parameters to be included. Lastly, each process runs in relation to the environment sensu largo (i.e., endogenous and exogenous), using specific functions which can be rather complex. Thus, the total number of parameters may vary greatly, from 20 to 200. However, the total number of parameters only gives a limited vision on this question, the actual objec-

371

tives of the modeling approach being more conclusive. Moreover, it seems essential to distinguish on one hand the parameters that are globally fitted, matching overall outputs with data, and on the other hand those parameters that are measured directly, or fitted locally on particular pieces of data, independently from one another. This last type is mostly used in architectural models, and therefore the large number of parameters is a less acute disadvantage. Moreover, these parameters, which have a concrete biological meaning, are easier to use and compare in relation to various situations. Among the concrete parameters, we find for example branching densities, or time lag before branching. Parallel to this calibration step, it is generally appropriate to use sensitivity analyses, which help to establish a hierarchy among these parameters, and to yield a better idea of the expected accuracy according to the objectives determined for the main variables of interest. Measuring empirical parameters, i.e., those directly derived from actual measurements, refers to the methodological problems that are associated to root system studies (see also Chapter 18 by Polomski and Kuhn in this volume). Some of these parameters are typically geometrical (e.g., branching density, trajectory parameters, branching angles) and are therefore directly measurable on single observations (destructively if necessary), whereas others include a temporal dimension (e.g., time lag, growth rate or duration, life expectancy) therefore requiring continuous or repeated observations using temporal markers. Tardieu and Pellerin (1990), Pellerin and Page`s (1994), and Page`s and Pellerin (1994) illustrated these procedures, applied to calibration of a model of maize root system architecture under field conditions. The parameters defining the emergence rate of nodal roots according to cumulated temperatures were estimated from counting nodal roots at the base of plants which were periodically excavated during the season. Periodic excavation of primary roots (seminal and nodal) made it possible to estimate the parameters concerning their trajectories and branching density, as well as growth rates using both length (measured directly on each excavated root) and age (estimated from the emission module). Similarly, the geometrical parameters of lateral roots (angles, trajectories, branching densities) were estimated from photographs of individually excavated roots, whereas growth parameters were calibrated from the length and position relative to the apex of the mother root. The distance to the apex made it possible to estimate the date of emergence, and therefore their age at excavation time. To

Page`s

372

complement these data obtained in the field, other data from the literature or independent observations in rhizotrons were used. Tsegaye et al. (1995a,b) have used a very similar approach to estimate the parameters of the Rootmap model (Diggle, 1988) for potted plants. B.

Evaluation of Outputs (Validation)

The validation of these models, although it is a vital part of the modeling approach, has not been much dealt with (Page`s et al., 1992; Tsegaye et al., 1995b; Page`s and Pellerin, 1996; Pellerin and Page`s, 1996). There are probably several reasons for that which are worth analyzing. Irrespective of the model, validation cannot be made in the absolute, but has to be related to a predetermined objective, the model being considered valid as long as it satisfies this initial objective. In many cases, however, the objectives assigned to the model were numerous, and not always well specified, because architectural models are potentially rich tools whose abilities are to be further investigated. Consequently, it is impossible to define clear validation criteria. Moreover, even when the objectives are predetermined, the formalization of validation criteria is not always a simple and straightforward step per se. A further difficulty arises from the comparison with data. The data sets used to test such models are rather scarce because they are difficult and costly to obtain. The diversity of the conditions required to refine validation makes it an acute problem, and limited data sets do not permit a powerful (discriminative) evaluation, given the intrinsic variations of the root system and the rhizosphere. Moreover, the data may also be subjected to numerous errors or uncertainties. Although subject to difficulties, validation is made, at least partially, by evaluating separately parts of the model. Some of these aspects can be illustrated by the work of Pellerin and Page`s (1996) and Page`s and Pellerin (1996), whose specific objective was to evaluate the capacity of a root system architectural model to predict the dynamics of the spatial distribution of maize roots. Previous studies showed the root distribution heterogeneity in various crops (e.g., Van Noordwijk and Brouwer, 1991) and more specifically in maize (Tardieu, 1988; Logsdon and Allmaras, 1991). Root distribution exhibits a spatial structure: at a large scale (dm to m), gradual variations can be observed, versus depth and distance from the row, and at a small scale (mm to cm), strong local variations are related to the aggregated pattern of root distribution. The question was to determine whether a simple model of the

root system architecture (Page`s et al., 1989) could simulate these main features of spatial distribution, as observed in the field. The data compared were horizontal root maps at different depth levels and at different dates (Pellerin and Page`s, 1996) as well as vertical maps at the flowering stage (Page`s and Pellerin, 1996). These works have shown the general ability of the model to simultaneously represent the various aspects of the data set, containing both spatial and temporal variations (Fig. 4). They have also shown that the main discrepancies between observed and simulated patterns were related to local compaction phenomena (especially in the plough pan). This pointed out that a better agreement could be expected by integrating the mechanical effect in the model. Using the model, it is also possible to evaluate the effect of the method used to draw vertical root maps. For example, some millimeters of soil are generally excavated around the roots to make them more visible, which increases the risk of overestimation of the root density, in comparison to what would be obtained on a strictly plane cross section. C.

Addressing Methodological Problems

Methodological problems encountered in the study of the root system come from its complexity, its fineness and fragility, and the opacity and cohesion of the medium in which it develops. In order to obtain data on its spatial distribution or development, many methods have been suggested, in which sampling procedures (and subsequent extrapolations) and indirect evaluation of target variables (via easier methods of measurement) take an important place. Such procedures rely more or less explicitly on theories and a priori assumptions on root system architecture, some of which have not been validated. The dynamic model of root system which can be produced by simulation models can help to improve the data acquisition procedures. Bengough et al. (1992) and Grabarnik et al. (1998) have studied some major geometrical characteristics of the root systems using model root systems and have evaluated the consequences of these characteristics on sampling schemes and indirect evaluation procedures. Page`s and Bengough (1997) tried to model root depth observations, as they are obtained from minirhizotron, using model root systems combined with cylinders of variable configuration. Their combined model made it possible to test theoretically several sampling designs considering variations in diameter, position, and inclination of the tubes. The simulation approach allowed

Figure 4 Distributions of the number of primary root intersections with horizontal trench walls in a maize stand versus the orthogonal distance to the plant row, at three dates after sowing, and three depths. These distributions were obtained from observed maps (solid lines) after excavation of horizontal trench walls, and from simulated maps (dashed lines) calculated from three-dimensional simulated root systems. (From Pellerin and Page`s, 1996.)

Modeling Root System Architecture 373

Page`s

374

comparison of a large number of different theoretical situations. In addition to their contribution to the design of sampling schemes, architectural models can help to fill the gaps in observations that are often scattered, partial, or at least discontinuous, either in time or space. Spek and Van Noordwijk (1994), for example, showed the relevance of their fractal model to simulate whole root systems using local measurements, achieved on very limited parts of the real root systems. This is particularly relevant for long-living and large root systems, such as tree root systems. D.

Consequences of Soil Constraints

The highly dynamic and interactive character of the soil–root system justifies developing integrated models among which are the characteristics of the root system architecture, its functioning, and its responses to soil constraints. The questions raised by modern agriculture, in a context of environmental protection, require new models with increased performance. The strong mechanical constraint imposed by soils on the root systems is an example of interaction between the root and the soil dynamics, with water acting as a mediator. In Bengough’s model (1997), some of the major dynamic features are combined in order to study the contribution of the soil mechanical and hydrodynamic properties, climatic demand, and root uptake and response within the overall system functioning. The model could account for great differences between crops regarding their rooting depth. The model of Somma et al. (1998), an extension of that of Clausnitzer and Hopmans (1994), generalized this approach in three-dimensional space and included more details about the effects of the mechanical constraint on root growth. It revealed additional aspects of the architectural plasticity, which emerged from the integrated system functioning. A complementary approach was suggested by Doussan et al. (1998a,b) for studying the consequences of the hydraulic organization of the root system architecture, in relation to the heterogeneity of water availability in the soil. This model permits evaluation of the spatial heterogeneity of water uptake and water potential within the root system, as a function of the variations in the soil water potential. E. Episodic Growth of the Root System Unlike episodic development of the shoot, which is well known and has been evidenced in many plants,

the temporal variations (phenology) of the root development are much less documented. Growth rate variations, not explainable by the heterogeneity of the rhizosphere, have been noticed on several species—on trees particularly (Head, 1967; Lyr and Hoffman, 1967; Riedacker, 1976; Atkinson, 1983). In young rubber trees, having a typical rhythmic shoot development (Halle´ and Martin, 1968), timedependent growth variations were also marked in some roots which exhibited alternations between elongation and rest periods, whereas others roots had a nearly continuous growth (Thaler and Page`s, 1996a). These variations also had consequences on the shape of the whole architecture, which also exhibited some episodic features. Along the main axes, zones where roots were dense and long alternated with zones where roots were rather short and scarce (Thaler and Page`s, 1996a). These complex phenomena with major architectural consequences emphasize the need to include within the models of the root system architecture some way of representing interactions between the root and the shoot system. Following this objective, Thaler and Page`s (1998) developed a model based on competition for carbon assimilates within the plant. This model combines a dynamic architectural description of the whole plant, with rules that govern the allocation of carbon assimilates among sink organs, in the shoot and root systems. It enables quantification of the organ requirements and consumption, and the kinetics of carbohydrate availability. It therefore helps to validate the hypothesis of carbohydrate competition in the determinism of an episodic root system development. F.

Simulating Infection and Disease Transmission

The root system is also the support and vector of many plant diseases. Several authors have shown the interest of combining models of the root system architecture with models of pathogen distribution and growth to investigate this complex and dynamic system (Bloomberg, 1979; Reynolds et al., 1986; Brown and Kulasiri, 1994, Gilligan et al., 1994; Page`s et al., 1995). In Hevea brasiliensis, for example, the Fomes fungus, causing root rot, is transmitted from one root system to another, through interroot contacts or close proximity. Then, it can develop along the roots and reach very quickly other parts of the root system (Tran Van Canh, 1982). In young plantations, the symptoms of the disease can generally be observed on some trees, from which the disease then spreads

Modeling Root System Architecture

as soon as root systems are large enough to be in contact with their neighbors. Such a phenomenon has been simulated by Page`s et al. (1995) using a root system architecture model which made it possible to calculate outbreak zones where roots from neighboring trees were very close to each other, and therefore predict the theoretical spread of the disease from the first infected parts. Such a model helps to evaluate the risks and measures that should be taken in order to control the disease.

IV.

FUTURE TRENDS AND NEW ISSUES

Modeling of root system architecture is aimed at understanding either root system development, function, or the methodological aspects required for root research. These models have now reached a level of maturity which makes their development and utilization credible in relation to several scientific issues. It is also a rather new approach, with a strong potential of development in several complementary directions that we shall now discuss briefly. A.

Integration of the Root System into the Soil–Plant System

When referring to former models developed in the 1970s and 80s to describe the development of root systems, we have mentioned the strong duality between ‘‘root density’’ models, starting from the soil and oversimplifying the root system, and root system architecture models, based on morphogenetic processes described from the plant, hardly tackling the interactions between the roots and their environment. This duality was justified, because morphology of soils with high constraints and heterogeneity clearly affects the root systems that show plasticity, whereas the internal processes of architectural genesis provide the structure of the root system, which is evident mostly in favorable or homogeneous soils. Since then, the two points of view (soil and plant) have got closer, especially because of the advances achieved in simulating root system architecture based on a morphogenetic approach. Gradually, these models have become capable of integrating the interactions between root components and their surrounding environment, with the later being either exogenous (soil, including biota) or endogenous (plant). However, there is still much to do in this area of soil–plant interactions.

375

1.

Merging with Models of the Soil and Biotic Environment

Regarding the soil properties to be first considered, we have seen that some variables have been emphasized because they play a major role in many situations, but not universally of course. Simulation in other specific conditions, for which the limiting factors are different, e.g., salinity, nitrogen, or oxygen availability, should involve the formalization and integration of other interactive mechanisms. Moreover, the soil variables considered up to now exhibit almost continuous variations in the soil. Therefore, in addition to a simple substitution, the type of variable considered can also differ in its distribution and effect, leading to a possibly new approach (Greene, 1991). For example, the integration of biotic interactions requires taking into account the architectural diversity within the root system, as well as the complex and time-dependent distribution of soil organisms. The interaction with soil structure elements, such as different pore types (cracks, textural pores, biopores) which have dimensions close to that of root tips, may result in exciting and promising modules, in which architectural diversity in root diameters is a major factor. Another great challenge regards the description of the spatial rhizosphere environment (Hinsinger, 1998), instead of an average environment (bulk soil), as it can be measured at the decimeter scale. Local measurements, when possible, or local models reveal large gradients between overall water and mineral availability and availability at the soil–root contact area, experienced by the root. From both the experimental and modeling points of view, it raises many methodological problems. The finite element method, as suggested by Lafolie et al. (1991) and Clausnitzer and Hopmans (1994), calls for division of the soil space into very fine elements around the roots, i.e., according to a highly complex geometry, which is a complicated numerical problem. Therefore, alternative approaches are needed in association with several levels of details during simulation. 2.

Merging with Models of ‘‘Endogenous Environment’’ Within the Plant

The concept of environment can easily be extended to internal conditions, which are found and perceived by the meristems (Page`s, 2000). Thus, the ‘‘endogenous environment’’ includes resource availability for growth (mainly carbohydrates, water, and nitrogen compounds) and possibly the presence of signaling compounds.

Page`s

376

Current models have started to link the fate of root components to carbon resource availability, and therefore contribute to specify a carbohydrate availability environment (Clausnitzer and Hopmans, 1994; Thaler and Page`s, 1998). Other resources needed for the development of new structures should be included in the future, especially water and nitrogen. The approach of Doussan et al. (1998a,b) shows some means of including the endogenous water environment by integrating the combined effects of climatic demand, availability in the soil, and transport within the root system. The endogenous environment has been considered as an overall environment by Clausnitzer and Hopmans (1994) and by Thaler and Page`s (1998). In these models, all the root components experience identical assimilate availability at the same time. To account for local competition phenomena, or apparent priority rules between meristems, it seems necessary to consider the spatial variation of the availability variable. Bidel et al. (2000a) proposed an approach in this direction for carbohydrates by modeling carbohydrate transport and consumption throughout the root system architecture.

about carbohydrates (Muller et al., 1998; Bidel et al., 2000b), water (Fraser et al., 1990; Durand et al., 1997), and nitrogen (Gastal and Nelson, 1994). Regarding mechanical constraints, architectural models have mainly integrated empirical relationships as submodels. New insight could probably be gained using architectural models to provide an organized local information on root tips, such as diameter and turgor pressure (from their endogenous environment). The influence of the apical movement, as well as exudation, is probably a major key for a better understanding of elongation and growth direction of the roots. Architectural models have highlighted the need for modules describing root growth direction. A number of studies have dealt with the positive gravitropism in radicles just emerging from the seed, whereas other roots, which seem to exhibit numerous ways of orienting themselves (Coutts, 1989; see Chapter 29 by Porterfield in this volume), have not been studied so extensively.

B.

A large majority of existing architectural models classify emerging roots according to predetermined types, which tightly control their subsequent development. The plasticity of root systems tends to be underestimated by this simulation method, particularly in very heterogeneous and constraining soils, where lateral roots are much affected by the fate of their mother root. It is very important to make significant progress in modelling the genesis of the type of axis, or even to re-evaluate these constraining categories. This point requires the study of early developmental stages of the root meristems, from initiation to emergence. Most of the variations in meristematic volume are acquired during these early stages, with significant consequences on subsequent growth of axes (Coutts, 1987; Page`s, 1995; Thaler et Page`s, 1998). The availability of carbon and nitrogen resources is probably critical during these stages, contributing both to the number and growth characteristics of the promeristems (Zhang and Forde, 1998).

Modeling the Response of Root Meristems to Their Local Environment

Modeling the architecture also raises new relevant questions for studies at the lower organization level—that of the root system components, and particularly meristems. Architectural models should be considered as tools for integrating knowledge on individual meristems, as well as for specifying variations in the situations experienced by meristems in a root system. This variability is built at the root system level, during its development and functioning (see also Chapter 9 by Waisel and Eshel in this volume). It regards both the types of roots emerging during development, and the actual environmental conditions (exogenous and endogenous) in which they develop. 1. Root Growth In terms of main resources, architectural models need to specify the root component requirements according to developmental stages and the effects of availability on these stages. Therefore, studies are required at the root segment level (Nielsen et al., 1994; Bidel et al., 2000b) or even at finer scales, using the theory and techniques of spatial analysis (Silk and Erickson, 1979; Silk, 1994). Such studies have been carried out recently, and have provided many interesting results

2.

3.

Root Initiation and Primordium Development

Interactions with Microorganisms, Mycorrhizae

There is much to gain by considering the root with its surrounding microorganisms, and particularly mycorrhizal fungi, which have a considerable functional significance in a large number of plant families (see

Modeling Root System Architecture

Chapter 50 by Kottke in this volume). No doubt that in this field also, the valuable dialog between specialists and the description of interactions between roots and symbionts require a detailed description of the root system architecture. Such a description should make it possible to specify favorable sites on the root system and the specialized effects of the microorganisms on root development (Schellenbaum et al., 1991; Wullscheleger et al., 1994). C.

New Prospects of Development and Applications

In parallel to these possible improvements, modeling the root system architecture makes it possible to investigate new fields of application. For example, modeling the anchorage ability of the root system should benefit from combining the architecture of the main root axes, the variation of their mechanical properties, and their response to the mechanical constraints transmitted from the shoot to the root system (cambium functions especially). Such models should also renew the study of various fluxes, either within the plant, from soil to plant, or from plant to soil, thus permitting a spatial representation of phenomena occurring at various scales in the plant and the soil. In addition to quantitative overall aspects, regarding water and carbon budget especially, this refined integration of fluxes and signals from the root system should be helpful to better understand some behaviors of the shoot organs, with various possible applications to morphogenesis, stomatal regulation, fruit quality, etc. In the general context of a modern agriculture, stressing environmental protection, such models should also contribute to genetic studies and to selecting adapted genotypes to many adverse situations. To develop models able to fulfill this crucial issue, it is necessary to contribute to the identification of most genetically controlled mechanisms and to integrate them into models. Provided such an integration is successful, architectural models will become major tools to make the necessary link between fundamental mechanisms and the architectures generated under sets of environmental constraints. Further advances will be made by increasing the power of computing tools. These tools should make biological (or biophysical) models easier to formalize, and particularly the consistent association between plant structures and their specific functioning traits (including genetic regulation) and the simulation of various interactions with the environment. Ergonomics and graphical visualization are obviously

377

required in order to get a fast and integrated rendering of properties emerging at the plant or crop level, based on the knowledge distributed on elementary components. Because of powerful and user-friendly computer facilities, it may be tempting to integrate the effect of many factors in many processes, so as to cover more fields of application. However, to be useful, models should remain easy to control and specifically oriented. Partial integration, by keeping a particular viewpoint on the system, and leading to a given hierarchy among the relevant mechanisms, is the most promising way for the future. Simplicity and parsimony are of major importance for predictive goals. This point should be the subject of specific research, aiming at simplifying architectural models as much as possible to facilitate and extend their use in related fields using plant simulators. ACKNOWLEDGMENTS I would like to thank Gilles Vercambre for drawing Fig. 2, and Ge´raldine Rigou for her language corrections. REFERENCES Adiku SGK, Braddock RD, Rose CW. 1995. Simulating root growth dynamics of cowpea under varying soil conditions. Proceedings of MODSIM95, International Congress on Modelling and Simulation, University of Newcastle, Newcastle, Australia, November 1995, pp. 27–30, 57–61. Aguirrezabal A, Pellerin S, Tardieu F. 1993. Carbon nutrition root branching and elongation: can the present state of knowledge allow a predective approach at a whole-plant level? Environ Exp Bot 33:121–130. Aguirrezabal A, Tardieu F. 1996. An architectural analysis of the elongation of field-grown sunflower root systems. Elements for modelling the effects of temperature and intercepted radiation. J Exp Bot 47:411–420. Amin T, Beissalah Y, El Hajzein B, Neville P. 1987. Variations de la re´ge´ne´ration du pivot de jeunes plants de cheˆne vert (Quercus ilex L.) apre`s divers traumatismes. Ecol Medit 13:61–76. Asseng S, Richter C, Wessolek G. 1997. Modelling root growth of wheat as the linkage between crop and soil. Plant Soil 190:267–277. Atger C, Edelin C. 1992. Premie`res donne´es sur l’architecture compare´e des syste`mes racinaires et caulinaires des arbres. Can J Bot 72:963–975. Atkinson D. 1983. The growth, activity, and distribution of the fruit tree root system. Plant Soil 71:23–36.

378 Atzmon N, Salomon E, Reuveni O, Riov J. 1994a. Lateral root formation in pine seedlings. I. Sources of stimulating and inhibitory substances. Trees 8:268–272. Atzmon N, Reuveni O, Riov J. 1994b. Lateral root formation in pine seedlings. II. The role of assimilates. Trees 8:273–277. Baldwin JP, Nye PH, Tinker PB. 1973. Uptake of solutes by multiple root systems from soil. III. A model for calculating the solute uptake by a randomly dispersed root system developing in a finite volume of soil. Plant Soil 38:621–635. Barthe´le´my D, Edelin C, Halle´ F. 1989. Architectural concepts for tropical trees. In: Holm-Nielsen LB, ed. Tropical Forests: Botanical Dynamics, Speciation and Diversity. London; Academic Press, pp 31–42. Bengough AG. 1997. Modelling rooting depth and soil strength in a drying soil profile. J Theor Biol 186:327–338. Bengough AG, Mullins CE. 1990. Mechanical impedance to root growth: a review of experimental techniques and root growth responses. J Soil Sci 41:341–358. Bengough AG, Mackenzie CJ, Diggle AJ. 1992. Relations between root length densities and root intersections with horizontal and vertical planes using root growth modelling in 3 dimensions. Plant Soil 145:245–252. Berntson GM. 1994a. Modelling root architecture: are there tradeoffs between efficiency and potential or resource acquisition? New Phytol 127:483–493. Berntson GM. 1994b. Fractal analysis of plant root systems: how reliable are calculations of fractal dimensions. Ann Bot 73:281–284. Bidel LPR, Page`s L, Rivie`re L-M, Pelloux G, Lorendeau JY. 2000a. MassFlowDyn I: a carbon transport and partitioning model for root system architecture. Ann Bot 85:869–886. Bidel LPR, Renault P, Page`s L, Rivie`re L-M. 2000b. Mapping meristem respiration of Prunus persica (L.) Batsch roots: potential respiration of the meristems, O2 diffusional constraints, and combined effects on root growth. J Exp Bot 51:755–768. Bloomberg WJ. 1979. A model of damping-off and root rot of Douglas-fir seedlings caused by Fusarium oxysporum. Phytopathology 69:1072–1077. Bo¨hm W. 1979. Methods of Studying Root Systems. Ecological Studies 33. Berlin; Springer-Verlag. Brown TN, Kulasiri D. 1994. Simulation of Pinus radiata root system structure for ecosytem management applications. Simulation 62:286–294. Bruckler L, Lafolie F, Tardieu F. 1991. Modelling root water potential and soil-root water transport. II. Field comparisons. Soil Sci Soc Am J 55:1213–1220. Cahn MD, Zobel RW, Bouldin DR. 1989. Relationship between root elongation rate and diameter and duration of growth of lateral roots of maize. Plant Soil 119:271–279.

Page`s Cannell MGR. 1985. Dry matter partitioning in tree crops. In: Cannell MGR, Jackson JE, eds. Trees as Crop Plants. Huntington, New York; Titus Wilson & Son, pp 160–193. Champagnat P. 1974. Introduction a` l’e´tude des complexes de corre´lations. Rev Cytol Biol Veg 37:175–208. Champagnat P, Baba J, Delaunay M. 1974. Corre´lations entre le pivot et ses ramifications dans le syste`me racinaire de jeunes cheˆnes cultive´s sous un brouillard nutritif. Rev Cytol Biol Veg 37:407–418. Chapman SC, Hammer GL, Meinke H. 1993. A sunflower simulation model. I. Model development. Agron J 85:725–735. Charlton WA. 1967. The root system of Linaria vulgaris mill. II. Differentiation of root types. Can J Bot 45:81–91. Clausnitzer V, Hopmans JW. 1994. Simultaneous modeling of transient three-dimensional root growth and soil water flow. Plant Soil 164:299–314. Coutts MP. 1983. Root architecture and tree stability. Plant Soil 71:171–188. Coutts MP. 1987. Developmental processes in tree root systems. Can J For Res 17:761–767. Coutts MP. 1989. Factors affecting growth direction of tree roots. Ann Sci For 46:277–287. Coutts MP, Philipson JJ. 1976. The influence of mineral nutrition on the root development of trees. I. The growth of sitka spruce with divided root systems. J Exp Bot 27:1102–1111. Diggle AJ. 1988. ROOTMAP—a model in three-dimensional coordinates of the growth and structure of fibrous root systems. Plant Soil 105:169–178. Doussan C, Page`s L, Vercambre G. 1998a. Modelling the hydraulic architecture of root systems: an integrated approach of water absorption. I. Model description. Ann Bot 81:213–223. Doussan C, Vercambre G, Page`s L. 1998b. Modelling the hydraulic architecture of root systems: an integrated approach of water absorption. II. Distribution of conductances and consequences. Ann Bot 81:225–232. Durand J-L, Gastal F, Etchebest S, Bonnet AC, Ghesquie`re M. 1997. Interspecific variability of plant water status and leaf morphogenesis in temperate forage grasses under summer water deficit. Eur J Agron 7:99–107. Drew MC. 1975. Comparison of the effects of a localized supply of phosphate, nitrate, ammonium and potassium on the growth of the seminal root system, and the shoot, in barley. New Phytol 75:479–490. Drew MC, Saker LR. 1975. Nutrient supply and the growth of seminal root system in barley. II. Localized, compensatory increases in lateral root growth and rates of nitrate uptake when nitrate uptake is restricted to only part of the root system. J Exp Bot 26:79–90. Dyanat-Nejad H, Neville P. 1972. Etude expe´rimentale de l’initiation et de la croissance des racines late´rales pre´coces du cacaoyer (Theobroma cacao L.). Ann Sci Nat Bot 13:211–246.

Modeling Root System Architecture Ennos AR, Fitter AH. 1992. Comparative functional morphology of the anchorage systems of annual dicots. Funct Ecol 6:71–78. Esau K. 1989. Anatomy of Seed Plants. 2nd ed. New York; John Wiley. Eshel A. 1998. On the fractal dimensions of a root system. Plant Cell Environ 21:247–251. Favre JM. 1985. Modification expe´rimentale de l’architecture racinaire d’ Hevea brasiliensis. Phytomorphology 35:35–46. Fitter AH. 1982. Morphometric analysis of root systems: application of the technique and influence of soil fertility on root system development in two herbaceous species. Plant Cell Environ 5:313–322. Fitter AH. 1986. The topology and geometry of plant root systems: influence of watering rate on root system topology in Trifolium pratense. Ann Bot 58:91–101. Fitter AH. 1987. An architectural approach to the comparative ecology of plant root systems. New Phytol 106(suppl):61–77. Fitter AH. 1996. Characteristics and functions of root systems. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. 2nd ed. New York; Marcel Deker, pp 1–20. Fitter AH, Stickland TR. 1992. Fractal characterization of root system architecture. Funct Ecol 6:632–635. Fitter AH, Stickland TR, Harvey ML, Wilson GW. 1991. Architectural analysis of plant root systems. 1. Architectural correlates of exploitation efficiency. New Phytol 118:375–382. Fraser TE, Silk WK, Rost TL. 1990. Effects of low water potential on cortical cell length in growing regions of maize roots. Plant Physiol 93:648–651. Gastal F, Nelson CJ. 1994. Nitrogen use within the growing leaf blade of tall fescue. Plant Physiol 105:191–197. Gilligan CA, Brasset PR, Campbell A. 1994. Modelling of early infection of cereal roots by the take-all fungus: a detailed mechanistic simulator. New Phytol 128:515– 537. Grabarnik P, Page`s L, Bengough AG. 1998. Geometrical properties of simulated maize root systems: consequences for length density and intersection density. Plant Soil 200:157–167. Granato TC, Raper CD. 1989. Proliferation of maize (Zea mays L.) roots in response to localized supply of nitrate. J Exp Bot 40:263–275. Greene N. 1991. Detailing tree skeletons with voxel automata. SIGGRAPH ’91 Course Notes on Photorealistic Volume Modeling and Rendering Techniques, Association for Computing Machinery, New York. Guingo E, He´bert Y. 1997. Relationships between mechanical resistance of the maize root system and root morphology and their genotypic and environmental variation. Maydica 42:265–274.

379 Hackett C. 1969. Quantitative aspects ot the growth of cereal root systems. In: Whittington WJ, ed. Root Growth. London; pp 134–147. Hackett C, Rose DA. 1972a. A model of the extension and branching of a seminal root of barley, and its use in studying relations between root dimensions. I. The model. Aust J Biol Sci 25:669–679. Hackett C, Rose DA. 1972b. A model of the extension and branching of a seminal root of barley, and its use in studying relations between root dimensions. II. Results and inferences from manipulation of the model. Aust J Biol Sci 25:681–690. Halle´ F, Martin R. 1968. Etude de la croissance rythmique chez l’he´ve´a. Adansonia 8:475–501. Halle´ F, Oldeman RAA. 1970. Essai sur l’Architecture et la Dynamique de Croissance des Arbres Tropicaux. Paris; Masson. Halle´ F, Oldeman RAA, Tomlinson PB. 1978. Tropical Trees and Forests. Berlin; Springer. Hansen GK. 1975. A dynamic continuous simulation model of water state and transpiration in the soil-plant-atmosphere system. I. The model and its sensitivity. Acta Agric Scand 25:129–149. Head GC. 1967. Effects of seasonal changes in shoot growth on the amount of unsuberized root on apple and plum trees. J Hort Sci 42:169–180. Henderson R, Ford ED, Renshaw E, Deans JD. 1983a. Morphology of the structural root system of sitka spruce. 1. Analysis and quantitative description. Forestry 56:121–135. Henderson R, Ford ED, Renshaw E. 1983b. Morphology of the structural root system of sitka spruce. 2. Computer simulation of rooting patterns. Forestry 56:137–153. Hinsinger P. 1998. How do plant roots acquire mineral nutrients? Chemical processes involved in the rhizosphere. Adv Agron 64:225–265. Ho LC. 1988. Metabolism and compartmentation of imported sugars in relation to sink strength. Annu Rev Plant Physiol 39:355–378. Hoogenboom G, Huck MG. 1986. ROOTSIMU V4.0 A dynamic simulation of root growth, water uptake, and biomass partitioning in a soil-plant-atmosphere continuum: update and documentation. Agronomy and Soils Departmental Series 109, Auburn University, Auburn, AL. Huck MG, Hillel D. 1983. A model of root growth and water uptake accounting for photosynthesis, respiration, transpiration, and soil hydraulics. Adv Irrig 2:273–333. Jones CA, Kiniry JR, eds. 1986. CERES-Maize: A Simulation Model of Maize Growth and Development. Texas A&M University Press. Jones CA, Bland WL, Ritchie JT, Williams JR. 1991. Simulation of root growth. In: Hanks J, Ritchie JT, eds. Modeling Plant and Soil Systems. Agron Monogr 31:92–124.

380 Jourdan C, Rey H, Gue´don Y. 1995. Architectural analysis and modeling of the branching process of the young oil-palm root system. Plant Soil 177:63–72. Jourdan C, Rey H. 1997. Modelling and simulation of the architecture and development of the oil-palm (Elaeis guinensis Jacq.) root system. I. The model. Plant Soil 190:217–233. Kahn F. 1978. Analyse structurale des syste`mes racinaires des plantes ligneuses de la foreˆt tropicale dense humide. Candollea 32:321–358. Klepper B, Belford RK, Rickman RW. 1984. Root and shoot development in winter wheat. Agron J 76:117–122. Klepper B, Rickman RW. 1990. Modeling crop root growth and function. Adv Agron 44:113–132. Kurth W. 1994. Growth grammar interpreter GROGRA 2.4: a software tool for the three-dimensional interpretation of stochastic, sensitive growth grammars in the context of plant modeling. Forschungszentrum Waldo¨kosysteme der Universita¨t Go¨ttingen, Go¨ttingen, Germany. Kutschera L. 1960. Wurzelatlas mitteleuropa¨ischer Ackerunkra¨uter und Kulturpflanzen. Frankfurt am Main, Germany: DLG-Verlag. Lafolie F, Bruckler L, Tardieu F. 1991. Modelling root water potential and soil-root water transport. I. Model presentation. Soil Sci Soc Am J 55:1203–1212. Lamond M, Takavol R, Riedacker A. 1983. Influence d’un blocage de l’extre´mite´ du pivot d’un semis de cheˆne, sur la morphogene`se de son syste`me racinaire. Ann Sci Forest 40:227–250. Legay J-M. 1997. L’Expe´rience et le Mode`le. Un Discours sur la Me´thode. Sciences en Question. Paris; INRA Editions. Le Roux Y. 1994. Mise en place de l’architecture racinaire d’Hevea brasiliensis. Etude compare´e du semis et de la microbouture. PhD Thesis, Universite´ Aix-Marseille, Marseille, France. Le Roux Y, Page`s L. 1994. De´veloppement et polymorphisme racinaire chez de jeunes semis d’he´ve´a (Hevea brasiliensis). Can J Bot 72:924–932. Le Roux Y, Page`s L. 1996. Re´action ge´otropique des diffe´rents types de racines chez l’he´ve´a (Hevea brasiliensis). Can J Bot 74:1910–1918. Lindenmayer A. 1968. Mathematical models for cellular interaction in development. Parts I and II. J Theor Biol 18:280–315. Lindenmayer A. 1971. Developmental systems without cellular interactions, their languages and grammars. J Theor Biol 30:455–484. Logsdon SD, Allmaras RR. 1991. Maize and soybean clustering as indicated by root mapping. Plant Soil 131:169–176. Lungley DR. 1973. The growth of root systems: a numerical computer simulation model. Plant Soil 38:145–159.

Page`s Lyford WH, Wilson BF. 1964. Development of the root system of Acer rubrum L. USDA Forest Service, Harvard Forest Paper 10:1-17 Lynch J. 1995. Root architecture and plant productivity. Plant Physiol 109:7–13. Lynch JP, Nielsen KL. 1996. Simulation of root system architecture. In: Waisel Y, Eshel A, Kafkafi U, eds. Plant Roots: The Hidden Half. 2nd ed. New York; Marcel Deker, pp 247–257. Lynch JP, Nielsen KL, Davis RD, Jablokow AG. 1997. Simroot: modelling and visualisation of root systems. Plant Soil 188:139–151. Lyr H, Hoffmann G. 1967. Growth rates and growth periodicity of tree roots. Int Rev For Res 2:181–236. Mandelbrot B. 1983. The Fractal Geometry of Nature. New York; Freeman. Muller B, Stosser M, Tardieu F. 1998. Spatial distributions of tissue expansion and cell division rates are related to irradiance and to sugar content in the growing zone of maize roots. Plant Cell Environ 21:149–158. Muller P-A. 1997. Mode´lisation objet avec UML. Paris; Eyrolles. Nemoto K, Yamazaki K. 1986. The successive stem growth and its relations to the diameters and the numbers of primary roots on main axes of rice plants. Jpn J Crop Sci 55:352–359. Nielsen KL, Lynch J, Jablokow AG, Curtis PS. 1994. Carbon cost of root systems: an architectural approach. Plant Soil 165:161–169. Oldeman RAA. 1974. L’Architecture de la Foreˆt Guyanaise. Doc 73. Paris; Orstom. Ozier-Lafontaine H, Lecompte F, Sillon J-F. 1999. Fractal analysis of the root architecture of Gliricidia sepium for the spatial prediction of root branching, size and mass. Model development and evaluation in agroforestry. Plant Soil 209:167–180. Page`s L. 1995. Growth patterns of the lateral roots in young oak (Quercus robur L.) trees. Relationship with apical diameter. New Phytol 130:503–509. Page`s L. 2000. How to include organ interactions in models of the root system architecture? The concept of endogenous environment. Ann For Sci 57:513–520. Page`s L, Aries F. 1988. SARAH : mode`le de simulation de la croissance, du de´veloppement, et de l’architecture des syste`mes racinaires. Agronomie 8:889–896. Page`s L. Bengough AG. 1997. Modelling minirhizotron observations to test experimental procedures. Plant Soil 189:81–89. Page`s L, Pellerin S. 1994. Evaluation of parameters describing the root system architecture of filed grown maize plants (Zea mays L.). II. Density, length, and branching of first-order lateral roots. Plant Soil 164:169–176. Page`s L, Pellerin S. 1996. Study of differences between vertical root maps observed in a maize crop and simulated

Modeling Root System Architecture maps obtained using a three-dimensional model of the root system architecture. Plant Soil 182:329–337. Page`s L, Serra V. 1994. Growth and branching of the taproot of young oak trees. A dynamic study. J Exp Bot 45:1327–1334. Page`s L, Jordan MO, Picard D. 1989. A simulation model of the three-dimensional architecture of the maize root system. Plant Soil 119:147–154. Page`s L, Chadoeuf J, Kervella J. 1992. Mode´lisation stochastique de la croissance et du de´veloppement du syste`me racinaire de jeunes peˆchers. I. Estimation et validation du mode`le. Agronomie 12:447–458. Page`s L, Le Roux Y, Thaler P. 1995. Mode´lisation de l’architecture racinaire de l’he´ve´a. Plant Rech Dev 2:19–34. Passioura JB. 1996. Simulation models: science, snake oil, education or engineering? Agron J 88:690–694. Pellerin S. 1993. Rate of differentiation and emergence of nodal maize roots. Plant Soil 148:155–161. Pellerin S, Page`s L. 1994. Evaluation of parameters describing the root system architecture of field-grown maize plants (Zea mays L.). I. Elongation of seminal and nodal roots and extension of their branched zone. Plant Soil 164:155–167. Pellerin S, Page`s L. 1996. Evaluation in field conditions of a three-dimensional architectural model of the maize root system: comparison of simulated and observed horizontal root maps. Plant Soil 178:101–112. Pellerin S, Tabourel F. 1995. Length of the apical unbranched zone of maize axile roots: its relationship to root elongation rate. Environ Exp Bot 35:193–200. Phillips M. 1996. Geomview. Manual. Minneapolis, MN: Geometry Center. Picard D, Jordan MO, Trendel R. 1985. Rythme d’apparition des racines primaires du maı¨ s (Zea mays L.). I. Etude de´taille´e pour une varie´te´ en un lieu donne´. Agronomie 5:667–676. Porter JR, Klepper B, Belford RK. 1986. A model (WHTROOT) which synchronizes root growth and development with shoot development for winter wheat. Plant Soil 92:133–145. Prusinkiewicz P. 1998. Modeling of spatial structure and development of plants: a review. Sci Hort 74:113–149. Prusinkiewicz P, Lindenmayer A. 1990. The Algorithmic Beauty of Plants. New York; Springer-Verlag. Reynolds JF, Thornley JHM. 1982. A shoot:root partitioning model. Ann Bot 49:585–597. Reynolds KM, Gold HJ, Bruck RI, Benson DM, Campbell CL. 1986. Simulation of the spread of Phytophtora cinnamoni causing a root rot of Fraser fir in nursery beds. Phytopathology 76:1190–1201. Riedacker A. 1976. Rythmes de croissance et de rege´ne´ration des racines des ve´ge´taux ligneux. Ann Sci Forest 33:109–138.

381 Riedacker A, Dexheimer J, Takavol R, Alaoui H. 1982. Modifications expe´rimentales de la morphogene`se et des tropismes dans le syste`me racianire de jeuneˆs cheˆnes. Can J Bot 60:765–778. Ritchie JT, Godwin DC, Otter S. 1985. CERES-Wheat: A User Oriented Wheat Yield Model. Preliminary Documentation. AGRISTARS Publication No. YMU3-04442-JSC-18892. East Lansing, MI: Michigan State University. Robinson D. 1994. The responses of plants to non uniform supply of nutrients. New Phytol 127:635–674. Rumbaugh J, Blaha M, Premerlani W, Eddy F, Lorensen W. 1991. Object Oriented Modeling and Design. Englewood Cliffs, NJ: Prentice Hall. Schellenbaum L, Berta G, Ravolanirina F, Tisserand B, Gianinazzi S, Fitter AH. 1991. Influence of endomycorrhizal infection on root morphology in a micropropagated woody plant species (Vitis vinifera L.). Ann Bot 68:135–141. Silk WK. 1994. Quantitative descriptions of development. Annu Rev Plant Physiol Plant Mol Biol 35:479–518. Silk WK, Erickson RO. 1979. Kinematics of plant growth. J Theor Biol 76:581–601. Shibusawa S. 1994. Modelling the branching growth fractal pattern of the maize root system. Plant Soil 165:339– 347. Shinozaki K, Yoda K, Hozumi K, Kira T. 1964. A quantitative analysis of plant form—the pipe model theory. I. Basic analyses. Jpn J Ecol 14:97–105. Skiles JW, Hanson JD, Parton WJ. 1982. Simulation of above- and below- ground carbon and nitrogen dynamics of Bouteloua gracilis and Agrophyron smithii. In: Lavenroth WK, Skogerboe GV, Flug M, eds. Analysis of Ecological Systems: State of the Art in Ecological Modelling. Amsterdam; Elsevier, pp 467– 473. Somma F, Hopmans JW, Clausnitzer V. 1998. Transient three-dimensional modeling of soil water and solute transport with simultaneous root growth, root water and nutrient uptake. Plant Soil 202:281–293. Spek LY, Van Noordwijk M. 1994. Proximal root diameters as predictors of total root system size for fractal branching models. II. Numerical model. Plant Soil 164:119–128. Tardieu F. 1988. Analysis of the spatial variability of maize root density. II. Distances between roots. Plant Soil 107:267–272. Tardieu F, Pellerin S. 1990. Trajectory of the nodal roots of maize in fields with low mechanical constraints. Plant Soil 124:39–45. Tardieu F, Bruckler F, Lafolie F. 1992. Root clumping may affect the root water potential and the resistance to soil-root water transport. Plant Soil 140:291–301.

382 Tatsumi J, Yamauchi A, Kono Y. 1989. Fractal analysis of plant root systems. Ann Bot 64:499–503. Thaler P, Page`s L. 1996a. Periodicity in the development of the root system of young rubber trees. Relationship with shoot development. Plant Cell Environ 19:56–64. Thaler P, Page`s L. 1996b. Root apical diameter and root elongation rate of rubber seedlings (Hevea brasiliensis) show parallel responses to photoassimilate availability. Physiol Plant 97:365–371. Thaler P, Page`s L. 1998. Modelling the influence of assimilate availability on root growth and architecture. Plant Soil 201:307–320. Torrey JG. 1976. Root hormone and plant growth. Annu Rev Plant Physiol 27:435–459. Tran Van Canh. 1982. Lutte contre le Fomes. Nouvelle me´thode d’e´tude. Rev Gen Caout Plast 59:61–74. Tsegaye T, Mullins CE, Diggle AJ. 1995a. An experimental procedure for obtaining input parameters for the ‘‘ROOTMAP’’ root simulation program for peas (Pisum sativum L.). Plant Soil 172:1–16. Tsegaye T, Mullins CE, Diggle AJ. 1995b. Modelling pea (Pisum sativum) root growth in drying soil. A comparison between observations and model predictions. New Phytol 131:179–189. Van Noordwijk M, Brouwer G. 1991. Quantitative root length data in agriculture. In: McMichael BL, Persson H, eds. Plant Roots and their Environment. Amsterdam; Elsevier, pp 515–525. Van Noordwijk M, Spek LY, De Willigen P. 1994. Proximal root diameters as predictors of total root system size for fractal branching models. I. Theory. Plant Soil 164:107–118. Varney GT, McCully ME. 1991. The branch roots of Zea. II. Developmental loss of the apical meristem in fieldgrown roots. New Phytol 118:535–546.

Page`s Varney GT, Canny MJ, Wang XL, McCully ME. 1991. The branch roots of Zea. I. First-order branches, their number, sizes and division into classes. Ann Bot 67:357–364. Vercambre G, Page`s L. 1998. Architecture racinaire du peˆcher en conditions de verger. Utilisation d’un mode`le pour lier des observations statiques et simuler une dynamique de mise en place. In: Architecture et Mode´lisation en Arboriculture Fruitie`re: 11e`me Colloque sur les Recherches Fruitie`res, INRACTIFL, Montpellier, France, 05-06/03/98:286–292. Whisler FD, Acock B, Baker DN, Fye RE, Hodges HF, Lambert JR, Lemmon HE, McKinion JM, Reddy VR. 1986. Crop simulation models in agronomic systems. Adv Agron 40:141–208. Wightman F, Thimann KV. 1980. Hormonal factors controlling the initiation and development of lateral roots. I. Sources of primordia-inducing substances in the primary root of pea seedlings. Physiol Plant 49:13–20. Wilcox H. 1962. Growth studies of the root of incense cedar, Libocedrus decurrens. II. Morphological features of the root system and growth behavior. Am J Bot 49:237– 245. Wilcox H. 1968. Morphological studies of the root of red pine, Pinus resinosa. I. Growth characteristics and patterns of branching. Am J Bot 55:247–254. Wullscheleger SD, Lynch JP, Berntson GM. 1994. Modeling the belowground response of plants and soil biota to edaphic and climatic change—what can we expect to gain? Plant Soil 165:111–121. Zhang H, Forde BG. 1998. An Arabidopsis MADS Box gene that controls nutrient changes in root architecture. Science 279:407–409.

23 Auxins in the Biology of Roots Thomas Gaspar, Odile Faivre-Rampant, Claire Kevers, and Jacques Dommes University of Lie`ge, Lie`ge, Belgium

Jean-Franc¸ois Hausman CRP-Gabriel Lippman, Luxembourg

I.

INTRODUCTION

cycles (cf. John et al., 1993; Ormrod and Francis, 1993). Moreover, the level of any one hormone affects the levels of the others by affecting their biosynthesis, degradation, conjugation, or transport (Itai and Birnbaum, 1991). This is called the hormonal crosstalkings (Rodrigues-Pousada et al., 1999). It seems that determining the level and effect of one hormone may yield information with a limited value. No single hormone has an overriding role. Hormone metabolism and action cannot be dissociated from the primary metabolic pathways with reciprocal influences (Gaspar et al., 2000a,b). This means that the effects of an externally applied hormone, or of an analog, cannot be interpreted simply through an increase of its endogenous bulk (Pilet, 1996) but that changes in metabolism and the role as an exterior signaling molecule have to be considered. Auxins and cytokinins were originally thought to produce growth responses at distances from their sites of synthesis, and thus to fit the definition of transported chemical messengers. It is now clear that none of the recognized five main classes of phytohormones (auxins, cytokinins, gibberellins, abscisic acid, ethylene) fulfill the requirements of a hormone in the mammalian sense, i.e., chemical messengers at low concentration, involving a localized site of synthesis, transport to a target tissue and control of a precise physiological response in a target tissue via the concentration. The synthesis of all plant hormones, as a rule, occurs or can occur in any type of living cells even if

The concepts in plant hormonology have changed dramatically with the progressive discovery of new phytohormones. It can now hardly be claimed that a single hormone is responsible for one growth or development process. Growth and development processes have been dissected into successive interdependent physiological phases with different requirements. Moreover, in many cases, it was shown that the control of these events is due to the simultaneous interaction of different plant hormones, acting synergistically or antagonistically, rather than to the effect of a single hormone. Distinct cell types respond differentially to various signals. It is clearer now that the hormonal controls act in a developmental and in a tissue-dependent manner. Thus, the former claimed specificity of one hormone may simply be the result of its preponderance in a balance with another one; for example, the ratio of auxins to cytokinins is determining growth and development. In neoplastic tissues, the sensitivity to this couple of hormones is shifted to the tandem polyamines/ethylene (Kevers et al., 1999c; Gaspar et al., 2000a). The possibility that different hormonal receptors control growth and development (De Klerk et al., 1997) or that different hormones compete for a common receptor or operate in separate signalling pathways (Timpte et al., 1995) is being investigated. Apparently, different hormones play in very tightened sequential events in the control of cell division

383

384

Gaspar et al.

certain tissues are privileged sites of synthesis and export for some hormones (e.g., aerial meristemic tissues for auxins or growing root parts for cytokinins). Thus, phytohormones may be transported and participate in some precise physiological processes at a distance. However, it is far from being the general case; they also act in the tissue or even within the cell in which they are synthesized. Furthermore, experimental results strongly argue that phytohormone control is not only by concentration but also by changes in sensitivity of the cells to the compounds (Trewavas and Cleland, 1983). Thus, the responses evoked by plant hormones are rarely proportional to their concentration. Furthermore, the countercurrents of hormones between aerial and underground plant parts create varying gradients and continuously change organ cross-talks. II.

THE DIVERSITY OF AUXINS AND THEIR METABOLISM IN ROOTS

A.

Naturally Occurring Auxins

No root-specific auxin has been discovered, and the most common natural auxin is indole-3-acetic acid (IAA). However, depending on the species, age of the plant, season, and the conditions under which it has been growing, other natural auxins have been identified, such as 4-chloroindole-3-acetic acid, indole-3acrylic acid, and indole-3-butyric acid (Marumo, 1986; Gaspar et al., 1996). In addition to these indolic auxins, various phenolic acids (such as phenylacetic acid) that appear also in roots have low auxin activity. However, a physiological role for such nonindolic compounds in auxin regulation has not been established (Bandurski et al., 1995). Auxin precursors may also have auxinlike properties and can sometimes replace IAA. Sometimes they may be more effective than auxin itself in stimulating growth or inducing organised development. Auxins are found in plants both as the free acid (which is thought to be the primary ‘‘active form’’) and as conjugated forms. Conjugation appears to be a mechanism for storing auxin in cells and stabilizing the level of free auxin by metabolizing its excess. Auxin in conjugated molecules is protected from oxidative breakdown and may provide a readily accessible and easily source of free IAA without de novo synthesis. One type of conjugated form is linked through carbon–oxygen–carbon bridges and these compounds are referred to generically as ‘‘esters,’’ although some 1-O sugar conjugates such as indole-3-acetyl-1-O--D-

glucose (1-O-IAGluc) are actually linked by acyl alkyl acetal bonds. True esters include compounds such as 6O-1AGluc and indole-3-acetyl-myo-inositol (IAInos). The other type of conjugates are linked through carbon–nitrogen–carbon amide bonds, as in the IAA– amino acid and peptide conjugates. All native auxins are found in both free forms and conjugated forms. However, in most tissues the conjugated forms predominate. Various conjugates of IAA, both ester and amide, have been used as ‘‘slow-release’’ forms of IAA for tissue cultures and for rooting of cuttings. IAA conjugates, each differing in ease of hydrolysis by the plant’s enzymes and having conjugating moieties of varying degrees of lipophilicity, could be used to ‘‘target’’ the IAA to a particular tissue or particular cell organelle with delivery of the hormone at the required rate. The conjugating moiety might thus be used as a ‘‘zip code,’’ to bring the IAA to the desired location, with simultaneous protection against peroxidative attack (Bandurski et al., 1995).

B.

Auxinlike Growth Regulators

Several indole derivatives, both naturally occurring and synthetic, are active in plant cultures. For instance, indole-3-acetaldehyde, indole-3-acetamide, indole-3acetonitrile, indole-3-lactic acid, indole-3-propionic acid, indole-3-pyruvic acid, indole-3-glycolic acid, 5OH-tryptamine (serotonin), and tryptophan were shown to support root, callus, or shoot formation and growth (Gaspar and Hofinger, 1969; Maeda and Thorpe, 1979; Gatineau et al., 1997). Several other substances can mimic auxin activity, in roots notably: this is the case of acetylcholine (Penel et al., 1976), which has been found in many plant roots (Tretyn et al., 1992). The most commonly used synthetic auxins are 2,4dichlorophenoxyacetic (2,4-D) and 1-naphthaleneacetic acid (NAA), but others like dicamba (3,6-dichloroo-anisic acid) and pichloram (4-amino-3,5,6-trichloropyridine-2-carboxylic acid) have applications in tissue cultures or as selective herbicides. BSAA [benzo(b)selenienyl-3 acetic acid] and its chloro and methoxy forms are new synthetic auxins with powerful effects on root growth and adventitious root formation (Hofinger et al., 1980; Lamproye et al., 1990; Gaspar, 1995; Kevers et al., 1997). Such a powerful auxin as BSAA is able to induce and sustain the growth of hairy roots in the absence of Agrobacterium rhizogenes (Kevers et al., 1999a).

Auxins

Auxins affect root formation and/or growth either directly or indirectly by affecting the level of endogenous auxin. The greater efficiency of IBA versus IAA in root formation is probably due to its progressive conversion (-oxidation) into IAA, thus working as slowrelease source for IAA. This has led Van der Krieken et al. (1997) to design synthetic slow-release auxin sources (IAA bound to bovine serum albumine, indolehexanoic acid, IBA-anhydride, IBA-aminoacids, IBA-polyamine-IBA, or IAA-polyamine-IAA). Most of these slow-release compounds are stable and can be autoclaved. They are more effective than the standard auxins in the induction of adventitious roots. C.

Auxin Protectors and Elicitors of Auxin Action

Monophenolic substances enhance the so-called IAAoxidase system but polyphenolic substances inhibit the same system (Gaspar, 1965; Pilet and Gaspar, 1968). Substances such as the soil humic acids can influence growth of the roots indirectly by modifying their auxin level. Other substances, e.g., breakdown products of the cell membrane (nonanoic acid and jasmonate), of the cell wall (lignosulfonates), or of fungal origin (Pythium extract), have the capacity to increase the sensitivity to auxins and thus the rooting of ‘‘recalcitrant’’ cutting types (Van der Krieken et al., 1997; Kevers et al., 1999b). D.

Auxin Metabolism and Distribution in the Roots

Bandurski et al. (1995) have addressed the different aspects of auxin metabolism. Apparently the pathways of auxin anabolism and catabolism of roots are the same as in other plant organs. However, it can be expected that the rates of turnover, the pool sizes, and their inputs and outputs are different, namely through the contribution of bacteria and mycorrhizae. The auxin–oxidase system, through its adaptation capacity, plays an important role in the regulation of the auxin levels. This seems particularly true along the roots where an inverse relationship between the auxin activity and the auxin level has been shown (Pilet and Gaspar, 1968). Some peroxidases, at least, are involved in auxin catabolism and hence in growth (Gaspar et al., 1982, 1991). It is therefore not astonishing that peroxidase activity and secretion also vary along roots (Bouchet et al., 1980). Plants synthesize, inactivate, and catabolize IAA by multiple pathways, and multiple genes can encode a

385

particular enzyme within each of the pathways. A number of these genes are now cloned, which greatly facilitates the future deciphering of IAA metabolism (Normanly and Bartel, 1999).

III.

CONTRIBUTION OF RHIZOBACTERIA AND MYCORRHIZAE TO ROOT AUXIN METABOLISM AND GROWTH

Saprophytic free-living bacteria and fungi, as well as symbiotic endophytic or nonendophytic rhizobacteria and mycorrhizae, exert a beneficial effect on plant growth. They increase the tolerance to abiotic constraints and to phytopathogens. This is generally an indirect result of an improved development of the root system (Mosse, 1957; Nelsen and Safir, 1982; Gianinazzi-Pearson, 1996). Root-colonizing bacteria and fungi produce plant growth substances, including auxins, gibberellins, and cytokinins (Brown, 1974). IBA concentrations in mycorrhizal and in nonmycorrhizal maize roots differed (Ludwig-Muller et al., 1997). Young roots of maize colonized with Glomus intraradices had a higher content of IBA while the total IAA content of such roots was lower. In older roots, total IBA content was slightly lower in arbuscular mycorrhizae-infected roots than in controls. However, the activity of IBA synthetase was threefold higher in infected roots than in control ones (Ludwig-Muller et al., 1997). Since it is unknown whether fungi possess similar enzymes as higher plants, it was assumed that IBA is synthesized by the host plants, though the observed increase was probably induced by the arbuscular mycorrhizal fungus (Kaldorf and Ludwig-Muller, 2000). However, in ectomycorrhizal symbiosis, IAA is produced by the fungal partner, resulting in an increased initiation of lateral roots (Karabaghli-Degron et al., 1998). Such production may also contribute to the enhancement of plant growth (Frankenberger and Arshad, 1995). Esch et al. (1994) showed that hyphae of Glomulus intraradices produce abscisic acid (ABA). This plant growth regulator increases the IBA synthesis (Ludwig-Muller et al., 1995). Thus it was hypothesized that auxin production may be increased via fungal ABA. On the other hand, the higher endogenous content of ABA could lead to an increase in the number of receptor sites, as indicated by the induction of auxinbinding protein 1 by IBA (Kaldorf and LudwigMuller, 2000). Another interesting example is the free-living, plant growth–promoting rhizobacterium Pseudomonas.

386

Gaspar et al.

These bacteria enhance plant growth by one or several mechanisms including production of plant growth regulators. Following the binding of P. putida to the root, the bacterial deaminase degrades ACC that is released from the plant’s root. This would lower the level of rhizogenic ACC as well as that of ethylene, and promote plant growth (Glick et al., 1994). Moreover, it was shown that some strains of P. putida overproduce IAA that is able to enter the plant’s root cells. This fungal IAA is then able to promote cell elongation and proliferation as well as to stop ethylene synthesis, by a combined decrease of ethylene and increase of endogenous auxin content enhance rooting (Xie et al., 1995; Glick et al., 1994). Nitrogen-fixing bacteria inducing the formation of root nodules produce auxins, but the involvement of the excreted auxins in nodulation remains debated. Nodulation may result either from production of phytohormones by the microsymbiont or the promotion of hormone synthesis in the cells of the host roots (Crozier et al., 1988).

IV.

AUXIN-RELATED GENES IN ROOTS

Questions regarding which genes are involved in the biosynthesis of auxin, how the level of auxin is regulated in the plant, and what is the basis for the differential sensitivity of different tissues to auxin remain to be answered. The isolation of mutants with altered responses to auxin and the cloning of the corresponding genes are a valuable strategy with which to address such questions. The following genes are involved in IAA metabolism. 1. PIN2 (protease inhibitor II). The protein PIN2 was found to be similar to members of the major facilitator family of transport proteins (Mu¨ller et al., 1998). It was localized in membranes of cortical and epidermal cells of the meristemic and of the elongation zones. The polar localization of such proteins and its function was specific in roots. The loss of PIN2 function impairs basipetal auxin transport. The authors have suggested that PIN2 plays an important role in control of gravitropism regulating the redistribution of auxin from the stele toward the elongation zone of roots. PIN2 was shown to encode a root-specific member of a novel membrane protein family, supporting the idea about its role in the transport of auxin (Chen et al., 1998; Utsuno et al., 1998; Luschnig et al., 1998). 2. AXR1 and AXR2 (altered-auxin response 1 and 2). AXR1 encodes a protein related to ubiquitin-activating enzyme E1, which catalyzes the first step in the

biosynthesis of ubiquitin–protein conjugates. The existence of relationships between AXR1 and E1 suggests that through the action of AXR1, auxin may stimulate ubiquitin-mediated degradation of a putative repressor of auxin-regulated genes (Abel et al., 1994). AXR1 gene is required very early in an auxin response pathway in mature roots. Sabatini et al. (1999) have shown that a reduced AXR1 activity is correlated with an incomplete cell division program in the distal root cap. AXR2 protein is likely to play a role very early in signal transduction (Wilson et al., 1990; Timpte et al., 1992, 1994). Karlsson et al. (1996) have shown that the AXR2 gene product has a different function in roots than in hypocotyls but may be more important for IAA responses in the hypocotyls than in roots. Since responses to IAA differ in roots and hypocotyls of wild-type plants, it seems likely that they should be regulated differently and the AXR2 gene product may be subject to or involved in this differential regulation. 3. AUX1 (auxin-resistant 1). The AUX1 polypeptide exhibits sequence similarities to a family of plant and fungal amino acid permeases (proteins facilitating transport of amino acids), suggesting that AUX1 mediates the transport of an amino acid–like signaling molecule. IAA, which is structurally similar to the amino acid tryptophan, is thus a likely substrate (Bennett et al., 1996). It was proposed that AUX1 may mediate proton-driven IAA uptake, as plant amino acid permeases mechanically function as proton-driven symporters (Bush, 1993). Bennett et al. (1998) showed that AUX1 is expressed in root apical tissues that control the root gravitropic response and also in root epidermal cells. According to Marchant et al. (1999), the AUX1 gene is expressed predominantly in the cap of lateral roots and in the epidermis of the root meristem of Arabidopsis thaliana. 4. RSI-1 (root system inducible-1). This gene is inducible by auxin and is expressed very early in lateral root development (Taylor and Scheuring, 1994). A specific role for RS1-1 is still unknown. 5. TIR1 (auxin transport inhibitor resistant 1). Expression of TIR1 is strongest in the primary root— and in lateral root—meristems, which is consistent with a key role for the protein in root development (Del Pozo and Estelle, 1999). TIR1 is required for pericycle cells to respond to the inductive signal, presumably auxin. 6. RML (root meristemless). The RM1 genes are involved in signalling cell proliferation at the root tip (Cheng et al., 1995). Their products are involved specifically in activating the cell division cycle in the root apical cells. They are required for cell proliferation

Auxins

during postembryonic root growth. RML genes regulate cell proliferation not only of primary roots but also of laterals and adventitious roots. Functional differentiation along the root axis may be accomplished by the regulation of the cell division cycle via the RNL gene products. Several other mutants overproducing IAA have been isolated, but the corresponding genes have not been cloned yet. Boerjan et al. (1995) have described the isolation of Arabidopsis mutants overproducing free and conjugated IAA designated superroot (sur). Seven allelic Arabidopsis mutants sur1-1 to sur1-7 developed excess adventitious and lateral roots. The authors hypothesized that the SUR1 gene encodes a regulator of auxin biosynthesis in Arabidopsis. Another Arabidopsis mutant exhibiting a phenotype similar to superroot mutant has been selected by King et al. (1995) and called rooty (rty). The authors suggested that the regulation of auxin levels is done by the wild-type RTY gene product. Another promising candidate for analysis of auxinregulated growth and development is the tomato mutant diageotropica (dgt). dgt plants failed to form lateral roots and do not respond to gravity (Zobel, 1974). According to Kressin Muday et al. (1995), the defect in dgt is in the ability to respond to auxin, rather than in auxin uptake or altered endogenous auxin concentrations. The analysis of Arabidopsis mutants has been instrumental in linking root hair formation to the action of plant hormones, particularly ethylene and auxin. Mutations affecting the CONSTITUTIVE TRIPLE RESPONSE1 (CTR1) locus, which encodes a Raf-like protein kinase negatively regulate the ethylene signal transduction pathway (Kieber et al., 1993), cause root hairs to form on epidermal cells that normally are hairless (Dolan et al., 1994). The hairless root phenotype of the dwarf (dwf; auxin resistant) and auxin resistant2 (axr2; auxin, ethylene, and abscisic acid resistant) mutants implicate that auxin is another possible regulator of root hair formation (Mizra et al., 1984; Wilson et al., 1990). In addition, the hairless phenotype of the root hair defective6 (rhd6) mutant can be suppressed by the inclusion of 1-aminocyclopropane-1-carboxylic acid (ACC; an ethylene precursor) or IAA in the growth media. This is another implication of ethylene and auxin action in root hair initiation (Fig. 1; Masucci and Schiefelbein, 1994). Given the involvement of hormones in epidermal cell differentiation, one attractive possibility is that ethylene and/or auxin may act as a diffusible signal responsible for epidermal cell–type patterning. According to

387

Figure 1 Tentative model for the control of cell-type differentiation in Arabidopsis root epidermis. T-bars indicate negative regulation, question marks indicate unclear relationships. (From Schiefelbein et al., 1997.)

this notion, TTG and GL2 may represent downstream transcription factor genes that are regulated by the hormones in a cell position–dependent manner. Masucci and Schiefelbein (1996) have shown that the ethylene/auxin pathway does not regulate the TTG/ GL2 pathway. It acts upstream of or independently from the ethylene/auxin pathway to define the pattern of cell types in the root epidermis (Fig. 1). All newly formed epidermal cells in the Arabidopsis root may initially have the capacity to differentiate into root hair cells, and the action of the TTG/GL2 pathway may be required to induce the hairless cell fate (i.e., a root hair cell may represent the default fate). Other root auxin-related genes are those introduced by plasmids, notably the ones of Agrobacterium rhizogenes. The best studied plasmid is the agropine pRiA4 plasmid that carries and can mediate the transfer of two T-DNAs, denoted TL and TR, to the plant cell (White et al., 1985). Hairy roots induced by agropine strains frequently contain only the TL-DNA (Jouanin et al., 1987). White et al. (1985) showed that insertions in only four of the 18 potential loci on the TL-DNA noticeably affected the morphology of the hairy roots that were produced. These loci were denoted root locus A-D (rolA-D). The gene rolB (ORF 11) contains an ORF of 777 bp encoding a 259-amino acid protein with a molecular mass of 30 kDa. It has been observed that the biological effects of the rolB expression, like root initiation, are reminiscent of auxin-mediated effects (Schmu¨lling et al., 1988; Estruch et al., 1991). Two main theories

388

Gaspar et al.

concerning the mechanism of RolB action were proposed. The first theory hypothesized that RolB acts by increasing the pool of free, active auxin in transgenic plants (Estruch et al., 1991). This was shown by using a RolB-mediated hydrolysis of inactive IAA conjugates. The second theory for RolB action suggests that RolB is regulating the cell sensitivity to IAA—i.e., making normally unresponsive cells responsive to IAA (Maurel et al., 1994). Further support for a role of RolB in an IAA signal perception/transduction pathway is the recent finding that RolB has a tyrosine phosphatase activity (Filippini et al., 1996). The rolB gene showed a tissue-specific expression pattern being mainly confined to root meristems. In hybrid aspen, the rolB promoter exhibited highly specific expression in groups of pericycle cells prior to and during lateral root initiation (Nilsson et al., 1997) and strong expression during the initiation and growth of tobacco adventitious roots (Altamura et al., 1991). This exhibits a high correlation between the expression of this gene and the initiation of root growth. V. AUXIN CONTROL OF ROOT GROWTH A.

Auxin Dose–Growth Response Curve

Root growth is stimulated by ‘‘low concentrations’’ of auxins, whatever their nature is, and is inhibited by ‘‘high concentrations.’’ This classical hormonal dose response curve differs from what is known for buds and stems (cf. Fig. 2). This has led to the consideration

Figure 2 Schematic representation of the growth responses of roots, buds, and stems to a range of auxin concentrations, each organ having a promotive and an inhibitory range. (From Thimann, 1969.)

of roots as more sensitive to auxins than the other organs, but still left some questions unanswered: 1. Do root and stem growth responses really correspond to different levels of endogenous auxins? The natural levels of auxin in these two organs have not really been investigated; when treated by the same exogenous auxin concentrations, do these organs absorb auxin in the same proportion and at the same rate? Does their adaptative auxin-oxidase system react in the same manner? 2. Are the different responses caused by changes in the number of receptors, changes in receptor‘s affinity, or changes in the subsequent chain of events, including possible changes in the level of other endogenous hormones that affect the response? For example, growth and ethylene production by lentil root tips, treated with varying concentrations of indolylacetic and indolylacrylic acids were measured. Stimulation of growth by low concentrations of auxin was accompanied by a reduced ethylene production. Growth inhibition by high levels of auxin induced increased ethylene evolution (Fig. 3A). If the ethylene synthesis is prevented by ethylene synthesis inhibitors, if ethylene is removed by hypobaric conditions, or if the action of ethylene is opposed by silver ions, then auxin is no longer inhibitory. This suggests that root growth stimulation by auxins may be mediated by ethylene (Davies, 1995). But simultaneous application of auxin and rhizobitoxine analog on the lentil root reduced ethylene evolution even though growth was inhibited (Fig. 3B). On the contrary, this might prove that auxin-controlled root growth is independent of ethylene production (Dubucq et al., 1978). In any case, this means that growth of roots as well as growth of aerial organs or calluses results in sequential events where auxins and ethylene play some roles (Kevers et al., 1984; Gaspar et al., 2000a). Thus, root growth inhibition can no longer be explained simply by a decrease in auxin content. Kinetin, light, and dinitrophenol, three factors inhibiting root growth, cause a decrease in auxin content together with a rise in the activity of the so-called inhibitor (Gaspar, 1973). Methyleneoxindole, the main degradation product of IAA, chromatographically runs next to ABA, and was shown to inhibit root growth. Thus, physical or chemical treatments which enhance auxin biodestruction may cause root growth inhibition through the regulation of an auxin ‘‘inhibitor’’ balance. Auxin oxidase must therefore be considered not only as a regulatory destructive system (Gaspar, 1965; Pilet and Gaspar, 1968) but also as a system that generates inhibitory substances. In this context, the role of peroxidases pre-

Auxins

389

B.

Figure 3 (A) Effect of varying concentrations of indolylacrylic acid (IAcrA) on growth (*—*) and ethylene production ~  ~) by lentil root tips. Results expressed in percent of the control. (B) Compared effect of IAA and rhizobitoxine (AAR), alone or combined, on root growth and ethylene production. Open column, growth; hatched column, ethylene. (From Dubucq et al., 1978.)

sent in the roots and not in the aerial parts (Gaspar et al., 1974) should be further considered. Moreover, it is still unclear whether the concentrations measured in extracts of whole plant organs or tissues reflect active hormone pools. There is an urgent need for a reliable method for differentiation between active hormone pools and pools that are inactive either because of compartmentation or because of conjugation.

Auxin–Cytokinin Interactions

Skoog and Miller (1957) discovered that both auxins and cytokinins can act synergistically in the induction of cell division and growth in plant tissue cultures but can also antagonistically control lateral bud and root outgrowth (Fig. 4). Molecular work on the links between hormone action and cell division led to the cloning of several genes that are responsive to auxins, to cytokinins, or to both hormones. However, the function of the majority of these genes either is not yet known or has no obvious connection with cell cycle control. A more direct link between plant hormones and cell cycle control is now being uncovered by analyzing the expression patterns and activity of proteins that are homologous to those that control the cell cycle in yeast (Coenen and Lomax, 1997). Lateral root primordia are generally initiated through the commencement of cell divisions in the pericycle opposite the xylem arches of the root vascular system. The antagonistic relationships of auxins and cytokinins in this process is similar to their interaction in regulation of the expression of a cdc2-like protein (Fig. 5; John et al., 1993). Although auxins increase immunologically detectable cdc2-like protein in extracts, cytokinins reduce the levels of the cdc2kinase. Recent papers on apical dominance (Bangerth, 1994; Li et al., 1995) have demonstrated that decapitation, and thus removal of the endogenous auxin source, leads to a large (up to 40-fold) increase in the cytokinin content of xylem exudate. Such an increase can be eliminated by application of the synthetic auxin -NAA to the apex of the decapitated plants. This effect of auxin on cytokinin concentrations in the xylem suggests that auxin can influence apical dominance via inhibition of cytokinin synthesis or export from the roots. However, bud outgrowth can also be inhibited by apically applied auxin in isolated stem segments, which indicates that this is not the only mechanism available. Coenen and Lomax (1997) have proposed the scheme of Fig. 6 with the potential points of control of active cytokinin pools by auxin, namely through a control of cytokinin oxidase by auxin. The auxin and cytokinin interactions in root growth were interpreted through a cytokinin control of isoperoxidases with the latter supposed to function as auxin oxidases (Darimont et al., 1971). However, increases in free IAA are observed both in roots of cytokinin-overproducing lines of Nicotiana glutinosa transformed with the ipt gene (Binns et al., 1987), after exogenous

390

Figure 4

Gaspar et al.

Control of different organogenic programs by the balance between auxins and cytokinins. (From George, 1993.)

Figure 5 Speculative model for control of the cell cycle through auxins and cytokinins. Both cytokinin and auxin have been reported to regulate the expression of cdc2 kinases and the cyclins that are required for their activation. The interaction between auxin and cytokinin in regulating the cell cycle is synergistic in undifferentiated cells, such as callus or protoplasts, with both auxin and cytokinin stimulating expression of the cdc2 kinase and cytokinin treatment increasing expression of a cyclin. In lateral root primordia the interaction is antagonistic: auxins stimulate and cytokinins reduce levels of the cdc2 kinase, and the expression of at least one cyclin is increased by auxin. (From Coenen and Lomax, 1997.)

Auxins

391

Figure 6 Potential points of control of active cytokinin pools by auxin. Open arrows indicate the steps where auxin or auxin conjugates are thought to regulate enzyme activity, resulting in changes in conjugate or metabolite levels. (From Coenen and Lomax, 1997.)

application of cytokinins to maize (Bourquin and Pilet, 1990) or in pea roots (Bertell and Eliasson, 1992). The different possibilities of a mutual control of auxin and cytokinin abundance can be explained even before the enzymes involved in their metabolism have been isolated and cloned and while a debate persists on the anabolic and catabolic pathways of both hormone types (cf. Schmu¨lling et al., 1997).

VI.

AUXIN TRANSPORT WITHIN AND FROM THE ROOTS

A phytohormone does not act like an animal hormone—i.e., as a signal that carries information from source cells to specific target cells or tissues. The observation that auxin replaces all the correlative effects of a shoot apex led to the conclusion that the activity profile of auxin may be better interpreted as the integrative signal by which the growing shoot tissues influence the development of the rest of the plant, including adventitious and lateral root formation, growth, and vascular patterning (Davies, 1995; Aloni, 1995; Palme and Ga¨lweiler, 1999; Jouve et al., 1999). Auxin acts in plants over long distances and throughout the plant’s life. Auxin triggers all developmental steps, beginning with the differentiation of the embryonic axis, and later with the transport to growing shoot and root (Bandurski et al., 1995). Auxin is transported acropetally from the shoot through the stele up to the meristemic region of the root tip. There it is redistributed into the various root tissues

and is transported basipetally toward the elongation and differentiation zone (Palme and Ga¨lweiler, 1999). Specific growth processes can be manipulated using naturally occurring or synthetic auxin transport inhibitors (Lomax et al., 1995). (For additional information see Chapters 31 by Poovaiah et al., 29 by Porterfield, and 30 by Pilet, in this volume). The idea that transport was an essential part of the role of plant hormones originated from experiments on the control of tropisms. Such growth movements, including the root gravitropic response, are based on an altered lateral auxin transport within plant tissues. Current models indicate that the gravity signal is perceived at the root cap and transmitted via the meristem to the zone of elongation. There, at the lower flank of the tissue (oriented toward the gravity stimulus), an increase in the auxin concentration can be found, leading to an overall increase of growth rate of the upper side and a decrease in the lower side (see also Chapter 30 by Pilet in this volume). The basipetal directionality of auxin transport from shoot apices and young leaves to roots is thought to result from the polar distribution of specialised carrier molecules in the plasma membranes. The discovery of mutants affecting polar transport or root gravitropism allows a molecular approach of the auxin efflux and influx carriers (Palme and Ga¨lweiler, 1999). This can answer the question whether such auxin efflux carriers represent the elusive auxin receptors. Furthermore, the new possibilities to modulate auxin distribution and response using genetic tools will allow direct tests of the role of auxin in patterning through the threshold-

392

Gaspar et al.

dependent activation of secondary responses (Sabatini et al., 1999; Tsukegi and Fedoroff, 1999). It must finally be recalled that auxin transport and metabolism are interdependent. This is illustrated by the relationship shown in Fig. 7 between auxin transport and degradation, in stems and roots. VII.

ROOT AUXINS AND THE GROWTH AND DEVELOPMENT OF AERIAL PLANT PARTS

Evidently, the development of seminal and adventitious roots sustains and enhances growth and vigour of the aerial plant parts. The importance of root-toshoot communication has been specifically emphasized in studies of root responses to environmental stresses (Davies and Jeffcoat, 1990). The contribution of root auxins, in addition to water and nutrients provision, has not been much addressed in the literature, probably because the problem is complex. Each hormone interacts with the others, and the development of the root system is also dependent on the actively growing aerial parts. Textbooks still teach that the ratio of shoot-derived auxin to root-derived cytokinin controls apical dominance and branching. Reexamination of this problem, using mutants, reveals that hormonelike signals other than auxin and cytokinin are also involved (Beveridge et al., 1997; see also Chapter 26 by Hose et al. in this volume).

A.

Contribution of Root Auxins in Stem Vascular Differentiation

Roots do not induce vascular differentiation nor must they be present in order to obtain vascular tissues in stems. However, the roots have two major functions in vascular differentiation, namely: (1) roots orient the pattern of vascular differentiation towards their tip by acting as a sink for the flow of auxin derived from young leaves; and (2) root apices are sources of inductive stimuli that promote vascular development (Aloni, 1995). The major developmental signals of roots are cytokinins and ABA. It should be emphasized however, that cytokinin alone, or root apices in the absence of an auxin source, do not induce vascular differentiation in stem tissues. B.

Auxin-Induced Adventitious Root Formation and Wood Formation

The lignin content of walnut shoots during their in vitro multiplication did not practically change but started to increase as soon as they were transferred to a rooting medium supplied with auxin (Fig. 8) (Kevers et al., 2000). Exogenous auxin provoked a temporary elevation of the endogenous free IAA level which allowed the completion of the rooting inductive phase before any visible histological event. This means that either exogenous or endogenous

Figure 7 IAA transport and IAA degradation compared in the stem and in the root. (From Pilet and Gaspar, 1968.)

Auxins

393

Figure 8 Variation of the lignin content of a whole population (*) of micropropagated walnut shoots, compared with the 20% shoots (*) which will not form roots, in the course of the multiplication (M) phase followed by the successive rooting inductive (Id), initiative (In) and expressive (Ex). (From Kevers et al., 2000.)

IAA, besides its rooting-inducing role, serves as a signal for an increased lignification. Indeed, auxin was shown to be a determining factor in cell wall lignification and differentiation (Bolwell, 1997). Sustained lignification in the shoot depends on the initiation and expression of rooting, since lignification was positively correlated with root emergence (Fig. 8). Continued increase of stem lignification with accentuated xylem proliferation and diameter increase were further shown to depend on the development of the root system (Fig. 9). This means that roots not only bring about essential substances for wood formation but that they also serve as coordinating organs. The cytokinins exported from growing roots are also known to be involved in cell wall growth and differentiation (Montague, 2000; see also Chapter 25 by Emery and Atkins in this volume), but this is probably part of a more complex interacting network between roots and aerial organs.

VIII.

AUXIN INVOLVEMENT IN ADVENTITIOUS ROOT FORMATION

It happened that the first discovered and identified phytohormone, IAA, was shown to promote adventitious rooting. Later identified natural auxins and synthetic compounds of this category had the same effects (Jackson, 1986; Altman and Waisel, 1997). Rooting property of auxins appeared to be specific to this

Figure 9 Evolution of peroxidase activity (A), and lignin level (B), and stem diameter (C) of micropropagated walnut shoots growing with a different number of roots. (From Kevers et al., 2000.)

class of growth regulators since no clear-cut effect was obtained by exogenous application of other phytohormones. Some hormones, such as cytokinins and gibberellins, were even classified as rooting inhibitors (Jackson, 1986; Davis et al., 1988; Davis and Haissig, 1994). This matter needs further investigation because the necessity of cytokinins and gibberellins for rooting, under certain circumstances, was also reported (Letham, 1978; Gaspar et al., 1977). Moreover, the application of exogenous auxins resulted in a series of wrong concepts: that auxin is the major triggering agent in rooting, that the application of exogenous auxin is needed to augment the endogenous bulk of

394

auxin, that rooting necessitates the maintenance of a ‘‘high’’ level of endogenous auxin for a certain time, etc. Because there are inductive/adaptative enzymes that regulate the exogenously fed hormones and because application of a hormone may induce modifications in the metabolism of other ones, no simplistic conclusions can be drawn. Another associated error was to consider rooting as a single developmental process. One of the main achievements in the studies of adventitious root formation has been the recognition of successive interdependent physiological phases (Mitsuhashi-Kato et al., 1978; Jarvis et al., 1983; Moncousin et al., 1988; Blakesley, 1994). These were generally called induction, initiation, and expression (Fig. 10) (Gaspar et al., 1992, 1994). Although some other terminologies can be used (De Klerk et al., 1995), a consensus tacitly emerged to define the rooting inductive phase as the time necessary for the biochemical events to precede the initiation of cell divisions and which lead to the formation of root primordia (Jarvis, 1986; Moncousin, 1991). Another definition, and a practical way to estimate the duration of the inductive phase, is the minimum time required for the presence of the external signal for rooting to proceed from competent cells (Hand, 1994). Competence itself is defined as a cell reactivity state allowing response to a stimulus that finally leads to a specific developmental pathway. Induced cuttings which do no longer require the rooting signals are said to be determined, even if other environmental factors are required for the completion of the successive developmental phases (Mohnen, 1994). The inductive rooting period sometimes appears to be very short, being achieved in