Plant Breeding Reviews (Volume 27)

PLANT BREEDING REVIEWS Volume 27

Plant Breeding Reviews is sponsored by: American Society for Horticultural Science Crop Science Society of America Society of American Foresters National Council of Commercial Plant Breeders International Society for Horticultural Science

Editorial Board, Volume 27 M. Gilbert I. L. Goldman C. H. Michler

PLANT BREEDING REVIEWS Volume 27

edited by

Jules Janick Purdue University

John Wiley & Sons, Inc.

This book is printed on acid-free paper. Copyright © 2006 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, e-mail: [email protected]. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: ISBN-13: 978-0-471-73213-6 ISBN-10: 0-471-73213-3 ISSN: 0730-2207 Printed in the United States of America 10

9

8

7

6

5

4

3

2

1

Contents

Contributors 1. Dedication: Fredrick A. Bliss Teacher, Researcher, and Director of Plant Breeding

viii 1

Thomas C. Osborn

2. Sugarcane Improvement through Breeding and Biotechnology

15

Ray Ming, Paul H. Moore, Kuo-Kao Wu, Angélique D’Hont, Jean C. Glaszmann, Thomas L. Tew, T. Erik Mirkov, Jorge da Silva, John Jifon, Mamta Rai, Raymond J. Schnell, Stevens M. Brumbley, Prakash Lakshmanan, Jack C. Comstock, and Andrew H. Paterson I. Introduction II. Sugarcane Breeding III. Sugarcane Improvement Through Biotechnology Literature Cited

18 20 57 100

3. Breeding for Resistance to Maize Foliar Pathogens

119

Richard C. Pratt and Stuart G. Gordon I. II. III. IV. V.

Introduction Diseases Incited by Fungal Pathogens Diseases Incited by Viral Pathogens Diseases Incited by Bacterial Pathogens Summary Literature Cited

121 125 142 156 159 162

v

vi

CONTENTS

4. Synteny in the Rosaceae

175

Pere Arús, Toshiya Yamamoto, Elisabeth Dirlewanger, and Albert G. Abbott I. II. III. IV.

Introduction Genetic Maps in the Main Rosaceae Species Map Comparisons Other Genetic Resources of Interest for Map Comparison V. Future Prospects Literature Cited

5. Genetic Mapping and Molecular Breeding in Cucurbits

176 177 191 202 203 205

213

Yi-Hong Wang, Ralph A. Dean, and Tarek Joobeur I. II. III. IV. V. VI. VII. VIII. IX.

Introduction Classic Genetic Maps Molecular Genetic Maps Gene Tagging QTL Mapping Molecular Breeding Gene Cloning Cucurbit Genomics Future Prospects Literature Cited

6. Breeding Douglas-Fir

214 215 216 219 226 233 235 236 237 239

245

Glenn T. Howe, Keith Jayawickrama, Marilyn Cherry, G. R. Johnson, and Nicholas C. Wheeler I. II. III. IV. V. VI. VII. VIII.

Abbreviations Introduction Distinctive Characteristics of Forest Trees Douglas-Fir: The Species Factors That Influence Douglas-Fir Breeding Breeding Goals and Objectives Overview of Tree Breeding Methods Breeding Programs

246 247 249 251 254 279 286 289

CONTENTS

IX. X. XI. XII. XIII.

vii

Breeding and Testing Methods Production of Improved Materials for Reforestation Biotechnology Gene Conservation Acknowledgments Literature Cited

296 319 331 337 339 339

Subject Index

354

Cumulative Subject Index

355

Cumulative Contributor Index

371

Contributors

Albert G. Abbott, Department of Genetics, Biochemistry and Life Science Studies, Clemson University, Clemson, South Carolina 29634, USA Pere Arús, Department de Genètica Vegetal, Laboratori de Genètica Molecular Vegetal, CSIC-IRTA, Carretera de Cabrils s/n; 08348, Cabrils, Spain, pere.arus@ irta.es Stevens M. Brumbley, BSES Limited, Indooroopilly, Brisbane QLD 4068, Australia Marilyn Cherry, Department of Forest Science, Oregon State University, 321 Richardson Hall, Corvallis, Oregon 97331-5752, USA Jack C. Comstock, USDA-ARS, 12990 U.S. Hwy 441 N., Canal Point, Florida, 33438, USA Ralph A. Dean, Fungal Genomics Laboratory, Department of Plant Pathology, North Carolina State University, Campus Box 7251, Raleigh, NC 27695-7251, USA Angélique D’Hont, CIRAD, UMR1096, TA 40/03, Avenue Agropolis, 34398 Montpellier Cedex 5, France Elisabeth Dirlewanger, INRA, Unité de Recherches sur les Espèces Fruitières et la Vigne, B.P. 81, F-33 883 Villenave d’Ornon cedex, France Jean C. Glaszmann, CIRAD, UMR1096, TA 40/03, Avenue Agropolis, 34398 Montpellier Cedex 5, France Stuart G. Gordon, Department of Plant Pathology, Ohio Agricultural Research and Development Center, The Ohio State University, Wooster, OH 44691, USA Glenn T. Howe, Department of Forest Science, Oregon State University, 321 Richardson Hall, Corvallis, Oregon 97331-5752, USA, [email protected] Keith Jayawickrama, Department of Forest Science, Oregon State University, 321 Richardson Hall, Corvallis, Oregon 97331-5752, USA John Jifon, The Texas A&M University System, Agricultural Research and Extension Center, 2415 E. Hwy. 83, Weslaco, Texas 78596, USA G. R. Johnson, U.S.D.A. Forest Service, Pacific Northwest Research Station, 3200 SW Jefferson Way, Corvallis, Oregon 97331-4401, USA Tarek Joobeur, Fungal Genomics Laboratory, Department of Plant Pathology, North Carolina State University, Campus Box 7251, Raleigh, NC 27695-7251, USA Prakash Lakshmanan, BSES Limited, Indooroopilly, Brisbane QLD 4068, Australia viii

CONTRIBUTORS

ix

Ray Ming, Hawaii Agriculture Research Center, 99-193 Aiea Heights Drive, Aiea, Hawaii 96701, USA T. Erik Mirkov, The Texas A&M University System, Agricultural Research and Extension Center, 2415 E. Hwy. 83, Weslaco, Texas 78596, USA Paul H. Moore, USDA-ARS, PBARC, 99-193 Aiea Heights Drive, Aiea, Hawaii 96701, USA Thomas C. Osborn, Seminis Vegetable Seeds, Inc., 37437 State Highway 16, Woodland, California 95695, USA, [email protected] Andrew H. Paterson, Plant Genome Mapping Laboratory, University of Georgia, 111 Riverbend Road, Rm 228, Athens, Georgia 30602, USA, [email protected] Richard C. Pratt, Department of Horticulture and Crop Science, Ohio Agricultural Research and Development Center, The Ohio State University, Wooster, OH 44691, USA, [email protected] Mamta Rai, The Texas A&M University System, Agricultural Research and Extension Center, 2415 E. Hwy. 83, Weslaco, Texas 78596, USA Raymond J. Schnell, USDA-ARS, 13601 Old Cutler Rd., Miami, Florida, 33158, USA Jorge da Silva, The Texas A&M University System, Agricultural Research and Extension Center, 2415 E. Hwy. 83, Weslaco, Texas 78596, USA Thomas L. Tew, USDA-ARS, 5883 USDA Road, Houma, Louisiana 70360, USA Yi-Hong Wang, Biology Department, Penn State University Behrend College, 5091 Station Rd., Erie, PA 16563, USA, [email protected] Nicholas C. Wheeler, 21040 Flumerfelt Rd. S.E., Centralia, Washington 98531, USA Kuo-Kao Wu, Hawaii Agriculture Research Center, 99-193 Aiea Heights Drive, Aiea, Hawaii 96701, USA Toshiya Yamamoto, National Institute of Fruit Tree Science, Tsukuba, Ibaraki 305-8605, Japan

Fredrick A. Bliss

1 Dedication: Fredrick A. Bliss Teacher, Researcher, and Director of Plant Breeding Thomas C. Osborn Seminis Vegetable Seeds, Inc. 37437 State Highway 16 Woodland, California 95695

Fred Bliss has always maintained an abiding appreciation for diversity. This is a valuable asset, especially for a plant breeder, because genetic diversity is so important for crop improvement. In Fred’s case, this appreciation of diversity has extended to many other aspects of his career, and undoubtedly contributed to the tremendous success he enjoyed. In fact, this appreciation of diversity probably under-pinned his decision to assume very different roles during his nearly 40-year career, including teacher and researcher at two land grant universities, chair of a university department, and research administrator at a major vegetable seed company. In each role, Fred maintained a broad interest in plant breeding, applying his talents to a wide range of crop species and research topics, and utilizing a variety of technical approaches. His ability to work with many different people and to integrate diverse ideas and approaches were essential elements to his success. All of these assets contributed to a career marked by important scientific discoveries, excellent student training, and valuable guidance on national and international issues.

Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 1

2

T. OSBORN

EARLY YEARS Fred was born in 1938 and raised on a farm near Red Cloud, Nebraska. His interest in agriculture began at an early age, and as a youngster he learned the methods of farming and skills required to manage a farm enterprise. In 1956, he entered the University of Nebraska–Lincoln as an undergraduate in engineering but soon changed his major to agronomy. Although this background as an agronomist from the heart of Nebraska would later create some suspicion among his horticultural colleagues, the education he received at U.N.–Lincoln provided a solid foundation for his future career in the plant sciences. Fred received the B.S. degree from U.N.–Lincoln in 1960 and entered graduate school the same year at the University of Wisconsin–Madison, studying under one of the leading horticultural plant breeders of the time, Warren H. Gabelman. His thesis research focused on cytoplasmic-genic male sterility in table beets, a trait still used today for development of commercial F1 hybrids. This graduate study experience at Wisconsin played an important role in Fred’s professional career, and he has maintained a life-long personal and professional association with Professor Gabelman. After receiving the Ph.D. degree in Horticulture–Genetics, Fred was awarded an NIH Postdoctoral Fellowship to study at the University of Minnesota, where he spent a year on the St. Paul campus working with Charles Gates and Ralph Comstock on computer simulation of selection in selfpollinated crops. By the 1960s, the University of Wisconsin–Madison had developed an outstanding program in plant breeding. This program was strengthened even further in 1966, by the hiring of the young, talented Fred Bliss, as an Assistant Professor. His appointment was in the Department of Horticulture, and his research focus was on breeding self-pollinated crops, especially common bean, Phaseolus vulgaris. However, before settling into academic life in Madison, Fred took an overseas assignment as a member of a U.W.-USAID team stationed at the University of Ife in Western Nigeria. This would prove to be an important experience. Here, Fred saw firsthand the pervasive problems and challenges facing researchers in developing countries. That experience kindled an interest in working to solve some of those problems, and provided a research focus for much of his career. He also gained an appreciation for the difficulties one faces adapting to a foreign culture, which he would later use to help the students from foreign countries who trained with him. After completing a two-year assignment in Nigeria, Fred and the family went to the University of Goettingen in West Germany in 1968 where he stayed for six months as a visiting scientist working with Gerhard

1. DEDICATION: FREDERICK A. BLISS

3

Robbelin. His research topic was a bit before its time: Development of Arabidopsis thaliana as a model system for studying quantitative expression. While in Europe, he traveled to Cambridge University to visit John Thoday, who introduced Fred to recent work showing that single genetic factors contributed to quantitative variation. This stimulated his interest in the genetic control of quantitative expression, which Fred would later explore in his research on common bean. Upon returning to Madison in 1969, Fred became immersed in campus life, initiating a research program in breeding self-pollinated crops and a teaching program for graduate and undergraduate students. With Warren (Buck) Gabelman, he co-instructed the introductory plant breeding course and taught a course on vegetable production. He became an important contributor to the newly created graduate program in Plant Breeding and Plant Genetics, and soon was a popular member of many graduate student committees. Although his research would focus on Phaseolus vulgaris, it would also encompass a wide range of topics and many different approaches, from traditional hybridization and selection procedures to implementation of new methods, such as rocket immunoelectrophoresis to quantify protein and development of inbred backcross lines to identify loci contributing to quantitative trait expression. This program provided training for scientists from all parts of the world.

BREEDING PROGRAMS AT THE UNIVERSITY OF WISCONSIN The faculty breeding position at U.W.–Madison was initiated, in part, to support a snap bean production industry that was expanding in the central sands area of Wisconsin. Fred’s research helped advance this industry through development of breeding methods for self-pollinated crops and improved bean germplasm. Initially, this effort was focused on resistance to important diseases, such as root rot caused by Rhizoctonia, Pythium and Phytophthora, halo blight, and bacterial brown spot, and on improvement of pod quality. Later, his work on biological nitrogen fixation would be applied to snap bean production in Wisconsin, as well as dry bean production elsewhere in the world. Many people contributed to this effort, key among whom was Ken Kmiecik, who also participated in many other projects conducted by Fred and his students, post-doctorals, and visiting scholars. Another notable area of research involved improvement of dry edible beans for traits important in developing countries, such as nutritional quality and reduced production inputs. This interest began with his

4

T. OSBORN

earlier work in Nigeria and led not only to practical innovations but also to pioneering work on genetic control of seed protein accumulation and biological nitrogen fixation. The seed protein work benefited from close collaboration with Timothy Hall in the Department of Horticulture; with their students and post docs, they created a comprehensive body of knowledge on the physiology, genetics and molecular biology of bean seed proteins. Fred’s research on nitrogen fixation was the first to utilize genetic variation for this trait in Phaseolus vulgaris to improve efficiency of nitrogen fixation through breeding. The nitrogen fixation work benefited from collaborations with Winston Brill and Robert Burris, also at the University of Wisconsin. Researchers in Brazil and at the International Center for Tropical Agriculture (CIAT) in Colombia played important roles in applying the findings from these studies to regions of the world where they could have the most impact. Fred’s association with Dermot Coyne on the Bean-Cowpea CRSP also was an influential factor in the success of these projects, and this collaboration was part of a life-long association with Dermot. One theme that has persisted throughout Fred’s research is the importance of identifying sources of useful genetic variation and developing efficient ways to utilize those resources for crop plant improvement. While working to improve Phaseolus, he realized that development of inbred backcross lines was an efficient method for introgressing genes controlling quantitative traits, and demonstrated the utility of this approach in several projects. He also was an early innovator in the use of molecular methods for studying genetic variation. The contributions of Fred’s research group on seed protein variation in cultivated and wild beans led to important insight about domestication of Phaseolus vulgaris and the discovery and utilization of arcelin seed protein as a mechanism that confers insect resistance. Again, collaboration with other scientists, such as entomologists Caesar Cardona and Art VanSchoonhoven at CIAT, was key to success in these projects.

UNIVERSITY OF CALIFORNIA, DAVIS In 1988, Fred was offered and accepted the first Will W. Lester Endowed Chair in the Department of Pomology at the University of California in Davis. The move to California represented a huge transition in research focus, and is a testament to his ability to adapt to and thrive in diverse environments. For the next ten years, he led collaborative genetic and breeding research on tree fruits, initiating a rootstock breeding program that continues at present, and developing molecular tools for genetic and

1. DEDICATION: FREDERICK A. BLISS

5

breeding applications in stone fruits (peach and almonds). At U.C.– Davis, he served as chair of the Plant Biology Graduate Program from 1990 to 1992 and chair of the Department of Pomology from 1991 to 1994. This was a difficult time to be an administrator. The weak California economy had forced budget and personnel reductions, and Fred provided much-needed leadership during this tough period of downsizing. In 1995, he established the Laboratory for Genetic Identification and directed that facility until 1998 when he became Professor Emeritus at U.C.–Davis.

SEMINIS VEGETABLE SEEDS The next transition in Fred’s career was perhaps the most challenging, but also one of the most rewarding. In 1998, he accepted the position of Director of Worldwide Plant Breeding for Seminis Vegetable Seeds, Inc., in Woodland, California. Seminis had been formed through the merger of several smaller seed companies, and one of Fred’s primary responsibilities was to integrate the 100+ different breeding programs into a cohesive strategy for developing new vegetable cultivars. He managed this diverse group of plant breeders located at research stations throughout the world with seven associate directors who reported directly to him. In 1990, Fred was part of a small senior management team that provided leadership in optimizing activities of the Research and Development Division of Seminis. When that group completed work, he assumed responsibility as Senior Director of Support Technology, which included areas of cell biology, molecular markers and applied genomics, pathology, vegetable quality, and foundation seed. In December 2003, Fred began a reduced-time appointment with responsibility for special projects, a position he currently holds.

ACCOMPLISHMENTS Throughout his career, Fred has provided his knowledge and expertise in the service of various research and professional organizations. He was a member of international review teams that evaluated research programs at EMBRAPA, CIAT, USAID, and FAO. His service to national organizations included membership on committees and review panels at USDA/NRI, USDA/ARS, NRC/NAS, and the U.S. National Plant Genetic Resources Program. Fred has held important posts in professional organizations, including Chair of the Strategic Planning Committee,

6

T. OSBORN

Vice President–Research, and President and Chairman of the Board of Directors of the American Society for Horticultural Science (ASHS). He served the Pomology Society as member of the Wilder Committee, and the Crop Science Society of America (CSSA) as a member of several committees. Fred’s editorial skills were widely recognized and put to good use as Associate Editor of the Journal and Chair of the Publication Committee for ASHS, and as Scientific Editor for the Kluwer book series on Biotechnology in Agriculture. Fred has received many awards for his professional accomplishments, including Outstanding Graduate Educator, the Asgrow Award and B.Y. Morrison Memorial Lectureship from the ASHS, and the Meritorious Service Award from the Bean Improvement Cooperative. In 1986 he was awarded the ASSINSEL Grand Prize from the International Association of Plant Breeders for the Protection of Plant Varieties. Fred was elected as a Fellow of the ASHS in 1985, the CSSA in 1986, and the American Association for the Advancement of Science in 1990. He was also presented with many consulting opportunities throughout his career. Receiving the Master’s Week Alumni Award from his alma mater, the University of Nebraska–Lincoln, in 1987 was especially appreciated. One of the greatest professional and personal rewards for Fred has come from the many interactions he enjoyed with students and colleagues. I believe that these interactions have been even more rewarding for those of us who had the pleasure to work with Fred. Fred’s mentoring skills were highly appreciated by his students (and recognized by many who wish they had been his students!). He always provided excellent ideas and boundaries for thesis research, but he also gave plenty of room and encouragement for students to explore their own ideas. One of his proudest achievements is that nearly every student who began a degree program with him finished their degree with him. His colleagues also greatly enjoyed their interactions with Fred, and they frequently comment on this in mentioning their association with him. In a research environment that can often be extremely competitive, Fred is a shining example of how it is possible to do good science and be a good person. While Fred’s many attributes can be mentioned only briefly in this short dedication, one that deserves special attention is his ability to see the larger picture and to integrate pieces of a complex problem. This is evident from the many different lines of research that he undertook during his career, and from his multifaceted approach to research problems. Fred was always comfortable incorporating new research techniques, such as those involving new physiology or molecular biology methods; and he encouraged collaboration among scientists from different disciplines. One example of his integrative skills that is particularly vivid to

1. DEDICATION: FREDERICK A. BLISS

7

me involved the discovery of arcelin seed protein as the factor conferring resistance to seed-boring bruchid insects. A group of students (including myself) had discovered this unusual seed protein in some accessions of wild beans, but our interest did not go further until Fred recognized that the same accessions also had been listed as highly resistant to bruchids in a CIAT report. This observation, and Fred’s connection to scientists that were eager to collaborate with him, led to a series of experiments that revealed a very interesting evolutionary history for arcelin and a practical solution for developing regions of the world that suffer from the devastation caused by this insect pest. I am sure that every student and colleague who worked with Fred has similar stories. I would be remiss not to point out that Fred has always led a balanced lifestyle and that his family and friends have played important roles in helping maintain this balance. He raised three sons and enjoyed many activities with them while they were growing up, including coaching soccer. Today, Fred and his wife Mary live in Davis, California, where they enjoy travel, antiquing, walking, art collecting, and each other. They also enjoy visiting the families of their five grown children, including five grand children. Fred is still very active in the plant breeding world, and he continues to participate in diverse activities related to this profession. He is always thrilled to see former students and colleagues, and to meet new ones. Those of you who know him undoubtedly look forward to these connections. I encourage those of you who have not met Fred to seek out his wealth of knowledge and wisdom.

PUBLICATIONS OF FREDRICK A. BLISS Bliss, F. A., and W. H. Gabelman. 1965. Inheritance of male sterility in beets, Beta vulgaris L. Crop Sci. 5:403–406. Bliss, F. A. 1968. Plant introduction to improve Nigerian crops. Nigerian Agriculture J. 5(2):82–87. Bliss, F. A. 1968. Onion varieties for the Western State of Nigeria. Agr. News Western State Nigeria 7:13–17. Bliss, F. A., and C. E. Gates. 1968. Directional selection in simulated population of selfpollinated plants. Australian J. Biol/ Sci. 21:705–719. Bliss, F. A. 1969. The efficiency of developing male-sterile and male-fertile inbred components by backcrossing. HortScience 4:49–51. Bliss, F. A., and D. C. Arny. 1971. Breeding tomatoes for resistance to leaf mould (Cladosporium fulvum) in Western Nigeria. Expt. Agr. 7:177–181. Bliss, F. A. 1971. Inheritance of growth habit and time of flowering in certain lines of beans, Phaseolus vulgaris L. J. Am. Soc. Hort. Sci. 96:715–717. Bliss, F. A., and D. G. Robertson. 1971. Genetics of host reaction to cowpea to cowpea yellow mosaic virus and cowpea mottle virus. Crop Sci. 11:258–262.

8

T. OSBORN

Hall, T. C., R. C. McLeester, and F. A. Bliss. 1972. Electrophoretic changes during the development of the French bean fruit. Phytochemistry 11:647–649. McLeester, R. C., T. C. Hall, S. M. Sun, and F. A. Bliss. 1973. Comparison of globulin proteins from Phaseolus vulgaris with those from Vicia faba. Phytochemistry 2:85–93. Bliss, F. A., P. T. Onesirosan, and D. C. Arny. 1973. Inheritance of resistance in tomato to target leaf spot. Phytopathology 63:837–840. Bliss, F. A., L. N. Barker, J. D. Franckowiak, and T. C. Hall. 1973. Genetic and environmental variation of seed yield, yield components and seed protein quantity and quality of cowpea (Vigna unguiculata (L.) Walp.). Crop Sci. 13:656–660. Sun, S. M., R. C. McLeester, F. A. Bliss, and T. C. Hall. 1974. Reversible and irreversible dissociation of globulins from Phaseolus vulgaris seed. J. Biol. Chem. 249:2118– 2121. Bliss, F. A. 1975. In cowpeas—Nigeria. p. 151–158. In: M. Milner (ed.), Nutritional improvement of food legumes by breeding. Wiley, New York. Renish, W. J., P. T. Onesirosan, F. A. Bliss, and D. C. Arny. 1975. Reaction of tomato to target leaf spot using spore inoculation and a toxin test. HortScience 10:163–165. Kelly, J. D., and F. A. Bliss. 1975. Quality factors affecting the nutritive value of bean seed protein. Crop Sci. 15:753–757. Kelly, J. D., and F. A. Bliss. 1975. Heritability estimates of percentage seed protein and available methionine and correlations with yield in dry beans, Phaseolus vulgaris L. Crop Sci. 15:753–757. Romero, J., S. M. Sun, R. C. McLeester, F. A. Bliss, and T. C. Hall. 1975. Heritable variation in a polypeptide subunit of the major storage protein of the bean, Phaseolus vulgaris L. Plant Physiol. 56:776–779. Bliss, F. A. 1976. Use of hill plots for genetic and breeding studies of bean. J. Am. Soc. Hort. Sci. 10:77–80. Hall, T. C., F. A. Bliss, D. S. Ryan, and S. M. Sun. 1976. The subunit structure and cellfree synthesis of the major storage protein from bean (Phaseolus vulgaris) seeds. Acides Nucleiques et Synthese des Proteines Chez Les Vegetaux. Bliss, F. A., and T. C. Hall. 1977. Food legume-compositional and nutritional changes induced by breeding. Cereal Foods World 22:106–113. Boomstra, A. C., F. A. Bliss, and S. E. Beebe. 1977. New sources of Fusarium root rot resistance in Phaseolus vulgaris L. J. Am. Soc. Hort. Sci. 102:182–185. Boomstra, A. G., and F. A. Bliss. 1977. Inheritance of resistance to Fusarium solani f.sp. phaseoli in beans Phaseolus vulgaris L.) and breeding strategy to transfer resistance. J. Am. Soc. Hort. Sci. 102:186–188. Ladeinde, T. A. O., and F. A. Bliss. 1977. A preliminary study of the production of plantlets from anthers of cowpea. Tropical Grain Legume Bul. 8:13. Hall, T. C., R. C. McLeester, and F. A. Bliss. 1977. Equal expression of the maternal and paternal loci for the polypeptide subunits of the major storage protein of the bean, Phaseolus vulgaris L. Plant Physiol. 59:1122–1124. Hall, T. C., F. A. Bliss, D. S. Ryan, and S. M. Sun. 1977. The subunit structure and cellfree synthesis of the major storage protein from bean (Phaseolus vulgaris L.) seed. p. 59–101. In: Roy Baker (ed.), Nucleic acids and protein synthesis in plants, Plenum Press, London. Ma, Y., and F. A. Bliss. 1978. Tannin content and inheritance in common bean (Phaseolus vulgaris L.). Crop Sci. 18:201–204. Ma, Y., and F. A. Bliss. 1978. Seed proteins of common bean. Crop Sci. 17:431–437. Sun, S. M., M. A. Mutschler, F. A. Bliss, and T. C. Hall. 1978. Protein synthesis and accumulation in bean cotyledons during growth. Plant Physiol. 61:918–923.

1. DEDICATION: FREDERICK A. BLISS

9

Hall, T. C., Y. Ma, B. U. Buchbinder, J. W. Pyne, S. M. Sun, and F. A. Bliss. 1978. Messenger RNA for G1 protein of French bean seeds: Cell-free translation and product characterization. Proc. Natl. Acad. Sci. (USA) 75:3196–3200. Knudson Butler, L. K., T. W. Tibbitts, and F. A. Bliss. 1979. Inheritance of resistance to ozone in Phaseolus vulgaris L. J. Am. Soc. Hort. Sci. 104:211–213. Hall, T. C., S. M. Sun, Y. Ma, R. C. McLeester, J. W. Pyne, F. A. Bliss, and B. U. Buchbinder. 1979. The major storage protein of French bean seeds: Characterization in vivo and translation in vitro. p. 3–26. In: I. Rubenstein, R. L. Phillips, C. E. Green, and B. G. Gengenbach (eds.), The plant seed. Academic Press, New York. Mutschler, M. A., F. A. Bliss, and T. C. Hall. 1980. Variation in the accumulation of seed storage protein among genotypes of Phaseolus vulgaris L. Plant Physiol. 65:627–630. Mutschler, M. A. and F. A. Bliss. 1980. Genic male sterility in the common bean (Phaseolus vulgaris L.). J. Am. Soc. Hort. Sci. 105:202–205. Brown, J. W. S., F. A. Bliss, and T. C. Hall. 1980. Microheterogeneity of globulin-1 storage protein from French bean with isoelectrofocusing. Plant Physiol. 66:838–840. Ma, Y., F. A. Bliss, and T. C. Hall. 1980. Peptid mapping reveals considerable sequence homology among the three polypeptide subunits of G1 storage protein from French bean seed. Plant Physiol. 66:897–902. Bliss, F. A. 1980. Breeding legumes for nutritional quality. p. 179–185. In: R. J. Summerfield and A. H. Bunging (eds.), Advances in legume science. Royal Botanic Gardens, Kew. Bliss, F. A. 1980. Common bean. p. 273–284. In: W. R. Fehr and H. H. Handley (eds.), Hybridization of crop plants. Am. Soc. Agron.-Crop Sci. Soc. Am., Madison, WI. Hall, T. C., S. M. Sun, B. U. Buchbinder, J. W. Pyne, F. A. Bliss, and J. D. Kemp. 1980. Bean seed globulin mRNA: Translation, characterization, and its use as a probe towards genetic engineering of crop plants. p. 259–272. In: C. J. Leaver (ed.), Genome organization and expression in plants. Plenum Press, New York. Beebe, S. E., F. A. Bliss, and H. F. Schwartz. 1981. Root rot resistance in common bean germ plasm of Latin American origin. Plant Dis. 65:485–489. Mutschler, M. A., and F. A. Bliss. 1981. Inheritance of bean seed globulin content and its relationship to protein content and quality. Crop Sci. 21:289–294. Bliss, F. A. 1981. Utilization of vegetable germplasm. HortScience 16:129–132. Sullivan, J. G., and F. A. Bliss. 1981. Compensation for missing plants in field experiments with the common bean. HortScience 16:185–186. Brown, J. W. S., F. A. Bliss, and T. C. Hall. 1981. Linkage relationships between genes controlling seed proteins in French bean. Theor. Appl. Gen. 60:251–259. Brown, J. W. S., Y. Ma, F. A. Bliss, and T. C. Hall. 1981. Genetic variation in the subunits of globulin-1 storage protein of the French bean. Theor. Appl. Gen. 59:83–88. Brown, J. W. S., T. C. Osborn, F. A. Bliss, and T. C. Hall. 1981. Genetic variation in the subunits of globulin-2 and albumin seed proteins of French bean. Theor. Appl. Genet. 60:245–250. Bliss, F. A., and J. W. S. Brown. 1982. Genetic control of phaseolin protein expression in seeds of common bean (Phaseolus vulgaris L.). Qual. Plant Foods Human Nutr. 31:269–279. Brown, J. W. S., T. C. Osborn, F. A. Bliss, and T. C. Hall. 1982. Bean lectins I: Relationships between agglutinating activity and electrophoretic variation in the lectincontaining G2/albumin seed proteins of French bean (Phaseolus vulgaris L.). Theor. Appl. Gen. 62:263–271. Brown, J. W. S., T. C. Osborn, F. A. Bliss, and T. C. Hall. 1982. Bean lectins, Part 2: Relationship between qualitative lectin variation in Phaseolus vulgaris and previous observations on purified bean lectins. Theor. Appl. Gen. 62:361–367.

10

T. OSBORN

Brown, J. W. S., J. R. McFerson, F. A. Bliss, and T. C. Hall. 1982. Genetic divergence among commercial classes of Phaseolus vulgaris in relation to phaseolin pattern. HortScience 17:752–754. McFerson, J., F. A. Bliss, and J. C. Rosas. 1982. Selection for enhanced nitrogen fixation in common bean, Phaseolus vulgaris. p. 39–43. In: P. Graham and S. Harris (eds.), Biological nitrogen fixation technology for tropical agriculture. Centro International de Agricultura Tropical, Cali, Colombia. Leal, N. R., I. A. Hamad, and F. Bliss. 1982. Avaliacao dos progenitores e linhas avancadas. Brasilia 17(2):225–231. Sullivan, J. G., and F. A. Bliss. 1983. Recurrent mass selection for increased seed yield and seed protein percentage in the common bean (Phaseolus vulgaris L.) using a selection index. J. Am. Soc. Hort. Sci. 108:42–46. Sullivan, J. G., and F. A. Bliss. 1983. Genetic control of quantitative variation in phaseolin seed protein of common bean. J. Am. Soc. Hort. Sci. 108:782–787. Sullivan, J. G., and F. A. Bliss. 1983. Expression of enhanced seed protein content in inbred backcross lines of common bean. J. Am. Soc. Hort. Sci. 108:787–791. Osborn, T. C., K. A. Ausloos, J. W. S. Brown, and F. A. Bliss. 1983. Bean lectins III. Evidence for greater complexity in the structural model of Phaseolus vulgaris lectin. Plant Sci. Lett. 31:193–203. Bliss, F. A., and John W. S. Brown. 1983. Breeding common bean (Phaseolus vulgaris L.) for improved quantity and quality of seed protein. Plant Breed. Rev 1:59–102. Bliss, F. A. 1984. The application of new plant biotechnology to crop improvement. HortScience 19:43–48. Thomas, R. J., J. R. McFerson, L. E. Schrader, and F. A. Bliss. 1984. Composition of bleeding sap nitrogen from lines of field-grown Phaseolus vulgaris L. Plant & Soil 79:77–88. Osborn, T. C., J. F. Manen, J. W. S. Brown, and F. A. Bliss. 1984. Bean lectins IV: Genetic variation in the non-denatured structure of lectins from different Phaseolus vulgaris L. cultivars. Theor. Appl. Gen. 67:547–552. Estrada, D. S., M. A. Mutschler, and F. A. Bliss. 1984. Temperature-influenced instability in a genic male sterile common bean. HortScience 19:401–402. Gepts, P., and F. A. Bliss. 1984. Enhanced available methionine concentration associated with higher phaseolin levels in common bean seeds. Theor. Appl. Gen. 69:47–53. Osborn, T. C., J. W. S. Brown, and F. A. Bliss. 1985. Bean lectins V: Quantitative genetic variation in seed lectins of Phaseolus vulgaris L. and its relationship to qualitative lectin variation. Theor. Appl. Gen. 70:22–31. Osborn, T. C., and F. A. Bliss. 1985. Effects of genetically removing lectin seed protein on horticultural and seed characteristics of common bean. J. Am. Soc. Hort. Sci. 110:484–488. Owens, K. W., F. A. Bliss, and C. E. Peterson. 1985. Genetic analysis of fruit length and weight in two cucumber populations using the inbred backcross line method. J. Am. Soc. Hort. Sci. 110:431–436. Owens, K. W., F. A. Bliss, and C. E. Peterson. 1985. Genetic variation within and between two cucumber populations derived via the inbred backcross line methods. J. Am. Soc. Hort. Sci. 110:437–441. Gepts, P., and F. A. Bliss. 1985. F1 hybrid weakness in the common bean. Differential geographic origin suggests two gene pools in cultivated bean germplasm. J. Hered. 76:447–450. Romero-Andreas, J., and F. A. Bliss. 1985. Heritable variation in the phaseolin protein of nondomesticated common bean Phaseolus vulgaris L. Theor. Appl. Gen. 71:478–480.

1. DEDICATION: FREDERICK A. BLISS

11

Bliss, F. A. 1985. Relationships of genetic engineering to conventional genetic technology and plant breeding. p. 65–87. In: A. M. Altschul and H. L. Wilcke (eds.), New protein foods, Vol. 5. Academic Press, New York. Norton, G., F. A. Bliss, and R. Bressani. 1985. Biochemical and nutritional attributes of grain legumes. p. 73–114. In: R. J. Summerfield and E. H. Roberts (eds.), Grain legume crops. Collins and Sons, Ltd. Bliss, F. A. 1985. Breeding for enhanced dinitrogen fixation potential of common bean, (Phaseolus vulgaris L.). p. 303–310. In: P. Ludden and J. Burris (eds.), Nitrogen fixation and CO2 metabolism. Proc. 14th Steenbock Symp. Elsevier Publ., New York. Attewell, J., and F. A. Bliss. 1985. Host plant characteristics of common bean lines selected using indirect measures of N2 fixation. p. 3–9. In: P. J. Bottomley and W. E. Newton (eds.), Nitrogen fixation research progress. Martinus Nijhoff Publ., Dordrecht, The Netherlands. Rosas, J. C., and F. A. Bliss. 1985. Improvement of the nitrogen fixation potential of common bean in Latin America. Proc. Interciencia Symp. “Biotechnology in the Americas— II. Applications in Tropical Agriculture.” Rosas, J. C., and F. A. Bliss. 1986. Host plant traits associated with estimates of nodulation and nitrogen fixation in common bean. HortScience 21:287–289. Romero-Andreas, J., B. S. Yandell, and F. A. Bliss. 1986. Bean arcelin. 1. Inheritance of a novel seed protein of Phaseolus vulgaris L. and its effect on seed composition. Theor. Appl. Gen. 72:123–128. Osborn, T. C., T. Blake, P. Gepts, and F. A. Bliss. 1986. Bean arcelin. 2. Genetic variation, inheritance and linkage relationships of a novel seed protein of Phaseolus vulgaris L. Theor. Appl. Gen. 71:847–855. Gepts, P., T. C. Osborn, K. Rashka, and F. A. Bliss. 1986. Phaseolin-protein variability in wild forms and landraces of the common bean (Phaseolus vulgaris): Evidence for multiple centers of domestication. Econ. Bot. 40:451–468. Gepts, P., and F. A. Bliss. 1986. Phaseolin variability among wild and cultivated common beans (Phaseolus vulgaris) from Colombia. Econ. Bot. 40:469–478. Araujo, R. S., J. Maya-Flores, D. Barnes-McConnell, C. Yokoyama, F. B. Dazzo, and F. A. Bliss. 1986. Semienclosed tube cultures of bean plants (Phaseolus vulgaris L.) for enumeration of Rhizobium phaseoli by the most-probable-number technique. Appl. Environ. Microbiol. 52:954–956. Rosas, J. C., and F. A. Bliss. 1986. Improvement of the nitrogen fixation potential of common bean in Latin America. CEIBA 27:245–259. Bliss, F. A. 1987. Host plant control of symbiotic N2 fixation in grain legumes. pp. 479–493. In: W. H. Gabelman and B. C. Loughman (eds.), Genetics aspects of plant mineral nutrition. Martinus Nijhoff Publ., Boston. Wynne, J. C., F. A. Bliss, and J. C. Rosas. 1987. Principles and practices of field designs to evaluate symbiotic nitrogen fixation. p. 371–389. In: G. H. Elkan (ed.), Symbiotic nitrogen fixation technology. Marcel Dekker, New York. Pereira, P. A. A., and F. A. Bliss. 1987. Nitrogen fixation and plant growth of common bean (Phaseolus vulgaris L.) at different levels of phosphorus availability. Plant & Soil 104:79–84. Bliss, F. A., and J. C. Miller, Jr. 1988. Selecting and breeding grain legumes for enhanced nitrogen fixation. In: R. J. Summerfield (ed.), World crops: Cool season food legumes. Marinus Nijhoff Publ., Dordrecht, The Netherlands. Quebedeaux, B., and F. A. Bliss (eds.). 1988. Horticulture and human health—Contributions of fruits and vegetables. Prentice-Hall, Englewood Cliffs, N.J.

12

T. OSBORN

Buso, G. S. C., and F. A. Bliss. 1988. Variability among lettuce cultivars grown at two levels of available phosphorus. Plant & Soil 111:67–73. St. Clair, D. A., D. J. Wolyn, J. DuBois, R. H. Burris, and F. A. Bliss. 1988. Field comparison of dinitrogen fixation determined with nitrogen-15-depleted and nitrogen-15enriched ammonium sulfate in selected inbred backcross lines of common bean. Crop Sci. 28:773–778. Gepts, P., K. Kmiecik, P. Pereira, and F. A. Bliss. 1988. Dissemination paths of the common bean (Phaseolus vulgaris) deduced from phaseolin electrophoretic variability. I. The Americas. Econ. Bot. 42:73–85. Gepts, P., and F. A. Bliss. 1988. Dissemination paths of the common bean (Phaseolus vulgaris) deduced from phaseolin electrophoretic variability. Econ. Bot. 42:86–104. Osborn, T. C., M. Burow, and F. A. Bliss. 1988. Purification and characterization of arcelin seed protein from common bean. Plant Physiol. 86:399–405. Osborn, T. C., D. C. Alexander, S. S. M. Sun, C. Cardona, and F. A. Bliss. 1988. Insecticidal activity and lectin homology of arcelin seed protein. Science 240:207–240. Pereira, P., R. H. Burris, and F. A. Bliss. 1989. 15N-determined dinitrogen fixation potential of genetically diverse bean lines (Phaseolus vulgaris L.). Plant & Soil 120:171–179. Wolyn, D. J., J. Attewell, P. W. Ludden, and F. A. Bliss. 1989. Indirect measures of N2 fixation in common bean (Phaseolus vulgaris L.) under field conditions: The role of lateral root nodules. Plant & Soil 113:181–187. Bliss, F. A. 1989. Utilization of genetic resources for crop improvement: The common bean. p. 320–336. In: A. D. H. Brown, M. T. Clegg, Alex L. Kahler, and B. S. Weir (eds.), Population genetics, plant breeding and gene conservation. Sinauer Assoc., Sunderland, MA. Bliss, F. A. 1989. Plant breeding, crop cultivars, and the nature of genetic variability. In: B. E. Caldwell and J. A. Schillinger et al. (eds.), Intellectual property rights associated with plants. Am. Soc. Agron. Special Publ. 52:69–89. Bliss, F. A., P. A. A. Pereira, R. S. Araujo, R. A. Henson, K. A. Kmiecik, J. R. McFerson, M. G. Teixeira, and C. C. Da Silva. 1989. Registration of five high nitrogen fixing common bean germplasm lines. Crop Sci. 29:240–241. Pereira, P. A. A., and F. A. Bliss. 1989. Selection of common bean (Phaseolus vulgaris L.) for N2 fixation at different levels of available phosphorus under field and environmentallycontrolled conditions. Plant & Soil 115:75–82. Delaney, D. E., and F. A. Bliss. 1991. Selection for increased percentage phaseolin in common bean. 1. Comparison of selection for seed protein alleles and S1 family recurrent selection. Theor. Appl. Gen. 81:301–305. Delaney, D. E., and F. A. Bliss. 1991. Selection for increased percentage phaseolin in common bean. 2. Changes in frequency of seed protein alleles with S1 family recurrent selection. Theor. Appl. Gen. 81:306–311. St. Clair, D. A., and F. A. Bliss. 1991. Intrapopulation recombination for 15N-determined dinitrogen fixation ability in common bean. Plant Breed. 106:215–225. Josephson, K. L., D. P. Bourque, F. A. Bliss, and I. L. Pepper. 1991. Competitiveness of Kim 5 and Viking 1 bean Rhizobia: Strain by cultivar interactions. Soil Biol. Biochem. 23:249–253. Wolyn, D. J., D. A. St. Clair, J. DuBois, J. C. Rosas, R. H. Burris, and F. A. Bliss. 1991. Distribution of nitrogen in common bean (Phaseolus vulgaris L.) genotypes selected for differences in nitrogen fixation ability. Plant & Soil 138:303–311. Miranda, B. D., and F. A. Bliss. 1991. Selection for increased seed nitrogen accumulation in common bean: Implications for improving dinitrogen fixation and seed yield. Plant Breed. 106:301–311.

1. DEDICATION: FREDERICK A. BLISS

13

Henson, R. A., and F. A. Bliss. 1991. Effects of N fertilizer application timing on common bean production. Fertilizer Res. 29:133–138. Shellie-Dessert, K. C., and F. A. Bliss. 1991. Genetic improvement of food quality factors. p. 649–677. In: O. Voysest and A. van Schoonhoven (eds.), Common beans: Research for crop improvement. CIAT, Cali, Colombia. C.A.B. Int., Wallingford, UK. Meyer, M. L., and F. A. Bliss. 1991. Pollination potential of staminate kiwifruit plants in vineyards at three locations in California. Acta Hort. 297:283–289. Bliss, Fredrick A. 1992. Breeding plants for enhanced beneficial interactions with soil microorganisms. p. 251–273. In: H. T. Stalker and J. P. Murphy (eds.), Plant breeding in the 1990’s. C.A.B. Int., Wallingford, Oxon, UK. Pereira, P. A. A., B. D. Miranda, J. R. Attewell, K. A. Kmiecik, and F. A. Bliss. 1993. Selection for increased nodule number in common bean (Phaseolus vulgaris L.). Plant & Soil 148:203–209. Bliss, F. A. 1993. Biotechnology and plant breeding. Acta Hort. 336:23–31. Henson, R. A., P. A. A. Pereira, J. E. S. Carneiro, and F. A. Bliss. 1993. Registration of ‘Ouro Negro’, a high dinitrogen-fixing, high-yielding common bean. Crop Sci. 33:644. Bliss, F. A., and G. Hardarson (eds.). 1993. Enhancement of biological nitrogen fixation of common bean in Latin America, Vol. 52. Developments in plant and soil sciences. Kluwer Academic Publ., Dordrecht, The Netherlands. Hardarson, G., F. A. Bliss, M. R. Cigales-Rivero, R. A. Henson, J. A. Kipe-Nolt, L. Longeri, A. Manrique, J. J. Pena-Cabriales, P. A. A. Pereira, C. A. Sanabria, and S. M. Tsai. 1993. Genotypic variation in biological nitrogen fixation by common bean. Plant & Soil 152:59–70. Bliss, F. A. 1993. Breeding common bean for improved biological nitrogen fixation. Plant & Soil 152:71–79. Bliss, F. A. 1993. Utilizing the potential for increased nitrogen fixation in common bean. Plant & Soil 152:157–160. Lewis, M. E., and F. A. Bliss. 1994. Tumor formation and β-glucuronidase expression in Phaseolus vulgaris inoculated with Agrobacterium tumefaciens. J. Am. Soc. Hort. Sci. 119:361–366. Foolad, M. R., S. Arulsekar, and F. A. Bliss. 1995. A genetic map of Prunus based on an interspecific cross between peach and almond. Theor. Appl. Genet. 91:262–269. Pereira, P. A., M. Yokoyama, E. D. Quintela, and F. A. Bliss. 1995. Controle do caruncho Zabrotes Subfasciatus (Bohemann, 1833) (Coleoptera-Bruchidae) pelo uso de proteina de semente em linhagens quase-isogenicas do feijoeiro. Pesq. Agropec. Bras. 30:1031–1034. Warburton, M. L. and F. A. Bliss. 1996. Genetic diversity in peach (Prunus persica L. Batch) revealed by randomly amplified polymorphic DNA (RAPD) markers and compared to inbreeding coefficients. J. Am. Soc. Hort. Sci. 121:1012–1019. Warburton, M. L., V. L. Becerra-Velásquez, J. C. Goffreda, and F. A. Bliss. 1996. Utility of RAPD markers in identifying genetic linkages to genes of economic interest in peach. Theor. Appl. Genet. 93:920–925. Bliss, F. A., A. A. Almehdi, A. M. Dandekar, P. L. Schuerman, and N. Bellaloui. 1999. Crown gall resistance in Accessions of 20 Prunus species. HortScience 34:326–330. Pereira, P. A. A., A. Oliver, F. A. Bliss, L. Crowe and J. Crowe. 2002. Preservation of rhizobia by lyophilization with trehalose. Pesq. Agropec. Bras., Brasilia 37:831–839. Bliss, F. A., S. Arulsekar, M. R. Foolad, V. Becerra, A. M. Gillen, M. L. Warburton, A. M. Dandekar, G. M. Kocsisne, and K. K. Mydin. 2002. An expanded genetic linkage map of Prunus based on an interspecific cross between almond and peach. Genome 45:520–529.

14

T. OSBORN

Osborn, T. C., L. M. Harweck, R. H. Harmsen, R. D. Vogelzang, K. A. Kmiecik, and F. A. Bliss. 2003. Registration of Phaseolus vulgaris genetic stocks with altered seed protein compositions. Crop Sci. 43:1570–1571. Gillen, A. M., and F. A. Bliss. 2005. Identification and mapping of markers linked to the Mi gene for root-knot resistance in peach. J. Am. Soc. Hort. Sci. 130:24–33.

2 Sugarcane Improvement through Breeding and Biotechnology Ray Ming Hawaii Agriculture Research Center 99-193 Aiea Heights Drive Aiea, Hawaii 96701 USA Paul H. Moore USDA-ARS, PBARC 99-193 Aiea Heights Drive Aiea, Hawaii 96701 USA Kuo-Kao Wu Hawaii Agriculture Research Center 99-193 Aiea Heights Drive Aiea, Hawaii 96701 USA Angélique D’Hont and Jean C. Glaszmann CIRAD UMR1096, TA 40/03, Avenue Agropolis 34398 Montpellier Cedex 5 France Thomas L. Tew USDA-ARS 5883 USDA Road Houma, Louisiana 70360

USA

Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 15

16

MING ET AL.

T. Erik Mirkov, Jorge da Silva, John Jifon, and Mamta Rai The Texas A&M University System Agricultural Research and Extension Center 2415 E. Hwy. 83 Weslaco, Texas 78596 USA Raymond J. Schnell USDA-ARS 13601 Old Cutler Rd. Miami, Florida, 33158

USA

Stevens M. Brumbley and Prakash Lakshmanan BSES Limited Indooroopilly, Brisbane QLD 4068 Australia Jack C. Comstock USDA-ARS 12990 U.S. Hwy 441 N. Canal Point, Florida, 33438

USA

Andrew H. Paterson* Plant Genome Mapping Laboratory University of Georgia 111 Riverbend Road, Rm 228 Athens, Georgia 30602 USA

I. INTRODUCTION II. SUGARCANE BREEDING A. History 1. Origin and Domestication 2. Early Breeding 3. Prospects B. Genetic Resources 1. Collections 2. Passport and Descriptor Information 3. Production of True Seed

*Corresponding author We thank Melinda Moore for editing the manuscript. We dedicate this manuscript to the late Dr. James Irvine, who was a mentor to many of us, and an inspiration to us all.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY 4. Phenotypic Evaluation 5. Core Collection 6. Disease Status C. Breeding for Sugar Yield 1. Parental Selection 2. Hybridization 3. Progeny Selection 4. Breeding Achievements D. Breeding for Biomass Production 1. Basis for Biomass Breeding 2. Biomass Breeding Background 3. Biomass Breeding Strategies 4. Looking Forward E. Breeding for Stress Tolerance 1. Terminology and Diversity of Stress Resistance Mechanisms 2. Physiological and Molecular Basis of Stress Resistance 3. Considerations for Breeding 4. Cost-Benefit Analysis of Stress Resistance 5. Concluding Remarks and Future Directions III. SUGARCANE IMPROVEMENT THROUGH BIOTECHNOLOGY A. Physiology of Sucrose Accumulation 1. Background 2. Plant Systems Biology 3. Crop Factors and Yields 4. Sucrose Accumulation Processes 5. Global Analysis of Sucrose Accumulation 6. Improving Physiological Traits B. Genetics and Genomics 1. Genetic Diversity 2. Genetic Mapping 3. Mapping Quantitative Trait Loci for Economic Traits 4. Synteny with Other Members of the Grass Family 5. Map-Based Cloning of a Rust Resistance Gene 6. Application of Sugarcane ESTs C. Molecular Cytogenetics 1. Determination of Basic Chromosomes Numbers 2. Origin of S. barberi and S. sinense 3. Genome Structure of Modern Cultivars 4. Related Genera 5. Chromosome Pairing D. Genetic Engineering 1. Sugarcane: a Suitable Crop for Genetic Engineering 2. Initial Genetic Transformation Technologies 3. Recent Advances in Sugarcane Genetic Engineering 4. Development of Commercially Useful Transgenic Sugarcane 5. Transgene Silencing 6. Sugarcane Biofactory Research E. Coordinating International Progress in Sugarcane Improvement, and Linking Biotechnology to Application: the ICSB LITERATURE CITED

17

18

MING ET AL.

I. INTRODUCTION Sugarcane (Saccharum spp., Poaceae) is a large, perennial, tropical or subtropical grass widely grown in a zone around the world within 30° of the equator. It is usually vegetatively propagated from axillary buds on stem (or stalk) cuttings. The first, “plant,” crop is generally harvested from 12 to 24 months after planting; thereafter, “ratoon” crops may be harvested at shorter to equal time periods. Ratoon crops may be grown from one to several cycles. The large, mature stalks contain juice of 9 to 18% sucrose. The juice is extracted by crushing the stalks with highpressure rollers in a mill. Sucrose is crystallized from the juice after water is removed by boiling to produce a brown-colored raw sugar. White sugar is produced by re-crystallization from raw sugar in a refinery. About 75% of the world’s sugar (sucrose) supply is from sugarcane and the other 25% from sugar beet (Beta vulgaris L., Chenopodiaceae). Sugarcane also is used for ethanol and biomass production as an alternative source of energy. More than 80 countries grow sugarcane, producing 111.8 million metric tones (t) of sugar for 2002/2003 (F. O. Licht 2003). Three countries, Brazil, India, and China, each produced more than 10 million t of cane sugar in the year 2002/2003. Other major cane sugar producing countries (each ≥ 2 million t) are Australia, Mexico, Thailand, Pakistan, USA, South Africa, Colombia, Cuba, and Philippines. The highest annual sugar yields recorded were 23.8 t/ha-yr in Leeward Oahu, Hawaii, and 17.4 t/ha-yr in the Central District, Australia (Heinz 1987b; Osgood 2003). Six species are often recognized in this genus based on classical taxonomic classification. An alternate view on classification of this and associated genera (Erianthus Michx. and Narenga Bor) has been proposed by Clayton and Renvoize (1986) on the basic characters that are more appropriate to infra-generic categories. Saccharum spontaneum L. is the most primitive species, with its center of origin and diversity in India and with a broad distribution in tropical and subtropical regions. It typically has pencil-thin stalks and very low sucrose content and is widely used in sugarcane breeding programs for its resistance to diseases, insects, and abiotic stresses. Saccharum robustum Brandes and Jewiet ex Grassl, the other wild species with its center of diversity in New Guinea, has small to large stalks and low sucrose content. Saccharum officinarum L. is the primary sugar producing species and is found only in cultivation. It has thick stalks with high sucrose content. Saccharum barberi Jeswiet has been cultivated in India and Saccharum sinense Roxb. in China since pre-historic times. These two species are often grouped together as a single species with thin to medium stalks,

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

19

low to moderate sucrose content, and higher fiber and greater tolerance to stress as compared to S. officinarum. Saccharum edule Hassk. has an aborted and edible inflorescence and is cultivated from New Guinea to Fiji as a vegetable. Three species were used for sugar production in the past—S. officinarum in New Guinea, S. barberi in India, and S. sinense in China. Modern sugarcane cultivars are Saccharum spp. hybrids mostly containing 90% S. officinarum and 10% S. spontaneum. Sugarcane may have evolved from a primitive form of S. spontaneum in the foothills of Himalaya in northern India (Stevenson 1965). Subsequent selection, movement, and introgression resulted in two centers of diversity—S. spontaneum in India and S. robustum in New Guinea. S. robustum is possibly derived from hybridization between S. spontaneum and other genera in Wallacea/New Guinea (Daniels and Roach 1987). The species S. officinarum, with high sucrose content, is believed to have been derived from S. robustum in New Guinea. Recent chromosome in situ hybridization revealed that S. barberi and S. sinense are derived from interspecific hybridization between S. officinarum and S. spontaneum (D’Hont et al. 2002). S. edule is thought to be an intergeneric hybrid between either S. officinarum or S. robustum and other genera (Daniels and Roach 1987). It appears that the five Saccharum species besides S. spontaneum may each contain more than 50% of the S. officinarum genome, due to the high chromosome number of S. officinarum and the 2n + n transmission in S. officinarum × S. spontaneum crosses, a phenomenon known as female restitution (Bremer 1923; Price 1957). This led to a suggestion to group the genus Saccharum into two species: S. spontaneum by itself, and S. officinarum including the other five species and their hybrid derivatives (Irvine 1999). Sugarcane has been cultivated and improved over thousands of years, beginning in prehistoric times with selection on natural variations and continuing to the current techniques of hybridization and genetic engineering. Enormous yield increase has been achieved in the last century by breeding for yield, disease and insect resistance, and stress tolerance. While sugarcane farmers throughout the world face constant challenges to sustain profitability and protect the environment (Glaz 2003), breeders face not only those challenges but also a biological constraint as the gap between average farm yield and genetic yield potential is narrowed through improved agronomic practices (Cassman 1999). Recently developed genomic resources and acquired molecular tools in sugarcane have the potential to further improve sugar and biomass production, but these must be used as tools in support of what is known as “traditional” crop improvement. Genetic transformation of sugarcane into a bio-factory for producing high value proteins shows promise.

20

MING ET AL.

This review is intended to summarize recent work on sugarcane crop improvement through conventional breeding and the rapid advancement in molecular biology and biotechnology in sugarcane that has occurred since publication of the landmark book Sugarcane Improvement through Breeding (Heinz 1987a). Considerable progress has been made since then in breeding for sugar yield, biomass production, and stress tolerance. A new field of sugarcane genomics has been established. New findings in molecular cytogenetics have clarified some century-old questions. A large sugarcane expressed sequence tags (EST) database is accessible to all researchers. A new era for sugarcane improvement has started with a new set of tools available to sugarcane breeders.

II. SUGARCANE BREEDING A. History The cultivated sugarcanes of today are mainly complex interspecific hybrids primarily between Saccharum officinarum, known as the noble cane, and Saccharum spontaneum with contributions from S. robustum, S. sinense, S. barberi, and related grass genera such as Miscanthus, Narenga, and Erianthus (Daniels and Roach 1987; Sreenivasan et al. 1987; Irvine 1999). For example, one of the Hawaiian cultivars, H371933, has in its ancestry S. officinarum (through cultivars ‘Lahaina’, ‘Cheribon’, ‘Yellow Caledonia’, ‘Badila’, and others); S. barberi (through POJ 213); S. spontaneum (through Kassoer); S. robustum (through Mol. 1231); and S. sinense (through Loethers) (Edgerton 1958; Tew 1987). S. sinense and S. barberi have been cultivated for sugar production in China and India for centuries (Blume 1985; Heinz et al. 1994). Following the rediscovery of S. officinarum in the 18th century, the sugar industry rapidly spread throughout the tropics and subtropical areas (Heinz et al. 1994). The widespread appearance of new sugarcane diseases caused great damage to the noble varieties and led to the search for new noble canes. The discovery of sexual fertility in sugarcane in the 19th century in the Bahamas and Java opened the door for establishing breeding programs (Stevenson 1965; Blackburn 1983). 1. Origin and Domestication The Origin of Sugarcane. The proposed origin of S. officinarum from a domesticated thick-stalked, high sugar, low fiber form of S. robustum in New Guinea is accepted by most sugarcane breeders (Daniels and Roach 1987). From New Guinea, S. officinarum spread to Indonesia, Malay,

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

21

China, India, Micronesia, and Polynesia during prehistoric times. An officer of Alexander the Great was the first to mention sugarcane in India, in 325 B.C. Its distribution from Polynesia to Hawaii took place with native migrations around 500–1000 A.D. and from Indonesia to southern Arabia and East Africa probably before 500 A.D. The Dutch breeders in Java called S. officinarum the “noble” cane, and “nobilization” the process of backcrossing of S. spontaneum hybrids to S. officinarum. Ancient Cultivars. S. sinense cultivated in China and Pansahi India was used for chewing as well as for sugar production, whereas the thinner, harder stalks of S. barberi cultivated in northern India were used only for crushing. The two cultivated sugarcanes were probably the result of natural hybrids of S. officinarum and S. spontaneum that occurred about 1000 B.C.E. S. barberi subsequently spread from India to the Middle East, Mediterranean, and to the New World beginning with the second voyage of Columbus in 1493. The most important Indian cultivar had many names: Creole in French, Criola in Spanish, or Crioula in Portuguese. Today, S. sinense and S. barberi exist only in collections (Stevenson 1965; Blume 1985; Rossi et al. 1987; Heinz et al. 1994). Cultivated Noble Canes for Sugar Production. Creole was quickly replaced in cultivation by the noble cultivar ‘Otaheite’ when it was brought to Jamaica from Tahiti by Bligh in 1793 (Rossi et al. 1987). From there it was distributed throughout the Caribbean and the Americas. Original noble canes collected from the Pacific Islands were the only source of cultivars for plantations for the world’s sugar production for over a hundred years. Before sugarcane breeding programs were started, the most important noble cultivars were the ‘Otaheite’ (Bourbon, Lahaina) of Tahiti, ‘Cheribon’ (Louisiana Purple) of Java, and ‘Caledonia’ of New Hebrides. ‘Bourbon’ was extremely susceptible to root rot, mosaic, and gumming disease; ‘Cheribon’ to sereh, mosaic, and root rot; and ‘Caledonia’ to mosaic (Edgerton 1958; Stevenson 1965). These initial cultivars were replaced by new ones selected from the recently established breeding programs. Today, clones of S. officinarum are in breeding collections and/or cultivated as garden canes for chewing. 2. Early Breeding. The first sugarcane breeding program began in Java and Barbados in 1888, following the observations independently in Java (1858) and Barbados (1859) that sugarcane was capable of producing viable seed (Stevenson 1965; Kennedy and Rao 2000). Other countries soon followed Java’s lead. In this review, the history of breeding has been separated into five periods based on the use of the major breeding materials and the type of progeny produced. They are: (1) breeding of noble

22

MING ET AL.

canes to produce noble cultivars, (2) breeding through nobilization to produce nobilized cultivars, (3) breeding of nobilized canes to produce hybrid cultivars, (4) breeding hybrid canes to produce the current cultivars, and (5) breeding to broaden the genetic base. Breeding of Noble Canes to Produce Noble Cultivars. Progenies of openpollinated noble canes were selected for sugar production. Each selected seedling was assigned a call sign followed with a seedling number. ‘Otaheiti’ (‘Lahaina’, ‘Bourbon’) produced the EK seedlings in Java, ‘H109’ in Hawaii, and ‘B716’ in Barbados. In Queensland, ‘Q813’ came from Badila, the famous chewing cane, originating in New Guinea. These selected noble cultivars were important for sugar production in the early 1900s. The original noble canes and selected noble progenies were found susceptible to disease and insects and limited to particular tropical environments. Breeders soon realized that the genetic base of the noble canes needed to be broadened to improve their adaptabilities and disease and insect resistance (Stevenson 1965). Breeding through Nobilization to Produce Nobilized Cultivars. Nobilization is the pollination of noble cane with its wild relative S. spontaneum followed by repeated backcrosses to the noble canes. The wild relative is nobilized through the breeding process, and the selected hybrid progenies are the “nobilized canes” (Bremer 1961). The key event was the production of a nobilized cultivar, ‘POJ2878’, of Proefstation Oost, Java, in 1921 (Jeswiet 1929). The nobilization process involved chromosome non-reduction plus introgression of additional genes through a system of crossing the noble clones with wild clones of S. spontaneum (x = 8, 2n = 32-128) (Irvine 1999; Ha et al. 1999), having thin stalks, low sucrose content, and high fiber with disease resistance and cold tolerance. More than 90% of the accessions classified as S. officinarum have 2n = 80, x = 10 chromosomes, whereas the most frequent count of chromosome number in S. spontaneum is 2n = 64 (Irvine 1999). Using these two chromosome complements as an example, one can envision nobilization in the following simplified crossing scheme where NN = noble, 2n = 80; SS = spontaneum, 2n = 64. Female × male → progeny (chromosomes) (% spontaneum) NN(80) × SS(64) → F1 (80 + 32 = 112) (S% = 29 = 100 × 32/112) NN(80) × F1(112) → BC1 (80 + 56 = 136) (S% = 12) NN(80) × B1(136) → BC2 (40 + 68 = 108) (S% = 7) Progeny of F1 and BC1 have the non-reduced somatic complement (2n) of the female parents plus the gametic number (n) of the male. Most cul-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

23

tivated nobilized canes (BC2s, BC3s, etc.) have chromosome numbers in the range 100–130, with about 5–10% consisting of the wild S. spontaneum contribution. Clones with chromosome numbers outside of this range are rarely suited for commercial production. In Coimbatore, India, nobilization of S. barberi and S. spontaneum with S. officinarum produced the famous early nobilized tri-species hybrids of “Co” seedlings. Cultivars ‘Co 213’, ‘Co 281’, and ‘Co 290’ gained wide acceptance in subtropical regions in India, South African, Australia, Louisiana, Argentina, and Brazil. The “Co” cultivars also were used on the poorer soils and under marginal growth conditions in the tropics. Nobilization breeding also was practiced in Barbados and other breeding stations to produce many important “B” and other “call sign” series cultivars. But after 1925–1930, nobilization breeding was seldom used (Stevenson 1965; Simmonds 1976; Ethirajan 1987). Breeding of Nobilized Canes to Produce Hybrid Cultivars. Crosses among nobilized canes in the 1930s produced many important hybrid cultivars for sugar production in the next three decades. Breeding of ‘POJ2878’ with other nobilized POJ canes produced cultivars ‘POJ3016’ and ‘POJ3067’. Together they occupied more that 85% of the cane area of Java in 1960. Crossing of ‘Co312’ and ‘POJ2978’ produced Hawaii’s most important cultivar, ‘H32-8560’, which was responsible for 65% of the cane area of Hawaii in 1945 (Fig. 2.1). ‘POJ2878’ × ‘Co290’ produced ‘Co419’ for the tropical area of India. Breeding of nobilized canes in Barbados in 1937 produced ‘B37161’ and ‘B37172’, both of which were important in the 1950s (Stevenson 1965). The cross of ‘Co421’ × ‘Co312’ was made in Coimbatore in 1938 to produce progeny of the cross that was grown in Natal, South Africa, in the same year. One of the progeny selected in 1939, assigned the number ‘NCo310’, became the most important cultivar of the world in the 1950s and 1960s (Anonymous 1945; Nuss and Brett 1995). Even as late as the 1980s, ‘NCo310’ still ranked fourth in growing area in Ecuador, second in Gabon, first in Japan, second in Malawi, third in Mexico, first in Texas, USA, and first in Uruguay (Tew 1987). Breeding of Hybrid Canes to Produce Current Cultivars. Commercial cultivars and hybrid canes selected from advanced selection stages have been the main breeding materials for the development of current cultivars since the 1950s. The names of the current cultivars, rank in percent of area occupied, immediate parents, and breeding stations of the world are listed in the Sugar Cane Variety Notes (Rossi 2001), the Sugarcane Varieties (Rossi 2002), and the World Sugarcane Variety Census—Year 2000 (Tew 2003).

LEADING HAWAII CULTIVARS and TONNES 96 DA SUGAR PER HECTARE PER YEAR (TSHY) (1915-2003)

70.0 65.0 60.0 55.0 50.0 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 1915

H109

Hybrid cultivars

H32-8560

16.00

H37-1933

14.00

H44-3098 H50-7209

12.00

H59-3775

10.00 8.00 6.00

TSHY

% ACREAGE

Noble cultivars

Lahaina Y. Caledonia

H65-7052 H73-6110 H74-4527

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

4.00

H77-4643

2.00

H78-4153

0.00

H78-3567

2000

YEAR Fig. 2.1.

H62-4671

Sugar yield of leading Hawaii cultivars in tonnes 96 DA sugar per hectare per year (TSHY) from 1915 to 2003.

H78-7750 TSHY

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

25

Breeding to Broaden the Genetic Base. Modern cultivars are essentially derivatives of no more than 15–20 genotypes of nobilized cultivars that can be traced back to the initial nobilized genetic base developed in Java and India (Roach 1989). The genetic base for cane and sugar yield in the advanced breeding populations of today is expected to be somewhat narrower than that in the initial germplasm following more than 100 years of directional selection within the original nobilized germplasm (Walker 1987). Attempts to increase this narrow base, called the base broadening program (BB-program) were started in Barbados in 1965 using clones different from those initially used in Java and India. The BB-program has been run in parallel to the Barbados contemporary cultivars production program. The BB-program started with nobilization crosses followed with hybridization of nobilized canes. The BB-program has produced many semi-commercial type clones since the late 1980s. In recent years, the gene pool of the semi-commercial BB-clones is being incorporated into the gene pool of advanced breeding populations (Kennedy and Rao 2000). Several other countries have tried BB-programs over the past 50 years by crossing wild canes with their commercial cultivars. However, none of these efforts was as long term and broad based as the BB-program of Barbados and Louisiana. In Barbados, cultivar ‘B79474’ is a progeny of ‘B67-106’ × ‘B71852’, which is a progeny of a second nobilization of ‘SES49’. ‘B79474’ is clearly not a hybrid of BC 2 generation. It is a hybrid of the third generation of introgression cross (I3). Cultivar ‘B80251’ is a progeny of ‘B74634’, which is a progeny of both S. spontaneum ‘Moentai’ and ‘SES84/58’ (Rossi 2001). In Louisiana, cultivar ‘LCP85-384’ was harvested from 88% of the state acreage in 2003 (Bischoff 2004). This cultivar is a I4 derivative from S. spontaneum ‘US56-15-8’, and ‘Ho95988’, a product of the USDA-ARS basic breeding program that has been released for commercial planting in 2004 (Tew 2004). In Australia, ‘QN66-2008’, an I3 progeny from a cross of ‘POJ2878’ × ‘Mandalay’ (S. spontaneum), is an important parent that has produced 22 ‘Q’ cultivars in the BSES program. Some of them, ‘Q138’, ‘Q154’, and ‘Q158’, are acquiring major cultivar status in the industry in 2004 (N. Berding, pers. comm.). In Argentina, ‘TUC(CP)77-42’, the leading cultivar in 1999 (Tew 2003), is a progeny of ‘CP71-321’ × ‘US72-19’, which is a progeny of ‘CP64-313’ × ‘SES 147B’ (S. spontaneum). This is a good example that both BC1 and I2 are proper descriptions for this cultivar ‘TUC(CP)77-42’. In Hawaii, ‘H59-3775’, a leading cultivar from 1975 to 1985 and responsible for 50% of the Hawaii cane area in 1980, has 1/32 of its germplasm from S. robustum, ‘Mol 1242’ (Fig. 2.1).

26

MING ET AL.

3. Prospects. Our inability to trace or follow the incorporated germplasm into the germplasm of the advanced breeding population through visual selection is perhaps the main reason for the failure of base broadening programs. Large favorable genetic variation exists among clones of Saccharum species (Tai and Miller 2002). What is missing is a breeding tool to assist breeders in incorporating useful genes from any source into the gene pool of the advanced cultivars. Recent developments in biotechnology are beginning to yield information and technologies that undoubtedly will help the breeders to broaden the gene pool of their advanced breeding populations and produce higher yielding cultivars in the future. B. Genetic Resources 1. Collections. Early in the last century, sugarcane geneticists realized that a large diverse germplasm collection was essential for sustained crop improvement. At least 31 separate collecting expeditions across the complete natural distribution of the species were made from 1892 through 1985. The objectives of these expeditions were to collect genotypes that were resistant to diseases, were highly productive, or had high sugar content (Berding and Roach 1987). Clones from these collections have been deposited in the two replicate world collections, one maintained in India and one in the United States. These collections serve as genetic reservoirs to be mined and used in breeding new cultivars for specific agronomic needs, for pest and disease resistance, and to broaden the genetic base of commercially grown cultivars. Traditional sugarcane taxonomy divides the genus Saccharum into six species. Disagreement exists concerning the taxonomy and, according to Irvine (1999), only two species are valid, S. officinarum and S. spontaneum, the other species being interspecific hybrids between these two. The separation of clones into the various groups or species provides a useful method of classification for the management of the germplasm collection. The number of accessions for each species in the World Collection of Sugarcane and Related Grasses located at Miami, Florida are listed in Table 2.1. The sugarcane germplasm collection is maintained in a four-hectare block on the USDA-ARS Subtropical Horticulture Research Station (SHRS) that has drip irrigation/fertigation and overhead sprinklers for cold protection. The collection requires approximately two hectares of growing area and is planted in 32 rows, each with 31 plots. Plots are 2 m long with 3 m between plots, and five stalks are cut and planted for rotation of the germplasm. The collection is rotated from east to west and

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY Table 2.1.

27

Sugarcane accessions contained in the collection in Miami.

Species Saccharum officinarum S. barberi S. sinense S. robustum S. spontaneum Commercial hybrids S. spp.z Erianthus Narenga Miscanthus Total

Number of accessions 397 58 42 85 348 193 229 23 1 18 1394

z

Includes new introductions not yet classified and unknown clones with labels lost in Hurricane Andrew.

back every three years. The old block is retained over the first winter to ensure successful establishment of the new planting. Approximately 360 (25%) are accessions of S. spontaneum. These accessions must be maintained on trellis rows with concrete sidewalks to prevent rampant spreading because of the strongly rhizomatous nature of the species and to control flowering. The plants are maintained in 45 L pots tied to the trellis rows to prevent wind from blowing over the plants (Fig. 2.2). The S. spontaneum accessions are cut to the base of the pot twice a year because of their vigorous growth habit and to prevent flowering that is induced by decreasing day length. The seeds of S. spontaneum are easily dispersed by wind and have the potential to become a noxious weed if transported to the nearby Everglades National Park. 2. Passport and Descriptor Information. Passport data, providing the basic documentation, including taxonomic designation and information about where an accession was collected, are available for most of these clones. Descriptor information is complete and the data are available from the Germplasm Resources Information Network (GRIN) database maintained by the National Plant Germplasm System (NPGS) of the USDA. The descriptor information for sugarcane can be found at http://www.ars-grin.gov/cgi-bin/npgs/html/crop.pl?101. The 102 descriptors are classified into eight categories. The two most important categories contain descriptors for diseases and morphology. The disease category contains descriptors for 19 diseases, while 69 descriptors relate

28

Fig. 2.2.

MING ET AL.

Trellis rows used for the maintenance of S. spontaneum.

to morphology. These data are useful for classification of accessions, but they are subject to environmental influences and most have a significant genotype × environment interaction.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

29

3. Production of True Seed. In the late 1980s, the USDA-ARS supported a project for the production of true seed of S. officinarum. Over 580 clones of S. officinarum were imported into Hawaii for the production of true seed at the Maunawili Breeding Station of the Hawaiian Sugar Planters’ Association (HSPA). This location was chosen because it provides the most ideal conditions for S. officinarum flowering found in the USA. Seed was produced from 1980 through 1993 using a polycross design. Only 215 clones flowered, the remaining 371 clones did not. Of these 215 clones that flowered, sufficient quantities of seed were obtained from 194, and this seed is stored at the National Center for Genetic Resource Preservation (NCGRP) at Fort Collins, Colorado. In addition, 47 clones of S. spontaneum were successfully polycrossed and the seed sent to the NCGRP. Seed was sent to NCGRP based on a germination test with a minimum of 100 seed germinating per gram of fuzz and a minimum of 10 g of fuzz. In 1994 the USDA-ARS supported a similar project to produce true seed of S. spontaneum at the Sugarcane Breeding Station (SBS) at Canal Point, Florida. True seed resulted mainly from self-pollination of 246 S. spontaneum accessions. Germination tests ensured that over 1,000 seeds from each accession were deposited at the NCGRP. These true seed samples represent nearly all major regions of the geographical distribution of this species. In 2003, a new project was funded to produce true seed of S. sinense, S. barberi, and S. robustum. This project is currently underway at the SBS at Canal Point and will require three years to complete. By 2007 all of the aforementioned species should have a significant representation of their genetic diversity stored at the NCGRP. 4. Phenotypic Evaluation. The evaluation of germplasm held in collections is a high priority for the NPGS. This information facilitates the use of germplasm and makes more efficient management possible. Evaluation is a lengthy and costly process of examining accessions for traits of significance; however, it does add tremendous value to germplasm collections. Clones of S. sinense, S. barberi, and S. robustum were evaluated for agronomic and quality characters and to estimate the genetic diversity within and between the populations. Thirty clones of each species were evaluated in two environments for eight phenotypic characteristics. Significant differences were found between the three species as well as for clones within species. The genetic repeatability for every character, except stalk number, was high, indicating that this information would be useful for breeders interested in using the material in commercial

30

MING ET AL.

crosses (Brown et al. 2002). Additional phenotypic evaluation is in progress to estimate the tolerance to environmental stresses. 5. Core Collection. Limited genetic variation exists among the S. officinarum, restricting the usefulness of a core collection for this group. With the exception of S. spontaneum all other species have a limited number of accessions, and the entire collection can be considered as core. Tai and Miller (2001) evaluated 11 methods for the production of a core collection of the S. spontaneum clones using 11 quantitative traits and limiting the number in the core to 75. The first nine methods involved a cluster analysis followed by various selection strategies for accessions within each cluster. The last two methods involved random selection over the entire collection and random selection based on geographic region. Although the authors did not suggest that any one core was the best, they did list the members of the core based on cluster analysis within each geographic region based on retained principal components for morphological traits and random selection of entries within each cluster (Method 9) (Tai and Miller 2001). 6. Disease Status. The presence of two systemic pathogens, sugarcane yellow leaf virus (SCYLV) and Leifsonia xyli subsp. xyli (Lxx), was determined in sugarcane clones that are maintained in field plots in the World Collection of Sugarcane and Related Grasses. Both pathogens have a wide distribution in the world and have been shown to cause yield losses. Most of the accessions in the collection were assayed for the presence of SCYLV in 2003. Five leaves of each clone from different sections and/or plants of the 2 m plot were sampled and assayed using a tissue blot immunoassay to detect SCYLV (Schenck et al. 1997). The incidence of SCYLV infection differed among Saccharum spp. S. officinarum had the highest incidence of infection, 78.2% of 325 clones sampled, followed by 60.0% of 55 S. robustum clones sampled, 45.9% of 37 S. sinense clones sampled, 12.7% of 55 S. barberi clones sampled, and 5.6% for 252 S. spontaneum clones sampled. Infected clones were assumed to be susceptible to the virus. Uninfected clones were assumed either to be resistant to infection or to have escaped being infected. The status of ratoon stunt, caused by Lxx, was determined for part of the collection in 2004. Five stalks were sampled from clones of S. officinarum, S. robustum, S. sinense, and S. barberi taken from different locations of each two meter plot, the presence of Lxx was determined using a tissue blot immunoassay specific for the pathogen (Comstock et al. 2001; Harrison et al. 1988). The test was performed on stalks from the

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

31

basal section, 1 cm in diameter. Clones of S. spontaneum have not been tested at this time since most have stalk diameters of less than one cm and will require a different methodology. The incidence of Lxx infection varied among Saccharum spp. The highest incidence of infection was in the S. sinense clones, in which 62.5% of 40 clones tested were infected. The incidence of infection was 42.6% of 54 S. barberi clones, 14.6% of 302 S. officinarum clones, and 17.2% of 58 S. robustum clones tested. Clones that tested positive for Lxx are assumed to be susceptible. Clones that tested negative may have either escaped infection or are resistant to the pathogen. The presence of these pathogens indicates that material shipped from the collection may be infected with these pathogens. This highlights the requirement of adequate therapeutic treatments and quarantine of sugarcane leaving the collection. C. Breeding for Sugar Yield There are approximately 23 sugarcane breeding stations in the world (Rossi 2002). Most maintain a large number of clones selected from local breeding programs, clones imported from other stations, and clones of basic species imported from world germplasm collections. Australia and Barbados make substantial use of basic species in their programs, whereas other stations use less than 10%. Most of the stations use foreign parents (60–68%), whereas Barbados, India, South Africa, and the USA use mostly their locally bred materials (90–100%) (Ramdoyal et al. 2003). Breeders have developed individualized crossing and selection procedures to produce elite clones for both parental selection and selection of cultivars that are adapted to local environments, are resistant to local disease and insects, increase yield potential under favorable conditions, and are suited to continually changing agronomic and management practices (Berding and Skinner 1987). Controlled bi-parental crosses (70–100%) are more popular in most of the breeding stations, but polycrosses also are made as an economic and efficient means to produce seeds at a number of stations. The percentage of polycrosses made in Brazil, Indonesia, and Mauritius ranges from 40 to 46%, in Barbados and Colombia it is 73 to 89%, and in Hawaii almost 100% (Ramdoyal et al. 2003). Traditional sugarcane breeding methods consist of three steps: (1) parental selection from a source population, (2) hybridization using biparental crosses and poly-crosses, and (3) progeny selection at several stages during clonal propagation. Selected clones will be grown in the

32

MING ET AL.

breeding station as new parent sources, whereas parents used substantially over many crossing seasons or having poor progeny performance records will become inactive or be discarded (Berding and Skinner 1987). Many source populations exist for parental selection for various crossing programs, such as source populations for base-broadening (Walker 1987), for experimental crosses, for high-quality recurrent selection (Breaux 1987), for specific environment and stress selection (Lo 1987; Moore 1987), and for cultivar selection (Skinner et al. 1987). These source populations overlap with each other, with some clones occurring in each. For each crossing program, the parental selection criteria vary depending on the goal and selection criteria of the crossing program. Progenies produced from all source populations are to be selected for backcrossing and/or for development as a cultivar. 1. Parental Selection. Selection of parents from a source population for crossing is based on the performance data of each parent and its progeny. Locally bred sources are initially selected from the advanced clonal stages of selection. In Hawaii, a clone that had a sugar yield record equal to or better than the commercial check in the same test would be selected for the source population. Other breeding stations may use different selection procedures. Locally bred and foreign clones new in the source population are allowed to produce small numbers of progeny for evaluation. Most breeding stations evaluate progeny performance on the basis of a selection rate. If the progeny performs better than a standard, the progeny’s parent will be identified as a proven parent and the cross that produced the progeny a proven cross. To repeat proven crosses, proven parents will be crossed again in the following crossing seasons to produce a large number of progenies for selection (Breaux 1987; Hogarth and Skinner 1987; Berding and Skinner 1987; Heinz and Tew 1987). Parental selection for bi-parental crossing, which is the proven cross system used in Australia, is based not only on selection rates from original seedlings but also on performance of clones from each family at subsequent stages of selection. A data bank involving the most recent 10 years of original seedling plantings is a part of the system (Hogarth and Skinner 1987; Skinner et al. 1987). The parental selection for polycrosses in Hawaii is based on both accumulated yield records of the parents and accumulated selection rates from the original seedlings over many years. The number of tassels used in a polycrossing area is set by the proven status of the selected parent. During a crossing day, a new parent will have four tassels, while a best proven parent will have 16 tassels, randomly arranged in the cross-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

33

ing area. The source population has many more new parents than best proven parents. Therefore, at the end of a crossing season, a proven parent will produce a large number of seeds, enough for both seed germination and seed storage. The large number of seedlings from seed germination is used for seedling selection. If the selection rate is high, the stored seeds of the same parent will be germinated again in the following season (Heinz and Tew 1987). Most countries favor the proven bi-parental system (Ramdoyal et al. 2003). Many breeding stations use their own criteria for parental selection and have made considerable progress in breeding for sugar yield (Breaux 1987). 2. Hybridization. Crossing procedures and techniques vary among breeding stations. Basic pollination procedures consist of harvesting tasseling stalks from field plots as flower anthesis begins, then moving the harvested stalks to a crossing shelter where they are placed in a weak acid solution for prolonging flower life to enable making either biparental crosses or polycrosses (Verret 1925). More than two-thirds of the breeding stations of the world use excised marcotted stalks alone or in combination with the crossing solution (Ramdoyal et al. 2003). In a bi-parental crossing shelter, male parents are discarded after 14 days and female parents are retained for an additional 10 to 20 days until the seed ripens. In a polycrossing shelter, tassels (panicles) with seed are harvested from 21 days to 35 days. The acid solutions must be changed twice a week to keep the flowering stalks alive. Flowering stalks are shaken each morning at about 08:00, shortly after pollen is shed, to maximize seed set. Seeds are germinated in a greenhouse and seedlings are transplanted to fields two to three months after germination for selection (James 1980). 3. Progeny Selection. Procedures for selecting cultivars vary among breeding stations, depending primarily on the crop cycle length and number of ratoon cycles practiced by the local cane growers (Mamet and Domainque 1999). For a long cropping cycle (24 months plant and 24 months ratoon), there are three stages (8 months each) based on visual selection followed by two stages (24 months each) based on highly replicated yield trials in various ecological regions. Selection criteria for long cycles include disease reactions and insect resistance, high sucrose, high cane tonnage with second year suckers, and intermediate fiber (12–15%). For short cropping cycles (18 months plant and 12 month ratoon), several ratoon cycles are evaluated in addition to visual and yield trials of the plant crop. Selection criteria for short cycles are mainly

34

MING ET AL.

disease reactions and insect resistance, high sucrose, low fiber (10–13%), and erect growth with few suckers. A commercial cultivar may take 12 to 15 years to identify, whereas a new breeding parent may take 7 to 10 years to select (Heinz 1987b). 4. Breeding Achievements. Crop yield is a result of genetic improvements and management practices (Cook 2001). Changing of cultivars was a significant factor contributing to Hawaii’s sugar yield from 1915 through 2003 (Fig. 2.1) (Wu and Arcinas 2004). The decreasing trend of sugar yield over the last 20 years was mainly due to the closure of plantations and introduction of new diseases and pests. Annual production at the end of 2003 was 14.73 t/ha. The rate of increase in yield appears to have slowed over the past 50 years, especially in the 1970s and 1980s, partly because sugarcane was closer to the perceived maximum theoretical yield than any other crop (Heinz 1987a). In Australia, sugar yield was static for 20 years from 1970 to 1989. The static trend was likely due to production factors, such as increase in area per grower, industry expansion onto marginal land, and increased losses due to diseases and pests. Sugar yield increased to 12.0 t/ha-yr between 1990 and 1995. The increase is believed largely due to increased breeding efforts over the whole industry (Berding et al. 1997). Sugar yield in 2001 (estimated from BSES, 2001–2002 annual report) was 12.6 t/ha-yr. In Colombia, sugar yield increased from approximately 5 t/ha-yr at the end of the 1950s to 8 t/hayr at the end of the 1970s and reached 12 t/ha-yr at the end of 1999 (Cook 2001). Thus, static yields or yield plateaus experienced by several countries may be overcome by intensive breeding for new cultivars accompanied by development of appropriate cultural practices. D. Breeding for Biomass Production 1. Basis for Biomass Breeding. Biomass supplies about one-seventh of current global primary energy (Kheshgi et al. 2000). The potential for expanded use of biomass energy appears to be excellent and sustainable over the long term, notwithstanding short-term fluctuations in interest dictated by economic and political considerations. Several important realities contribute to this long-term view. First, there is the growing desire on the part of most nations to have a dependable, renewable energy source of internal origin. Second, there is the growing worldwide concern over CO2 emissions produced from fossil fuels and the desire to have energy sources that would lead to a lowering of global CO2 emissions. Third, with advances in biotechnology, there is the growing recog-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

35

nition that highly energy-efficient plants can provide a platform for the production of a vast array of high-value products, in addition to the production of energy, per se. Sugarcane has long been recognized as one of the world’s most efficient crops in converting solar energy into chemical energy. Like its tropical grass relatives maize (Zea mays L.) and sorghum (Sorghum bicolor L.), sugarcane is a C4 plant. In C4 plants, CO2 is initially added to a 3carbon acid to form a 4-carbon acid that is then transported to a region of the leaf where ribulosebisphosphate carboxylase is located. Here, carboxylation is reversed so that the local CO2 concentration is enhanced, and photorespiration is dramatically decreased. The theoretical maximum net efficiency of the photosynthetic process of converting solar energy into biomass in C4 plants is estimated to be 6 to 7%; by comparison, photorespiration in C3 plants leads to maximum efficiency of around 3%. Photosynthetic efficiencies can be translated to biomass yields. At maximum efficiency, the theoretical upper limit for sugarcane biomass (total solids) production is estimated to be 281 t/hayr (Loomis and Williams 1963). Of course, maximal efficiencies are not even closely approached in the real world for an array of reasons, such as incomplete canopy closure early in the crop cycle, overcast weather, sub-optimal ambient temperatures, sub-optimal water and nutrient conditions in the soil, and losses to diseases and insect pests. Sugarcane, though limited to tropical and subtropical climates, is a worldwide crop, commercially grown in 2004 in more than 80 nations. On a fresh weight basis, a larger mass of this crop is harvested and transported for processing than any other crop in the world, exceeding 1.3 billion tonnes (1,300 Mt) in 2003; sugarcane production in Brazil alone reached 386 Mt, from which 14.8 Mt of sugar and 15.3-million m3 of ethanol were produced (FAOSTAT Database: http://apps.fao.org). There are very few agronomic crops that rival sugarcane in energy conversion efficiency. In 2003, the average yield of sugarcane on a worldwide basis was 65 t/ha-yr, which translates to roughly 25 t total solids (sugar and fiber)/ha/yr. Highest average cane yields, in the range of 90–110 t/ha-yr, occurred in a few African nations (Burkina Faso, Chad, Egypt, Ethiopia, Malawi, Senegal, Swaziland, Tanzania, Zambia, and Zimbabwe), where sugarcane is intensively cultivated. Hawaii, which grows a two-year crop, normally produces about 200 t cane/ha, or 100 t cane/ha-yr, which translates to about 37 t total solids/ha/yr. The highest recorded yield for sugarcane on a commercial field scale of which we are aware occurred at Kekaha Sugar Company, Island of Kauai, in Hawaii, USA, in 1982. In one 60-ha field that had just been converted from furrow to drip irrigation and harvested at 24 months of age, the

36

MING ET AL.

fresh weight yield of the cane crop was 380 t/ha, total sugar was 47.5 t/ha, and estimated weight of total solids (soluble and insoluble) was about 140 t/ha or 70 t/ha-yr (Osgood 2003). Alexander (1985) argued, based on his experience in Puerto Rico, that biomass yields could be in the range of 2- to 3-fold that of present expectations, with a reorientation of cane management involving (1) utilization of biomass-oriented genotypes, (2) utilization of the whole plant including tops and leaves that have traditionally been burned just prior to harvest, and (3) growth orientation from planting to harvest. Heichel (1974) cited sugarcane, sweet sorghum, and field corn among those crops having the most favorable input/output ratios. For sugarcane grown in Hawaii in 1978, Tew (1980) reported an energy balance that showed a total annual energy output potential (including sugar, molasses, bagasse, and leafy trash) of 1,252 GJ/ha and an energy input (field and factory) of 377 GJ/ha, or greater than a 1:3 input/output ratio. During 1978, Hawaii’s sugar industry processed 8.1 Mt cane from 40,235 ha. From 2.6 Mt bagasse (1.3 Mt dry wt.) burned in boilers, 15.4 PJ (1015) of heat (384 GJ/ha at 60% efficiency) was produced. In 1978, sugar mills supplied 42%, 38%, 23%, and 2% of the total electricity generated on Hawaii’s four sugar islands of Hawaii, Kauai, Maui, and Oahu, respectively (Tew 1980). Macedo (2000) summarized energy utilization/production figures for the cane agro-industry in Brazil for the year 1985 and reported an overall input/output ratio of 1:9. Personnel at COPERSUCAR developed an online energy balance sheet for Brazil’s cane industry (http://www.mct .gov.br/clima/ingles/comunic_old/coperal5.htm), indicating a similar overall input/output ratio. 2. Biomass Breeding Background. Sugarcane breeding and selection has traditionally focused on traits that result in consistently high total sugar yield and sugar content at harvest and in minimal planting, crop maintenance, and harvesting costs. Milling considerations also come increasingly into play as cultivars reach the more advanced stages of evaluation. For example, fiber levels are required to be within a fairly narrow range. The mills need sufficient fiber to efficiently process cane and to have enough bagasse as a fuel source for their boilers to meet their own electrical needs and whatever electrical generation commitments they may have to the surrounding community. Too much fiber results in reduced milling efficiency. Penalties and incentives from the mill have tended to reinforce the stringent selection criteria that breeders have traditionally used, especially in terms of fiber and sucrose content. Formulas have been developed for the benefit of the breeder so that the most economical cultivars are advanced (Deren et al. 1995).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

37

With increased interest in biomass as a renewable energy resource and recognition that large tropical grasses are highly efficient converters of solar energy into chemical energy, some sugarcane breeding programs are investing a significant portion of their total effort toward biomass breeding. Contrasted with traditional sugarcane breeding and selection objectives, biomass breeding will more closely reflect the value of sugarcane and other energy cane candidates from the standpoint of their net caloric output potential. To most sugarcane breeders, biomass breeding closely parallels the involvement of wild species (S. spontaneum, S. robustum) and species indigenous to North India and China and cultivated from prehistoric times (S. barberi, S. sinense). The greatest quantum jump in the modern day genetic improvement of sugarcane occurred in the early 1900s when high-sucrose noble canes (S. officinarum) were hybridized with wild canes possessing host plant resistance to diseases in Java and India. Following their efforts to achieve interspecific crosses between Saccharum officinarum and S. spontaneum, early sugarcane breeders realized that resultant F1 progeny were distinctively more robust than either parent. When S. officinarum clones were used as the female parent, progeny tended to be larger stalked, higher in sucrose levels, and generally more vigorous than when S. spontaneum clones were used as the female parent. Reciprocal differences in vigor were eventually explained by the cytological phenomenon of a high frequency of “2n + n” progeny in S. officinarum (female) × S. spontaneum (male) crosses (Bremer 1923). For example, ‘Ashy Mauritius’ (2n = 80) × Indian S. spontaneum (2n = 64) produced ‘Co 206’ (2n = 112). Interspecific clones were not only more vigorous but were more widely adapted, more resistant to diseases, had greater ratooning ability, and produced progeny with greater genetic diversity than the noble canes from which they were derived. However, hybrids were unacceptable from the standpoint of low sugar and high fiber content, so early breeding efforts involved backcrossing using noble canes as the recurrent parent, until such universal clones as ‘POJ 2878’ (BC2 S. spontaneum) were developed. Success in involving S. spontaneum clones prior to 1930 generally satisfied most sugarcane breeders in terms of utilizing exotic germplasm to further improve sugarcane (Roach 1978). Utilization of wild germplasm within the Saccharum genus, particularly utilization of S. spontaneum to attain higher overall biomass yield potential, increased in the 1960s and became an integral part of some breeding programs, most notably in Australia (Berding and Roach 1987), Barbados (Walker 1972), India (Panje 1972), Taiwan (Shang et al. 1969), Louisiana (Dunckelman and Breaux 1972), and Hawaii (Heinz 1967).

38

MING ET AL.

3. Biomass Breeding Strategies. The objectives of biomass breeding should be clearly defined at the outset of any biomass-breeding program. Most biomass breeding programs to date have focused on maximizing both the sucrose and fiber potential of the crop by considering the nonsucrose fraction as an increasingly valuable product. A second possible focus would be the maximization of total soluble solids (Brix) while relaxing the standard for the sucrose fraction (purity) so that a larger group of candidates, including so-called “syrup canes” and sweet sorghum, could come under consideration. A third possible focus would be to maximize the insoluble solids fraction of the total energy package with the expectation that a harvestable product could remain in the field for an expanded period beyond killing frosts and/or natural senescence with minimal loss from deterioration. The third focus encompasses a far wider range of potential candidates extending beyond Saccharum to allied genera in the Saccharum complex, most notably Erianthus and Miscanthus. Large grass candidates, such as napiergrass (Pennisetum purpureum Schum.), giant reed (Arundo donax L.), and switchgrass (Panicum virgatum L.), are recognized to have considerable biomass potential, but are beyond the scope of this review. Focus 1: Maximizing Soluble and Insoluble Solids. Early-generation Saccharum hybrids involving S. spontaneum appear to have the greatest potential as energy cane candidates. When commercial sugarcane is crossed with S. spontaneum, maximum expression of vigor generally occurs in the F1 generation. Levels of vigor observed in this generation were extremely difficult to retain through successive generations of backcrossing (BC1, BC2, etc.), regardless of selection intensity for cane yield, as sugar and fiber percentages were brought to an acceptable standard (Roach 1978; Legendre and Burner 1996). Exploit of the vigor of early-generation hybrids involving S. spontaneum germplasm to develop biomass canes should be possible (Alexander 1985). A much greater effort to recombine clones in the F1 generation should result in F2 clones that display similar vigor and an increase in genetic diversity. This generation could form the foundation for subsequent intercrossing and limited backcrossing to achieve the specific objectives of a biomass-breeding program, the most important being maximizing net energy output. One of the keys to success in efforts to utilize exotic germplasm has been in not imposing the same selection criteria on wild-derivative clones as on clones originating from commercial × commercial crosses (Breaux and Dunkelman 1969). A biomass cane ideotype may be distinct from that of commercial sugarcane. For example, biomass clones are

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

39

likely to have a much higher population of slender stalks than typical sugarcane and perhaps have a rhizomatous habit that would confer extended ratooning capability. The rate of genetic gain may be more favorable for total biomass yield than for sugar yield. First, adoption of agronomic practices intended to optimize growth throughout the life cycle of the crop should result in a more rapid rate of genetic gain than would be expected under a cropping regime designed to limit growth as a means of improving sugar content. As a general rule, heritability for yield tends to be higher where growth is not restricted (Allen et al. 1978). Second, because stringent standards for sucrose and fiber levels are relaxed, a wider array of germplasm of potential benefit should be available to the breeder. Third, high-fiber cultivars are likely to remain more erect, so that the breeder can conduct selection closer to harvest time and thus be more effective. Selection in a biomass-breeding program requires an understanding of the contributions of all aboveground parts of the cane plant. Sugarcane biomass can be arbitrarily divided into three fractions, namely (1) the cane stalk free of tops and leaves, i.e., millable cane, (2) green immature cane tops and leaves removed from the cane during harvest, and (3) dead and dry leaves, known as trash. Alexander (1985) reported a relative weight fraction of cane in Puerto Rico of 69% stalk, 17% green immature cane tops and leaves, and 14% trash. Beeharry (1996) reported similar respective numbers of 69%, 21%, and 10% for cane grown in Mauritius. Mature cane stalks usually consist of about 25% total solids or 13% soluble solids (>90% sucrose), and about 12% fiber. Following the sugar extraction process, the fibrous residue (bagasse) contains about 50% fiber and 50% water. Green cane tops and leaves consist of 20–25% solids (largely fiber) and 75–80% water. Dry leaves consist of about 80% fiber and 20% water (Beeharry 1996). Recent first-ratoon data from an energy cane test in Louisiana, with candidate clones selected on the basis of exceptional total solids content, look encouraging (T. L. Tew, unpublished results). Sugarcane was partitioned into recoverable sugar (pol), additional soluble solids (Brixpol), bagasse dry weight, leafy trash dry weight, and water content. Insoluble solids content (bagasse and leafy trash) ranged from 19% to 38% among the energy cane candidates that were evaluated. Total solids content ranged from 33% to 47%. The best-performing clone, in terms of cane yield, fiber content, and total solids content was the only F1 hybrid included in the test, namely L 79-1002 (parents: ‘CP 52-68’ × Tainan S. spontaneum). Any biomass-breeding program must consider the energy (and economic) cost associated with transporting and later removing water from

40

MING ET AL.

the total cane mass. Candidates with the most favorable solids/water ratios are expected to have a competitive advantage over those candidates superior only in total solids content at harvest. Net energy output, rather than gross output, should be the ultimate selection criterion. Focus 2: Maximizing Soluble Solids. Sweet sorghum has the capability of producing high levels of total soluble solids in a relatively short period and may play a role where sugarcane cannot be grown year around. In Louisiana, the cane harvesting campaign is exceptionally short, only three months (October–December), due to the high risk of damaging freezes beyond December. Use of sorghum as a biofuel could extend the harvest period two additional months (August–September) and fit well within an integrated energy cane system. Sorghum cultivars and sorghum × sudangrass hybrids that do not flower in long daylengths (photoperiod sensitive gene) may have the greatest potential in this scenario because active growth could be extended to upward of 180 days, well beyond a crop life cycle traditionally defined by earliness of flowering. Focus 3: Maximizing Insoluble Solids. In an effort to minimize soluble solids and maximize insoluble solids, genera closely related to Saccharum may be of greatest interest to sugarcane breeders; these include Erianthus and Miscanthus. Both genera are within the so-called Saccharum complex, a term originally coined by Mukherjee (1957) and commonly used by sugarcane breeders to describe a small set of genera that are sufficiently closely related to Saccharum possibly to be involved in its phylogenetic background. A review of the taxonomy of Erianthus and Miscanthus by Daniels and Roach (1987) addressed the possible roles of both genera in the origin of sugarcane. The genus Miscanthus is widely distributed in eastern Asia from Indonesia to Japan, and through the sub-continent. Species most commonly placed within the genus include M. floridulus (Labill.) Warb. (2n = 38), M. sinensis Anderss. (2n = 38), and M. sacchariflorus (Maxim.) Benth. (2n = 76). Naturally occurring sterile triploid interspecific hybrids (2n = 57), designated as M. × giganteus GREEF et DEU, were identified in Japan and very likely involved the parents M. sacchariflorus and M. sinensis (Greef et al. 1997). Breeding efforts to improve Miscanthus and to achieve inter-generic crosses with Saccharum have occurred since the early 1900s, primarily in Taiwan and Japan. In 1975–1976, sugarcane breeders in Taiwan collected 181 accessions of Miscanthus indigenous to that island. Primary interest in collecting these accessions was to obtain germplasm that could confer greater cold tolerance and greater resistance to smut and downy mildew diseases to sugarcane (Lo et al. 1978).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

41

In 1935, an accession of M. × giganteus, collected in Yokohama, Japan, was planted in Denmark (Lewandowski et al. 2000). This clone, propagated by rhizomes, was found to grow vigorously and to produce a stubble crop following the winter season in Denmark’s temperate climate. Based on this and subsequent observations, M. × giganteus was eventually proposed as a biofuel in Europe. Extensive field trials have been performed in several European nations since 1983. Parent species, namely M. sacchariflorus and M. sinensis, have been identified with some traits superior to those of M. × giganteus, which indicate the large genetic diversity and potential of this genus in the development of breeding lines for energy production (Greef et al. 1997; Clifton-Brown et al. 2001). Of the two species, M. sinensis is more cold tolerant, hence is better adapted in more northerly climates such as Germany and Denmark; M. sacchariflorus genotypes are generally larger in stature and outperform M. sinensis in the more southerly climates such as those in the southern regions of Portugal and Spain. High biomass productivity potential of Miscanthus in a temperate climate does not appear to be limited to Europe. Heaton et al. (2004) reported that M. × giganteus compared favorably to switchgrass (Panicum virgatum) cultivars in yield trials in Illinois and speculated that Miscanthus may hold greater promise for biomass cropping than switchgrass under a wide range of environmental conditions. Similar to Miscanthus, Erianthus Michx sect. Ripidium Henrard is native to Southeastern Asia. Among the species most commonly reported to be within the genus, E. arundinaceus is considered to be the most robust and for this reason is perhaps the most interesting candidate from the standpoint of biomass production potential. Less is known about the biomass potential of Erianthus than Miscanthus. In a yield test conducted over two years at a former phosphatic clay settling pond site in Polk County, Florida, USA, Erianthus arundinaceum (Retz.) Jeswiet. accession IK 76-63 produced 148 t/ha dry matter, compared to yields of 42, 52, and 59 t/ha-yr for the highest-yielding genotype among the forage sorghum (Sorghum bicolor), elephantgrass (Pennisetum purpureum), and energy-cane/sugarcane (Saccharum spp.) candidates included in the test, respectively (Stricker et al. 1993). 4. Looking Forward. Breeders must take the long-term view and recognize the enormous potential of sugarcane and its relatives as an energy crop. As discussed elsewhere in this chapter, molecular breeders will be successful in inserting genes that will further widen the uses of sugarcane, particularly higher-value uses. While there remains an aversion to genetically modifying sugarcane as a common food source, genetic

42

MING ET AL.

engineering to produce pharmaceuticals, high-value proteins, etc., may not carry the same stigma. Concurrent sugar and energy cane breeding programs should be symbiotic, rather than adversarial. In the process of breeding more energyefficient biomass-oriented genotypes, the resulting germplasm should be usable, advantageously, in ongoing sugarcane breeding programs aimed at increasing sugar yield potential. E. Breeding for Stress Tolerance Most sugarcane production is from land prone to various environmental stresses that frequently limit yield potential (defined as the yield of a cultivar when grown in an environment to which it is adapted; with nonlimiting environmental conditions; Evans 1993). Increases in crop yields have traditionally been obtained through genetic manipulation, expansion of cultivated land, or modification of the crop environment (e.g., supplementary irrigation and fertilization) to alleviate the negative effects of stresses. The need for improved productivity coupled with limited resources for environmental modification and an availability of a finite amount of land for expansion point to genetic improvement as the major means of significantly increasing crop yields. Breeding for improved yield potential is the stated goal of most plant breeding programs. Nevertheless, because of the large impact that stresses have on crop yields (Boyer 1982), many plant breeders also target increased resistance to stress as a major route to crop improvement. One difficulty in breeding for stress tolerance per se is that genetic advancement is often evaluated in terms of yield performance, and in the case of sugarcane there may be tradeoffs between stress tolerance and yield. For instance, sugarcane is a long season crop whose primary product is sucrose stored in vegetative tissues of the stalk; sucrose yield is a function of stalk yield per unit area and sucrose concentration in the stalk (Moore and Maretzki 1996). Early in the season, stress resistance is essential for rapid stalk development, whereas “ripening” or accumulation of sucrose in the stalk late in the season is promoted by environmental stress. A stress resistance trait that results in rapid vegetative growth might be desirable early in the season but it would be undesirable during the ripening period. Another issue that arises in assessing the role of stress tolerance in achieving the high yield potential of improved cultivars is whether the higher yields are the product of concurrent selection for stress tolerance as well as yield potential, or simply a consequence of selection for higher yield potential alone (Bell et al. 1995; Evans and Fischer 1999).The bottom line seems to be that there is a role for considering stress tolerance in developing modern cultivars since the range of natural stresses occur-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

43

ring in field trials over time provides unrecognized selection pressures and thus an opportunity for plant expression of useful traits. 1. Terminology and Diversity of Stress Resistance Mechanisms. Biological stress has been defined as any external force (environmental pressure) that has the potential to cause aberrant changes in metabolic processes to eventually produce an injury (strain) to a living organism (Levitt 1980, 1982). For example, a water deficit (stress) reduces the rate of cell expansion and growth. Reduced cell expansion and growth constitute a strain in response to the water deficit stress. A plant’s ability to survive and reproduce in the presence of the stress constitutes resistance to the stress. The multiplicity of stress factors can be divided into two broad groups, namely, biotic and abiotic stresses. Biotic stresses are those caused by interaction with other living organisms, e.g., bacterial, fungal, and viral pathogens, insects, and animals. Abiotic stresses are the chemical and physical features of the environment that are capable of interfering with optimum plant function. Most plant species are likely to have evolved in environments with growing conditions that are periodically extreme, and genetic mechanisms probably evolved to confer stress tolerance. These traits and mechanisms have been broadly categorized as escape, avoidance, or tolerance. Some mechanisms may involve several traits, whereas others may comprise only a single trait (Nilsen and Orcutt 1996). Escape mechanisms enable a plant to complete its life cycle before the stress effect develops. For instance, early-maturing genotypes of short-season crops, such as cereals, escape periods of potentially damaging temperatures and water deficits by flowering and fruiting early before stress appears. Avoidance mechanisms enable a plant to exclude a stress or minimize its effect by means of a physical, metabolic, or developmental barrier. For example, plant leaf traits such as small canopy volumes, small leaves, thick cuticles, low stomatal density, and stomatal resistance can limit the rate of water loss and prevent the development of plant water deficits. Similarly, glaucousness, pubescence, and leaf rolling, all of which minimize radiation absorption, also can limit the rate of water loss and prevent development of drought stress. Tolerance mechanisms allow the plant to sustain physiological activity (albeit at reduced levels) without suffering irreparable damage. Metabolic shifts such as osmotic adjustment help maintain cell turgor and counteract the loss of water due to osmotic imbalance. Stresses such as heat and drought often trigger responses at subcellular and molecular levels, long before physiological and morphological symptoms occur. For instance, one of the early responses to nearly all types of stress is generation of potentially lethal free radicals

44

MING ET AL.

or reactive oxygen species (ROS), which leads to cellular oxidative stress (Ogren and Sjostrom 1990; Perl-Treves and Perl 2002). Oxidative stress is a widespread phenomenon in plants under environmental stress and is indicative of over-excitation of the photosynthetic apparatus (Baker 1993; Ort 2001; Foyer 2002). As part of an integrated response to environmental stresses, plants have evolved an array of scavengers (antioxidants) to detoxify the ROS. These include enzymes and low molecular weight compounds such as ascorbate, glutathione, superoxide dismutase, and xanthophyll cycle pigments (Demmig-Adams 1990; Foyer 2002). Some of the xanthophyll cycle antioxidants (e.g., zeaxanthin) are produced in elevated amounts under drought and high light stress (Demmig-Adams 1990; Long et al. 1994; Demmig-Adams and Adams 2002). Plants with elevated levels of these antioxidants seem to tolerate stresses better than those with low levels (Demmig-Adams 1990; Foyer et al. 1995). Recent work on sugarcane expressed sequence tags (ESTs) has led to the identification of several genes involved in antioxidant defense such as metal chelators, low molecular weight compounds, antioxidant enzymes, and repair systems (Sorares-Netto 2001). Signal transduction is an important component of the stress perceptionreaction cascade. The abundance and activity of signal transduction reactions may determine the level of stress resistance of a genotype. One of the early reactions to stress is an elevation in the level of free calcium in the cytosol (Souza et al. 2001). Plants also contain a large family of Ca2+dependent protein kinases (CDPKs) that are implicated in signaling pathways in response to stresses such as drought, wounding, and cold. Transgenic analyses of stress-responsive gene expression and characterization of CDPK activation suggest that some stress-induced Ca2+ signals are perceived and translated by CDPKs. In Arabidopsis, one gene coding for a calcineurin B-like protein is strongly induced by stress signals (Kudla et al. 1999). Ectopic expression of CDPK induced the expression of a rice stress-responsive gene RAB16 (Saijo et al. 2000) and HVA1 in maize protoplasts (Sheen 1996). Sugarcane EST clusters corresponding to calcium channels and calmodulin-regulated channels have been identified (Souza et al. 2001). A search in the SUCEST database showed 169 clusters coding for CDPKs, with possible functions in cellular communication/signal transduction and/or RNA metabolism and transcription. Other signaling molecules such as the jasmonates (JAs) and inositol triphosphate also play critical roles in plant responses to many biotic and abiotic stresses (Reymond et al. 2000). Sugarcane EST clusters encoding proteins involved in jasmonate, GA, ABA, ethylene, and auxin biosynthesis have been identified (Souza et al. 2001). This suggests the existence of major genes that could serve as breeding goals. One

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

45

approach for sugarcane improvement could be the utilization of quantitative trait loci (QTL) information, to accumulate favorable alleles in a single genotype. Such an approach has proved useful in rice where three major genes for rice blast (caused by the fungal pathogen Magnaporthe grisea) resistance have been pyramided (Hittalmani et al. 2000; Zhou et al. 2004). Identification of such QTLs by molecular markers should facilitate such gene accumulation. The diversity of stress resistance mechanisms discussed above also reflects responses at different levels of organization (molecular, cellular, anatomical, morphological, whole-plant, or crop levels), over different timescales (immediate responses, acclimation, and adaptation). Within seconds to days of exposure to stress, there is usually a decline in the activity of the affected physiological processes. Processes involved in the immediate response to stress are usually rapidly reversible depending upon the intensity and duration of the stress. For example, diurnal changes in vapor pressure deficit often result in midday depression of photosynthesis in many species (Jifon and Syvertsen 2003). As the stress persists, plants may compensate for the reduced performance through acclimation (hardening or morphological and physiological adjustments) within days to weeks. Acclimation, which is indicative of a genotype’s phenotypic plasticity, allows the plant to adjust physiologically or morphologically to rapid daily and seasonal changes in environmental parameters (Jones 1992). Tolerance mechanisms usually involve a strong metabolic acclimation. Osmotic adjustment, heat- and cold-shock protein synthesis, ethylene response during hypoxia, and photosynthetic induction during sun flecks are examples of tolerance mechanisms established through physiological acclimation (Nilsen and Orcutt 1996). These different strategies give an indication of the level of genetic variability in a given trait and dictate the techniques available for screening and breeding. 2. Physiological and Molecular Basis of Stress Resistance. Understanding the physiological and molecular basis of stress resistance is crucial in crop improvement for performance in stressful environments (Jackson et al. 1996). Considerable progress has been made in understanding the molecular, biochemical, and physiological bases of traits that enable plants to cope with individual stresses. Incorporating these traits into commercial germplasm has been slow, partly because of the complex inheritance patterns of stress resistance traits and yield components and the lack of efficient selection techniques (Flowers and Yeo 1995). Among the small grain cereals, incorporation of the dwarfing genes led to increased resistance to lodging and made the Green Revolution

46

MING ET AL.

possible. From a physiological perspective, this shift in biomass allocation from stem growth to grain growth resulted in higher harvest index and increased yield potential (Evans and Fischer 1999). The use of physiological traits in a breeding program, either by direct selection or through a surrogate selection such as for specific molecular markers, can result in more precise targeting of factors limiting yield and may result in faster rates of genetic improvement. A key step in identifying the physiological or molecular basis for stress resistance is to verify existence of a cause-and-effect relationship between the physiological process or presence of a particular protein and whole-plant stress tolerance. This can be achieved by analyzing differential gene products (proteins) associated with stress response, followed by isolation and characterization of the genes responsible for the proteins. The cause-andeffect relationship is then verified by transforming the gene into a stress sensitive plant that does not possess the stress tolerance trait and determining if the genetically modified plant acquires resistance. Molecular techniques such as MAS might offer a more efficient means for selection of physiological traits, particularly if they are used early in a breeding program to facilitate culling of low potential lines and selection of high potential lines for hybridization (Lee 1995). Molecular markers might also provide a better understanding of the genetic basis of the trait. Continued refinements in the genetic makeup of sugarcane and other species should facilitate use of MAS and genetic transformation for improving stress resistance and yield potential in sugarcane. 3. Considerations for Breeding. The effective identification of useful stress resistance traits and characterization of their genetic complexity are important steps in the breeding process. Trait identification can be difficult depending on the level of heterozygosity for the trait in a population. Under stressful environments, selection pressures affecting the genome of a population can result in a decrease or increase in the range of the stress resistance traits that can be used for improvement purposes. The genomes of species that have evolved in environments with extreme conditions are likely to contain genes that confer stress resistance. In contrast, the genomes of many modern crop plants selected under more optimal conditions may contain a smaller collection of such traits due to the artificial selection process employed in their domestication. Given that environmental stress is the major constraint to crop yield, it is prudent to search for resistance traits and genes in a wide array of both wild and domesticated genotypes. Screening and Selection Methods. Development of rapid and practical screening methods for identifying and characterizing genetic variation

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

47

for traits of interest can expedite the crop improvement. Drought stress, for example, is one of the most limiting growth factors in sugarcane production. Many indices (e.g., tissue water potential, relative water content, stomatal conductance, ABA, free amino acid accumulation, root development, etc.) have been used to measure the development of sugarcane water stress as a way of screening for drought tolerance among genotypes (Meinzer et al. 1994; Sheu 1996). Water use efficiency (WUE or amount of water required for growth and production; see discussion below) is another index that integrates plant characteristics (such as photosynthetic capacity, stomatal behavior, and leaf characteristics) with climatic factors. Paraquat (methyl viologen) resistance has been used successfully to screen barley and wheat varieties for drought tolerance (Altinkut et al. 2001). Resistant genotypes showed reduced damage (chlorophyll degradation) from paraquat-induced oxidative stress (increased superoxide generation), presumably by having efficient antioxidant systems. A combination of paraquat resistance and chlorophyll fluorescence has been used effectively to analyze drought tolerance of the sugarcane cultivars TCP87-3388, ‘CP72-1210’, and ‘NCo-310’ (da Silva and Jifon, unpubl.). Combining paraquat resistance and chlorophyll fluorescence has the advantage of speeding the screening process because chlorophyll fluorescence is a sensitive (pre-visual) indicator of oxidative stress (van Kooten and Snell 1990; Demmig-Adams et al. 1997). These measurements are non-intrusive, reliable, fast, and easy to collect in both field and laboratory studies, making the technique potentially useful for stress tolerance screening in sugarcane cultivar development programs (Earl and Tollenaar 1999; Fracheboud et al. 1999). Advances in molecular, physiological, and biochemical techniques as well as increased cooperation between physiologists, molecular biologists, and breeders have led to a better understanding of the diversity of stress resistance mechanisms and to more efficient ways of identifying, characterizing, and incorporating them into commercial germplasm. The molecular improvement approach holds promise in being able to selectively incorporate specific traits into superior yielding cultivars in a fraction of the time and reducing the effort required to achieve similar results through conventional breeding. Genome sequencing has great potential in facilitating crop improvement for stress resistance thanks to advances in automation and the computing/storage capacities of modern computers. It is becoming realistic, for example, to combine large amounts of DNA sequence data with high-throughput molecular biology methods to identify genes that are differentially expressed under specific environmental conditions (such as drought). These genes are especially promising candidates for further study by genetic and breeding approaches.

48

MING ET AL.

Identifying molecular markers for sugarcane applications has been slow as a consequence of the complex polyploid nature of its genome (Moore 1996). Extensive work with molecular markers in grass species has identified QTLs controlling response to abiotic stresses. A search of the Gramene database identified 408 of these QTLs. Future improvements in the knowledge of the genetic repertoire of sugarcane that has been achieved, thanks to programs such as the sugarcane EST Project (SUCEST 2001), is expected and will continue to uncover potential stress resistance-related genes and chaperones that could be incorporated into commercial germplasm (Rossi et al. 2003). The possibility of using genetic transformation to engineer plant tolerance to stresses is appealing even if its application to sugarcane may be beyond current technology. Methods needed to transform sugarcane are well developed and useful applications for major gene traits are underway as will be discussed later under genetic engineering. Stress tolerance is quantitative in nature so there may be a large suite of genes needed for improving this trait. In addition, stress tolerance genes may be affected by stage of plant development and other genes through epistasis and pleotropy, complicating the engineering of stress tolerance. 3a. Improving Drought Tolerance. There is a high correlation between the amount of water used by the sugarcane plant and the amount of dry matter produced (Rozeff et al. 1998). In many production areas, the requirement for ample water to produce a sugarcane crop is seldom met by rainfall alone so that supplemental irrigation is required. Given the high cost of irrigation water, drought tolerance could be a major component of sugarcane improvement programs to achieve higher and more stable yields. In addition to reducing yields, drought also predisposes plants to other stresses such as stalk borers and diseases (Mattson and Haack 1987). The root traits of abundance, distribution, and more importantly the total absorbing surface area are generally considered good for drought resistance (Fitter and Hay 1987). Roots contribute to drought resistance by exploring the soil volume more thoroughly for water, thereby delaying the onset of stress related reduced water supply. The deep-rooted S. spontaneum genotypes are generally considered more drought tolerant than the more shallow-rooted, drought susceptible S. officinarum genotypes (Evans 1935; Rao 1951). Leaf characteristics such as short, narrow, and erect leaves, low stomatal density, sunken stomates, thick cuticles, glaucousness, and pubescence play important roles in a plant’s ability to conserve water during drought. These traits generally contribute to drought resistance by lim-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

49

iting the rate of transpirational water loss. Rao (1951) presented a range in values, indicating genetic variation that could be useful in an improvement program, of leaf characters associated with sugarcane genotypes differing in drought tolerance. Stomatal conductance, for instance, has been used to predict yields of irrigated cotton and bread wheat grown at high temperatures (Lu et al. 1998). In sugarcane, stomatal conductance decreases with plant size and age, thus downregulating whole plant water loss as the canopy increases (Meinzer and Grantz 1990). A balance between transpirational water loss and root water transport (hydraulic conductance) mediated by chemical signals in the transpiration stream (e.g., ABA) also has been reported and may represent a homeostatic mechanism for maintaining near constant leaf water status as environmental conditions change (Meinzer and Grantz 1991). The difficulty in using root hydraulic conductance as a trait is that it is tightly linked to other plant parameters and there are no rapid simple methods for its assessment. The differential influence of stomatal conductance on CO2 uptake and water loss has important effects on water use efficiency (WUE). Leaf-level WUE refers to the ratio of photosynthetic carbon assimilation (A) to evapo-transpiration water loss (E), whereas WUE of productivity refers to the ratio between biomass accumulation and water used during production of that biomass. Since A and E are both affected by stomatal conductance, plant WUE also depends on stomatal conductance and the factors influencing stomatal conductance such as the leaf-air vapor pressure difference (LAVPD, i.e., the difference between the leaf’s intercellular air spaces and the bulk air). Temperature also has a pronounced effect on plant WUE, since it affects LAVPD (Lambers et al. 1998). High water-use efficiency (WUE) is an important drought resistance trait for regions where water supply is limited. Leaf-level WUE (or photosynthetic WUE) refers to the ratio of photosynthesis (A) to transpiration (E), whereas WUE of productivity refers to the ratio of biomass accumulation to water used during production of that biomass. The CO2-concentrating mechanism makes C4 plants highly efficient in utilizing available water compared to C3 plants. The productivity WUE for C4 cereals varies from 2 to 4 g dry matter/liter of water transpired compared to 0.9 to 2.5 g/liter for C3 cereals (Larcher 1981; Jones 1992). Few studies have reported genetic variation of WUE in sugarcane (Zhang et al. 2001). Reasons for the paucity of studies include the substantial effort required to conduct WUE trials and the limited application of WUE information to plant breeding because of its marked temporal and spatial variability. Nevertheless, strong correlations between leaf carbon isotope (12C or 13C) discrimination (∆) and WUE means that ∆ can be a

50

MING ET AL.

surrogate measure of WUE that is more meaningful and easier to obtain. ∆ provides an integrated measure of the photosynthetic WUE during the total time that tissue carbon (C) was assimilated. C3 and C4 plants differ in their ability to discriminate against the heavier 13C isotope through physiological processes such as diffusion (stomatal conductance) and carboxylation (see Farquhar et al. 1989 for a detailed review). Values of ∆ in C3 species range from 15–28 parts per million (‰) compared to 2–10‰ in C4 species (Farquhar et al. 1989; Griffith 1993). The lower 13C discrimination in C4 plants is related to increased CO2 diffusion into the leaf, carboxylation by Rubisco (ribulose-1,5-bisphosphate carboxylase/oxygenase), dissolution of CO2 to HCO3– and fixation by PEPC (phosphoenolpyruvate-carboxylase), ratio of leaf internal to external CO2 partial pressure (pi/pa), and leakage of previously fixed CO2 or HCO3– out of the bundle sheath cells back into the mesophyll cells (Farquhar 1983; Bowman et al. 1989; Meinzer et al. 1994; Williams et al. 2001). Recent studies characterizing the genetic variation in ∆ among C4 and C3 plant species report heritable relationships among ∆, WUE, and yield (Hubick et al. 1986; Smedley et al. 1991; Araus et al. 1993; Martin et al. 1999). These findings suggest that ∆ has good potential for use as a screening index in breeding for high WUE and drought resistance. Although a significant correlation between genetic variation in ∆ and grain yield of sorghum (a C4 crop) has been reported (Hubick et al. 1990), an analysis of 120 genotypes of corn (Zea mays) did not indicate any significant genetic variation in ∆ (O’Leary 1988). Meinzer and coworkers (1994) reported a small (0.4‰) but significant genotypic variation in ∆ associated with differences in shoot growth rate of two sugarcane cultivars known to be susceptible (‘H65-7052’) or resistant (‘H69-8235’) to salinity stress. Their data led to the conclusion that “∆ may prove useful as a screening criterion in a breeding and selection program for resistance to salinity and possibly other stresses in sugarcane.” However, since ∆ is strongly influenced by both environmental and genetic factors, timing of the sample collection is an important consideration (Meinzer et al. 1994; Williams et al. 2001). If ∆ is to be used as a screening tool for identifying stress-resistant genotypes, we suggest that the inherently narrow range of ∆ for C4 species could be a potential limitation in resolving genetic variability in ∆ among closely related genotypes. Accumulation of compatible solutes (e.g., proline, glycinebetaine, and trehalose) is an important physiological mechanism, termed osmoregulation, for maintaining cell turgor during water deficits (Nilsen and Orcutt 1996). Proline is one of the best-known examples of compatible solutes used in osmoregulation and has been associated with sugarcane drought and salt tolerance (Rao and Asokan 1978; Paquet et al. 1994;

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

51

Wahid 2004). In addition, Burcer reported increased drought tolerance and yield for sugarcane cultivars that accumulated free-proline. However, such associations are not proof of a causative role for proline. It is necessary to measure tissue water potential components alongside proline accumulation to confirm that proline indeed has a role in sugarcane osmotic adjustment. Therefore, the usefulness of proline accumulation as a drought resistance trait is still debatable and more work is required to evaluate proline accumulation as a useful trait in breeding for stress resistance. Trehalose, a non-reducing disaccharide of glucose, is another example of a compatible solute (osmoprotectant) that plays an important role in stress protection in organisms ranging from bacteria to fungi, higher plants, and invertebrate animals. It may have potential as an osmoprotectant in sugarcane, as trehalose is known to be effective in stabilizing dehydrated enzymes and lipid membranes (Pilon-Smits et al. 1998; Goddijn and van Dun 1999; Wingler et al. 2000; Garg et al. 2002). Despite the wide distribution of trehalose in microorganisms and invertebrates, trehalose had until recently been found in only a few plant species, notably the highly desiccation-tolerant, resurrection plants (club mosses Selaginella lepidophylla Hook and Grev. and the angiosperm Myrothamnus flabellifolius Welw.), so named because of their unique ability to fully recover from a state of almost complete loss of water. Genes that encode enzymes of trehalose synthesis, i.e., trehalose-6-phosphate synthase (TPS) and trehalose6-phosphate phosphatase (TPP), have been identified recently in a number of higher plants. Transgenic rice and tobacco plants expressing these genes were shown to exhibit increased drought tolerance (Pilon-Smits et al. 1998; Garg et al. 2002). However, although transgenic tobacco plants were more drought tolerant, they also tended to show developmental anomalies such as stunted growth (Goddijn and van Dun 1999). Phytohormones, e.g., indole acetic acid, gibberellic acid (GA), cytokinin, abscisic acid (ABA), and ethylene, also play important roles, not only in plant development, but also in the response to environmental stresses. Ethylene and ABA, generally considered to be growth inhibitors, are commonly produced in plant response to stresses such as drought. ABA is ubiquitous in plants and plays an important role in regulating many processes, including gene expression, closure of stomata, photosynthesis, and adaptation to environmental stresses, such as drought, salinity, and cold (Zeevaart and Creelman 1988; Gowing et al. 1993; Tardieu and Davies 1993). Stressed sugarcane leaves have been shown to accumulate ABA. However, while ABA was correlated with leaf water potential, the correlation with genotypic differences in drought resistance was weak (Kuhnle et al. 1979). The signal transduction pathway involved in

52

MING ET AL.

ABA-regulated gene expression has been the subject of numerous research efforts (Gómez-Cadenas et al. 1999). For instance, the serine/ threonine protein kinase, PKABA1 mRNA, which is up-regulated by ABA in seeds, has been shown to act as a key intermediate in the signal transduction pathway leading to the suppression of GA-inducible gene expression in cereal aleurone layers (Gómez-Cadenas et al. 1999). Approaches aimed at up-regulating the production of such phytohormone-mediated kinases may enhance the plants’ ability to withstand environmental stress. Molecular markers associated with water deficit stress have been identified in sugarcane (SUCEST 2001) and in other monocots. Numerous QTLs involved in drought resistance have been found, for instance, in rice. The significance of this for sugarcane improvement relies on the possibility of genetic mapping of stress-related QTLs. Molecular marker genetic analysis of drought tolerance in maize led to the identification of genomic segments responsible for components of yield tolerance to drought (Frova et al. 1999). Data from this study suggested that drought tolerance for yield components is largely associated with genetic and physiological factors that are independent from those determining the yield component traits per se. More importantly, as could be deduced from their largely different positions in the genome, the majority of the drought-tolerance QTLs were unrelated to those controlling the basic traits. Since many favorable alleles were contributed by the less-tolerant line, a fair amount of transgressive segregation was found. These results are important to the breeder because they indicate, as pointed out by Frova et al. (1999), that “tolerance QTLs can be exploited in a selection program, since the positive alleles could be introduced in a wellperforming genotype without breaking the existing favorable genetic combinations that control the basic trait.” Taking into account the close similarity between sugarcane and maize (Freeling 2001), a comparative mapping approach might be utilized to expedite the identification of analogous sugarcane QTLs. 3b. Improving Low Temperature Stress Tolerance. The sensitivity of C4 plants (e.g., sugarcane and maize with tropical and subtropical origins) to low temperature stress is well documented (Moore 1987; Wolfe 1991; Boese et al. 1997; Ebrahim et al. 1998 a, b; Du et al. 1999; Du and Nose 2002). The optimum temperature for sugarcane growth is about 35°C, and non-freezing temperatures under 20°C will significantly curtail growth and yield (Moore 1987). The risk of freezing damage is one of the major limitations to growing sugarcane at latitudes more than 30° from the equator. Freezing temperatures damage mature plants prior to

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

53

harvest and can kill re-growth of young plants and ratoons. Typically, this kind of damage results in delayed crop development and reduced stalk populations. In the continental United States, winter freezes have forced the industry to adopt short growing and milling seasons. Growers and processors try normally to complete the harvest before January. With record crops harvested during the past three years in the state of Louisiana, many growers and processors had to extend the harvest until mid January, thus increasing the risk of freeze losses . Exposure to sub-optimal temperatures at any growth stage can decrease sugarcane growth and productivity (cane yield and sugar content). The photosynthetic apparatus, which is ultimately responsible for sugar production, is highly sensitive to both chilling and freezing temperatures (Kratsch and Wise 2000). The mesophyll chloroplasts of three grass species showed significant swelling at 0°C when plants were exposed to temperatures below 25°C. Leaves exposed to low temperature typically exhibit a wide range of physiological effects, including decreased net photosynthesis. At the photochemical level, as indicated by changes in chlorophyll a fluorescence parameters, low temperature exposure caused a reduction in the quantum yield of CO2 assimilation, a reduction in the fluorescence yield from dark-adapted samples, and the onset of photo-inhibition (Long et al. 1994). In vivo chlorophyll fluorescence monitoring provides a very sensitive method of probing the photosynthetic apparatus and has been used as a screening method for cold tolerance in maize (Earl and Tollenaar 1999). This method has been considered as an efficient selection tool for improving cold tolerance of maize through breeding (Fracheboud et al. 1999). Improved cold tolerance would give sugarcane growers additional insurance against the harmful effects of early winter freezes during harvest. Superior freeze tolerance has been reported in high-fiber genotypes and selections of S. spontaneum, S. sinense, and Miscanthus (Irvine 1977). Native to subtropical India, S. spontaneum is highly polymorphic and is well adapted to a wide range of climatic conditions as might be experienced from sea level up to about 2700 m (Roach and Daniels 1987). Nevertheless, with the possible exception of Louisiana, where S. spontaneum germplasm has been used frequently in the commercial breeding program (Legendre, pers. comm.), the major sugarcane breeding programs in the U.S. have traditionally relied primarily on nobilized hybrids of S. spontaneum and S. officinarum as sources of breeding material (Walker 1987). Making the basic inter-specific crosses involving different species of Saccharum and further “nobilization” to commercial status would likely require at least a decade to introgress positive alleles for cold tolerance.

54

MING ET AL.

Using the MAS approach may offer a better solution (Wiegand et al. 1996; SUCEST 2001). Thirty-four cold-induced sugarcane ESTs were identified among the total of 263,000 ESTs in the SUCEST 2001 sugarcane EST project database (Nogueira et al. 2003). In this work, five cold-induced sugarcane ESTs encoding polyubiquitin proteins were identified. In addition, genes encoding PPDK (an important enzyme in C4 plant photosynthesis) and NADP-ME (a major decarboxylating enzyme) were found to be induced by cold exposure. These results suggest mechanisms for maintenance of photosynthesis at low temperatures and illustrate the importance of integrated stress response systems. Since chilling also induces oxidative stress (Du et al. 1999; Fracheboud et al. 1999; Ort 2001; Du and Nose 2002), it seems that the best approach for producing cold tolerant sugarcane might be to selectively incorporate traits such as increased antioxidant content to confer an integrated tolerance to multiple stresses. 3c. Improving Disease Tolerance. Increased resistance to pests and diseases is a goal of most crop improvement programs worldwide and has resulted in effective control of many sugarcane diseases. For a majority of sugarcane diseases, there seems to be ample genetic variation for resistance/tolerance in wild genotypes, especially those of S. spontaneum, which appear to be the major source of resistance to leaf scald, red rot, smut, and mosaic diseases (Walker 1987). Genes conferring pathogen race-specific resistance are commonly clustered, often arranged in plant genomes as direct tandem repeats, which is congruent with their continuous evolution through unequal exchange (Richter and Ronald 2000). Resistance genes (RGs) evolve more rapidly than the rest of the genome through sequence divergence and unequal recombination of microsatellites present within these genes to alter the number of RG family members and increase variation for resistance within the population. Recombination provides a mechanism for generating new race specificities by creation of novel resistances to biotypes that neither parental allele encoded. Microsatellites found within sugarcane resistance genes (da Silva 2001) and within signal transduction ESTs (J. da Silva, unpubl.) can be used for gene tagging that has potential to improve the efficiency in breeding for disease resistance. Molecular marker identification of the chromosomal location of sugarcane RGs should enhance the ability to assess the sugarcane germplasm for the presence of genetic resistance and allow the introgression of positive alleles into commercial hybrids (Lee 1995; Berding et al. 2004). This strategy has been implemented by mapping amplified fragment length polymorphism (AFLP) markers at 1.9 and 2.2 cM flanking

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

55

each side of the sugarcane rust resistance gene Bru1 (Asnaghi et al. 2004). Mapping this gene should allow for the evaluation of map-based cloning and plant transformation for the development of resistant cultivars. The RNA fingerprinting technique of cDNA-AFLP also might be used to assess disease response of different cultivars since pathogen attack often triggers specific patterns of gene expression. Thokoane and Rutherford (2001) used this technique to identify differentially expressed genes (chitin receptor kinase, Pto ser/thr protein kinase interactor, and retrotransposon) in a smut resistant sugarcane cultivar that had been challenged with smut inoculation. In a similar study, sequence cultivar of genes differentially expressed in response to challenge by smut identified putative receptors involved in the signaling of resistance mechanisms, transcription factors, and enzymes involved in phenylpropanoid metabolism (Heinze et al. 2001). 4. Cost-Benefit Analysis of Stress Resistance. The ultimate goal of any crop improvement and selection program is to make direct genetic gains for yield as well as indirect increases in yield via improvements in other characters such as quality or increased resistance to environmental and biological stresses. However, producing genetic gains for performance in a stressful environment is often difficult to justify economically (Bell et al. 1995). One reason is that improved stress resistance is often attained by changing the allocation of photosynthetic carbon resources among plant organs so that the harvested component is reduced under non-stress conditions. Resistance/tolerance to environmental stress necessarily involves a balance between the benefit of allocating resources toward one process (e.g., stress resistance) and the cost of reducing those resources from other competing processes (e.g., sugar production/ storage) (Robertson et al. 1996). For instance, defenses against herbivory damage may require substantial investment of energy, carbon, and nitrogen resources for the synthesis of secondary metabolites at the expense of primary metabolism and growth that are necessary for high productivity. Natural selection may favor those genotypes that are most effective at adjusting the trade-off between primary and secondary metabolism in order to optimize productivity in stressful environments. Related to the cost-benefit analysis of stress resistance is the concept of resource use efficiency, which can be considered in terms of adaptation (a shift in allele frequency) or acclimation (phenotypic regulation or differential gene expression) (Nilsen and Orcutt 1996). For example, the nutrient use efficiency of plants adapted to low nutrient environments may exceed that of plants from high nutrient environments;

56

MING ET AL.

however, plants from the low nutrient environments also may be less able to utilize fluctuations in resource availability compared to plants adapted to high nutrient environments. In environments characterized by short-term temporal patchiness of resource availability, a genotype’s ability to rapidly adjust to the changing resource status (phenotypic plasticity) may be the most efficient mode of stress resistance. In environments characterized by spatially consistent patchiness in resource availability, a change in allele frequency in populations as opposed to individuals (i.e., adaptation) would be most efficient. Recognizing the potential trade-offs between stress resistance and yield potential should help breeders target specific traits in their breeding programs in a sitespecific manner rather than treating resistance traits as being universally beneficial. 5. Concluding Remarks and Future Directions. Plants have evolved a suite of mechanisms to cope with environmental stresses. The challenge is to transfer appropriate stress resistance traits into domesticated crops. Some of these mechanisms responsible for resistance traits represent adaptations, while others are manifested as phenotypic plasticity. Conventional breeding/selection for some of these traits has already contributed to substantial gains in crop yield potential (Bell et al. 1995), but this process is laborious and time consuming. As the genomic data for sugarcane continue to expand, genes conferring stress tolerance probably will be uncovered and incorporated into commercial sugarcane germplasm. Recent advances in MAS and genetic transformation technologies can expedite this process. In addition, developing methods for coordinated up-regulation of expression of existing genes or particular points in a pathway also may be used to enhance stress resistance. Ultimately, the degree of genetic linkage among traits of interest will determine the number of traits that can be managed during selection for cultivar improvement. Where linkages are rare, simultaneous selection for several traits will be necessary. The extent of such linkages between yield/quality traits and vigor/stress tolerance characters is still not known. Hallauer and Miranda-Filho (1981) noted that such correlations are not strong in maize. Much more research has been conducted on the genetic mechanisms of stress tolerance for species such as maize, rice, wheat, Arabidopsis, and soybean than has been done for sugarcane. However, the extensive genetic and metabolic conservation that exists within the grass family should help in efforts to tag RGs in sugarcane (Asnaghi et al. 2004). Genomic programs such as the SUCEST project will likely provide information that is a valuable foundation for new approaches to improve stress resistance of sugarcane.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

57

III. SUGARCANE IMPROVEMENT THROUGH BIOTECHNOLOGY A. Physiology of Sucrose Accumulation 1. Background. Contemporary plant research is rapidly revealing the potential of new methods for genetic improvement of higher plants. These methods include various procedures involved in defined genetic selection systems and recombinant DNA technologies. To use these techniques effectively, clear understanding is necessary of how plant growth and development, coupled with physiological and biochemical processes, are regulated by the environment to control crop yield. Since about 1727, when Stephen Hales published Vegetable Staticks explaining how plants grow, people have referred to the science describing the interaction of plant structure and function as plant physiology. However, following the discovery by Watson and Crick of the structure of DNA in 1953, plant physiology has evolved into a highly integrated field that now includes not only physiology, biochemistry, and structural biology, but also aspects of developmental biology, molecular biology, genetics, and genomics in an attempt to explain how plants function. Modern plant physiology has become such an interdisciplinary activity, operating across hierarchical scales from the gene to the ecosystem, that there has been a move to substitute “plant biology” for what used to be “plant physiology.” In fact, the American Society of Plant Physiologists, established in 1924 to become the world’s premier professional organization of plant physiologists, changed its name in 2001 to the American Society of Plant Biologists. Here, we discuss the role of plant physiology in increasing sucrose accumulation for improving sugarcane through breeding and biotechnology in the spirit of modern plant physiology. 2. Plant Systems Biology. In plants, a “system” can be described at the level of gene action, a biochemical pathway, an organelle, a cell, an organ, a whole plant, or a community of plants (crop or ecosystem). Systems biology attempts to discover and understand the biological properties that emerge from the interactions of many system elements (Kitano 2002). This is contrasted with the reductionist approach that, prior to the development of high-power computing and modeling, was forced to consider one or a few variables at a time. This is not an argument against the reductionist approach (one gene, one enzyme, or one mutant at a time), for it has been extremely successful. For example, Mendel’s laws, discovered because he used the reductionist approach of crossing pairs of inbred lines with only a single contrasting character, became the

58

MING ET AL.

foundation for the genetics employed by plant breeders for the past 80 to 100 years. Despite the power of the reductionist approach, it has limitations in that: (1) it works only for systems that have minimal interactions, (2) the laws discovered are applicable only to a narrow set of conditions, (3) because of the first two limitations, it would take a long time and be prohibitively costly to conduct enough independent onedimensional experiments to develop laws governing a whole system. The discovery of the structure of DNA (Watson and Crick 1953) has become the basis for the many concepts and technologies of molecular biology that link information about life’s physical entities such as DNA or proteins to the genetic traits of organisms. The specificity, speed, and minuteness of scale of DNA molecular interactions have led to the development of high-throughput technologies that quickly provide us spectacular breakthroughs such as full sequencing of genomes, including those of the model plant Arabidopsis (Arabidopsis thaliana L.) and the crop plant rice (Oryza sativa L.). In addition, post-genomic developments, such as using microarrays and proteomic techniques, are enabling us to determine which genes are activated or inactivated during development or in response to an environmental change. Amassing large molecular datasets is one thing; producing an understanding of what they all mean about plant growth and development is quite another. Collecting and analyzing the information from even a single high-throughput experiment quickly reveals the complexity and magnitude of effort required to build a sufficiently comprehensive dataset to predict how the plant, or tissue, or cell, or biochemical pathway will perform under many different sets of conditions. One can scarcely envision the magnitude of systematically establishing perhaps 500 different environmental or developmental stages over which to compare the mRNA expression of a genome of 25,000 genes. What is needed in place of a traditional bottom-up approach, i.e., analyzing the details of small sub-systems, is a top-down approach that models the entire system from the behavior of its many sub-systems (Minorsky 2003). In molecular biology, a top-down or systems approach is based on models that are developed to integrate the exhaustive descriptions of biological systems to predict how the different levels or hierarchical scales interact to form higher functional units like coordinated pathways, regulatory networks, or complex structures such as cells or tissues. The ultimate goal is to understand the biological system in sufficient detail to enable accurate, quantitative predictions about its behavior when we somehow manage to introduce or block the expression of a suite of genes. This will give us the ability to engineer the design of a crop plant predictably. The challenge is to improve our understanding of how plants function at all scales of complexity to such an extent that we can produce models that

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

59

will predict how a crop will respond to any given genetic manipulation or environmental perturbation. 3. Crop Factors and Yields. Crop productivity is a conceptually simple system driven by photosynthesis that involves plant interception of photosynthetically active radiation (PAR) and the conversion of that radiant energy to electron energy that is used in a multistep enzymatic pathway to convert CO2 and water into carbohydrates. The carbohydrates produced may be converted to other organic compounds (biosynthesis), used for energy to maintain the plant as a living entity (respiration), or be assembled into new plant structures (growth and development). The harvested yield of a crop is the result of a complex interaction of plant structures and the processes involved in the amount of radiant energy captured, the efficiency with which the captured energy is used in the production of new dry matter, the proportion of dry matter partitioned among the different plant parts, the amount of dry matter lost through all causes (physiological, pathological diseases, and pests), and the accumulated time over which these processes act. Viewed in this way, crop productivity is deceptively simple, much like describing the flight of a Boeing 777 as the result of air pressure lift generated by air flow over a particularly shaped wing structure. However, just as there are tens of thousands of variables involved in the successful controlled flight of a giant aircraft, there are even more interactions of genes, enzymes, pathways, cells, organs, and environmental signals involved in the productivity of a crop. A major goal of plant physiology is to link crop production in a meaningful way to factors that can be selected and manipulated by plant breeders to develop new germplasm that has been somehow improved for increased productivity. Factors that maximize the amount of light absorbed by a sugarcane crop are the optimum leaf area per stalk, the number of stalks per unit of land area, and the absorbance characteristics of the leaf. Although these factors vary considerably with age, environment, and cultivar, we can envision an average commercial hybrid cultivar that might support 65,000 to 80,000 mature stalks/ha throughout most of the crop cycle and be harvested. Each stalk would have 10 to 14 fully expanded leaves with each blade having an area of about 0.05 m2 so that each stalk and tiller would have a leaf area of 0.5 to 0.7 m2 to give the crop a leaf area index (LAI) of 3 to 7. Once crop development proceeds beyond the early growth phase, the LAI may be in the 6 to 8 range but then declines near the time for harvest. Changes in canopy size occur with changes in rate of leaf formation at the shoot apical meristem, the rate of leaf emergence and expansion from the leaf whorl, and the rate of leaf senescence. Canopy size is

60

MING ET AL.

regulated by genotype, biotic stresses of disease, and abiotic stresses including diminished sunlight low in the leaf canopy, low temperatures, and drought. In general, crop development is more sensitive to temperature than it is to sunlight, which is only infrequently a limiting factor. Sugarcane, the plant in which C-4 photosynthesis was discovered, is a leading performer in rates of photosynthesis. Sugarcane has been reported to have carbon fixation rates as high as 28 mg CO2m–1s–1 (63 µmolm–1s–1) (Irvine 1975). For limited periods of time, the total dry matter produced by the sugarcane crop may reach 41 gm–1d–1 (Muchow et al. 1994). If this rate of dry matter accumulation were constant for an entire year, it would attain a dry matter yield of 150 t/ha-yr, about half of which would be partitioned into the harvested stalk. Since the stalk dry matter is 45% to 55% sugar, the sugar yield could potentially reach 33 to 41 t/ha-yr. Yields in this range have not been achieved on a commercial field basis but they have been achieved in highly managed experimental plots (Moore et al. 1997). As noted in the breeding section, maximum sugar yields on a commercial field basis in Hawaii have been recorded at 24.2 t/ha-yr. This record yield is 35% of a theoretical maximum sugar yield of 70 t/ha-yr (Moore et al. 1997) that was calculated on the basis of the model of Loomis and Williams (1963) (Loomis and Amthor 1999). Experimental plots have attained sugar yields of 35.2 t/ha-yr in Hawaii and 32.8 t/ha-yr in Australia, which, although phenomenal, are still only about half of the theoretical maximum. 4. Sucrose Accumulation Processes. Sucrose synthesis and accumulation in higher plants is the product of an extensive network of interactions that can be analyzed from several perspectives. At the gross level, sucrose accumulation is simply the difference between the amount of sucrose produced in the leaf by photosynthesis and the amount of this sucrose that is removed by metabolism to produce carbon and energy for growth and other components of the plant. But this is not all that sucrose does. There is increasing evidence that sucrose is involved in signaling to modulate expression of genes controlling cell division and differentiation, transporters and storage proteins, induction of flowering, differentiation of vascular tissue, seed development, and accumulation of storage products (Lunn and MacRae 2003; Koch 2004). Each of the reactions involved is controlled by activation of specific genes by an interaction among the genotype of the plant, the environment under which it is growing, and its developmental stage at that instant. Thus a global analysis of gene expression in sucrose accumulation will identify pathways directly involved as well as pathways that are in turn regulated by the change in sucrose.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

61

Modern sugarcane cultivars are interspecific hybrids that, under ideal conditions, are capable of storing sucrose in the parenchyma tissues of the stem up to 62% of the dry weight or 25% of the fresh weight (Bull and Glasziou 1963). These are approximately the levels obtained in S. officinarum, the major source of commercial hybrid germplasm. On the other hand, some of the wild relatives of sugarcane store less than 2% of the fresh weight as sucrose. These striking differences in sink accumulation and storage activity cannot be explained by differences in photosynthetic rates in the leaves. The photosynthetic rates on an area basis of S. spontaneum have been reported as nearly twice those of S. officinarum and 30% greater than that of hybrid cultivars (Irvine 1975). Since photosynthetic rates cannot explain the differences in sucrose storage, they might be explained by differences occurring within the stem or within the transport system(s) between the source of photoassimilates in the leaves and the deposition of those photoassimilates in the stalk sink. Multiple pathways involved in sucrose accumulation in which there might be rate-limiting physio-biochemical reactions include: • Leaf reactions including photosynthetic rate, sucrose synthesis, metabolism, and carbon partitioning across various membranes into different pools; • Phloem reactions including loading in the leaf and translocation to and unloading in various sink tissues including primary storage in parenchyma cells of the culm; • Stalk reactions including membrane transport, sucrose metabolism, carbon partitioning into different pools, and remobilization of stored sucrose; • Genetic and developmental controls including timing of maturation; • Environment perception and signal transduction pathways to coordinate plant development. The sucrose accumulation pathway includes metabolic and physical processes in cells and tissues that are involved in sucrose synthesis in the leaf, short-distance and long-distance transport tissues for the export of sucrose from the leaf to various sites in the plant, and parenchyma tissues of the stem storage or meristem tissues located primarily at the shoot and root apices. Sucrose accumulation rates vary over a wide range as a function of plant genotype, developmental stage, and the environment in which the plant is grown. The plant responds to its environment through a complex network for the perception and transduction of information to coordinate plant development including the accumulation of sucrose.

62

MING ET AL.

The size and complexity of the numerous networks involved in regulating sucrose accumulation have until very recently forced scientists to confine their studies to only a small subset of conditions presumed to be critical to sucrose accumulation. Sugarcane literature abounds with reports of the level of sucrose determined at the crop level or in plants or parts of plants as a function of various factors such as genotype and age of the plant and sometimes a set of differing environmental conditions such as different levels of mineral nutrients (frequently nitrogen), amounts of water, and temperature. Some studies include reports correlating the activity of a small set of leaf or stem enzymes to the level of sucrose measured. There is also considerable literature evaluating enzyme activity and cell membrane properties based on studies with stem tissue slices, cell cultures, and protoplasts isolated from cell cultures. Results of the numerous studies have produced a few simple models aimed at identifying bottlenecks or constraints on sucrose accumulation of sugarcane. Although these studies have produced considerable data and interesting hypotheses about sucrose accumulation, they explain only how a certain genotype under a prescribed set of conditions once behaved. While the data obtained thus far have value for establishing an initial hypothesis about sucrose accumulation, what is needed is a global analysis of the metabolic networks involved in sucrose accumulation. 5. Global Analysis of Sucrose Accumulation. The national and international efforts underway to develop and catalog ESTs for major food crops are paralleled by independent sugarcane EST projects in Australia, Brazil, South Africa, and the United States. Most notable among them, accounting for >90% of reported sugarcane ESTs, is the SUCEST program of the Brazil ONSA consortium. An initial report about this herculean effort was published as a special issue of Genetics and Molecular Biology (Vol. 24, No. 1-4, 2001) entitled “Sugarcane Transcriptome: A Landmark in Plant Genomics in the Tropics.” It contains 34 research articles from 74 sequencing and data mining laboratories relating sugarcane ESTs to factors such as flowering, signal transduction, plant development, aluminum toxicity, pest and pathogen defense systems, mitochondria and chloroplast functions, membrane transport and secretion, cell wall metabolism, and cytoplasm metabolism. These analyses were based on the SUCEST database containing 238,000 ESTs from 26 sugarcane cDNA libraries constructed from several tissues—shoot apical meristem, flowers, lateral vegetative buds, unfurled immature leaves, mature leaves, roots, stem (culm) rind, culm internodes, seed, and tissue culture calli—at different developmental stages (Vettore et al. 2001). The ESTs similar in sequence were assembled into 43,000 clusters of

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

63

which 38% had no matches in existing public sequence databases. Fiftythree percent of the clusters were formed by ESTs in more than one library and thus delimit a group of genes that are coordinately expressed in different tissues; and 47% of the clusters were formed by ESTs expressed in only one library and thus delimit tissue-specific expressed genes. A more thorough analysis of the SUCEST database (Vettore et al. 2003) reassembled the 43,000 sequences to report a 22% redundancy and tentative identity of 33,620 unique genes. Annotation of the 43,000 assembled sequences showed almost 50% of them were associated with protein metabolism, cellular communication/signal transduction, bioenergetics, and stress response. Because cane sugar is a major, internationally traded commodity, there is often a substantial lag in release of information that might lead to competitive advantages. The GenBank database for sugarcane currently lists more than 300,000 EST sequences with 33,620 UniGenes. Interest in the possibility of discovering agronomically important sugarcane genes led an international consortium of sugarcane research institutions (the International Consortium of Sugarcane Biotechnology, ICSB), Australia, and South Africa to establish smaller independent sugarcane EST programs. The ICSB program to date is that of a single laboratory analyzing three cDNA libraries (apex, mature leaf, and mature internode) to develop 9,216 ESTs that were clustered into 3,400 nonredundant tags (Ma et al. 2004). About 57% of these ESTs were assigned a putative function based on statistically significant similarity to previously characterized proteins or sequences. Another 28% corresponded to previously identified, but uncharacterized, sequences. Some of the remaining sequences were predicted to be genes that may be new or unique to sugarcane. Comparisons of the sugarcane ESTs to a large sorghum EST database revealed similar compositions of expressed genes between different tissues, suggesting applicability of the more abundant sorghum data. Curiously, a substantial fraction of the ICSB ESTs were absent from SUCEST, suggesting that genotype × environment interactions play an important role in the samples of sequences expressed. The published sugarcane EST research has profiled gene expression differences between immature culm internodes that are not storing sucrose versus internodes that are mature and storing sucrose (Carson and Botha 2002; Carson et al. 2002), or it has focused on internodes that are in the process of maturing and most active in accumulating sucrose (Casu et al. 2003, 2004, 2005). The underlying hypothesis of both approaches is that knowledge about gene expression associated with high storage of sucrose will be revealed through global analysis and can

64

MING ET AL.

contribute to a systems approach for increasing yield potential. Approximately one-third of the 400 cDNAs analyzed, 200 from each of the two reciprocal subtractive cDNA libraries, were preferentially expressed in either the immature or mature internodes (Carson and Botha 2002). ESTs generated from all 132 differentially expressed clones revealed 95 unique transcripts of which, based on homology, two-thirds were assigned functions such as cell wall metabolism, carbohydrate metabolism, stress responses, and regulatory proteins. ESTs directly associated with sucrose metabolism were found not to be developmentally regulated, suggesting that growth and maturation of the sugarcane culm is associated with the expression of genes for a cultivar of processes other than sucrose metabolism. Likewise, a sequence survey of 7,242 ESTs derived from a sucrose-accumulating, maturing culm revealed that transcripts for carbohydrate metabolism gene sequences (CMGs) were relatively rare in this tissue (Casu et al. 2003). Nevertheless, ESTs of sugar transporter homologues were highly abundant CMG transcripts. The most abundant of the sugar transporter ESTs was associated with phloem companion cells and nearby parenchyma, suggesting a critical role for partitioning and translocation in sucrose accumulation. The coordinated expression of genes encoding enzymes involved in sucrose synthesis and cleavage as well as glycolysis and the pentose phosphate pathway points toward the need for systems-level approaches to understand sucrose accumulation in sugarcane. Such systems-level approaches are just beginning (Casu et al. 2004; Casu et al. 2005; and Watt et al. 2005). Scientists in South Africa are focusing on a selected subset of genes, proteins, and other small molecules that might be active in sucrose accumulation. This information will be used to develop initial models that will be tested with data from high-throughput technologies, including array analyses in comparative studies, to detect and quantify the molecular responses to each variable (age, development, and genotype: Watt et al. 2005). Later research will focus on perturbing a single genotype with environmental variables (temperature, water, and mineral nutrients) known to effect biochemical changes at the source or sink and then quantify the molecular responses to these perturbations to evolve new hypotheses. Scientists in Australia are focusing on bioinformatic analysis of EST collections from mature and immature stem tissues to suggest functions for genes expressed in the sucrose storing stem and potential interactions among them. They have reported an abundance of several classes of sequence associated with fiber biosynthesis in the maturing stem (Casu et al. 2004). Down regulation of some of these genes might redirect photoassimilate from fiber to increased sucrose accumulation. Gene expression

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

65

profiles, obtained from DNA array analyses, were compared between high sucrose progeny and low sucrose progeny of a segregating population during maturation and sucrose accumulation to develop a “genetical genomics” strategy for identifying candidate genes that may control sugar accumulation (Casu et al. 2004). The programs of both South Africa and Australia are examples in which recent rapid expansion of sugarcane molecular datasets and the beginning of a systems approach to metabolic modeling of sucrose accumulation are pointing toward future applications for raising sucrose yields of sugarcane. 6. Improving Physiological Traits. Fifty years of sugarcane research have identified and characterized a suite of physiological processes and enzymes involved in sucrose accumulation (for reviews see Moore 1995; Moore and Maretzki 1996; Moore et al. 1997). Except for storing high concentrations of sucrose in the vegetative stem tissue, the enzymology and physiology of sucrose production, transport, and accumulation in sugarcane do not differ in significant ways from those in most other higher plants, so one might expect that the genes involved also may be the same. Thus far, this seems to be the case. Prokaryotic and eukaryotic genes encoding sucrose synthesizing and metabolizing enzymes have been isolated, cloned, and used in experiments to transform sugarcane to increase or decrease expression of specific enzymes towards the ultimate goal of increasing sucrose accumulation (Ma et al. 2000). However, results of this reductionist approach have fallen short of expectations, apparently because of the complex interactions among the multitude of simultaneous processes. Recent rapid expansion of sugarcane molecular datasets and the beginning of a systems approach to metabolic modeling of sucrose accumulation point the way for future research efforts to integrate processes from gene to crop performance. In a classical paper, Donald (1968) pointed out that plant breeding is based on either “defect elimination” or “selection for yield.” Defect elimination might be achieved by selection for pest and disease resistance/tolerance and for morphological attributes such as leaf characters to reduce drought susceptibility. Many defect traits are controlled by a small number of genes that can be mapped in segregating populations and then tagged to follow during introgression into the desired genetic background to accelerate germplasm enhancement. On the other hand, if there are no non-defect genes (e.g., disease resistance genes) in compatible germplasm, one might obtain such genes from non-related sources for transgenic insertion into cultivars needing to be improved (see later section on genetic engineering). In either case, eliminating defects is relatively easy because the small effect of environment on the

66

MING ET AL.

expressed trait means that there is not a large network of genes involved. This differs drastically from the case of traits related to growth, development, and yield because these quantitative traits are encoded by suites of genes (complex polygenic nature) whose expression is regulated by constantly changing environments. In the latter case, environmental factors interact with complex gene networks and transcription factors that control signaling to communicate in intricate ways among the multiple metabolic networks underlying the total developmental and physiological processes of the plant. Crop simulation modeling began as a way to integrate weather data (temperature, solar radiation, precipitation, relative humidity, wind direction and speed), soil data (nutrients, pH, ion exchange capacity, salinity/sodicity; and bulk), and managerial operations (date of planting, plant density, irrigation and fertilizer amounts and dates of application) with plant physiological processes regulating growth, development, and yield (de Wit 1965). Crop simulation models proved to be quite good at revealing areas where more research and data were needed to simulate the physiological processes leading to yield. They also pointed out differences among genotypes in such parameters as length of time spent in a specific developmental stage or the level of physiological response to an environmental variable. Subsequently, genotype differences were addressed by developing genotype specific coefficients called “genetic coefficients” as inputs for the models. Such coefficients, intended to summarize different aspects of the genetic make-up of the individual, represent in numerical terms either the presence or absence of a group of genes that operate together as an interconnecting network, or the presence or absence of specific genes (Hunt et al. 2003). Thus, for the past few years, there have been efforts to scale down from the crop or plant phenotype level to the level of genes regulating the expression of the phenotype so that the rapidly accumulating genomics information might be used as estimators of performance of select genotypes in specific environments (Cooper et al. 2002; White and Hoogenboom 2003; Yin et al. 2003). Modeling the genotype (15 genes controlling four adaptive traits) performance of sorghum grown under water-limited conditions suggested an optimum breeding strategy for using genomics and markerassisted selection in crop improvement for dryland adaptation (Chapman et al. 2003). Concurrent with efforts toward scaling down from crop to gene by process-based models are efforts to scale up from genes and gene expression levels to the level of the transcriptome, proteome, and metabolome of cellular pathways. While this approach has succeeded in linking the transcriptome and proteome in highly conserved components of devel-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

67

opment in fruit fly (von Dassow et al. 2000) and sea urchin embryo (Davidson et al. 2002) and in galactose utilization by yeast (Ideker et al. 2001), it has not yet reached the level of the transcriptome nor has it been applied to the much more complex problems of plant development. Hammer and co-workers (2004) see limitations of this approach for analyzing plant growth and have suggested that a strictly “bottom up” approach towards understanding plant growth and development will not likely succeed. The challenge then remains how to proceed to fill the vast middle ground between models built up from descriptions of gene sequences and the comparatively simple crop simulation models now in use (White and Hoogenboom 2003). A better understanding of sugarcane physiology, and thus improved chances for increasing sucrose accumulation, will be achieved ultimately through a combination of experimental and computational analyses of the comprehensive datasets currently being developed. B. Genetics and Genomics In the last 15 years, genetic and genomic research in sugarcane has yielded enormous amounts of new knowledge and information due to the successful use of single dose markers for linkage mapping (Wu et al. 1992) and the rapid advance of molecular techniques (reviewed by D’Hont and Glaszmann 2001; Butterfield et al. 2001; Grivet and Arruda 2001). In this short time, the genome structure of modern sugarcane cultivars has been elucidated; the phylogenetic relationship of Saccharum species has been clarified; multiple genetic maps have been constructed; over 30,000 unique genes have been sequenced; the basic chromosome numbers of Saccharum have been resolved; and map-based cloning of a rust resistance gene is getting very close to the target. 1. Genetic Diversity. Taxonomy of the sugarcane complex, based on morphology, chromosome numbers, and geographical distribution, has been controversial since the original classification of Saccharum officinarum by Linnaeus in 1753. Recent genomic data for evaluating genetic diversity within the genus is beginning to suggest new relationships among accessions and may ultimately produce a definitive classification for the sugarcane species. The first molecular evidence came from restriction fragment patterns of nuclear ribosomal DNA that was used to separate accessions of S. spontaneum, which showed the widest within-species variation, from accessions of four other taxa often afforded species status: S. robustum, S. officinarum, S. barberi, and S. sinense (Glaszmann et al. 1990). RFLP analyses of the mitochondrial genome showed an

68

MING ET AL.

identical pattern among 18 S. officinarum clones and 15 of 17 S. robustum clones (D’Hont et al. 1993). RFLP patterns were similar among S. officinarum, S. barberi, S. sinense, and S. edule, all of which were distinctively different from S. spontaneum. Restriction patterns of the chloroplast genome suggested that, except for S. spontaneum, the Saccharum species all have the same chloroplast restriction sites (Sobral et al. 1994). RFLP analyses of nuclear genomic DNA confirmed observations about the cytoplasmic genomes that suggested distinctively greater diversity within S. spontaneum than among the four other species that were highly similar (Burnquist et al. 1992; Lu et al. 1994a; Nair et al. 1999). The most recent analysis, based on genomic in situ hybridization, is compatible with the hypothesis that S. barberi and S. sinense were derived from interspecific hybridization between S. officinarum and S. spontaneum (D’Hont et al. 2002). These authors conclude that genetic similarities among S. barberi and S. sinense accessions do not support the present classification of these being two distinct taxa. S. robustum has been postulated to be the progenitor of S. officinarum. By all lines of evidence, these two species were the most closely related among six Saccharum species studied in the multiple independent DNA diversity studies mentioned above. Although not proven, S. edule is thought to be an intergeneric hybrid between either S. officinarum or S. robustum and a related genus that may account for its aborted infloresence (Daniels and Roach 1987). Because of the phenomenon of female restitution, hybrids derived from interspecific or intergeneric crosses involving a female S. officinarum conserve the entire genome of S. officinarum, which becomes the genetic basis for the high similarity detected among the five species other than S. spontaneum. The latest molecular data based on in situ hybridization verify Irvine’s (1999) classification of the genus Saccharum into two species: S. spontaneum, as it is presently classified, and S. officinarum, which includes the other four Saccharum species and their interspecific hybrids. Because of its polyploid nature, interspecific origin, and vegetative propagation, high levels of heterozygosity were detected among modern sugarcane cultivars using RFLP markers (Lu et al. 1994b; Jannoo et al. 1999a). The major part of this diversity was attributed to the 15–25% chromosome complement that was inherited from S. spontaneum by random assortment of half its chromosomes, and which have the greatest intraspecific diversity (see above) (D’Hont et al. 1996). Similar patterns of molecular diversity also were detected using AFLP markers (Lima et al. 2002). On the other hand, modern sugarcane cultivars, derived from a small germplasm base contributed by only a few genotypes, show strong linkage disequilibrium. Some haplotypes are conserved in segments extending for at least 10 cM (Jannoo et al. 1999b).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

69

This is in contrast to the situation in diploid species such as Arabidopsis, in which linkage disequilibrium decays down to less than 1 cM or 250 kb (Nordborg et al. 2002). The large-scale linkage disequilibrium of modern sugarcane cultivars is likely a consequence of its narrow genetic base and the few generations of hybridization from the original crosses. The information generated from genetic diversity studies has been applied to sugarcane crop improvement. DNA fingerprinting is widely use in management of large clonal field collections to ensure trueness to label, identification of cultivars to protect intellectual property, and the identification of introgression progeny as to trueness-to-label. RFLP markers were successfully used to identify two mislabeled clones in Hawaii (R. Ming, K. K. Wu, P. H. Moore, unpubl.). 2. Genetic Mapping. Linkage mapping in an autopolyploid is a complex task due to the random pairing of multiple homologous chromosomes, identification of multiple bands by a DNA probe or a pair of primers, and the segregation of alleles with different dosage level. It only became possible when single dose DNA markers were proposed for linkage mapping (Wu et al. 1992). The single dose DNA markers segregate 1:1 (presence : absence) in an F1 population of two heterozygous parents, behaving like a pseudo test cross, or 3:1 in a self-pollinated population of a heterozygous genotype. Fortunately, this type of single dose marker is abundant in the sugarcane genome, accounting for 70% of the detectable polymorphic loci, as expected from the segregation ratios of different plex loci (da Silva et al. 1993). Linkage mapping in sugarcane has been conducted on five populations producing nine linkage maps, mostly based on single-dose DNA markers (Wu et al. 1992). The first sugarcane linkage map was constructed from the progeny of a cross between S. spontaneum (2n = 64) and its doubled haploid (da Silva et al. 1993; Al-Janabi et al. 1993). This map consists of 64 linkage groups assembled into eight homologous groups (HGs) based on 276 RFLPs and 208 single dose (SD) arbitrarily primed polymerase chain reaction (PCR) loci (da Silva et al. 1995). The second map was derived from the progeny of a self-pollinated hybrid cultivar (2n = 107 – 115). This map consists of 408 RFLP loci on 96 linkage groups and ten putative homologous groups (Grivet et al. 1996). More recently, AFLP markers were used to construct another linkage map based on the same population but different individuals, resulting in 120 linkage groups. Thirty-four of the linkage groups could be assembled into ten homologous groups (Hoarau et al. 2001). A map of S. officinarum (2n = 80) is based on a cross with S. robustum and consists of 160 randomly amplified polymorphic DNA (RAPD) markers plus one morphological marker assembled in 51 linkage groups (Mudge et al.

70

MING ET AL.

1996). Four maps were constructed for each of the four parents from two interspecific crosses: S. officinarum (GG, 2n = 97 – 117) × S. spontaneum (IND, 2n = 52 – 56) and S. spontaneum (PIN, 2n = 96) × S. officinarum (MJ, 2n = 140). A total of 72, 69, 72, and 69 linkage groups were assembled from 615, 536, 575, and 418 RFLP markers for GG, IND, MJ, and PIN, respectively (Ming et al. 1998). Two additional maps were constructed for each of the two parents in a third mapping population S. officinarum (LAP, 2n = 80) and S. robustum (MOL, 2n = 80?) with arbitrarily primed-PCR, RFLP, and AFLP markers, resulting in 74 and 65 linkage maps, respectively (Guimarães et al. 1999). A comparative analysis between the sorghum linkage map and the five sugarcane linkage maps indicates that every one of the sugarcane maps is incomplete (Ming et al. 1998). Although the basic chromosome numbers of Saccharum are known, complete linkage maps reflecting these basic chromosome numbers are still not available. Most linkage groups in the nine existing maps represent short chromosomal segments with some linkage groups assembled from as few as two markers. Burdened by the large number of chromosomes to be mapped, all nine maps are partial, with most maps representing less than 50% of the genome of the particular genotype selected for mapping. The five sugarcane maps aligned with the sorghum map collectively covered about 70% of the sorghum genome. Homologybased RFLP markers would have the advantage of inferring homologous group and chromosome alignment with the more extensively studied close relatives sorghum and maize; however, RFLP maps are labor intensive and slow to generate. Adding PCR-based AFLP markers to existing RFLP maps is a cost-effective approach to increase the genome coverage of sugarcane genetic maps. Genome coverage by any of the sugarcane maps is negatively correlated with chromosome numbers to be mapped. The sugarcane genetic map with best genome coverage is that of IND 81146, which has the lowest number of chromosomes (26–28) to be mapped, with about 58% genome coverage (Ming et al. 1998). If sugarcane breeders desire a saturated genetic map of the basic chromosomes, it would be most efficient to work with mapping populations generated from parents having the fewest number of chromosomes. One such population, already partially mapped, derived from a cross between S. officinarum (LAP, 2n = 80) and S. robustum (MOL, 2n = 40) might be worth additional efforts (Mudge et al. 1996; Guimarães et al. 1999). 3. Mapping Quantitative Trait Loci for Economic Traits. Mapping quantitative trait loci (QTL) in autopolyploids is complicated by the potential for segregation of three or more alleles at a locus and by the lack of preferential pairing. As a consequence, different parental alleles of

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

71

autopolyploids are not mutually exclusive alternatives. For the subset of polymorphic alleles that show simplex segregation ratios, the effect of an allele substitution can be estimated from the average phenotypic difference between the two possible genotypes (presence versus absence). Large-scale QTL mapping was conducted in two interspecific populations (Ming et al. 2001; 2002a,b) and in a segregating population from a selfed hybrid R570 (Hoarau et al. 2002). Most QTL alleles for sugar content showed phenotypic effects consistent with the parental phenotypes. However, the occasional transgressive QTLs revealed opportunities to purge unfavorable alleles from cultivars or to introgress valuable alleles from exotics (Ming et al. 2001). In many cases, QTLs controlling a given trait were mapped to corresponding genomic locations within the same genotype, across genotypes, and across species. This complex mapping of a given trait suggests that at least some QTLs on the same cluster might be different forms of the same gene or conserved homologous genes (Ming et al. 2001; 2002a,b). Several QTLs mapped for sugar content correspond to approximate genomic locations of previously mapped maize mutants and QTLs for sugar content. This correspondence between mapped loci for sugar content in sugarcane and maize suggests that stem storage and seed storage crops may share a partly-overlapping basis of genetic variation for carbohydrate storage (Ming et al. 2001). The commercial hybrid R570 revealed small effects of individual QTLs for sugar yield components that were not conserved between crop cycles (Hoarau et al. 2002). Multiplex segregation at QTL loci may be partly responsible for phenotypic buffering that is an important factor in the success of many autopolyploid crops. In several cases, two or more loci detected by the same DNA probe were each associated with variation in sugar content and plant height, enabling us to investigate the possibility of multiplex phenotypic buffering in sugarcane. “Stacking” of multiple doses of chromosomal segments containing favorable QTLs generally produced diminishing effects on phenotype, especially in cases where high-order duplications could be tested (Ming at al. 2001; 2002a). This is similar to the results reported from stacking unlinked QTLs in the diploid tomato. The tomato results were attributed to epistasis (Eshed and Zamir 1996). Evaluating epistasis in sugarcane is complicated by the possibility of non-linear interactions between alleles at homologous loci, in addition to non-linear interactions between unrelated loci (Eshed and Zamir 1996). Detecting this type of phenotypic buffering has potential for cultivar improvement through marker-assisted selection in autopolyploid crops. Although diagnostic DNA markers enable us to pyramid multiple QTLs in a polyploid, incorporating just one copy of the multiple alleles may be sufficient to achieve most of the desired effect in the breeding

72

MING ET AL.

population. Non-additive gene action in multiple dose QTLs also may have contributed to evolutionary opportunities. If a single copy of a gene/QTL is physiologically sufficient, the additional copies are “extra” and thus free to collect mutations, often becoming nonfunctional, but perhaps occasionally resulting in a distinctive new function that improves fitness. 4. Synteny with Other Members of the Grass Family. The conservation of gene repertoire and colinearity of gene order in the genomes of diverse grasses are well established (e.g., Freeling 2001). For sugarcane, the small diploid genome of sorghum has proven an especially facile model. Sorghum is the closest relative of sugarcane and the two grasses diverged from a common ancestor as little as five million years ago. Sorghum and sugarcane genomes share more extensive genome-wide colinearity, and fewer chromosomal rearrangements (Dufour et al. 1997; Guimarães et al. 1997; Ming et al. 1998), than either share with any other known grass. Comparative mapping to establish co-linearity between sugarcane and maize is complicated by segmental polyploidy of the maize genome and the resulting mapping of many sugarcane loci to two duplicated loci in maize (Grivet at al. 1996; Dufour et al. 1997). Although it has not been through a genomic duplication event subsequent to its divergence from sugarcane, rice is much more distantly related and numerous chromosomal rearrangements are found when attempting to align their genomes. Colinearity has been employed to evaluate the correspondence of QTLs affecting related traits in sugarcane and other grasses. Corresponding QTLs controlling plant height and flowering were found in sorghum and sugarcane (Ming et al. 2002a). Several previously mapped maize and rice mutants and QTLs of the sugar metabolic pathway might be candidate genes for controlling sugar content in sugarcane (Ming et al. 2001). Sorghum, rice, and maize linkage maps and physical maps were used to identify potential markers for fine mapping and chromosome walking towards cloning the rust resistance gene in sugarcane (Asnaghi et al. 2000); sorghum RFLP markers played a key role in mapping this gene to a small interval. The close relationship among these grasses, a high degree of co-linearity, and cross-hybridization of DNA probes are compelling reasons for using the more abundant information from the small genome of sorghum to guide molecular mapping and positional cloning in sugarcane. 5. Map-Based Cloning of a Rust Resistance Gene. The first major gene of sugarcane to be mapped was the gene for resistance to brown rust (Puccinia melanocephala H&P Syd.) in ‘R570’ (Daugrois et al. 1996).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

73

Mapping this gene with sugarcane cDNA probe CDSR29 provided the first opportunity to evaluate the potential for map-based cloning in a complex polyploid plant. With support from the ICSB, a bacterial artificial chromosome (BAC) library was constructed with 14 × basic genome or 1.3 × total genome coverage using genomic DNA from R570 (Tomkins et al. 1999). Meanwhile, a fine mapping project began to saturate the region surrounding the rust resistance gene. Using the synteny relationships between sugarcane and sorghum, maize, and rice, and selecting probes in the surrounding regions, this unlinked rust resistance gene was mapped to the end of a linkage group corresponding to sorghum linkage group D (Asnaghi et al. 2000). Bulk segregant analysis added eight markers surrounding the rust resistance gene with the two closest flanking markers placed 1.9 and 2.2 cM from the resistance gene (Asnaghi et al. 2004). Flanking markers were narrowed down to 0.1 and 0.3 cM on each side of the target gene, by chromosome walking using sugarcane, sorghum, and rice BAC resources (A. D’Hont, pers. comm.). Beginning with an unlinked rust resistance gene with a tagged marker 10 cM away to produce a fine mapped target gene flanked by sugarcane BACs, this project demonstrates the rapid advancement of sugarcane genomics that will ultimately be applied by breeders for the benefit of the sugarcane industry. 6. Application of Sugarcane ESTs. The sugarcane EST database is among the largest for monocot species, with more than 300,000 ESTs generated by research groups in Brazil (Vettore et al. 2003), the USA (Ma et al. 2004), South Africa (Carson and Botha 2000), and Australia (Casu et al. 2001). This enormous EST collection has been cataloged, characterized, and deposited in GenBank (see special issue in Genetics and Molecular Biology 2001 Vol. 24 No. 1-4; Carson and Botha 2002; Carson et al. 2002; Casu et al. 2003; Ma et al. 2004; Vincentz et al. 2004). Comparative analysis of 42,982 sugarcane aligned sequences (SAS) with the protein and DNA sequences from Arabidopsis and rice provided the first detailed estimates of the degree of conservation/divergence between a monocot and a eudicot (Vincentz et al. 2004). The 42,982 SASs represent possibly 33,620 unique genes (Vettore et al. 2003). Among them, 70.5% have homologous sequences in Arabidopsis, 2% in other eudicots, 14% in monocots, and 13.5% no matches. The 14% monocotspecific cDNA sequences may represent novel genes or fast-evolving sequences that diverged from their eudicot counterparts. Another noticeable application of the EST resources is the identification of resistance gene analogs (RGAs: Rossi et al. 2003). A total of 88 RGAs were identified based on their sequence homology to typical disease resistance

74

MING ET AL.

genes. These sugarcane RGAs included representatives of the three major groups of resistance genes with a nucleotide-binding site (NBS), leucinerich repeats (LRR), and a serine-threonine (S/T Kinase) domain. Fiftyfive RGAs were used as RFLP probes for genetic mapping and identified 148 single dose loci. Several RGA clusters were found, including one cluster of two loci mapped close to the sugarcane brown rust resistance gene. Detailed sequence analyses of these two RGAs with their rice and maize orthologs suggested a polyphyletic origin. These sugarcane RGAs are a useful resource for identifying and cloning disease resistance genes. C. Molecular Cytogenetics Sugarcanes are characterized by numerous (from 36 to more than 200) small and variably sized chromosomes. Classical cytogenetic studies have been essential in establishing a classification of the Saccharum genus and in understanding the nobilization process (reviewed by Sreenivasan et al. 1987). Modern sugarcane cultivars are derived from a few interspecific crosses performed a century ago between S. officinarum (2n = 80), the domesticated sugar producing species, S. barberi (2n = 81–124), a group of old Indian cultivars, and the wild species, S. spontaneum (2n = 36 to 128). These interspecific crosses were followed by a few backcrosses to S. officinarum clones to recover types adapted to cultivation (Arceneaux 1965; Price 1965). During these crosses, breeders selected the results of a transmission of 2n chromosomes by the female parent, a phenomenon commonly observed in this type of combination (Bremer 1923, 1961) and which facilitated the recovery of clones adapted to sugar production. These crosses were successful in introgressing disease resistance, vigor, and adaptability into the sugar producing lines. This process yielded composite interspecific genomes the complexity of which probably exceeds that of any other major crop. Modern cultivars, which evolved from several generations of recombination since the initial interspecific hybrids, are highly polyploid and aneuploid, with between 100 and 130 chromosomes (Simmonds 1976). A breakthrough in our understanding of sugarcane cytogenetics has been achieved over the last 10 years by using molecular cytogenetics in conjunction with diversity and genetic mapping studies. Molecular cytogenetics was used to determine the origin of S. barberi, a group of canes involved in the origin of modern sugarcane cultivars. In addition, molecular cytogenetics revealed the size of the basic chromosome sets in S. officinarum and S. spontaneum and the genome structure of modern cultivars and related genera. Segregation of molecular markers in progenies was used to analyze chromosome pairing in S. officinarum, S. spontaneum, and modern cultivars.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

75

1. Determination of Basic Chromosomes Numbers. The size of the basic chromosome set (1X) in sugarcane and related germplasm has been actively debated for a long time. Basic chromosome numbers of x = 5, 6, 8, 10, and 12 have been proposed for the Saccharum species (reviewed by Sreenivasan et al. 1987) and the possibility of several basic chromosome numbers in this genus has been suggested. The chromosome number of S. officinarum has been established as 2n = 80. Clones with morphology of S. officinarum but with higher chromosome numbers are considered as atypical or hybrids (reviewed by Sreenivasan et al. 1987). For S. officinarum and its wild progenitor, S. robustum, which exhibits from 60 to 200 chromosomes with major cytotypes of 2n = 60 or 80, the most likely basic chromosome number is x = 10. This is consistent with the most common number in the Andropogoneae tribe (Bremer 1961), and the major cytotypes are more likely to represent euploid forms. For S. spontaneum, which displays a wide range of chromosome numbers from 2n = 36 to 2n = 128 with five major cytotypes: 2n = 64, 80, 96, 112, and 128 (Panje and Babu 1960), the series suggests a basic chromosome number of x = 8. However, because of the high polyploidy and the difficulty of differentiating the chromosomes based on their morphology, these hypotheses could not be tested with classical cytogenetics. Fluorescence in situ hybridization, FISH, of ribosomal gene clusters was used to address the question of basic chromosome numbers in S. officinarum and S. spontaneum, the two species involved in the origin of modern cultivars, either directly or through S. barberi. In plants, the 18S-5.8S-25S (later referred to as “45S”) and 5S rRNA genes are arranged in long tandem arrays of repeat units containing the coding sequences and intergenic spacers. These two multigene families are organized in separate clusters, each one being located at one locus or several loci in the genome (Appels and Honeycutt 1986). The major 45S rDNA sites are usually associated with secondary constrictions and nucleolus organizer regions as opposed to non-expressed minor sites. The 45S rRNA loci consist of many copies (up to several thousand) of a DNA unit of several kbp. The 5S genes are smaller (several hundred bp) and less repeated. Location of the 45S and the 5S rRNA Genes on S. officinarum and S. robustum. The 45S rRNA and the 5S genes were physically mapped by FISH on chromosomes of two S. officinarum (2n = 80) clones (‘Black Cheribon’, ‘BNS3066’) and three S. robustum clones, one with 2n = 80 (‘NG77230’) and two with 2n = 60 (‘Mol4503’, ‘IM76234’). The rationale was to determine the basic chromosome number by dividing the total number of chromosomes by the number of homologous rDNA sites. For both gene families, sites of various intensities were detected, reflecting

76

MING ET AL.

the presence of major as well as minor sites. Eight 45S rDNA sites were localized in the terminal position of eight chromosomes of 2n = 80 clones and six 45S rDNA sites were localized in the terminal position of six chromosomes in 2n = 60 clones. Eight and six 5S rDNA sites were localized at an interstitial position on eight and six other chromosomes in these clones, respectively (Table 2.2 and Plate 2.1a). The most straightforward interpretation for each type of rDNA was for the presence of eight and six homologous rDNA sites borne by eight and six homologous chromosomes for the cytotypes with 2n = 80 and 2n = 60, respectively. This suggested a basic chromosome number of n = 10 for both S. officinarum and S. robustum by dividing the total number of chromosomes by the number of homologous rDNA sites (D’Hont et al. 1996, 1998). Location of the 45S rRNA and the 5S rRNA Genes on S. spontaneum. The 45S and the 5S rRNA genes were located on chromosomes of different S. spontaneum cytotypes. In the various clones studied, the 18S25S and the 5S rRNA genes were localized at an interstitial position on different sets of chromosomes. The cytotypes with 2n = 64 (SES14 and SES 106B) displayed eight sites for both types of genes, whereas the cytotypes with 2n = 80 (NG 51-2 and Mol 5801) displayed ten sites (Plate 2.1b). This was interpreted, for both types of genes, as the presence of one locus per basic genome and 8 and 10 copies for the cytotypes with 2n = 64 and 2n = 80, respectively. The clones ‘Mandalay’ (2n = 96) and Table 2.2.

Number of rDNA sites in the various clones studied.

Clone S. officinarum BNS 3066 Black Cheribon S. spontaneum Haploid of SES208 SES 14 SES106B* Mol 5801 NG 51-2 Mandalay Glagah S. robustum NG77230 Mol 4503 IM 76234

Chromosome number

No. 18S-25S sites

No. 5S sites

Source

80 80

8 8

8

D’Hont et al. 1996 D’Hont et al. 1998

32 64 64 80 80 96 112

4 8 8 10 10 12 10

10 12 14

Ha et al. 1999 D’Hont et al. 1996 D’Hont et al. 1998 D’Hont et al. 1998 D’Hont et al. 1998 D’Hont et al. 1998 D’Hont et al. 1998

80 60 60

8 6 6

8 6 6

D’Hont et al. 1998 D’Hont et al. 1998 D’Hont et al. 1998

4 8

*Clone which must be mislabeled since it is reported with 2n = 48 (Rao and Babu 1955).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

77

‘Glagah’ (2n = 112) displayed 12 and 14 5S rDNA sites, respectively. This was interpreted for the 5S rRNA genes as the presence of one locus per basic genome and 12 and 14 copies for the cytotypes with 2n = 96 and 2n = 112, respectively. However, only 10 45S rDNA sites were detected in the clone Glagah (2n = 112) and 12 45S rDNA sites were observed in ‘Mandalay’ (2n = 96), five of which were tenuous and barely detectable. Ha and co-workers (1999) analyzed a “haploid” S. spontaneum clone of n = 32 derived from anther culture on a 2n = 64 clone and observed four 45S rDNA sites and four 5S rDNA sites. In the various clones studied, minor as well as major 45S rDNA sites were detected, probably reflecting a reduction of the number of repeats and/or suppression of the activity. In clones with very high chromosome numbers, this phenomenon may have been accentuated to the point where sites may have been completely deleted or at least are very difficult to detect. Except in the latter cases, the total number of chromosomes of the different cytotypes is proportional to the number of rDNA sites, in congruence with a basic chromosome number of x = 8 for S. spontaneum (D’Hont et al. 1996, 1998). In summary, these results revealed a basic chromosome number of x = 10 for S. officinarum and S. robustum and a basic chromosome number of x = 8 for S. spontaneum. Two distinct chromosome organizations coexist in modern cultivars. The genetic maps available so far suggest that the parental genomes are colinear and probably differ by only a small number of rearrangements (Grivet et al. 1996; Ming et al. 1998). One such case may reside on homology group (HG) VIII. The 45S rRNA genes were genetically mapped by Grivet and co-workers in cultivar R570 at an interstitial position on S. spontaneum cosegregation groups of HG VIII. This HG comprises two large S. spontaneum cosegregation groups together with two separate sets of smaller S. officinarum cosegregation groups that could not be merged. The structure of the chromosomes of this HG VIII may thus be different in S. officinarum and S. spontaneum. In the three Saccharum species analyzed, the 45S and 5S gene clusters were located on different sets of chromosomes. For each gene cluster and each species, the different sites were located in a constant position within the chromosomes. The 45S sites were terminal for S. officinarum and S. robustum and interstitial for S. spontaneum. The 5S sites were interstitial for the three species. These results were interpreted, for each gene family, as the presence of one gene cluster in several copies, borne by homologous chromosomes. This supports the current hypotheses of autopolyploidy for these species. The basic chromosome numbers, together with the position of 45S rDNA sites and DAPI bands (double stranded DNA stained with 4′,6-diamino-2-phenylindole) within the chromosomes, support classical phylogenetic schemes (reviewed by Daniels and Roach 1987). Meanwhile, molecular marker

78

MING ET AL.

studies (D’Hont et al. 1993; Lu et al. 1994a) also suggested that S. officinarum is directly derived from S. robustum, whereas S. spontaneum is a more distant species. Flow cytometry has been used to estimate the genome size of sugarcane (D’Hont and Glaszmann 2001). The size of the total genome is 7.7 pg (7,440 Mbp) for S. officinarum (2n = 8x = 80), 6.2 pg (5,990 Mbp) for S. spontaneum (2n = 8x = 64), and 11 pg (10,000 Mbp) for a typical modern sugarcane cultivar (R570, 2n = 115). This is much larger than in rice, 860 Mbp (2n = 24); sorghum, 1,600 Mbp (2n = 20); or maize, 5,500 Mbp (2n = 20). However, taking into account the ploidy level, the size of the basic genome (1X), 930 Mbp (0.96 pg) for S. officinarum and 750 Mpb (0.78 pg) for S. spontaneum are close to sorghum, with 800 Mb (x = 10), as compared to 430 Mbp for rice (x = 12) and 2,750 Mbp for maize (x = 10). 2. Origin of S. barberi and S. sinense. Sugarcanes indigenous to North India and China and cultivated from prehistoric times are referred to as S. barberi (2n = 81 – 124) and S. sinense (2n = 116 – 120), respectively. These genera contrast with S. officinarum for floral characteristics and having thinner stalks with lower sucrose content and higher fiber and greater tolerance to biotic and abiotic stresses. Extraction of sugar from these clones was most probably developed in India and China (Daniels and Daniels 1975). Although these clones are no longer cultivated, S. barberi was involved in the origin of modern cultivars. It is thus important when studying the genome of modern cultivars to know the origin and genome structure of S. barberi. Many hypotheses on the origins of S. barberi and S. sinense have been suggested (reviewed by Daniels and Daniels 1975; Paton et al. 1978; Daniels and Roach 1987). The main ones are that: (1) S. barberi and S. sinense arose from introgression of S. officinarum with S. spontaneum in India and China, respectively, in prehistoric times; (2) S. barberi was developed from S. spontaneum in India; and (3) S. barberi and S. sinense arose through introgression between S. officinarum, S. spontaneum, or other genera such as Erianthus and Miscanthus. These hypotheses were tested by genomic in situ hybridization (GISH) performed using S. spontaneum total genomic DNA and S. officinarum total genomic DNA as probes on chromosome preparations of genotypes representative of S. barberi and S. sinense. In all clones analyzed, GISH clearly identified two distinct populations of chromosomes or chromosome fragments, thus revealing the interspecific origin of S. barberi and S. sinense (Plate 2.2a; D’Hont et al. 2002). The in situ hybridization experiments showed a homogeneous labeling intensity of the chromosomes. There were no genomic regions lacking coloration, nor was there a third color pattern. This would not

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

79

have been the case if a third species were involved, especially if it belonged to another genus. For example, GISH performed on intergeneric hybrids between S. officinarum × Erianthus or S. officinarum × Miscanthus showed that total genomic DNA of one genus gave a very weak hybridization signal on the other genus (D’Hont et al. 1995; Piperidis et al. 2000 and unpublished results of these workers). These results are corroborated by the absence of Erianthus or Miscanthus genus specific sequences in S. barberi and S. sinense on Southern hybridization patterns (Alix et al. 1998, 1999). The interspecific origin of S. barberi and S. sinense thus demonstrated by GISH had been long suggested but never proven due to the inability to differentiate chromosomes from parental species by classical cytogenetics. These results, together with recent cytoplasmic (D’Hont et al. 1993) and nuclear molecular marker analyses (Glaszmann et al. 1990; Lu et al. 1994a), are in agreement with the origin of both S. barberi and S. sinense from hybridizations between S. officinarum (female) and S. spontaneum (male). The proportion of chromosomes from the two species in the clones studied was variable, with 61% S. officinarum: 39% S. spontaneum for 2n = 82 clones, 68% : 32% for a clone with 2n = 91, and 66% : 33% for a clone with 2n = 116. From 0 to 4 chromosomes per cell appeared to result from interspecific intrachromosomal exchanges (Table 2.3, Plate 2.2a, D’Hont et al. 2002). Considering the frequency of such exchanges

Table 2.3. Chromosome complement of the North Indian and Chinese sugarcanes studied by GISH.

Taxa S. barberi

S. sinense z

Clones

Groupsz

Chunnee Saretha Xw Pathri Paunra

Saretha-Katha Saretha-Katha

Uba Naguin

Pansahi

Sunnabile-Dhaulu Mungo

2ny

Chromosome complement revealed by GISH 2n = S. off + S. spont + recx

91 91–92 106 82 82

91 = 61 + 29 + 1 91 = 61-63 + 28-29 + 0-2 82 = 45-50 + 26-32 + 3-4 82-83 = 50 + 32 + 0 82 = 47-50 + 32-33 + 0-3

117

116 = 75-76 + 39-40 + 0-2

Classification according to the review of Paton et al. 1978 and Daniels et al. 1991. According to Price 1968. x (2n): Total number of chromosomes, (S. off ): number of S. officinarum chromosomes, (S. spont): number of S. spontaneum chromosomes, (rec): number of interspecific recombinant chromosomes. w This clone referred to as Hullu Kabu in our sample is mislabeled since the observed chromosome number does not correspond to the one reported. y

80

MING ET AL.

in modern cultivars (see below), this indicates that a very small number of meiotic events must have occurred since interspecific hybridization. Further RFLP analyses indicated that the S. barberi and S. sinense clones are clustered into a few groups, each derived from a single interspecific hybrid that has subsequently undergone somatic mutations. These groups correspond quite well with those already defined based on morphological characters and chromosome numbers (reviewed by Daniels et al. 1991). However, the calculated genetic similarities do not support the existence of two distinct taxa. The “North Indian” and “Chinese” sugarcanes thus represent a set of horticultural groups rather than established species (D’Hont et al. 2002). 3. Genome Structure of Modern Cultivars. The contribution of S. officinarum versus S. spontaneum chromosomes to the karyotype of modern cultivars was determined by GISH of total genomic DNA of S. officinarum and S. spontaneum (D’Hont et al. 1996). In ‘R570’ (2n = ca. 115), 10% of the chromosomes appeared to be contributed by S. spontaneum, 80% by S. officinarum, and 10% were clearly derived from recombination between the two species with a few cases of double recombination (D’Hont et al. 1996). In ‘NCo376’ (2n = ca. 112), the distribution of the above categories was 20:70:10 (Plate 2.2b). These results demonstrated for the first time the occurrence of exchanges between S. officinarum and S. spontaneum chromosomes, contradicting the common assumption that no exchange occurs between the chromosomes of the two species (Price 1963, 1965; Berding and Roach 1987). Whether these exchanges occur by translocations or by meiotic recombination was assessed by molecular mapping in the cultivar R570 (Grivet et al. 1996; Hoarau et al. 2001). In these studies, determination of the specificity (S. spontaneum versus S. officinarum) of the linkage groups was attempted through comparison with RFLP or AFLP banding patterns of a set of S. officinarum and S. spontaneum clones involved in the genealogy of current sugarcane cultivars. On this basis, several linkage groups consisted of segments bearing S. spontaneum “specific” markers, as well as segments bearing “specific” markers from S. officinarum, suggesting that interspecific chromosome exchanges had occurred between the two species. These groups display a global colinearity with S. officinarum or S. spontaneum linkage groups, suggesting that these chromosome exchanges occurred through recombinations (as opposed to translocations) between the chromosomes of the two species. Additional GISH analyses on seven cultivars by Piperidis and D’Hont (2001) and on three cultivars by Cuadrado and co-workers (2004) showed that the proportion of complete S. spontaneum chromosomes

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

81

remains between 10% and 20%, while the proportion of recombinant chromosomes between the two parental species varies from 5% to 17%. Collectively, these data allowed us to propose a schematic representation of the organization of cultivar genomes as represented in Fig. 2.3.

S. spontaneum S. officinarum

recombinants

Fig. 2.3. Schematic representation of the genome of current sugarcane cultivars as deduced from FISH and GISH experiments. Each bar represents a chromosome. White and gray bars correspond to S. officinarum and S. spontaneum chromosomes or chromosome segments respectively. Chromosomes from the same row are homologous (or homeologous). The main characteristics of this genome are the high polyploidy, the aneuploidy, the bispecific origin of the chromosomes, the occurrence of interspecific recombinant chromosomes and the existence of structural differences between the chromosomes of the two species.

82

MING ET AL.

4. Related Genera. Saccharum, Erianthus sect. Ripidium, Sclerostachya (2n = 30), Narenga (2n = 30), and Miscanthus (2n = 38 and 40) were assembled into the “Saccharum complex” as a closely related interbreeding group (Mukherjee 1957; Daniels et al. 1975). Erianthus sect. Ripidium includes chromosome numbers of 2n = 20, 30, 40, and 60 with a basic chromosome set of x = 10. They show one 45S rRNA gene cluster and one 5S rRNA gene cluster on two different chromosomes (Jenkin et al. 1995; D’Hont et al. 1995). Miscanthus clones displayed 2n = 38 and 40 chromosomes; Sclerostachya and Narenga have 2n = 30. FISH analysis of one Narenga clone revealed two terminal 45S rRNA sites and four interstitial 5S rRNA sites borne by two pairs of chromosomes of different size (Plate 2.1c). The basic chromosome number for Narenga thus appears to be x = 15, with one 45S rRNA gene cluster and two 5S rRNA gene clusters per basic chromosome set. Sugarcane breeders have long tried to exploit agronomic characteristics of Miscanthus and Erianthus arundinaceus. However, in particular for E. arundinaceus, they not only have difficulties in producing progeny, but it has been nearly impossible to identify true hybrids on the basis of morphological characters and to monitor subsequent generations of introgression. Molecular diagnostic tools including species-specific DNA markers and GISH were developed to overcome the difficulties of identifying true hybrids at the seedling stage and following the introgressed genes (D’Hont et al. 1995; Alix et al. 1998, 1999; Piperidis and D’Hont 2001). GISH allowed analysis of the chromosome complement of intergeneric hybrids involving Erianthus and Miscanthus (Plate 2.2c, D’Hont et al. 1995) and revealed that chromosome elimination occurs in Saccharum × E. arundinaceus hybrids (D’Hont et al. 1995; Piperidis and D’Hont 2001). GISH revealed a high contrast between the chromosomes of the two genera in Saccharum × E. arundinaceus hybrids as compared to that of S. officinarum × S. spontaneum hybrids. Since GISH is based on the presence of species-specific repeated sequences that evolved quickly during speciation, the greater contrast between Saccharum and Erianthus could reflect a greater genetic distance between these two genera than might be expected based on morphological characteristics, which may explain the occurrence of chromosome elimination and the difficulties encountered by breeders attempting to exploit this genus. Several Erianthus and Miscanthus specific repeated sequences have been cloned. Their distribution on the chromosomes was analysed by FISH and revealed two sub-telomeric families (Alix et al. 1998), one centromeric family, and one family apparently dispersed along the genome (Alix et al. 1999).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

83

5. Chromosome Pairing. Modern sugarcane cultivars constitute a particularly complex case of polyploidy. The mode of chromosome pairing influences interspecific recombination and, more generally, phenotypic inheritance. Despite the high ploidy level, both parental species mainly form bivalents at meiosis (Bremer 1929; Sreenivasan 1975). In interspecific cultivars, meiotic metaphase spreads reveal swarms of chromosomes that are difficult to interpret, but show mostly bivalents. The most detailed studies of pollen mother cells in a range of cultivars (Price 1963; Burner 1994) report on average less than one multivalent (tri- or quadri-valent) and three to four univalents per cell. New insights into meiosis were brought about by molecular marker segregations that due to linkages in the repulsion phase allow the identification of chromosome pairing affinities. This approach conducted on progeny from S. spontaneum, although limited by partial genome coverage, suggested polysomic inheritance on the basis of a general lack of linkage in repulsion (Al-Janabi et al. 1993; da Silva et al. 1993). Studies that used an S. officinarum parent detected some instances of repulsion, indicating preferential pairing between a subset of chromosomes (Al-Janabi et al. 1994; Ming et al. 1998). The first segregation studies conducted in a hybrid cultivar, namely R570 (Grivet et al. 1996; Hoarau et al. 2001), were based on unsaturated RFLP and AFLP genetic maps that revealed several instances of repulsions between chromosomes, indicating that these are strongly associated at meiosis. At least two of these couples displayed systematic pairing. The chromosomes involved appeared to be of mixed origin and several instances of pairing between regions derived from the two ancestral species were detected. An approach based on RFLP analysis using a particular probe (BNL12.06) was chosen because it enabled access to all chromosomes of a given HG and characterized their segregation according to their dosage (Jannoo et al. 2004). Eleven alleles were identified that correspond to the 11 alleles detected by Grivet and co-workers (1996). These alleles belong to 11 cosegregation groups, of which 10 are homologous (or homoeologous when inherited from S. officinarum and S. spontaneum), for they share markers derived from the same series of RFLP probes. The eleventh group clearly belongs to another HG. The species specificity of these alleles, as well as of markers closely linked to them, resulted in the tentative assignment of an S. officinarum origin to five of the 10 tagged chromosomes, an S. spontaneum origin to one chromosome, and a mixed origin to two. The rest of the chromosomes were of unclear origin. Given that the single dose restriction fragments are present as a single dose in the parent, the dosages in progeny individuals derived from

84

MING ET AL.

self-pollination can either be zero (no band), one, or two. A description of most individuals by the number of doses (0, 1, or 2) of all 10 alleles present was obtained by analyzing the RFLP patterns. Estimation of the frequency of pairing between two chromosomes bearing two alleles was performed using a maximum likelihood method. For each pair of alleles, the most likely pairing frequency could be estimated (Table 2.4), with a standard deviation close to 6%. The pij values are between 0 and 39.1%. Null values highlighted sets of chromosomes that never pair. Values higher than the 11% (1/9) that would be expected from random assortment identify couples that tend to pair preferentially. Although this preferential pairing (more frequent than random) is observed often, there is no case of systematic pairing, which would be identified by a value of 100%. The total value for a given allele should be 100% if it were involved in a bivalent, irrespective of the specificity of the association. Lower values indicate pairing with heterologous chromosomes, no pairing (univalent), or pairing in a trivalent. The totals for one allele ranged from 39.0% to 95.2%. The mean of the total for the 10 alleles is 75.1% (Table 2.4); this figure, compared to 100%, characterizes a global deviation from a typical 5 × 2 bivalent meiosis caused by particular cases such as univalents or trivalents. The differentiation within the population of chromosomes is illustrated by the results of a factor analysis or the mutual pairing frequenTable 2.4. Maximum likelihood estimates for the pairing frequencies pij (percent) between the chromosomes as revealed from allele segregation patterns at locus BNL12.06. Allele 1

2

3

4

5

6

7

8

9

10

o

Total

70.3

751.2

Originz Allele

u

u

u

o

o

s

i

o

i

2 3 4 5 6 7 8 9 10 Total 1 to 10

27.4 39.1 0.0 0.0 1.9 6.6 3.2 0.8 2.8

0.0 4.4 12.4 13.2 13.3 8.6 0.0 3.8

1.1 12.3 15.6 9.2 1.1 9.3 7.5

0.0 4.4 11.6 13.5 4.5 26.3

18.0 22.5 0.0 8.1 0.0

22.4 0.5 7.7 5.5

4.0 2.6 1.4

6.0 23.0

0.0

81.8

83.1

95.2

65.8

73.3

89.2

93.6

59.9

39.0

z Suspected origin of chromosome segment, either S. spontaneum (s), S. officinarum (o), interspecific (i), or unclear (u).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

85

cies. The distribution of the chromosomes along these axes is summarized in Fig. 2.4. The global interpretation is that the chromosomes are differentiated into several subsets: • One chromosome (I-9) seems often left as a univalent. This chromosome has a clear interspecific origin. However, there is no proof this is the cause of its isolation at meiosis.

Axis 2 (18%) X1 X3 X2

10 8

Axis 1 (23%) 9

4

6

7

: S. officinarum

: S. spontaneum

5

: hybrid

X : unclear

Axis 3 (14%) X2 7 X1

8 10

4

6

5

Axis 1 (23%)

X3 9 Fig. 2.4. Distribution of ten alleles representing ten interacting chromosomes in planes (1,2) and (1,3) of a factor analysis that summarizes the matrix of distances (1-pij), where pij is the frequency of pairing estimated from 282 individuals. The specificity of the chromosome segments is indicated.

86

MING ET AL.

• Three chromosomes (I-4, I-8, I-10) seem to be involved in preferential mutual pairing. Whether these chromosomes assemble into a trivalent or a bivalent plus a univalent cannot be determined from the present data; all probably have a pure S. officinarum origin. • Within a subset of six chromosomes, one chromosome (I-1, of unclear origin) seems to preferentially form a bivalent with one of two particular chromosomes (I-2 and I-3, of unclear origin), while the four chromosomes left form two random bivalents. This subset includes chromosomes or chromosomal regions originating from the two ancestral species. The (at least) two couples of chromosomes mentioned earlier that form systematic pairs (Hoarau et al. 2001) and the variation in pairing frequencies observed here collectively illustrate a remarkable range of affinities among chromosomes. Homolog juxtaposition is thus clearly determined by some features of the individual chromosomes that seem partly, although incompletely, related to the species origin of the chromosomes. Specific understanding of the structure of the genome, i.e., the basic chromosome number, ploidy level and rate of recombination, and the pairing dynamics at meiosis will be essential for drawing the best information from the results of genetic studies such as QTL analyses for complex traits. In cases of disomic inheritance, one essentially compares the value of the two alleles borne by the regions that segregate in repulsion phase. In cases of polysomic inheritance, the value of a particular allele is basically compared to the average value of all other alleles, with little or no inference about the mode of gene action in terms of additivity or dominance. Intermediate situations such as those that characterize sugarcane require a specific investigation for each HG of the respective affinities among the various homo(eo)logous chromosomes. Global QTL analyses based on a general scanning of the genome probably cannot provide the resolution that would be needed for this. However, it seems important that candidate QTLs be a subject of additional studies to resolve as many alleles as possible at the corresponding linked marker loci and to assess the mode of assortment so that all useful parameters can be estimated in view of subsequent applications. D. Genetic Engineering Plant pathologists, agronomists, entomologists, and plant breeders have expended enormous effort towards producing improved sugarcane cultivars over the past century. Development of new cultivars through tra-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

87

ditional breeding methods relies upon observations of plant phenotypes or genotypes and studies on their inheritance. Plant breeders have identified numerous beneficial traits and, through controlled breeding programs, have transferred the genes that encode these traits into cultivars. However, in the process of transferring desirable characteristics, undesirable traits also can be transferred. Breeders must then perform numerous, time-consuming crossings that may or may not produce the desired phenotype. The effort to transfer a single disease resistance gene, for example, and to separate this gene from undesirable genes that have accompanied it, may take from 5 to 15 years. Recombinant DNA techniques and transgenic plant technologies can be integrated as one method by which to overcome this limitation of conventional breeding techniques, enabling the rapid transfer of one to several genes from one species to another or from other organisms into sugarcane. 1. Sugarcane: a Suitable Crop for Genetic Engineering. Sugarcane has several features that make it an ideal candidate for improvement via genetic engineering. First, genetic improvement of elite sugarcane clones by conventional breeding is difficult due to the complex aneuploid, polyploid genome, variable fertility and genotype × environment interactions. Traditional back-crossing to recover elite genotypes with desired agronomic traits is difficult in sugarcane. In this context, genetic engineering could be a very valuable tool to introduce commercially important traits into elite germplasm. Second, reasonably efficient sugarcane transformation systems are available (Birch 1997; Arencibia et al. 1998; Enriquez-Obregon et al. 1998) and useful transgenic lines can be maintained indefinitely by vegetative propagation. In addition, unlike other food crops, genetically engineered sugarcane could be accepted by the public and regulatory authorities, as refined white sugar is the most chemically pure food derived from plants and has been found to be free from DNA and proteins expressed from the introduced transgene (Klein et al. 1998; Taylor et al. 1999). 2. Initial Genetic Transformation Technologies. In the last decade, substantial research effort has been expended to develop efficient genetic transformation systems for sugarcane (Chen et al. 1987; Bower and Birch 1992; Rathus and Birch 1992; Gambley et al. 1993; Birch 1997; Arencibia et al. 1998; Enriquez-Obregon et al. 1998). Different transformation techniques such as electroporation (Rathus and Birch 1992), polyethyleneglycol (PEG) treatment (Chen et al. 1987), and particle bombardment (Bower and Birch 1992) were used to introduce marker genes in sugarcane cells and calli. The first transgenic sugarcane cell cultures or calli

88

MING ET AL.

were obtained following PEG-mediated DNA transfer into protoplasts of cultivar F164 (Chen et al. 1987). Despite being a simple procedure requiring little specialized equipment, PEG-mediated transformation did not receive further attention due to its low efficiency and poor reproducibility. A transformation efficiency of one per 106 treated protoplasts was reported for the cultivar F164 (Chen et al. 1987). Sugarcane transformation by electroporation was found to be more efficient and reproducible than the PEG method. Stably transformed calli have been obtained from electroporated protoplasts of different sugarcane cultivars (Rathus and Birch 1992; Chowdhury and Vasil 1992). Rathus and Birch (1992) documented one transformation event per 102 to 104 electroporated protoplasts for the Australian cultivars Q63 and Q96, a frequency sufficient for the useful generation of transgenic plants. However, a lack of regeneration from protoplasts prevented production of transgenic plants with this approach as well. Nonetheless, transgenic sugarcane plants have been obtained following electroporation of intact embryogenic cells (Arencibia et al. 1995) and of meristematic tissues of in vitro grown plants (Arencibia et al. 1992). 3. Recent Advances in Sugarcane Genetic Engineering. Transformation Systems. Improvements in the genetic transformation of sugarcane have been considerable since the first report of the production of transgenic callus expressing the gene for resistance to the antibiotic chloramphenicol via electroporation of protoplasts (Chen et al. 1987) and apparently the first production of transgenic sugarcane plants expressing a chimeric beta-glucuronidase (GUS) as a reporter (Irvine and de Almedia 1991) (Fig. 2.5). These plants were produced using plasmid-coated tungsten particles shot into sugarcane callus with an airless paint sprayer. Improvements in biolistic methods and fine-tuning of tissue culture methods (see Birch 1997) has allowed transformation and regeneration of sugarcane plants to become fairly routine in many laboratories around the world. However, many leading elite cultivars are recalcitrant to regeneration, and no transgenic versions of these cultivars have been produced to date, even though considerable time and effort have gone into the endeavor (T. E. Mirlov, unpubl.; M. Irey, pers. comm.). Beginning in about 2000, several groups doing research that relies on transgenic sugarcane have been exploring new methodologies to improve the efficiency of transformation using biolistic procedures. In the past, the most common biolistic method relied on somatic embryogenesis from callus, which can lead to somaclonal variation and phenotypic off-types. In a recent field study (Vickers et al. 2005a), transgenic plants produced from callus showed significantly lower yield compared to non-transformed tissue culture-derived plants. Interestingly, this

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

Fig. 2.5.

89

Timeline for the major milestones in genetic transformation of sugarcane.

study also showed that some components of the transformation process, other than tissue culture, are at least in part responsible for the yield loss. However, to minimize the incidence of somaclonal variation, a method for rapid and high efficiency regeneration of commercially important sugarcane cultivars was recently developed and demonstrated with a

90

MING ET AL.

diverse collection of sugarcane genotypes (Geijskes et al. 2003). With this method, using 1–2 mm thick transverse sections of leaf rolls as explants, shoots can be regenerated directly either through adventitious shoot meristems or by somatic embryogenesis within 4–5 weeks. This type of regeneration system formed the basis for a new biolistic transformation method that avoids the callus stage, which is now in place in several laboratories (Snyman et al. 1998; Geijskes et al. 2003). Not only does this method of transformation have the potential to reduce somaclonal variation and phenotypic off-types, it greatly reduces the amount of time from the initial bombardment to the regeneration of plantlets. More importantly, this method also has the potential to generate transgenic versions of elite sugarcane cultivars for which regeneration via somatic embryogenesis after a callus stage has not been possible. One drawback of the direct somatic embryogenesis method is that it can produce chimeric plants (Barbara Huckett, pers. comm.) Utilizing the method of somatic embryogenesis from bombarded callus, several groups (Vickers et al. 2000; Liebbrandt and Snyman 2003; J. E. Irvine and T. E. Mirkov, unpubl.) have demonstrated via field trials that it is possible to produce transgenic sugarcane plants that have yields equivalent to or higher than the starting non-transgenic version. However, these types of events account for only about 20 to 50% of the independent transformants (T. Erik Mirkov, unpubl.). Therefore, using transgenic cultivars as parents in breeding programs may have some advantages over the direct transformation of existing elite cultivars for certain traits. Transformation of an existing cultivar requires several years of field evaluation to ensure that the introduced gene is stably expressed and that the transformation process induces no other detrimental phenotypic effects. In contrast, large numbers of transgenic progeny can be produced from crosses involving transgenic parents and can be evaluated during the normal stages of a selection program. No extra evaluation is required, which could save several years in the cultivar development program. This, however, is dependent on the stable transmission and expression of transgenes from parents to offspring and the breeding value of the parent genotypes. Butterfield and co-workers (2002) used transgenic sugarcane parents containing multiple copies of herbicide resistance and virus resistance genes as males for crossing with non-transgenic sugarcane clones. Segregation of the transgenes in the progeny was determined using Southern blot analysis; herbicide resistance and virus resistance was assessed using bioassays. The segregation data were used to infer linkage relationships between transgenes in the parent plants. In two of the parents, all transgene insertions were linked in one position in the genome,

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

91

although some recombination between insertion events did occur. In the third parent, transgene insertion had occurred in two independent, unlinked loci. Analysis of progeny of this parent indicated that rearrangement or mutation occurred in both loci, resulting in non-parental transgene DNA fragments in some progeny. Most transgenic progeny containing the bar gene maintained resistance to herbicide. This work has demonstrated that transgenic sugarcane parents showing stable inheritance of transgenes can be effectively used in breeding programs (Butterfield et al. 2002). Work by Liebbrandt and Snyman (2003) has clearly demonstrated that transgenes for herbicide resistance also are expressed stably under field conditions during multiple cycles of vegetative propagation. Furthermore, this work demonstrated that morphological and agronomic characteristics such as stalk height, diameter, population, fiber, disease resistance, and yield were the same in the transgenic line and its nontransgenic counterpart when measured in the first ratoon. Regardless of the route of somatic embryogenesis, biolistic transformation invariably leads to the genomic insertion of multiple copies of the transgene cassette, which in turn can lead to gene silencing. Agrobacterium-mediated transformation of sugarcane may be another alternative to reduce somaclonal variation and phenotypic off-types and for overcoming gene silencing. Agrobacterium-mediated transformation of sugarcane has been achieved (Arencibia et al. 1998; Elliott et al. 1998; Enriquez-Obregon et al. 1998), but it is far from being routine in most laboratories. Nevertheless, recent work by Manickavasagam et al. (2004) demonstrated that highly efficient Agrobacterium-mediated transformation of sugarcane axillary buds is possible. The authors reported 50% transformation efficiency and a protocol that eliminates chimeric plants. If this method can be easily reproduced in other laboratories and with different elite sugarcane cultivars, this will indeed be a tremendous advancement in the methodologies available for generation of transgenic sugarcane. Promoters. In addition to an efficient transformation system, the availability of a range of promoters with varied strengths and tissue specificities is critical to the continued success of transgenic approaches to sugarcane improvement. Historically, the number of promoters available for sugarcane transformation has been very limited. Of the promoters initially tested in sugarcane, the maize ubiquitin 1 promoter (Christensen and Quail 1996), the rice actin 1 promoter, and the synthetic Emu promoter (Last et al. 1991) were shown to drive higher levels of reporter gene expression than the CaMV 35S promoter (McElroy et al. 1991;

92

MING ET AL.

Gallo-Meagher and Irvine 1993; Rathus et al. 1993). More recent publications report on the use in sugarcane of promoters derived from plant viruses and on the use of sugarcane and rice ubiquitin promoters (Schenk et al. 1999; Wei et al. 1999; Schenk et al. 2001; Liu et al. 2003; Wei et al. 2003). It appears that the rice ubiquitin promoter will be useful in sugarcane transformation. In production of transgenic sugarcane plants via bombardment of sugarcane callus, reporter gene expression levels driven by the rice ubiquitin promoter were nearly two-fold over those levels driven by the maize ubiquitin 1 promoter (Liu et al. 2003). Although the two sugarcane ubiquitin promoters drove high levels of reporter gene expression in sugarcane callus, this expression dropped to very low or undetectable levels in the resulting transgenic plants (Wei et al. 2003). This study also determined that the drop in expression was due to posttranscriptional gene silencing (PTGS), indicating that these promoters may still be useful if PTGS can be avoided. Recently, Yang and co-workers (2003) isolated and characterized sugarcane promoters from a gene for a sugarcane elongation factor 1a and a sugarcane proline-rich protein-encoding gene. Expression levels of a reporter gene driven by these promoters in sugarcane callus were essentially equivalent to those of the maize ubiquitin 1 promoter. However, expression studies in transgenic plants were not conducted. Researchers at BSES Limited have recently cloned and characterized a promoter from a sugarcane homolog of the sugarbeet Hs1pro1 gene that confers resistance to the beet cyst nematode. In tobacco, the promoter drives constitutive reporter gene expression; however, no reporter gene product could be detected in sugarcane, possibly due to gene silencing (Hermann et al., pers. comm.). To further improve the range of promoters useful in sugarcane, several promoters have been isolated from viral sources. For instance, a promoter from Banana Streak Virus drove GFP expression up to three-fold higher (1.06% of total soluble leaf protein) than the expression levels obtained using the maize ubiquitin 1 promoter (Schenk et al. 2001) in sugarcane. Braithwaite et al. (2004) amplified three Sugarcane Bacilliform Virus (SCBV) promoter regions from three separate SCBV infected S. officinarum plants and cloned a fourth one from a purified SCBV isolate. The promoter obtained directly from the virus isolate drove reporter gene (GUS) expression in leaves, meristems, and roots of plants grown in vitro at levels equal to or higher than the maize ubiquitin promoter. In terms of tissue-specific promoters for use in sugarcane, the molecular biologist’s “tool box” is virtually empty. Although there are several groups pursuing this line of research, the only published work on tissue specific promoters in transgenic sugarcane is from Hansom and co-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

93

workers (1999), who cloned a stem-specific promoter, UQ67P, from sugarcane and showed that it drove strong, tissue specific GUS expression in sugarcane. This, however, was found not to be repeatable for driving expression of other transgenes in sugarcane ‘Q117’, possibly because of promoter silencing or PTGS (S. Brumbley, pers. comm.). An ongoing effort is underway by CSIRO Plant Industries and BSES Limited, funded by the Cooperative Research Centre for Sugar Industry Innovation through Biotechnology, to identify tissue-specific promoters that will be useful for the biofactory research program. Ongoing work at the South African Sugar Research Institute (SASRI) is involved in developing a suite of tissue-specific promoters for sugarcane transformation (B. Huckett, pers. comm.). They have a number of test lines in the field; new promoter constructs are being delivered to sugarcane callus, and others are in construction. More recent work at SASRI is focused on accessing the promoter sequences from maize or sorghum in an attempt to avoid silencing. Recent work at Texas A&M University has identified and isolated from sugarcane two stem specific promoters that drive high levels of reporter gene expression in stems of transgenic sugarcane when reintroduced back into the original cultivar (T. E. Mirkov, unpubl.). 4. Development of Commercially Useful Transgenic Sugarcane. Following the development of a microprojectile transformation system (Bower and Birch 1992), efforts were focused on engineering economically important traits into commercially grown sugarcane cultivars. In the last few years, transgenic sugarcane plants with improved resistance to Sugarcane Mosaic Virus (SCMV) (Joyce et al. 1998), Sorghum Mosaic Virus (SrMV) (Ingelbrecht et al. 1999), Fiji leaf gall (McQualter et al. 2001), leaf scald disease (Zhang et al. 1999), cane grubs (Nutt et al. 1999), Mexican rice borer and sugarcane borer (Legaspi and Mirkov 2000; Sétamou et al. 2003), and herbicide (Gallo-Meagher and Irvine 1996; Enriquez-Obregon et al. 1998; Leibbrandt and Snyman 2003) have been produced. Efforts also are underway to engineer sugarcane for lowcolor, raw sugar and high-value products (Groenewald et al. 1995; Brumbley et al. 2002). Although currently published field data are available only for sugarcane plants engineered to resist sugarcane borer (Arencibia et al. 1999), Mexican rice borer (Legaspi and Mirkov 2000) infestation, and herbicide resistance (Liebbrandt and Snyman 2003), glasshouse trial results obtained to date with transgenic sugarcane plants resistant to leaf scald (Zhang et al. 1999), SCMV (Joyce et al. 1998), SrMV (Ingelbrecht et al. 1999), cane grubs (Allsopp et al. 2000), and herbicide (GalloMeagher and Irvine 1996) are encouraging. For instance, the ‘Q117’ transgenic lines engineered with the potato proteinase inhibitor II or

94

MING ET AL.

snowdrop lectin genes exhibited marked antibiosis to the cane grub species Antitrogus consanguineus (Allsopp et al. 2000). Similarly, a significant increase in resistance to SCMV was shown by plants engineered with the SCMV coat protein gene in glasshouse trials (Joyce et al. 1998). Other examples of transgenic sugarcane for herbicide resistance, insect resistance, viral resistance, bacterial resistance, genes that improve sugar purity and color, and genes for increasing sucrose yields that are in glasshouse or field include: Gallo-Meagher and Irvine 1996; Arencibia et al. 1997; Irvine and Mirkov 1997; Snyman et al. 1998; Falco et al. 2000; Vickers et al. 2000, 2005b; Botha et al. 2001; Groenewald and Botha 2001; Nutt et al. 2001; Butterfield et al. 2002; Falco and Silva 2003; Sétamou et al. 2003; McQualter et al. 2004. 5. Transgene Silencing. Scientists have made significant progress in the development and application of recombinant DNA techniques and transgenic plant technologies for the improvement of sugarcane. However, in comparison to the current status for some of the major monocot crops and dicot plants, there is much room for improvement. Although many laboratories are able to obtain transgenic sugarcane plants on a regular basis, a majority of these plants become “silenced” for the expression of the transgenes. This has been shown both at the transcriptional (promoter methylation) and posttranscriptional (mRNA degradation) levels. One area of research that could lead to a solution to these problems would be to develop methods that result in the generation of hundreds of independent transgenic events per shot. For some cultivars, it has been possible in some laboratories to obtain 10 transgenic events per shot; in other laboratories, the best experiments result in one transgenic event for every 10 to 20 shots for certain elite cultivars that can be regenerated from callus. However, a sheer numbers, brute force approach is unlikely to succeed. We need to understand the pathways that lead to promoter silencing and PTGS in sugarcane. By unraveling the pathway(s) of gene silencing, we will identify the key sugarcane proteins involved in this process. This knowledge may then be applied toward improved strategies for engineered virus resistance, enable us to increase our success in expressing transgenes by inhibiting silencing, especially from native sugarcane promoters, and to improve our success in intentional silencing of native genes to eliminate undesirable traits. It is possible that gene silencing may be overcome through the use of plant viruses, or coexpression of certain plant viral genes that encode suppressors of PTGS (Marathe et al. 2000; Moissiard and Voinnet 2004). The p19 protein of Tomato Bushy Stunt Virus has been successfully used for this purpose in tobacco (Voin-

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

95

net et al. 2003). Expression levels of a range of proteins were increased up to 50-fold or more in the presence of Tomato Bushy Stunt Virus p19. Work at Texas A&M University indicates that SrMV and Sugarcane yellow leaf virus both encode strong suppressors of PTGS and that it may be possible to use these suppressors to prevent PTGS in transgenic sugarcane (T. E. Mirkov, unpubl.). 6. Sugarcane Biofactory Research. Sugarcane has the potential to be a key crop in the biological revolution that will gradually shift the world’s economies from being petrochemical-based to carbohydrate-based. Sugarcane is a fast-growing tropical grass, produces a large biomass, partitions carbon into sucrose at up to 42% of the dry weight of the stalk, and has a mobile pool of hexose sugars through most of its life cycle. It is difficult to produce viable seed from commercial cultivars, which greatly reduces problems associated with genetic drift from an agricultural crop engineered to produce industrial chemicals. In addition, sugarcane regrows from the underground portions, or stools, left in the ground after harvesting and therefore can be harvested multiple times before replanting, saving reduced plow-out and replanting costs. At least four or five groups are actively involved in the development of sugarcane as a biofactory for the transgenic production and subsequent purification of large quantities of high-value proteins, plastics, and carbohydrates. Work conducted at the Hawaii Agriculture Research Center in Aiea, Hawaii (Wang et al. 2005) is focused on using transgenic sugarcane to produce human granulocyte macrophage colony stimulating factor (GM-CSF), which is used in clinical applications for the treatment of neutropenia and aplastic anemia. This work has culminated in the first field trial in the U.S. using transgenic sugarcane to produce a human pharmaceutical protein. In this study, accumulation of the GM-CSF protein was up to 0.02% of total soluble protein. Research at Texas A&M University is aimed at using sugarcane to produce pharmaceutical-grade human structural proteins for human therapeutic uses (Holland-Moritz 2003). To test the ability of sugarcane to be a biofactory for the production of high-value products other than proteins, Brumbley and co-workers (2002) chose to engineer the genetic pathway for poly-3-hydroxybutyrate (PHB) (Schubert et al. 1988; Peoples and Sinskey 1989a,b) into sugarcane. Polyhydroxyalkanoates (PHAs), of which PHBs are one class, are polyesters of 3-hydroxyacids used by many bacterial genera as carbon storage compounds (Steinbüchel and Valentin 1995; Madison and Huisman 1999; Rehm and Steinbüchel 1999). Because PHAs have thermoplastic properties and are biodegradable, they are attractive alternatives

96

MING ET AL.

to petrochemically derived plastics. To reduce the costs to produce these compounds by fermentation and thereby make these plastics economically competitive with petrochemically derived plastics and to maximize environmental benefits, attempts have been made to produce PHAs in plants. These efforts have been recently reviewed (Poirier 1999; Poirier 2002; Snell and Peoples 2002). The three-gene pathway, phaA, phaB, and phaC, has been successfully engineered into a number of plant species. However, either levels of PHB accumulation were low or one or more of the products from the PHB biosynthetic pathway had adverse effects on the transgenic plants (Poirier et al. 1992; Nawrath et al. 1994; Slater et al. 1999; Bohmert et al. 2000; Lössl et al. 2003). Brumbley and co-workers (2002) targeted the products from the Ralstonia eutropha PHB biosynthetic pathway to several subcellular compartments of sugarcane. In the best producing line, polymer accumulated in the leaves at 1.2 % of dry weight, 0.4% in the stem rind, and 0.004% in the stem pith (S. Brumbley, pers. comm.). Analysis of height, weight, and sugar levels revealed no significant difference between transgenic and wild type lines. Sugarcane also was evaluated as a production platform for a major component of liquid crystal polymers (LCP), p-hydroxybenzoic acid (pHBA), using two different bacterial proteins, a chloroplast-targeted version of Escherichia coli chorismate pyruvate-lyase (CPL) (Siebert et al. 1996) and a Pseudomonas fluorescens 4-hydroxycinnamoyl-CoA hydratase/lyase (HCHL) (Gasson et al. 1998). Both provide a one-enzyme pathway from a naturally occurring plant intermediate. The substrates for these enzymes are chorismate, a shikimate pathway intermediate that is synthesized in plastids (Herrmann and Weaver 1999), and 4hydroxycinnamoyl-CoA, a cytosolic phenylpropanoid intermediate (Mayer et al. 2001). Siebert et al. (1996) demonstrated that pHBA can be produced in tobacco plants and tobacco cell cultures using a chloroplasttargeted version of CPL. Mayer et al. (2001) demonstrated that HCHL could be used to produce pHBA in transgenic tobacco; however, the plants had numerous physiological anomalies. McQualter and co-workers (2005) demonstrated that HCHL is the superior catalyst for production of pHBA in sugarcane leaf and stem tissue. p-Hydroxybenzoic acid was quantitatively converted to glucose conjugates by endogenous UDP-glucosyltransferases and was presumably stored in the vacuole. The largest amounts detected in sugarcane leaf and stem tissue were 7.3 % and 1.5 % dry weight, respectively, yet there were no discernible phenotypic abnormalities in any of the sugarcane lines tested (McQualter et al. 2005). However, as a result of diverting carbon away from the phenylpropanoid pathway, there was a reduction in leaf chlorogenic acid, subtle changes in lignin composition,

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

97

and an apparent compensatory upregulation of phenylalanine ammonialyase (McQualter et al. 2005). In addition to pHBA, vanillic acid also was produced in sugarcane plants expressing the HCHL genes (McQualter et al. 2005). Other work in Australia is focused on developing sugarcane as a production platform for a range of higher-value sugars, other biopolymers, and edible vaccines (S. Brumbley, pers. comm.). Biofactory research being carried out at SASRI involves some work on novel carbohydrate production in transgenic sugarcane. Unfortunately, public information on this technology is limited at this time (B. Huckett, pers. comm.). Using sugarcane as a biofactory is an exciting area of research that could have a tremendous impact on the sugarcane industries around the globe. However, there are many hurdles that must be overcome before this can become a commercial reality. To be competitive with commercial protein production systems that use corn, alfalfa, rice, or tobacco for the production of high-value proteins, we will need to achieve much higher levels of protein expression in the transgenic sugarcane stalk. This will require the continued identification of new promoters, development of viral vectors, and overcoming both transcriptional and posttranscriptional gene silencing. Furthermore, the details of how best to extract and purify proteins on a large scale from plants are not trivial either. Experience in this area, even with large companies (particularly for sugarcane) is very scarce. Biotechnology will not replace conventional sugarcane breeding, but it is another tool that can be used in conjunction with marker-assisted selection and traditional breeding for the continued improvement of sugarcane. E. Coordinating International Progress in Sugarcane Improvement and Linking Biotechnology to Application: the ICSB A July 2004 survey of more than 1,100 registered members of sugarcane’s professional society, the International Society of Sugar Cane Technologists (ISSCT), showed that 77 individuals identified themselves as a “breeding technologist” located at one of 52 sugarcane research organizations (SRO) in 41 sugarcane producing countries of the world (pers. comm. of Dr. C. Ricaud, Permanent Secretary of the ISSCT). This statistic is evidence that SROs worldwide recognize the need for continued development of new sugarcane cultivars and are expending considerable resources in breeding and selection programs towards this goal. Throughout this chapter, we have referred to sugarcane’s large polyploid genome that greatly curtailed the use of genetic principles in breeding for the improvement of this crop. Without knowing the ploidy

98

MING ET AL.

level or type, there could be no prediction of phenotype segregation and thus no characterization of sugarcane genes. Nevertheless, for the first 100 years of sugarcane breeding, the principles of quantitative genetics used to resolve issues about heritability of yield traits, coupled with the prevailing philosophy of evaluating the maximum number of progeny that one can afford, was sufficient to make good genetic advancements. The first 100 years of sugarcane improvement through hybridization was very successful in increasing crop yields. Every country around the world that invested in developing locally adapted cultivars experienced yield increases that paid for the investment in supporting local breeding programs. However, after a century of success, indications arose that dependence on brute force production of millions of seedlings to be serially eliminated by screening phenotypes for performance is producing fewer new cultivars, with smaller yield increases, at ever increasing costs. The diminishing return on breeding program investments indicated a need to evaluate the potential for using the evolving technologies of molecular genetics and biotechnology for improving sugarcane germplasm. It should be noted that countries of the tropical zone where sugarcane is grown are, on average, poorer, less developed, and with a lower level of science and technology than countries at higher latitudes. One consequence of this development disparity is that crop geneticists and breeders of the temperate zone had early access to the technologies of molecular biology and were able to apply them for the improvement of temperate crops such as tomato, maize, and soybean as early as the late 1980s. On the other hand, the tropical crop plant breeders had less access to new tools and technologies and were thus slower to investigate molecular approaches for the improvement of sugarcane. The International Society of Plant Molecular Biology, formed in 1983, held its first congress in Savannah in 1985. The meeting highlighted molecular genetics research on “model plants” that gave hope that such approaches might work with sugarcane. In 1988, Paul Moore, then head of the Plant Physiology Section of the International Society of Sugar Cane Technologists, arranged with his committee member, James Irvine, to hold an ISSCT joint Plant Breeding and Physiology Workshop in Honolulu. U.S. plant molecular geneticists were keynote speakers at that workshop, which was intended to fuel interest in sugarcane scientists in the potential of the molecular approach to crop improvement. During that workshop, Moore and Irvine arranged meetings between experiment station directors Don Heinz of the Hawaii Sugar Planters’ Association (HSPA) and Manoel Sobral of the Copersucar Technology Center (CTC) of Brazil, with the purpose of collaboratively funding research by Steven Tanksley and Mark Sorrells of Cornell University to

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

99

evaluate the feasibility of using RFLP markers to map the sugarcane genome. The HSPA/CTC pact included an arrangement for each institution to place one of its staff members in the Cornell laboratory to work on the project to facilitate the transfer of acquired technologies back to their respective industries. By late 1990, K. K. Wu (HSPA) and William Burnquist (CTC) had worked more than a year at Cornell and developed the concept that complex polyploids such as sugarcane could indeed be genetically mapped if one restricted the analysis to only single or double dose DNA restriction fragments that could be identified by their fit to expected diploid segregation ratios (Wu et al. 1992). This breakthrough looked so promising that Irvine arranged a First International Workshop on Sugarcane Genome Analysis in Beltsville, Maryland, in March of 1991. That workshop had the participation of the original HSPA/CTC directors and scientists plus directors from five additional SROs: American Sugar Cane League (ASCL, Louisiana), Bureau of Sugar Experiment Stations (BSES, now known as BSES Limited, Brisbane, Australia), South Africa Sugar Experiment Stations (SASEX, now known as the South Africa Sugar Research Institute SASRI, Mt Edgecombe, South Africa), Texas A&M Agricultural Experiment Station (TAES, Weslaco, TX), and the U. S. Department of Agriculture, Agricultural Research Service (ARS) with scientists from Florida, Hawaii, and Louisiana. The outcome of this workshop was agreement with a memorandum of understanding that these seven SROs would collaborate to expand research efforts to better understand sugarcane genomics and apply this knowledge to crop improvement. As first steps in that direction they agreed to freely share information, meet annually to discuss progress and issues, and collaboratively fund further research. The first project funded was one to contract the California Institute of Biological Research in La Jolla, California, USA, to evaluate use of molecular markers and pulsed-field gel electrophoresis for map-based cloning of sugarcane genes. The second International Workshop on Sugarcane Genome Analysis, held in 1992 in Albany, California, USA, added three more SRO members: Cenicaña (CENICAÑA, Colombia), Florida Sugar Cane League (FSCL, Florida), and Mauritius Sugar Industry Research Institute (MSIRI, Reduit, Mauritius). In addition, action at the workshop produced a new memorandum of understanding to include the new members and to name the growing organization as the International Consortium for Sugarcane Biotechnology (ICSB). The members pledged support to collaboratively fund new projects on linking markers to agronomically important traits, conduct genetic mapping with RAPD markers, and to genetically transform sugarcane for resistance to mosaic disease.

100

MING ET AL.

In November 1992, the ICSB joined as a founding member of the International Plant Genome Conference held annually in San Diego, California, USA, to conduct sugarcane biotechnology workshops. From the 1992 formalization of an organization, the ICSB has held at least annual meetings and has funded a large part of the fundamental sugarcane biotechnology research worldwide. The ICSB currently comprises 18 SROs worldwide that represent 11 countries responsible for about half of the world’s production of cane sugar. In the 12 years since the establishment of the ICSB, this organization has played a critical role in the development of much of the sugarcane specific knowledge and approaches presented in this chapter. The international collaboration to develop new knowledge for local application for improving crop production is a model worthy of consideration by other crop professional organizations. The initial goal of developing molecular genetic research capability in SROs has been met to a considerable degree. However, the field of plant molecular biology is changing rapidly and it increasingly depends on using expensive, high-throughput instrumentation not generally affordable by individual SROs. Therefore, the ICSB mission of international cooperation and coordination to share expenses of highcost, cutting-edge research in plant molecular biology may be valid well into the future. The world’s leading sugarcane producers seem well positioned to take advantage of future developments through the ICSB’s established cooperative efforts. LITERATURE CITED Anonymous. 1945. A newly released cane: some notes on NCo310. S. Afr. Sugar J. 30:91. Alexander, A. G. 1985. The energy cane alternative. Sugar series, Vol. 6. Elsevier, Amsterdam. Alix, K., F. C. Baurens, F. Paulet, J. C. Glaszmann, and A. D’Hont. 1998. Isolation and characterization of a satellite DNA family in the Saccharum complex. Genome 41:854–864. Alix, K., F. Paulet, J. C. Glaszmann, and A. D’Hont. 1999. Inter-Alu like species-specific sequences in the Saccharum complex. Theor. Appl. Genet. 6:962–968. Al-Janabi, S. M., R. J. Honeycutt, M. McClelland, and B. W. S. Sobral. 1993. A genetic linkage map of Saccharum spontaneum L. ‘SES 208’. Genetics 134:1249–1260. Al-Janabi, S. M., R. J. Honeycutt, C. Peterson, and B. W. S. Sobral. 1994. Phylogenetic analysis of organellar DNA sequences in the Andropogoneae: Saccharum. Theor. Appl. Genet. 88:933–944. Allen, F. L., R. E. Comstock, and D. C. Rasmussen. 1978. Optimal environments for yield testing. Crop Sci. 18:747–751. Allsopp, P. G., K. A. Nutt, J. Geijskes, and G. R. Smith. 2000. Transgenic sugarcane with increased resistance to canegrubs. p. 63–67. In: P. G. Allsopp and W. Suasa-ard (eds.), Sugarcane pest management strategies in the new millennium, Proc. IV Int. Soc. Sugar Cane Technol. Sugarcane Entomol. Workshop.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

101

. Altinkut, A., K. Kazan, Z. I pekci, and N. Gozukirmizi. 2001. Tolerance to paraquat is correlated with the traits associated with water stress tolerance in segregating F2 populations of barley and wheat. Euphytica 121:81–86. Appels, R., and R. Honeycutt. 1986. rDNA evolution over a billion years. p. 81–125. In: S. K. Dutta (ed.), DNA systematics. CRC Press, Boca Raton, New York, London, Tokyo. Araus, J. L., H. R. Brown, A. Febrero, J. Bort, and M. Serret. 1993. Ear photosynthesis, carbon isotope discrimination and the contribution of respiratory CO2 to differences in grain mass in durum wheat. Plant Cell Environ. 16:383–392. Arceneaux, G. 1965. Cultivated sugarcanes of the world and their botanical derivation. Proc. Int. Soc. Sugar Cane Technol. 12:844–854. Arencibia, A., E. Carmona, M. T. Cornide, S. Castiglione, J. O’Relly, A. Cinea, P. Oramas, and F. Sala. 1999. Somaclonal variation in insect-resistant transgenic sugarcane (Saccharum hybrid) plants produced by cell electroporation. Transgenic Res. 8:349–360. Arencibia, A. D., E. R. Carmona, P. Tellez, M. T. Chan, S. M. Yu, L. E. Trujillo, and P. Ormas. 1998. An efficient protocol for sugarcane (Saccharum spp. L.) transformation mediated by Agrobacterium tumefaciens. Transgenic Res. 7:213–222. Arencibia, A., P. Molina, C. Gutierrez, A. Fuentes, V. Greenidge, E. Menendez, G. De la Riva, and G. Selman. 1992. Regeneration of transgenic sugarcane (Saccharum officinarum L.) plants from intact meristematic tissues transformed by electroporation. Biotechnologia Aplicada 9:156–165. Arencibia, A., P. Molina, G. de la Riva, and G. Selman-Houssein. 1995. Production of transgenic sugarcane (Saccharum officinarum L.) plants by intact cell electroporation. Plant Cell Rep. 14:305–309. Arencibia, A. D., R. I. Vazquez, D. Prieto, P. Tellez, E. R. Carmona, A. Coego, L. Hernandez, G. A. Delariva, and G. Selmanhousein. 1997. Transgenic sugarcane plants resistant to stem borer attack. Mol. Breeding. 3:247–255. Asnaghi, C., F. Paulet, C. Kaye, L. Grivet, M. Deu, J. C. Glaszmann, and A. D’Hont. 2000. Application of synteny across Poaceae to determine the map location of a sugarcane rust resistance gene. Theor. Appl. Genet. 101:962–969. Asnaghi, C., D. Roques, S. Ruffel, C. Kaye, J.-Y. Hoarau, H. Télismart, J. C. Girard, L. M. Raboin, A. M. Risterucci, L. Grivet, and A. D’Hont. 2004. Targeted mapping of a sugarcane rust resistance gene (Bru1) using bulked segregant analysis and AFLP markers. Theor. Appl. Genet. 108:759–764. Baker, N. R. 1993. Light-use efficiency and photoinhibition of photosynthesis in plants under environmental stress. p. 221–233. In: J. A. C. Smith and H. Griffiths (eds.), Water deficits: plant responses from cell to community. Bios Scientific Publishers, Oxford, U.K. Beeharry, R. P. 1996. Extended sugarcane biomass utilisation for exportable electricity production in Mauritius. Biomass and Bioenergy 11:441–449. Bell, M. A., R. A. Fischer, D. Byerlee, and K. Sayre. 1995. Genetic and agronomic contributions to yield gains: a case study for wheat. Field Crops Res. 44:55–65. Berding, N., D. M. Hogarth, and M. C. Cox. 2004. Plant improvement of sugarcane. p. 20–53. In: J. Glyn (ed.), Sugarcane. Blackwell Science, Oxford, U.K. Berding, N., and B. T. Roach. 1987. Germplasm collection, maintenance, and use. p. 143–210. In: D. J. Heinz (ed.), Sugarcane improvement through breeding. Elsevier Press, Amsterdam. Berding, N., and J. C. Skinner. 1987. Traditional breeding methods. p. 269–320. In: Anonymous (eds.), Copersucar International Sugarcane Breeding Workshop. Copersucar, Brazil.

102

MING ET AL.

Berding, N., P. H. Moore, and G. R. Smith. 1997. Advances in breeding technology for sugarcane. p. 189–220. In: B. A. Keating, and J. R. Wilson (eds.), Intensive sugarcane production, meeting the challenges beyond 2000. CAB International. Birch, R. G. 1997. Transgenic sugarcane: opportunities and limitations. p. 125–140. In: B. A. Keating, and J. R. Wilson (eds.), Intensive sugarcane production: meeting the challenges beyond 2000. CAB Int., Wallingford, U.K. Bischoff, K. P. 2004. Excepts from the president’s message. Sugar Journal, 67:13. Blackburn, F. 1983. Sugar-cane. Longman, London and New York. p. 104–114. Blume, H. 1985. Geography of sugar cane: environmental, structural and economical aspects of cane sugar production. Verlag Dr. Albert Bartens. Berlin. p. 21–36. Boese, S. R., D. W. Wolfe, and J. J. Melkonian. 1997. Elevated CO2 mitigates chillinginduced water stress and photosynthetic reduction during chilling. Plant Cell Environ. 20:625–632. Bohmert, K., I. Balbo, J. Kopka, V. Mittendorf, C. Nawrath, Y. Poirier, G. Tishcendorf, R. N. Trethewey, and L. Willmitzer. 2000. Transgenic Arabidopsis plants can accumulate polyhydroxybutyrate to up to 4% of their fresh weight. Planta 211:841–845. Botha, F. C., B. J. B. Sawyer, and R. G. Birch. 2001. Sucrose metabolism in the culm of transgenic sugarcane with reduced soluble acid invertase activity. Proc. Int. Soc. Sugar Cane Technol. 24:588–591. Bower, R., and R. G. Birch. 1992. Transgenic sugarcane plants via microprojectile bombardment. Plant J. 2:409–416. Bowman, W. D., K. T. Hubick, S. von Caemmerer, and G. D. Farquhar. 1989. Short-term changes in leaf carbon isotope discrimination in salt- and water-stressed C4 grasses. Plant Physiol. 90:162–166. Boyer, J. S. 1982. Plant productivity and environment. Science 218:443–448. Braithwaite, K. S., R. J. Geijskes, and G. R. Smith. 2004. A variable region of the sugarcane bacilliform virus (SCBV) genome can be used to generate promoters for transgene expression in sugarcane. Plant Cell Rep. 23:319–326. Breaux, R. D. 1987. Some breeding strategies with bi-parental and polycrosses. p. 71–86. In: Anonymous (eds.), Copersucar International Sugarcane Breeding Workshop. Copersucar, Brazil. Breaux, R. D., and P. H. Dunckelman. 1969. Breeding for resistance to sugarcane mosaic with interspecific hybrids. Proc. Int. Soc. Sugar Cane Technol. 13:927–932. Bremer, G. 1923. A cytological investigation of some species and species-hybrids of the genus Saccharum. Genetica 5:273–326. Bremer, G. 1929. Remarks on the cytology of Saccharum. Facts about sugar 24:926–927. Bremer, G. 1961. Problems in breeding and cytology of sugar cane. Euphytica 10:59–78. Brown, J. S., R. J. Schnell, P. Y. P. Tai, and J. D. Miller. 2002. Phenotypic evaluation of Saccharum barberi, S. robustum, and S. sinense germplasm from the Miami, FL, USA world collection. Sugar Cane Int. 20:2002. Brumbley, S. M., L. Petrasovits, M. Purnell, M. G. O’Shea, J. Geijskes, P. Lakshmanan, G. R. Smith, and L. K. Nielsen. 2002. Application of biotechnology for future sugar industry diversification. Proc. 24th Australian Soc. Sugar Cane Technol. 24:40–46. Bull, T. A., and K. T. Glasziou. 1963. The evolutionary significance of sugar accumulation in Saccharum. Austral. J. Biol. Sci. 16:737–742. Burner, D. M. 1994. Cytogenetic and fertility characteristics of elite sugarcane clones. Sugar Cane 1:6–10. Burnquist, W. L., M. E. Sorrells, and S. Tanksley. 1992. Characterization of genetic variability in Saccharum germplasm by means of restriction fragment length polymorphism (RFLP) analysis. Proc. Int. Soc. Sugar Cane Technol. 21:355–365.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

103

Butterfield, M. K., A. D’Hont, and N. Berding. 2001. The sugarcane genome: a synthesis of current understanding, and lessons for breeding and biotechnology. Proc. Soc. Afr. Sugarcane Technol. Assn. 75:1–5. Butterfield, M. K., J. E. Irvine, M. Valdez Garza, and T. E. Mirkov. 2002. Inheritance and segregation of virus and herbicide resistance transgenes in sugarcane. Theor. Appl. Genet. 104:797–803. Carson, D. L., and F. C. Botha. 2000. Preliminary analysis of expressed sequence tags for sugarcane. Crop Sci. 40:1769–1779. Carson, D. L., and F. C. Botha. 2002. Genes expressed in sugarcane maturing internodal tissue. Plant Cell Rep. 20:1075–1081. Carson, D. L., B. I. Huckett, and F. C. Botha. 2002. Sugarcane ESTs differentially expressed in immature and maturing internodal tissue. Plant Sci. 162:289–300. Cassman, K. G. 1999. Ecological intensification of cereal production systems: yield potential, soil quality, and precision agriculture. Proc. Natl. Acad. Sci. (USA) 96:5952–5959. Casu, R. E., C. M. Dimmock, S. C. Chapman, C. P. L. Grof, L. McIntyre, G. D. Bonnett, and J. M. Manners. 2004. Identification of differentially expressed transcripts from maturing stem of sugarcane by in silico analysis of stem expressed sequence tags and gene expression profiling. Plant Mol. Biol. 54:503–517. Casu, R., C. Dimmock, M. Thomas, N. Bower, D. Knight, C. Grof, L. McIntyre, P. Jackson, D. Jordan, V. Whan, J. Drenth, Y. Tao, and J. Manners. 2001. Genetic and expression profiling in sugarcane. Proc. Int. Soc. Sugar Cane Technol. 24:542–546. Casu, R. E., C. P. L. Grof, A. L. Rae, C. L. McIntyre, C. M. Dimmock, and J. M. Manners. 2003. Identification of a novel sugar transporter homologue strongly expressed in maturing stem vascular tissues of sugarcane by expressed sequence tag and microarray analysis. Plant Mol. Biol. 52:371–386. Casu, R. E., J. M. Manners, G. D. Bonnett, P. A. Jackson, L. McIntyre, R. Dunne, S. C. Chapman, A. L. Rae, and C. P. L. Grof. 2005. Genomics approaches for the identification of genes determining important traits in sugarcane. Field Crops Res. (In press). Chapman, S. C., M. Cooper, D. W. Podlich, and G. L. Hammer. 2003. Evaluating plant breeding strategies by simulating gene action and dryland environment effects. Agron. J. 95:99–113. Chen, W. C., K. M. A. Gartland, M. R. Davey, R. Sotak, J. S. Gartland, B. J. Mulligan, J. B. Power, and E. C. Cocking. 1987. Transformation of sugarcane protoplasts by direct uptake of a selectable chimeric gene. Plant Cell Rep. 6:297–301. Chowdhury, M. K. U., and I. K. Vasil. 1992. Stably transformed herbicide resistant callus of sugarcane via microprojectile bombardment of cell suspension cultures and electroporation of protoplasts. Plant Cell Rep. 11:494–498. Christensen, A. H., and P. H. Quail. 1996. Ubiquitin promoter-based vectors for high-level expression of selectable and/or screenable marker genes in monocotyledonous plants. Transgenic Res. 5:213–318. Clayton, W. D., and S. A. Renvoize. 1986. Genera graminum—Grasses of the world. Kew Bulletin, Additional Series. 13:1-389 Clifton-Brown, J. C., I. Lewandowski, B. Andersson, G. Basch, D. G. Christian, J. B. Kjeldsen, U. Jørgensen, J. V. Mortensen, A. B. Riche, K.-U. Schwarz, K. Tayebi, and F. Teixeira. 2001. Performance of 15 Miscanthus genotypes at five sites in Europe. Agron. J. 93:1013–1019. Comstock, J. C., J. M. Shine, P. Y. P. Tai, and J. D. Miller. 2001. Breeding for ratoon stunting disease resistance: Is it both possible and effective? Proc. Int. Soc. Sugar Cane. Technol. 24:471–476. Cook, J. H. 2001. Sugarcane growth and development. Sugar Cane Int. August issue. 5–15.

104

MING ET AL.

Cooper, M., S. C. Chapman, D. W. Podlich, and G. L. Hammer. 2002. The GP problem: Quantifying gene-to-phenotype relationships. In Silico Biol. 2:151–164. Cuadrado, A., R. Acevedo, S. Moreno Dias de la Espina, N. Jouve, and C. de la Torre. 2004. Genome remodelling in three modern S. officinarum × S. spontaneum sugarcane cultivars. J. Expt. Bot. 55:847–854. D’Hont, A., and J. C. Glaszmann. 2001. Sugarcane genome analysis with molecular markers, a first decade of research. Proc. Int. Soc. Sugar Cane Technol. 24:556–559. D’Hont, A., L. Grivet, P. Feldmann, P. S. Rao, N. Berding, and J. C. Glaszmann. 1995. Identification and characterization of sugarcane intergeneric hybrids, Saccharum officinarum × Erianthus arundinaceus, with molecular markers and DNA in situ hybridization. Theor. Appl. Genet. 91:320–326. D’Hont, A., L. Grivet, P. Feldmann, P. S. Rao, N. Berding, and J. C. Glaszmann. 1996. Characterisation of the double genome structure of modern sugarcane cultivars (Saccharun spp.) by molecular cytogenetics. Mol. Gen. Genet. 250:405–413. D’Hont, A., D. Ison, K. Alix, C. Roux, and J. C. Glaszmann. 1998. Determination of basic chromosome numbers in the genus Saccharum by physical mapping of ribosomal RNA genes. Genome 41:221–225. D’Hont, A., Y. H. Lu, P. Feldmann, and J. C. Glaszmann. 1993. Cytoplasmic diversity in sugarcane revealed by heterologous probes. Sugar Cane 1:12–15. D’Hont, A., Y. H. Lu, F. Paulet, and J. C. Glaszmann. 2002. Oligoclonal interspecific origin of ‘North Indian’ and ‘Chinese’ sugarcanes. Chromosome Res. 10:253–262. da Silva, J. A., R. J. Honeycutt, W. Burnquist, S. M. Al-Janabi, M. E. Sorrells, S. D. Tanksley, W. S. Sobral. 1995. Saccharum spontaneum L. ‘SES 208’ genetic linkage map combining RFLP- and CR-based markers. Mol. Breeding 1:165–179. da Silva, J. A. 2001. Preliminary analysis of microsatellite markers derived from sugarcane ESTs. Genetics and Mol. Biol. 24:155–159. da Silva, J., M. E. Sorrells, W. Burnquist, and S. D. Tanksley. 1993. RFLP linkage map and genome analysis of Saccharum spontaneum. Genome 36:782–791. Daniels, J., and C. Daniels. 1975. Geographical, historical and cultural aspect of the origin of the Indian and Chinese sugarcanes S. barberi and S. sinense. Int. Soc. Sugar Cane Technol. Sugarcane Breed. Newslett. 36:4–23. Daniels, J., and B. T. Roach. 1987. Taxonomy and evolution in sugarcane. p. 7–84. In: D. J. Heinz (ed.), Sugarcane improvement through breeding. Elsevier Press, Amsterdam. Daniels, J., B. Roach, C. Daniels, and N. Paton. 1991. The taxonomic status of Saccharum barberi Jesweit and S. sinense Roxb. Sugar Cane 3:11–16. Daniels, J., P. Smith, N. Paton, and C. Williams. 1975. The origin of the genus Saccharum. Sugarcane Breed. Newslett. 36:24–39. Daugrois, J. H., L. Grivet, D. Roques, J. Y. Hoarau, H. Lombard, J. C. Glaszmann, and A. D’Hont. 1996. A putative major gene for rust resistance linked with a RFLP markers in sugarcane cultivar ‘R570’. Theor. Appl. Genet. 92:1059–1064. Davidson, E. H., J. P. Rast, P. Oliveri, A. Ransick, C. Calestani, C. H. Yuh, T. Minokawa, G. Amore, V. Hinman, C. Arenas-Mena, O. Ochan, C. T. Brown, C. B. Livi, P. Y. Lee, R. Revilla, A. G. Rust, Z. J. Pan, M. J. Schilstra, P. J. C. Clarke, M. I. Arnone, L. Rowen, R. A. Cameron, D. R. McClay, L. Hood, and H. Bolouri. 2002. A genomic regulatory network for development. Science 295:1669–1678. Demmig-Adams, B. 1990. Carotenoids and photoprotectants in plants: a role for the xanthophyll zeaxanthin. Biochem. Biophys Acta 1020:1–24. Demmig-Adams, B., and W. W. Adams III. 2002. Antioxidants in photosynthesis and human nutrition. Science 298:2149–2153.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

105

Demmig-Adams, B., W. W. Adams III, and S. C. Grace. 1997. Physiology of light tolerance in plants. Hort. Rev. 18:215–246. Deren, C. W., J. Alverez, and B. Glaz. 1995. Use of economic criteria for selecting clones in a sugarcane breeding program. Proc. Int. Soc. Sugar Cane Technol. 21:437–447. de Wit, C. T. 1965. Photosynthesis of leaf canopies. Agric. Res. Rep. 663. Wageningen, The Netherlands. Donald, C. M. 1968. The breeding of crop ideotypes. Euphytica 17:385–403. Du, Y.-C., and A. Nose. 2002. Effect of chilling temperature on the activity of enzymes of sucrose synthesis, and the accumulation of saccharides in leaves of three sugarcane cultivars differing in cold sensitivity. Phosynthetica 40:389–395. Du, Y.-C., A. Nose, and K. Wasano. 1999. Effects of chilling temperature on photosynthetic rates, photosynthetic enzyme activities, and metabolite levels in leaves of three sugarcane species. Plant Cell Environ. 22:317–324. Dufour, P., M. Deu, L. Grivet, A. D’Hont, F. Paulet, A. Bouet, C. Lanaud, J. C. Glaszmann, and P. Hamon. 1997. Construction of a composite sorghum genome map and comparison with sugarcane, a related complex polyploid. Theor. Appl. Genet. 94:409–418. Dunckelman, P. H., and R. D. Breaux. 1972. Breeding sugarcane varieties for Louisiana with new germplasm. Proc. Int. Soc. Sugar Cane Technol. 14:233–239. Earl, H. J., and M. Tollenaar. 1999. Using chlorophyll fluorometry to compare photosynthetic performance of commercial maize (Zea mays L.) hybrids in the field. Field Crops Res. 61:201–210. Ebrahim, M. K., G. Vogg, M. N. E. H. Osman, and E. Komor. 1998a. Photosynthetic performance and adaptation of sugarcane at suboptimal temperatures. J. Plant Physiol. 153:587–592. Ebrahim, M. K., O. Zingsheim, M. N. El-Shourbagy, P. H. Moore, and E. Komor. 1998b. Growth and sugar storage in sugarcane grown at temperatures below and above optimum. J. Plant Physiol. 153:593–602. Edgerton, C. W. 1958. Sugarcane and its diseases. LSU, Baton Rouge. p. 43–61. Elliott, A. R., J. A. Campbell, R. I. S. Brettell, and C. P. L. Grof. 1998. Agrobacteriummediated transformation of sugarcane using GFP as a screenable marker. Austral. J. Plant Physiol. 25:739–743. Enriquez-Obregon, G. A., R. I. Vazquezpadron, D. L. Prietosamsonov, G. A. Delariva, and G. Selmanhousein. 1998. Herbicide-resistant sugarcane (Saccharum officinarum L.) plants by Agrobacterium-mediated transformation. Planta 206:20–27. Eshed, Y., and D. Zamir. 1996. Less-than-additive epistatic interactions of quantitative trait loci in tomato. Genetics 143:1807–1817. Ethirajan, A. S. 1987. Sugarcane hybridization techniques. p. 129–148. In: Anonymous (eds.), Copersucar International Sugarcane Breeding Workshop. Copersucar, Brazil. Evans, H. 1935. Investigation on the root-system of sugarcane varieties. Mauritius Dep. Ag. Res. Sta. Bull. 6, 44pp. Evans, L. T. 1993. Crop evolution, adaptation and yield. Cambridge Univ. Press, Cambridge. Evans, L. T., and R. A. Fischer. 1999. Yield potential: its definition, measurement, and significance. Crop Sci. 39:1544–1551. Falco, M. C., A. T. Neto, and E. C. Ulian. 2000. Transformation and expression of a gene for herbicide resistance in a brazilian sugarcane. Plant Cell Rep. 19:1188–1194. Falco, M. C., and M. C. Silva. 2003. Expression of soybean proteinase inhibitors in transgenic sugarcane plants: effects on natural defense against Diatraea saccharalis. Plant Phys. Biochem. 41:761–766.

106

MING ET AL.

Farquhar, G. D. 1983. On the nature of carbon isotope discrimination in C4 species. Austral. J. Plant Physiol. 10:205–225. Farquhar, G. D., J. R. Ehleringer, and K. T. Hubick. 1989. Carbon isotope discrimination and photosynthesis. Ann. Rev. Plant Physiol. Plant Mol. Biol. 40:503–537. Fitter, A. H., and R. K. M. Hay. 1987. Environmental physiology of plants. Academic Press, San Diego, USA. 423pp. Flowers, T. J., and A. R. Yeo. 1995. Breeding for salinity resistance in crop plants: where next? Austral. J. Plant Physiol. 22:875–884. F. O. Licht’s International sugar report. 2003. 135(7):108–119. Foyer, C. H. 2002. The contribution of photosynthetic oxygen metabolism to oxidative stress in plants. p. 33–68. In: D. Inzé and M. Van Montagu (eds.), Oxidative stress in plants. Taylor & Francis, New York. Foyer, C. H., N. Souriau, S. Perret, M. Lelandais, K. J. Kunert, C. Pruvost, and L. Jouanin. 1995. Overexpression of glutathione reductase but not glutathione synthetase leads to increases in antioxidant capacity and improved photosynthesis in poplar (Populus tremula × P. alba) trees. Plant Physiol. 109:1047–1057. Fracheboud, Y., P. Haldimann, J. Leipner, and P. Stamp. 1999. Chlorophyll fluorescence as a selection tool for cold tolerance of photosynthesis in maize (Zea mays L.). J. Expt. Bot. 50:1533–1540. Freeling, M. 2001. Grasses as a single genetic system. Plant Physiol. 125:1191–1197. Frova, C., P. Krajewski, N. di Fonzo, M. Villa, and M. Sari-Gorla. 1999. Genetic analysis of drought tolerance in maize by molecular markers. I. Yield components. Theor. Appl. Genet. 99:280–288. Gallo-Meagher, M., and J. E. Irvine. 1993. Effects of tissue type and promoter strength on transient GUS expression in sugarcane following particle bombardment. Plant Cell Rep. 12:666–670. Gallo-Meagher, M., and J. E. Irvine. 1996. Herbicide resistant transgenic sugarcane plants containing the bar gene. Crop Sci. 36:1367–1374. Gambley, R. L., R. Ford, and G. R. Smith. 1993. Microprojectile transformation of sugarcane meristems and regeneration of shoots expressing β-glucuronidase. Plant Cell Rep. 12:343–346. Garg, A. K., J. K. Kim, T. G. Owens, A. P. Ranwala, Y. D. Choi, L. V. Kochian, and R. J. Wu. 2002. Trehalose accumulation in rice plants confers high tolerance levels to different abiotic stresses. Proc. Natl. Acad. Sci. (USA) 99:15898–15903. Gasson, M. J., Y. Kitamura, W. R. McLauchlan, A. Narbad, A. J. Parr, E. L. H. Parsons, J. Payne, M. J. C. Rhodes, and N. J. Walton. 1998. Metabolism of ferulic acid to vanillin. A bacterial gene of the enoyl-SCoA hydratase/isomerase superfamily encodes an enzyme for the hydration and cleavage of a hydroxycinnamic acid SCoA thioester. J. Biol. Chem. 273:4163–4170. Geijskes, R. J., L. Wang, P. Lakshmanan, M. G. McKenon, N. Berding, R. S. Swain, A. R. Elliot, C. P. L. Grof, J. A. Jackson, and G. R. Smith. 2003. Smartsett™ seedlings: tissue cultured seed plants for the Australian sugar industry. Proc. Australia Society Sugar Cane Technologists. 25: p6. Glaszmann, J. C., Y. H. Lu, and C. Lanaud. 1990. Variation of nuclear ribosomal DNA in sugarcane. J. Genet. Breed. 44:191–198. Glaz, B. 2003. Integrated crop management for sustainable sugarcane production: recent advances. Int. Sugar J. 105:175–186. Goddijn, O. J. M., and K. van Dun. 1999. Trehalose metabolism in plants. Trends Plant Sci. 4:315–319.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

107

Gómez-Cadenas, A., S. D. Verhey, L. D. Holappa, Q. Shen, T.-H. D. Ho, and M. K. WalkerSimmons. 1999. An abscisic acid-induced protein kinase, PKABA1, mediates abscisic acid-suppressed gene expression in barley aleurone layers. Proc. Natl. Acad. Sci. (USA) 96:1767–1772. Gowing, D. J. G., H. G. Jones, and W. J. Davies. 1993. Xylem-transported abscisic acid: the relative importance of its mass and its concentration in the control of stomatal aperature. Plant Cell Environ. 16:453–460. Greef, J. M., M. Deuter, C. Jung, and J. Schondelmaier. 1997. Genetic diversity of European Miscanthus species revealed by AFLP fingerprinting. Genet. Resources Crop Evol. 44:185–195. Griffith, H. 1993. Carbon isotope discrimination. p. 181–192. In: D. O. Hall, J. M. O. Scurlock, H. R. Bolhar-Nordenkampf, R. C. Leegood, and S. P. Long (eds.), Photosynthesis and production in a changing environment: a field and laboratory manual. Chapman & Hall, NY. Grivet, L., and P. Arruda. 2001. Sugarcane genomics: depicting the complex genome of an important tropical crop. Curr. Opinions Plant Biol. 5:122–127. Grivet, L., A. D’Hont, D. Roques, P. Feldmann, C. Lanaud, and J. C. Glaszmann. 1996. RFLP mapping in cultivated sugarcane (Saccharum spp.): Genome organization in a highly polyploid and aneuploid interspecific hybrid. Genetics 142:987–1000. Groenewald, J. H., and F. C. Botha. 2001. Down regulating pyrophospahate-dependent phosphofrukto-kinase (PFP) in sugarcane. Proc. Int. Soc. Sugar Cane Technol. 24:592–594. Groenewald, J. H., S. Groenewald, A. Whittaker, B. I. Huckett, and F. C. Botha. 1995. Molecular agriculture: prospects for production of alternative commodities in sugarcane through genetic engineering. Proc. S. Afr. Sugar Technol. 69:14–20. Guimarães, C. T., G. R. Sills, and B. W. S. Sobral. 1997. Comparative mapping of Andropogoneae: Saccharum L. (sugarcane) and its relation to sorghum and maize. Proc. Natl. Acad. Sci. (USA) 94:14261–14266. Guimarães, C. T., R. J. Honeycutt, G. R. Sills, and B. W. S. Sobral. 1999. Genetic maps of Saccharum officinarum L. and Sacharum robustum Brandes & Jew. Ex Grassl. Genetics Mol. Biol. 22:125–132. Ha, S., P. H. Moore, D. Heinz, S. Kato, N. Ohmido, and K. Fukui. 1999. Quantitative chromosome map of the polyploid Saccharum spontaneum by multicolor fluorescence in situ hybridization and imaging methods. Plant Mol. Biol. 39:1165–1173. Hallauer, A. R., and J. B. Miranda-Filho. 1981. Breeding plans. p. 407–462. In: Quantitative genetics in maize breeding. Iowa State Univ. Press, Ames. Hammer, G. L., T. R. Sinclair, S. C. Chapman, and E. van Oosterom. 2004. On systems thinking, systems biology, and the in silico plant. Plant Physiol. 134:909–911. Hansom, S., R. Bower, L. Zhang, B. Potier, A. Elliot, S. Basnayake, B. Cordeiro, D. M. Hogarth, M. Cox., N. Berding, and R. G. Elliot. 1999. Regulation of transgene expression in sugarcane. Proc. Int. Soc. Sugar Cane Technol. 23:278–289. Harrison, N. A., and M. J. Davis. 1988. Colonization of vascular tissues by Clavibacter xyli subsp. xyli in stalks of sugarcane cultivars differing in susceptibility to ratoon stunting disease. Phytopathology 78:722–727. Heaton, E., T. Voigt, and S. P. Long. 2004. A quantitative review comparing the yields of two candidate C4 perennial biomass crops in relation to nitrogen, temperature and water. Biomass Bioenergy 27:21–30. Heichel, G. H. 1974. Comparative efficiency of energy use in crop production. Bull. Conn. Ag. Expt. Sta., New Haven, 739:1–26.

108

MING ET AL.

Heinz, D. J. 1967. Wild Saccharum species for breeding in Hawaii. Proc. Int. Soc. Sugar Cane Technol. 12:1037–1043. Heinz, D. J. 1987a. Sugarcane improvement through breeding. Elsevier, New York. Heinz, D. J. 1987b. Sugarcane improvement: Current productivity and future opportunities. p. 55–70. In: Anonymous (ed.), Copersucar Int. Sugarcane Breeding Workshop. Copersucar, Brazil. Heinz, D. J., and T. Tew. 1987. Hybridization procedures. p. 313–342. In: D. J. Heinz (ed.), Sugarcane improvement through breeding. Elsevier, Amsterdam. Heinz, D. J., R. V. Osgood, and P. M. Moore. 1994. Sugarcane. Encycl. Ag. Sci. 4:225– 238. Heinze, B. S., L. N. Thokoane, N. J. Williams, J. M. Barnes, and R. S. Rutherford. 2001. The smut-sugarcane interaction as a model system for the integration of marker discovery and gene isolation. Proc. S. Afr. Sugar Cane Technol. 75:88–93. Herrmann, K. M., and L. M. Weaver. 1999. The shikimate pathway. Ann. Rev. Plant Physiol. Plant Mol. Biol. 50:473–503. Hittalmani, S., A. Parco, T. V. Mew, R. S. Zeigler, and N. Huang. 2000. Fine mapping and DNA marker-assisted pyramiding of the three major genes for blast resistance in rice. Theor. Appl. Genet 100:1121–1128. Hoarau, J.-Y., L. Grivet, B. Offmann, L.-M. Raboin, J.-P. Diorflar, J. Payet, M. Hellmann, A. D’Hont, and J.-C. Glaszmann. 2002. Genetic dissection of a modern sugarcane cultivar (Saccharum spp.). II. Detection of QTLs for yield components. Theor. Appl. Genet. 105:1027–1037. Hoarau, J. Y., B. Offmann, A. D’Hont, A. M. Risterucci, D. Roques, J. C. Glaszmann, and L. Grivet. 2001. Genetic dissection of a modern cultivar (Saccharum spp.). I. Genome mapping with AFLP. Theor. Appl. Genet. 103:84–97. Hogarth, D. M., and J. C. Skinner. 1987. Computerisation of parental selection. p. 87–102. In: Anonymous (eds.), Copersucar Int. Sugarcane Breeding Workshop. Copersucar, Brazil. Holland-Moritz, P. 2003. Sugar cane-derived therapeutic proteins avoid contamination issues. Drug Discovery Develop. Sept. 2003. Hubick, K. T., G. D. Farquhar, and R. Shorter. 1986. Correlation between water-use efficiency and carbon isotope discrimination in diverse peanut (Arachis) germplasm. Austral. J. Plant Physiol. 13:803–816. Hubick, K. T., G. L. Hammer, G. D. Farquhar, L. J. Wade, S. von Caemmerer, and S. A. Henderson. 1990. Carbon isotope discrimination varies genetically in C4 species. Plant Physiol. 91:534–537. Hunt, L. A., M. P. Reynolds, K. D. Sayre, S. Rajaram, J. W. White, and W. Yan. 2003. Crop modeling and the identification of sable coefficients that may reflect significant groups of genes. Agron. J. 95:20–31. Ideker, T., V. Thorsson, J. A. Ranish, R. Christama, J. Buhler, J. K. Eng, R. Bumgarner, D. R. Goodlett, R. Aebrsold, and L. Hood. 2001. Integrated genomic and proteomic analyses of a systematically perturbed metabolic network. Science 292:929–934. Ingelbrecht, I. L., J. E. Irvine, and T. E. Mirkov. 1999. Posttranscriptional gene silencing in transgenic sugarcane. Dissection of homology-dependent virus resistance in a monocot that has a complex polyploid genome. Plant Physiol. 119:1187–1197. Irvine, J. E. 1975. Relations of photosynthetic rates and leaf and canopy characters to sugarcane yield. Crop Sci. 15:671–676. Irvine, J. E. 1977. Identification of cold tolerance in Saccharum and related genera through refrigerated freezing screening. Proc. Int. Soc. Sugar Cane Technol. 16:147–156. Irvine, J. E. 1999. Saccharum species as horticultural classes. Theor. Appl. Genet. 98:186–194.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

109

Irvine, J. E., and C. G. de Almedia. 1991. Delivery of plasmid DNA through particle bombardment with an airless sprayer. Texas A&M Misc. Publ. 1726. Irvine, J. E., and T. E. Mirkov. 1997. The development of genetic transformation of sugarcane in Texas. Sugar J. 60:25–29. Jackson, P., M. Robertson, M. Cooper, and G. Hammer. 1996. The role of physiological understanding in plant breeding; from a breeding perspective. Field Crops Res. 49:11–37. James, N. I. 1980. Sugarcane. p. 617–629. In: W. R. Fehr and H. H. Hadley (eds.), Hybridization of crop plants. ASA and CSSA, Madison, Wisconsin, USA. Jannoo, N., L. Grivet, J. David, A. D’Hont, and J. C. Glaszmann. 2004. Differential chromosome pairing affinities at meiosis in polyploid sugarcane revealed by molecular markers. Heredity 93:460–467. Jannoo, N., L. Grivet, A. Dookun, A. D’Hont, and J. C. Glaszmann. 1999a. Linkage disequilibrium among modern sugarcane cultivars. Theor. Appl. Genet. 99:1053–1060. Jannoo, N., L. Grivet, M. Seguin, F. Paulet, R. Domaingue, P. S. Rao, A. Dookun, A. D’Hont, and J. C. Glaszmann. 1999b. Molecular investigation of the genetic base of sugarcane cultivars. Theor. Appl. Genet. 99:171–184. Jenkin, M., S. Reader, K. Purdie, and T. Miller. 1995. Detection of rDNA sites in sugarcane by FISH. Chromosome Res. 3:444–445. Jeswiet, J. 1929. The development of selection and breeding of the sugarcane in Java. Proc. Int. Soc. Sugar Cane Technol. 3:44–57. Jifon, J. L., and J. P. Syvertsen. 2003. Moderate shade can increase net gas exchange and reduce photoinhibition in citrus leaves. Tree Physiol. 23:119–127. Jones, H. G. 1992. Plants and microclimate. A quantitative approach to environmental plant physiology. 2nd ed., Cambridge Univ. Press, Cambridge. Joyce, P. A., R. B. McQualter, M. J. Bernard, and G. R. Smith. 1998. Engineering for resistance to SCMV in sugarcane. Acta Hort. 461:385–391. Kennedy, A. J., and P. S. Rao. 2000. Handbook 2000. West Indies Central Sugar Cane Breeding Stn, Groves, St. George, Barbados. p. 1–10. Kheshgi, H. S., R. C. Prince, and G. Marland. 2000. The potential of biomass fuels in the context of global climate change: focus on transportation fuels. Annu. Rev. Energy Environ. 25:199–244. Kitano, H. 2002. Systems biology: a brief overview. Science 295:1662–1664. Klein, J., J. Altenbuchner, and R. Mattes. 1998. Nucleic acid and protein elimination during the sugar manufacturing process of conventional and transgenic sugar beets. J. Biotech. 60:145–153. Koch, K. 2004. Sucrose metabolism: regulatory mechanisms and pivotal roles in sensing and plant development. Curr. Opinions Plant Biol. 7:235–246. Kratsch, H. A., and R. R. Wise. 2000. The ultrastructure of chilling stress. Plant, Cell Environ. 23:337–350. Kudla, J. Q., K. Xu, K. Harter, W. Gruissem, and S. Luan. 1999. Genes for calcineurin Blike protein in Arabidopsis are differentially regulated by stress signals. Proc. Natl. Acad. Sci. (USA) 96:4718–4723. Kuhnle, J. A., P. H. Moore, W. L. Yauger, and W. F. Haddon. 1979. Drought induced abscisic acid changes in three sugarcane cultivars. Proc. Plant Growth Reg. Work. Group, 7:221–222. Lambers, H., F. S. Chapin III, and T. L. Pons. 1998. Plant physiological ecology. Springer, New York. p. 540. Larcher, W. 1981. Physiological plant ecology. Springer-Verlag, New York. p. 231. Last, K. I., R. I. S. Brettell, D. A. Chamberlain, A. M. Chaudhury, P. J. Larkin, E. L. Marsch, W. J. Peacock, and E. S. Dennis. 1991. pEmu: an improved promoter for gene expression in cereal cells. Theor. Appl. Genet. 81:581–588.

110

MING ET AL.

Lee, M. 1995. DNA markers and plant breeding programs. p. 265–344. In: D. L. Sparks (ed.), Advances in agronomy. Academic Press, San Diego. Legaspi, J. C., and T. E. Mirkov. 2000. Evaluation of transgenic sugarcane against stalkborers. In: P. G. Allsopp and W. Suasa-ard (eds.), Sugarcane pest management strategies in the new millennium. IV Int. Soc. Sugar Cane Technol. Entomology Workshop. p. 68–71. Legendre, B. L. and D. M. Burner. 1996. Biomass production of sugarcane cultivars and early-generation hybrids. Biomass Bioenergy 8:55–61. Leibbrandt, N. B., and S. J. Snyman. 2003. Stability of gene expression and agronomic performance of a transgenic herbicide-resistant sugarcane line in South Africa. Crop Sci. 43:671–677. Levitt, J. 1980. Responses of plants to environmental stresses. Vol I. Chilling, freezing and high temperature stresses. Academic Press, New York, 497 pp. Levitt, J. 1982. Stress terminology. p. 437–439. In: N. C. Turner and P. J. Kramer (eds.), Adaptations of plants to water and high temperature stress. Wiley-Interscience, New York. Lewandowski, I., J. C. Clifton-Brown, J. M. O. Scurlock, and W. Huisman. 2000. Miscanthus: European experience with a novel energy crop. Biomass Bioenergy 19:209–277. Lima, M. L. A., A. A. F. Garcia, K. M. Oliveira, S. Matsuoka, H. Arizono, C. L. de Souza Jr., and A. P. de Souza. 2002. Analysis of genetic similarity detected by AFLP and coefficient of parentage among genotypes of sugar cane (Saccharum spp.). Theor. Appl. Genet. 104:30–38. Liu, D. W., S. V. Oard, and J. H. Oard. 2003. High transgene expression levels in sugarcane (Saccharum officinarum L.) driven by the rice ubiquitin promoter RUBQ2. Plant Sci. 165:743–750. Lo, C. C. 1987. Sugarcane breeding for different environments. p. 189–202. In: Anonymous (eds.), Copersucar International Sugarcane Breeding Workshop. Copersucar, Brazil. Lo, C. C., Y. H. Chia, and W. H. Chen. 1978. Collecting Miscanthus germplasm in Taiwan. Proc. Int. Soc. Sugar Cane Technol. 16:59–69. Long, S. P., S. Humphries, and P. G. Falkowski. 1994. Photoinhibition of photosynthesis in nature. Annu. Rev. Plant Physiol. Plant Mol. Biol. 45:633–662. Loomis, R. S., and J. S. Amthor. 1999. Yield potential, plant assimilatory capacity and metabolic efficiencies. Crop Sci. 39:1584–1596. Loomis, R. S., and W. A. Williams. 1963. Maximum crop productivity: An estimate. Crop Sci. 3:67–72. Lössl, A., C. Eibl, H.-J. Harloff, C. Jung, and H.-U. Koop. 2003. Polyester synthesis in transplastomic tobacco (Nicotiana tabacum L.): significant contents of polyhydroxybutyrate are associated with growth reduction. Plant Cell Rep. 21:891–899. Lu, Y. H., A. D’Hont, F. Paulet, L. Grivet, M. Arnaud, and J. C. Glaszmann. 1994b. Molecular diversity and genome structure in modern sugarcane varieties. Euphytica 78:217–226. Lu, Y. H., A. D’Hont, D. I. T. Walker, P. S. Rao, P. Feldmann, and J. C. Glaszmann. 1994a. Relationships among ancestral species of sugarcane revealed with RFLP using single copy maize nuclear probes. Euphytica 78:7–8. Lu, Z., R. G. Percy, C. O. Qualset, and E. Zeiger. 1998. Stomatal conductance predicts yields in irrigated pima cotton and bread wheat grown at high temperatures. J. Exp. Bot. 49:453–460. Lunn, J. E., and E. MacRae. 2003. New complexities in the synthesis of sucrose. Curr. Opinions Plant Biol. 6:208–214. Ma, H., H. H. Albert, R. Paull, and P. H. Moore. 2000. Metabolic engineering of invertase activities in different subcellular compartments affects sucrose accumulation in sugarcane cells. Austral. J. Plant Physiol. 27:1021–1030.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

111

Ma, H., S. Schulze, S. Lee, E. Mirkov, J. Irvine, P. Moore, and A. Paterson. 2004. An EST survey of the sugarcane transcriptome. Theor. Appl. Genet. 108:851–863. Macedo, I. C. 2000. Energy production from biomass: the sugar cane agro-industry in Brazil. p. 119–128. In: C. E. Rocha-Miranda (ed.), Transition to global sustainability: the contributions of Brazilian science. Academia Brasiliera de Ciências, Rio de Janeiro. Madison, L. L., and G. W. Huisman. 1999. Metabolic engineering of poly(3hydroxyalkanoates): from DNA to plastic. Microbiol. Mol. Biol. Rev. 63:21–53. Mamet, L. D., and R. Domaingue. 1999. Shortening the selection process for sugarcane. Expt. Ag. 35:391–405. Manickavasagam, M., A. Ganapathi, V. R. Anbazhagan, B. Sudhakar, N. Selvaraj, A. Vasudevan, and S. Kasthurirengan. 2004. Agrobacterium-mediated genetic transformation and development of herbicide-resistant sugarcane (Saccharum species hybrids) using axillary buds. Plant Cell Rep. 23:134–143. Marathe, R., R. Anandalakshmi, T. Smith, G. Pruss, and V. B. Vance. 2000. RNA viruses as inducers, suppressors and targets of post-transcriptional gene silencing. Plant Mol. Biol. 43:295–306. Martin, B., C. G. Tauer, and R. K. Lin. 1999. Carbon isotope discrimination as a tool to improve water-use-efficiency in tomato. Crop Sci. 39:1775–1783. Mattson, W. J., and R. A. Haack. 1987. The role of drought in outbreaks of plant-eating insects. BioScience 7:110–118. Mayer, M. J., A. Narbad, A. J. Parr, M. L. Parker, N. J. Walton, F. A. Mellon, and A. J. Michael. 2001. Rerouting the plant phenylpropanoid pathway by expression of a novel bacterial enoyl-CoA hydratase/lyase enzyme function. Plant Cell 13:1669–1682. McElroy, D., A. D. Blowers, B. Jenes, and R. Wu. 1991. Construction of expression vectors based on the rice actin 1 (Act1) 5′ region for use in monocot transformation. Mol. Gen. Genet. 231:150–160. McQualter, R. B., J. L. Dale, R. M. Harding, J. A. McMahon, and G. R. Smith. 2004. Production and evaluation of transgenic sugarcane containing a Fiji disease virus (FDV) genome segment S9-derived synthetic resistance gene. Austral. J. Agr. Res. 55: 139–145. McQualter, R. B., B. Fong Chong, M. O’Shea, K. Meyer, D. E. Van Dyk, P. V. Viitanen, and S. M. Brumbley. 2005. Initial evaluation of sugarcane as a production platform for phydroxybenzoic acid. Plant Biotechnol. J. (In press). McQualter, R. B., R. M. Harding, J. L. Dale, and G. R. Smith. 2001. Virus derived resistance to Fiji disease in transgenic plants. Proc. Int. Soc. Sugar Cane Technol. 24:584–585. Meinzer, F. C., and D. A. Grantz. 1990. Stomatal and hydraulic conductance in growing sugarcane: stomatal adjustment to water transport capacity. Plant Cell Environ. 13:383–388. Meinzer, F. C., and D. A. Grantz. 1991. Coordination of stomatal, hydraulic, and canopy boundary layer properties: do stomata balance conductances by measuring transpiration? Physiologia Plant. 83:324–329. Meinzer, F. C., Z. Plaut, and N. Z. Saliendra. 1994. Carbon isotope discrimination, gas exchange and growth of sugarcane cultivars under salinity. Plant Physiol. 104: 521–526. Ming, R., S. C. Liu, P. H. Moore, J. E. Irvine, and A. H. Paterson. 2001. QTL analysis in a complex autopolyploid: genetic control of sugar content in sugarcane. Genome Res. 11:2075–2084. Ming, R., S. C. Liu, Y. R. Lin, D. Braga, J. da Silva, A. van Deynze, T. F. Wenslaff, K. K. Wu, P. H. Moore, W. Burnquist, M. E. Sorrells, J. E. Irvine, and A. H. Paterson. 1998. Alignment of Sorghum and Saccharum chromosomes: comparative organization of closely-related diploid and polyploid genomes. Genetics 150:1663–1682.

112

MING ET AL.

Ming, R., T. DelMonte, P. H. Moore, J. E. Irvine, and A. H. Paterson. 2002a. Comparative analysis of QTLs affecting plant height and flowering time among closely-related diploid and polyploid genomes. Genome 45:794–803. Ming, R., Y.-W. Wang, X. Dryer, P. H. Moore, J. E. Irvine, and A. H. Paterson. 2002b. Molecular dissection of complex traits in autopolyploid: mapping QTLs influencing sugar yield and related traits in sugarcane. Theor. Appl. Genet. 105:332–345. Minorsky, P. V. 2003. Achieving the in silico plant. Systems biology and the future of plant biological research. Plant Physiol. 132:404–409. Moissiard, G., and O. Voinnet. 2004. Viral suppression of RNA silencing in plants. Mol. Plant Path. 5:71–82. Moore, P. H. 1987. Breeding for stress resistance. p. 503–542. In: D. J. Heinz (ed.), Sugarcane improvement through breeding. Elsevier, Amsterdam. Moore, P. H. 1995. Temporal and spatial regulation of sucrose accumulation in the sugarcane stem. Austral. J. Plant Physiol. 22:661–679. Moore, P. H. 1996. Progress in sugarcane molecular biology. Proc. Int. Soc. Sugar Cane Technol. 22:353–362. Moore, P. H., and A. Maretzki. 1996. Sugarcane. p. 643–669. In: E. Zamski and A. A. Schaffer (eds.), Photoassimilate distribution in plants and crops: source-sink relationships. Marcel Dekker, New York, Basel, Hong Kong. Moore, P. H., F. C. Botha, R. Furbank, and C. Grof. 1997. Potential for overcoming physiobiochemical limits to sucrose accumulation. p. 141–155. In: B. A. Keating and J. R. Wilson (eds.), Intensive sugarcane production: meeting the challenges beyond 2000. CAB Int., Wallingford, UK. Muchow, R. C., M. F. Spillman, W. W. Wood, and M. R. Thomas. 1994. Radiation interception and biomass accumulation in a sugarcane crop under irrigated tropical conditions. Austral. J. Agr. Res. 45:3–49. Mudge, J., W. R. Anderson, R. L. Kehrer, and D. J. Fairbanks. 1996. A RAPD genetic map of Saccharum officinarum. Crop Sci. 36:1362–1366. Mukherjee, S. K. 1957. Origin and distribution of Saccharum. Bot. Gaz. 119:55–56. Naidu, K. M., and T. V. Sreenivasan. 1987. Conservation of Sugarcane Germplasm. p. 3–52. In: Copersucar Int. Sugarcane Breeding Workshop. Copersucar, Brazil. Nair, N. V., S. Nair, T. V. Sreenivasan, and M. Mohan. 1999. Analysis of genetic diversity and phylogeny in Saccharum and related genera using RAPD markers. Genet. Res. Crop Evol. 46:73–79. Nawrath, C., Y. Poirier, and C. Somerville. 1994. Targeting of the polyhydroxybutyrate biosynthetic pathway to the plastids of Arabidopsis thaliana results in high levels of polymer accumulation. Proc. Natl. Acad. Sci. (USA) 91:12760–12764. Nilsen, E. T., and D. M. Orcutt. 1996. Physiology of plants under stress: abiotic factors. Wiley, New York. Nogueira, F. T. S., V. E. de Rosa Jr., M. Menossi, E. C. Ulian, and P. Arruda. 2003. RNA expression profiles and data mining of sugarcane response to low temperature. Plant Physiol. 132:1811–1824. Nordborg, M., J. O. Borevitz, J. Bergelson, C. C. Berry, J. Chory, J. Hagenblad, M. Kreitman, J. N. Maloof, T. Noyes, P. J. Oefner, E. A. Stahl, and D. Weigel. 2002. The extent of linkage disequilibrium in Arabidopsis thaliana. Nat. Genet. 30:190–193. Nuss, K. J., and P. G. C. Brett. 1995. The release of cultivar NCo310 in 1945 and its impact on the sugar industry. Proc. S. Afr. Sug. Technol. Assoc. 69:3–8. Nutt, K. A., P. G. Allsopp, R. J. Geijskes, M. G. McKeon, and G. R. Smith. 2001. Canegrub resistant sugarcane. Proc. Int. Soc. Sugar Cane Technol. 24:582–583.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

113

Nutt, K. A., P. G. Allsopp, T. K. McGhie, K. M. Shepherd, P. A. Joyce, G. O. Taylor, R. B. McQualter, and G. R. Smith. 1999. Transgenic sugarcane with increased resistance to canegrubs. Proc. Austral. Soc. Sugar Cane Technol. 21:171–176. Ogren, E., and M. Sjostrom. 1990. Estimation of the effect of photoinhibition on the carbon gain in leaves of a willow canopy. Planta 181:560–567. O’Leary, M. H. 1988. Carbon isotopes in photosynthesis. Bioscience 38:328–336. Ort, D. R. 2001. When there is too much light. Plant Physiol. 125:29–32. Osgood, R. 2003. Sugarcane yield and record yield. Sugar J. (Dec):6–7. Panje, R., and C. Babu. 1960. Studies in Saccharum spontaneum distribution and geographical association of chromosome numbers. Cytologia, 25:152–172. Panje, R. R. 1972. The role of Saccharum spontaneum in sugarcane breeding. Proc. Int. Soc. Sugar Cane Technol. 14:217–223. Paquet, L., B. Rathinasabapathi, H. Saini, L. Zamir, D. A. Gage, Z.-H. Huang, and A. D. Hanson. 1994. Accumulation of the compatible solute 3-dimethylsulfoniopropionate in sugarcane and its relatives, but not in other gramineous crops. Austral. J. Plant Physiol. 21: 37–48. Paton, N., J. Daniels, and P. Smith. 1978. A study of S. sinense Roxb. and S. barberi Jesw. Sugarcane Breed. Newslett. 41:33–50. Peoples, O. P., and A. J. Sinskey. 1989a. Poly-beta-hydroxybutyrate biosynthesis in Alcaligenes eutrophus H16. Characterization of the genes encoding beta-ketothiolase and acetoacetyl-CoA reductase. J. Biol. Chem. 264:15293–15297. Peoples, O. P., and A. J. Sinskey. 1989b. Poly-beta-hydroxybutyrate (PHB) biosynthesis in Alcaligenes eutrophus H16. Identification and characterization of the PHB polymerase gene (phbC). J. Biol. Chem. 264:15298–15303. Perl-Treves, R., and A. Perl. 2002. Oxidative stress: an introduction. p. 1–32. In: D. Inzé, M. Van Montagu (eds.), Oxidative stress in plants. Taylor & Francis, New York. Pilon-Smits, E. A. H., N. Terry, T. Sears, H. Kim, A. Zayed, S. Hwang, K. van Dun, E. Voogd, T. C. Verwoerd, R. W. H. H. Krutwagen, and O. J. M. Goddijn. 1998. Trehaloseproducing transgenic tobacco plants show improved growth performance under drought stress. J. Plant Physiol 152:525–532 Piperidis, G., and A. D’Hont. 2001. Chromosome composition analysis of various Saccharum interspecific hybrids by genomic in situ hybridisation (GISH). Proc. Int. Soc. Sugar Cane Technol. 11:565–566. Piperidis, G., M. J. Christopher, B. J. Carroll, N. Berding, and A. D’Hont. 2000. Molecular contribution to selection of intergeneric hybrids between sugarcane and the wild species Erianthus arundinaceus. Genome 43:1033–1037. Poirier, Y. 1999. Production of new polymeric compounds in plants. Curr. Opinions Biotechnol. 10:181–185. Poirier, Y. 2002. Polyhydroxyalknoate synthesis in plants as a tool for biotechnology and basic studies of lipid metabolism. Progress Lipid Res. 41:131–155. Poirier, Y., D. Dennis, K. Klomparens, C. Nawrath, and C. Somerville. 1992. Perspectives on the production of polyhydroxyalkanoates in plants. FEMS Microbiol. Rev. 103:237–246. Price, S. 1957. Cytological studies in Saccharum and allied genera II. Chromosome numbers in interspecific hybrids. Bot. Gaz. 118:146–159. Price, S. 1963. Cytogenetics of modern sugar canes. Economic Bot. 17:97–105. Price, S. 1965. Interspecific hybridization in sugarcane breeding. Proc. Int. Soc. Sugar Cane Technol. 12:1021–1026. Ramdoyal, K., G. H. Badaloo, and P. Goburdhun. 2003. An international survey of sugar cane crossing programmes. Sugar Cane Int. (July):3–11.

114

MING ET AL.

Rao, J. T. 1951. Xeromorphic adaptations in sugarcane for resistance to drought. Proc. Int. Soc. Sugar Cane Technol. 7:82–89. Rao, K. C., and S. Asokan. 1978. Studies on free proline association to drought resistance in sugarcane. Sugar J. 40(1):23–24. Rathus, C., and R. G. Birch. 1992. Stable transformation of callus from electroporated sugarcane protoplasts. Plant Sci. 82:81–89. Rathus, C., R. Bower, and R. G. Birch. 1993. Effects of promoter, intron and enhancer elements on transient gene expression in sugarcane and carrot protoplasts. Plant Mol. Biol. 23:613–618. Rehm, B. H. A., and A. Steinbuchel. 1999. Biochemical and genetic analysis of PHA synthases and other proteins required for PHA synthesis. Inter. J. Biol. Macromol. 25:3–19. Reymond, P., H. Weber, M. Damon, and E. E. Farmer. 2000. Differential gene expression in response to mechanical wounding and insect feeding in Arabidopsis. Plant Cell 12:707–720. Richter, T. E., and P. C. Ronald. 2000. The evolution of disease resistance genes. Plant Mol. Biol. 42:195–204. Roach, B. T. 1978. Utilization of Saccharum spontaneum in sugarcane breeding. Proc. Int. Soc. Sugar Cane Technol. 16:43–58. Roach, B. T. 1989. Origin and improvement of the genetic base of sugarcane. Proc. Austral. Soc. Sugar Cane Technol. 11:34–47. Roach, B. T., and J. Daniels. 1987. The Saccharum complex and the genus Saccharum. p. 1–33. In: Copersucar Int. Sugarcane Breeding Workshop. Copersucar, Brazil. Robertson, M. J., R. C. Muchow, A. W. Wood, and J. A. Campbell. 1996. Accumulation of reducing sugars by sugarcane: effects of crop age, nitrogen supply and cultivar. Field Crops Res. 49:39–50. Rossi, G. 2001. Sugarcane variety notes “an international directory”, 7th rev. Piracicaba, SP, Brazil. Rossi, G. 2002. Sugarcane varieties. Sugar J. Oct.:6–7. Rossi, G., W. M. da Silva, and J. E. Irvine. 1987. Sugarcane breeding in Brazil: the Copersucar program. p. 217–232. In: Copersucar International Sugarcane Breeding Workshop. Copersucar, Brazil. Rossi, M., P. G. Araujo, F. Paulet, O. Garsmeur, V. M. Dias, H. Chen, M.-A. Van Sluys, and A. d’Hont. 2003. Genomic distribution and characterization of EST-derived resistance gene analogs (RGAs) in sugarcane. Mol. Gen. Genomics 269:406–419. Rozeff, N., J. Amador, and J. Irvine. 1998. South Texas sugarcane production handbook. Texas A&M Univ. System, Rio Grande Valley Sugar Growers, Inc. Weslaco, TX. Saijo, Y., S. Hata, J. Kyozuka, K. Shimamoto, and K. Izui. 2000. Overexpression of a single Ca2+ dependent protein kinase confers both cold and salt / drought tolerance on rice plants. Plant J. 23:319–327. Schenck, S., B. E. Lockhart, and J. S. Hu. 1997. Use of a tissue blot immunoassay to determine the distribution of sugarcane yellow leaf virus in Hawaii. Sugar Cane 4:5–8. Schenk, P. M., L. Sági, T. Remans, R. G. Dietzgen, M. J. Bernard, M. W. Graham, and J. M. Manners. 1999. A promoter from sugarcane bacilliform badnavirus drives transgene expression in banana and other monocot and dicot plants. Plant Mol. Biol. 39:1221–1230. Schenk, P. M., T. Remans, L. Sági, A. R. Elliot, R. G. Dietzgen, R. Swennen, P. R. Ebert, C. P. Grof, and J. M. Manners. 2001. Promoters for pregenomic RNA of banana streak bandnavirus are active for transgene expression in monocot and dicot plants. Plant Mol. Biol. 47:399–412.

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

115

Schubert, P., A. Steinbuchel, and H. G. Schlegel. 1988. Cloning of the Alcaligenes eutrophus genes for synthesis of poly-beta-hydroxybutyric acid (PHB) and synthesis of PHB in Escherichia coli. J. Bacteriol. 170:5837–47. Sétamou, M., J. S. Bernal, T. E. Mirkov, and J. C. Legaspi. 2003. Effects of snowdrop lectin (GNA) on Mexican rice borer (Lepidoptera: Pyralidae) life history parameters. J. Econ. Entomol. 96:950–956. Shang, K. C., P. Y. Juang, T. L. Chu, and S. T. Huang. 1969. A study on the transmission of some important characteristics of Taiwan-originated wild cane Saccharum spontaneum L. Proc. Int. Soc. Sugar Cane Technol. 13:968–974. Sheen, J. 1996. Ca2+-dependent protein kinase in stress signal transduction. Science 274:1900–1902. Sheu, Y. S. 1996. Rating of drought resistance in sugarcane. Proc. Int. Soc. Sugar Cane Technol. 22(2):75–79. Siebert, M., S. Sommer, S. Li, Z. Wang, K. Severin, and L. Heide. 1996. Genetic engineering of plant secondary metabolism. Accumulation of 4-hydroxybenzoate glucosides as a result of the expression of the bacterial ubiC gene in tobacco. Plant Physiol. 112:811–819. Simmonds, N. W. 1976. Sugarcanes. p. 104–108. In: N. W. Simmonds (ed.), Evolution of crop plants. Longmans, London. Skinner, J. C., D. M. Hogarth, and K. K. Wu. 1987. Selection methods, criteria, and indices. p. 409–435. In: D. J. Heinz (ed.), Sugarcane improvement through breeding. Elsevier, Amsterdam. Slater, S., T. A. Mitsky, K. L. Houmiel, M. Hao, S. E. Reiser, N. B. Taylor, M. Tran, H. E. Valentin, D. J. Rodriquez, D. A. Stone, S. R. Padgette, G. Kishore, and K. J. Gruys. 1999. Metabolic engineering of Arabidopsis and Brassica for poly(3-hydroxybutyrate-co3-hydroxyvalerate) copolymer production. Nature Biotechnol. 17:1011–1016. Smedley, M. P., T. E. Dawson, J. P. Comstock, L. A. Donovan, D. E. Sherrill, C. S. Cook, and J. R. Ehleringer. 1991. Seasonal carbon isotope discrimination in a grassland community. Oecologia 85:314–320. Snell, K. D., and O. P. Peoples. 2002. Polyhydroxyalkanoate polymers and their production in transgenic plants. Metabolic Engineering 4:29–40. Snyman, S. J., N. B. Leibbrandt, and F. C. Botha. 1998. Buster resistant sugarcane. Proc. S. African Sugar Technol. Assoc. 72:138–139. Sobral, B. W. S., D. P. V. Braga, E. S. LaHood, and P. Keim. 1994. Phylogenetic analysis of chloroplast restriction enzyme site mutations in the Saccharinae Griseb. Subtribe of the Andropogoneae Dumort. Tribe. Theor. Appl. Genet. 87:843–853. Sorares-Netto, L. E. 2001. Oxidative stress response in sugarcane. Genetics Mol. Biol. 24:93–102. Souza, G. M, A. C. Q. Simoes, K. C. Oliveira, H. M. Garay, L. C. Fiorini, F. S. Gomes, M. Y. Nishiyam-Junior, and A. M. da Silva. 2001. The sugarcane signal transduction (SUCAST) catalogue: prospecting signal transduction in sugarcane. Genetics Mol. Biol. 24:25–34. Sreenivasan, T. V. 1975. Cytogenetical studies in Saccharum spontaneum L. Proc. Indian Acad. Sci. 81:131–144. Sreenivasan, T. V., B. S. Ahloowalia, and D. J. Heinz. 1987. Cytogenetics. p. 211–253. In: D. J. Heinz (ed.), Sugarcane improvement through breeding. Elsevier, Amsterdam. Steinbüchel, A., and H. E. Valentin. 1995. Diversity of bacterial polyhydroxyalkanoic acids. FEMS Microbiol. Lett. 128:219–228. Stevenson, G. C. 1965. Genetics and breeding of sugar cane. Longman, London.

116

MING ET AL.

Stricker, J. A., G. M. Prine, D. L. Anderson, D. B. Shibles, and T. C. Riddle. 1993. Biomass yield of tall grass energy crops on phosphatic clay in central Florida. Circular 1084, 4–6. Coop. Ext. Service, Inst. Food Ag. Sci., Univ. Florida, Gainesville. SUCEST. 2001. The Sugarcane EST Project (SUCEST) http://sucest.lbi.ic.unicamp.br/en/. Tai, P. Y. P, and J. D. Miller. 2001. A core collection for Saccharum spontaneum L. from the World Collection of Sugarcane. Crop Science, Vol. 41:879–885. Tai, P. Y. P., and J. D. Miller. 2002. Germplasm diversity among four sugarcane species for sugar composition. Crop Sci. 42:958–964. Tardieu, F., and W. J. Davies. 1993. Integration of hydraulic and chemical signaling in the control of stomatal conductance and water status of droughted plants: opinion. Plant Cell Environ. 16:341–350. Taylor, G. O., P. A. Joyce, J. M. Sedl, and G. R. Smith. 1999. Laboratory crystallised sugar from genetically engineered sugarcane does not contain transgene DNA. Proc. Austral. Soc. Sugar Cane Technol. 21:502. Tew, T. L. 1980. Genetic engineering for greater energy efficiency in sugarcane. Conf. Renewable Energy Technol. Dec 5, 1980. Honolulu. Tew, T. L. 1987. New varieties. p. 559–594. In: D. J. Heinz (ed.), Sugarcane improvement through breeding. Elsevier, Amsterdam. Tew, T. L. 2003. World sugarcane variety census–Year 2000. Sugar Cane Int. March/April 2003, p. 12–18. Tew, T. L. 2004. Ho 95-988, a product of the USDA-ARS basic breeding effort in Louisiana. Sugar Journal, 67:21–22. Thokoane, L. N., and R. S. Rutherford. 2001. cDNA-AFLP differential display of sugarcane (Saccharum spp. Hybrids) genes induced by challenge with the fungal pathogen Ustilago scitaminea (sugarcane smut). Proc. S. Afr. Sug. Technol. 75:104–107. Tomkins, J. P., Y. Yu, H. Miller-Smith, D. A. Frisch, S. S. Woo, and R. Wing. 1999. A bacterial artificial chromosome library for sugarcane. Theor. Appl. Genet. 99:419–424. van Kooten, O., and J. F. H. Snell. 1990. The use of chlorophyll fluorescence nomenclature in plant stress physiology. Photosyn. Res. 25:147–150. Verret, J. A. 1925. A method of handling cane tassels for breeding work. Hawaiian Planters’ Records 29:84–94. Vettore, A. L., F. R. da Silva, E. L. Kemper, and P. Arruda. 2001. The libraries that made SUCEST. Gen. Mol. Biol. 24:1–7. Vettore, A. L., F. R. da Silva, E. L. Kemper, G. M. Souza, A. M. da Silva, M. I. T. Ferro, F. Henrique-Silva, E. A. Giglioti, M. V. F. Lemos, L. L. Coutinho, M. P. Nobrega, H. Carrer, S. C. Franca, M. Bacci Jr., M. H. S. Goldman, S. L. Gomes, L. R. Nunes, L. E. A. Camarfgo, W. J. Siqueira, M.-A. Van Sluys, O. H. Thiemann, E. E. Kuramae, R. V. Santelli, C. L. Marino, M. L. P. Targon, J. A. Ferro, H. C. S. Silveira, D. C. Marini, E. G. M. Lemos, C. B. Monteiro-Vitorello, J. H. M. Tambor, D. M. Carraro, P. G. Roberto, V. G. Martins, G. H. Goldman, G. C. de Oliveira, D. Truffi, C. A. Colombo, M. Rossi, P. G. de Araujo, S. A. Sculaccio, A. Angella, M. M. A. Lima, L. Grivet, S. M. Z. Di Mauro, F. G. Nobrega, S. A. C. F. M. Menck, M. D. V. Braga, G. P. Telles, F. A. A. Cara, G. Pendrosa, J. Meidanis, and P. Arruda. 2003. Analysis and functional annotation of an expressed sequence tag collection for tropical crop sugarcane. Genome Res. 13:2725–2735. Vickers, J. E., A. E. R. Kartikasari, D. P. Knight, P. A. Jackson, T. E. Morgan, S. J. Attard, G. D. Bonnett, and C. P. L. Grof. 2000. Transgenic sugarcane clones can show normal agronomic performance in field trials. Proc. Austral. Soc. Sugar Cane Technol. 22:510. Vickers, J. E., C. P. L. Grof, G. D. Bonnett, P. A. Jackson, and T. E. Morgan. 2005a. Tissue culture and biolistic transformation of callus have independent effects on yield and sugar content of transgenic sugarcane in the field. Crop Sci. (In press).

2. SUGARCANE IMPROVEMENT THROUGH BREEDING AND BIOTECHNOLOGY

117

Vickers, J. E., C. P. L. Grof, G. D. Bonnett, P. A. Jackson, D. P. Knight, S. E. Roberts, and S. P. Robinson. 2005b. Directed over-expression and general evaluation of PPO activity in transgenic sugarcane is related to juice and raw sugar colour. Crop Sci. (In press). Vincentz, M., F. A. A. Cara, V. K. Okura, F. R. da Silva, G. L. Pedrosa, A. S. Hemerly, A. N. Capella, M. Marins, P. C. Ferreira, S. C. Franca, L. Grivet, A. L. Vettore, E. L. Kemper, W. L. Burnquist, M. L. P. Targon, W. J. Siqueira, E. E. Kuramae, C. L. Marino, L. E. A. Camargo, H. Carrer, L. L. Coutinho, L. R. Furlan, M. V. F. Lemos, L. R. Nunes, S. L. Gomes, R. V. Santelli, M. H. Goldman, M. Bacci Jr., E. A. Giglioti, O. H. Thiemann, F. H. Silva, M.-A. V. Sluys, F. G. Nobrega, P. Arruda, and C. F. M. Menck. 2004. Evaluation of monocot and eudicot divergence using the sugarcane transcriptome. Plant Physiol. 134:951–959. Voinnet, O., S. Rivas, P. Mestre, and D. Baulcombe. 2003. An enhanced transient expression system in plants based on suppression of gene silencing by the p19 protein of tomato bushy stunt virus. Plant J. 33:949–956. von Dassow, G., E. Meir, E. M. Munro, and G. M. Odel. 2000. The segment polarity network is a robust developmental module. Nature 406:188–192. Wahid, A. 2004. Analysis of toxic and osmotic effects of sodium chloride on leaf growth and economic yield of sugarcane. Bot. Bull. Acad. Sin. 45:133–141. Walker, D. I. T. 1972. Utilization of noble and Saccharum spontaneum germplasm in the West Indies. Proc. Int. Soc. Sugar Cane Technol. 14:224–232. Walker, D. I. T. 1987. Manipulating the genetic base of sugarcane. p. 321–334. In Anonymous (eds.), Copersucar International Sugarcane Breeding Workshop. Copersucar, Brazil. Wang, M. L., C. Goldstein, W. Su, P. H. Moore, and H. H. Albert. 2005. Production of biologically active GM-CSF in sugarcane: a secure biofactory. Transgenic Res. (In press). Watson, J. D., and F. H. C. Crick. 1953. A structure for Deoxyribose Nucleic Acid. Nature 171:737. Watt, D., A. McCormick, C. Govender, M. Cramer, B. Huckett, and F. Botha. 2005. Increasing the utility of genomics in unraveling sucrose accumulation. Field Crops Res. (In press). Wei, H., H. H. Albert, and P. H. Moore. 1999. Differential expression of sugarcane polyubiquitin genes and isolation of promoters from two highly-expressed members of the gene family. J. Plant Physiol. 155:513–519. Wei, H., P. H. Moore, and H. H. Albert. 2003. Comparative expression analysis of two sugarcane polyubiquitin promoters and flanking sequences in transgenic plants. J. Plant Physiol. 160:1241–1251. White, J. W., and G. Hoogenboom. 2003. Gene-based approaches to crop simulation: past experiences and future opportunities. Agron. J. 95:52–64. Wiegand, C., G. Anderson, S. Lingle, and D. Escobar. 1996. Soil salinity effects on crop growth and yield—Illustration of an analysis and mapping methodology for sugarcane. J. Plant Physiol. 148:418–424. Williams, D. G., V. Gempko, A. Fravolini, S. W. Leavitt, G. W. Wall, B. A. Kimball, P. J. Pinter, R. LaMorte, and M. Ottman. 2001. Carbon isotope discrimination by Sorghum bicolor under CO2 enrichment and drought. New Phytol. 150:285–293. Wingler, A., T. Fritzius, A. Wiemken, T. Boller, and R. A. Aeschbacher. 2000. Trehalose induced the ADP-glucose pyrophosphorylase gene, ApL3, and starch synthesis in Arabidopsis. Plant Physiol. 124:105–114. Wolfe, D. W. 1991. Low temperature effects on early vegetative growth, leaf gas exchange and water potential of chilling-sensitive and chilling-tolerant crop species. Ann. Bot. 67:205–212.

118

MING ET AL.

Wu, K. K., and A. Arcinas. 2004. Acreage census of cane varieties 2003. Varieties Report 462, Hawaii Ag. Research Center. Wu, K. K., W. Burnquist, M. E. Sorrells, T. L. Tew, P. H. Moore, and S. D. Tanksley. 1992. The detection and estimation of linkage in polyploids using single-dose restriction fragments. Theor. Appl. Genet. 83:294–300. Yang, M., R. Bower, M. D. Burow, A. H. Paterson, and T. E. Mirkov. 2003. A rapid and direct approach to identify promoters that confer high-levels of gene expression in monocots. Crop Sci. 43:1805–1813. Yin, X., P. Stam, J. J. Kropff, and H. C. M. Schapendonk. 2003. Crop modeling, QTL mapping, and their complementary role in plant breeding. Agron. J. 95:90–98. Zeevaart, J. A. D., and R. A. Creelman. 1988. Annu. Rev. Plant Physiol. Plant Mol. Biol. 39:439–473. Zhang, L. H., J. L. Xu, and R. G. Birch. 1999. Engineered detoxification confers resistance against a pathogenic bacterium. Nature Biotechnol. 17:1021–1024. Zhang, M.-Q., L. Guo-Jing, and C. Ru-Kai. 2001. Photosynthesis characteristics in eleven cultivars of sugar cane and their responses to water stress during the elongation stage. Proc. Int. Soc. Sugarcane Technol. 24:642–643. Zhou, J. H., J. L. Wang, J. C. Xu, C. L. Lei, and Z. Z. Ling. 2004. Identification and mapping of a rice blast resistance gene Pi-g(t) in the cultivar Guangchangzhan. Plant Pathol. 53:191–196.

3 Breeding for Resistance to Maize Foliar Pathogens* Richard C. Pratt Department of Horticulture and Crop Science Ohio Agricultural Research and Development Center The Ohio State University Wooster, OH 44691 Stuart G. Gordon Department of Plant Pathology Ohio Agricultural Research and Development Center The Ohio State University Wooster, OH 44691

I. INTRODUCTION II. DISEASES INCITED BY FUNGAL PATHOGENS A. Gray Leaf Spot 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress B. Northern Corn Leaf Blight 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress

*Journal paper HCS 04-08. The authors wish to express their sincere appreciation to Pat Lipps, and two anonymous reviewers, who provided valuable assistance by critically reviewing the manuscript.

Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 119

120

R. PRATT AND S. GORDON

C. Southern Corn Leaf Blight 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress D. Common Rust 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress E. Southern Rust 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress F. Downy Mildew 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress III. DISEASES INCITED BY VIRAL PATHOGENS A. Maize Dwarf Mosaic 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress B. Maize Chlorotic Dwarf 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress C. Maize Streak 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress D. Mal de Rio Cuarto 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress E. Maize Rayado Fino

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

121

1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress IV. DISEASES INCITED BY BACTERIAL PATHOGENS A. Stewart’s Bacterial Wilt 1. Disease Description, Importance, Range, and History 2. Sources of Resistance 3. Genetics of Resistance 4. Quantitative Trait Loci and Marker-assisted Selection 5. Progress V. SUMMARY AND CONCLUSIONS LITERATURE CITED

List of abbreviations: amplified fragment length polymorphism (AFLP); Centro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT; the international center for improvement of maize and wheat); gray leaf spot (GLS); International Institute of Tropical Agriculture (IITA); maize chlorotic dwarf virus (MCDV); maize dwarf mosaic virus (MDMV); maize rayado fino virus (MRFV); maize río cuarto virus (MRCV); markerassisted selection (MAS); northern corn leaf blight (NCLB); quantitative trait locus (QTL); restriction fragment length polymorphism (RFLP); recombinant inbred line(s) (RIL); sorghum downy mildew (SDM); southern corn leaf blight (SCLB); Stewart’s bacterial wilt (SBW). I. INTRODUCTION Foliar diseases are arguably the primary biotic constraints to worldwide yields of maize (Zea mays L.) (Simmonds and Smartt 1999). Most causal agents are widespread, but some tend to be more or less prevalent in particular regions, or during certain seasons, due to environmental factors (Smith 1999). Widespread epiphytotics are difficult to predict due to the inability to forecast long-term environmental conditions. In addition to periodic changes in the weather, changes in host-pathogen genetic interrelationships also may be induced by introduction of new cultivars or the occurrence of favorable mutations in the pathogen. Susceptible new cultivars may be severely affected by what were once considered minor pathogens, or new races of a pathogen could emerge and successfully attack previously resistant cultivars. The unexpected “eruption” of diseases is not a hypothetical scenario. Maize dwarf mosaic virus (MDMV)

122

R. PRATT AND S. GORDON

and maize chlorotic dwarf virus (MCDV) came out of nowhere in the early 1960s and became widespread pathogens in North America (Louie 1999). They are now controlled through planting of resistant hybrids and improved weed management (Duvick and Cassman 1999; Louie 1999) but they have spread to other parts of the world, including both Europe (Kuntze et al. 1997) and China (Smith 1999). In the well-known case of southern corn leaf blight (SCLB), causal agent Cochliobolus heterostrophus (Drechs.) Drechs. [anamorph: Bipolaris maydis (Nisikado & Miyake) Shoemaker], the disease increased dramatically during a period of favorable weather in the early 1970s because of widespread use of “T” cytoplasmic male sterility in production of commercial hybrids. The adoption of reduced tillage practices in the United States helped reduce erosion by leaving crop residues in the field, but also gradually created a niche in the U.S. Corn Belt that was exploited by Cercospora zeaemaydis (Tehon & E. Y. Daniels), causal agent of gray leaf spot (GLS). GLS has now spread to many regions of the world (Ward et al. 1999). Breeders and their colleagues have actively responded to outbreaks such as these, and in most instances, the effects of the pathogens have gradually been reduced through development of more resistant cultivars (Duvick and Cassman 1999). Both producers and breeders must constantly refine their priorities concerning the pathogens they consider most likely to cause economic damage. Host-plant-resistance breeding is practical, cost-effective, and environmentally friendly (Fehr 1987; Pratt et al. 2003) and should remain an integral part of, if not the core, of future disease management strategies. A trend is presently underway that could shift the host-pathogen dynamics in many parts of the world. Total world production of maize has overtaken that of rice (Oryza sativa L.) and wheat (Triticum aestivum L.). In developing countries, maize production is projected to exceed the combined production of rice and wheat by the year 2020 (Pingali and Pandey 2001). The increasing production of maize reflects both population growth and a growing need for livestock feed as consumers purchase more meat products. To meet the rising demand, an increasing shift from small-scale, complex, cropping systems to larger, more intensive ones will be inevitable in diverse regions of the world (Rosegrant et al. 2001). Expanded cultivation into less ideal environments, and changes in management practices—such as increased plant densities, double cropping, alteration in planting dates and fertility regimes, and the amount of crop residue left in the field, can influence disease incidence and severity. Increases in culture of genetically uniform cultivars also will create more favorable host-parasite population dynamics for the pathogens (Simmonds and Smartt 1999).

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

123

Some of the major pathogens of maize may incite disease in only a few crops, whereas others attack many crops. The causal agents may be characterized as either specialists or generalists, respectively (Parlevliet 2002). The defense mechanisms displayed by maize reflect the manner in which the host combats these different types of pathogens. Resistance to many pathogens that are generalists is usually of a quantitative nature. Quantitative resistance is also referred to as partial-resistance, and the terms will be used interchangeably in this review. Resistance to pathogens that are specialists is usually of a more qualitative (all or nothing) nature wherein “major” or “R” genes confer high levels of resistance, but may be more readily overcome following favorable mutation by the pathogen. Major-gene resistance is ephemeral and typically fails within a period of time following deployment, although the longevity is difficult, if not impossible, to predict. Prolonged selection of the host through modern breeding typically narrows the genetic base over time and predisposes the crop to genetic vulnerability. The reliance on relatively few resistance alleles in a small number of genotypes is well illustrated by the narrow genetic base of the U.S. maize crop (Kraja et al. 2000). The fact that field maize is grown on more than 28 million hectares in the United States creates a scenario of vulnerability that predisposes the maize crop to possible changes in pathogen populations. An example of the defeat of major-gene resistance in the U.S. maize crop that is not yet published, but is common knowledge among researchers, is the loss of effective control of northern corn leaf blight (NCLB) using Ht genes (Leonard et al. 1989). The causal agent of NCLB is Setosphaeria turcica (Luttrell) K. J. Leonard & E. G. Suggs [anamorph: Exserohilum turcicum (Pass.) K. J. Leonard & E. G. Suggs]. The Ht resistance was widely deployed and virulent races of E. turcicum developed within a period of approximately two decades (Smith 1999). Awareness of the need to emphasize partial resistance (or quantitative resistance, as opposed to qualitative resistance) appears to be increasing. Redirection of breeding efforts toward increasing partial resistance is a positive development in long-term disease management. Hopefully, breeders in regions with gradually intensifying maize production also will adopt partial-resistance breeding strategies. Nevertheless, deployment of major-gene resistance is likely to continue, and new selection methods that enable the combination of major-genes with partial resistance are needed (Parlevliet 1983). Disease-resistance breeding, especially for partial resistance, is complicated by the variable nature inherent in both the incidence and severity of natural disease epiphytotics. The complexity arising from a host

124

R. PRATT AND S. GORDON

organism, a pathogenic microorganism, the interaction between them, and their interaction with the environment can make expedient and efficient selection difficult. Testing across locations or seasons and/or artificial inoculation systems are usually necessary to accurately characterize host responses. The situation is further complicated, especially with viruses, when an insect vector carries the pathogen and may itself be subject to various environmental vagaries. Development of highly efficient inoculation techniques or identification of sites with consistent, severe disease pressure may be used to overcome fluctuations in disease pressure (Geiger and Heun 1989). Selection for improved disease resistance must be undertaken in the overall context of the goals of a breeding program. Other traits may be considered just as important, or more important, than disease resistance. Combining competitive yielding ability and other agronomic characteristics with disease resistance in a cultivar can be quite difficult. Gain through selection is diminished for each additional trait under selection, and some disease-resistance factors may be linked to genes that detract from yield or other agronomic traits. Breeders are forced to prioritize the traits that will be emphasized in their program, and the cost-benefit relationship between added disease resistance and its relationship to other traits is always scrutinized closely. Discovery of new resistance genes will be critical to stop the advance of diseases as pathogens evolve to overcome those currently deployed in the field. The development of modern commercial hybrids may yet undergo re-direction using genetic engineering of single genes, and this may create opportunities for combining them with those identified through traditional methods. Focusing on gene-stacking in the same elite genetic backgrounds deployed over vast regions could increase vulnerability to changes in pathogen populations (Michelmore 2003). Durability, or the longevity of resistance in the presence of dynamic pathogen populations in the long run, can be improved by attention to the manner in which new cultivars are developed and deployed in a region (Johnson 1984). Many factors can influence the durability of resistance, and it has been observed to be highly variable from one pathovar to another (Hogenboom 1993). An emphasis on quantitative resistance breeding should enhance durability (Parlevliet 1992, 2002). Simmonds (1985) correctly asserted that plant breeders working in developing countries have a moral and professional responsibility to breed for durable resistance because decreased yields or crop failures may result in more than economic hardships alone. If breeders are successful in developing useful cultivars that can meet the challenges for

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

125

increased production in diverse agro-ecosystems, and reduce losses to prevailing diseases through resistance breeding, everybody wins. This review will emphasize the progress realized during the last 20 years in resistance breeding and research concerning a dozen foliar diseases of maize. We will also address the prospects and limitations of marker-assisted selection (MAS) for improvement of resistance. It is not possible to cover all diseases of importance. We hope that at least some of the diseases we have chosen will be of interest to the majority of those concerned with breeding for disease resistance.

II. DISEASES INCITED BY FUNGAL PATHOGENS A. Gray Leaf Spot 1. Disease Description, Importance, Range, and History. Cercospora zeae-maydis (Tehon and E. Y. Daniels) is the causal pathogen of GLS. Two sibling types have been identified (Wang et al. 1998) and type II is considered to be the most prevalent in the United States (Carson et al. 2002) and sub-Saharan Africa (Okori et al. 2003). The pathogen is an imperfect fungus for which there is no known sexual stage. Isolates of the fungus vary in aggressiveness, but no races have been observed (Bair and Ayers 1986; Dunkle and Carson 1998; Carson et al. 2002). Rectangular lesions on the foliage make GLS easily recognizable. Some genotypes display different types of lesions during disease progress, e.g., chlorotic flecks vs. large, tan, necrotic lesions (Freppon et al. 1994, 1996). GLS causes economic damage by destroying the photosynthetically active leaf surface and it also results in reduced stalk quality (Lipps et al. 1997b). The disease was once largely confined to the mountainous regions, and low-lying riverbank areas, of the eastern United States. Since the early 1990s, it has increased in incidence, particularly in North America, due primarily to an increase in conservation tillage practices (Ward et al. 1999). Reduced tillage practices enable survival and build up of spores of C. zeae-maydis because the pathogen readily overwinters on crop residue left in the field (de Nazareno et al. 1993; Dunkle and Carson 1998). GLS is now a primary foliar disease of maize in the United States and sub-Saharan Africa (Ngwira et al. 1999; Okori et al. 1999; Ward et al. 1999) and it is also present in South America and Asia (Carson 1999b; Ward et al. 1999; Smith 1999). Yield losses in excess of 50% have been reported during GLS epidemics in the United States (Latterell

126

R. PRATT AND S. GORDON

and Rossi 1983; Lipps 1987). Research conducted in the Republic of South Africa (RSA) has demonstrated GLS yield reductions from 30 to 60%, depending on the hybrid and environmental conditions (Ward et al. 1999). 2. Sources of Resistance. Resistance has been identified in temperate U.S. germplasm; however, there have not been high levels of resistance available in maturity classes suitably adapted to the U.S. Corn Belt (Lipps et al. 1997a) or in Iowa Stiff-Stalk Synthetic-related germplasm (Graham et al. 1993). Publicly available U.S. sources of resistance characterized in earlier studies include: Mo18W, NC250, NC250A, NC258, NC290, Pa875, T222, Va14, Va17, Va59, and Va85 (Thompson et al. 1987; Ulrich et al. 1990; Bubeck et al. 1993; Saghai-Maroof et al. 1993). A study by Coates and White (1994) examined 1,396 inbreds and the average rating of the population (1 to 5 scale, 5 is completely susceptible) was 3.5 during the more intense of two screening seasons. Several inbreds (Ky128, Mo18W, Mp311, and Va3a) were rated highly resistant, but were considered unsuitable for breeding purposes in the Corn Belt because of their late maturity. Several other resistant accessions were selected for further studies in testcrosses with elite Corn Belt testers (Coates and White 1998), and those that imparted the highest levels of resistance were considered suitable as resistance donors. The selected lines were derived from diverse germplasm: B37HtN (PI 406119), from South African breeding material (it is unclear what proportion of the pedigree is actually B37); DS74:1071, from an unknown synthetic population; and 061 (from PI 320061—an accession from Brazil that is no longer in the collection of the USDA-ARS National Plant Genetic Resources Program) (USDA-ARS-NGRP 2004). Recent evaluations in Illinois and Virginia of 1,237 accessions from the USDA germplasm collection have revealed that approximately 2% of the genotypes are resistant and another 2% are partially resistant (USDA-ARS-NGRP 2004). Several accessions derived from South African germplasm related to the B37HtN accession identified by Coates and White (1994), and one South African accession apparently derived from inbred K64HtN breeding materials, were highly resistant. Additional accessions considered partially resistant were Branco Redondo × 366 4-7-5 (Angola); a breeding line perhaps derived from inbred C103 (PI 406109, South Africa); inbred Va102; and Florida synthetic population FS8B(T) (USDA-ARS-NGRP 2004). Additional resistant germplasm adapted to southern Africa includes modified opaque-2 maize inbreds KO54W and SO507W characterized by Gevers et al. (1994) and VO613Y characterized by Gordon et al. (2004). Temperate-adapted lines derived from diverse germplasm at

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

127

North Carolina State University that are reported to display high levels of GLS resistance include NC262A, NC264, NC270, NC288, NC290, NC290A, NC318, NC320, NC332, NC334, NC334, and NC354 (Major Goodman, NCSU, pers. comm.). Other resistant lines include CML440, CML443, and CML445 (CIMMYT 2000b). Composite populations, e.g., PA Plt. Path. Composite (GLS)C1 (LD), with resistance to GLS are also available (Johnson and Ayers 1988c). 3. Genetics of Resistance. Researchers examined the genetic basis of C. zeae-maydis resistance using diallel crosses and generation means analysis (Thompson et al. 1987; Elwinger et al. 1990; Ulrich et al. 1990; Gevers et al. 1994; Coates and White 1998). From those studies it was concluded that resistance is under additive genetic control with some dominance effects. Gevers and Lake (1994) presented evidence for single-gene resistance to C. zeae-maydis in South African germplasm, but subsequent studies suggested two factors were associated with resistance (Gordon et al. 2004). Freppon et al. (1996) examined the chlorotic lesion phenotype conferred by inbred NC250A in segregating progeny lines at mid-epiphytotic and concluded a single gene governed the chlorotic lesion response; however, transition of the lesion phenotype from chlorotic to necrotic in some progenies later in the season was not consistent with a single gene model. Most resistant genotypes are considered to have two or more factors associated with resistance. 4. Quantitative Trait Loci and Marker-assisted Selection. Genetic mapping experiments have identified quantitative trait loci (QTL) for resistance to C. zeae-maydis on all 10 chromosomes (Bubeck et al. 1993; Saghai-Maroof et al. 1996; Clements et al. 2000; Lehmensiek et al. 2001; Gordon et al. 2004). Bubeck et al. (1993) used inbreds NC250A and ADENT as sources of resistance in three F2:3 mapping populations. They identified significant QTL on five different chromosomes. Only the one on the short arm of chromosome 2 was found consistently over three environments. Saghai-Maroof et al. (1996) employed selective genotyping to identify three QTL on chromosomes 1, 4, and 8 that collectively explained 44 to 64% of the variation across two generations, F2 and F2:3, across two seasons in one location. The resistant parent, inbred Va14, contributed QTL on chromosomes 1 and 8. The susceptible parent, inbred B73, contributed the QTL on chromosome 4. In tests for epistatic interactions, the authors demonstrated the chromosome 4 QTL had little or no effect when the QTL on chromosome 1 was homozygous for the Va14 allele. The QTL on chromosome 8 displayed recessive gene action. Using the inbred 061 as a resistant parent, Clements et al. (2000) evaluated a BC1S1 population for two years at one site and one year at a second site.

128

R. PRATT AND S. GORDON

Five QTL, all from the resistant parent, were significantly associated with resistance to C. zeae-maydis in both years and locations. Bulked segregant analysis was used by Lehmensiek et al. (2001) to identify QTL on chromosomes 1, 3 and 5 associated with resistance in an F2 population derived from South African proprietary parental lines. The QTL were confirmed in sister F2 populations evaluated in two environments. Gordon et al. (2004) used a selective genotyping strategy to examine the linkage of molecular markers to resistance in segregating progenies obtained from a cross between the resistant South African inbred line VO613Y and Pa405 (susceptible). Disease plots were located in both Ohio and South Africa. Kruskal-Wallis analysis (Gordon et al. 2004) and composite interval mapping (Gordon et al. 2004) identified significant marker intervals near umc137 on chromosome 2L (bnlg1520-umc36) and an interval on chromosome 4L containing umc127. The previous studies collectively utilized five resistant inbreds, and a resistance QTL on chromosome 1S (bin 1.05–1.06) is the nearest to a consensus QTL, with three of the five inbreds contributing resistance at that location. 5. Progress. Breeders in the United States have undertaken a concerted effort to improve the resistance of elite germplasm and these efforts are ongoing. Substantial improvement has been noted in the degree of resistance present in elite U.S. Corn Belt hybrids. A marker-assisted selection (MAS) strategy could accelerate the breeding process by allowing indirect selection of resistant individuals based on marker genotype. MAS for recessive alleles, such as the QTL on chromosome 2 identified by Gordon et al. (2004), would be especially useful. Coupling MAS with phenotypic selection would be likely to yield the greatest gain, although evidence to that effect is unavailable at this time. Developing maize inbreds and populations with resistance to C. zeaemaydis is a high priority in many areas where the pathogen has more recently become endemic. Large areas of sub-Saharan Africa are experiencing increased maize production in environments that are favorable to disease development. Development of durable resistance (Simmonds 1985) should be the top priority. Identification of mapped C. zeae-maydis resistance loci from different resistance sources now provides the opportunity for combining resistance loci. Such resistance QTL combinations could provide more durable protection against C. zeae-maydis infection than might be obtained with resistance conferred by only a single source. Breeders presently rely on field screening to assess disease resistance. If controlled inoculation protocols could be developed that would allow off-season screening to take place in the greenhouse, this would be a very useful tool. Some initial progress has been made in this regard, but further refinements are needed (Asea et al. 2003).

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

129

B. Northern Corn Leaf Blight 1. Disease Description, Importance, Range, and History. Northern corn leaf blight is caused by Exserohilum turcicum (Pass) K. J. Leonard and E. G. Suggs [teliomorph: Setosphaeria turcica (Luttrell) Leonard and Suggs] and at least seven pathogenic races have been described (Carson 1999a). Infection of the host causes characteristic gray-green, elliptical or cigar-shaped lesions that are typically 3–15 cm in length. Major resistance genes condition long, chlorotic, linear steaks that may sometimes result in confusion with symptoms caused by Stewart’s wilt. Mycelia and conidia can overwinter in the leaf residue and infect subsequent crops. Conidiophores can also be transported long distances by wind. NCLB occurs throughout maize-producing regions wherever moderate temperatures and humidity prevail (Carson 1999a: Smith 1999). It has been increasingly visible in the eastern United States during the past two seasons (Patrick Lipps, OSU, pers. comm.). NCLB can significantly reduce maize yield in many production regions (Latterell and Rossi 1983). It is a potentially devastating disease that routinely limits maize productivity in sub-Saharan Africa, especially in the humid mid-altitude and highland regions (Ngwira et al. 1999; Okori et al. 1999; Ward et al. 1999; Bigirwa et al. 2001; DeVries and Toenniessen 2001). 2. Sources of Resistance. Many sources of qualitative and quantitative resistance have been identified and they are presented in a review article published by Welz and Geiger (2000). In this section, only those lines that have been reported during the last 20 years will be highlighted. Promising resistant inbred lines C.I.66, C.H.586-12, H60, H95, H99, Mo42, Ms75, Oh509A, and W570 were identified in an inter-regional evaluation (Darrah 1985). Host-resistance research has demonstrated that partially resistant inbreds, such as H99, may confer high levels of resistance in crosses (Lipps et al. 1997a; Hakiza et al. 2004), even in tropical environments where major gene resistance is unstable and may provide only limited resistance (Welz and Geiger 2000; Hakiza et al. 2004). Warren (1981, 1982) reported inbred releases (H102, H103, H110, and H111) that are resistant. Inbreds Pa392, Pa860, Pa877, Pa879, and Pa891 also display resistance to E. turcicum (Johnson and Ayers 1988a; Johnson and Ayers 1988b; Johnson 1989, 1992). Resistant lines of earlier maturity adapted to the northern Corn Belt include SD40, SD41, SD42, SD44, SD46, SD47, SD48, and SD101 (Wicks et al. 1986a,b; 1987, 1989, 1990) and inbred ND256 (Cross 1984). A recent release from Iowa State University, inbred B115, also is reported to display higher-than-average resistance to NCLB (Hallauer et al. 2001). Evaluations of tropical germplasm and temperate inbreds adapted to tropical conditions (Brewbaker et al. 1989) revealed many lines that

130

R. PRATT AND S. GORDON

display resistance. Among those rated highly resistant were CM118 (India), Fla2AT115 and Fla2AT116 (Florida), ICA 127 (Columbia), H55 (Indiana), Hi39 (Hawaii), and F (Kenya; a parent of hybrid H632). Sharma and Payak (1990) reported on the durability of resistance displayed by inbred lines CM104 and CM105 in India. Welz et al. (1998a) indicated early-maturing, central-European lines 1511C, 493B, DI64, D305, and R2038 are resistant. Recent evaluations of the USDA collection revealed one highly resistant accession, ‘Lanzarote’, from the Canary Islands. In another evaluation of USDA germplasm accessions inoculated with race O in Iowa, Nariño 605 (PI 445237 from Colombia) and maize inbred lines B93 and B100 exhibited high levels of resistance. CIMMYT lines with resistance and adaptation to tropical conditions are CML434, CML437, CML438, and CML439 (CIMMYT 2000a). CIMMYT lines of sub-tropical adaptation that display resistance are CML102, CML117, and CML132. Resistant lines with adaptation to mid-altitude conditions are CML443, CML444, and CML445 (CIMMYT 2004). Breeding populations with resistance are OhS10(C1) (Pratt et al. 1994b) and RSSSCC6 (Lambert 1996). 3. Genetics of Resistance. The presence of both qualitative (major gene) and quantitative (polygene) modes of resistance to E. turcicum has been clearly identified (Raymundo and Hooker 1982; Welz and Geiger 2000). Quantitative resistance to E. turcicum is inherited polygenically (Pataky et al. 1986). Six dominant genes (Ht1, Ht2, Ht3, Htm1, Htn1, and NN) and one recessive gene (ht4) control resistance to specific races of E. turcicum (Welz and Geiger 2000). The ht4 resistance factor was ineffective in field trials in Kenya (Welz and Geiger 2000) and Ohio (R. Pratt, OSU, pers. obs.). Sporulation is suppressed or delayed in plants carrying major dominant genes singly or in combination. Minimum estimates of the number of resistance genes associated with partial resistance ranged from three to six (Hughes and Hooker 1971; Jenkins and Robert 1961). Hakiza et al. (2004) estimated the number of genes involved with quantitative resistance of maize inbred H99 and concluded there were over 20. Brewster et al. (1992) studied the genetics of components of resistance to NCLB and concluded that incubation period and lesion number appeared most important in expression of partial resistance. 4. Quantitative Trait Loci and Marker-assisted Selection. Several groups mapped QTL responsible for quantitative resistance to E. turcicum (Freymark et al. 1993, 1994; Dingerdissen et al. 1996; Welz et al. 1998a; Schechert et al. 1999). Freymark et al. (1993, 1994) concluded lesion

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

131

number in a segregating population obtained from the cross B52 × Mo17 was controlled by three QTL, while five QTL controlled disease severity in an environment of low disease severity. The three QTL affecting lesion number were the same as those affecting disease severity (Freymark et al. 1993, 1994). Dingerdissen et al. (1996) studied the same population in a tropical environment characterized by high disease severity (Kenya). Those authors conducted interval mapping and reported four of eight QTL for severity were consistent across locations. Four of the QTL were in agreement with those reported by Freymark et al. (1994). Three QTL from chromosomes 3, 5, and 8 were consistent across studies and may be considered consensus QTL. 5. Progress. It can perhaps be stated that the deployment of major gene resistance in U.S. Corn Belt hybrids during the 20th century was overdone. Races of E. turcicum capable of overcoming the Ht genes initially deployed have been discovered (Smith and Kinsey 1993). Quantitative resistance, alone or in combination with Ht genes, is now necessary to manage NCLB. Ongoing efforts to breed for resistance, with renewed focus on quantitative resistance, are being undertaken in many temperate zone breeding programs. Breeders in tropical environments emphasized quantitative resistance and have been successful in developing high levels of durable resistance (Sharma and Payak 1990; Paliwal 2000). NCLB is now seen as a problem only in unadapted, exotic, or introduced temperate germplasm in tropical environments (Paliwal 2000). A considerable body of scientific literature has been amassed for this cosmopolitan disease of maize. Much progress has been made in understanding the genetics of resistance, but it is not yet clear if an additive effect of qualitative and quantitative resistance can be realized (Lipps et al. 1997a; Carson 1999a). Several QTL studies have been completed, but it has not yet been reported whether MAS will enhance the efficiency of selection of resistance to E. turcicum. Welz and Geiger (2000) caution that the majority of QTL positions are still not well defined, and additional fine-mapping needs to be undertaken. C. Southern Corn Leaf Blight 1. Disease Description, Importance, Range, and History. The causal agent of SCLB, C. heterostrophus (Drechs.), is a hemi-necrotrophic plant pathogen (Carson 1999a). The fungus overwinters on maize debris in previously diseased fields. The disease cycle from primary infection of growing leaves to sporulation from lesions and re-infection can occur

132

R. PRATT AND S. GORDON

under ideal conditions in as little as two to three days. Pathogen infection causes foliar lesions that vary in size and shape depending on the race of the pathogen and the genetic background of the host. Typical lesions following infection by race O are small, tan in color, and show rectangular growth that is limited by the veins. Lesions resulting from race T infection are more yellow-green in color, larger in size, and spindle shaped or elliptical. Severe blighting of the leaves causes predisposition of stalks to rotting. Some races may produce toxins. SCLB is common in many maize-growing areas in both temperate and tropical regions. Disease epidemics are most frequent and severe in warm-to-hot, humid regions, such as the lowland tropics (Cardwell et al. 1997b), and may occur occasionally in more temperate regions of the U.S. Corn Belt (Carson 1999a). The disease has been most commonly caused by race O. SCLB was not considered an important pathogen until the early 1970s when race T developed. Race T incited disease that reached epidemic proportions during the period 1970 to 1971, starting in North America, because it was pathogenic on U.S. hybrid maize containing Texas malesterile cytoplasm (cms-T). Race T is now rare in North America because cms-T cytoplasm is no longer employed in hybrid maize production. Race C has been reported to be specifically virulent on cms-C sterile maize cytoplasm. It is considered to be present only in China (Carson 1999a), although one monitoring project has been unable to detect it (Smith 1999). Use of the cms source of male sterility has ceased and breeding for polygenic resistance has been emphasized. Resistance tends to be quantitative and additive in effect, although one qualitative recessive gene, rhm, which conditions resistance in immature plants, has been found (Carson 1999a). The disease is also an important problem in sweet maize production. 2. Sources of Resistance. The recessive gene rhm was isolated from the composite population ‘Ibadan A’ from Nigeria and it was incorporated in some inbred lines (Carson 1999a). The resistant inbred line NC250 contains two recessive genes (not allelic with rhm) that confer resistance that does not break down in mature plants. It has been used as a resistance donor in the development of additional resistant lines. High levels of partial resistance are found in tropical and southern U.S. germplasm (Carson 1999a). Resistant sources include Fla2AT114, Fla2AT115, FlaAT116, FlaBT106, Tzi3, Tzi5, Tzi8, and Tzi11 [Tzi lines were developed and released by the International Institute of Tropical Agriculture, (IITA)], Ibadan, Nigeria; KU1414 from Thailand (Brewbaker et al. 1989); and NC290A, NC318, NC320, and NC332, released by North

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

133

Carolina State University (M. Goodman, NCSU, pers. comm.). Improved inbred releases from North Carolina State University with NC250 derived SCLB resistance are NC270, NC274, NC292, NC314, NC316, NC324, NC326, NC328, and NC330 (M. Goodman, NCSU, pers. comm.). Evaluation of U.S. inbred lines from various maturity groups revealed B85, C.H. 586-12, H84, H95, H99, Mo42, Mp496, MS75, NY562, Pa762, Pa872, and T159 were resistant to infection by B. maydis (USDA 2004). Other resistant inbreds include DE811 (Hawk 1985), Mp313E (Scott and Zummo 1990), Pa891 (Johnson 1989), and W570 and W576 (Coors and Mardones 1989). Resistant lines from CIMMYT with tropical adaptation are CML247, CML248, CML251, CML254, CML258, CML259, CML274, CML275, CML276, CML295, CML298, CML304, and CML447 (CIMMYT 2002b; CIMMYT 2004). Breeding populations with resistance are FS8A(S), FS8A(T), FS8B(S), and FS8B(T) (Horner 1990) and RSSSCC6 and RBS10 C6 (Lambert 1996). 3. Genetics of Resistance. Resistance is both qualitative and quantitative in nature (Paliwal, 2000) and may differentially affect immature or mature plants (Carson 1999a). The recessive gene rhm confers chloroticlesion resistance to B. maydis race O in susceptible maize plants. A source of resistance related to the original Nigerian germplasm from which the recessive rhm gene was originally isolated was examined by Thompson and Bergquist (1984) and they were able to identify matureplant resistance. Resistance appeared to be controlled by two complementary recessive genes. Holley and Goodman (1984) identified and characterized additional partial resistance factors isolated by extracting inbreds from tropical hybrids of diverse origin. When resistant tropical inbreds were crossed with Corn Belt inbred B73 (susceptible), it was determined that hybrids ranged from moderately resistant to susceptible. They interpreted the range in resistance to likely indicate both additive and recessive types of gene action were operative. Evaluations of resistant phenotypes of segregating progenies derived from resistant × resistant crosses also suggested genes carried by some inbreds were different from those of other tropical inbreds. 4. Quantitative Trait Loci and Marker-assisted Selection. Zaitlan et al. (1993) mapped rhm in F3 families derived from the cross RH95rhm × B73. The expression of rhm in maize seedlings inoculated with conidia of B. maydis race O was observed under controlled conditions. The genotype at the rhm locus of each F2 parent was deduced from the reactions observed in the progeny seedlings. The results indicated the rhm locus maps to the short arm of chromosome 6. In a recent study by Cai

134

R. PRATT AND S. GORDON

et al. (2003), amplified fragment length polymorphism (AFLP) markers were used to map a large F2 population. Bulked-segregant analysis identified molecular markers linked to the rhm gene for resistance to SCLB. One co-dominant AFLP marker, p7m36, mapped to a position 1.0 cM from rhm, and was converted to a sequence-tagged site marker. 5. Progress. Deployment of resistance has considerably reduced the extent of damage by the pathogen (Paliwal 2000). High levels of partial resistance are now found in much of the tropical and southern U.S. germplasm, and acceptable levels may be found in many hybrids grown in the Corn Belt (Carson 1999a). The recently identified restriction fragment length polymorphism (RFLP) and AFLP markers may be useful for map-based cloning of the rhm gene and MAS for rhm (Cai et al. 2003). It will be important to ensure that adequate levels of partial resistance are also deployed to safeguard against the formation of new races of the pathogen. D. Common Rust 1. Disease Description, Importance, Range, and History. Common (or temperate) rust, caused by Puccinia sorghi Schwein, is a foliar disease that can affect maize whenever conditions are favorable. Historically, its importance has peaked and subsided. In 1925, yield losses were 76,272 Mg (metric tonnes or 3,000,000 bu at 56 lb/bu) in Iowa alone (Arthur 1929), but by the middle of the 20th century, common rust was considered a minor disease of maize in the United States. (Ullstrup 1953). In 1993, an epidemic of common rust again occurred in the United States (Smith 1999). Common rust should continue to be considered a potential threat because of the possibility of attack by new or unique races of the pathogen (Pataky 1999). Despite the scientific name, this pathogen does not infect sorghum [Sorghum bicolor (L.) Moench]. Grain yield losses are estimated at between 3 and 8% for every 10% of leaf area diseased (Pataky 1999). The fact that uredinia sporulate on the adaxial and abaxial leaf surfaces distinguishes it from southern rust (causal agent Puccinia polysora Underw.), which sporulates primarily on the adaxial surface. Wind currents carry uredinospores north to the U.S. Corn Belt from tropical and subtropical regions. Infection usually occurs during favorable conditions and later in the growing season. This typically results in first infection from mid-June to midJuly if temperatures are moderate (16° to 25°C) and periods of high relative humidity (95%) last for 6 h or more. Teliospores are produced in the rust

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

135

pustules as the host matures and both uredial and telial forms of P. sorghi are present everywhere maize is grown (Nyvall 1989). Young leaf tissue is more susceptible than mature tissue, and this may be due the higher amount of cuticular wax on mature leaf tissue. 2. Sources of Resistance. Several P. sorghi resistance genes from many different sources have been identified. In maize, more than 25 dominant, rust-resistance genes have been reported (Hu and Hulbert 1996). Public inbred maize lines have been converted to P. sorghi resistant lines via backcross conversions to introgress individual resistance genes (Groth et al. 1992). In a well-known conversion program, the resistant exotic donor population from South America, ‘Cuzco’, was utilized for introgression of resistance to elite line B14. The resultant inbred, B14A, displays high resistance to all known biotypes of corn leaf rust (Russell et al. 1971). Wilkinson and Hooker (1968) identified Rp genes in eight maize inbred lines from Europe and Africa. Sources of high partial resistance include Oh545 and CM111 (Kim and Brewbaker 1977). Highlandadapted lines CML239 through CML246 are resistant (CIMMYT 2004). 3. Genetics of Resistance. Resistance to P. sorghi falls under two categories: specific, which is controlled by single genes; and general, which is controlled by many genes acting in concert. Single, P. sorghi-resistance genes, or Rp genes, are usually dominant and have been identified at five different genetic loci (Pataky 1999). These Rp genes govern the racespecific hypersensitive response to P. sorghi infection. Many Rp genes reside at complex loci that have been utilized in studies of basic genetic mechanisms, especially the Rp1 locus (Hulbert et al. 1991; Hulbert 1997). General resistance is expressed as reduced severity of infection, lower number of rust pustules, reduced sporulation, and increased length of latent period (Pataky 1999). This type of resistance is controlled by multiple genes, is effective against all biotypes of P. sorghi, and is expressed mostly in mature plants (Hu and Hulbert 1996). Expression of general resistance to P. sorghi is very dependent on the environment, and general resistance that is effective in less conducive environments usually breaks down in tropical environments. Due to the abundance of singlegene resistance, limited attention has been paid to multi-gene resistance to P. sorghi. Kim and Brewbaker (1977) studied the inheritance of partial resistance in a diallel evaluation of 11 maize inbred lines. They found that the general combining ability effects for partial resistance were high for inbreds

136

R. PRATT AND S. GORDON

Oh545 and CM105, but specific combining ability, though statistically significant, was low. Inbreds Oh545, CM105, and CM111 contributed smaller lesion sizes to their progeny. In progenies of other inbreds, they reported variation in lesion number and amount of leaf chlorosis, and cited these findings as evidence of different mechanisms for expression of partial resistance. 4. Quantitative Trait Loci and Marker-assisted Selection. Lubberstedt et al. (1998) mapped QTL associated with partial-resistance to P. sorghi in four European flint maize populations across two locations. They found QTL on all 10 chromosomes, and the populations contained between four and 13 QTL. There was no association with the location of the Rp loci and the resistance QTL. The parents of the populations they employed are not especially susceptible or resistant, and the progeny transgressed the parental values for both resistance and susceptibility. The authors reported heritability estimates of partial resistance were high (0.64 and above), with the QTL explaining 33 to 71% of the phenotypic variance. QTL × environment interaction effects were frequently noted as well. Given these results, prospects should be good for MAS. 5. Progress. Mass selection by local farmers in East and West Africa (Pataky 1999; Paliwal 2000) successfully developed general resistance in their cultivars before qualitative resistance became available. Breeders in North America were fortunate to identify resistance factors that could be incorporated into temperate elite germplasm. Selection for resistance has been successfully carried out under both conditions of natural inoculation and using spore suspensions. Breeding for hostplant-resistance has been successful in preventing severe losses due to rust epidemics. Taking advantage of the complex genetic nature of the Rp1 locus, Hu and Hulbert (1996) demonstrated that “compound” rust resistance genes could be constructed with two or more resistance genes combined in coupling phase. Some constraints to this approach are the very small genetic distance separating the alleles, often less than 0.5 cM, and the availability of P. sorghi races with which to test the progeny of genetic crosses. Much progress has yet to be made in identification and manipulation of P. sorghi resistance QTL. This is no doubt due, in large part, to the abundance of race-specific resistance genes. However, a race of P. sorghi exists that can overcome any known Rp allele. Therefore, quantitative resistance should be identified and exploited to incorporate durable

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

137

resistance into cultivars. Kraja et al. (2000) identified alleles for resistance to multiple pathogens, including P. sorghi, from tropical and temperate inbred lines. They suggested selfing the progeny of resistant × susceptible crosses, rather than backcrossing, and developing new inbred lines as the best strategy for utilizing the resistance alleles they identified. E. Southern Rust 1. Disease Description, Importance, Range, and History. Southern rust is incited by Puccinia polysora Underw. Physiologic races of P. polysora are known. Symptoms of southern rust are quite similar to those of common rust; however, development of pustules on the lower side of the leaf is sparse in comparison with common rust (Pataky 1999). Uredinia are also common on husk leaves, shanks, stalks, and leaf sheaths. The telia resemble those of common rust, but they tend to remain covered by the epidermis for a longer time. Urediniospores are carried by wind to maize plants and act as both primary and secondary inoculum. They can be distinguished from those of common rust because they are darker and rounder. Scott and Zummo (1989) considered the most important difference between common and southern rust to be the distinction that southern rust can kill the host. Epiphytotics of southern rust are favored by warm, humid environments; hence, it is normally found in sub-tropical and tropical environments. Recent observations in Cameroon are in agreement: the highest incidence of P. polysora was in the lowland-tropical areas in a survey conducted throughout the country (Cardwell et al. 1997b). Southern rust has been reported in Africa, Southeast Asia, The Philippines, Australia, southern and central North America, South America, and the West Indies (Pataky 1999; Smith 1999). During the 1940s, epiphytotics occurred throughout West Africa. Southern rust does not frequently extend far to the north in North America. Several southern rust epidemics occurred in the southern United States during the 1970s (Bailey et al. 1987) and again during the 2003 season (Major Goodman, NCSU, pers. comm.). In Brazil, southern rust is one of the most common diseases of maize, especially in the midwest and southwest regions of the country (Godoy et al. 2003). The importance of southern rust has increased in Brazil during recent years, due to expanded cultivation of maize. In northeastern Tucuman Province of Argentina (a subtropical region), southern rust was observed and yield losses were experienced during the 2000–2001 growing season (Hernandez et al. 2002).

138

R. PRATT AND S. GORDON

2. Sources of Resistance. Many local farmer varieties in sub-Saharan Africa display general (partial) resistance (Paliwal 2000). Scott et al. (1984) were able to make selections of resistant lines from four populations containing tropical or broad-based tropical × U.S. germplasm. The populations designated ‘Nigeria Composite A’ and ‘Corn Belt × ‘Columbian’ yielded the largest number of lines homozygous for majorgene resistance to P. polysora. An evaluation of “slow-rusting” resistance of maize inbreds revealed several lines from Texas (Tx94, Tx601, and Tx706) and Corn Belt inbred Mo17 to be resistant (Bailey et al. 1987). In an evaluation of 120 openpedigree field maize inbreds in Nigeria and the United States (Texas and Hawaii), Brewbaker et al. (1989) showed general resistance in tropical germplasm to be common, with maize inbreds Hi34, ICA L223, and L219 (Colombia); CIM. T 11-ES (CIMMYT); Nariño 330-S5 (Thailand); INV 534 (Texas); and Tzi14 (IITA) showing the highest levels of resistance. Tallury and Goodman (1999) demonstrated that temperateadapted maize inbreds NC296 and NC300 derived from 100% tropical pedigrees produced single crosses with very high resistance to southern rust at a site with high disease severity in North Carolina. CIMMYT lowland tropical lines CML3, CML9, CML10, CML19, CML33, CML426, CML447, CML450, CML452, and CML453 are resistant (Vasal et al. 1997; CIMMYT 2004). Clerget et al. (1996) have described the tropical line CIRAD390 as “fairly tolerant of southern rust.” CIMMYT population no. 304 (‘Population Hoja Erecta’) is considered to offer good resistance (CIMMYT 2004). 3. Genetics of Resistance. Ullstrup (1965) obtained a resistant selection from the South African accession ‘Boesman Yellow Flint’ (PI 186208) with qualitative resistance to southern rust race PP.9. The single, dominant gene that conferred resistance was designated Rpp9. The location of that resistance factor (bin 10.01) is very near that of the Rp1 resistance gene (confers resistance to common rust) on the short arm of chromosome 10 (bin 10.01). Scott et al. (1984) concluded the resistant lines obtained from the tropical and tropical × U.S. germplasm populations contained either one or two genes with varying degrees of resistance. Eleven genes are known to control resistance to P. polysora and they are designated Rpp1 through Rpp11. Specific resistance is now frequently deployed using a combination of Rpp genes. Partial resistance to P. polysora also has been identified (Bailey et al. 1987; Zummo 1988). Scott and Zummo (1989) studied crosses between lines with major gene or partial resistance, and susceptible lines, and concluded crosses among slow-rusting (partial resistance) lines had

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

139

fewer pustules than either the slow-rusting × susceptible or susceptible crosses. 4. Quantitative Trait Loci and Marker-assisted Selection. Holland et al. (1998) studied inheritance of resistance to P. polysora in segregating, partially inbred, progeny lines derived from crosses between two temperate-adapted, tropical breeding lines with high levels of resistance. RFLP loci on three chromosomal regions known to be associated with resistance to either southern or temperate rust were used to localize genes affecting resistance and study their genetic effects. Multiplemarker locus models including loci from three individual regions on chromosomes 3, 4, and 10 affecting partial resistance, and their epistatic interactions, were able to account for nearly all of the variation in field scores. The authors concluded the results were consistent with a model wherein minor genes on chromosomes 3 and 4 interact epistatically with the major gene on chromosome 10S. Heritability of resistance estimates based on parent-offspring regression ranged between 0.30 and 0.50 in the two populations studied by Holland et al. (1998). Heritability estimates based on variances among partially-inbred progenies in the same populations were 0.25 to 0.35 units higher than those based on regression estimates. 5. Progress. Resistance has been the primary method of southern rust control in field maize. Deployment of major genes has been effective, but continued caution must be advised in the deployment of major-gene resistance so that the resistance will remain durable. Advances in our understanding of partial “slow rusting” resistance (Bailey et al. 1987; Zummo 1988) and the components of resistance (Zummo 1988) will likely prove invaluable in the continued management of the disease as new or unique variants of the pathogen arise (Smith 1999). The QTL studies of Holland et al. (1998) suggest that MAS may be feasible in future. F. Downy Mildew 1. Disease Description, Importance, Range, and History. Downy mildews are a diverse class of diseases that affect most grass crops. Ten species comprising three genera are pathogens of maize (Smith and Renfro 1999). Of these species, Peronosclerospora sorghi (Weston and Uppal) C. G. Shaw, causal agent of sorghum downy mildew (SDM), probably causes the most widespread damage because it can be found on every continent where maize is grown.

140

R. PRATT AND S. GORDON

Peronosclerospora sorghi spreads primarily through oospores, and conidia can also play an important role (Frederiksen 1980). Oospores are resting spores that can be dispersed by wind and infected debris. Conidia are produced on systemically infected plants and can contribute to significant disease spread within a field. P. sorghi produces few spores on maize, but readily sporulates on sorghum and Johnsongrass (Sorghum halapense) (Bigirwa et al. 1998). Infected plants usually are chlorotic and may display white stripes on the leaves. Leaves of infected plants are narrower and more upright than those of non-infected plants. Downy white growth may appear on infected leaves. In some environments, the tassels of infected plants may show a retrogradation of the floral organs to the condition of leaves (phyllody) (Smith and Renfro 1999). In Nigeria, the disease is widespread and increasing in incidence, due primarily to the continuous cultivation of maize year-round, even through the dry season (Kim et al. 1994). Bigirwa et al. (1998) found SDM in 11 of 22 Uganda districts surveyed (broader distribution than had been previously reported) and estimated that it reduced maize yields 4% in both 1994 and 1995. Though SDM has been reported as far north as Illinois and Kentucky in the United States, its economic impact occurs mainly in Texas (Frederiksen and Renfro 1977). Host-plant resistance is the most effective and economical means of controlling SDM. Among the alternative control measures, cultural and chemical controls prevail. Cultural practices such as eradication of alternative hosts (Johnsongrass and Sorghum arundinaceum), deep plowing to bury crop residue, and adjusting time of planting are the most effective. Chemical seed treatments with fungicides have been effective to a limited degree. Young plants can often outgrow the disease if they are not infected as seedlings (Frederiksen and Renfro 1977). All of these methods act to reduce the amount of primary inoculum. Planting maize in well-drained soils can prevent infection by oospores, although in Uganda, Bigirwa et al. (1998) reported higher incidences of SDM in sandy soils than in clay soils. 2. Sources of Resistance. Inbred T232 is reported to be resistant (Josephson et al. 1982), but it showed only an average response in a later evaluation (Darrah 1985). Other inbreds (Ga209, Mo17, NC248, T250, T254, Tx61M, Tx403, Tx601, Tx5855, and Tx6252) appeared to be free of infection in that evaluation. Several inbreds from Thailand derived from ‘Suwan-1’ (KU1403, KU1409, KU1414, KU1418) and Phil DMR-S6 and Oh514 were highly resistant to P. sorghi infection in trials conducted in Nigeria and Taiwan (Brewbaker et al. 1989). CIMMYT tropical inbreds

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

141

CML425, and CML427 through CML433, are resistant (CIMMYT 2000a). Adipala et al. (1999) also demonstrated resistance of ‘Suwan-1’ in Uganda. 3. Genetics of Resistance. Southeast Asia has the longest history of breeding for resistance to P. sorghi infection (Kim et al. 1994), and SDM has been effectively managed in many areas as a result of these activities (De Leon et al. 1993). Most of the progress has been made by using diallel studies with resistant and susceptible inbreds to identify lines with general combining ability for resistance and also combinations of inbreds that display significant interaction effects, or specific combining ability. In a diallel analysis using six maize inbred lines, Borges (1987) identified resistance as being mainly additive in nature, but nonadditive and maternal effects were also important. Maternal effects make the proper choice of seed parent for hybrid production an important consideration. Kim et al. (2003) reported similar genetic effects from their diallel study and identified a single cross that is being used as a commercial hybrid to manage SDM in West Africa. 4. Quantitative Trait Loci and Marker-assisted Selection. Only one QTL mapping study has been reported for resistance to P. sorghi. Agrama et al. (1999) utilized a population of 94 recombinant inbred lines (RIL) derived from a cross between resistant (G62) and susceptible (G58) maize inbred lines. Disease evaluations were carried out over two seasons in the field. Plants were inoculated with a conidial suspension, and after four weeks, reactions of the lines were recorded as the percentage of symptomatic plants per plot. The RIL population was genotyped at 106 RFLP loci yielding a map covering 1648 cM. Three QTL were detected, two on chromosome 1, and one on chromosome 9. All three were contributed by the resistant parent, and like most of the diallel results, the QTL displayed additive effects. Cumulatively, these QTL explained 54% of the phenotypic variation. These results are largely in agreement with the diallel studies that suggested quantitative inheritance of resistance. 5. Progress. In 1980, the maize breeding program at IITA began using inoculation protocols developed in Thailand and resistant germplasm from Thailand and The Philippines. Following 10 years of continuous selection, locally-adapted genotypes with resistance to P. sorghi had been developed (Cardwell et al. 1997a). Progress was impeded because disease escape rates of 20–30% were common with the conidial suspension inoculation protocols. This typically resulted in inconsistencies

142

R. PRATT AND S. GORDON

in ratings across trials. In 1992, a research project was designed to develop more reliable and cost-effective alternatives. Improvements in inoculation techniques have helped reduce escapes in SDM disease evaluations (Kim et al. 1994; Cardwell et al. 1997a). Kim et al. (1994) reduced the escape rate to 5–10%, which allowed for more accurate disease evaluations. Cardwell et al. (1997a) reported that directed inoculation of pre-germinated seed of spreader rows resulted in consistently high levels of P. sorghi-infected plants. The technique described also required substantially less labor and inoculum than the spray technique. One common finding regarding inoculations and disease assessments is that early and late disease assessment scores are highly correlated, so much so that effective disease assessments could be performed once, later in the epidemic (Ajala et al. 2003; Kim et al. 2003). Several researchers made significant progress in simultaneously improving P. sorghi resistance and agronomic traits of maize through recurrent selection (De Leon et al. 1993; Ajala et al. 2003). De Leon et al. (1993) utilized four maize populations and a combination of S1 and S2 recurrent selection to simultaneously improve grain yield and resistance to P. sorghi. After three selection cycles, they achieved an average reduction in SDM of 11% per cycle and grain yield increased over 500 kg/ha per cycle. Ajala et al. (2003) used S1 recurrent selection in six maize populations and selected for resistance to P. sorghi and increased grain yield. Reduction in SDM as much as 100% after four selection cycles and a grain yield increase of 10–98% were reported.

III. DISEASES INCITED BY VIRAL PATHOGENS A. Maize Dwarf Mosaic 1. Disease Description, Importance, Range, and History. Maize dwarf mosaic (MDM) is incited by MDMV, a member of the Potyviridae, the largest and most destructive family of plant viruses (Shukla et al. 1994). It was the first virus disease of maize to cause widespread damage throughout the United States and it is now common in many temperate and tropical regions of the world (Louie 1999; Paliwal 2000). MDMV was once thought to be distributed worldwide, but Shukla et al. (1994) have re-examined the taxonomic relationship between MDMV and sugarcane mosaic virus (SCMV). It was determined that some strains of MDMV (B, K, and O) were more closely related to SCMV and should be considered distinct from MDMV A, C, D, E, and F strains originating in Johnson-

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

143

grass, the over-wintering host of MDMV in the United States. MDMV and SCMV are now considered the most important viruses of maize in Europe (Alliot et al. 1995; Kuntze et al. 1997). MDMV naturally infects maize through aphid transmission. Most infected plants exhibit mosaic patterns; however, the developmental stage of the host at the time of infection may cause symptom expression to vary. In addition, specific genotypes may express unique symptom responses, e.g., inbred Ky61:2335 displays limited chlorotic streaks (Louie 1999). Stunting occurs in susceptible genotypes if they are infected early, but may not occur at all—especially in resistant hybrids infected during later stages of development (Louie 1999). The virus may also be transmitted mechanically and at low frequency through seed. MDM has been well managed in regions where private companies have incorporated resistance into susceptible cultivars and farmers have practiced sound weed management. Recent advances in the control of Johnsongrass have made a large impact on disease control. 2. Sources of Resistance. Resistance to MDMV has been identified in a variety of germplasm (Darrah 1985; Louie et al. 1990; Brewbaker et al. 1991; Scott and Louie 1996, Kuntze et al. 1997). Infection incidence near zero was reported for inbred lines Pa405, T232, and Tx601 in an interregional maize inbred evaluation (Darrah 1985). Twenty-three U.S. inbreds and one synthetic population were evaluated with MDMV-A and -B strains in field and greenhouse tests by Louie et al. (1990). That study revealed that apparent resistance to MDMV-A in the field was not always confirmed when inoculations were made on younger plants in the greenhouse. For example, MDMV-A (field-resistant) inbred Mo18W (Zuber 1973) was 100% infected in the 1990 greenhouse evaluation. This phenomenon was later studied experimentally by Scott and Louie (1996). In that study, resistant selections developed through controlled greenhouse inoculation of breeding lines during the winter season in Mississippi were evaluated in the field and later in the greenhouse using controlled inoculation procedures in each evaluation. The resistant lines developed from diverse breeding populations showed low levels of infection in the field, but many were rated susceptible in the subsequent greenhouse assay. The greenhouse assay also verified several breeding lines—one developed from PiraVEN-457, one from Early Caribbean MARs, and one selection from a Fusarium resistant synthetic population (MP:92:445)—were highly resistant. Other genotypes displaying resistance following controlled greenhouse inoculations with MDMV-A were B68, Oh1EP, Pa11, Pa405, and breeding population OhS2 (Louie et al. 1990). Kuntze et al. (1997) evaluated 122 European maize inbreds for

144

R. PRATT AND S. GORDON

their reaction to MDMV in field and greenhouse experiments following artificial inoculation. Three dent inbreds (D21, D32, and FAP1360A) displayed complete resistance against MDMV infection. Several other inbreds were partially resistant. Additional genotypes reported to display field resistance to MDMV are inbred lines Mp339 and Mp412 (Scott 1993), Ga203 (Roane et al. 1989a), and KyB110 (Scott 1989). Tennessee lines T61 and T232 were reported to display good field resistance to the virus complex (Josephson et al. 1982). The virus complex is now known to contain both MDMV and MCDV (Gordon et al. 1981). Findley et al. (1981) considered Oh7B and T232 to be parental inbreds that contributed high virus resistance to hybrids more consistently than others. Tropical lines that displayed high levels of resistance to both MDMV (strain A) were Hi31, Hi40, HIX4237, and KUCP3-X (Brewbaker et al. 1989). In a later study, inbreds Hi31 and Hi40 displayed high levels of resistance across studies conducted in Ohio and Hawaii (Brewbaker et al. 1991). Scott and Rosenkranz (1975) demonstrated resistant progenies could be selected in diverse breeding populations such as ‘Kenya Composite’ and PRMp1. Breeding populations with a higher proportion of temperate germplasm that are useful sources of MDMV resistance are available from Arkansas (York 1991), Kentucky (Poneleit 1990), and Ohio (Findley et al. 1981; Pratt et al. 1994b). 3. Genetics of Resistance. The interpretation of the phenotypic data has sometimes been confounded by variation in strains across locations and other factors beyond investigator control, such as field temperature and light intensity conditions (Roane and Tolin 1989). In early studies of inheritance, Scott and Rosenkranz (1982) concluded that inbreds Ga209, Mp339, Mp412, T240, and Va35 had one to three genes for resistance. Later genetic studies by those authors again suggested two to three genes in the majority of inbreds (AR254, Ga203, Mp71:22, and T232) with the presence of up to five resistance loci in Pa405 (Rosenkranz and Scott 1984). Research conducted by Mikel et al. (1984) led to the conclusion that three genes conferred resistance in inbreds B68 and Pa405. The ease of mechanical transmissibility has made inheritance studies of MDMV resistance less confounded by environmental factors and mixed infection. It has also been observed that parental crosses that differ in kernel color and MDMV resistance tend to produce progeny that show linkage between the yellow-color locus (yellow endosperm1 – y1; bin 6.01) near the centromere on chromosome 6 and MDMV resistance (Roane et al. 1989b; Scott 1989; Scott and Louie 1996). A major locus for resistance to MDMV des-

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

145

ignated mdm1 segregates as a dominant gene in Pa405 (McMullen and Louie 1989; Simcox et al. 1995). It maps to a position near the nucleolar organizer region (nor) locus on the short arm of chromosome 6 (bin 6.01). McMullen and Louie (1989) speculated that genetic modifiers of mdm1 activity might exist in different genotypes. The presence of genetic background effects may be associated with the different conclusions reached by different investigators. Scott and Louie (1996) also described resistance in a breeding line (Mp92:448) that does not appear to be attributable to the major factor near the nor on chromosome 6. 4. Quantitative Trait Loci and Marker-assisted Selection. Molecular mapping of chromosome 6 confirmed the presence of a dominant resistance factor (mdm1) identified by previous studies. Subsequent research examined segregating populations derived by crossing 42 MDMVresistant lines with three MDMV-susceptible inbreds (Redinbaugh et al. 2004). Results from that study indicated chromosome 6S markers were associated with resistance in 40 of the lines. Resistance was also associated with chromosomes 3 and 10, but only in 12, and seven inbreds, respectively. The authors suggested mdm1 is associated with MDMV resistance in most germplasm, but other loci might also contribute resistance in some genotypes. Breeding in the private sector has resulted in MDMV-resistant hybrids (Scott and Louie 1996) and it would hardly be surprising if conversion of susceptible inbreds to MDM resistance by transfer of mdm1 using MAS procedures were utilized. 5. Progress. Before 1960, virus diseases were of no economic importance in the United States (Findley et al. 1981). Breeding for resistance to MDMV enjoyed success in spite of initial problems in obtaining a high and uniform disease incidence, viral variation, and host environment (Findley et al. 1981) due very likely to the fact that resistance is simply inherited. Public-private cooperation in breeding efforts led to the rapid application of research findings to the development of elite commercial hybrids. Breeding for resistance to MDMV infection is now highly effective because it can be undertaken using controlled bioassays and infected plants can be classified with greater reliability. Selection of inbred lines (or partially inbred lines) is highly effective because a further “boost” in resistance is observed when vigor is expressed in hybrids created using the resistant lines (Scott et al. 1981). Researchers and breeders in Europe are now making rapid progress in identifying resistant lines and development of MAS for the improvement of MDMV resistance (Lubberstedt et al. 1999).

146

R. PRATT AND S. GORDON

B. Maize Chlorotic Dwarf 1. Disease Description, Importance, Range, and History. Maize chlorotic dwarf (MCD) is incited by MCDV, a member of the waikavirus group. It is a leafhopper-transmitted, isometric particle with single-stranded RNA (Louie 1999). Its range is not as wide as that of MDMV—it occurs only where its over-wintering host (Johnsongrass) and leafhopper vector [Graminella nigrifrons (Forbes)] overlap (Gordon et al. 1981). Thus, it is restricted to states primarily in the south, southeast, and the southern Corn Belt of the United States (Findley et al. 1984). The virus has also been detected in Texas and Arizona (Louie 1999). MCD epiphytotics are rare, but where control of Johnsongrass is inadequate, or susceptible hybrids follow sod, early infection can severely stunt plants. The symptoms of MCD vary depending on the time of infection, the strain of the virus, and the susceptibility of the host. The characteristic symptom that allows one to distinguish it from MDM is the “vein banding” or “vein clearing” of the tertiary veins of leaves. The leaves also assume a dull sheen and rough upper surface because the veins of infected leaves become swollen (Louie 1999). Leaves may yellow or redden, and may even become torn and twisted. Infected plants may also become stunted due to shortening of the upper internodes. 2. Sources of Resistance. High resistance or immunity to MCDV is rare and was not detected for decades (Findley et al. 1981; Pratt et al. 1994a; Redinbaugh et al. 2004). Accurate assessments of disease severity in the field and selection for host resistance to infection by MCDV have been hampered by the unpredictable occurrence, severity, and nonuniformity of epiphytotics (Findley et al. 1984; Scott 1983). A large part of the problem has been the confounding of diagnostic MCD symptoms caused by the presence of MDM symptoms (Louie and Knoke 1981). To illustrate the case in point, one can examine the data for inbred T232. It was thought to show tolerance in a cooperative field study conducted in three locations (Findley et al. 1981), but in a subsequent study using controlled inoculation in the greenhouse, it received a susceptible rating (Guthrie et al. 1982). Twenty-three U.S. inbreds and one synthetic population were evaluated for resistance to infection by MCDV using controlled inoculation tests in the greenhouse and all were susceptible (Louie et al. 1990). Only incidence data was recorded—disease severity based on symptom expression was not assessed. Pratt et al. (1994a) utilized an inoculation technique developed by Louie and Anderson (1993) and were able to demonstrate differential degrees of host tolerance to MCDV infection

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

147

based on symptom severity. Young plants of inbreds Mp705 and Mp707 were able to tolerate infection by a severe strain of MCDV, whereas other susceptible genotypes, such as inbred Va35, displayed severe symptoms and stunting. Consistent with observations in studies with MDMV inoculation (Louie et al. 1990), symptoms of plants grown in the field were not as severe as those on plants grown in the greenhouse. Other inbred lines reported to display resistance to MCDV infection are B77 (Russell and Hallauer 1974), Mp339 and Mp412 (Scott 1993), and T61 (Josephson et al. 1982). The identification of highly resistant, Caribbean germplasm, and the development of the MCDV-resistant inbred OH1V1, offer excellent opportunities for resistance breeding and research (Louie et al. 2002). Breeding populations from Arkansas (York 1991), Kentucky (Poneleit 1990), and Ohio (Findley et al. 1984; Pratt and Findley 1994) are potentially useful resources for MCDV-resistance breeding. 3. Genetics of Resistance. Due to the difficulties described above, no conclusive understanding regarding inheritance of resistance has been established. It has generally been agreed among experienced breeders that resistance to MCD infection involves several or more genes. Genetic studies are now being carried out using QTL analysis. 4. Quantitative Trait Loci and Marker-assisted Selection. Rufener et al. (1996) used RFLP analysis in an F2 population and with F2:3 progenies derived from Mp705 (tolerant) × Va35 (susceptible) hybrids. Phenotypic data were obtained using controlled inoculation procedures in a greenhouse. Five resistance QTL were identified on chromosomes 1, 3, 5, 7, and 10. More recently, two QTL were mapped in Oh1VI × Va35 F2 plants (Redinbaugh et al. 2004). The resistance factors co-segregated with markers on chromosome 3 near umc102 and chromosome 10 near umc44. These findings suggest the two QTLs on chromosomes 3 and 10 may be in common between the different resistance sources. 5. Progress. Progress in breeding for resistance to MCDV has not been as successful as MDMV-resistance breeding (Findley et al. 1984). Delays in identifying highly resistant genotypes have forestalled breeding efforts (Pratt et al. 1994a). Problems with incomplete virus transmission under field conditions and the presence of MDMV-infected plants limited selection for disease resistance and genetic studies. Improved techniques, such as rearing viruliferous leafhoppers and dropping them in the whorls of plants improved the success of breeding efforts (Findley et al. 1984). Refinements in methodology for virus transmission in

148

R. PRATT AND S. GORDON

controlled environments have virtually eliminated escapes and increased the consistency and accuracy of phenotypic classification (Louie and Anderson 1993; Pratt et al. 1994b). The improved inoculation and evaluation procedures, combined with QTL markers for consensus QTLs on chromosomes 3 and 10, should enable future MAS efforts. C. Maize Streak 1. Disease Description, Importance, Range, and History. Maize streak is incited by maize streak virus (MSV), a geminivirus that is transmitted by viruliferous leafhoppers of the Cicadulina genus found in Africa. MSV is a single-stranded DNA virus that displays geminate (twin) isometric particles when viewed with a scanning electron microscope. The vector-virus relationship, and distribution of the insect vector, probably account for restriction of the endemic virus almost exclusively to the African continent. Distinct strains of MSV are known and they have been shown to differ in their degree of sequence homology (Martin et al. 2001), serological variation (Pinner and Markham 1990; Konate and Traore 1994), and ability to induce different degrees of symptom severity (Mesfin et al. 1992; Konate and Traore 1994; Martin et al. 1999). Martin et al. (2001) grouped 85 MSV isolates into five main strain groupings designated A through E. A majority of isolates from maize were strain A. Different subtypes within the strain appeared to predominate in different parts of Africa and certain of these subtypes were determined to be more pathogenic in maize (MSV-A1, MSV-A2, and MSV-A5) than others (MSV-A3 and MSV-A4). These differences are of potential importance in screening and deployment of resistant genotypes in specific geographic regions. The characteristic elongated streaks of chlorotic tissue on the leaves arise from small chlorotic spots that appear on the basal portions of leaves that emerge following infection (Louie 1999). The virus moves systemically and occurs only in symptomatic regions. The time of inoculation and the degree of susceptibility of the host can markedly affect the degree of symptom expression (Bosque-Pérez et al. 1998). The effects of maize streak on grain yield are most pronounced when young plants are infected and decrease when infection occurs later. Maize streak is the most important and widespread disease of maize in Africa (Bosque-Pérez 2000). It is distributed throughout sub-Saharan Africa, from Sudan in the north to southern Africa and from East Africa to Senegal. It is present on adjacent islands in the Indian Ocean and it has been reported in Egypt and Yemen (Thottappilly et al. 1993). Maize

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

149

streak incidence is estimated at 60% across all African agroecosystems where maize is grown (DeVries and Toenniessen 2001) and epidemics may be devastating. Outbreaks of maize streak have been associated with the ecology and behavior of Cicadulina species in Zimbabwe (Rose 1978) and with drought and irregular, early rains in West Africa (Bosque-Pérez 2000). It is not clear which interacting factors foster the outbreak of epidemics, but it is likely the spread of maize production, and the increasing intensity of production (e.g., shorter fallow periods, less frequent rotations, and increased monoculture) have contributed to increased incidence of MSV epiphytotics (Bosque-Pérez 2000). 2. Sources of Resistance. MSV resistance was identified at IITA in 1975 in the maize population TZ-Y (Tropical Zea-Yellow), partly based on Tuxpeño Planta Baja (Bjarnason 1986) and unknown yellow germplasm from East Africa (Efron et al. 1989). MSV-resistant populations with white or yellow kernel color were developed using TZ-Y and they were designated TZSR-W and TZSR-Y, respectively. Resistance in the populations was fixed in inbred lines following continuous self-pollination and artificial MSV infection. One of the lines fixed from TZ-Y, inbred Ibadan 32 (IB32), was widely used as a donor of MSV resistance (Bjarnason 1986). In 1976, MSV resistance was also identified in the variety ‘La Revolution’ from Réunion Island (Department of Réunion) and in Tuxpeño × Ilonga composite from Tanzania (Bjarnason 1986). It is noteworthy that the vast majority of populations tested were considered susceptible, but a few individual plants, and in some instances populations with a low frequency of plants, showed tolerance (Soto et al. 1982). It is also important to note that selection for immunity (symptomless plants) was intentionally not undertaken in order to avoid escapes. Resistance from the sources presented above has been incorporated into many breeding populations through cooperative programs throughout sub-Saharan Africa. Large-scale screening using field and screenhouse infestation techniques made possible the development of the populations, varieties, and inbred lines with high levels of MSV resistance and superior agronomic performance. Sixteen inbred lines of diverse origin were publicly released from IITA (Kim et al. 1987) and the resistance in some of them (Tzi3, Tzi4, Tzi15, and TZi17) has been confirmed in subsequent evaluations conducted in Nigeria (Brewbaker et al. 1991). For a more complete list of germplasm developed at IITA during the 1970s and 1980s, the reader is referred to Efron et al. (1989). Breeding of MSV-resistant germplasm initially relied on the formation of base populations from crosses between resistant donors and locally

150

R. PRATT AND S. GORDON

adapted material. The donors arose from the cooperative effort in improvement of MSV-resistant germplasm between IITA and CIMMYT during 1980 to 1988 (Tang and Bjarnason 1993). Rearing of leafhoppers to enable controlled screening of germplasm for MSV resistance was initiated at the CIMMYT station near Harare, Zimbabwe in 1985 (K. Pixley, CIMMYT, pers. comm.). Pham (1992) reported breeding lines derived from MSV-resistant population Pop 49-SR (MSV-resistant conversion of Blanco Dentado-2) exhibited excellent yield and favorable streak ratings when combined with tropical and temperate testers and evaluated across several locations. Improvement of the base populations through S1 and S2 recurrent selection was successful for MSV resistance and increased emphasis on clear heterotic definition (K. Pixley, CIMMYT, pers. comm.). All CIMMYT mid-altitude inbreds from CML195 to CML215 now have average or above-average resistance to MSV. Recent CIMMYT mid-altitude inbred CML442 also shows resistance (CIMMYT 2000b). Tropical inbreds CML217 to CML238 also show average or above-average resistance (CIMMYT 2004). CIMMYT regional population ZM607 has been improved for MSV resistance and improved yield (CIMMYT 2004). The variety ‘La Révolution’ from Réunion Island, identified as resistant through cooperation between the Institute de Recherche Agronomiques Tropical (Tropical Agronomy Research Institute) and the Kenya Agricultural Research Institute (Efron et al. 1989), has been used as a resistance donor for improved populations such as CVR3-C3 (Rodier et al. 1995). The CVR-C3 population has been subsequently used as a resistance donor for an inbred (CIRAD 390) extracted from a CIMMYT population ‘Tocumen 7931’ (Clerget et al. 1996). Maize inbred D211, a resistant line from Réunion, has been used in QTL studies (Pernet 1999b). Another MSV-resistant inbred line, designated MSIRI 3B, has been released by the Mauritius Sugar Industry Research Institute (Govinden and Rummun 1996). The first screening of elite germplasm in South Africa by private company [Pannar (Pty) Ltd.] breeders revealed little good resistance, but some individual plants in populations and inbreds with intermediate resistance were selected (Barrow 1992). Selected plants were inter-mated in all combinations and pedigree selection was practiced under conditions of infection over several years. An assessment of MSV resistance in South African commercial maize hybrids during the 1984/85 to 1987/88 growing seasons showed varying degrees of resistance. Yellow hybrids generally displayed a wide range of variation and were more resistant than white hybrids (van Rensburg et al. 1991). Newer inbreds with higher levels of resistance have resulted from sustained breeding

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

151

efforts (Barrow 1992). Researchers in Zimbabwe and South Africa have also identified resistance in the population IRAT 297 from Réunion Island (Caulfield et al. 1994). 3. Genetics of Resistance. Reports have ascribed MSV resistance to a single gene lacking dominance (Storey and Howland 1967), simple inheritance with an apparently strong dominance component (Fourie and Piennar 1983), five dominant genes (Engelbrecht 1975), or quantitative inheritance (Gorter 1959). MSV resistance in inbred TZ-Y was considered to be simply inherited (Soto et al. 1982), but Kim et al. (1989) reported that resistance in inbred IB32 was quantitatively inherited through additive action of several genes. More recent studies of genetic control of MSV resistance in population CVR3-C3 have shown that multiple genetic systems for resistance to MSV may indeed exist (Rodier et al. 1995). Both unimodal and bi-modal frequency distributions of symptom ratings were observed when progeny developed by self-pollination within resistant, partially inbred lines were inoculated with MSV. This result suggested the possible existence of two different systems for genetic control of resistance; one with major genes controlling complete resistance, the other with minor genes controlling partial resistance. 4. Quantitative Trait Loci and Marker-assisted Selection. Kyetere et al. (1995) mapped a major resistance QTL to the short arm of chromosome 1 (bin 1.04) in a population of 87 RIL using 79 RFLP. The authors based their disease-severity assessments on percent leaf area affected at the height of the epidemic using a 1–5 scale. Their study employed both field (Zimbabwe) and greenhouse assessments (Uganda) and hence different viral isolates (Kyetere et al. 1999). The QTL on chromosome 1 acted like a single locus responsible for resistance and the F1 reacted similarly to the resistant parent, Tzi4, indicating dominant, or partially dominant, inheritance. The QTL locus on chromosome 1 was designated msv1. Several research groups, each using different sources of resistance, reached remarkably similar conclusions regarding the genomic location of a major QTL for resistance to MSV (Welz et al. 1998b; Kyetere et al. 1999; Pernet et al. 1999a,b; Lu et al. 1999). In all studies, the short arm of chromosome 1 was significantly associated with resistance despite variability of the phenotype conferred by the resistance source. All investigators have concluded that resistance sources examined to date contain msv1, or an allele of msv1. Minor QTL were also observed on chromosomes 2, 3, and 4 by Welz et al. (1998b), and chromosome 9 by Lu et al. (1999). Pernet et al. (1999a,b) noted minor QTL on chromosomes

152

R. PRATT AND S. GORDON

2 and 10 that were consistent across two populations studied. Additional study of putative minor QTL is needed. 5. Progress. Open-pollinated and hybrid maize cultivars that combine superior agronomic performance and MSV resistance have been developed by both public and private-sector programs. The breeding efforts have required interdisciplinary cooperation among entomologists, pathologists, agronomists, and breeders. Both improved and traditional varieties were converted to MSV resistance by IITA in cooperation with CIMMYT and national programs in Africa (Efron et al. 1989). Tang and Bjarnason (1993) evaluated progress in MSV-resistance breeding and concluded both modified full-sib recurrent selection and backcrossing were highly effective in improving MSV resistance without sacrificing yield. Maize inbred lines with high levels of resistance have been developed by several breeding programs including CIMMYT in Zimbabwe, IITA, IRAT in Réunion (Rodier et al. 1995), the Mauritius Sugar Industry Research Institute in Mauritius (Goviden and Rummun 1996), and Pannar (Pty) Ltd. and Seed Coop seed companies in South Africa and Zimbabwe, respectively. Studies have shown resistant IITA hybrids TZB-SR and 8321-21 (late maturing, white-grain types adapted to lowland and savanna regions, respectively) suffer little reduction in yield when infected by MSV. Those hybrids also showed reduced incidence of infection under equalchallenge opportunities (tolremicity), an indication that both tolerance and tolremicity contribute to overall resistance (Bosque-Pérez et al. 1998). In the future it may be possible to exploit tolremicity further. Hybrid combinations of improved parents (e.g., CML202 × CML208) have demonstrated high grain yield and very good levels of resistance to MSV in regional mid-altitude trials (K. Pixley, CIMMYT, pers. comm.). Barrow (1993) reported Pannar (Pty) Ltd. hybrids PAN6099, grown previously on a limited scale in South Africa and Swaziland, outyielded other hybrids across diverse growing environments in southern, western, and eastern Africa when MSV had been prevalent. A more recently released hybrid, PAN6195, displayed high levels of resistance where MSV was prevalent, and also yielded competitively in MSV-free trials in Nigeria. Because strains of MSV are known to occur in different localities, confirmation that resistance is effective across diverse locations suggests that germplasm developed in a national or regional program will likely be useful in broader areas as well. Martin et al. (1999) assessed the MSV resistance of 19 previously characterized genotypes using agroinoculation (transfer of infectious MSV isolates to host cells using Agrobac-

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

153

terium tumefacians as a vector) and demonstrated how differentially virulent isolates of MSV could be used to identify hosts with differing levels of susceptibility. Resistance rankings using the agroinoculation method with known isolates (individually) corresponded well with the rankings provided by breeders. CIMMYT genotypes Z491, Z470, and Z471, considered immune in field tests, were highly resistant, but they showed mild symptoms in response to controlled inoculation. Lines P612 and CIMMYT 459 (also considered immune) displayed susceptible responses when infected by one severe strain (from Zimbabwe) and one moderately severe strain (from South Africa). The authors proposed that MSV isolates should be specifically chosen for particular resistance breeding programs. It is true that using mixtures of isolates or races as inoculum can lead to imprecise conclusions regarding the genetic basis of resistance. Breeders have typically used mixed isolates of MSV to facilitate screening and this approach has thus far been successful for improvement of resistance in breeding lines (K. Pixley, CIMMYT, pers. comm.). D. Mal de Río Cuarto 1. Disease Description, Importance, Range, and History. Mal de Río Cuarto Virus (MRCV) is a member of the Fijivirus group (Distéfano et al. 2003). It is considered to be distinct from maize rough dwarf fijivirus that occurs in the Mediterranean region. The virus is organized as icosahedric particles composed of 10 double-stranded RNA segments. Symptoms of MRCV infection include shortening of the internodes, which causes stunting, and the formation of galls or enations on the leaves. The diagnostic symptom of MRCV infection is the overgrowth of tissue on the veins located on the underside of the leaves (enation). Veins may become swollen, creating an apparent roughness of the leaf. The leaves of susceptible cultivars may become deformed. Infected leaves sometimes undergo cross-sectional splits, trimming leaves to the point that the laminae may even disappear. The inflorescences of infected plants may undergo malformations. They typically become smaller and in greater number, and may even show total atrophy of both male and female flowers. The severity of symptoms is associated with plant genotype, age at which infection occurs, and weather conditions. MRCV causes one of the principal diseases of maize in Argentina and also infects a wide range of cultivated and wild species of the Graminae. Its spread in the Argentine maize-growing area coincides largely with the geographical distribution of its insect vector Delphacodes kuscheli (Fennah). The worst epidemic in Argentina occurred during 1996–97

154

R. PRATT AND S. GORDON

(Lenardon et al. 1998). The disease is also reported to be present in Brazil and Uruguay. Numerous species of grassy weeds and other cultivated cereals constitute inoculum reservoirs and hosts for the insect vector populations (Rodriguez Pardina et al. 1998). The vector populations typically increase in wheat and oats (Avena sativa L.), then migrate to maize seedlings and transmit the virus during feeding. 2. Sources of Resistance. Most commercial hybrids surveyed were apparently susceptible to the virus, although some were tolerant (Lenardon et al. 1998). A partially resistant flint line, BLS14, has been identified in Argentina (Di Renzo et al. 2002). 3. Genetics of Resistance. Moderate (broad-sense) heritability estimates, from 0.46 to 0.56, have been determined in populations arising from a cross between maize inbred line BLS14 and the susceptible line Mo17 (Di Renzo et al. 2002). Breeding for resistance has been hampered by a lack of highly-resistant germplasm and what appears to be recessive inheritance associated with resistance (Presello 1993 and Presello et al. 1995, cited in Di Renzo et al. 2002). Results of the first published inheritance studies suggest that resistance in BLS14 is governed by many genes (Di Renzo et al. 2002). 4. Quantitative Trait Loci and Marker-assisted Selection. No information is available. 5. Progress. Breeding for resistance has been laborious and time consuming (Di Renzo et al. 2002). Substantial genotype × environment interaction has been observed when disease evaluations are conducted in sites relying on natural infection, but some success has been achieved using partial-resistance sources. The authors also suggested that the timing of the planting of maize plots to coincide with the harvest of the winter cereal crop should provide an effective environment for selection. Attempts to utilize artificial inoculation have given infection rates of around 70%. The inoculation efficiency diminishes to less than half when the plants have five leaves and to practically zero when plants have seven leaves. Further refinement of controlled inoculation procedures would appear to be quite valuable. E. Maize Rayado Fino 1. Disease Description, Importance, Range, and History. Maize rayado fino virus (MRFV) is the type member of the genus Marafivirus

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

155

(Hammond and Ramirez 2001). Small isometric MRFV virions contain a single-stranded RNA genome that has recently been sequenced (León and Gámez 1981; Hammond and Ramirez 2001). It is a leafhopper-borne pathogen that is persistently transmitted by members of the genus Dalbulis. Widely distributed over diverse ecological areas, MRFV is found only in the American continents. MRFV was first recognized in El Salvador and Costa Rica over 20 years ago (Kogel et al. 1996). It has been detected in all Central American countries and isolates from Argentina, Bolivia, Brazil, Peru, and Uruguay have been characterized. It is also present in Texas and Florida in the United States (Louie 1999). MRFV frequently occurs in mixed infestations with corn stunt spiroplasma and maize bushy stunt phytoplasma (Gámez and León 1985). Typical symptoms of MRFV usually include small conspicuous spots that develop at the base and along the veins of young leaves in characteristic stippled stripes (Kogel et al. 1996). Additional symptoms of infection are considered variable depending on the susceptibility of the host. Susceptible genotypes frequently display wilting of young plants, general chlorosis, and stunting (Bustamante et al. 1998). In both temperate and subtropical areas of Argentina, MRFV was confirmed in plants that displayed fine chlorotic stipple-striping of the veins, chlorosis, numerous dots and stripes, and holes in the leaf blade (GiménezPecci et al. 2000). It is considered to be associated with corn stunt, but its contribution to stunting and leaf-reddening symptoms in “red stunt” described in Brazil is unknown (Hammond and Bedendo 2001). 2. Sources of Resistance. The degree of susceptibility among genotypes is variable, with Central American landraces considered to be more tolerant of MRFV infection and also to show lower incidence of infection (Gámez and León 1985). Bustamante et al. (1998) examined 20 maize genotypes obtained from CIMMYT, the University of Costa Rica, and the Consejo Nacional de Produccio in Costa Rica. The genotypes were considered representative of various races of maize. Using enzyme-linked immunosorbent assay (ELISA), and assessment of symptom severity, the investigators evaluated the responses of maize genotypes to MRFV infection under screenhouse conditions. Two accessions from CIMMYT (2980-93, a CIMMYT hybrid; and 3974 from St. Croix) displayed mild and delayed symptoms and low MRFV concentrations as determined by ELISA. 3. Genetics of Resistance. Inheritance of resistance is not known. 4. Quantitative Trait Loci and Marker-assisted Selection. No information is available.

156

R. PRATT AND S. GORDON

5. Progress. Progress has been made in characterization of the virus and understanding the epidemiology of infection. Some resistant germplasm has now been identified and further germplasm screening and genetic studies are being conducted.

IV. DISEASES INCITED BY BACTERIAL PATHOGENS A. Stewart’s Bacterial Wilt 1. Disease Description, Importance, Range, and History. The causal agent is a gram-negative bacterium, Erwinia stewartii (Smith) Dye (Claflin 1999) that over-winters in the gut of its primary vector, the corn flea beetle Chaetocnema pulicaria (Melsheimer). In the spring, the flea beetles feed on maize and other grasses, spreading the bacterium. Because several generations of the flea beetle are produced during the growing season, disease incidence may continue to increase during the summer months. The bacterium enters the host through feeding wounds and wilting subsequently occurs due to plugging of the xylem tissues. Stewart’s bacterial wilt (SBW) caused catastrophic losses in the United States during the 1930s. A concerted effort in research and breeding resulted in rapid improvements of host-plant-resistance. Since that time, the disease has been a prominent pathogen of sweet maize and continues to act as a sporadic economic pathogen of field maize, especially in the southern U.S. Corn Belt. It has been observed in the Americas as far north as Ontario, and it has been reported in Europe and Asia (Claflin 1999). In temperate regions, following warm winters that do not appreciably impact the survival of its insect vector, its range may be extended. It caused some losses during the 1980s and, because it is seed borne, it continues to be of concern for international seed exchange (Ming et al. 1999). Many countries require additional declarations (inspections or phytosanitary permits) to ensure that imported seed is free of contamination (Claflin 1999). Seed can be tested and certified infection-free by the Iowa State University Seed Testing Laboratory (http://www.seed lab.iastate.edu/seedtest/). Two phases of SBW occur, one in the juvenile phase and one that is manifested as leaf blight (Claflin 1999). The infection of young plants is not common in field maize, but systemic infection of susceptible plants results in rapid wilting. Leaves display linear, pale green-to-yellow streaks that parallel the veins and the lesions then desiccate and turn a brownish color. Infection that manifests itself as leaf blight later in the life cycle is more common. Foliar lesions are gray green to yellow and

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

157

develop a characteristic streaking. The streaks become a pale brown or straw color and severely infected leaves may die. 2. Sources of Resistance. Resistance was identified and incorporated into elite germplasm in the United States following devastating epidemics that occurred during the 1930s. Pataky et al. (2000) examined 2,000 accessions representing various geographic regions of the United States and the world. The authors concluded moderate levels of resistance could be found in accessions collected from nearly everywhere. High levels of resistance were somewhat more abundant in accessions collected from regions where the disease was endemic. Blanco et al. (1976) evaluated 240 genotypes, and 21% (including inbreds Mo17 and B37) were highly resistant. Warren (1981, 1982) reported several inbred releases that were resistant (H102, H103, and H111). Screening of more than 2,000 accessions in the U.S. National Germplasm Resources Program has been conducted. In one evaluation of 1,986 accessions, seven genotypes showed little or no spread following infection. They were B85, Pa891, Va17, Va21C, Va37, Va59, and SD40 (Pataky et al. 2000; USDA-ARS 2004). Other evaluations did not reveal genotypes with the highest resistance ratings; however, numerous genotypes displayed moderate resistance across tests. Those genotypes were H84, Mo17, Pa91, T232, V102, and W182BN. A sizeable number of popcorn accessions also were among the most resistant accessions identified in each evaluation. Kang (1990) showed that resistance in Louisiana inbreds L329 and L765 was excellent and that line L329 resistance also seemed to restrict lesions in susceptible backgrounds more than L765. Parker and Hooker (1993) concluded that Pa887P was the most resistant line among the four different sources of resistance used in their study. Ming et al. (1999) demonstrated resistance of inbred Ki14 from Thailand. A recent examination of maize lines from South Africa led investigators to conclude that levels of resistance were sufficient to minimize the effects of Stewart’s wilt on yield should the pathogen enter that region (Michener and Pataky 2002). 3. Genetics of Resistance. Early studies concluded that resistance was controlled by two to four genes. Studies conducted from the 1930s through the early 1970s have been reviewed by Parker and Hooker (1993) and Ming et al. (1999). Blanco et al. (1979) used Castle’s formula, as modified by Wright (1968), to estimate the number of genes using multiple F2 populations derived from crosses between susceptible (B14A) and intermediate (N28 and Va35) inbreds with three resistant inbreds

158

R. PRATT AND S. GORDON

(33-16, B37, and Mo17). They also examined the genetic effects of the resistant genotypes using diallel analysis. They were unable to discern discrete Mendelian classes. Their analysis of variance indicated that genotype × rating date disease-response interactions were significant. Environmental variance was higher than genetic variance at the first rating, and gene-number estimates were low. Environmental variance tended to diminish proportionally at the later ratings and genetic variance increased. The number of gene estimates rose to values as high as nine at the later ratings. The authors concluded that disease resistance was a highly heritable trait governed predominantly by additive effects conferred by several major genes, perhaps with some involvement by modifier genes. Kang (1990) studied inheritance of resistance in F2 and backcross generations derived from crosses between susceptible inbred L163 and two resistant inbreds, L329 and L765. The F2 segregation ratio of 15 resistant to 1 susceptible indicated that resistance was controlled by two dominant, duplicate genes. Segregation in the backcross populations also showed the expected ratio of 3 resistant to 1 susceptible. Parker and Hooker (1993) studied resistance in four different sources of resistance (Pa70, Pa83, Pa419P, and Pa887P) using generation mean analysis. They also estimated gene numbers governing resistance using Wright’s (1968) formula and Mendelian analysis. They concluded that approximately one to four genes conferred resistance and felt that the higher estimates obtained by Blanco et al. (1979) might have been affected by the rating scale employed in their study. They observed most of the variation among generation means was attributable to additive effects. An additive × dominance model accounted for 90% of the total variation observed in the populations. 4. Quantitative Trait Loci and Marker-assisted Selection. An RFLP marker study with tropical germplasm was conducted with RIL derived from resistant inbred Ki14 and susceptible inbred Hi31 (Ming et al. 1999). A highly significant QTL, designated sw1, was mapped on the short arm of chromosome 1 between the markers umc167 and umc67. Another significant region was identified on chromosome 9, but the authors felt that additional studies would be needed to confirm that minor QTL. The authors also concluded that the markers would be very useful for identifying escapes in the nursery and MAS would be useful for this trait. 5. Progress. Dramatic losses like those experienced in the United States during the 1930s have not been in evidence. Improved levels of resis-

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

159

tance have been incorporated into many elite U.S. Corn Belt inbreds. All hybrids are not resistant, and it is important for producers to select resistant cultivars in areas where winters may be mild. Blanco et al. (1979) observed that genotypes differ in both the extent and the rate at which resistant lines become diseased. Additional research on the components of resistance might prove informative. Research to improve our knowledge regarding the role of insect and environmental interaction with the host would also be useful.

V. SUMMARY In nature, hosts and pathogens co-exist in a complex equilibrium, and epidemics are rare. Modern agricultural practices provide the benefit of larger harvests, but they also alter the host-pathogen equilibrium in ways that can result in epidemics. Production of increasingly homogeneous cultivars results in a loss of genetic diversity. Expanded production of maize to the point where it dominates the landscape also results in the loss of overall crop diversity. Continuous maize crops further reduce the “spatial complexity” of the environment in space and time (Simmonds and Smartt 1999). Such fundamental changes in the agroecosystem can shift the interrelationship between host and pathogen in favor of the pathogen, and the risk of disease “outbreaks” will be heightened. If new pathogens arrive, or if weather patterns conducive to disease become prevalent, then crops with inadequate host-plant-resistance may be severely impacted. We have already witnessed a collapse in maize production during 1970 due to the SCLB epidemic in the southern United States. Such devastating epidemics simply cannot be sustained by farmers who already receive low returns or are without a safety net. Fortunately, severe epiphytotics have been the exception rather than the rule in the United States (Duvick and Cassman 1999), where the world’s largest aggregation of related maize germplasm is to be found. Each successive epidemic, such as the GLS epidemic of the 1994 and 1995 seasons that caused widespread losses, was effectively managed by replacement of highly susceptible cultivars, increased resistance breeding, and to some extent, changes in cultural practices. With the exception of suppression of viral diseases (MDM and MCD) through Johnsongrass control, cultural practices have not had a major impact on disease control. Historically, breeding for major-gene resistance has been the method of first choice because it is expedient and effective. Major-gene resistance

160

R. PRATT AND S. GORDON

has been utilized for management of common rust, NCLB, and MDM. In each of the first two instances, we have also witnessed the subsequent defeat of those resistance genes and the development of virulent strains. So far, we have largely managed to escape the “boom-bust” cycles whereby major genes are deployed and defeated by the pathogen in a repetitive cycle (Suneson 1960). Foresight on the part of researchers and breeders to maintain breeding for partial resistance (White and Carson 1999), combined with good partial resistance characteristics readily identifiable in germplasm sources such as many C103-related inbreds (Troyer 1999; MBS 2002) and southern and southeastern U.S. germplasm, have provided a level of protection. Emphasis on selection for “stay green” (late-season plant health) (Troyer 1999) may also have contributed to maintaining a “background” level of partial resistance in many commercial hybrids. The constantly changing interrelationships between maize and its pathogens will require continued attention and monitoring (Smith 1999). Hopefully, new research findings will continue to be integrated with breeding programs in a timely and productive manner. Successful identification of resistance donors is not a simple “numbers game,” the success of which is determined largely by chance. Improved disease management using host-plant-resistance can be attributed in large part to interdisciplinary cooperation that improves our understanding of disease epidemiology and the development and application of effective selection techniques. Refinements in bioassays, particularly for insect-vectored viral diseases such as MCD and MSV, have been paramount in identifying and exploiting high levels of tolerance. The progress in selection of MSV tolerance is a well-documented example of the strengths and rewards of interdisciplinary cooperation (Bosque-Pérez 2000). The wisdom of preserving the continuity of interdisciplinary crop improvement programs that can achieve such dramatic results should be obvious, but apparently it is not always so. Declining prioritization of practical breeding programs makes extension of the products of molecular-genetic research to improved crop cultivars increasingly unlikely in the public sector (Knight 2003). Some diseases can be induced under controlled conditions quite readily, whereas others cannot. Hence, plant phenotypes may be variable and the need for replication across locations or years arises. The need may also exist for exposure to multiple races or strains of a pathogen to assure effective deployment of resistance genes. Use of laborious or expensive inoculation protocols and the need for replication may be essential for productive selection. The breeder will often have no recourse but to reduce the total number of progenies that can be exam-

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

161

ined in order to meet the logistical demands necessary to assure accurate disease assessment. There appears to be a prevailing sentiment among breeders that it would be imprudent to rely on the quick, but too often transient, rewards of major-gene resistance. An understanding and appreciation that partialresistance breeding will play an important role in safeguarding the longevity of resistance appears to be growing (Parlevliet 2002). The challenge presented by the high genotype × environment interaction associated with multiple genes associated with partial resistance typically requires a greater logistical commitment for both breeding and research activities (Geiger and Heun 1989) and it will not be easy to sustain these efforts in an environment of declining resources. QTL mapping has also given us new approaches to dealing with both qualitative- and quantitative-resistance factors. Major-genes obviously emerge as major QTL, e.g., ht1, Rpp9, rhm, mdm1, msv1, and sw1, but so too do other major QTL that consistently account for a sizeable proportion of total variance, in concert with minor QTL. Thus, the “gray zone” of genes that may not clearly fit a qualitative or quantitative definition emerges as a very likely source of candidate QTL for transfer to susceptible genotypes because they are perhaps less “dependent” on the environment than are minor-genes. MAS should enhance selection of these “robust” quantitative-genes and facilitate their incorporation into elite cultivars (Lindhout 2002). The nature of minor genes and QTL is still elusive and will require considerable effort to understand. Improved methods are being developed for gene mining (Dong et al. 2003) using genomic and proteomic approaches. Genomic approaches may be of considerable value in helping us reveal the role of minor genes, but again, proof of impact on the phenotype will depend on accurate and reproducible disease induction and quantification. Additional information regarding climatic variables and genetic background will be needed in many instances if environmental and epistatic interactions are to be resolved. The role of individual gene products in host-plant-resistance responses will benefit from production of near isogenic lines and the further refinement of disease bioassays. Increases in our understanding of pathogen variation will also enable determination of specific interactions between host and pathogen and enable optimization of MAS strategies. In many regions of the world, threats from multiple diseases are ongoing (Bigirwa et al. 2001; DeVries and Toenniessen 2001) and must be addressed. Evaluation of many breeding lines at many locations to estimate genotype × environment effects and test stability, and challenging the lines with predominant races or highly aggressive isolates of the

162

R. PRATT AND S. GORDON

pathogen, can be a daunting task (Johnson 1984; Troyer 2001; Carson et al. 2002). Faced with additional logistical demands, breeders frequently inoculate with several or more pathogens in a given disease nursery (Troyer 2001). However, when the objective is to identify the most promising lines with partial resistance, or markers linked to resistance loci, this may be a reasonable approach. The use of markers to select lines with increased potential durability, especially when breeding, may encompass resistance to multiple diseases (Pratt et al. 2003). When genes are more precisely identified and characterized with the help of emerging technologies, introgression of resistance factors into susceptible genotypes should be enhanced via “marker-guided introgression” (Koorneef and Stam 2001).

LITERATURE CITED Adipala, E., G. Bigirwa, J. P. Esele, and K. F. Cardwell. 1999. Development of sorghum downy mildew on sequential plantings of maize in Uganda. Int. J. Pest Manag. 45:147–153. Agrama, H. A., M. E. Moussa, M. E. Naser, M. A. Tarek, and A. H. Ibrahim. 1999. Mapping of QTL for downy mildew resistance. Theor. Appl. Genet. 99:519–523. Ajala, S. O., J. G. Kling, S. K. Kim, and A. O. Obajimi. 2003. Improvement of maize populations for resistance to downy mildew. Plant Breed. 122:328–333. Alliot, B., E. Gellez, and P. A. Signoret. 1995. Correlation between greenhouse and field scores in testing resistance of inbred maize lines to maize dwarf mosaic virus-A (MDMVA). Agronomie 15:459–462. Arthur, J. C. 1929. p. 338–339. In: The plant rusts. Wiley, New York. Asea, G., P. E. Lipps, S. Gordon, E. Adipala and R. C. Pratt. 2003. Development of inoculation procedures for evaluation of partial resistance to Cercospora zeae-maydis in young maize plants. Sixth Biennial Conference of the African Crop Sci. Soc., Nairobi, Kenya. 12–17 Oct. 2003. Programme, Abstracts, and List of Participants. p. 112. Bailey, B. A., W. Schuh, R. A. Frederiksen, A. J. Bockholt, and J. D. Smith. 1987. Identification of slow-rusting resistance to Puccinia polysora in maize inbreds and single crosses. Plant Dis. 71:518–521. Bair, W., and J. E. Ayers. 1986. Variability in isolates of Cercospora zeae-maydis. Phytopathology 76:129–132. Barrow, M. R. 1992. Development of maize hybrids resistant to maize streak virus. Crop Protec. 11:267–271. Barrow, M. R. 1993. Increasing maize yields in Africa through the use of maize streak virus resistant hybrids. African. Crop. Sci. J. 1:139–144. Bigirwa, G., E. Adipala, and J. P. Esele. 1998. Occurrence of Peronosclerospora sorghi in Uganda. Plant. Dis. 82:757–760. Bigirwa, G., R. C. Pratt, E. Adipala, and P. E. Lipps. 2001. Assessment of gray leaf spot and stem borer incidence and severity on maize in Uganda. Proc. African Crop Sci. Conf. 4:469–474. Fourth African Crop Sci. Conf., Casablanca, Morocco. 11–14 Oct. 1999. Editorial Secretariat, African Crop Sci. Soc., Kampala, Uganda.

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

163

Bjarnason, M. 1986. Progress in breeding for resistance to the maize streak virus disease. p. 197–207. In: To feed ourselves: Proc. First Eastern, Central Southern Africa Regional Maize Workshop, Lusaka, Zambia. 10–17 March 1985. Sponsored by Govt. of Zambia and CIMMYT. CIMMYT, Mexico. Blanco, M. H., J. R. Wallin, D. V. Loonan, and M. S. Zuber. 1976. Comparison of seedling and mature plant resistance to Stewart’s bacterial wilt in maize. p. 46. In: 1976 Agronomy abstracts. ASA. Madison, WI. Blanco, M. H., M. S. Zuber, J. R. Wallin, D. V. Loonan, and G. F. Krause. 1979. Host resistance to Stewart’s disease in maize. Phytopathology 69:849–853. Borges, O. L. 1987. Diallel analysis of maize resistance to sorghum downy mildew. Crop Sci. 27:178–180. Bosque-Pérez, N. A. 2000. Eight decades of maize streak virus research. Virus Res. 71:107–121. Bosque-Pérez, N. A., S. O. Olojede, and I. W. Buddenhagen. 1998. Effect of maize streak virus disease on the growth and yield of maize as influenced by varietal resistance levels and plant stage at time of challenge. Euphytica 101:307–317. Brewbaker, J. L., S. K. Kim, and M. L. Logrono. 1991. Resistance of tropical maize inbreds to major virus and virus-like diseases. Maydica 36:257–265. Brewbaker, J. L., M. L. Logrono, and S. K. Kim. 1989. The MIR (maize inbred resistance) trials: performance of tropical adapted maize inbreds. Hawaii Agr. Expt. Stat., Hawaii Institute of Agr. and Human Resources Research Series 062. Brewster, V. A., M. L. Carson, and Z. W. Wicks III. 1992. Mapping components of partial resistance to northern leaf blight of maize using reciprocal translocations. Phytopathology 82:225–229. Bubeck, D. M., M. M. Goodman, W. D. Beavis, and D. Grant. 1993. Quantitative trait loci controlling resistance to gray leaf spot in maize. Crop Sci. 33:838–847. Burnette, D. C., and D. G. White. 1985. Inheritance of resistance to Bipolaris maydis race O in crosses derived from nine resistant inbred lines of maize. Phytopathology 75:1195–1200. Bustamante, P. I., R. Hammond, and P. Ramirez. 1998. Evaluation of maize germ plasm for resistance to maize rayado fino virus. Plant Dis. 82:50–56. Cai, H. W., Z. S. Gao, N. Yuyama, and N. Ogawa. 2003. Identification of AFLP markers closely linked to the rhm gene for resistance to southern corn leaf blight in maize by using bulked segregant analysis. Mol. Genet. Genom. 269:299–303. Cardwell, K. F., J. G. Kling, and C. Bock. 1997a. Methods for screening maize against downy mildew Peronosclerospora sorghi. Plant Breed. 116:221–226. Cardwell, K. F., F. Schulthess, R. Nidemah, and Z. Ngoko. 1997b. A systems approach to assess crop health and maize yield losses due to pests and diseases in Cameroon. Agric. Ecosys. Environ. 65:33–47. Carson, M. L. 1999a. “Helminthosporium” Leaf Spots and Blights. p. 15–20. In: D. G. White (ed.), Compendium of corn diseases, 3rd ed. APS Press, St. Paul, MN. Carson, M. L. 1999b. Gray Leaf Spot. p. 22–25. In: D. G. White (ed.), Compendium of corn diseases, 3rd ed. APS Press, St. Paul, MN. Carson, M. L., M. M. Goodman, and S. M. Williamson. 2002. Variation in aggressiveness among isolates of Cercospora from maize as a potential cause of genotype-environment interaction in gray leaf spot trials. Plant Dis. 86:1089–1093. Caulfield, M. J., J. G. du Plessis, J. B. J. van Rensburg, N. W. McLaren, and B. C. Flett. 1994. A field selection method to screen for resistance to maize streak virus. p. 46–49. In: Proc. Tenth South Afr. Maize Breed. Symp., Potchefstroom, South Africa. March 17–19, 1992. Tech. Comm. 238, Dept. of Agric., Pretoria, South Africa.

164

R. PRATT AND S. GORDON

CIMMYT. 2000a. Seventh announcement of 58 maize inbred lines by CIMMYT. http:// www.cimmyt.org/research/maize/announcement/announcement.doc. CIMMYT. 2000b. Eighth announcement of 30 maize inbred lines by CIMMYT. http:// www.cimmyt.org/research/maize/announcement/announcement.doc. CIMMYT. 2004. Maize germplasm listing. http://www.cimmyt.org/research/maize/ germplasmlist/htm/GermplasmList.html. Claflin, L. E. 1999. Stewart’s bacterial wilt. p. 3–4. In: D. G. White (ed.), Compendium of corn diseases, 3rd ed. APS Press, St. Paul, MN. Clements, M. J., J. W. Dudley, and D. G. White. 2000. Quantitative trait loci associated with resistance to gray leaf spot of corn. Phytopathology 90:1018–1025. Clerget, B., J. Dintinger, and B. Reynaud. 1996. Registration of CIRAD 390 maize inbred parental line. Crop Sci. 36:826. Coates, S. T., and D. G. White. 1994. Sources of resistance to gray leaf spot of corn. Plant Dis. 78:1153–1155. Coates, S. T., and D. G. White. 1998. Inheritance of resistance to gray leaf spot in crosses involving selected resistant inbred lines of corn. Phytopathology 88:972–982. Coors, J. G., and M. C. Mardones. 1989. Registration of W552, W570, W572, and W576 parental lines of maize. Crop Sci. 29:247–248. Cross, H. Z. 1984. Registration of ND101 and ND256 parental lines of maize. Crop Sci. 24:1223. Darrah, L. L. 1985. Report of the inter-regional maize inbred evaluation. Special Rep. 325, Univ. of Missouri-Columbia Agr. Expt. Sta. De Leon, C., G. Granados, R. N. Wedderburn, and S. Pandey. 1993. Simultaneous improvement of downy mildew resistance and agronomic traits in tropical maize. Crop Sci. 33:100–102. de Nazareno, N. R. X., P. E. Lipps, and L. V. Madden. 1993. Effect of levels of corn residue on the epidemiology of gray leaf spot. Plant Dis. 77:67–70. DeVries, J., and G. Toenniessen. 2001. Securing the harvest, biotechnology, breeding and seed systems for African crops. CABI, New York. Dingerdissen, A. L., H. H. Geiger, M. Lee, A. Schechert, and H. G. Welz. 1996. Interval mapping of genes for quantitative resistance of maize to Setosphaeria turcica, cause of northern leaf blight, in a tropical environment. Molecular Breed. 2:143–156. Di Renzo, M. A., N. C. Bonamico, D. D. Díaz, J. C. Salerno, M. M. Ibañez, and J. J. Gesumaria. 2002. Inheritance of resistance to Mal de Rio Cuarto (MRC) disease in Zea mays (L.). J. Agr. Sci. 139:47–53. Distéfano, A. J., L. R. Conci, M. Muñoz Hidalgo, F. A. Guzmán, H. E. Hopp, and M. del Vas. 2003. Sequence and phylogenetic analysis of genome segments S1, S2, S3 and S6 of Mal de Rio Cuarto virus, a newly accepted Fijivirus species. Virus Res. 92:113–121. Dong, Q., L. Roy, M. Freeling, V. Walbot, and V. Brendel. 2003. ZmDB, an integrated database for maize genome research. Nucleic Acids Res. 31:244–247. Dunkle, L., and M. Carson. 1998. Genetic variation in Cercospora and the potential impact on selecting for resistance to gray leaf spot of corn. p. 334–347. In: Proc. Fifty-third Ann. Corn & Sorghum Res. Conf., Chicago, IL. 9–10 Dec. 1998. Am. Seed Trade Assoc., Washington, DC. Duvick, D. N., and K. G. Cassman. 1999. Post-green revolution trends in yield potential of temperate maize in the north-central United States. Crop Sci. 39:1622–1630. Efron, Y., S. K. Kim, J. M. Fajemisin, J. H. Mareck, C. Y. Tang, Z. T. Dabrowski, H. W. Rossel, and G. Thottappilly. 1989. Breeding for resistance to maize streak virus: A multidisciplinary approach. Plant Breed. 103:1–36. Elwinger, E. F., M. W. Johnson, R. R. Hill, and J. E. Ayers. 1990. Inheritance of resistance to gray leaf spot of corn. Crop Sci. 30:350–358.

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

165

Engelbrecht, G. C. 1975. Streak, a major threat? South African Dept. Agr. Tech. Serv. Tech. Commun. 132:101–103. Fehr, W. R. 1987. Vols. 1 and 2, Principles of cultivar development. Macmillan, New York. Findley, W. R., L. M. Josephsen, and E. J. Dollinger. 1981. Breeding for disease resistance in corn. p. 137–140. In: D. T. Gordon, J. K. Knoke, and G. E. Scott (eds.), Virus and viruslike diseases of maize in the United States. So. Coop. Ser. Bul. 247. Wooster, OH. Findley, W. R., R. Louie, and J. K. Knoke. 1984. Breeding corn for resistance to corn viruses in Ohio. p. 52–67. In: Proc. Thirty-ninth Ann. Corn & Sorg. Res. Conf., Chicago, IL, 5–6 Dec. 1984. Publication 39, Edited and Published by D. Wilkinson and R. Brown, Am. Seed Trade Assoc., Washington, DC. Fourie, A. P., and J. H. Piennar. 1983. Breeding for resistance to maize streak virus: A report on the Vaalharts breeding programme. p. 44–50. In: Proc. 5th South African Maize Breeding Symposium, 23–24 March 1982. Dep. Agric. Tech. Commun. Potchefstroom, S. Africa. Frederiksen, R. A. 1980. Sorghum downy mildew in the United States: Overview and outlook. Plant Dis. 64:903–908. Frederiksen, R. A., and B. L. Renfro. 1977. Global status of maize downy mildew. Annu. Rev. Phytopathology 15:249–275. Freppon, J. T., P. E. Lipps, and R. C. Pratt. 1994. Characterization of the chlorotic lesion response by maize to Cercospora zeae-maydis. Plant Dis. 78:945–949. Freppon, J. T., R. C. Pratt, and P. E. Lipps. 1996. Chlorotic lesion response of maize to Cercospora zeae-maydis and its effect on gray leaf spot disease. Phytopathology 86:733–738. Freymark, P. J., M. Lee, C. A. Martinson, and W. L. Woodman. 1993. Quantitative and qualitative trait loci affecting host-plant response to Exserohilum turcicum in maize (Zea mays L.) Theor. Appl. Genet. 87:537–544. Freymark, P. J., M. Lee, W. L. Woodman, and C. A. Martinson. 1994. Molecular-markerfacilitated investigation of host-plant response to Exserohilum turcicum in maize (Zea mays L.): components of resistance. Theor. Appl. Genet. 88:305–313. Gámez, R., and P. León. 1985. Ecology and evolution of neotropical leafhopper-virus-maize association. p. 331–350. In: L. R. Nault and J. G. Rodriquez (eds.), The Leafhopper and planthopper. Wiley, New York. Geiger, H. H., and M. Heun. 1989. Genetics of quantitative resistance to fungal pathogens. Annu. Rev. Phytopathol. 27:317–341. Gevers, H. O., and J. K. Lake. 1994. GLS1- a major gene for resistance to gray leaf spot in maize. Suid-Afrikaanse Tydskrif vir Wetenskap 90:377–379. Gevers, H. O., J. K. Lake, and T. Hohls. 1994. Diallel cross analysis of resistance to gray leaf spot in maize. Plant Dis. 78:379–383. Giménez-Pecci, M. P., E. Oliveira, R. Resende, C. Borgogno, C. F. Nome, and I. G. Laguna. 2000. Occurrence of maize rayado fino virus in maize in Argentina. Plant Dis. 84:46. Godoy, C. V., L. Amorim, A. Bergamin-Filho, H. P. Silva, W. Silva, and R. D. Berger. 2003. Temporal progress of southern rust in maize under different environmental conditions. Fitopatol. Brasília 28:273–278. Gordon, D. T., O. E. Bradfute, R. E. Gingery, J. K. Knoke, R. Louie, L. R. Nault, and G. E. Scott. 1981. Introduction: History, geographical distribution, pathogen characteristics, and economic importance. p. 1–12. In: D. T. Gordon, J. K. Knoke, and G. E. Scott (eds.), Virus and viruslike diseases of maize in the United States. Southern Regional Coop. Bul. 247. Wooster, OH. Gordon, S. G., M. Bartsch, I. Matthies, H. O. Gevers, P. E. Lipps, and R. C. Pratt. 2004. Linkage of molecular markers to Cercospora zeae-maydis resistance in maize. Crop Sci. 44:628–636.

166

R. PRATT AND S. GORDON

Gorter, G. J. M. A. 1959. Breeding maize resistant to streak. Euphytica 8:234–240. Govinden, N., and K. Rummun. 1996. Registration of MSIRI 3B streak resistant germplasm. Crop Sci. 36:824. Graham, M. J., J. A. Hawk, R. B. Carroll, J. E. Ayers, K. R. Lamkey, and A. R. Hallauer. 1993. Evaluation of Iowa stiff stalk synthetic for resistance to gray leaf spot. Plant Dis. 77:382–385. Groth, J. V., J. K. Pataky, and G. R. Gingera. 1992. Virulence in eastern North American populations of Puccinia sorghi to Rp resistance genes in corn. Plant Dis. 76:1140–1144. Guthrie, W. D., C. T. Tseng, J. Knoke, and J. L. Jarvis. 1982. European corn-borer and maize chlorotic dwarf virus resistance-susceptibility in inbred lines of dent maize. Maydica 27:221–233. Hakiza, J. J., P. E. Lipps, S. St. Martin, and R. C. Pratt. 2004. Heritability and number of genes controlling partial resistance Exserohilum turcicum in maize inbred H99. Maydica 49:173–182. Hallauer, A. R., K. R. Lamkey, and P. R. White. 2001. Registration of B115 inbred line of maize. Crop Sci. 41:2012–2013. Hammond, R. W., and I. P. Bedendo. 2001. Role of Maize rayado fino virus in the etiology of “red stunt” disease in Brazil. Plant Dis. 85:99. Hammond, R. W., and P. Ramirez. 2001. Molecular characterization of the genome of maize rayado fino virus, the type member of the genus Marafivirus. Virology 282:338–347. Hawk, J. A. 1985. Registration of DE811 germplasm line of maize. Crop Sci. 25:716. Hernandez, J. R., M. Yasem de Romero, C. G. Díaz, and J. C. Ramallo. 2002. First report of Puccinia polysora on corn in Argentina. Plant Dis. 86:187. Hogenboom, N. G. 1993. Economic importance of breeding for disease resistance. In: T. Jacobs and J. E. Parlevliet (eds.), Durability of disease resistance, Proc. Int. Symp., Wageningen, The Netherlands. 24–28 Feb. 1992. Curr. Plant Sci. Biotech. Agr. 18:5–9. Kluwer Acad. Publ., Dordrecht, The Netherlands. Holland, J. B., D. V. Uhr, D. Jeffers, and M. M. Goodman. 1998. Inheritance of resistance to southern corn rust in tropical-by-corn-belt maize. Theor. Appl. Genet. 96:232–241. Holley, R. N., and M. M. Goodman. 1984. New sources of resistance to southern corn leaf blight from tropical hybrid maize derivatives. Plant Dis. 73:562–564. Horner, E. S. 1990. Registration of maize germplasms FS8A(S), FS8A(T), FS8B(S), and FS8B(T). Crop Sci. 30:964. Hu, G., and S. Hulbert. 1996. Construction of ‘compound’ rust resistance genes in maize. Euphytica 87:45–51. Hughes, G. R., and A. L. Hooker. 1971. Gene action conditioning resistance to northern corn leaf blight in maize. Crop Sci. 11:180–183. Hulbert, S. H. 1997. Structure and evolution of the rp1 complex conferring rust resistance in maize. Annu. Rev. Phytopathol. 35:293–310. Hulbert, S. H., P. C. Lyons, and J. L. Bennetzen. 1991. Reactions of maize lines carrying Rp resistance genes to isolates of the common rust pathogen, Puccinia sorghi. Plant Dis. 75:1130–33. Jenkins, M. T., and A. L. Robert. 1961. Further genetic studies of resistance to Helminthosporium turcicum Pass. in maize by means of chromosomal translocation. Crop Sci. 1:450–455. Johnson, M. W. 1989. Registration of Pa356, Pa376 and Pa891 parental lines of maize. Crop Sci. 29:1333–1334. Johnson, M. W. 1992. Registration of Pa778 and Pa860 parental lines of maize. Crop Sci. 32:1516. Johnson, M. W., and J. E. Ayers. 1988a. Registration of three parental lines of maize with improved resistance to gray leaf spot. Crop Sci. 28:876.

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

167

Johnson, M. W., and J. E. Ayers. 1988b. Registration of four parental lines of maize. Crop Sci. 28:876. Johnson, M. W., and J. E. Ayers. 1988c. Registration of eight maize germplasm sources for gray leaf spot (GLS) resistance. Crop Sci. 28:871–872. Johnson, R. 1984. A critical analysis of durable resistance. Ann. Rev. Phytopath. 22:309–330. Josephson, L. M., H. C. Kincer, and D. R. West. 1982. Registration of 10 germplasm lines of maize. Crop Sci. 22:696–697. Kang, M. S. 1990. Inheritance of resistance to Stewart’s bacterial wilt in maize. Pflanzenzuecht. 104:173–175. Kim, S. K., S. O. Ajala, and J. G. Kling. 2003. Combining ability of tropical maize germplasm in West Africa. IV. Inheritance of resistance to downy mildew (Peronosclerospora sorghi) infection. Maydica 48:9–14. Kim, S. K., and J. L. Brewbaker. 1977. Inheritance of general resistance in maize to Puccinia sorghi Schw. Crop Sci. 77:456–461. Kim, S. K., Y. Efron, J. M. Fajemisin, and I. W. Buddenhagen. 1989. Mode of gene action for resistance in maize to maize streak virus. Crop Sci. 29:890–894. Kim, S. K., Y. Efron, F. Khadr, J. Fajemisin, and M. H. Lee. 1987. Registration of 16 maizestreak virus resistant tropical maize parental inbred lines. Crop Sci. 27:824–825. Kim, S. K., J. G. Kling, J. E. Iken, K. F. Cardwell, A. O. Adenola, and V. Adenle. 1994. Breeding maize for downy mildew resistance. p. 26–33. In: K. F. Cardwell (ed.), Eradication strategy for downy mildew. In: Proc. of an African conference and workshop. 22–24 February 1994. IITA/FAO/FDA. IITA, Ibadan, Nigeria. Knight, J. 2003. Crop improvement: a dying breed. Nature 421:568–570. Kogel, R., R. W. Hammond, and P. Ramirez. 1996. Incidence and geographic distribution of maize rayado fino virus (MRFV) in Latin America. Plant Dis. 80:679–683. Konate, G., and O. Traore. 1994. Variability of the maize streak virus (MSV) in the SudanoSahelian region. Phytoprotection 75:91–99. Koorneef, M., and P. Stam. 2001. Changing paradigms in plant breeding. Plant Physiol. 125:156–159. Kraja, A., J. W. Dudley, and D. G. White. 2000. Identification of tropical and temperate maize populations having favorable alleles for disease resistance. Crop Sci. 40:948–954. Kuntze, L., E. Fuchs, M. Gruntzig, B. Schulz, D. Klein, and A. E. Melchinger. 1997. Resistance of early-maturing European maize germplasm to sugarcane mosaic virus (SCMV) and maize dwarf mosaic virus (MDMV). Plant Breed. 116:499–501. Kyetere, D., R. Ming, M. D. McMullen, R. C. Pratt, J. Brewbaker, T. Musket, K. V. Pixley, and H. G. Moon. 1995. Monogenic tolerance to maize streak virus maps to the short arm of chromosome 1. Maize Genet. Coop. Newslett. 69:136–137. Kyetere, D. T., R. Ming, M. D. McMullen, R. C. Pratt, J. Brewbaker, and T. Musket. 1999. Genetic analysis of tolerance to maize streak virus in maize. Genome 42:20–26. Lambert, R. J. 1996. Registration of RSSSCC6 and RBS10C6 maize germplasm. Crop Sci. 36:215. Latterell, F. M., and A. E. Rossi. 1983. Gray leaf spot of corn: a disease on the move. Plant Dis. 67:842–847. Lehmensiek, A., A. M. Esterhuizen, D. van Staden, S. W. Nelson, and A. E. Retief. 2001. Genetic mapping of gray leaf spot (GLS) resistance genes in maize. Theor. Appl. Genet. 103:979–803. Lenardon, S. L., G. J. March, S. F. Nome, and J. A. Ornaghi. 1998. Recent outbreak of mal de río cuarto virus on corn in Argentina. Plant Dis. 82:448. León, P., and R. Gámez. 1981. Some physiochemical properties of maize rayado fino virus. J. Gen. Virol. 56:67–75.

168

R. PRATT AND S. GORDON

Leonard, K. J., Y. Levy, and D. R. Smith. 1989. Proposed nomenclature for pathogenic races of Exserohilum turcicum on corn. Plant Dis. 73:776–777. Lindhout, P. 2002. The perspectives of polygenic resistance in breeding durable disease resistance. Euphytica 124:217–226. Lipps, P. E. 1987. Gray leaf spot epiphytotic in Ohio corn. Plant Dis. 71:281. Lipps, P. E., R. C. Pratt, and J. J. Hakiza. 1997a. Interaction of Ht and partial resistance to Exserohilum turcicum in maize. Plant Dis. 81:277–282. Lipps, P. E., P. R. Thomison, and R. C. Pratt. 1997b. Reactions of corn hybrids to gray leaf spot. p. 163–180. In: Proc. Fifty-first Ann. Corn & Sorg. Res. Conf., Chicago, IL, Dec. 1997. Publication 51. Am. Seed Trade Assoc., Inc., Washington, DC. Louie, R. 1999. Diseases caused by viruses. p. 49–55. In: D. G. White (ed.), Compendium of corn diseases. APS Press, St. Paul, MN. Louie, R., and R. J. Anderson. 1993. Evaluation of maize chlorotic dwarf virus resistance in maize with multiple inoculations by Graminella nigrifrons (Homoptera, Cicadellidae). J. Econ. Entomol. 86:1579–1583. Louie, R., M. W. Jones, R. A. Anderson, and M. G. Redinbaugh. 2002. Notice of release of Oh1VI. Crop Sci. 42:991. Louie, R., and J. K. Knoke. 1981. Symptoms and disease diagnosis. p. 13–18. In: D. T. Gordon, J. K. Knoke, and G. E. Scott (eds.), Virus and viruslike diseases of maize in the United States. Southern Coop. Ser. Bul. 247. Wooster, OH. Louie, R., J. K. Knoke, and W. R. Findley. 1990. Elite maize germplasm—reactions to maize dwarf mosaic and maize chlorotic dwarf viruses. Crop Sci. 30:1210–1215. Lu, X. W., J. L. Brewbaker, S. M. Nourse, H. G. Moon, S. K. Kim, and M. Khairallah. 1999. Mapping of quantitative trait loci conferring resistance to maize streak virus. Maydica 44:313–318. Lubberstedt, T., D. Klein, and A. E. Melchinger. 1998. Comparative quantitative trait loci mapping of partial resistance to Puccinia sorghi across four populations of European flint maize. Phytopathology 88:1324–1329. Lubberstedt, T., X. C. Xia, M. L. Xu, L. Kuntze, and A. E. Melchinger. 1999. Inheritance of resistance to SCMV and MDMV in European maize. p. 241–250. In: G. T. ScarasciaMugnozza, E. Porcedu, and M. A. Pagnotta (eds.), Genetics and breeding for crop quality and resistance. Proc. XV EUCARPIA Congress. Niterbo, Italy, Sept. 20–25, 1998. Martin, D. P., J. A. Willment, R. Billharz, R. Velders, B. Odhiambo, J. Njuguna, D. James, and E. P. Rybicki. 2001. Sequence diversity and virulence in Zea mays of maize streak virus isolates. Virology 288:247–255. Martin, D. P., J. A. Willment, and E. P. Rybicki. 1999. Evaluation of maize streak virus pathogenicity in differentially resistant Zea mays genotypes. Phytopathology 89:695–700. MBS. 2002. Genetic handbook. MBS, Inc., Story City, IA. McMullen, M. D., and R. Louie. 1989. The linkage of molecular markers to a gene controlling the symptom response in maize to maize dwarf mosaic virus. Mol. PlantMicrobe Interact. 2:309–314. Mesfin, T., N. A. Bosque-Perez, I. W. Buddenhagen, G. Thottappilly, and S. O. Olojede. 1992. Studies of maize streak virus isolates from grass and cereal hosts in Nigeria. Plant Dis. 76:789–795. Michelmore, R. W. 2003. The impact zone: genomics and breeding for durable disease resistance. Curr. Opin. Plant Biol. 6:397–404. Michener, P. M., and J. K. Pataky. 2002. Stewart’s wilt reactions of South African maize varieties inoculated with Erwinia stewartii in field and greenhouse trials. African Plant Protec. 8:33–40.

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

169

Mikel, M. A., C. J. D’Arcy, A. M. Rhodes, and R. E. Ford. 1984. Genetics of resistance of two dent corn inbreds to maize dwarf mosaic virus and transfer of resistance into sweet corn. Phytopathology 74:467–473. Ming, R., J. L. Brewbaker, H. G. Moon, T. A. Musket, R. D. Holley, J. K. Pataky, and M. D. McMullen. 1999. Identification of RFLP markers linked to a major gene, sw1, conferring resistance to Stewart’s wilt in maize. Maydica 44:319–323. Ngwira, P., E. M. Sibale, W. G. Nhlane, and V. W. Saka. 1999. An overview of the status of maize diseases in Malawi. Proc. African Crop Science Conf. 4:457–461. Fourth African Crop Sci. Conf., Casablanca, Morocco, 11–14. Oct. 1999. Editorial Secretariat, African Crop Sci. Soc., Kampala, Uganda. Nyvall, R. F. 1989. Field crop diseases handbook, 2nd edition. Van Nostrand Reinhold, New York. Okori, P., D. Kyetere, and E. Adipala 1997. Greenhouse reaction of maize plants to Exserohilum turcicum and maize streak virus concomitant infection. Proc. African Crop Science Conf. 3:927–931. Third African Crop Science Conf., Pretoria, South Africa, 1999. Editorial Secretariat, African Crop Sci. Soc., Kampala, Uganda. Okori, P., G. Asea, G. Bigirwa, and E. Adipala. 1999. An overview of status of maize diseases in Uganda. Proc. African Crop Sci. Conf. 4:463–468. Fourth African Crop Science Conf., Casablanca, Morocco. Oct. 11–14,1999. Editorial Secretariat, African Crop Science Society, Kampala, Uganda. Okori, P., J. Fahleson, P. R. Rubaihayo, E. Adipala, and C. Dixelius. 2003. Assessment of genetic variation among East African Cercospora zeae-maydis populations. African Crop Sci. J. 11:75–85. Paliwal, R. L. 2000. Maize diseases. p. 63–80. In: R. L. Paliwal, G. Granados, H. R. Lafitte, A. D. Violic, and J. P. Marathee (eds.), Tropical maize improvement and production. FAO Plant Production and Protection Series 28, FAO, Rome. Parker, G. B., and A. L. Hooker. 1993. Inheritance of resistance to Erwinia stewartii in four inbred lines of dent corn: qualitative and quantitative analyses. Maydica 38:223–229. Parlevliet, J. E. 1983. Can horizontal resistance be recognized in the presence of vertical resistance in plants exposed to a mixture of pathogen races? Plant Dis. 73:379. Parlevliet, J. E. 1992. Selecting components of partial resistance. p. 281–302. In: H. T. Stalker and J. P. Murphy (eds.), Plant breeding in the 1990s: Proc. of the Symposium on Plant Breeding in the 1990s, Raleigh, NC, March 1991. CABI, UK. Parlevliet, J. E. 2002. Durability of resistance against fungal, bacterial and viral pathogens; present situation. Euphytica 124:147–156. Pataky, J. K. 1999. Rusts. p. 35–38. In: D. G. White (ed.), Compendium of corn diseases, 3rd ed. APS Press, St. Paul, MN. Pataky, J. K., J. M. Perkins, and S. Leath. 1986. Effects of qualitative and quantitative resistance on the development and spread of northern leaf-blight of maize caused by Exserohilum turcicum race-1 and race-2. Phytopathology 76:1349–1352. Pataky, J. K., L. J. du Toit, and N. D. Freeman. 2000. Stewart’s wilt reactions of an international collection of Zea mays germ plasm inoculated with Erwinia stewartii. Plant Dis. 84:901–906. Pernet, A., D. Hoisington, J. Dintinger, D. Jewel, C. Jiang, M. Khairallah, P. Letourmy, J. L. Marchand, J. C. Glaszmann, and D. Gonzalez de Leon. 1999a. Genetic mapping of maize streak virus resistance from the Mascarene source II. Resistance in line CIRAD390 and stability across germplasm. Theor. Appl. Genet. 99:540–553. Pernet, A., D. Hoisington, J. Franco, M. Isnard, D. Jewel, C. Jiang, J. L. Marchand, B. Reynaud, J. C. Glaszmann, and D. Gonzalez de Leon. 1999b. Genetic mapping of maize

170

R. PRATT AND S. GORDON

streak virus resistance from the Mascarene source I. Resistance in line D211 and stability against different virus clones. Theor. Appl. Genet. 99:524–539. Pham, H. 1992. Characterization of tropical maize streak resistant breeding lines developed at CIMMYT Harare. In: 1992 Agronomy abstracts. ASA, Madison, WI. Pingali, P. L., and S. Pandey. 2001. Meeting world maize needs: technological opportunities and priorities for the public sector. p. 1–24. In: P. L. Pingali (ed.), 1999/2000 World maize facts and trends. CIMMYT, Mexico. Pinner, M. S., and P. G. Markham. 1990. Serotyping and strain identification of maize streak virus isolates. J. Gen. Virol. 71:1635–1640. Poneleit, C. 1990. Two virus resistant maize germplasms. Crop Sci. 30:756. Pratt, R. C., R. J. Anderson, R. Louie, M. D. McMullen, and J. K. Knoke. 1994a. Maize responses to a severe isolate of maize chlorotic dwarf virus. Crop Sci. 34:635–641. Pratt, R. C., and W. R. Findley. 1994. Registration of OhS11 (C1). Crop Sci. 34:1131. Pratt, R. C., W. R. Findley, A. Sotomayor-Rios, and S. Torres-Cardona. 1994b. Registration of OhS9(C1) and OhS10(C1) maize germplasm. Crop Sci. 34:1418–1419. Pratt, R. C., S. Gordon, P. Lipps, G. Asea, G. Bigirwa, and K. Pixley. 2003. Use of IPM in the control of multiple diseases of maize: strategies for selection of host resistance. African Crop Sci. J. 11:189–198. Raymundo, A. D., and A. L. Hooker. 1982. Single and combined effects of monogenic and polygenic resistance on certain components of northern corn leaf blight development. Phytopathology 72:99–103. Redinbaugh, M. G., M. W. Jones, and R. E. Gingery. 2004. The genetics of virus resistance in maize (Zea mays L.). Maydica 49:183–190. Roane, C. W., and S. A. Tolin. 1989. Problems in maize MDMV genetics. Maize Genet. Coop. Newslett. 63:41–41. Roane, C. W., S. A. Tolin, and H. S. Aycock. 1989a. Genetics of reaction to maize dwarf mosaic virus strain A in several maize inbred lines. Phytopathology 79:1364–1368. Roane, C. W., S. A. Tolin, H. S. Aycock, and P. J. Donahue. 1989b. Association of Rmd1, a gene conditioning reaction to maize dwarf mosaic virus, with genes conditioning endosperm color (y1) and type (su2) in maize. Phytopathology 79:1368–1372. Rodier, A., J. Assie, J. L. Marchand, and Y. Herve. 1995. Breeding maize lines for complete and partial resistance to maize steak virus (MSV). Euphytica 81:57–70. Rodriguez-Pardina, P. E., M. P. Gim-Pecci, I. G. Laguna, and G. Truol. 1998. Wheat: a new natural host for the Mal de Cuarto virus in the endemic disease area, Cuarto, Cordoba Province, Argentina. Plant Dis. 82:149–152. Rose, D. J. W. 1978. Epidemiology of maize streak disease. Annu. Rev. Entomol. 23:259–282. Rosegrant, M. W., M. S. Paisner, S. Meijer, and J. Witcover. 2001. Recent trends in food supply and demand. p. 3–49. In: Global food projections to 2020: Emerging trends and alternative futures. International Food Policy Res. Inst., Washington, DC. Rosenkranz, E., and G. E. Scott. 1984. Determination of the number of genes for resistance to maize dwarf mosaic virus strain A in five corn inbred lines. Phytopathology 74:71–76. Rufener, G. K., A. J. Balducchi, R. P. Mowers, R. C. Pratt, R. L. Louie, M. McMullen, and J. Knoke. 1996. United States Patent 5,563,316 Maize chlorotic dwarf virus resistant maize and the production thereof. http://www.uspto.gov/patft/index.html. Russell, W. A., L. H. Penny, G. F. Sprague, W. D. Guthrie, and F. F. Dicke. 1971. Registration of maize parental lines. Crop Sci. 11:143. Russell, W. A., and A. R. Hallauer. 1974. Registration of B76 and B77 parental lines of maize. Crop Sci. 14:778. Saghai-Maroof, M. A., S. W. Van Scoyoc, Y. G. Yu, and E. L. Stromberg. 1993. Gray leaf spot disease of maize: rating methodology and inbred line evaluation. Plant Dis. 77:583–587.

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

171

Saghai-Maroof, M. A., G. Y. Yue, Z. X. Xiang, E. L. Stromberg, and G. K. Rufener. 1996. Identification of quantitative trait loci controlling resistance to gray leaf spot. Theor. Appl. Genet. 93:539–546. Schechert, A. W., H. G. Welz, and H. H. Geiger. 1999. QTL for resistance to Setosphaeria turcica in tropical African maize. Crop Sci. 39:514–523. Scott, G. E. 1983. Breeding for and genetics of virus resistance in field corn. p. 248–254. In: D. T. Gordon, J. K. Knoke, L. R. Nault, and R. M. Ritter (eds.), Proc. Int. Maize Virus Disease Colloquium and Workshop, Wooster, OH. Aug. 2–6, 1982. Scott, G. E. 1989. Linkage between maize dwarf virus resistance and endosperm color in maize. Crop Sci. 29:1478–1480. Scott, G. E. 1993. Registration of Mp339 and Mp412 parental lines of maize. Crop Sci. 33:888. Scott, G. E., W. R. Findley, and E. J. Dollinger. 1981. Genetics of resistance in corn. p. 141–144. In: D. T. Gordon, J. K. Knoke, and G. E. Scott (eds.), Virus and virus-like diseases of maize in the United States. Southern Cooperative Series Bul. 247. Wooster, OH. Scott, G. E., S. B. King, and J. W. Armour, Jr. 1984. Inheritance of resistance to southern corn rust in maize populations. Crop Sci. 24:265–267. Scott, G. E., and R. Louie. 1996. Improved resistance to maize dwarf mosaic virus by selection under greenhouse conditions. Crop Sci. 36:1503–1506. Scott, G. E., and E. E. Rosenkranz. 1975. Frequency of progenies resistant to corn stunt and maize dwarf mosaic in maize populations. Crop Sci. 15:233–235. Scott, G. E., and E. E. Rosenkranz. 1982. A new method to determine the number of genes for resistance to maize dwarf mosaic in maize. Crop Sci. 22:756–761. Scott, G. E., and N. Zummo. 1989. Effect of genes with slow-rusting characteristics on southern corn rust in maize. Plant Dis. 73:114–116. Scott, G. E., and N. Zummo. 1990. Registration of Mp313E parental line of maize. Crop Sci. 30:1378. Sharma, R. C., and M. M. Payak. 1990. Durable resistance to two leaf blights in two maize inbred lines. Theor. Appl. Genet. 80:542–544. Shukla, D. D., C. W. Ward, and A. A. Brunt. 1994. The potyviridae. CABI, Wallingford, UK. Simcox, K. D., M. D. McMullen, and R. Louie. 1995. Co-segregation of the maize dwarf mosaic virus resistance gene, Mdm1, with the nucleolus organizer region in maize. Theor. Appl. Genet. 90:341–346. Simmonds, N. W. 1985. A plant breeder’s perspective of durable resistance. FAO Plant Prot. Bul. 33:13–17. FAO, Rome, Italy. Simmonds, N. W., and J. Smartt. 1999. Principles of crop improvement (2nd ed.). Blackwell Science, Oxford, UK. Smith, D. R. 1999. Global disease assessment of corn. p. 111–123. In: Proc. Fifty-fourth Ann. Corn & Sorghum Res. Conf., Chicago, IL. 9–10 Dec. 1999. Publication 54, Am. Seed Trade Assoc., Inc., Washington, DC. Smith, D. R., and J. G. Kinsey. 1993. Latent period—a possible selection tool for Exserohilum turcicum resistance in corn (Zea mays L.). Maydica 38:205–208. Smith, D. R., and B. L. Renfro. 1999. Downy mildews. p. 25–27. In: D. G. White (ed.), Compendium of corn diseases, 3rd ed. APS Press, St. Paul, MN. Soto, P. E., I. W. Buddenhagen, and V. L. Asnani. 1982. Development of streak virus– resistant maize populations through improved challenge and selection methods. Ann. Appl. Biol. 100:539–546. Storey, H. H., and A. K. Howland. 1967. Inheritance of resistance in maize to the virus of streak disease in East Africa. Ann. Appl. Biol. 59:429–436. Suneson, C. A. 1960. Genetic diversity—a protection against plant diseases and insects. Agron. J. 52:319–321.

172

R. PRATT AND S. GORDON

Tallury, S. P., and M. M. Goodman. 1999. Experimental evaluation of the potential of tropical germplasm for temperate maize improvement. Theor. Appl. Genet. 98:54–61. Tang, C. Y., and M. S. Bjarnason. 1993. Two approaches for the development of maize germplasm resistant to maize streak virus. Maydica 38:301–307. Thompson, D. L., and R. R. Bergquist. 1984. Inheritance of mature plant resistance to Helminthosporium maydis race O in maize. Crop Sci. 24:807–811. Thompson, D. L., R. R. Berquist, G. A. Payne, D. T. Bowman, and M. M. Goodman. 1987. Inheritance of resistance to gray leaf spot in maize. Crop Sci. 27:243–246. Thottappilly, G., N. A. Bosque-Perez, and H. W. Rossel. 1993. Viruses and virus diseases of maize in tropical Africa. Plant Pathol. 42:494–509. Troyer, A. F. 1999. Background of U.S. hybrid corn. Crop Sci. 39:601–626. Troyer, A. F. 2001. Temperate corn: Background, breeding and behavior. p. 393–466. In: A. R. Hallauer (ed.), Specialty corns. CRC Press, Boca Raton, FL. Ullstrup, A. J. 1953. Some smuts and rusts of corn. In: Plant diseases, The yearbook of agriculture. USDA, Washington, DC. Ullstrup, A. J. 1965. Inheritance and linkage of a gene determining resistance to an American race of Puccinia polysora. Phytopathology 55:425–428. Ulrich, J. F., J. A. Hawk, and R. B. Carroll. 1990. Diallel analysis of maize inbreds for resistance to gray leaf spot. Crop Sci. 30:1198–1200. USDA-ARS. 2004. Germplasm Resources Information Network—(GRIN). [Online Database] National Germplasm Resources Laboratory, Beltsville, Maryland. Available: http:// www.ars-grin.gov/cgi-bin/npgs/html/dno_eval_acc.pl?89009+205001+1780. van Rensburg, G. D. J., J. H. Giliomee, and K. L. Pringle. 1991. Resistance of South African maize hybrids to maize streak virus. S. African J. Plant Soil Sci. 8:38–42. Vasal, S. K., G. Srinivasan, S. D. McLean, N. Vergara, B. S. Dhillon, S. H. Zhang, F. San Vincente, and S. K. Ramanujam. 1997. Registration of 21 tropical yellow-endosperm lines of maize. Crop Sci. 37:1402–1403. Wang, J., M. Levy, and L. D. Dunkle. 1998. Sibling species of Cercospora associated with gray leaf spot of maize. Phytopathology. 88:1269–1275. Ward, J. M. J., E. L. Stromberg, D. C. Nowell, and F. W. Nutter. 1999. Gray leaf spot: A disease of global importance in maize production. Plant Dis. 83:884–895. Warren, H. L. 1981. Registration of H102 and H103 maize germplasm. Crop Sci. 21:477. Warren, H. L. 1982. Registration of H110 and H111 maize germplasm. Crop Sci. 22:1270–1271. Welz, H. G., and H. H. Geiger. 2000. Genes for resistance to northern corn leaf blight in diverse maize populations. Plant Breed. 119:1–14. Welz, H. G., T. Lubberstedt, P. Basseti, H. Y. Geiger, and A. E. Melchinger. 1998a. Resistance of early-maturing European maize inbred lines to Setosphaeria turcica. Maydica 43:27–33. Welz, H. G., A. Schechert, A. K. Pernet, K. Pixley, and H. H. Geiger. 1998b. A gene for resistance to the maize streak virus in the African CIMMYT maize inbred line CML202. Mol. Breed. 4:147–154. White, D. G., and M. L. Carson. 1999. History of corn. p. 1–2. In: D. G. White (ed.), Compendium of corn diseases, 3rd ed. APS Press, St. Paul, MN. Wicks III, Z. W., M. L. Carson, H. Rahman, and G. G. Scholten. 1990. Registration of SD45, SD47, and SD48 parental lines of maize. Crop Sci. 30:243. Wicks III, Z. W., M. L. Carson, and D. L. Robbins. 1986a. Registration of SD40 parental line of maize. Crop Sci. 26:1267. Wicks III, Z. W., M. L. Carson, and D. L. Robbins. 1986b. Registration of SD41 parental line of maize. Crop Sci. 26:1267–1268.

3. BREEDING RESISTANCE TO MAIZE FOLIAR PATHOGENS

173

Wicks, Z. W., M. L. Carson, and D. L. Robbins. 1987. Registration of SD42 parental line of maize. Crop Sci. 27:1324. Wicks III, Z. W., M. L. Carson, and G. G. Scholten. 1989. Registration of SD46 parental line of maize. Crop Sci. 29:840. Wilkinson, D. R., and A. L. Hooker. 1968. Genetics of reaction to Puccinia sorghi in ten corn inbred lines from Africa and Europe. Phytopathology 58:605–608. Wright, S. 1968. Evolution and the genetics of populations. Genetics and biometrics foundations. Vol. 1. The University of Chicago Press, Chicago, IL. York, J. O. 1991. Registration of Ark. SWCB Syn., Ark. Leaf Feed Res. Syn., and Ark. CornBorer Comp. #1, #2, #3 maize germplasm. Crop Sci. 31:501–502. Zaitlin, D., S. DeMars, and Y. Ma. 1993. Linkage of rhm, a recessive gene for resistance to southern corn leaf blight, to RFLP marker loci in maize (Zea mays) seedlings. Genome 36:555–564. Zuber, M. S. 1973. Registration of 20 maize parental lines. Crop Sci. 13:779–780. Zummo, N. 1988. Components contributing to partial resistance in maize to Puccinia polysora. Plant Dis. 72:157–160.

4 Synteny in the Rosaceae Pere Arús Department de Genètica Vegetal, Laboratori de Genètica Molecular Vegetal, CSIC-IRTA Carretera de Cabrils s/n; 08348, Cabrils, Spain Toshiya Yamamoto National Institute of Fruit Tree Science Tsukuba, Ibaraki 305-8605, Japan Elisabeth Dirlewanger INRA, Unité de Recherches sur les Espèces Fruitières et la Vigne B.P. 81, F-33 883 Villenave d’Ornon cedex, France Albert G. Abbott Department of Genetics, Biochemistry and Life Science Studies, Clemson University Clemson, South Carolina 29634, USA

I. INTRODUCTION II. GENETIC MAPS IN THE MAIN ROSACEAE SPECIES A. Subfamily Prunoideae 1. Subgenus Amygdalus 2. Subgenus Prunophora 3. Subgenus Cerasus B. Subfamily Maloideae 1. Apple 2. Pear C. Subfamily Rosoideae III. MAP COMPARISONS A. Within the Prunus Genus B. Between Apple and Pear C. Between Apple and Prunus D. Between Prunus and Arabidopsis Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 175

176

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

IV. OTHER GENETIC RESOURCES OF INTEREST FOR MAP COMPARISON A. The Genome Database for Rosaceae (GDR) B. The Peach Physical Map C. EST Functional Genomics Database Development V. FUTURE PROSPECTS LITERATURE CITED

I. INTRODUCTION The Rosaceae is a large and diverse family that includes deciduous and evergreen trees, shrubs, and herbs. It consists of about 100 genera and more than 2,000 species distributed worldwide, although it is most abundant in the colder and temperate Northern regions. The family includes numerous economically important crops, grown for their fruits, nuts, and timber or for their ornamental value. The characteristic flowers of this family have radial symmetry and usually have five sepals, five petals, and numerous stamens. The number of carpels and the ovary position varies, giving rise to different fruit types (achenes, drupes, pomes, or follicles), which are important for sub-family classification. Flowers are usually insect-pollinated and frequently large and showy: a high percentage of all species are actual or potential garden ornamentals. Most species have a gametophytic self-incompatibility system that prevents selfing and requires the presence of two inter-compatible genotypes for fruit production. The family is divided into four subfamilies: Spiraeoideae, Maloideae, Prunoideae, and Rosoideae (Rehder 1947). The three latter subfamilies encompass major cultivated species and will be described in more detail. The ability of biochemical and DNA-based markers to identify homologous loci in different species is one of their most important properties. Comparing the positions of homologous markers in the linkage maps of different species allows the degree of resemblance between their genomes to be established. Synteny, initially defined as the occurrence of two or more genes on the same chromosome, but more recently expanded to describe the similarity between the chromosomes of two species, was studied in the early days of marker discovery, when isozymes were almost the only kind of markers available for these studies (Tanksley 1983). With the development of DNA markers, such as restriction fragment length polymorphisms (RFLPs), maps covering the whole genome could be produced relatively quickly, and the use of the same DNA probes in mapping populations of different species yielded comparable maps. Tanksley et al. (1992) analyzed the tomato and potato genomes, and found that the two genomes have an essentially identical constitution, with the exception of five paracentric inversions. The com-

4. SYNTENY IN THE ROSACEAE

177

parison of rice with maize, which are more distant, was undertaken by Ahn and Tanksley (1993). In spite of the differences in genome size (about 6 times larger in maize) and chromosome number (x = 10 in maize and x = 12 in rice), important syntenic regions between these two species were identified, accounting for 62% of the maize and 70% of the rice linkage maps. Later results established that in general there is a high level of genetic conservation between members of the same family, as demonstrated in the Poaceae (Devos and Gale 2000), Solanaceae (Doganlar et al. 2003), Brassicaceae (Lukens et al. 2003), and Fabaceae (Choi et al. 2004), but synteny decreases considerably between species of different families (Dominguez et al. 2003). Information on comparative mapping in the Rosaceae has been very limited, until recently. The use of markers and map construction started later in this family than in others, and the first saturated maps with transferable markers were produced in the late 1990s (Joobeur et al. 1998; Maliepaard et al. 1998), resulting in a delay with respect to other herbaceous crops, more easily amenable to genetic studies than fruit trees that are woody perennials and have a long intergeneration period. The development of simple-sequence repeat (SSR), or microsatellite markers, has been widespread in the Rosaceae crops (Cipriani et al. 1999; Liebhardt et al. 2002; Sargent et al. 2004; Graham et al. 2004) in the last five years. In addition to their good properties as markers, such as codominance, polymorphism, abundance, and suitability for automation (Weber and May 1989), SSRs have a good rate of transferability among closely related Rosaceae species (Dirlewanger et al. 2002), and occasionally among genera within the same subfamily (Yamamoto et al. 2001b), and have been used for synteny studies. Comparisons between species of different subfamilies require more transferable markers, such as RFLPs, which have been used to a more limited extent for map construction in species of this family. Our objective in this review is to summarize the recent progress in this area, discuss the importance and application of the results in plant breeding, and identify where additional research is needed to completely elucidate the genome relationships of this economically important plant family. II. GENETIC MAPS IN THE MAIN ROSACEAE SPECIES A. Subfamily Prunoideae The five genera in the subfamily Prunoideae are woody plants, either trees or shrubs. The crop species of this subfamily belong to its largest genus, Prunus, and produce drupes as fruits, commonly called “stone fruits.” The most important species of this genus belong to the three

178

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

subgenera, Amygdalus, Prunophora, and Cerasus, in which it is divided, including peach (Prunus persica L. Batsch), almond (Prunus dulcis Mill.), apricot (Prunus armeniaca L.), European plum (Prunus domestica L.), Japanese plum (Prunus salicina Lindl.), sweet cherry (Prunus avium L.), and sour cherry (Prunus cerasus L.). Several other species, such as myrobalan plum (Prunus cerasifera Ehrh.) or Sainte Lucie cherry (Prunus mahaleb L.), are used mainly as Prunus rootstocks. Production of all stone fruits equals 32 million tonnes (FAOSTAT data, 2003, http://faostat.fao.org), ranking second in importance of all temperate fruits, after apple. The base chromosome number of Prunus is x = 8, with most of the cultivated species being diploids, with the exception of sour cherry (2n = 4x = 32) and the European plum (2n = 6x = 48). Peach, the most economically important species of the genus Prunus, has distinct advantages that make it more suitable than others for genetic and genomic analysis. Peach has a short juvenile phase (2 to 3 years) compared to most other fruit tree species, and a small haploid genome of approximately 290 Mbp (Baird et al. 1994), only about twice the size of the Arabidopsis thaliana genome (Arumuganathan and Earle 1991). Moreover, peach is genetically the best characterized Prunus species, with many Mendelian genes controlling morphological traits (Hesse 1975; Scorza and Sherman 1996; Monet et al. 1996), and well developed genomic tools that will be detailed in this review. These attributes make peach a good model species for the Rosaceae (Abbott et al. 2002). 1. Subgenus Amygdalus. The most important crops of this subgenus are peach and almond. They can be intercrossed and produce fertile hybrids but have gross differences in other respects: peach is a self-compatible species with a low level of variability and is used for its fruit, whereas almond is a highly polymorphic and self-incompatible species used for its seed. The first published map in Prunus (Chaparro et al. 1994) was constructed with 83 RAPDs, one isozyme gene, and four morphological single-gene characters in 96 F2 progeny, obtained from the cross between two peach lines, NC174RL and ‘Pillar’. This map covered a total distance of 396 cM and identified 15 of the 16 linkage groups expected. Most markers were dominant and linkage could be detected only for markers in coupling phase. Interspecific F2 or backcross populations between peach and other species have also been used successfully, due to their high degree of segregation, compared to the low variability found in peach intraspecific populations. Foolad et al. (1995) were the first to publish one such map, using an almond × peach F2 population. The first map of almond was constructed entirely with codominant markers (RFLPs and isozymes)

4. SYNTENY IN THE ROSACEAE

179

using the F1 progeny between ‘Ferragnès’ and ‘Tuono’ (Viruel et al. 1995). In total, 16 maps, constructed with 14 populations involving only species of the Amygdalus subgenus (peach, almond, P. ferganensis, and P. davidiana), have been published to date and their characteristics are summarized in Table 4.1. Six of these maps, and 10 more in other Prunus subgenera, have been obtained from crosses between two partially heterozygous and generally unrelated trees. These full-sib populations, also termed F1 segregating progenies, are the common type of breeding populations for many fruit tree species, which explains their abundance and interest. Given the generally self-incompatible nature of some species, such as almond, cherry, plum, or apple, this is often the only possible kind of population. Map construction in these progenies with predominantly dominant markers is usually done with the pseudo-testcross or two-way pseudo-testcross method (Hemmat et al. 1994; Grattapaglia and Sederoff 1994) that implies the construction of two maps, one for each parent of the cross. Each map contains all the markers heterozygous in each parent. When using codominant markers (RFLPs or SSRs), the majority of them segregate as a backcross (Viruel et al. 1995), but a certain number of markers are heterozygous in both parents, giving rise to 1:2:1 or 1:1:1:1 segregation ratios. These latter markers can be mapped in both parents and thus be used as anchors to establish the connection between the two parental maps. As peach is self-compatible, F2 populations can be obtained, allowing a simpler and more standard mapping procedure. The map constructed with the F2 of the cross between ‘Texas’ almond and ‘Earlygold’ peach has been adopted by the stone fruit community as the reference for the genus. This map, originally constructed by a consortium of European groups (Joobeur et al. 1998), was obtained only with codominant markers (235 RFLPs and 11 isozymes). The ‘Texas’ × ‘Earlygold’ map (abbreviated T×E) was considered to be saturated, as all markers could be placed on the expected eight linkage groups (G1 to G8), average marker density was high (2.0 cM/marker), and gaps were scarce and small (only two gaps >10 cM, the largest 12 cM). New high-quality markers were added later to T×E (Aranzana et al. 2003, Dirlewanger et al. 2004a) to produce the current map with 562 markers (361 RFLPs, 185 SSRs, 11 isozymes, and 5 STSs), covering a total distance of 519 cM, with high density (150) this range is approximately 500–800 cM. The total distance of most of the maps was smaller than that commonly found in other species, of approximately 100 cM per chromosome, which may be explained in part by the small size of the Prunus genome. The relationship between chromosomes and linkage groups has not been established yet, but linkage group 1 (G1) of Prunus is longer and more populated with markers than the other linkage groups in most maps (i.e., 121 markers and a distance of 87 cM of G1 vs. an average of 63 markers per linkage group and a distance of 61 cM for the rest of the linkage groups in T×E), and one of the chromosomes of Prunus is clearly longer than the rest (Salesses and Mouras 1977; Corredor et al. 2004), suggesting that G1 corresponds to chromosome 1. 2. Subgenus Prunophora. Linkage map construction started eight years later in this subgenus, which includes apricot and plum, than in Amygdalus. The first published map was in apricot and used mainly AFLPs (Hurtado et al. 2002). A detailed map, constructed in part with markers selected from T×E, was later produced with the F1 population of the cross between ‘Polonais’ and ‘Stark Early Orange’, and is the basis for the comparison between the species of these two subgenera. The unique map involving a plum species was obtained by Dirlewanger et al. (2004b) in one of the parents of a three-way cross between myrobalan plum and the almond × peach rootstock ‘Felinem’. This map was constructed mainly with SSRs, most common to the T×E population. The characteristics of the 6 maps obtained with 4 populations of crosses between species of this subgenus are summarized in Table 4.2. They are similar to those of the Amygdalus subgenus, with distances ranging from 467 to 699 cM. The number of linkage groups is reasonably close to the expected eight in the maps using codominant markers.

4. SYNTENY IN THE ROSACEAE

183

3. Subgenus Cerasus. Six maps have been published in species of this subgenus using five populations (Table 4.2). The earliest was constructed in sweet cherry with RAPDs (Stockinger et al. 1996). Another map created exclusively with isozyme genes was obtained using data from two interspecific cherry progenies (Bos˘kovi´c and Tobutt 1998). This map includes a total of 47 segregating isozyme genes, from which 34 were aligned into seven linkage groups. Sour cherry, an important crop in this subgenus, is tetraploid. Isozyme analysis detected a clear allopolyploid behaviour (Beaver and Iezzoni 1993) and the map obtained by Wang et al. (1998), with RFLPs in an F1 progeny of this species, confirmed these findings, although the segregation of a few loci suggested that a low degree of intergenomic pairing and recombination may occur (Wang et al. 1998). Map sizes in Cerasus are consistent with those obtained in species of the other subgenera, but given that the total number of markers is generally lower, the number of linkage groups is different than that expected, and the size of the largest gaps and the proportion of unlinked markers are higher than in other more populated maps. B. Subfamily Maloideae This subfamily, characterized by a distinctive pome fruit, includes approximately 1,000 species in 30 genera (Westwood 1978), some of which are important fruit tree species, such as apple (Malus spp.), pear (Pyrus spp.), quince (Cydonia oblonga Mill.), loquat (Eryobotrya japonica (Thunb.) Mill.), medlar (Mespilus germanica L.), hawthorn (Crataegus spp.), and others (Kovanda 1965; Westwood 1978; Luby 2003). About 58 million tonnes of apple fruits are produced worldwide in more than 90 countries (FAOSTAT data, 2003, http://faostat.fao.org), and account for 12.1% of all fruit production. Seventeen million tonnes of pears are produced (3.6% of world fruit production), and the other fruit species belonging to Maloideae account for less than 1%. The basic chromosome number is x = 17 for all Maloideae genera (Sax 1931, 1932; Kovanda 1965). While triploid and tetraploid plants have been found in Malus and Pyrus, only diploids are known in Eryobotrya and Cydonia (Sax 1932; Kovanda 1965). The genome size of genera in Maloideae ranges from 450 to 800 Mbp/haploid genome, which is 2 to 3 times larger than species in the other subfamilies, consistent with their polyploid origin (Dickson et al. 1992). 1. Apple. The genetic linkage maps of apple are listed in Table 4.3. Apple has a long juvenile period and is self-incompatible, and genetic

184

Table 4.2.

Prunus linkage maps of the Prunophora and Cerasus subgenera.

Populationz Prunophora P. cerasifera × Felinem (101) Polonais × Stark Early Orange (SEO) (142)

Type

F1

P. cerasifera

F1

Apricot (Polonais) Apricot (SEO)

Goldrich × Valenciano (81)

F1

SEO × Tyrinthos (76)

F2

Cerasus Napoleon × P. incisa (63) and Napoleon × P. nipponica (47) Emperor Francis (56)

Species

F1

Apricot (Goldrich) Apricot (Valenciano) Apricot

Sweet cherry, P. incisa and P. nipponica

Micro- Sweet cherry sporederived calli

No. markersy

93 110 (212) 141 132 (176) 80 211

Most common marker typesx

SSR (98), SCAR (2) RFLP (45), AFLP (31) RFLP (40), AFLP (38) AFLP (62), RAPD (25) AFLP (60), RAPD (24) AFLP (85), SSR (14)

34

Isozymes (100)

89

RAPD (98), isozyme (2)

No. linkage groups

Total distance (cM)w

Longest gap (cM)

% unlinked lociv

References

8

525

24

1

8

538

33

0

8

699

31

0

8

511j

24

22

7

467j

28

35

11

602j

29

20

Vilanova et al. (2003)

174ru

24ru

28

Bo˘skovi´c and Tobutt (1998)

503

27

3

Stockinger et al. (1996)

7

10

Dirlewanger et al. (2004b) Lambert et al. (2004)

Hurtado et al. (2002)

Rheinische Schattenmorelle (RS) × Erdi Botermo (EB) (86)

F1

Régina × Lapins (133)

F1

z

Sour cherry (RS) Sour cherry (EB)

Sweet cherry (Régina) Sweet cherry (Lapins)

126 (126)

RFLP (100)

95

68 (99) 54

SSRs (100)

19

462j

21

12

16

279j

21

19

11

639

26

1

9

495

30

10

Wang et al. (1998)

Dirlewanger et al. (2004a)

Population size is in parenthesis. Number of mapped markers. For F1 progenies, in parenthesis after the female parent number of markers, total number of markers located on the map. x In parenthesis, percentage of the two most common kinds of molecular markers used for mapping. For F1 progenies this percentage refers to the total number of markers used. w j = maps constructed with JoinMap (van Ooijen and Voorrips, 2002) software; ru = maps constructed with LINKEM (Vowden et al. 1995) and linkage measured in recombination units; the rest were constructed with MapMaker (Lander et al. 1987). Distances are Kosambi. v Proportion of markers assayed not falling within any of the linkage groups. y

185

186 Table 4.3.

Linkage maps of apple and pear.

Populationz Apple Rome Beauty × White Angel (56)

Cultivar name

Rome Beauty White Angel

No. linkage groups

Total distance (cM)w

Longest gap (cM)

% unlinked lociv

References

RAPDs, Isozymes, RFLPs RAPDs, Isozymes, RFLPs

21

~682

~20

8

Hemmat et al. (1994)

No. markersy

Most common marker typesx

156 (427) 253

24

950

~28

2

Wijcik McIntosh (WM) × NY 75441-67 (114),

Wijcik McIntosh

238

RAPDs, Isozymes

21

1,206j

27

11

Conner et al. (1997)

WM × NY 75441-58 (172)

NY 75441-67 NY 75441-58

110 183

RAPDs, Isozymes RAPDs, Isozymes

21 20

692j 898j

24 23

14 4

Conner et al. (1997)

Iduna × A679-2 (95)

Iduna A679-2

65 135

RAPDs (100) RAPDs (100)

9 14

386j 627j

22 22

23 7

Seglias and Gessler (1997)

Prima × Fiesta (152)

Prima

194 (290)

17

842j

24

13+

Maliepaard et al. (1998)

Fiesta

163

RFLPs (48), RAPDs (41) RFLPs (53), RAPDs (33)

17

984j

33

Fiesta × Discovery (267)

Pear Kinchaku × Kosui (82) Bartlett × Housui (63)

z

Fiesta

439 (840)

Discovery

499

Kinchaku Kosui Housui

120 78 180

Bartlett

256

2+

AFLPs (50), SSRs (26) AFLPs (51), SSRs (22)

17

1,144j

26

17

1,455j

26

RAPDs (100) RAPDs (100) AFLPs (61), SSRs (36) AFLPs (70), SSRs (30)

18 22

768j 508j

27 21

10 15

20

995j

23

6

19

1,020j

24

3

Liebhard et al. (2003a)

Iketani et al. (2001)

Yamamoto et al. (2002, 2004a)

Population size is in parenthesis. All populations are F1 segregating progenies. Number of mapped markers. In parenthesis after the female parent number of markers, total number of markers located on the map. x In parenthesis, percentage of the two most common kinds of molecular markers used for mapping referred to the total number of markers used. w j = maps constructed with JoinMap (van Ooijen and Voorrips, 2002) software; the rest were constructed with MapMaker (Lander et al. 1987). Distances are Kosambi. v Proportion of markers assayed not falling within any of the linkage groups. When “+” this proportion was calculated from the total number of markers used in both parents. y

187

188

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

analysis typically is performed on the full-sib progeny of a single cross. The first genetic linkage map of apple was created from a ‘Rome Beauty’ × ‘White Angel’ F1 population combining RAPDs, isozymes, and RFLPs (Hemmat et al. 1994). The linkage map for ‘White Angel’ consisted of 253 markers arranged in 24 linkage groups and extended for 950 cM. The map of ‘Rome Beauty’ consisted of 156 markers in 21 linkage groups. RAPDs were also the predominant markers in the maps constructed by Conner et al. (1997) and Seglias and Gessler (1997). Two saturated maps have been published in apple. The first, reported by Maliepaard et al. (1998), was based on the F1 progeny of the cross between the cultivars ‘Prima’ and ‘Fiesta’ and was constructed using a majority of transferable markers (RFLPs, isozymes, and SSRs). The maps of each parent were well aligned with 67 multi-allelic molecular markers, in which 17 linkage groups were found, putatively corresponding to the basic chromosome number. Scab resistance (Vf ) and rosy leaf curling aphid resistance (Sd1) genes were identified at the bottom of linkage group 1 and at the top of linkage group 7, respectively, in these well-organized reference maps. The fruit acidity locus, Ma, was located at the top of group 16, while the self-incompatibility locus S was found at the bottom of group 17. The second saturated apple map is based on 267 F1 progeny from a cross of ‘Fiesta’ × ‘Discovery’ (Liebhard et al. 2002; Liebhard et al. 2003a). The maps of ‘Fiesta’ and ‘Discovery’, including 115 and 112 SSRs, respectively, could be integrated and anchored by ca. 100 SSR loci, and were aligned to the maps of Maliepaard et al. (1998). The total distance of these maps ranged from 842 to 1455 cM, approximately twice the distance found in Prunus, as expected considering the tetraploid nature of apple. SSR markers are currently the best choice for comparing and aligning different genetic linkage maps within a species and they have been used to align apple maps initially constructed with RAPDs or AFLPs. Hemmat et al. (2003) established the homology between linkage groups of different apple maps of ‘Rome Beauty’, ‘White Angel’, ‘Wijcik McIntosh’, and NY 75441-58 (Hemmat et al. 1994; Conner et al. 1997), by using 41 SSR primer sets. Their maps could also be partially aligned to that of Maliepaard et al. (1998), to which 13 out of 17 linkage groups were anchored. Ten SSR markers were also mapped by Gianfranceschi et al. (1998) in the population used by Seglias and Gessler (1997). Eight of them segregated in both parents, allowing 6 homologous linkage groups to be identified. More than 60 major genes in apple have been identified for pest and disease resistances, and fruit, flower, reproductive, and plant attributes (Brown 1992; Janick et al. 1996; Alston et al. 2000). However, only a

4. SYNTENY IN THE ROSACEAE

189

small number of these phenotypic characteristics have been identified in genetic linkage maps. The use of common SSRs in different populations may help to establish a consensus apple map with which it would be possible to compare the position and function of genes of interest found in different populations. 2. Pear. The genetic linkage maps of pear are also included in Table 4.3. The first molecular linkage maps were constructed in the Japanese pear (Pyrus pyrifolia Nakai) cultivars ‘Kinchaku’ and ‘Kosui’ using their 82 F1 progenies (Iketani et al. 2001). The linkage maps of ‘Kinchaku’ and ‘Kosui’ were constructed only with RAPDs and allowed the detection of 18 and 22 linkage groups, respectively. It is believed that these 2 maps cover at least half of the total pear genome. The resistance to pear scab disease (Vn) and susceptibility to black spot disease (A) were identified in the genetic map of ‘Kinchaku’. Several RAPD markers had significant linkage to pear scab resistance and black spot susceptibility. Genetic linkage maps of the European pear (Pyrus communis L. ‘Bartlett’) and the Japanese pear (P. pyrifolia (Burm.) Nakai ‘Housui’) were constructed using their interspecific F1 progenies (Yamamoto et al. 2002; Yamamoto et al. 2004a). The map of the seed parent, ‘Bartlett’, consisted of 256 loci, distributed on 19 linkage groups. Out of 76 SSRs mapped, 32, 39, and 5 were derived from Pyrus, Malus, and Prunus, respectively. The map of ‘Housui’ contains 180 loci, including 64 SSRs (29 pear, 29 apple, 6 Prunus SSRs) on 20 linkage groups. The two pear maps were aligned using 37 codominant markers with segregating alleles in both parents. These pear maps may cover more than 80% of the total pear genome. C. Subfamily Rosoideae There are three main crops in the Rosoideae subfamily, with a basic chromosome number of x = 7: strawberry, rose, and raspberry. Strawberry is a species of the genus Fragaria, which includes 12 species with different degrees of ploidy, from the diploid wild strawberry (F. vesca) to the octoploid modern garden strawberry (F. × ananassa), synthesised in the middle of the 18th century from the cross between two octoploid wild species, F. chiloensis and F. virginiana (Jones 1976). The edible part of strawberries consists of an enlarged, fleshy fruit receptacle that supports the tiny true fruits (achenes). Diploid strawberry species have the smallest genomes within the cultivated Rosaceae, with 164 Mbp in the F. vesca genome (Akiyama et al. 2001). In spite of its economic importance, the cultivated strawberry is poorly characterised genetically. This is in part because of its complex genetic

190

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

background. It is an octoploid species with 2n = 56, with an unknown genomic composition. Only one map has been published (LecerteauKöhler et al. 2003) based on an F1 progeny between two octoploid lines. This map was made with a large number of AFLPs (789) that coalesced into 58 linkage groups. Based on the presence of linkage groups composed only of single-dose restriction fragments (SDRF) in coupling, and the frequency of multiplex vs. simplex markers (Wu et al. 1992; Da Silva et al. 1993), the authors concluded that the octoploid strawberry has a mixed diploid/polyploid behavior and that at least two of the component genomes are duplicated. One way of simplifying the complexity of the strawberry genome would be to develop detailed maps in diploid relatives and use these maps as references for polyploid map construction, as has been done in other polyploid crops, including alfalfa (Diwan et al. 2000) and potato (Milbourne et al. 1998). Two maps have been constructed so far in the diploid wild strawberry, F. vesca, which seems the ideal organism for this purpose. The first one was with an F2 progeny of the cross between two F. vesca accessions by Davis and Yu (1997) using RAPDs (75), isozymes (3), and morphological characters (2). All these markers mapped to the expected 7 linkage groups covering a total distance of 445 cM. From the nature of most of the markers used (RAPDs), it is unlikely that this map can be used for genome comparisons in Fragaria. A more adequate map for this purpose is the one constructed with 78 markers [68 SSRs, 6 gene-specific markers, one sequence characterized amplified region (SCAR), and three morphological characters] by Sargent et al. (2004), using an interspecific F2 population (F. vesca × F. nubicola). Seventy-six of these markers could be placed on seven linkage groups spanning a distance of 448 cM. Given its high level of polymorphism and the work already done on mapping, this population may become the reference for strawberry in the future. Rose (Rosa spp.) cultivars are a complex of different hybrids between various diploid and tetraploid species with different ploidy levels— diploid, triploid, and tetraploid. The rose achenes are surrounded by the hypanthium (formed by the bottom of the petals, sepals, and stamens stuck together), giving a more or less fleshy, fruit-like structure, called the hip. Rose maps have been elaborated in three populations with the objective of establishing the location of major genes or QTLs responsible for the inheritance of some of the most important characters of flower quality and disease resistance. These maps were constructed almost entirely with dominant markers (RAPDs and AFLPs), which implies that they cannot be compared or used for synteny analysis with other members of the Rosaceae. Two of these maps were obtained in diploid F1 populations (Debener and Mattiesch 1999; Crespel et al. 2002) and one in an F2 between two tetraploid genotypes (Rajapakse et al. 2001).

4. SYNTENY IN THE ROSACEAE

191

Map sizes were generally small, as in other Rosaceae species, ranging from 238 to 370 cM in the diploid genotypes, to 628 to 902 for the tetraploids. The raspberry belongs to the genus Rubus, which also includes other berries, such as dewberries, brambles, and blackberries. The flowers of Rubus are structurally rather similar to those of strawberries; however, in Rubus each carpel develops into a small drupe (drupelet), with the mesocarp becoming fleshy and the endocarp hardening and forming a tiny pit that encloses a single seed. Since there are many carpels per flower, there are many drupelets, and the “fruit” of a blackberry or raspberry is really an aggregate of drupelets. Raspberries are diploid, with a genome of similar size to Prunus (294 Mbp) (Arumuganatan and Earle 1991), but blackberries have different levels of ploidy, from tetraploid to octoploid. A map with 273 markers, including 34 SSRs, was constructed by Graham et al. (2004) in a diploid red raspberry (Rubus idaeus) F1 population. The map included 9 linkage groups, two more than the seven expected, covering a distance of 789 cM. Several QTLs for spiny phenotype and root sucker production were placed on this map, the first produced in Rubus.

III. MAP COMPARISONS A. Within the Prunus Genus Common markers mapped in the reference T×E Prunus and in other Prunus populations have been used to compare their map positions in different species and interspecific hybrids. Table 4.4 summarizes these results for the 11 populations (corresponding to 16 maps) having more than 25 markers in common with T×E. These maps allow comparisons between seven species: peach, almond, apricot, sweet cherry, myrobalan plum, P. ferganensis, and P. davidiana. The part of the genome of T×E covered with these comparisons is on average 57%, with a range of 21–78%. Markers used for synteny analysis are of three kinds: RFLPs and isozymes, both known to be highly transferable across genus and families, and more recently, SSRs. The excellent properties of SSRs and the development of hundreds since the first set reported by Cipriani et al. (1999) have made SSRs the markers of choice for many uses in Prunus genetics and breeding. Although systematic studies on SSR transferability among Prunus species have not been made, when SSRs developed in one species have been used in another, they often have been useful (i.e., give amplified polymorphic DNA fragments of about the

192 Table 4.4.

Comparison of Prunus linkage maps with the reference ‘Texas’ × ‘Earlygold’ (T×E) map.

Map typez

Population

Anchors with T×Ey

% Same group as T×Ex

Noncolinear markersw

% of T×E distancev

% of common distanceu

Paired t test comparison with T×Et

References

Texas × Earlygold

F2

562

100

0

100

100

—

Garfi × Nemared(8)

F2

51

100

0

78

117

3.57**

Jáuregui et al. (2001)

Summergrand × P. davidiana

F1 (P. davidiana)

52

96

2

64

113

0.98

Foulongne et al. (2003)

Summergrand × P. davidiana

F2

57

100

0

70

196

4.62**

Foulongne et al. (2003)

IF7310828 × P. ferganensis

BC1

32

100

1

41

121

1.54

Dettori et al. (2001)

P. cerasifera ×

F1 (P. cerasifera)

43

93

3

53

190

6.32**

Dirlewanger et al. (2004b)

F1 (Felinems)

87

98

2

66

163

5.47**

Dirlewanger et al. (2004b)

F1 (Ferragnès)

Joobeur et al. (2000)

Felinem Ferragnès × Tuono Felisia × Bertina

53 (72)

100

3

64

100

0.01

F1 (Tuono)

41

100

1

48

111

0.76

F1 (Felisia)

32 (43)

100

0

57

111

0.80

F1 (Bertina)

28

100

2

57

112

1.23

Dirlewanger et al. (2004a)

Ballester et al. (2001)

Polonais × Stark Early Orange (SEO)

49 (81)

93

2

63

115

2.19

Lambert et al. (2004)

F1 (SEO)

61

95

4

79

152

3.59**

Lambert et al. (2004)

Ferjalou Jalousia × Fantasia

F2

49

95

2

57

142

4.21**

Etienne et al. (2002)

Akame × Juseitous

F2

45

98

2

52

109

0.79

Yamamoto et al. (unp)

Régina × Lapins

F1 (Régina) F1 (Lapins)

30 28

97 96

1 1

34 21

233 323

3.34** 3.19**

Dirlewanger et al. (2004a)

R

z

F1 (Polonais)

In parenthesis, name of the parent map for F1 segregating populations. Only maps with more than 25 anchor points with T×E have been considered. For F1 progenies, in parenthesis after the female parent marker number, total number of markers studied in the cross. x Percentage of anchor markers located on the same linkage group as T×E. w Number of markers placed on the same linkage group as that of T×E but in different order (generally pairs of markers in inverted order; only one of the two markers is considered). v Percentage of the T×E map covered by the other map: distance of T×E covered*100/total T×E distance. u Relative size of the common maps: common distance covered in a map*100/common distance covered in T×E. t Paired t-test of the comparison between the distances of the two most separated common markers of each linkage group. ** P≤0.01. s For the populations segregating for the G6-G8 reciprocal translocation, the T×E configuration has been taken for map comparisons. y

193

194

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

expected size). This has occurred generally with peach SSRs used in other species, because peach SSRs were developed first and there are more available than in any other Prunus species. For example, 55% and 45% of the peach SSRs used for variability analysis in apricot (Hormaza 2002) or cherry (Dirlewanger et al. 2002), respectively, amplified and were polymorphic in a sample of cultivars of these species. The distribution of markers to different linkage groups and their order within each linkage group in all Prunus species comparisons results in a general pattern of complete synteny between all species, suggesting that the Prunus genome can be treated as a single genetic entity. This result is in agreement with the affinity between different species, which can be crossed within the same subgenus, in some cases between subgenera (Amygdalus and Prunophora), and that these crosses occasionally produce fertile offspring (Scorza and Sherman 1996). Some of the figures of Table 4.4 clearly show this trend, with a percentage of markers located in the same group as T×E, generally of 100 or very close to 100, and a very small number of non-colinear markers among those that are in the same linkage group. Given that RFLP probes frequently detect more than one locus in Prunus (Viruel et al. 1995; Joobeur et al. 1998), it is reasonable to think that most of the markers of one map that are not located in the expected linkage group of T×E correspond to copies of this RFLP that were not mapped (probably because they were not segregating) in T×E (Lambert et al. 2004). Moreover, the markers that have a changed order within a linkage group are frequently contiguous and separated by only a few cM, suggesting that these discrepancies are more likely due to sampling errors, leading to slight differences in locus order, than to actual chromosomal rearrangements between the two compared genomes. The genetic distances between maps were compared with a paired ttest (two-tailed) of the difference between the map distances of the two most separate anchor markers of each linkage group of T×E, and that of the map compared with it. All maps were constructed with Mapmaker EXP/3.0 (Lander et al. 1987) software with the exception of that of Dettori et al. (2001), where JoinMap (van Ooijen and Voorrips 2002) was used instead. JoinMap usually produces shorter maps than those of MapMaker (Liebhard et al. 2003a). This may be due to differences in the calculation of linkage when using the Kosambi mapping function (Stam 1993), which was employed in all maps listed in Table 4.4. Thus, we cannot discard that the paired t-test between T×E and the peach × P. ferganensis BC1 map of Dettori et al. (2001) would have been significant should a Mapmaker version of this map be used. An interesting observation is that some maps are significantly longer than T×E, suggesting that some genotypes have higher rates of meiotic

4. SYNTENY IN THE ROSACEAE

195

recombination than others. In some maps total genetic distances are double or even triple the size of the T×E map. Comparisons between intraspecific maps and interspecific maps have generally produced shorter maps in the latter than in the former (Gebhardt et al. 1991). This can be explained by a decrease in crossing over frequencies between distant genomes, compared to those that occur between chromosomes of the same species. In the Prunus progenies studied, some of the maps significantly longer than T×E were those obtained with intraspecific crosses (peach, one of the apricot parents, and cherry), and the P. cerasifera parent of the P. cerasifera × ‘Felinem’ F1 progeny. Some of the shortest maps were also obtained with interspecific populations, such as T×E and the BC1 of the cross between peach and P. ferganensis. Nevertheless, exceptions were also very important, including all 4 almond maps, the P. davidiana map of the F1 with ‘Summergrand’, the peach ‘Akame’ × ‘Juseitou’ F2, and one of the parents of the apricot cross, which were expected to have longer maps, and the interspecific progenies ‘Garfi’ × ‘Nemared’ and ‘Felinem’ (a seedling of ‘Garfi’ × ‘Nemared’) and the F2 progeny of ‘Summergrand’ × P. davidiana, which were also expected to produce longer maps. De Vicente and Tanksley (1991) found significant differences in recombination rates between male and female gametes in tomato. If this happened in Prunus, the maps obtained with male and female individuals in F1 progenies would have different sizes. This was not the case in four of the five progenies studied, in all but the apricot F1 ‘Polonais’ × ‘Stark Early Orange’, but in this case the female parent had a lower level of recombination, whereas in tomato the reduction of recombination occurred in the male gametes. The observed pattern seems to fit with a model in which recombination rates would be associated to specific genotypes, more than to the distance between genomes or to sex. In the simple hypothesis that one or a few genes could be responsible for the level of recombination of a certain individual, our results agree with a situation in which the alleles that increase recombination of these genes are absent or at a very low frequency in almond and present in cherry. Apricot and peach would be polymorphic for these genes, and each individual would behave according to its genetic composition. Given that recombination is one of the driving forces of plant breeding, characterization of its intensity in different genotypes may be an important additional element in the selection of parents for specific selection purposes. The pattern of synteny observed in the species of the Prunus genus has an exception. A reciprocal translocation was detected by Jáuregui et al. (2001), between ‘Nemared’ peach and ‘Garfi’ almond. Only seven linkage groups could be obtained in the map of the ‘Garfi’ × ‘Nemared’ F2

196

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

population, and one of these groups included markers placed on groups 6 and 8 of T×E. The hypothesis of pseudolinkage between markers of these two chromosomes due to a reciprocal translocation was validated with cytogenetic studies and segregation of the semi-sterile and fully fertile individuals in the progeny. Although these data did not establish which of the parents (‘Garfi’ or ‘Nemared’) carried the translocation, indirect evidence points to ‘Nemared’ as the most probable translocated genotype (Jáuregui et al. 2001). ‘Nemared’ is a red-leaved genotype and the breakpoint of the translocation was located in the same region of the Gr/gr gene that determines the anthocyanin coloration of leaves, suggesting that there may be a relationship between the cytogenetic configuration and the morphology. Given that the ‘standard’ chromosome configuration has been found in the rest of the almond and peach genotypes studied so far, it can be concluded that this chromosomal rearrangement is not characteristic of peach, but only present in some of its germplasm. The same reciprocal translocation was found in the map of ‘Felinem’ (Dirlewanger et al. 2004b), an expected result given that this rootstock is an offspring of ‘Garfi’ × ‘Nemared’, and in the ‘Akame’ × ‘Juseitou’ peach F2 (Yamamoto et al. 2001a; Yamamoto, pers. comm.). Given that ‘Akame’ is a red-leaved genotype, this indicates that the translocation may occur in the group of red-leaved peaches. There are two more maps constructed with one red-leaved parent, the F2s NC174RL × ‘Pillar’ and ‘Lovell’ × ‘Nemared’ used by Chaparro et al. (1994) and Lu et al. (1998). Both maps are difficult to compare with others because they were constructed with dominant markers, but the addition of a few SSR markers of linkage groups 6 and 8 would easily demonstrate the presence of the translocation. The general synteny between the genomes of Prunus and the existence of a network of maps anchored with T×E allows positioning of all markers, genes, or QTLs obtained in these maps in a common “consensus” map. Using the data available, it was possible to establish the position of 28 major genes obtained in different species on a single map (Dirlewanger et al. 2004a). These major genes included 19 genes from peach, 6 from almond or almond × peach crosses, 2 from apricot, and one from Myrobolan plum. A large number of QTLs from different Prunus progenies, affecting characters such as disease resistance, fruit quality, blooming or fruit maturity times or tree architecture, have been identified (Testolin 2003). Those that are located in the populations with maps anchored with T×E (Table 4.4) could be positioned in the Prunus “consensus” map using the same approach. Given the high marker density of T×E and the possibility of finding additional markers

4. SYNTENY IN THE ROSACEAE

197

for specific regions using the information of homologous regions of other maps, there is a high probability of finding markers sufficiently close to these genes or QTLs to be used in marker-assisted selection. B. Between Apple and Pear Although apple and pear belong to different genera, it is believed that they are genetically very close to each other because there are natural inter-generic hybrids between them (Sax 1931; Weber 1964). In addition, a relatively large number of natural inter-generic hybrids have been found in the Maloideae, including Pyrus × Cydonia, Malus × Cydonia, Pyrus × Sorbus, Amelanchier × Sorbus, and Crataegus × Mespilus (Sax 1931; Weber 1964). Yamamoto et al. (2001b) reported that SSR markers developed in apple produced discrete amplified fragments in several Pyrus genotypes, indicating that apple SSRs could be applicable in pear. Nucleotide repeats were detected in amplified fragments in pear, and the DNA sequence of flanking regions in apple was highly conserved in pear, indicating that SSR markers are good tools to compare genetic linkage maps obtained from different species as well as different genera of the Maloideae. Liebhard et al. (2002) noted that apple SSRs successfully amplified in species of other Maloideae genera (Amelanchier, Cotoneaster, Crataegus, Cydonia, Mespilus, Pyrus, and Sorbus). The SSR markers developed in apple and pear have been utilized as a reliable tool for identifying quince varieties (Yamamoto et al. 2004b). Thirtynine and 29 SSR markers derived from apple produced segregating loci in the genetic maps of ‘Bartlett’ and ‘Housui’, respectively (Yamamoto et al. 2004a). Sixty-six apple SSRs were also mapped in a genetic map of the European pear cultivar ‘La France’ constructed from the three-way interspecific hybrid progeny of ‘Shinsei’ × 282-12 (‘Housui’ × ‘La France’) (T. Yamamoto, unpubl. data). When the pear maps of ‘Bartlett’, ‘Housui’, and ‘La France’ were compared with those of ‘Fiesta’ and ‘Discovery’, all pear linkage groups could be successfully aligned to the apple reference map by at least one apple SSR, suggesting that positions and linkages of SSR loci were well conserved between pear and apple. As Fig. 4.1 shows, on linkage groups 10, 12, 14, and 15, there are 8, 7, 7, and 5 SSRs, respectively, located in both maps of apple and pear. Positions, order, and linkage of SSR loci found in genetic maps of apple and pear were almost completely conserved in these 4 linkage groups. The self-incompatibility locus (S locus) was mapped to linkage group 17 in Japanese and European pears as well as in apple (Yamamoto et al. 2004a). In all, these results indicate a high level of conservation between

198

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

L10A L10P

L12A

L12P

L15A

L15P

CH05d04 CH05d04

50 cM

CH02a08 CH02a08 CH01f12 CH02c11 CH01f12

CH04g04 CH05g07 CH05d11 CH04d02 CH01f02 CH03c02

CH04g04 CH05g07 CH05d11 CH04d02 CH01f02 CH03c02

NZ02b1 CH05a02

NZ02b1 CH05a02

CH01d08 CH01d08

CH02c11 CH04g09

CH01f07a CH02b03b MS02a01 MS06g03

CH04g09

CH02d11 CH02d11

CH01f07a CH02b03b MS02a01 MS06g03

L14P L14A

CH02c09 CH02c09

CH05g07 CH01g05 CH04f06 CH03d08 CH05g11 CH04c07 CH05d03

CH05g07

CH01g05 CH04f06 CH03d08 CH05g11 CH04c07 CH05d03

Fig. 4.1. Comparison between linkage groups 10, 12, 14, and 15 of apple and pear. Groups are labelled with L followed by the group number and “A” for apple or “P” for pear. Linkage groups of apple are from the genetic map of ‘Discovery’ described in Liebhard et al. (2003a). Linkage groups of pear are from an integrated map of ‘Bartlett’, ‘Housui’, and ‘La France’ (Yamamoto et al. 2004a, unpublished data). Anchor SSRs are indicated by the dotted lines.

the pear and apple genomes. This information may be applicable to other Maloideae genera and will help to advance genome research in lesser-known fruit tree species such as quince, loquat, and medlar. About 10–20% of SSR markers in apple and pear are multi-locus (Liebhard et al. 2002; Liebhard et al. 2003a; Yamamoto et al. 2004a). Maliepaard et al. (1998) identified several homeologous linkage groups of apple using the markers detected by duplicated RFLPs. Similarly, Liebhard et al. (2002, 2003a) pointed out duplication patterns of multilocus SSRs in the linkage group pairs 1-3, 1-7, 4-12, 5-10, 8-15, 9-17, 1213, and 12-14 of apple. In pear, duplication of the linkage group pairs 1-3, 2-15, 5-10, 8-15, 9-17, 10-17, 12-14, and 13-16 were revealed by multi-locus SSRs (Yamamoto et al. 2004a). Duplications of linkage

4. SYNTENY IN THE ROSACEAE

199

groups 1-3, 5-10, 8-15, 9-17, and 12-14 were observed in both apple and pear. C. Between Apple and Prunus Data for the comparison of these two genomes are limited to the 30 common loci (24 RFLPs and 6 isozymes) between the Prunus T×E and the apple ‘Prima’ × ‘Fiesta’ maps (Dirlewanger et al. 2004a). A comparison between the three Prunus linkage groups having three or more anchor markers with apple linkage groups is shown in Fig. 4.2. Four

G1 L4

MC013 AG53

L13

L16

G3

AG53a

MC115 MC041b

MC041a

L9

L17

MC115b

MC115a 10 cM

Aat-1

(Aat-2) Me-1

MC038b MC038a

L2

MC038A MC007

G4

LY5b

MC001

L15

Pgm-2 LY21

(Aat-1) Lap-2 LY37, MC022 LY16A

L10

MC028b

MC001b

LY5B

L1, L7

L5

MC028B MC007a

MC028a

MC030A

MC030b

MC024A

MC024a

MC030a

MC001a LY5a

L8 LY21a

Aat-2

Lap-2, MC022 LY16a

MC024b

Fig. 4.2. Comparison between Prunus (Joobeur et al. 1998) and apple (Maliepaard et al. 1998) linkage maps. Only the position of anchor loci is shown. Linkage groups in Prunus are labelled G, and apple groups L, followed by a number. The positions of markers in parentheses in Prunus were inferred from other maps. Marker positions in apple were obtained using the maps of both parents of the F1 cross ‘Prima’ × ‘Fiesta’. Two parallel oblique lines indicate that only a fragment of the linkage group is included. Arrows pointing to the left in the Prunus map are anchors to markers located in the indicated linkage groups of the apple map. Reproduced with permission from Dirlewanger et al. (2004) Comparative mapping and marker assisted selection in Rosaceae fruit crops. Proc. Natl. Acad. Sci. (USA) 101:9891–9896 (copyright (2004) National Academy of Sciences, U.S.A.).

200

P. ARÚS, T. YAMAMOTO, E. DIRLEWANGER, AND A. ABBOTT

more Prunus groups had anchor points with the apple map, three of them with two markers each: G2, corresponding to apple groups L3 and L11, G8, to L5, L3 and L10 and G7 to L14, and one with one marker, G5, which had its homologue in apple group L6 (Joobeur 1998). For the linkage groups with more common markers, the position of most loci appears colinear and the distances between contiguous loci are similar, suggesting that the synteny between the chromosomes compared is important. G3 and G7 are homologous to two apple linkage groups each, as expected if they correspond to two homeologous chromosomes. The comparison of G1 is especially interesting because its upper part corresponds to two homeologous apple groups (L13 and L16), whereas the lower part is syntenic to one more apple group (L8), suggesting that either two of the chromosomes of the ancestor species of apple and Prunus fused in the Prunus lineage, or one chromosome of this ancestor split into two in the Malus lineage. This agrees with the cytogenetic observation that apple does not have the large chromosome observed in Prunus (Bouvier et al. 2000). Recent molecular genetic studies refuted the hypothesis that the allopolyploid genome of the Maloideae (x = 17) included one Prunoideae (x = 8) and one Spiraeoideae (x = 9) genome, but supported autopolyploidy or hybridization between closely related members of a single lineage, with species of the Spiraeoideae subfamily being the most probable parental lineages (Morgan et al. 1994; Evans and Campbell 2002). Phylogenetic analysis of the rbcL gene sequence does not provide close links between Maloideae and Prunoideae but between Maloideae and some genera of Spiraeoideae (Morgan et al. 1994). Based on GBSSI (granule-bound starch synthase, waxy) genes, Evans and Campbell (2002) showed that the subfamily Maloideae originated from polyploidy involving only members of a lineage that contained the ancestors of Gillenia, of the Spiraeoideae. If one of the differences between the genomes of Prunoideae and Spiraeoideae is that the long chromosome of the former corresponds to two in the latter, our observation that the long G1 group of Prunus appears to be split into two in apple is in agreement with the origin of the Maloideae genome being composed of two Spiroideae genomes without the inclusion of a Prunoideae genome. The level of transferability of SSRs between Maloideae and Prunoideae was rather low, suggesting that these markers are inadequate for map comparisons between species of these two subfamilies. Cipriani et al. (1999) found that only 18% of peach SSRs were amplified in apple. Yamamoto et al. (2004a) noted that only about one-tenth of the Prunus SSRs could be mapped in maps of ‘Bartlett’ and ‘Housui’. Only one out

4. SYNTENY IN THE ROSACEAE

201

of 15 apple SSR markers was transferable to Prunoideae (Liebhard et al. 2002). D. Between Prunus and Arabidopsis Most of the probes used for RFLP mapping in the Prunus T×E reference map have been obtained with probes of known sequence. For some of the RFLPs of this map (detecting 111 loci), Arabidopsis probes with a high level of homology with rice sequences were used, and for others we used probes mostly from almond, peach, and other Rosaceae DNA libraries, some of which (detecting 116 additional loci) have a high level of sequence homology (TBLASTX value 25%) were detected for all traits; 61% of them were detected in two or more experiments. QTLs for fruit shape were detected in more trials than QTLs for fruit weight and sugar content, confirming that fruit shape was under highly heritable polygenic control, similar to what was reported by Périn et al. (2002b). External color segregated as yellow:green in both experimental populations. Flesh color segregated as white: green:orange. The locus responsible for the green flesh color was mapped on linkage group 1, and corresponded to the locus gf. The genetic

232

Y. WANG, R. DEAN, AND T. JOOBEUR

control of orange flesh color was complex: two loci in linkage groups 2 and 12 were associated with orange flesh. QTLs conditioning disease resistance, in addition to fruit traits, were also mapped. Dogimont et al. (2000) analyzed 122 RILs from Védrantais × PI 161375 for resistance to cucumber mosaic cucumovirus (CMV) in melon. Dates of appearance and intensity of symptoms were recorded upon infection with three CMV strains: two strains from Subgroup I (TL and P9) and one from Subgroup II (To 72). A molecular map was constructed from the same population and was used for detecting QTLs. Seven genomic regions were found to carry factors for resistance to CMV. One of them, located in linkage group XII and accounting for a large amount of the phenotypic variation, was common to the three CMV strains (Table 5.3). C. Watermelon QTL mapping in watermelon has also concentrated on fruit traits. Fan et al. (2000) evaluated F2’s from 97103 × PI 296341 in a nonreplicated greenhouse trial for total soluble solids content, rind hardness, fruit, and seed weight. 97103 had higher soluble solids content and a thin rind but was susceptible to fusarium wilt, while PI 296341 was a wild germplasm with lower soluble solids content, a thick rind, and was resistant to fusarium wilt (race 1). Using mostly RAPD and SSR markers, they detected four QTLs for soluble solids content, five for rind hardness, two for rind thickness, three for fruit weight, and six for seed weight (Table 5.3). Total soluble solid QTLs qSSC-2 and qSSC-4 showed overdominance; qSSC6 was close to Wf, a gene conferring red flesh color in watermelon. This was confirmed by Hashizume et al. (2003), who also found that loci for red flesh color and sugar content are very close in their map. Rind hardiness QTL qHR and sugar content QTL qSSC are close to each other in linkage groups 2 and 4 and overlap in linkage group 6 next to Wf. Hashizume et al. (2003) analyzed 60 BC individuals from H-7 × SA-1 for rind hardness, soluble solids, flesh color (red and yellow), and rind color. Five QTLs were detected for the four traits (Table 5.3). D. Squash Genetic mapping in squash lags behind that in the other cucurbits. QTL mapping conducted by Brown and Myers (2002) defined regions in their map for fruit shape and leaf indentation using RAPD markers. This has been the only recent effort to map the squash genome for either qualitative or quantitative traits. Other marker systems such as AFLP and SSR

5. GENETIC MAPPING AND MOLECULAR BREEDING IN CUCURBITS

233

have been tested in squash (Brown and Myers 2000; Katzir et al. 2000; Paris et al. 2003) and may be incorporated into future mapping efforts.

VI. MOLECULAR BREEDING Markers linked to a trait of interest are very powerful selection tools in cucurbit breeding, especially in selection for disease resistance, as has been demonstrated by a number of reports discussed below. However, before a marker can be effectively used in a breeding program, it needs to be converted in most cases to PCR-based user-friendly markers amenable to high-throughput production and tested for its robustness in different germplasms. A. Marker Conversion Most traits in cucurbits are mapped with RFLP, RAPD, or AFLP markers. These markers are not very cost-efficient to use in molecular breeding. RFLP and AFLP are both time-consuming and involve use of radioactive materials, whereas RAPD is notoriously inconsistent. Converting these to PCR-based dominant or co-dominant SCAR or STS markers is appropriate if they are to be used in molecular breeding. The conversion process often involves sequencing the marker DNA from both parents, identifying polymorphic segments, and designing PCR primers according to the sequence difference. Horejsi et al. (1999) attempted to convert 75 RAPD markers to SCAR markers and successfully converted 48 of them, of which only 11 reproduced the polymorphism detected by original RAPD markers. This may be due to the short length of SCAR primers designed in their studies (16–22 nucleotides), which may have amplified non-specific fragments. Primers longer than 24 nucleotides tend to be more robust and stable under different PCR conditions (Wechter et al. 1995; Wechter et al. 1998). Xu et al. (2000) converted a RAPD marker OPP01/700, which is linked to Fo-1 in watermelon at a distance of 1.3 cM, to a SCAR marker SCP01/700 (24 nucleotide-primer), which unlike its RAPD counterpart only amplified one band. For the RAPD marker 596 linked to melon Fom-2 (Wechter et al. 1995), Wechter et al. (1998) designed two 24-mer primers, MUSKFOM I and MUSKFOM II, which only amplified a product from resistant cultivars, hence produced a dominant SCAR marker. Subsequently, marker 596 and an AFLP marker were found to co-segregate with Fom-2 and both markers were converted to co-dominant PCR markers that amplified bands

234

Y. WANG, R. DEAN, AND T. JOOBEUR

from both parents and were named FM and AM (Wang et al. 2000). The advantage of this in molecular breeding is that a failed PCR reaction can be easily diagnosed, unlike dominant markers. These two markers were later validated on a diverse collection of melon germplasm by Burger et al. (2003) (see VI B. Germplasm Screening). Additional PCR-based markers were later developed using end sequences of BAC clones identified using AM and FM markers as probes (Joobeur et al. 2004). Zheng et al. (1999) converted RAPD markers E07 and G17 linked to Fom-2 in repulsion (Baudarcco-Arnas and Pitrat 1996) to CAPS markers, i.e., amplified PCR products from both parents were sequenced to identify their sequence polymorphism and digested with selected restriction enzymes to produce the CAPS marker. These markers were used to evaluate different melon germplasms (Zheng et al. 1999) (see VI B. Germplasm Screening). B. Germplasm Screening Germplasm screening provides additional test for robustness of molecular markers. Using the co-dominant FM and AM markers linked to Fom-2, Wang et al. (2000) surveyed 45 melon genotypes from different sources. They found that FM correctly predicted disease phenotypes from 37 genotypes and AM from 41 genotypes. A test of 48 commercial RILs showed that both markers predicted all lines correctly (Y.-H. Wang, J. J. King and, R. A. Dean, unpubl. data). Burger et al. (2003) tested 24 genotypes with FM and AM markers and found that FM correctly predicted disease phenotype in 22, while AM correctly predicted phenotypes for all of them and differentiated heterozygotes from homozygotes, indicating that these markers were suitable for use in molecular breeding. The same 24 melon genotypes were also used to evaluate the CAPS marker developed from RAPD marker E07, which is linked to Fom-2 at a distance of 1.6 cM (Baudarcco-Arnas and Pitrat 1996) by Zheng et al. (1999). The CAPS marker detected homozygous resistant melon lines but failed to differentiate between heterozygous resistant and susceptible lines (Burger et al. 2003). Zheng and Wolff (2000a) evaluated 48 resistant and 41 susceptible melon lines with RAPD markers E07, G17 (Baudarcco-Arnas and Pitrat 1996), and 596 (the RAPD marker from which co-dominant FM was derived) (Wechter et al. 1995) that were tightly linked to Fom-2. The correlation between molecular and phenotypic data was quite low: E07 and G17 correctly predicted disease phenotypes for only 88% and 81% of all genotypes, respectively, while 596 only matched 70% (Zheng and Wolff 2000a). An explanation for this result may be that RAPD markers were used in this study. The CAPS developed from E07 SCAR marker

5. GENETIC MAPPING AND MOLECULAR BREEDING IN CUCURBITS

235

correctly matched 81 disease phenotypes out of 90 tested (Zheng et al. 1999). The authors also pointed out that scoring CAPS may be complicated by incomplete digestion of PCR products, a problem also encountered by Burger et al. (2003) with CAPS markers. Although not genetically mapped, RFLP markers based on ACC oxidase or ACC synthase were used by Zheng and Wolff (2000b) to evaluate the association between these markers and ethylene production in 63 melon genotypes. They found that ACC oxidase (RFLP-MEL1) allele A0 was associated with low ethylene production and green/white flesh color. However, RFLP-MEL1 allele B0, when homozygous, was associated with high ethylene production and orange flesh color. An RFLP marker from ACC synthase (RFLP-MEACS1) revealed five fragments (named A, B, C, D, and E, from large to small in sizes) when digested with Nde I. Two-fragment pattern (AB) and three-fragment pattern (ACE) were associated with low and high ethylene production, respectively. ACE was also associated with orange flesh color, similar to homozygous B0 (Zheng and Wolff 2000b). C. Marker Assisted Selection (MAS) Molecular markers tightly linked to monogenic traits have been shown to be effective in molecular breeding. Wang et al. (2000) and Burger et al. (2003) have provided evidence that melon Fom-2 markers are robust and can be used for phenotypic screening in a breeding program. In fact, most markers linked to monogenic traits reported in this review can be used in MAS. MAS with QTLs is not as efficient because markers tightly linked to a QTL are not easily defined. But Fazio et al. (2003a) indicated otherwise. Based on markers linked to MLB QTLs, they compared MAS with phenotypic selection in BC1 and BC2 plants and found no difference between the two, suggesting that MAS is at least as efficient as traditional selection methods for QTL selection (Fazio et al. 2003a). To date, this has been the only report using molecular markers as a selection tool for QTLs in cucurbits.

VII. GENE CLONING The first gene cloned in cucurbits by map-based cloning strategy was the disease resistance gene. Although extensive efforts have been put into cloning disease genes from many plant species, such efforts have just begun in earnest in cucurbits. Taler et al. (2004) reported cloning of two partially dominant complementary genes (At1 and At2) that confer a broad-spectrum resistance to downy mildew in melon. They identified

236

Y. WANG, R. DEAN, AND T. JOOBEUR

these genes using an isolated protein P45 associated with the resistance to downy mildew (Balass et al. 1992). P45 is a peroxisomal aminotransferase enzyme. Transferring this protein into a susceptible melon line turned it into highly downy mildew resistant. The first gene in cucurbit cloned by the map-based cloning approach is melon Fom2 (Joobeur et al. 2004). Unlike At1 and At2, Fom-2 confers a racespecific resistance. Map-based cloning of nsv is in progress (Morales et al. 2003). To clone Fom-2, Joobeur et al. (2004) further mapped the gene using markers presented in Wang et al. (2000) in 159 RILs derived from ‘Védrantais’ and PI 161375. SSR154 (FM) and STS178 (AM) flanked Fom-2 at 0.7 and 0.6 cM, respectively. Screening a melon BAC library (Luo et al. 2001) with ACT/CAT1 marker (Wang et al. 2000) identified 23 BAC clones that belonged to the same contig. One SSR marker SSR138 derived from BAC end sequence was found to co-segregate with Fom-2. Chromosome walking was initiated with two flanking (SSR154 and STS178) and one co-segregating (SSR138) markers. Two BAC clones encompassing Fom-2 were identified and sequenced. Fom-2 position was then delimited by two PCR based markers, STS411 and STS296, developed using the BAC clones sequences. Three putative genes were found in this 75 kb interval and only two were found to be complete. One of the two shared similarity to Arabidopsis AtCPSF73-II, which was primarily expressed in flower tissue (Xu et al. 2004). The other putative gene shared significant similarity to the previously characterized NBSLRR class of resistance genes such as I2 in tomato and was thus designated as Fom-2. The characterization of Fom-2 and its physical interval will facilitate the identification of useful markers for MAS in different breeding material. Markers such as SSR138 and SSR430 were used to screen resistant (MR-1 and PI 161375) and susceptible material (Védrantais, AY, and Durango) and showed the expected association between resistance phenotype and markers bands (Joobeur et al. 2004). Screening of additional cultivars may facilitate the identification of Fom-2 co-dominant marker, which will be the ultimate molecular selection tool for Fom-2 genotypes. Fom-2 can also be transferred to other cucurbits to see if heterologous expression of this gene affects host resistance.

VIII. CUCURBIT GENOMICS Among the molecular tools developed for cucurbit crops, mapping tools such as DNA markers, especially SSRs, should be mentioned. Develop-

5. GENETIC MAPPING AND MOLECULAR BREEDING IN CUCURBITS

237

ing SSRs requires a heavy up-front investment because genomic sequences have to be determined before primers can be designed and used for actual genetic mapping. However, once they are identified and tested, SSRs constitute good quality markers. Melon SSR markers have been largely developed by Katzir et al. (1996), Danin-Poleg et al. (2001), Chiba et al. (2003), and Ritschel et al. (2004), while SSRs in cucumber and watermelon have been studied by Fazio et al. (2002) and Jarret et al. (1997), respectively. SNP markers are also being developed for melon by Morales et al. (2004) and have been used in cucumber mapping (Fazio et al. 2003b). Work on cucurbit genomics has also begun. Cucurbit large-insert genomic libraries are available in melon (Luo et al. 2001; van Leeuwen et al. 2003) and watermelon. EST projects have also been initiated. Several groups are generating ESTs in cucumber as listed in GenBank. In melon, Garcia-Mas et al. (2002) produced 1085 ESTs from three cDNA libraries constructed of seedling and immature fruits. Other EST projects have just started in melon. These ESTs will be useful in genetic mapping, comparative genome analysis, gene discovery, and construction of genome array for molecular breeding. Yariv et al. (2002) constructed a melon macroarray consisting of 5000 cDNA clones from four libraries of cultivar ‘Noy Yizre’el’, a climacteric Ha’Ogen-type melon. They found 616 differentially expressed clones from mature and 384 from young fruits. Some of the 59 clones subjected to RNA gel analysis were found to be specifically expressed in the fruit. These include MEL 7 major latex protein, acetoacetyl coenzyme-A thiolase, expansin, MEL2 alcohol acetyltransferase, and jasmonic acid regulatory protein and some unknown proteins (Yariv et al. 2002). This represents a significant step forward for genomic analysis of cucurbits.

IX. FUTURE PROSPECTS Cucurbit mapping did not start in earnest until the 1990s when genetic mapping of other crop plants was already well established. New marker technologies such as AFLP and SSR has greatly facilitated cucurbit mapping in recent years. With high-density maps now available, efforts are being focused on comparative mapping of different cucurbit genomes. Cucurbits are very similar to each other at the DNA level, which makes comparative mapping more useful for gene identification and marker development. Homology at the DNA level is 78% between cucumber and watermelon, 84% between watermelon and melon, and 90% between cucumber and melon (Pasha and Sen 1998). In fact, SCAR primers (Wang

238

Y. WANG, R. DEAN, AND T. JOOBEUR

et al. 2000) designed for melon amplified a product of similar size and sequence in cucumber (Wang 1999). Danin-Poleg et al. (2001) compared six PCR fragments amplified by SSR primers and found that the sequences between cucumber and melon share 76–94.8% homology. Chiba et al. (2003) tested 31 melon SSR markers and found that 20 of them can be used with cucumber. Ritschel et al. (2004) found that 33 of 67 melon SSR markers amplified cucumber and 16 amplified watermelon genomic DNA. Thus, SSR markers from melon and cucumber were used to cross-map the two genomes (Danin-Poleg et al. 2000). In addition, cucumber cDNA has been routinely used to probe melon genomes for phylogenetic studies and genetic mapping (Neuhausen 1992; Silberstein et al. 1999; Brotman et al. 2000; Klingler et al. 2001; Oliver et al. 2001; Silberstein et al. 2003). In order to utilize information generated from model plant systems, a comparison of melon genome with that of Arabidopsis thaliana has been performed by Monfort et al. (2004) and van Leeuwen et al. (2003). Monfort et al. (2004) mapped 204 melon ESTs and detected macrosynteny between 30 regions of Arabidopsis (16% of whole genome) corresponding to 22 melon regions (20% of melon genome). In fact, the two species are close enough that Arabidopsis ESTs can be used to map the melon genome (Oliver et al. 2001). Currently, a comprehensive comparative map between melon and watermelon, based on SSR markers, is being constructed (T. Joobuer and R. A. Dean, unpubl. data). Conserved orthologous sequences (COS markers) are also being used to anchor this linkage map to Arabidopsis and tomato genomes. These comparative maps will allow the transfer of valuable information between these species. For example, it should be possible to find cucurbit homologs for the tomato fruit size and shapes that are under extensive study in this species. EST projects underway in several labs will facilitate the discovery of important novel genes and the saturation of these comparative maps. These mapped ESTs may constitute an excellent source of markers for the important trait mapped so far in cucurbits and be exploited as markers in molecular breeding. One of the most important breeding objectives in cucurbits is the development of disease-resistant cultivars. Fortunately, a number of known resistance genes in cucurbits are mapped and readily adapted for molecular breeding. Other genes that can be further explored for commercial values should also be focused on, such as genes conditioning fruit quality and sex expression (gynoecy and parthenocarpy). While mapping of fruit traits has been initiated, more studies need to be conducted on sex expression that could be used in commercial hybrid production, critical for cucurbit growers.

5. GENETIC MAPPING AND MOLECULAR BREEDING IN CUCURBITS

239

LITERATURE CITED Andres, T. C. 2004. Web site for the plant family Cucurbitaceae & home of The Cucurbit Network. http://www.cucurbit.org/family.html. Arumuganathan, K., and E. D. Earle. 1991. Nuclear DNA content of some important plant species. Plant Mol. Biol. Rep. 9:211–215. Balass, M., Y. Cohen, and M. Bar-Joseph. 1992. Identification of a constitutive 45 kD soluble protein associated with resistance to downy mildew in muskmelon (Cucumis melo L.) line PI 124111F. Physiol. Mol. Plant Pathol. 41:387–396. Baudracco-Arnas, S., and M. Pitrat. 1996. A genetic map of melon (Cucumis melo L.) with RFLP, RAPD, isozyme, disease resistance and morphological markers. Theor. Appl. Genet. 93:57–64. Bradeen, J. M., J. E. Staub, C. Wye, R. Antonise, and J. Peleman. 2001. Towards an expanded and integrated linkage map of cucumber (Cucumis sativus L.). Genome 44:111–119. Brotman, Y., L. Silberstein, I. Kovalski, J. Klingler, G. Thompson, N. Katzir, and R. PerlTreves. 2000. Linkage groups of Cucumis melo, including resistance gene homologues and known genes. Acta Hort. 510:441–448. Brotman, Y., L. Silberstein, I. Kovalski, C. Perin, C. Dogimont, M. Pitrat, J. Klingler, G. A. Thompson, and R. Perl-Treves. 2002. Resistance gene homologues in melon are linked to genetic loci conferring disease and pest resistance. Theor. Appl. Genet. 104:1055– 1063. Brown, R. N., and J. R. Myers. 2000. Searching for molecular markers linked to ZYMV resistance in squash. Cucurbit Genet. Coop. Rep. 23:69–70. Brown, R. N., and J. R. Myers. 2002. A genetic map of squash (Cucurbita sp.) with randomly amplified polymorphic DNA markers and morphological markers. J. Am. Soc. Hort. Sci. 127:568–575. Burger, Y., N. Katzir, G. Tzuri, V. Portnoy, U. Saar, S. Shriber, R. Perl-Treves, and R. Cohen. 2003. Variation in the response of melon genotypes to Fusarium oxysporum f.sp. melonis race 1 determined by inoculation tests and molecular markers. Plant Path. 52:204–211. Chiba, N., K. Suwabe, T. Nunome, and M. Hirai. 2003. Development of microsatellite markers in melon (Cucumis melo L.) and their application to major cucurbit crops. Breed Sci. 53:21–27. Danin-Poleg, Y., N. Reis, S. Baudracco-Arnas, M. Pitrat, J. E. Staub, M. Oliver, P. Arus, C. M. deVicente, and N. Katzir. 2000. Simple sequence repeats in Cucumis mapping and map merging. Genome 43:963–974. Danin-Poleg, Y., N. Reis, G. Tzuri, and N. Katzir. 2001. Development and characterization of microsatellite markers in Cucumis. Theor. Appl. Genet. 102:61–72. Danin-Poleg, Y., Y. Tadmor, G. Tzuri, N. Reis, J. Hirschberg, and N. Katzir. 2002. Construction of a genetic map of melon with molecular markers and horticultural traits, and localization of genes associated with ZYMV resistance. Euphytica 125:373–384. Dijkhuizen, A., and J. E. Staub. 2002. QTL conditioning yield and fruit quality traits in cucumber (Cucumis sativus L.): effects of environment and genetic background. J. New Seeds 4:1–30. Dogimont, C., L. Leconte, C. Perin, A. Thabuis, H. Lecoq, and M. Pitrat. 2000. Identification of QTLs contributing to resistance to different strains of cucumber mosaic cucumovirus in melon. Acta Hort. 510:391–398. Esquinas, J. T. 1981. Alloenzyme variation and relationships among Spanish land-races of Cucumis melo L. Kulturpflanze 29:337–352.

240

Y. WANG, R. DEAN, AND T. JOOBEUR

Fan, M., Y. Xu, H.-Y. Zhang, H.-Z. Ren, G.-B. Kang, Y.-J. Wang, and H. Chen. 2000. Identification of quantitative trait loci associated with fruit traits in watermelon (Citrullus lanatus) and analysis of their genetic effects (in Chinese with English summary). Acta Genet. Sinica 27:902–910. Fanourakis, N. E., and P. W. Simon. 1987. Analysis of genetic linkage in cucumber. J. Hered. 78:238–242. FAO, FAOSTAT (2004) http://faostat.fao.org/faostat/. Accessed June 2004. Fazio, G., J. E. Staub, and S. M. Chung. 2002. Development and characterization of PCR markers in cucumber. J. Am. Soc. Hort. Sci. 127:545–557. Fazio, G., S. M. Chung, and J. E. Staub. 2003a. Comparative analysis of response to phenotypic and marker-assisted selection for multiple lateral branching in cucumber (Cucumis sativus L.). Theor. Appl. Genet. 107:875–883. Fazio, G., J. E. Staub, and M. R. Stevens. 2003b. Genetic mapping and QTL analysis of horticultural traits in cucumber (Cucumis sativus L.) using recombinant inbred lines. Theor. Appl. Genet. 107:864–874. Fehér, T. 1993. Watermelon Citrullus lanatus (Thunb.) Matsum. & Nakai. p. 295–311. In: G. Kalloo and B. O. Bergh (eds.), Genetic improvement of vegetable crops. Pergamon Press, New York. Fukino, N., M. Taneishi, T. Saito, T. Nishijima, and M. Hirai. 2002. Construction of a linkage map and genetic analysis for resistance to cotton aphid and powdery mildew in melon. Acta Hort. 588:283–286. Garcia-Mas, J., M. Morales, M. A. Monforte, P. Arús, A. Monfort, N. Sánchez, and P. Puigdomènech. 2002. EST sequencing and SNP discovery in melon. p. 20–24. In: Cucurbitaceae 2002. ASHS Press, Alexandria, VA. Guner, N., and T. C. Wehner. 2003. Gene list for watermelon. Cucurbit Genet. Coop. Rep. 26:76–92. Hashizume, T., I. Shimamoto, Y. Harushima, M. Yui, T. Sato, T. Imai, and M. Hirai. 1996. Construction of a linkage map for watermelon (Citrullus lanatus (Thumb.) Matsum & Nakai) using random amplified polymorphic DNA (RAPD). Euphytica 90:265–273. Hashizume, T., I. Shimamoto, and M. Hirai. 2003. Construction of a linkage map and QTL analysis of horticultural traits for watermelon [Citrullus lanatus (Thunb.) Matsum & Nakai] using RAPD, RFLP and ISSR markers. Theor. Appl. Genet. 106:779–785. Hawkins, L. K., F. Dane, T. L. Kubisiak, B. B. Rhodes, and R. L. Jarret. 2001. Linkage mapping in a watermelon population segregating for fusarium wilt resistance. J. Am. Soc. Hort. Sci. 126:344–350. Horejsi, T., J. M. Box, and J. E. Staub. 1999. Efficiency of randomly amplified polymorphic DNA to sequence characterized amplified region marker conversion and their comparative polymerase chain reaction sensitivity in cucumber. J. Am. Soc. Hort. Sci. 124:128–135. Horejsi, T., J. E. Staub, and C. E. Thomas. 2000. Linkage of random amplified polymorphic DNA markers to downy mildew resistance in cucumber (Cucumis sativus L.). Euphytica 115:105–113. Jarret, R. L., L. C. Merrick, T. Holms, J. Evans, and M. K. Aradhya. 1997. Simple sequence repeats in watermelon (Citrullus lanatus (Thunb.) Matsum. & Nakai). Genome 40:433–441. Jeffrey, C. 1990. Systematics of the Cucurbitaceae: an overview. p. 3–9. In: D. M. Bates, R. W. Robinson and C. Jeffrey (eds.), Biology and utilization of the Cucurbitaceae. Cornell Univ. Press, Ithaca, NY. Joobeur, T., J. J. King, S. J. Nolin, C. E. Thomas, and R. A. Dean. 2004. The fusarium wilt resistance locus Fom-2 of melon contains a single resistance gene with complex features. Plant J. 39:283–297.

5. GENETIC MAPPING AND MOLECULAR BREEDING IN CUCURBITS

241

Katzir, N., Y. Danin-Poleg, G. Tzuri, Z. Karchi, U. Lavi, and P. B. Cregan. 1996. Length polymorphism and homologies of microsatellites in several Cucurbitaceae species. Theor. Appl. Genet. 93:1282–1290. Katzir, N., N. Mozes-Daube, Y. Danin-Poleg, and H. S. Paris. 2000. Potential usefulness of SSR markers for studying infraspecific variability in Cucurbita pepo. Cucurbit Genet. Coop. Rep. 17:61–65. Kennard, W. C., and M. J. Havey. 1995. Quantitative trait analysis of fruit quality in cucumber: QTL detection, confirmation, and comparison with mating-design variation. Theor. Appl. Genet. 91:53–61. Kennard, W. C., K. Poetter, A. Dijkhuizen, V. Meglic, J. E. Staub, and M. J. Havey. 1994. Linkages among RFLP, RAPD, isozyme, disease-resistance, and morphological markers in narrow and wide crosses of cucumber. Theor. Appl. Genet. 89:42–48. Klingler, J., I. Kovalski, L. Silberstein, G. A. Thompson, and R. Perl-Treves. 2001. Mapping of cotton-melon aphid resistance in melon. J. Am. Soc. Hort. Sci. 126:56–63. Knerr, L. D., and J. E. Staub. 1992. Inheritance and linkage relationships of isozyme loci in cucumber (Cucumis sativus L.). Theor. Appl. Genet. 84:217–224. Lee, Y. H., H. J. Jeon, K. H. Hong, and B. D. Kim. 1995. Use of random amplified polymorphic DNAs for linkage group analysis in interspecific hybrid F2 generation of Cucurbita. J. Kor. Soc. Hort. Sci. 36:323–330. Levi, A., C. E. Thomas, T. Joobeur, X. Zhang, and A. Davis. 2002. A genetic linkage map for watermelon derived from a testcross population: (Citrullus lanatus var. citroides × C. lanatus var. lanatus) × Citrullus colocynthis. Theor. Appl. Genet. 105: 555–563. Levi, A., C. E. Thomas, X. Zhang, T. Joobeur, R. A. Dean, T. C. Wehner, and B. R. Carle. 2001. A genetic linkage map for watermelon based on Randomly Amplified polymorphic DNA (RAPD) markers. J. Am. Soc. Hort. Sci. 126:730–737. Liou, P. C., Y. M. Chang, W. S. Hsu, Y. H. Cheng, H. R. Chang, C. H. Hsiao, and R. A. Drew. 1998. Construction of a linkage map in Cucumis melo (L.) using random amplified polymorphic DNA markers. Acta Hort. 461:123–131. Luo, M., Y.-H. Wang, D. Frisch, T. Joobeur, R. Wing, and R. A. Dean. 2001. Melon BAC library construction using improved methods and identification of clones linked to the locus conferring resistance to melon fusarium wilt (Fom-2). Genome 44:154–162. McCreight, J. D., H. Nerson, and R. Grumet. 1993. Melon Cucumis melo L. p. 267–294. In: G. Kalloo and B. O. Bergh (eds.), Genetic improvement of vegetable crops. Pergamon Press, New York. Meglic, V., and J. E. Staub. 1996. Inheritance and linkage relationships of allozyme and morphological loci in cucumber (Cucumis sativus L.). Theor. Appl. Genet. 92:865–872. Monfort, A., J. Garcia-Mas, N. Sanches-Coll, M. Morales, H. van Leeuwen, E. Graziano, A. J. Monforte, P. Puigdomènech, and P. Arús. 2004. Developing tools for melon genome analysis: DNA sequence variability and macrorsynteny with Arabidopsis. Plant & Animals Genome XII, San Diego, CA. Jan. 10–14, 2004. Monforte, A. J., M. Oliver, M. J. Gonzalo, J. M. Alvarez, R. Dolcet-Sanjuan, and P. Arús. 2004. Identification of quantitative trait loci involved in fruit quality traits in melon (Cucumis melo L.). Theor. Appl. Genet. 108:750–758. Morales, M., M. Luis-Arteaga, J. M. Alvarez, R. Colcet-Sanjuan, A. Monfort, P. Arus, and J. Garcia-Mas. 2002. Marker saturation of the region flanking the gene NSV conferring resistance to the melon necrotic spot Carmovirus (MNSV) in melon. J. Am. Soc. Hort. Sci. 127:540–544. Morales, M., H. van Leeuwen, P. Puigdomènech, P. Arús, A. Monfort, and J. Garcia-Mas. 2003. Towards the positional cloning of the nsv resistance gene against MNSV in melon. Plant & Animals Genome XI, San Diego, CA. Jan. 11–15, 2003.

242

Y. WANG, R. DEAN, AND T. JOOBEUR

Morales, M., E. Roig, A. J. Monforte, P. Arús, and J. Garcia-Mas. 2004. Single-nucleotide polymorphisms detected in expressed sequence tags of melon (Cucumis melo L.). Genome 47:352–360. Navot, N., and D. Zamir. 1987. Isozyme and seed protein phylogeny of the genus Citrullus (Cucurbitaceae). Plant Syst. Evol. 156:61–67. Navot, N., M. Sarfatti, and D. Zamir. 1990. Linkage relationships of genes affecting bitterness and flesh color in watermelon. J. Hered. 81:162–165. Neuhausen, S. L. 1992. Evaluation of restriction fragment length polymorphism in Cucumis melo. Theor. Appl. Genet. 83:379–384. Oliver, M., J. Garcia-Mas, M. Morates, R. Doleet-Sanjuan, M. C. de Vicente, H. Gómez, H. van Leeuwen, A. Monfort, P. Puigdomènech, and P. Arús. 2000. The Spanish melon genome project: construction of a saturated genetic map. Acta Hort. 510:375–378. Oliver, M., J. Garcia-Mas, M. Cardús, N. Pueyo, A. I. López-Sesé, M. Arroyo, H. GómezPaniagua, P. Arús, and M. C. de Vicente. 2001. Construction of a reference linkage map for melon. Genome 44:836–845. Paris, H. S., N. Yonash, Y. Portnoy, N. Mozes-Daube, G. Tzuri, and N. Katzir. 2003. Assessment of genetic relationships in Cucurbita pepo (Cucurbitaceae) using DNA markers. Theor. Appl. Genet. 106:971–978. Park, S. O., and K. M. Crosby. 2004. Development of RAPD markers linked to the malesterile ms-3 gene in melon. Acta Hort. 637:243–249. Park, Y. H., S. Sensoy, C. Wye, R. Antonise, J. Peleman, and M. J. Havey. 2000. A genetic map of cucumber composed of RAPDs, RFLPs, AFLPs, and loci conditioning resistance to papaya ringspot and zucchini yellow mosaic viruses. Genome 43:1003–1010. Pasha, M. K., and S. P. Sen. 1998. Numerical and chemical approaches to taxonomy and relationships. p. 19–32. In: N. M. Nayar and T. A. More (eds.), Cucurbits. Science Publ., Enfield, NH. Périn, C., L. Hagen, C. Dogimont, V. De Conto, and M. Pitrat. 1998. Construction of a genetic map of melon with molecular markers and horticultural traits. p. 370–376. In: J. D. McCreight (ed.), Cucurbitaceae ‘98. ASHS Press, Alexandria, VA. Périn, C., L. Hagen, V. De Conto, N. Katzir, Y. Danin-Poleg, V. Portnoy, S. Baudracco-Arnas, J. Chadoeuf, C. Dogimont, and M. Pitrat. 2002a. A reference map of Cucumis melo based on two recombinant inbred line populations. Theor. Appl. Genet. 104:1017–1034. Périn, C., L. S. Hagen, N. Giovinazzo, D. Besombes, C. Dogimont, and M. Pitrat. 2002b. Genetic control of fruit shape acts prior to anthesis in melon (Cucumis melo L.). Mol. Genet. Genomics 266:933–941. Périn, C., M. C. Gomez-Jimenez, L. Hagen, C. Dogimont, J.-C. Pech, A. Latché, M. Pitrat, and J.-M. Lelièvre. 2002c. Molecular and genetic characterization of a non-climacteric phenotype in melon reveals two loci conferring altered ethylene response in fruit. Plant Physiol. 129:300–309. Perl-Treves, T., D. Zamir, N. Navot, and E. Galun. 1985. Phylogeny of Cucumis based on isozyme variability and its comparison with plastome phylogeny. Theor. Appl. Genet. 71:430–436. Pierce, L. K., and T. C. Wehner. 1990. Review of genes and linkage groups in cucumber. HortScience 25:605–615. Pitrat, M. 1991. Linkage groups in Cucumis melo L. J. Hered. 82:406–411. Pitrat, M. 2002. 2002 gene list for melon. Cucurbit Genet. Coop. Rep. 25:76–93. Ritschel, P. S., T. C. de Lima Lins, R. L. Tristan, G. S. C. Buso, J. A. Buso, and M. E. Ferreira. 2004. Development of microsatellite markers from an enriched genomic library for genetic analysis of melon (Cucumis melo L.). BMC Plant Biol. 4:9. Robinson, R. W. 1999. Rationale and methods for producing hybrid cucurbit seed. J. New Seeds 1:1–47.

5. GENETIC MAPPING AND MOLECULAR BREEDING IN CUCURBITS

243

Robinson, R. W., and D. S. Decker-Walters. 1997. Cucurbits. CAB Int., New York. Serquen, F. C., J. Bacher, and J. E. Staub. 1997. Mapping and QTL analysis of a narrow cross in cucumber (Cucumis sativus L.) using random amplified polymorphic DNA markers. Mol. Breed. 3:257–268. Sherf, A. F., and A. A. Macnab. 1986. Vegetable diseases and their control. 2nd ed. Wiley, New York. p. 334–337. Silberstein, L., I. Kovalski, Y. Brotman, C. Perin, C. Dogimont, M. Pitrat, J. Klingler, G. Thompson, V. Portnoy, N. Katzir, and R. Perl-Treves. 2003. Linkage map of Cucumis melo including phenotypic traits and sequence-characterized genes. Genome 46:761–773. Silberstein, L., I. Kovalski, R. Huang, K. Anagnostou, M. M. K. Jahn, and R. Perl-Treves. 1999. Molecular variation in melon (Cucumis melo L.) as revealed by RFLP and RAPD markers. Sci. Hortic. 79:101–111. Staub, J. E., and V. Meglic. 1993. Molecular genetic markers and their legal relevance for cultivar discrimination: a case study in cucumber. HortTechnology 3:291–300. Staub, J. E., and F. C. Serquen. 2000. Towards an integrated linkage map of cucumber: map merging. Acta Hort. 510:357–366. Taler, D., M. Galperin, I. Benjamin, Y. Cohen, and D. Kenigsbuch. 2004. Plant eR genes that encode photorespiratory enzymes confer resistance against disease. Plant Cell 16:172–184. Tatlioglu, T. 1993. Cucumber Cucumis sativus L. p. 197–234. In: G. Kalloo and B. O. Bergh (eds.), Genetic improvement of vegetable crops. Pergamon Press, New York. Touyama, T., I. Asami, T. Oyabu, K. Yabe, S. Sugahara, and M. Kanbe. 2000. Development of PCR-based markers linked to a long-shelf-life-character ‘non-yellowing’ in muskmelon. Res. Bul. Aichi ken Agr. Res. Center 32:47–52. Trebitsh, T., J. E. Staub, and S. D. O’Neill. 1997. Identification of a 1-aminocyclopropane1-carboxylic acid synthase gene linked to the female (F) locus that enhances female sex expression in cucumber. Plant Physiol. 113:987–995. Vakalounakis, D. J. 1992. Heart leaf; a recessive leaf shape marker in cucumber: Linkage with disease resistance and other traits. J. Hered. 83:217–221. van Leeuwen, H., A. Monfort, H.-B. Zhang, and P. Puigdomènech. 2003. Identification and characterisation of a melon genomic region containing a resistance gene cluster from a constructed BAC library. Microcolinearity between Cucumis melo and Arabidopsis thaliana. Plant Mol. Biol. 51:703–718. Wang, Y.-H. 1999. Marker development and genome mapping for localizing disease resistance genes in melon (Cucumis melo L.). Ph.D Dissertation, Clemson University, Clemson, SC. Wang, Y.-H., C. E. Thomas, and R. A. Dean. 1997. A genetic map of melon (Cucumis melo L.) based on amplified fragment length polymorphism (AFLP) markers. Theor. Appl. Genet. 95:791–798. Wang, Y.-H., C. E. Thomas, and R. A. Dean. 2000. Genetic mapping of a fusarium wilt resistance gene (Fom-2) in melon (Cucumis melo L.). Mol. Breed. 6:379–389. Wechter, W. P., M. P. Whitehead, C. E. Thomas, and R. A. Dean. 1995. Identification of a randomly amplified polymorphic DNA marker linked to the Fom-2 fusarium wilt resistance gene in muskmelon MR-1. Phytopathology 85:1245–1249. Wechter, W. P., R. A. Dean, and C. E. Thomas. 1998. Development of sequence-specific primers that amplify a 1.5-kb DNA marker for race 1 fusarium wilt resistance in Cucumis melo L. HortScience 33:291–292. Weeden, N. F., and R. W. Robinson. 1986. Allozyme segregation ratios in the interspecific cross Cucurbita maxima × C. ecuadorensis suggest that hybrid breakdown is not caused by minor alteration in chromosome structure. Genetics 114:593–609.

244

Y. WANG, R. DEAN, AND T. JOOBEUR

Whitaker, T. W., and R. W. Robinson. 1986. Squash breeding. p. 211–242. In: M. J. Basset (ed.), Breeding vegetable crops. Avi Publishing Co. Inc., Westport, CT. Xie, J. H., and T. C. Wehner. 2001. Gene list 2001 for cucumber. Cucurbit Genet. Coop. Rep. 24:110–136. Xu, R., X. Ye, and Q. Q. Li. 2004. AtCPSF73-II gene encoding an Arabidopsis homolog of CPSF 73 kDa subunit is critical for early embryo development. Gene 324:35–45. Xu, Y., X.-X. Ouyang, H.-Y. Zhang, G.-B. Kang, and Y.-J. Wang. 1998. Identification of a molecular marker linked to chilling tolerance in watermelon wild germplasm using random amplified polymorphic DNA (in Chinese with English summary). Acta Hort. Sinica 25:397–398. Xu, Y., X.-X. Ouyang, H.-Y. Zhang, G.-B. Kang, and Y.-J. Wang, and H. Chen. 1999. Identification of a RAPD marker linked to fusarium wilt resistance gene in wild watermelon germplasm (Citrullus lanatus var. citroides) (in Chinese with English summary). Acta Bot. Sinica 41:952–955. Xu, Y., H. Y. Zhang, G. B. Kang, Y. J. Wang, and H. Chen. 2000. Studies of molecular marker-assisted selection for resistance to fusarium wilt in watermelon (Citrullus lanatus) breeding (in Chinese with English summary). Acta Genet. Sinica 27:151–157. Yariv, Y., V. Portnoy, Y. Burger, H. Debbi, G. Tzuri, N. Mozes-Daube, E. Lewinsohn, N. Katzir, J. Giovannoni, R. White, and A. Schaffer. 2002. The isolation and characterization of fruit ripening-related genes in melon (Cucumis melo)—A geneomic approach. p. 421–426. In: Cucurbitaceae 2002, ASHS Press, Alexandria, VA. Zamir, D., N. Navot, and J. Rudich. 1984. Enzyme polymorphism in Citrullus lanatus and C. colocynthis in Israel and Sinai. Plant Syst. Evol. 146:163–170. Zheng, X. Y., and D. W. Wolff. 2000a. Randomly amplified polymorphic DNA markers linked to fusarium wilt resistance in diverse melons. HortScience 35:716–721. Zheng, X. Y., and D. W. Wolff. 2000b. Ethylene production, shelf-life and evidence of RFLP polymorphisms linked to ethylene genes in melon (Cucumis melo L.). Theor. Appl. Genet. 101:613–624. Zheng, X. Y., D. W. Wolff, S. Baudracco-Arnas, and M. Pitrat. 1999. Development and utility of cleaved amplified polymorphic sequences (CAPS) and restriction fragment length polymorphisms (RFLPs) linked to the Fom-2 fusarium wilt resistance gene in melon (Cucumis melo L.). Theor. Appl. Genet. 99:453–463.

6 Breeding Douglas-fir Glenn T. Howe, Keith Jayawickrama, Marilyn Cherry Department of Forest Science Oregon State University 321 Richardson Hall Corvallis, Oregon 97331-5752 G. R. Johnson U.S.D.A. Forest Service Pacific Northwest Research Station 3200 SW Jefferson Way Corvallis, Oregon 97331-4401 Nicholas C. Wheeler 21040 Flumerfelt Rd. S.E. Centralia, Washington 98531

I. ABBREVIATIONS II. INTRODUCTION A. Rationale B. Objectives III. DISTINCTIVE CHARACTERISTICS OF FOREST TREES A. Biological Characteristics B. Ecological Characteristics IV. DOUGLAS-FIR: THE SPECIES A. Taxonomy B. Range C. Silvical Characteristics V. FACTORS THAT INFLUENCE DOUGLAS-FIR BREEDING A. Genetics and Life-History B. Abiotic Environment C. Insects and Disease D. Genecology

Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 245

246

VI.

VII.

VIII.

IX.

X.

XI.

XII. XIII.

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER E. Long-Term Provenance Tests F. Quantitative Genetics and Inheritance G. Sociological and Political Factors BREEDING GOALS AND OBJECTIVES A. Primary Breeding Objectives B. Secondary Breeding Objectives OVERVIEW OF TREE BREEDING METHODS A. Breeding Zones B. Population Improvement and Genetic Diversity C. Early Selection BREEDING PROGRAMS A. North America B. Europe C. New Zealand BREEDING AND TESTING METHODS A. Overall Breeding Strategies B. Breeding Zones—Theoretical Considerations C. Tree Breeding Populations D. Mating Designs—Theoretical Considerations E. Field Test Designs—Theoretical Considerations F. First-Generation Strategies G. Second-Generation Strategies PRODUCTION OF IMPROVED MATERIALS FOR REFORESTATION A. Introduction B. Seed Orchards C. Vegetative Propagation D. Deployment BIOTECHNOLOGY A. Vegetative Propagation B. Transgenics C. Molecular Genetic Markers D. Genomics GENE CONSERVATION ACKNOWLEDGMENTS LITERATURE CITED

I. ABBREVIATIONS BCMoF, British Columbia Ministry of Forests CMP, controlled mass pollination EST, expressed sequence tag GCA, general combining ability H2, broad-sense heritability h2, narrow-sense heritability IETIC, Inland Empire Tree Improvement Cooperative IFA, Industrial Forestry Association IUFRO, International Union of Forest Research Organizations

6. BREEDING DOUGLAS-FIR

247

MAS, marker-aided-selection N, census population size Ne, effective population size NWTIC, Northwest Tree Improvement Cooperative QTL, quantitative trait locus RAPD, random amplified polymorphic DNA RFLP, restriction fragment length polymorphism SCA, specific combining ability SMP, supplemental mass pollination SNP, single nucleotide polymorphism SSR, simple sequence repeat

II. INTRODUCTION A. Rationale Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco] is one of the most important timber species in western North America (Fig. 6.1). In 2002, over 27.8 million m3 (8 billion board feet) of Douglas-fir lumber was produced in the U.S., and in 1999, it comprised a third of all U.S. log exports (Howard 2001; Anon. 2003). Douglas-fir is planted as an exotic timber species in Europe, New Zealand, Australia, and Chile, and as a Christmas tree throughout the northern U.S. It is the most important North American tree species introduced into Europe, and European forests of Douglas-fir cover 600,000 ha, 350,000 of which are in France (Hermann and Lavender 1999). New Zealand has the largest area of Douglas-fir in the southern hemisphere, where it ranks second in importance to radiata pine (Pinus radiata). Douglas-fir is ideal for structural lumber and veneer—it is among the strongest and stiffest of all North American softwoods, dimensionally stable, and moderately durable (Forest Products Laboratory 1999). Although Douglas-fir is used for producing paper and cardboard, it excels as lumber for heavy-duty construction and veneer for plywood and engineered products (e.g., laminated veneer lumber). The finegrained appearance of large, slow-growing trees made Douglas-fir a prized softwood for interior use, although these trees are rarely harvested anymore. As Douglas-fir crops shifted from large, old-growth trees from natural stands to smaller trees from planted forests, the use of Douglas-fir in engineered products increased. Douglas-fir trees are easy to establish, fast growing, and mostly free from major insect and disease pests, making them an ideal forest crop.

248

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Figure 6.1. Upper panel: Douglas-fir branch with seed cone (drawing by Gretchen Bracher). Lower panel: Mature seed orchard of Douglas-fir (photo by Terrance Ye).

One of the distinctive characteristics of Douglas-fir in western North America (compared to most agronomic crops) is that it is extensively planted among natural stands of the same species. Douglas-fir is an ecologically important, dominant tree species, particularly in the Pacific Northwest (Oregon, Washington, and British Columbia). In the U.S., it

6. BREEDING DOUGLAS-FIR

249

grows on nearly 17 million ha, 3 million of which are in plantations (Smith et al. 2001). Another 4.5 million ha of Douglas-fir are found in Canada (Hermann and Lavender 1999). Over much of its range, Douglasfir is considered a keystone species (Lipow et al. 2003). Douglas-fir exhibits high levels of genetic variation for all economic and adaptive traits studied, providing a rich foundation for genetic improvement. Applied tree breeding efforts in Douglas-fir are among the most extensive in the world, with more than 4 million progeny from nearly 34,000 selections growing on almost 1,000 test sites in western North America (Lipow et al. 2003). Douglas-fir tree improvement is also being practiced in New Zealand, Australia, Chile, and Europe. Despite extensive breeding efforts and wide-scale planting of genetically improved Douglas-fir, the public has little knowledge about tree breeding concepts, practices, and implications. Nonetheless, the public often voices strong concerns about the management of Douglas-fir forests. B. Objectives In this paper, we describe the (1) key elements that distinguish tree breeding from other types of crop breeding, (2) differences between Douglas-fir breeding and other types of tree breeding, (3) environmental, biological, and sociological factors that make Douglas-fir breeding unique, (4) current state of Douglas-fir breeding, and (5) future of Douglas-fir breeding and research needs. Although Douglas-fir improvement is practiced throughout the world, we will focus on Douglas-fir breeding in western North America (i.e., California, Oregon, Washington, British Columbia, Idaho, and Montana), and the Pacific Northwest in particular—where most of the Douglas-fir breeding is occurring.

III. DISTINCTIVE CHARACTERISTICS OF FOREST TREES A. Biological Characteristics The distinctive biological and ecological characteristics of trees present both challenges and opportunities for tree breeders. The biggest challenges are that trees are large and slow to reach reproductive and economic maturity. Douglas-fir trees are infrequently harvested before they are 40 years old, and rotation lengths greater than 70 years are possible. Therefore, genetic selections must be made well before the final crop can be evaluated. Although final selections can be made when the trees are about 10 years old (Johnson et al. 1997), this still makes breeding

250

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

progress slow. Although the time it takes to reach reproductive maturity is often cited as a constraint, natural stands of coastal Douglas-fir begin producing seed when they are about 7 to 10 years old (Stein and Owston 2002) and it is possible to induce flowering on slightly younger trees. Therefore, the breeding cycle is constrained mostly by the final selection age, rather than by the age to reproductive maturity. In contrast, the long life spans of trees can be an advantage because desirable genotypes can be maintained essentially indefinitely using grafting or other methods of vegetative propagation. Although artificial methods of vegetative propagation are possible, natural reproduction occurs exclusively by seed. Because trees are large and must be grown for a long time, genetic field tests are both large and expensive, and this is exacerbated by the large number of families or clones that are often tested. A Douglas-fir progeny test might include 150 to 300 families planted in single-tree plots at 3 to 12 sites of 2.5 to 5 hectares each. Because trees occupy so much space, seedlots are infrequently tested in large block-plot experiments. Thus, the yield of single families or orchard seedlots in large production plantations is usually extrapolated from experiments in which these seedlots were actually grown in single-tree-plots or small row-plots. On the other hand, because individual trees can produce thousands of offspring, it is possible to precisely estimate their genetic worth using progeny tests. Most forest trees have a mixed mating system that is predominantly outcrossing, and inbreeding depression is usually severe (Sorensen 1999). This is one reason why few naturally occurring mutants have been characterized in forest trees. Inbreeding depression and the long generation interval make it difficult or impossible to develop inbred lines, use recurrent backcrossing, or recover recessive mutants in otherwise healthy trees. Therefore, mutational approaches to gene discovery, which have been so valuable in plants such as Arabidopsis and tomato, are not practical. In many agronomic crops, grain yields have been dramatically increased by breeding plants that allocate more of their photosynthate (and perhaps nitrogen) to the reproductive parts of the plant (Sinclair 1998). In contrast, genetic improvement in forest trees mostly comes from increasing vegetative growth, improving resistance to biotic or abiotic stresses, and enhancing wood and stem quality. Despite the challenges, genetic improvement of Douglas-fir is biologically and economically attractive. Although a generation may take 15 years, gains in stem volume of 20 to 30% may be possible in a single generation. Furthermore, Douglas-fir is in the earliest stages of domestication, and tree

6. BREEDING DOUGLAS-FIR

251

breeders can draw on a vast resource of untapped genetic variation for future genetic improvement. B. Ecological Characteristics The ecological characteristics of forest trees also impact breeding programs. Many forest tree crops are planted within their native ranges on sites that are only modestly improved at best. A typical program of “intensive” site improvement might include cultivation and weed control at planting, followed by occasional applications of nitrogen fertilizer and thinning during the rotation. Therefore, medium-term (i.e., rotation length) adaptation to the natural environment is critical. Furthermore, because many forest crops are planted on lands that are expected to remain forested indefinitely, it is important to maintain long-term (i.e., microevolutionary-scale) adaptation to both current and future environments. Therefore, maintenance of genetic diversity and adaptability are important goals in most tree breeding programs.

IV. DOUGLAS-FIR: THE SPECIES A. Taxonomy The taxonomy of Pseudotsuga is unclear. In North America, two species are well recognized, Douglas-fir and bigcone Douglas-fir [P. macrocarpa (Vasey) Mayr] (Hermann 1982). Both species are native to western North America, but their ranges do not overlap. Bigcone Douglas-fir is found only in southwestern California—34 km from the southernmost population of coastal Douglas-fir (Griffin 1964). Not surprisingly, no natural hybrids between these species have been found, although an artificial hybrid was reported by Ching (1959). Pseudotsuga populations in Mexico are either included with P. menziesii var. glauca, or considered additional species (Li and Adams 1989; Gernandt and Liston 1999). A variable number of other species have been described in China, Taiwan, and Japan (Gernandt and Liston 1999). Some DNA-based analyses suggest that Douglas-fir and bigcone Douglas-fir are more closely related to each other than they are to the Asian species (Strauss et al. 1990; Gernandt and Liston 1999), and attempts to hybridize Douglas-fir with the Asian species have been unsuccessful (Silen 1978). Within the Pinaceae, Pseudotsuga is considered to have a sister relationship with the genus Larix (Gernandt and Liston 1999).

252

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

B. Range Douglas-fir has one of the widest natural ranges of any tree species. It ranges from the Pacific coast to the eastern slopes of the Rockies and from Canada to Mexico (almost 4,500 km), occurring from sea level to over 3,000 m (Hermann and Lavender 1990; Fig. 6.2). Within this area, Douglas-fir grows on over 20 million ha (Hermann and Lavender 1999; Smith et al. 2001). Two botanical varieties of Douglas-fir are recognized—the coastal variety (var. menziesii) and the interior, or Rocky Mountain variety (var. glauca)—which are physiologically, morphologically, and chemically distinct (Silen 1978; Hermann and Lavender 1990; Fig. 6.2). Although phenotypic differences are pronounced in the south, the varieties intergrade in areas of contact from the northern half of Oregon northward into central British Columbia (von Rudloff 1972; Sorensen 1979). In contrast, RAPD markers amplified from mitochondrial DNA showed a rather abrupt genetic discontinuity in this same area (Aagaard et al. 1995). The northern and southern populations of Rocky Mountain Douglasfir are well separated geographically and almost genetically distinct (Wright et al. 1971; Li and Adams 1989). Based on rangewide patterns of variation in morphology, physiology, isozymes, and terpene chemistry, these northern and southern subgroups may deserve varietal status (Li and Adams 1989).

C. Silvical Characteristics Coastal Douglas-fir dominates the landscape over much of its range. On the best sites, coastal Douglas-firs can become huge—up to 120 m tall and 5 m in diameter, and may live up to 1,400 years (Silen 1978; Farrar 1995; Hermann and Lavender 1990). The coastal variety is usually an early seral component of forests. That is, large continuous stands tend to regenerate after disturbances such as fire, grow rapidly in full sun, then eventually give way to other species that regenerate and grow better beneath the canopies of mature Douglas-fir. This process, however, often takes hundreds of years. Interior Douglas-fir reaches maximum heights of only about 49 m on the best sites, and rarely grows older than about 400 years (Hermann and Lavender 1990). On moist sites, it may function as an early seral species, but on warmer, drier sites, it may be a component of the climax forest (Hermann and Lavender 1990). In many parts of the Rocky Mountains, it occurs in mixed stands with other species, rather than in extensive pure stands, as is more common in the coastal region. Compared to coastal Douglas-fir, the Rocky Mountain variety tends to be slower grow-

6. BREEDING DOUGLAS-FIR

253

Figure 6.2. Natural range of Douglas-fir (Pseudotsuga menziesii) in the U.S. and Canada. The dashed line represents the approximate location of the transition zone between the coastal (var. menziesii) and Rocky Mountain (var. glauca) varieties. Isolated populations of putative Mexican populations of Douglas-fir are not shown. The range map was downloaded from the USGS web site entitled “Digital Representations of Tree Species Range Maps from Atlas of United States Trees by Elbert L. Little, Jr. (and other publications)” (http://climchange.cr.usgs.gov/data/atlas/little/).

254

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

ing, more cold hardy, more shade tolerant, and less tolerant of Rhabdocline (pathogen, Rhabdocline pseudotsugae; Stephan 1973) and Swiss needle cast (pathogen, Phaeocryptopus gaeumannii; Ferre 1955; Rohmeder 1956) diseases. It also has shorter needles, bluer foliage, and cone characteristics that distinguish it from the coastal variety.

V. FACTORS THAT INFLUENCE DOUGLAS-FIR BREEDING A. Genetics and Life-History Douglas-fir is unique among the Pinaceae in having a diploid chromosome number of 26. The remaining species in the Pinaceae, including other species of Pseudotsuga, have a diploid chromosome number of 24 (Doerksen and Ching 1972). As in most other conifers, its nuclear genome is large and complex (3.7 × 1010 bp), chloroplast DNA is inherited paternally, and mitochondrial DNA is inherited maternally (Neale et al. 1986; Marshall and Neale 1992; O’Brien et al. 1996). Flowering and seed production begin at age 7 to 10 in coastal Douglasfir, and a little later in the Rocky Mountain variety (Stein and Owston 2002). In mature stands, good seed crops are produced every 2 to 11 years (Stein and Owston 2002). Douglas-fir does not naturally reproduce by vegetative propagation. Pollination occurs in the spring, and mature seeds are shed in late summer or early fall of the same year. Douglas-fir is monoecious and has a mixed mating system (selfing and outcrossing), but it is mostly an outcrosser. Selfing at the mature seed stage is usually less than 10% because of high mortality of the selfed embryos (Sorensen 1999). Inbreeding depression is manifested throughout its life cycle, and selfed progeny rarely become part of the mature, reproductive population (Sorensen 1999). In seedlings and saplings, inbreeding depression in growth traits seems to be linearly related to the inbreeding coefficient (F). Under nursery conditions, for example, inbreeding depression in height increased 5 to 6% for each 0.1 increase in F from 0.0 to 0.5 (Woods and Stoehr 1993; Sorensen 1997; Woods et al. 2002). Campbell (1979) estimated that a single tree in an old-growth stand is probably selected from 2,000 or more germinated seedlings. Seeds are easily dispersed several hundred feet or more (Isaac 1930; Bever 1954), and because Douglas-fir is wind-pollinated, pollen dispersal is extensive (Adams 1992). This contributes to large effective population sizes, low population differentiation for neutral genetic markers, and low linkage disequilibrium (Neale and Savolainen 2004). The nutritive tissue (megagametophyte) of Douglas-fir seed is haploid, and genetically identical to the egg cell (i.e., maternal contribution to the

6. BREEDING DOUGLAS-FIR

255

embryo). This facilitates genetic analyses using protein (allozyme) or DNA-based molecular markers. The maternal haplotype can be easily measured, and the paternal contribution can be deduced by comparing the genotype of the embryo with the haplotype of the seed megagametophyte. This makes it easy to use open-pollinated or controlled-cross seed to study segregation, construct linkage maps, determine paternity, measure gene flow, and use association genetics to link genes to phenotypes (Adams et al. 1997; Krutovskii et al. 1998; Neale and Savolainen 2004). Douglas-fir is a long-lived tree that undergoes pronounced developmental changes during its life cycle. Compared to juvenile trees, mature trees are competent to flower, have shorter growing seasons (i.e., earlier bud set), slower rates of shoot elongation, less annual height growth, greater plagiotropism, and lower capacities to be propagated via rooted cuttings or tissue cultures (Ritchie and Keeley 1994; Stein and Owston 2002). B. Abiotic Environment The natural range of Douglas-fir is large, mountainous, and environmentally diverse, resulting in correspondingly diverse selection pressures and genecological differentiation. The climate of the Douglas-fir region changes dramatically from west to east because of the Pacific Ocean and a series of mostly north-south mountain ranges. Moving from west to east, warm, moist air from the Pacific Ocean rises, cools, and releases moisture as it encounters each of these mountain ranges. These storms mostly arrive during the late fall, winter, and spring. This results in wet conditions along the coast, pronounced rain shadows on the eastern sides of the mountain ranges, and a gradient in precipitation and humidity from west to east. The western part of the region is warmed by the Pacific and protected from the influx of cold, continental air from the interior by the north-south mountain ranges. As one moves from the maritime climate near the coast to the continental climate in the Rocky Mountains, the sites are generally drier and colder, with shorter frost-free periods (Hermann and Lavender 1990). North-south trends are also observed throughout the region, but they are much less dramatic. The southern areas tend to be drier and warmer, with longer frost-free periods and longer photoperiods in the fall and winter (i.e., when short days promote growth cessation, cold acclimation, and dormancy induction). Throughout the region, additional environmental heterogeneity is imposed by large differences in elevation. The higher elevation sites tend to be colder and wetter, with larger diurnal fluctuations in temperature (Hermann and Lavender 1990).

256

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

1. Coastal Environment. Populations along the coast experience a truly maritime climate, characterized by mild wet winters, cool dry summers, and long growing seasons. The climate becomes increasingly continental toward the east, particularly at higher elevations. The major environmental gradients within the region are associated with distance from the ocean (i.e., degree of continentality), latitude, and elevation. The Klamath Mountains in southern Oregon and northern California are particularly hot and dry; whereas more northern areas on Vancouver Island, British Columbia, and the Olympic Peninsula of Washington include temperate rainforests with up to 444 cm (175 inches) of rain a year. Throughout the coastal region, much of the precipitation occurs as winter rain, although snow is prevalent at the higher elevations, particularly in the Cascades and Sierra Nevada (Hermann and Lavender 1990). Despite the high annual rainfall, the summers are remarkably dry, leading to the frequent description of the climate as simply “winter-wet, summer-dry.” Droughts in southern Oregon and northern California can last from May until September (Hermann and Lavender 1990). 2. Rocky Mountain Environment. The environment of the Rocky Mountain region can be broken down into northern, central, and southern subregions. The northern subregion has a mild continental climate (except for a dry period in midsummer), the central subregion has a continental climate, and the southern subregion has some of the driest areas and shortest frost-free periods (Hermann and Lavender 1999). Compared to coastal Douglas-fir, the Rocky Mountain variety has a more mosaic distribution with numerous disjunct populations, particularly in the south. 3. Environmental Challenges. The huge diversity of environments that Douglas-fir inhabits presents numerous physiological and developmental challenges. In forest trees, geographic patterns of genetic variation are molded by three major environmental factors—temperature, water availability, and photoperiod—and Douglas-fir is no exception (Morgenstern 1996; Aitken and Hannerz 2001; Howe et al. 2003). Temperature and water availability play two important roles. First, they are environmental constraints that limit survival and growth. Second, they are environmental signals that trees use to adjust their physiology and development to acclimate to unfavorable conditions. Unlike temperature and water, photoperiod acts mainly as an environmental signal. For both varieties, low temperatures are the major limiting factor within the northern part of the range, whereas lack of moisture is the predominant limiting factor in the south (Hermann and Lavender 1990).

6. BREEDING DOUGLAS-FIR

257

Cold temperatures cause damage and death from spring and fall frosts, heavy snow loads, ice storms, and winter exposure (Duffield 1956; Silen 1978). Trees are particularly susceptible to cold after they deacclimate and begin growing in the spring, and before they become fully cold hardy in the fall (Howe et al. 2003). In western Washington, for example, a major cold spell in November 1955 killed or damaged millions of Douglas-fir trees, leaving trees with forked tops and crooked stems that can still be seen today (Duffield 1956). Despite the importance of fall frosts, damage from spring frosts is predicted to be two to three times greater, at least for coastal Douglas-fir (Timmis et al. 1994). Because bud flush occurs progressively later in the spring as trees age (IrgensMoller 1967b), the potential for spring frost damage should also decrease as trees become older. Although spring and fall frosts are an important cause of damage and death, low temperatures in the middle of winter are rarely a problem because the trees are generally fully dormant and cold hardy at this time. Because spring and fall frosts limit the duration of seasonal growth, the length of the frost-free growing season is a particularly important environmental constraint. Summer droughts, which are pronounced in many portions of the range, also damage and kill trees (Silen 1978). Because trees are more susceptible to droughts when they are actively elongating, trees from drier climates tend to set bud earlier in the summer, have shorter growing seasons, and less annual height growth. Trees that have stopped elongating and set a terminal bud in response to dry summer conditions occasionally begin growing again (second flush) if water becomes available later in the growing season. High temperatures are also a problem because they contribute to evapotranspiration and drought stress. Temperature, moisture, and photoperiod also act as important environmental signals that help trees acclimate to adverse conditions (Howe et al. 1999, 2003). Short days and low night temperatures promote bud set, cold acclimation, and winter dormancy in the fall. Temperatures just below freezing induce the second stage of cold acclimation, temperatures just above freezing promote dormancy release by fulfilling the chilling requirement, and warm temperatures promote cold deacclimation and bud flush in the spring. Moisture stress may also act as an environmental signal because it promotes growth cessation and bud set, which enhance drought hardiness (Griffin and Ching 1977). C. Insects and Disease Although insects and disease cause economic losses, no insect or disease pests are serious enough to warrant the development of large

258

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

resistance breeding programs. Swiss needle cast (Phaeocryptopus gaeumannii) is a problem in some areas along the Oregon Coast. The root and heart rots caused by Armillaria ostoyae and Phellinus weirii can also cause problems, particularly in young stands. Insects attack Douglas-fir forests, but major problems occur only during infrequent outbreaks. The Douglas-fir tussock moth (Orgyia pseudotsugata) and the western spruce budworm (Choristoneura fumiferana), for example, can cause serious defoliation in some areas and years (Hermann and Lavender 1990). Cone and seed insects can destroy seed orchard crops, but they can be controlled with insecticides. D. Genecology Genecology is the study of genetic differences in relation to the environment. As discussed above, the key environmental factors to consider are temperature, water, and photoperiod. Because gradients in these environmental factors are associated with location variables (e.g., elevation, latitude, and longitude), genetic differences have been mostly described in relation to these easily measured variables. Nonetheless, location variables are imperfect surrogates for the underlying climatic variables that drive patterns of genetic variation. Mathematical models are now available to predict the climate at any location from geographic and topographic information. Thus, there is an increasing trend toward describing fine-scale patterns of genetic variation in relation to the predicted climate, rather than to location variables alone (Balduman et al. 1999). The genecology of Douglas-fir has been studied at many scales—using wide ranging provenance tests (Wright et al. 1971) to studies of variation within a single watershed (Campbell 1979). A hierarchical approach is often used to describe genetic variation in forest trees. Variation in Douglas-fir may be partitioned among varieties, provenances-withinvarieties, populations-within-provenances, and individuals-withinpopulations. We use “seed source” as a synonym for “provenance” to denote the original geographic source of seed, pollen, or propagules. Furthermore, the geographic area described by the term “provenance” (or “seed source”) is assumed to contain multiple populations. In this section, we describe the patterns of genetic variation seen in Douglas-fir, starting at the species level, and then progressing to finer scales. An understanding of these patterns can be used to help design effective breeding programs and evaluate the potential effects of climate change. 1. Adaptive Traits. Studies of genecology focus on “adaptive traits”— traits believed to be under strong natural selection because they confer

6. BREEDING DOUGLAS-FIR

259

adaptation to the environment and enhance individual fitness. In most cases, indirect evidence is used to infer that a trait is under strong natural selection, including (1) physiological data that link the trait to enhanced survival and growth and (2) patterns of genetic variation that suggest the trait has been influenced by natural selection, including levels of population differentiation that exceed those observed for neutral genetic markers, and consistent associations between genotypes and location or environmental variables (Howe et al. 2003). Studies of adaptive traits typically include survival, height growth, fall and spring frost hardiness, drought hardiness, vegetative bud phenology (i.e., the timing of fall bud set or spring bud flush), and the frequency of second flushing. In natural stands, early height growth is important because Douglasfir grows poorly in heavy shade (e.g., compared to western hemlock [Tsuga heterophylla], or western redcedar [Thuja plicata]), often regenerates in dense stands following fire, and is browsed by deer and elk (Hermann and Lavender 1990). At the sapling stage, rapid height growth is needed to become a dominant part of the forest canopy, thereby facilitating survival, as well as early and prolific seed production. The importance of frost and drought hardiness is obvious, given some of the harsh environments in which Douglas-fir grows. Vegetative bud phenology is important because it is related to annual height growth and the ability to withstand frosts and droughts. Trees that begin elongating too early in the spring are likely to be damaged by late spring frosts, whereas trees that grow too late into the fall are likely to be damaged by summer droughts, early fall frosts, and winter cold (Campbell and Sorensen 1973; Griffin and Ching 1977; White 1987; Howe et al. 2003). A delay in bud set of a single week can increase fall frost damage by 18 to 25% (Campbell and Sorensen 1973; Rehfeldt 1983a). Because early bud set also limits annual shoot elongation, there is an important tradeoff between height growth and fall frost hardiness (Howe et al. 2003). The timing of bud flush is affected by chilling and flushing requirements (Howe et al. 1999; 2003). High chilling requirements help prevent mid-winter bud flush and cold injury in climates with winter temperatures that fluctuate above and below freezing. In contrast, early bud flush may be advantageous in cold environments that have very short growing seasons (i.e., to maximize seasonal growth), whereas late bud flush may be advantageous in mild climates that have particularly variable spring temperatures and high risks of late spring frosts (Campbell and Sugano 1979). Second flushing occurs when a tree stops elongating and sets a bud, then flushes a second time in the same growing season. Young trees may second flush if temperature and moisture conditions are favorable in late

260

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

summer. Second flushing promotes annual height growth, but also predisposes trees to damage from fall frosts (Rehfeldt 1979a; Anekonda et al. 1998). In summary, a large component of climatic adaptation involves the synchronization of a tree’s annual growth cycle to the prevailing frostfree and drought-free growing seasons. Poor synchronization may lead to poor height growth, inadequate lignification, needle and twig damage, mortality caused by freezing of buds, and increased susceptibility to disease (Campbell 1986). Therefore, the lengths of the frost-free and drought-free growing seasons influence much of the geographically based genetic variation we see in forest trees (Morgenstern 1996). Douglas-fir exhibits high levels of genetic variation for each of these adaptive traits—within and among varieties, provenances, and populations (Silen 1978; Sorensen 1979; Campbell 1986; Rehfeldt 1989). Population differences in other adaptive traits, such as resistances to diseases, insects, and mammals are also present (e.g., McDonald 1979; Hoff 1987), but not well characterized. 2. Long-Term Field Tests vs. Short-Term Seedling Tests. Patterns of genetic variation in adaptive traits have been studied using long-term field tests and short-term experiments in outdoor nurseries, greenhouses, or growth chambers (hereafter referred to as seedling tests). The longterm field tests consist of everything from wide-ranging provenance tests to progeny tests of parent trees from limited geographic areas (i.e., single breeding zones). The seedling tests often include imposed environmental treatments (such as cold or drought) and detailed physiological analyses. Long-term field tests and seedling tests have a few key differences that must be considered when their results are interpreted (White and Ching 1985). First, the long-term field tests provide direct information, whereas the seedling tests provide indirect information for designing seed transfer guidelines and breeding zones (Adams and Campbell 1981). Second, short-term seedling experiments focus on the juvenile stage of tree development, which is substantially different from that of older, more mature trees (Ritchie and Keeley 1994). Compared to mature trees, Douglas-fir seedlings set bud much later in the growing season, are more prone to second flush, and are less cold hardy (Rehfeldt 1983b). Therefore, patterns and amounts of genetic variation change as trees mature (White and Ching 1985). Third, seedlings in short-term tests are frequently subjected to cold and drought treatments to specifically measure their resistances to these environmental stresses. In contrast, trees in long-term field tests may never experience severe frosts or droughts, or may be exposed for only a small fraction of their life span (e.g., single growing

6. BREEDING DOUGLAS-FIR

261

season). Fourth, seedling tests are often more precise (i.e., have higher heritabilities) than long-term field tests, often include more genotypes, and are generally used to measure many more traits (e.g., Campbell 1986). Although many of the same traits could be measured in the field, it would be enormously more difficult and expensive, although the results may be more relevant. Compared to the large number of mostly adaptive traits measured in seedling experiments, survival and growth (e.g., height, diameter, or volume) are often the only traits reported from long-term field tests (e.g., Silen and Mandel 1983). Therefore, it is usually easier to uncover subtle genetic differences using seedling experiments. Finally, non-genetic differences in seed weight, environmental preconditioning (Johnsen 1988), or other maternal effects that are influenced by the environment in which the seed was produced play a larger role in seedling tests when they are present. These factors may lead to inflated estimates of the associations between parental genotypes and their environment. Overall, seedling experiments typically involve a detailed assessment of adaptive traits in the first one to three years, whereas long-term field tests measure overall tree growth over a much longer period of time. In the following sections, we describe climatic, geographic, and topographic patterns of genetic variation in Douglas-fir based on results from long-term field tests and seedling experiments. 3. Varietal Differences. Compared to the coastal variety, the Rocky Mountain variety of Douglas-fir experiences a colder, drier climate with shorter growing seasons. These environmental differences, however, are not abrupt, and west to east clines in genetic variation have developed along corresponding temperature and moisture gradients. Based on long-term provenance tests, there is a corresponding decrease in height growth from west to east. Pacific coast populations grow fastest, populations from interior British Columbia and the northern and southern Rocky Mountains are intermediate, whereas populations from the intermountain region and east slope of the Rocky Mountains grow the slowest (Silen 1978). Compared to coastal populations, interior populations set bud earlier and grow slower in a wide range of environments (IrgensMoller 1967a; Wright et al. 1971; Rehfeldt 1977; Silen 1978; Sorensen 1979). At the seedling stage, inland populations are also better able to tolerate fall frosts, winter cold, and droughts (Ferrell and Woodward 1966; Pharis and Ferrell 1966; Wright et al. 1971; Rehfeldt 1977). Furthermore, the poor survival of coastal sources planted in the Rocky Mountains, Michigan, and Eastern Europe has been attributed to poor winter hardiness (Wright et al. 1971; Silen 1978). Because interior seedlings are more sensitive to daylength (Irgens-Moller 1962), some of these differences may be tied to differences in short-day-induced bud set and

262

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

cold acclimation. Compared to interior provenances, coastal provenances are more tolerant of both Rhabdocline and Swiss needle cast diseases (Ferre 1955; Rohmeder 1956; Stephan 1973). 4. Within-Variety Patterns of Genetic Variation. Detailed genecological studies have mostly focused on specific regions, breeding zones, or watersheds of either coastal Douglas-fir or Rocky Mountain Douglas-fir. In the largest study, wind-pollinated families from about 1,300 parents in Oregon and Washington are being studied in seedling nursery tests (J. B. St.Clair, pers. comm.). Overall, genecological studies confirm the overriding importance of temperature and moisture regimes in shaping genetic variation throughout the species range. Compared to species such as western white pine (Pinus monticola) and western redcedar, Douglas-fir is particularly responsive to these selective forces, and is considered an adaptive specialist. Mean genetic differences in adaptive traits can be detected among populations that are separated by as little as 100 to 200 m in elevation (Rehfeldt 1979a,b). Our discussion of adaptive genetic variation will focus on these two selective forces separately. 5. Genecological Models and Maps. Many of the relationships described below were uncovered using genecological modeling—a set of multivariate statistical approaches commonly used to study the relationships between adaptive traits and either climatic or location variables (Campbell 1986). The genotypic data for genecological modeling comes from common garden studies of wind-pollinated seedlots collected from wild parents. The variation among populations is usually partitioned using principal components analysis (PCA). PCA is a statistical method for finding linear combinations of adaptive traits that can be grouped together and described by a single composite variable, or principal component. PCA is useful for reducing a large set of interrelated traits into a smaller number of principal components (often 2–3) that explain most of the variation. The first principal component (PC1), for example, may largely reflect variation in height growth, second flushing, and the timing of bud set, whereas the second (PC2) may largely reflect variation in the timing of bud flush and spring frost damage. Population means for the composite traits (i.e., factor scores) are calculated, and multiple regression is used to determine which location variables are significantly associated with the population means. The resulting genecological models can then be used to: (1) quantify the degree of genetic change that is associated with various environmental, geographic, or topographic gradients and (2) produce genecological maps, which are visual representations of the relationships between adaptive trait variation

6. BREEDING DOUGLAS-FIR

263

(i.e., factor scores) and location or climatic variables (e.g., O’Neill and Aitkin 2004). These genecological maps (which are analogous to topographic maps) can be constructed using geographic information systems (GIS). 6. Temperature-Related Patterns of Variation. Colder environments are associated with a number of highly correlated climatic variables, such as shorter frost-free growing seasons, colder temperatures during the winter months, and colder average annual temperatures. Populations from colder locations typically set bud earlier in the fall, have fewer growth flushes, less annual height growth, and are more tolerant of fall frosts and winter cold (Sweet 1965; Griffin 1974; Griffin and Ching 1977; Aitken et al. 1996; Balduman et al. 1999; J. B. St.Clair, pers. comm.). Therefore, the suite of traits described above is characteristic of populations from higher elevations, more northern latitudes, and (at least for coastal Douglas-fir) greater distances from the Pacific Ocean (Sweet 1965; Kung and Wright 1972; Griffin and Ching 1977; Larsen 1981; Sorensen 1983; Balduman et al. 1999). Patterns of variation have also been described for the timing of bud flush and spring frost hardiness, but these patterns are neither as strong nor consistent as for the traits described above (Sweet 1965; Sorensen 1967; Griffin and Ching 1977; Howe et al. 2003). In general, it seems that trees from colder locations, higher elevations, and greater distances from the ocean flush slightly earlier and are, therefore, somewhat more susceptible to cold damage in the spring (Campbell and Sugano 1979; Balduman et al. 1999). Contradictory or inconclusive results from other studies probably result from insufficient sampling (Munger and Morris 1936; Sweet 1965). 7. Moisture-Related Patterns of Variation. Patterns of genetic variation are also associated with the availability of water, which is mainly a function of rainfall and temperature, but also humidity, solar radiation, and wind. Southern populations of both varieties are exposed to hot, dry climates with early summer droughts (Silen 1978). In Douglas-fir seedlings, there is a corresponding trend toward increasing drought hardiness from north to south—particularly from the northern, mesic areas along the coast to the hot, dry region of southwestern Oregon and California (coastal variety), and from the intermountain region to the southern Rockies (Rocky Mountain variety) (Pharis and Ferrell 1966; Heiner and Lavender 1972; Kung and Wright 1972; Larsen 1981). In coastal Douglas-fir, populations from southern Oregon flush earlier, set bud earlier, and grow slower than more northern populations (Lavender et

264

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

al. 1968; Lavender and Overton 1972; Heiner and Lavender 1972). Genetic clines are also associated with the pronounced west to east decrease in precipitation and humidity—eastern populations set bud earlier, have less annual height growth, and are more drought hardy (Pharis and Ferrell 1966; Sorensen 1979). Much of our detailed knowledge about genetic variation in drought hardiness traits comes from studies that focused on genetic variation in southwestern Oregon and northern California. Because of the harsh environments and difficulty with plantation establishment, southwestern Oregon received extra attention through the Forest Intensive Research (FIR) program (e.g., Campbell 1986, 1991). In a genecological study of southwestern Oregon populations, trees from higher elevations, farther south, and farther east were shorter, set bud earlier, had less winter damage, and were more drought hardy at the seedling stage (Campbell 1986; White 1987). Furthermore, clines in drought hardiness followed gradients in summer precipitation (White 1987). In other studies of trees from southwestern Oregon, inland populations from drier environments had slower rates of shoot elongation, earlier bud set, and higher root:shoot ratios compared to coastal populations (Campbell 1986; Joly et al. 1989). Interestingly, however, genetic variation in drought hardiness was more closely related to temperature regimes than to rainfall patterns in this region (White 1987). Because early bud set is strongly correlated with drought hardiness, this relationship may result from an adaptation to the colder temperatures and short growing seasons at high elevations, rather than to drought per se (Sorensen 1983; White 1987). High-elevation populations are also better able to survive winter drought (Larsen 1981). 8. Disease- and Insect-Related Patterns of Variation. Resistances to a few fungal pathogens and insect pests also show geographic patterns of genetic variation. Douglas-fir populations from northern California are more resistant to the Cooley spruce gall adelgid (Gilletteella cooleyi) than are populations from Oregon, Washington, and British Columbia (Stephan 1987). As discussed above, the coastal variety of Douglas-fir is more tolerant of Rhabdocline and Swiss needle cast diseases than is the Rocky Mountain variety (Ferre 1955; Rohmeder 1956; Stephan 1973). Furthermore, within the coastal variety, tolerance to Swiss needle cast increases near the coast, at lower elevations, and as one moves northward from California (reviewed in Johnson 2002). 9. Summary and Implications. A large body of information demonstrates that patterns of genetic variation in adaptive traits are associated

6. BREEDING DOUGLAS-FIR

265

with temperature and moisture regimes. In general, the best growth potential occurs in populations from the best growing environments (i.e., mild climates with plenty of moisture) (Silen and Mandel 1983; Monserud and Rehfeldt 1990; Hernandez et al. 1993). Although climate models are now being used to study adaptive traits in relation to predicted climatic variables, these climatic variables are only slightly better than location variables for explaining patterns of genetic variation in Douglas-fir (Balduman et al. 1999; J. B. St.Clair, pers. comm.). The numerous reports of associations between adaptive traits and location variables support the conclusion that temperature and moisture are the major selective forces influencing patterns of adaptive variation. Differences in elevation are particularly important because genetic differences have been reported within small geographic areas (Balduman et al. 1999; Rehfeldt 1989). Within both varieties, west to east clines are substantial (Kung and Wright 1972; Rehfeldt 1978; Griffin and Ching 1977; Campbell and Sorensen 1978; Sorensen 1983; J. B. St.Clair, pers. comm.). Differences in latitude are subtler, and genetic differentiation is less pronounced. The latitudinal clines that do exist seem to be stronger in the Rocky Mountains than in the coastal region, presumably because of the moderating effects of the Pacific Ocean (Hernandez et al. 1993). Moisture regimes seem to be particularly important drivers of genetic differentiation in the southern parts of the species range where the climates are hot and dry, and the growing seasons are limited by summer drought. Each of these patterns of genetic variation must be carefully considered when breeding and deployment strategies are designed. Although genecological modeling consistently detects statistically significant relationships between population means and either climatic or location variables, the proportion of population variation that remains unexplained is often large. In the long-term studies of coastal Douglas-fir that were reported by Balduman et al. (1999), the amount of unexplained variation in cold hardiness and growth phenology ranged from 65 to 83%. In seedling studies of Rocky Mountain Douglas-fir, however, the amount of unexplained variation was much lower—only about 21% (Rehfeldt 1989). Although better results might be achieved if climatic variability and the frequency of extreme climatic events were taken into account (Balduman et al. 1999), the nature of this unexplained variation remains an important area of research. What are the implications of these patterns of genetic variation? First, large elevational transfers of seed (which are possible within relatively small geographic areas) can have adverse effects. Based on field tests of 6- to 7-year-old trees, for example, Balduman et al. (1999) predicted that (on average) seed transferred to a site 600 m higher in elevation would

266

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

experience a 15% increase in fall cold damage. Rehfeldt (1989) studied seedlings from northern Idaho and western Montana and detected genetic differences between populations separated by only 240 m in elevation—which corresponded to a difference of 20 frost-free days. Based on data from two coastal breeding zones, Silen and Mandel (1983) predicted that a transfer of seed 305 m (1,000 feet) upward would result in a 5 % increase in height at ages 10 to 12. A similar (but weak) trend has been seen in other breeding zones as well (K. Jayawickrama, unpubl. data). Balduman et al. (1999) also predicted that damage from spring frosts would increase if genotypes were moved toward the coast, whereas damage from fall frosts would increase if genotypes were moved away from the coast. North-south transfers within a given elevation are likely to be less risky, especially along the coast. These patterns, however, reflect differences among population means, yet large amounts of genetic variation also exist within populations. In fact, within-population variation for adaptive traits often equals or exceeds the variation found among populations (Wheeler et al. 1990; Rehfeldt 1979a; Howe et al. 2003). In contrast, White (1987) reported that differences in drought tolerance among populations were much larger than differences among families. Although local patterns of adaptive variation can be used for designing seed transfer guidelines and breeding zones (Campbell 1986; Silen and Mandel 1983), they are not useful for making selections within breeding populations. For example, Balduman et al. (1999) used long-term field tests to study the relationship between parent tree environment and cold hardiness traits within two breeding zones, and concluded that these associations were too weak to be an effective tool for tree improvement. E. Long-Term Provenance Tests In addition to providing an understanding of genecology, provenance tests can shed light on two important questions: what is the best provenance to plant in a particular region, and what is the extent of provenance by environment interaction. In this section, we describe what is known about the long-term performance of seed sources in those regions that have large Douglas-fir breeding programs—i.e., western North America, Europe, and New Zealand. In contrast to the results from seedling genecological studies, which have been used to justify relatively small breeding zones in the Pacific Northwest (Silen 1978), results from long-term provenance tests have generally been viewed as evidence that larger breeding zones are appropriate (Woods 1993; Stonecypher et al. 1996).

6. BREEDING DOUGLAS-FIR

267

1. Pacific Northwest. Two provenance tests have been established in western Washington and western Oregon. The first test was established in 1912 by the U.S. Forest Service. Wind-pollinated progeny from 120 parents from 13 locations were planted at five test sites in western Oregon and Washington (Munger and Morris 1936). Unfortunately, this test has produced few robust analyses and conclusions because of problems with its initial design (Adams and Campbell 1981). Nonetheless, early measurements (e.g., 17 to 18 years old from seed) indicated that provenance by plantation interactions were small and that seed transfer among mild sites in the region would be successful (Munger and Morris 1936; Ching and Hinz 1978). For example, several sources grew well at all five test plantations that spanned an elevational range of 1,000 m and a latitudinal range of about 3° (Munger and Morris 1936; reviewed in Stonecypher et al. 1996). In contrast, interactions seemed to be present when the performance of high and low elevation sources were compared at high and low elevation sites, a result that is not surprising given that elevation is consistently identified as an important factor in genecological studies. Based on later measurements, Silen suggested that these elevational interactions increased over time, and that increased mortality and reduced volume growth of some of the better seed sources had resulted in changes in rank that became apparent around age 30, and were large by age 50 (Silen 1965, 1966a, 1978). The significance of these reports, however, has been questioned. Stonecypher et al. (1996), for example, pointed out that Silen did not support these conclusions with “complete and detailed analyses.” The only widely distributed, long-term provenance test in the Pacific Northwest was planted in 1959 in Oregon, Washington, British Columbia, and northern California (Ching 1965). Unfortunately, only 16 seed sources were tested, the experiment has low statistical precision, and results from many of the 17 plantations were never published (White and Ching 1985). Because results from plantations damaged by animals, fire, frost, and drought were not reported (Ching and Hinz 1978), provenance by environment interactions may be underestimated. Nonetheless, measurements over 25 years indicate that growth rates vary significantly among seed sources, and interactions between source and planting site are mostly unimportant (White and Ching 1985). Furthermore, Ching and Hinz (1978) found no evidence to suggest that a provenance is best adapted to the location from which it came. Three of four seed sources from Vancouver Island were always among the tallest, whereas two others (i.e., the southernmost source and one high elevation source) were consistently among the shortest (White and Ching 1985). However, this test also provides evidence that seed source rankings may change over

268

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

time. One provenance that performed poorly on five test sites at age nine was average or superior by age 25 (White and Ching 1985). This test also suggests that plantations and sources from southwestern Oregon should be considered distinct, which is consistent with results from genecological studies (White and Ching 1985). In an archive plantation in Oregon, the local provenance had one of the largest mean diameters, but sources from as far away as Vancouver Island and northwest Washington were not significantly different from the local source (Gamble et al. 1996). Because of the limitations of the provenance tests in Washington and Oregon, it is difficult to draw definitive conclusions about the extent of genotype by environment interactions. This, and other factors, has resulted in two very different conclusions about the appropriate size of breeding zones in this region (discussed below). Despite their limitations, these tests are widely viewed as providing little evidence for strong provenance by environment interaction, unless seed transfer involves large elevational distances (Adams and Campbell 1981; White and Ching 1985; Stonecypher et al. 1996; Johnson 1997; but see Silen 1978). In British Columbia, provenance tests tell the same general story. Ching and Hinz (1978) summarized 20-year results from five provenance test plantations in British Columbia, whereas White and Ching (1985) reported on two low-elevation coastal plantations at age 25. At age 25, the five fastest-growing provenances were from British Columbia, Washington, and Oregon, but not from California (White and Ching 1985). Illingworth (1978) analyzed data from four series of provenance tests in coastal British Columbia and concluded that there is no clear relationship between growth and latitude or elevation for sources collected from an extensive region along the coast. In general, the sources from the areas with optimal growing conditions (i.e., warm and moist) grew the best, the performance of local sources was generally inferior, and sources from the north coast and transition zones sources were shorter overall. Another provenance test of 102 sources was planted in 1971 at a single site in coastal British Columbia and measured in 1986 (Sziklai 1990). In this test, four of the five fastest-growing provenances came from Washington, and the fifth came from British Columbia. Sziklai (1990) also noted that selection at age seven might be risky, but this was based on a consistently declining performance of a single source from the coast of southwestern Oregon. Overall, the results from coastal British Columbia provenance tests were used to conclude that (1) genetic differences are not strong among coastal British Columbia populations, (2)

6. BREEDING DOUGLAS-FIR

269

maladaptation does not seem to be a problem unless interior sources are planted along the coast, and (3) only a few breeding zones are needed in this region (Woods 1993). A provenance test of 64 interior provenances was established on a single site in the southern interior of British Columbia (Jaquish 1990). Provenances from a large area within British Columbia and Washington grew well, and the growth of the northern sources was closely associated with the elevation from which they were collected. 2. Europe. Extensive information is available from the IUFRO (International Union of Forest Research Organizations) provenance tests that were established in 33 countries during the 1960s, 1970s, and 1980s (Breidenstein et al. 1990). Interior provenances perform poorly in Europe, whereas provenances from a large area in Washington and Oregon (excluding southwest Oregon) combine good growth, good form, late flushing, and low susceptibility to fall frosts (Breidenstein et al. 1990; Anon. 1998; EURFORGEN 2003). One hundred and eight test sites in Europe and British Columbia were classified into four groups based on climatic conditions: (1) northeastern Europe, (2) northwestern Europe and southwestern British Columbia, (3) northwestern British Columbia, and (4) southern Europe. Analyses of height and survival indicated that (1) provenances from Washington grew the fastest in all site groups, (2) additional provenances from northern Oregon grew fast in areas with mild climates (i.e., groups 2 and 4), (3) some specific provenances grow fast and are stable across all groups of sites, and (4) height growth decreases as the elevation of origin increases. In Spain, most of the fastest-growing provenances came from a wide area covering 27 contiguous seed zones, including 4 zones in southwestern British Columbia, 13 zones in western Washington, and 10 zones in northwestern Oregon (Hernandez et al. 1993). In Europe (as in British Columbia), high-elevation provenances and provenances from latitudinal extremes grow slower (Illingworth 1978; Breidenstein et al. 1990; Hernandez et al. 1993). 3. New Zealand. Provenance trials in New Zealand indicate that coastal provenances from southern Oregon and northern California are the most productive (R. L. Knowles, pers. comm.). Based on these results, new seed collections were made in the best provenances in 1993, and new provenance-progeny trials were established in 1996. 4. Summary and Implications. Based on provenance tests from throughout the world, there is a consistent tendency for the best seed sources to

270

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

come from large areas that are climatically similar to the planting site. Except for plantations on harsh sites, the amount of seed source by environment interaction is often low for most provenances, suggesting that breeding zones can encompass large areas that have sufficient climatic homogeneity. Provenances that are likely to interact with planting site are likely to come from high elevations, latitudinal extremes, and southwestern Oregon (at least for coastal Douglas-fir). Transfers among mild environments can span large geographic distances, presumably because the environments in which a particular seed source does well recur at many locations across the landscape (Rehfeldt 1990). In short, the environmental distance is far more important than the geographic distance for evaluating seed transfer (Adams and Campbell 1981). This is important because most breeding programs concentrate on the lower-elevation, milder sites—and most of the high-elevation areas have been dropped from the second-generation breeding programs (discussed below). In many tree species, there is a consistent increase in growth when seed sources are transferred from milder to harsher climates (Namkoong 1969; Silen 1978). However, unless long-term field performance has been measured, this approach is risky because evidence of maladaptation may not show up for many years, and perhaps only after exposure to extreme climatic events (Silen 1978; Adams and Campbell 1981). Therefore, all of the major North American Douglas-fir breeding programs used local provenances as their first-generation breeding population, although the geographic extent of the parents varied dramatically. The lack of high-quality, region-wide provenance tests in Oregon and Washington has been a major impediment to answering fundamental questions about seed movement and breeding zones in this region. In retrospect, it is unfortunate that the comprehensive provenance tests that were established in Europe and British Columbia were never planted in Washington and Oregon, the region from which most of the provenances were collected. F. Quantitative Genetics and Inheritance An understanding of genetic and environmental variation is important for designing breeding strategies, picking suitable mating designs, designing field tests, and predicting genetic gains. Key pieces of information include relative amounts of additive vs. non-additive genetic variance, genetic and environmental variances, heritabilities, and genetic correlations (Namkoong and Kang 1990). Most information on quantitative genetic parameters is derived from analyses of wind-

6. BREEDING DOUGLAS-FIR

271

pollinated families collected from natural populations (as compared to advanced-generation breeding populations). Furthermore, genetic parameters differ depending on whether they are calculated among or within natural populations (e.g., Howe et al. 2003). The results presented below are based on within-population (or within-breeding-zone) analyses because these are most relevant for improving breeding populations. 1. Additive vs. Non-Additive Genetic Variance. Additive genetic variance, which is the variance of breeding values, is the main reason that progeny resemble their parents and the main determinant of a population’s response to selection (Falconer and Mackay 1996). Non-additive genetic variance consists of all other types of genetic variation, including dominance, interaction, and disequilibrium variance (Falconer and Mackay 1996). Non-additive genetic variance, which does not contribute to population improvement via recurrent selection, is the variation that causes the progeny of specific crosses to perform differently from what is predicted by the breeding values of their parents. For Douglas-fir growth traits, the ratio of non-additive to additive genetic variation is usually less than half (i.e., 50%). This is best illustrated by two large experiments in which the ratio of SCA (specific combining ability) variance to GCA (general combining ability) variance was estimated from a large number of 6-parent diallels planted across many sites in the Pacific Northwest. In one series of experiments planted in Oregon and Washington (65 6-parent disconnected diallels), the ratio of SCA to GCA variance averaged 46% for height growth at ages 6 or 8 (Stonecypher et al. 1996). In another series of experiments planted in British Columbia (36 half-diallels), the average ratio was 36% for three growth traits (age-7 height = 34%, age-12 height = 32%, and age-12 volume = 43%; Yanchuk 1996). Despite these values, Yanchuk (1996) concluded that the additional gain from using SCA (i.e., by making crosses between specific parents) would only be about 3% for volume and 1% for height (i.e., as compared to producing seed from the best parents based on GCA). Douglas-fir breeders are mainly interested in the additive genetic variance because most breeding strategies rely on improving populations via recurrent selection, and because most improved materials are produced via wind-pollinated seed orchards, which do not capture the nonadditive component. Non-additive genetic variation, however, can be captured if breeding programs deploy clones or full-sib families. Clones capture all of the non-additive variation, whereas full-sib families capture only a quarter or less of the various sources of non-additive variation.

272

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

2. Heritabilities and Amounts of Genetic Variation. Estimates of genetic gain are important for judging alternative breeding strategies and for financial analyses of breeding programs. Additive genetic gains are a function of the selection intensity (i), additive genetic variance (σ2addi2 2 tive), and the narrow-sense heritability (h ), or ratio of σ additive to total phenotypic variance (σ2phenotypic) (i.e., gain = ihσadditive). One of the most important indicators of potential breeding success is heritability, the proportion of observed variation (i.e., phenotypic variation) that is controlled by genetics. Broad-sense heritability (H2) is the ratio of total genetic variation (i.e., additive plus non-additive variation) to total phenotypic variation [H2 = (σ2additive + σ2non-additive) / (σ2additive + σ2non-additive + σ2environment)]. Narrow-sense heritability (simply called heritability, or h2) is the ratio of additive genetic variation to total phenotypic variation [h2 = σ2additive / (σ2additive + σ2non-additive + σ2environment)]. Trait heritabilities have received a great deal of attention because they integrate information on genetic and environmental variation, and because they can be altered to increase genetic gains, primarily by reducing environmental variability in genetic tests and by increasing family size to increase family heritabilities. Because heritability is the ratio of genetic to phenotypic variation, the amount of environmental variation has an important influence on breeding progress. In forestry, environmental variation is particularly large for growth traits, except when tests are conducted in carefully controlled nursery or greenhouse environments. Most progeny tests are planted on sites typical of plantation forestry, which are usually more variable than agricultural sites because of less site preparation, less site maintenance, and larger tests that encompass more land. Furthermore, heritabilities can vary considerably among progeny test sites for the same population. Johnson et al. (1997), for example, examined age-15 height in six breeding programs in which 90 to 150 families were planted across 6 to 12 sites in each program. Within each program, the heritabilities ranged enormously among sites (i.e., 0.00–0.33, 0.09–0.29, 0.05–0.39, 0.19–0.41, 0.09–0.41, and 0.07–0.18). In forest trees, heritabilities are low for most traits, typically averaging less than 0.30 (Table 6.1; Cornelius 1994). Nonetheless, traits vary in their degree of genetic control (h2) and the relative amount of genetic variation (i.e., additive genetic coefficient of variation = σadditive/mean). Some traits, such as wood density, have high heritabilities but low genetic variation, whereas other traits such as volume growth, stem defects, and some branching traits have low heritabilities but high levels of variation (Cornelius 1994).

6. BREEDING DOUGLAS-FIR

273

Table 6.1. Median and mean values of individual-tree narrow-sense heritability and additive genetic coefficient of variation (AGCV) for seven traits or trait types compiled from 67 published studies (from Cornelius 1994). Heritability Trait or trait type Height Diameter Volume Straightness Morphological and structural Specific gravity (wood density) Branching traits

AGCV

Median

Mean

Median

Mean

0.25 0.19 0.18 0.26 0.23 0.48 0.24

0.28 0.23 0.21 0.28 0.23 0.50 0.26

8.50 8.60 20.30 11.65 8.80 5.10 8.40

11.10 9.10 23.10 16.25 14.73 5.34 16.30

Douglas-fir heritabilities follow the same patterns as seen in other forest trees: low to moderate for growth traits, fall cold hardiness, drought hardiness, stem defects, and branch size—and moderate to high for wood density, bud flush, bud set, spring cold hardiness, and branch angle (Table 6.2). Douglas-fir breeding programs emphasize productivity, which is measured as height, diameter, or stem volume (estimated from height and diameter measurements). Because these are the most frequently measured traits, we know more about their heritabilities than we do for other traits. Heritabilities for height, diameter, and stem volume typically range from 0.10 to 0.30. Three studies examined the heritability of Douglas-fir height growth in detail. The heritability was 0.13 in a large series of 6-parent, half-diallels planted across 11 sites in British Columbia (Yeh and Heaman 1987). In a similar experiment of 6-parent diallels in Oregon and Washington, the heritability averaged 0.13 (0.00–0.40) at individual test sites at ages 6 to 8, and 0.16 (0.08–0.28) in other half-sib and full-sib tests at ages 8 to 12 (Stonecypher et al. 1996). Johnson et al. (1997) reported an average heritability of 0.16 (0.00–0.38) for 51 test plantations in six breeding programs. Compared to the first experiment (Yeh and Heaman 1987), the heritabilities from the latter two experiments are biased upward because they are single-site estimates in which the variation due to the genotype by environment interaction is confounded with the family variation (Stonecypher et al. 1996; Johnson et al. 1997). For growth traits, the genotype by environment interaction is typically about one-quarter to one-third as large as the additive genetic variance (Yeh and Heaman 1987; Stonecypher et al. 1996; Johnson 1997), and the

274

Table 6.2. Mean heritabilities for common traits measured in Douglas-fir progeny tests under field conditions. Only experiments in which the trees were at least 4 years old are included. Traitz HT

DIA

WD

BF

BS

2nd FLUSH

CH (spring)

CH (fall)

FK/ RB

SIN/ CROOK

BR ANGLE

BR DIA

TAPER

SNC

Age (sites)y

0.21

—

—

1.00

0.90

0.25

0.98

0.25

—

—

—

—

—

—

5–7 (m)

0.32

—

—

0.78

0.42

0.66

0.59

0.17

—

—

—

—

—

—

5–7 (s)

—

—

—

—

—

—

—

0.16

—

—

—

—

—

—

9 (m)

~0.22

~0.33

~0.74

~0.88

—

—

—

—

—

—

—

—

—

—

9–16 (s)

0.13t 0.30

— 0.31

— —

— 0.92

— —

— —

— —

— —

— —

— 0.39

— 0.49

— —

— —

— —

3–4 (m) 6–12 (s)

0.27

0.23

—

—

—

—

—

—

0.20

0.16

—

—

—

—

11 (m)

0.18

0.21

—

—

—

—

—

—

—

—

—

—

—

—

10–15 (s)

0.29

0.30

—

0.79

—

—

—

—

—

—

—

—

—

0.18

10–13 (s)

0.14

0.16

—

—

—

—

—

—

—

—

0.73

0.26

0.10

—

12 (m)

—

0.23

0.90

—

—

—

—

—

—

—

—

—

—

—

12 (s)

—

—

—

0.90

0.81

—

—

—

—

—

—

—

—

—

14 (s)

—

—

—

0.74

—

—

—

—

—

—

—

—

—

—

13–16 (m)

Citation Aitken and Adams 1995x,w Aitken and Adams 1995v,w Aitken et al. 1996 Bastien et al. 1985u Campbell 1972 Christophe and Birot 1979; Birot and Christophe 1983 Howe and Jayawickrama 2002 Johnson et al. 1997 Johnson et al. 2002 King et al. 1988a; 1992 King et al. 1988b Li and Adams 1993 Li and Adams 1993

—

0.19

—

0.87

—

—

—

—

—

—

—

—

—

—

15 (s)

~0.10s

—

—

—

—

—

—

—

—

—

—

—

—

—

12 (s)

—

—

—

0.85

0.42

—

0.77

0.21

—

—

—

—

—

—

7 (m)

— 0.17 0.15

— 0.27 —

— 0.52 —

0.52 — —

0.25 — —

— — —

— — —

— — —

— — —

— — —

— — —

— — —

— 0.24 —

— — —

4 (s) 18 (s) 6–12 (s)

—

0.08

—

—

—

—

—

—

0.08

0.13

—

—

—

—

12–13 (m)

0.10

0.12

—

—

—

—

—

—

—

—

—

—

—

—

6 (m)

0.13

—

—

—

—

—

—

—

—

—

—

—

—

—

7 (m)

z

Li and Adams 1994 Magnussen and Yanchuk 1994 O’Neill et al. 2000w Rehfeldt 1983br St.Clair 1994 Stonecypher et al. 1996 Temel and Adams 2000 Yeh and Heaman 1982 Yeh and Heaman 1987

275

Traits are total height (HT) (except as noted); stem diameter (DIA); wood density (WD); time of spring bud flush (BF); time of fall bud set (BS); second flushing (2nd FLUSH); cold hardiness (CH); forking and/or ramicorn branches (FK/RB); sinuosity or crookedness (SIN/CROOK); branch angle (BR ANGLE); branch diameter (BR DIA); stem taper (TAPER); and foliage health in the presence of the Swiss needle cast pathogen (SNC). “Sinuosity” has been used to describe two physiologically distinct traits, stem waviness in the first few interwhorls at the top of the tree (SIN, sensu Campbell 1965) and overall crookedness throughout the bole (CROOK). y Age at the time of measurement and indication of whether the heritabilities were calculated across multiple sites (m), or are applicable to single sites (s), in which case they would be inflated because the variance due to genotype by environment interaction is confounded with the genetic variance. x Two test sites in the coastal breeding zone. w Alternative analyses of some of these data are presented in Aitken and Adams (1995) and O’Neill et al. (2000). v One test site in the Cascade breeding zone. u These heritabilities were estimated from the figure on page 17 (ANNEX 8) of Bastien et al. (1985). t Total height increment during the third and fourth growing seasons. s Estimated from Figure 4 in Magnussen and Yanchuk (1994). r Nursery study.

276

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

genetic correlations among sites within breeding zones are about 0.6 to 0.8 (Johnson 1997). Because growth traits have low heritabilities, most breeding programs rely heavily on among-family selection to obtain genetic gain. Heritabilities (i.e., repeatabilities) of family means tend to be much higher (0.60–0.90) than individual tree heritabilities because families are typically planted on four or more sites and are usually represented by more than 60 progeny. As a result, advanced-generation selections are heavily weighted by the family performance compared to the within-family performance. The added benefit of replicating families over multiple sites is that it is possible to reduce the impact of genotype by environment interaction by finding families that perform well and are stable across a breeding zone (Stonecypher et al. 1996). Heritabilities must also be examined in the context of stand age. In Douglas-fir breeding programs, selections are often made between the ages of 10 and 20, well before the harvest age of 40 or more. Heritabilities for growth traits slowly increase with age (Johnson et al. 1997). Although some of this increase may be associated with intrinsic changes in the relative proportion of genetic to environmental variation, most of the increase that occurs after crown closure is probably artifactual. As stands age and inter-tree competition increases, the fundamental tenet of good experimental design (i.e., no correlation between genotype and environment) no longer holds. If the trees that grow larger during the early life of the stand do so because of their genetic makeup, then genotype is confounded with an environmental effect (i.e., degree of suppression), thereby upwardly biasing estimates of genetic variation, heritability, and gain (Yanchuk 1996). Because tree roots also spread with time, the subsurface environment may exert a similar influence. Because of competition, there is a window of opportunity during which it is possible to obtain unbiased estimates of genetic variance and heritability (i.e., before competition becomes a factor). Unless the tests are systematically thinned, estimates of genetic variance and heritabilities are likely to be inflated after crown closure occurs around age 8 to 15, depending on the site. Studies examining heritabilities for other traits are fewer, and often include fewer families and sites. Except for survival and growth, cold adaptation traits are the best-studied adaptive traits in Douglas-fir. These traits include vegetative bud phenology (i.e., bud set and bud flush), fall cold hardiness, and spring cold hardiness. In Douglas-fir and other conifers, the heritability of fall cold hardiness tends to be low to moderate, often about the same as for growth traits (O’Neill et al. 2000; Howe et al. 2003). The heritabilities of spring cold hardiness, bud set, and bud flush tend to be higher (i.e., moderate to high), with bud flush having the

6. BREEDING DOUGLAS-FIR

277

highest heritability of any measured trait in some studies (O’Neill et al. 2000; Howe et al. 2003). Heritabilities of stem defects (i.e., ramicorn branches, forks, and sinuosity) are low, about the same as for growth. Nonetheless, because these traits have high genetic variability, potential gains are large (Table 6.1). As for other tree species, the heritability of wood density is high but genetic variation is low. Therefore, proportional gains in wood density may be comparable to genetic gains in growth traits. The heritability of branch size (i.e., relative to stem diameter) is low, whereas the heritability for branch angle is high (Table 6.2). Although genetic gains are possible, these traits are not considered important enough to be treated as primary or secondary selection criteria. 3. Genetic Correlations. Genetic correlations are important because breeders may cause undesirable changes in some traits by selecting for other correlated traits, and because some traits can be used as indirect selection criteria for other traits that are harder or more expensive to measure. Selection is mostly based on growth, and adverse genetic correlations have been repeatedly demonstrated between growth and cold hardiness, bud set, ramicorn branches, and wood density (discussed below). Therefore, it might be wise for breeders to monitor these traits in their improved populations to ensure that undesirable changes are not occurring. In Douglas-fir and other species, increased growth is associated with increased second flushing, later bud set, and increased cold injury in the fall, and these adverse relationships are stronger among, as compared to within, populations (Howe et al. 2003). In contrast, there is no consistent correlation between growth and either spring cold injury or bud flush (Howe et al. 2003). Wood density consistently shows an adverse genetic correlation with growth, and this association is stronger for diameter growth than for height growth. In two studies in which the adverse genetic correlation was strong (< –0.50), wood density decreased by 3 to 6.5% by selecting for increased growth using a 10% selection intensity (King et al. 1988b; Vargas-Hernandez and Adams 1991). One reason for this modest loss in wood density is that density has a high heritability, but a small coefficient of additive genetic variation (σadditive/mean) (Table 6.1; Bastien et al. 1985; King et al. 1988b; St.Clair 1994). Increased growth is also positively associated with ramicorn branching, an important stem defect (Howe and Jayawickrama 2002). This relationship seems to exist, in part, because second flushing leads to greater seasonal height growth, but aberrant second flushing is also a major cause of ramicorn branches. Increased growth is associated to a lesser extent with a more sinuous stem and more forking.

278

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Genetic correlations are valuable when they allow a breeder to use indirect selection. Although stem volume is the primary trait of interest, selection is usually practiced on height and diameter measurements. Not surprisingly, these three traits are highly correlated, with genetic correlations usually 0.80 or higher (Birot and Christophe 1983; Yeh and Heaman 1982; Johnson et al. 1997). Nonetheless, much lower (but positive) genetic correlations between height and diameter are sometimes found (e.g., 0.45; King et al. 1988a). Other important relationships exist between vegetative bud phenology and frost hardiness. The timing of bud set in first-year seedlings (which have seasonally indeterminate growth) has a negative genetic correlation with fall frost hardiness—trees that set bud earlier have greater frost hardiness. In saplings (which exhibit seasonally determinate growth), bud set occurs much earlier in the summer and this relationship is weak (Li and Adams 1993; O’Neill et al. 2000). The positive correlation between spring bud flush and spring frost hardiness is stronger and less influenced by tree age. Trees that flush later in the spring are more frost hardy. Because of these relationships, it is possible to use bud phenology (which is easy to measure) as an indirect measure of fall and spring cold hardiness. Nonetheless, the use of bud set as an indirect selection criterion is limited because the correlation between bud set and fall cold hardiness is not very strong in older trees (O’Neill et al. 2000). Correlations between cold hardiness in the spring and fall seem to vary by population. In Douglas-fir seedlings and saplings, they were either uncorrelated (Cascade population), or had a weak to moderate negative correlation (coast population) (O’Neill et al. 2000, 2001). 4. Estimates of Genetic Gain. How much genetic gain is possible in Douglas-fir? Predicted gains in age-12 height and volume were reported for the British Columbia Ministry of Forests (BCMoF) program based on the assumption that the top 10% of parents (n = 22) are selected from all parents that were tested in four series of half-diallel tests (Yanchuk 1996). Gains in height and volume were predicted to be 8.8 and 22.4%, respectively, if the selected parents are placed in a seed orchard where mating is random and there is no pollen contamination or inbreeding. Furthermore, these gains would only increase slightly (i.e., up to 9.4 and 25.2%, respectively) if specific crosses were made to capture SCA (Yanchuk 1996). Genetic gains are expected to be even larger in the Northwest Tree Improvement Cooperative (NWTIC) program because selection intensities could be much higher (i.e., because of the large numbers of parents tested; K. J. Jayawickrama, unpubl. data). Realized gains may be lower than predicted gains, however, mostly because of

6. BREEDING DOUGLAS-FIR

279

pollen contamination (if controlled-pollination is not used), the effects of genotype by environment interactions that are not fully measured in the test plantations, and violations of other assumptions used to predict genetic gain. Results from a block-plot, realized gain trial established at five sites in the northern Oregon Cascades were recently published by St.Clair et al. (2004). Five years after planting, realized gains for selected groups of families were similar to the gains predicted from their performance in progeny tests using single-tree plots. Realized genetic gains were about 6% for height, 8% for diameter, and 28% for stem volume, compared to predicted genetic gains of about 8% for height, 7% for diameter, and 25% for stem volume. Age-7 results were reported for a trial established on five sites in British Columbia (Woods et al. 1995a; Stoehr and Bird 2003). Again, realized genetic gains were similar to predicted gains, with mid-gain (i.e., moderately selected) families showing a 10% gain in height over unimproved seedlots, and top-crosses (i.e., crosses among elite parents) showing a gain of nearly 17%. G. Sociological and Political Factors The public believes it has an important responsibility to maintain the long-term health of our forests. The public certainly feels this way about public lands in the Pacific Northwest—and sometimes extends these views to large industrial forests as well. Tree breeders must, therefore, operate within the social and political climates, as well as the physical and economic climates in which they operate (Howe et al. 2005). Despite these constraints, very large breeding programs exist for a handful of the world’s most important forest tree species, including Douglas-fir.

VI. BREEDING GOALS AND OBJECTIVES The main goals of Douglas-fir breeding programs are to improve the economic value of tree crops and maintain adaptability (Table 6.3). At the same time, most breeders also acknowledge the need to maintain flexibility in case markets (and breeding objectives) change, and to ensure that improved populations have sufficient genetic variation for continued gains in the future. Although the value of tree crops is determined on a per-hectare basis, the traits used for selection (i.e., specific breeding objectives) are typically measured on individual trees. This is because it is difficult and expensive to measure the performance of many seedlots (e.g., families) on a per-hectare basis. Nonetheless, breeders must be

280 Table 6.3.

Typical breeding goals, objectives, and traits of interest for Douglas-fir breeding programs in North America. Relative importancez

Breeding goal Increase crop value

Breeding objective (i.e., primary traits of interest)

Designing breeding zones (and populations)y

Selecting within breeding populations

Large stem volume

2

1

Stem height (y) Stem diameter (y) Estimated stem volume (y) Growth model parameters (n)

High stem quality

3

2

Ramicorn branches (y) Forks (y) Stem sinuosity (y) Second flushing (p)

Wood quality (i.e., stiff and strong wood)

3

2

Wood density (y) Microfibril angle (n) Modulus of elasticity (MOE) (n) Modulus of rupture (MOR) (n)

Small knots

3

3

Branch size (p) Branch angle (p)

Selection criterionx

Maintain adaptability

Disease resistance

z

Spring frost hardiness

1

2–3

Spring artificial freeze test (p) Spring bud flush (y)

Fall frost hardiness

1

2

Fall artificial freeze test (p) Fall bud set or growth cessation (y)

Drought hardiness

1

3

Summer artificial drought test (p)

Swiss needle cast resistance

3

2

Foliage health or diameter growth in the presence of Swiss needle cast (y)

Rhabdocline needle cast resistance

3

2

Visual assessment in the field (y)

Armillaria root rot resistance

3

2

Visual assessment in the field (y) Artificial inoculation (p)

Phellinus root rot resistance

3

2

Visual assessment in the field (y) Artificial inoculation (p)

1 = Always or almost always considered; 2 = sometimes considered, or used as a secondary criterion in two-stage selection; 3 = never or almost never considered. y First-generation breeding populations generally consist of genotypes collected from the breeding zone in which they will be used. x y = yes, trait is always or sometimes used; p = potential for use (trait is not used, but information exists to allow its use); n = no, trait has never been used and there is little information to guide its use.

281

282

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

mindful of the assumptions involved in defining breeding objectives on a per-tree basis when the real goal is to increase the value of the entire crop. A. Primary Breeding Objectives The two main breeding goals—increasing crop value and maintaining adaptability—are generally met in different ways. “Adaptability,” which usually refers to the ability to tolerate both frosts and droughts, is typically maintained through the use of appropriate breeding zones (Table 6.3). Most of the parents in first-generation breeding populations were selected from the breeding zones in which they will be used, so maladaptation should be not be a problem (i.e., damage from cold and drought should be comparable to that seen in natural populations). In contrast to adaptability, tree value is generally improved by selecting and breeding the most valuable genotypes within these well-adapted populations. Tree value is primarily determined by stem volume, and secondarily by stem quality and wood properties such as wood density. Therefore, the primary breeding objective for Douglas-fir, and most other forest trees, is to increase volume growth (Campbell 1964; Wheat and Silen 1977; Silen and Wheat 1979; Lester 1986). Greater growth would result in either greater yields at harvest, or permit the use of shorter rotations. The key traits used as predictors of rotation-age volume are diameter at breast height (DBH) and total height—typically measured anywhere from ages five to 20. Although there is interest in exploring measures of growth per se (e.g., growth model parameters) (Cherry and Howe 2004), the advantages of these approaches are not yet known for Douglas-fir. B. Secondary Breeding Objectives Improvements in wood quality (i.e., wood properties) and stem quality are secondary in importance because their impact on tree value is neither as great (nor as quantifiable) as it is for stem volume (Table 6.3). The single most important wood property is wood density (e.g., wood specific gravity) because dense wood is associated with stiffer and stronger wood, as well as increased pulp yields. Important stem defects include forks and ramicorn branches. Forked stems are often formed if the terminal leader is damaged or killed, and two branches subsequently assume equal dominance. Ramicorn branches are excessively large, upright branches. Forks and ramicorn branches either reduce stem volume and log length (by making a portion of the stem unmerchantable), or result in low-grade products because of the large knots that are formed. Stem sinuosity is stem waviness in the top few interwhorls of

6. BREEDING DOUGLAS-FIR

283

the tree (Campbell 1965). Although stem sinuosity is sometimes used as a selection criterion (Table 6.3), it has little impact on tree value because it only affects a small cylinder of wood in the center of the tree (Spicer et al. 2000). These secondary traits (i.e., wood density, ramicorn branching, forking, and sinuosity) are often considered in breeding programs, but receive much less weight than stem volume. Genotypes with many stem defects, for example, are unlikely to be included in future breeding populations (regardless of their growth rate), and selections for increased wood density are often made only among the fastest-growing genotypes (i.e., two-stage selection). Other undesirable traits include large branch diameters and upright branches (i.e., small branch angles), both of which cause large knots. These “tertiary” traits, however, are usually ignored when selection decisions are made (Table 6.3). Branch size and branch angle are clearly associated with knot size, but small differences in knot size have little impact on product value. The relatively small impact that the secondary and tertiary traits have on tree value is exacerbated by the dearth of quantitative data that could be used to derive economic weights for multi-trait selection indices. We have excellent information on the value of gains in stem volume, but little information on the value of improvements in stem quality and wood properties. Although milling studies could be used to derive accurate economic weights for breeding programs, these studies are expensive, and are unlikely to reduce the prime focus on stem volume. Because genetic gains in any single trait decrease as more traits are included in a breeding program, there is little rationale for including these tertiary traits in current breeding programs. Despite the current focus on stem volume, stem quality and wood properties should be monitored because breeding objectives can change. As new products and manufacturing processes arise—and if machine grading becomes more common—the importance of stem quality and wood properties is likely to increase. Furthermore, rotations are becoming shorter and more of the wood is being harvested from the juvenile core of the tree. Compared to mature wood produced outside of the juvenile core, juvenile wood is inferior for most solid wood products. Therefore, it may become necessary to focus more heavily on wood properties simply to maintain the overall quality of the wood that is harvested. Furthermore, we know little about the genetics of the “real” traits of interest—traits such as wood stiffness and strength. Wood density is used as an indirect measure of these mechanical properties, but other fundamental wood properties, such as microfibril angle, might be useful as well (MeGraw 2002; Table 6.3). Because breeding progress is slow, breeding objectives must be viewed within the context of the

284

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

products we intend to make in 50 to 100 years, not solely on the products we are making today. In addition to increasing tree value, most programs seek to maintain sufficient physiological adaptability and genetic variation. In fact, the existence of breeding programs specifically designed to maintain adaptability and variability is a hallmark of Douglas-fir breeding. The most important adaptive traits are cold and drought hardiness. Genotypes that can tolerate the normal, year-to-year frosts and droughts are likely to be included in advanced generations if genotypes are selected based on their superior growth in field tests at age 10 to 15. In addition, trees must be able to survive and prosper after the unusually severe frosts and droughts that sometimes occur (e.g., Duffield 1956). Unlike stem volume, adaptive traits are infrequently considered as specific traits when genotypes are selected for the next generation (Table 6.3). Instead, this breeding objective is typically pursued through the design of appropriate breeding zones. The reasons for this are three-fold. First, because severe frosts and droughts are rare, it is difficult to measure frost and drought hardiness under normal progeny test conditions. Second, it is difficult to predict how these rare events will affect production plantations and, thus, the value of improving cold and drought hardiness. Third, inexpensive, artificial tests can be used to measure cold and drought hardiness in seedling nursery tests and in the field (at least for cold hardiness), but they add complexity to testing programs. Although costs can be lessened by measuring other traits that are correlated with cold and drought hardiness (e.g., bud phenology; Stonecypher et al. 1996), this approach would be considerably less effective than selecting for cold and drought hardiness directly, except for the timing of spring bud flush, which has a strong positive genetic correlation with spring frost hardiness (O’Neill et al. 2000). Other traits that have low to moderate genetic correlations with cold and drought hardiness include the timing of fall bud set and the propensity for second flushing (Howe et al. 2003). Despite the potential for within population selection, adaptability is typically maintained through the design of appropriate breeding zones. Although frost and drought hardiness are rarely measured directly, it might be wise for breeding programs to monitor these traits because selection for increased growth (i.e., without considering adaptive traits) can reduce adaptability (see Genetic correlations) and because the rarity of severe frosts and droughts has the potential to instill a false sense of adaptive security. Few insect and disease problems have risen to the level where they form key components of Douglas-fir breeding programs. Nonetheless, tolerance to Swiss needle cast has become a breeding objective in some coastal areas where this disease is impacting Douglas-fir plantations.

6. BREEDING DOUGLAS-FIR

285

Production Breeding Inter-situ

Genetic diversity

Genetic gain

Recent studies suggest that Swiss needle cast tolerance can be improved via selection and breeding (Johnson 2002; Johnson et al. 2002). Root rots caused by Armillaria and Phellinus also cause problems, and studies are underway to determine whether there is genetic variation in resistance to these fungal pathogens (B. C. Jaquish, pers. comm.; R. Sturrock, pers. comm.). A major challenge to breeders is that the two primary breeding goals— improving value and maintaining adaptability—often conflict with one another. Increased growth, for example, is associated with increased cold damage, a greater number of stem defects, and lower wood density, particularly at the population and provenance levels (Howe et al. 2003; see Genetic correlations). Furthermore, genetic variation itself may be adaptive—and genetic gain can only be achieved by reducing genetic variation in the trait of interest (Fig. 6.3). Breeding strategies for Douglas-fir seek to balance these conflicting objectives.

Gene resource Relative size of population Figure 6.3. Tree breeding populations. The gene resource population is the population of individuals that could be included in current or future breeding populations; the intersitu population is the population of individuals in progeny tests, archives, and clone banks that are not included in the breeding or production populations; the breeding population is the group of parents that will be used to produce offspring for the next cycle of selection; and the production population is the subset of individuals from the breeding population that is used to produce propagules for reforestation (see text for details). This figure shows the inverse relationship between genetic diversity (and population size) vs. genetic gain (modified from R. D. Burdon, pers. comm.; cited in Johnson et al. 2001).

286

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

VII. OVERVIEW OF TREE BREEDING METHODS In this section, we describe the key elements of a typical conifer breeding program. In many respects, these programs resemble agronomic and horticultural crop breeding. Nonetheless, they contain elements that are often unfamiliar to breeders of maize, tomato, apple, and even landscape trees—particularly because of their long-term perspective, concentration on population improvement, and focus on maintaining genetic diversity. In fact, tree breeding is more similar to animal breeding than to many other types of plant breeding. In the absence of a breeding program, tree improvement typically involves the planting of source identified wild seed according to seed transfer guidelines or seed zone recommendations. Because we are focusing on breeding per se, we will not discuss seed zones and seed transfer guidelines in detail. Readers interested in these topics should consult reviews by Adams and Campbell (1981), Campbell (1986), Morgenstern (1996), and Randall (1996). A typical tree breeding program has four main steps, which may occur simultaneously (Wright 1976; Zobel and Talbert 1984; Morgenstern 1996). The first step is to delineate breeding zones. A breeding zone is a set of environments within which the genotypes from a particular breeding population can be safely planted—i.e., resulting in welladapted plantations that will meet forest management objectives. The second step is to develop one or more breeding populations for each breeding zone. In the first generation, breeding populations may be selected from wild stands within each breeding zone; from superior, non-local populations that have been identified based on provenance tests; or from landraces. The first approach is common for Douglas-fir within its native range, whereas the second approach has been used for Douglas-fir in Europe and the southern hemisphere. The third step is to field test the progeny of the selected parents and pursue advancedgeneration breeding within each breeding population. The fourth step is to produce genetically improved materials for outplanting. In most Douglas-fir programs, this involves establishing the best genotypes from each breeding population in separate wind-pollinated seed orchards. A. Breeding Zones Tree crops are typically planted across large, environmentally diverse landscapes, managed extensively, and harvested decades later. Compared to the loss of an annual crop, the loss of a forest crop could wipe out an investment that has lasted for decades. Therefore, biological and

6. BREEDING DOUGLAS-FIR

287

economic risks are generally greater for tree crops than for annual, agronomic crops—and maintenance of long-term adaptability is critical. The first defense against maladaptation is the design of appropriate breeding zones. A key decision in any breeding program is the appropriate number, size, and locations (or environmental characteristics) of the breeding zones. The greater the number of breeding zones, the easier it is to ensure long-term adaptability and manage genotype by environment interactions. However, breeding, testing, and selection programs must be replicated for each zone. Furthermore, seed orchards will be needed for each breeding zone, at least for programs that rely on wind-pollinated seed orchards. Therefore, one of the most important decisions in any breeding program is the appropriate size and number of breeding zones. This decision involves a balance between decreasing breeding program costs (i.e., by using fewer breeding zones) and decreasing the biological and financial risks associated with maladaptation (i.e., by using more breeding zones). B. Population Improvement and Genetic Diversity For species that are planted within their native range (like Douglas-fir), gene conservation and diversity issues are critically important. Therefore, most breeding programs rely on population improvement, rather than cultivar development. Unlike many agronomic and horticultural crops where a single genotype (or genetically homogeneous cultivar) may be planted over a vast number of hectares, most organizations seek to maintain substantial genetic variation within their breeding and planting programs. Therefore, tree breeders deal almost exclusively with populations, and rarely pursue the kinds of individual-focused, pedigree breeding that is commonly practiced in agronomic and horticultural crops. Although exceptions may exist for intensively managed, shortrotation species such as poplars (Populus) and willows (Salix), this is the rule for species such as Douglas-fir that are mostly propagated by seed and extensively managed on long rotations. Even when clonal deployment is considered, a mixture of clones is typically recommended—either as true mixtures, or as clonal blocks. This constrains the operational use of clones and genetically engineered trees. C. Early Selection Because of the long rotations in forest trees, selections must be made well before harvest age. In Douglas-fir, final selections are commonly made when the trees are about 10 to 15 years old, compared to harvest

288

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

ages of 40 to 70 years. Based on age-age correlations from more than 51 progeny test sites in Oregon, Johnson et al. (1997) concluded that peryear gains are maximized when selections are made for height at age 10, and for diameter at age 13. Magnussen and Yanchuk (1993) used stochastic simulation of Douglas-fir age-age correlations to evaluate the risks associated with different selection ages. If field tests are used to select the tallest families before age 15, they concluded that family sizes of at least 20 trees per family are needed, whereas family sizes should be above 40 if selections are made before age 10. For within-family selection, they concluded that the “safe” age for making selections is age 17 or older. Various scenarios have been investigated to further reduce the selection age (i.e., below what is optimal using standard progeny tests). These methods of “early selection” include measuring young seedlings (≤ 2 years old) in controlled environments (Lambeth et al. 1982), bareroot nurseries (Riitters and Perry 1987; Adams et al. 2001; Vargas-Hernandez et al. 2003), or container nurseries (Adams et al. 2001; Vargas-Hernandez et al. 2003). In these experiments, low to moderate correlations (e.g., 0.3–0.5) were found between family means for some seedling traits (e.g., height, dry-weight, bud set) and the height or volume of the same families in field tests at ages 6 to 15. Although these correlations are not large enough to permit final selections to be made at very young ages, they are large enough to practice “early culling” using two-stage selection. Adams et al. (2001), for example, concluded that the costs of standard progeny tests could be reduced by 18% (i.e., from $371,000 to $306,000) by culling the poorest 25% of families based on first-year height growth, then planting only the remaining 75% of families in long-term field tests. Other experiments suggest that early culling would also be effective for some branch traits (Vargas-Hernandez et al. 2003). Furthermore, to improve the tolerance of Douglas-fir to Swiss needle cast disease, early selection in the field at age two was 25 to 100% as efficient as waiting until age 10 or 12 (Temel and Johnson 2001). Another option is to practice early selection using older trees (up to age 7) in “farm-field” tests. Farm-field tests are progeny tests established using intensive site preparation, close spacing, and nearly complete weed control (Woods et al. 1995b). From ages 3 to 7, there was a high correlation (≥ 0.71) between family height in a farm-field test and both the height and volume of the same families in standard progeny tests. Unlike the very early tests described above, the farm-field tests would not be valuable for early culling. This is because the trees could not be transplanted to standard progeny tests after the test is complete (as is done with early culling). On the other hand, farm-field tests might pro-

6. BREEDING DOUGLAS-FIR

289

vide greater gains per year by allowing final selections to be made earlier, but (presumably) this would entail using them as replacements for standard progeny tests, which seems unlikely. Because it takes so long for the financial benefits of tree improvement to be realized, discount rates have a major impact on the economics of tree breeding. The financial benefits of alternative breeding programs are particularly sensitive to discount rates—and modest up-front costs must be balanced by large returns when the stands are harvested. These factors tend to favor low-cost approaches to genetic improvement and exclusion of secondary traits that have ill-defined impacts on tree value.

VIII. BREEDING PROGRAMS A. North America In North America, most Douglas-fir improvement is carried out by the Northwest Tree Improvement Cooperative (NWTIC), Inland Empire Tree Improvement Cooperative (IETIC), British Columbia Ministry of Forests (BCMoF), and Weyerhaeuser Company (Table 6.4). These four organizations are responsible for developing improved materials planted by private companies, tribal governments, and public agencies in the U.S. and Canada. The federal agencies include the USDA-Forest Service (USFS) and USDI-Bureau of Land Management (BLM). The state agencies include the Idaho Department of Lands, Montana Division of Forestry, Oregon Department of Forestry (ODF), and Washington Department of Natural Resources (WDNR). Because the approach to Douglasfir improvement differs among these four organizations, it is important to understand their organizational differences, tree improvement histories, and core philosophies. In subsequent sections, we will expand on their alternative approaches to selection, breeding, testing, seed orchards, vegetative propagation, and deployment of improved genotypes. Earlier summaries of the North American Douglas-fir breeding programs were given by Woods (1993) and Lipow et al. (2003). 1. Northwest Tree Improvement Cooperative (NWTIC). The NWTIC evolved from the Industrial Forestry Association (IFA) breeding program and the subsequent IFA-PNW Progressive Tree Improvement Program. In the 1950s, a small group of government agencies, forestry companies, and forestry associations began independent tree improvement programs (Hagenstein 1986). Booth-Kelly Lumber Company, Crown Zellerbach Corporation, the IFA, Port Blakely Mill Company, Simpson Timber

290

Table 6.4.

Major Douglas-fir genetic improvement programs in North America (updated from Woods 1993). Region and program U.S. Pacific Northwest NWTICz

Characteristic First generation No. of breeding zones No. of parents tested Mating designs No. of field tests Second generation No. of breeding zones No. of parents to be tested No. of families to be tested Mating designs

Weyerhaeusery

British Columbia BCMoF coastalx

U.S. Intermountain

BCMoF interiorx

IETICw

109 26,000u OP, polymix, singlepair full-sibs 835

6 (low elevation)v 3,500 OP, diallel, polymix, single-pair full-sibs 400

2 660 OP, half-diallel, factorial 130

8 1,661 OP

13 2,503 OP

32

43

8 2,000

3 (low elevation) 1,043

2 300

5 Not determined

Not planned Not planned

2,600t

4,886

930

Not determined

Not planned

Unstructured full-sibs

Tester, positive assortative matings, correctively-mated full-sibss 101

Polycross, halfdiallel, doublepair mating

Factorial

Not planned

36

Not determined

Not planned

Adaptability (BZ) Volume (HT, DIA) Wood quality (WD)

Adaptability (BZ) Volume (HT, DIA)

Adaptability (BZ) Volume (HT, DIA) Fall frost hardiness (BS)

No. of field tests

95

Breeding objectives Primary (selection criteria)r

Adaptability (BZ) Volume (HT, DIA)

Adaptability (BZ) Volume (HT, DIA) Stem quality (RB, FK, SIN)

Secondary (selection criteria)q

z

Stem quality (RB, FK, SIN) Wood quality (WD)

Wood quality (WD) Frost hardiness (BF) Disease resistance (NC)

Stem quality (RB, FK, SIN) Frost hardiness (FD)

Wood quality (WD) Stem quality (RB, FK, SIN)

Disease resistance (RR, NC)

NWTIC is Northwest Tree Improvement Cooperative. Data are from K. J. Jayawickrama. Weyerhaeuser data (U.S. breeding program up to 2003) are from C. A. Dean. x BCMoF is British Columbia Ministry of Forests. Coastal data are from M. U. Stoehr and interior data are from B. C. Jaquish. w IETIC is Inland Empire Tree Improvement Cooperative. Data are from M. L. Rust. v Although the Weyerhaeuser program originally had 6 low elevation and 6 high elevation breeding zones (Stonecypher et al. 1996), data for only the 6 low elevation zones were reported by Woods (1993). u Sum of the number of parents tested in each breeding zone. The actual number of parents tested is slightly less because some parents were tested in more than one breeding zone. t Including families growing in nurseries in 2004. s The number of crosses per parent was based on each parent’s relative performance and genetic diversity targets. r Primary breeding objectives are those that are almost always considered when making selection decisions. Adaptability refers to the goal of maintaining or enhancing both cold and drought hardiness (i.e., multiple breeding objectives). Selection criteria (in parentheses) refer to the traits on which selection is based. BZ indicates that the breeding objectives are mostly achieved through the design of appropriate breeding zones (and breeding populations) based on a complex suite of traits (e.g., multiple traits measured in field and nursery tests), as well as other geographic, physiographic, and climatological information (see Table 6.3 and text). Selection criteria that are typically used to select within breeding populations include stem height (HT), stem diameter (DIA), ramicorn branching (RB), forking (FK), stem sinuosity (SIN), wood density (WD), damage from natural frosts in the spring or fall (FD), timing of spring bud flush (BF), timing of fall bud set or growth cessation (BS), tolerance to Armillaria and Phellinus root rots (RR), and tolerance to Rhabdocline and Swiss needle cast diseases (NC). “Sinuosity” has been used to describe two physiologically distinct traits, stem waviness in the first few interwhorls at the top of the tree (Campbell 1965) and overall crookedness throughout the bole. See Table 6.3 for more information on breeding goals, breeding objectives, and selection criteria (i.e., traits of interest). q Secondary breeding objectives are those that are sometimes considered, or those used as secondary criteria in two-stage selection. Abbreviations are as described above. y

291

292

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Company, Timber Service Company, USFS, and Weyerhaeuser Company were among the first to select coastal Douglas-fir trees and establish grafted (i.e., clonal) seed orchards. The IFA played a key role in this process by hiring a forest geneticist, John Duffield, to guide their tree improvement efforts (Hagenstein 1986). During the 1960s, the USFS Pacific Northwest Forest and Range Experiment Station (PNW Station) played an important role in Douglas-fir improvement. In 1966, the PNW Station joined forces with the IFA to form the IFA-PNW Progressive Tree Improvement Program, which was described as a simple, low-cost program that would appeal to mediumand smaller-sized forest landowners in western Oregon and western Washington (Silen 1966b). The Progressive Program was first described by Roy Silen in 1966, and then summarized in subsequent publications (Silen 1966b; Wheat and Silen 1977; Silen and Wheat 1979). The IFAProgressive Program eventually evolved into the Northwest Tree Improvement Cooperative (NWTIC) in 1986. Although the USFS, BLM, Georgia-Pacific, Simpson, and WDNR operated independent programs for many years, these programs are now part of the NWTIC. As of 2004, the NWTIC is housed at Oregon State University and consists of 27 member organizations (i.e., forest industries, tribal governments, one federal agency, and state agencies within the U.S. and Canada). Distinctive features of the Progressive Program included the assumption that local seed sources are best (i.e., rather than choosing seed sources based on provenance tests), low-intensity selection of firstgeneration parents (rather than intensive phenotypic, or “plus-tree,” selection), use of many small breeding zones (mostly less than 60,703 ha; 150,000 acres), and the use of very large breeding populations (Silen 1966b; Silen and Wheat 1979). This conservative approach was intended to ensure adaptability of the first-generation breeding populations, which consisted of parents selected from natural stands within the breeding zone. Although breeding zones have been consolidated and breeding populations are being reduced (Table 6.4), the NWTIC program is still one of the largest tree breeding programs in the world. The Progressive Program was implemented by forming local, geographically based cooperatives to share the costs and benefits of tree improvement. The members of a particular cooperative consist of organizations that manage lands within the same geographic area and have common tree improvement needs. In the first generation, these individual cooperatives covered one to 20 breeding zones (Silen and Wheat 1979). When the establishment of first-generation tests was completed in 1993, 21 first-generation cooperatives had been formed, and the Douglas-fir zone west of the Cascades was blanketed with 109 breeding

6. BREEDING DOUGLAS-FIR

293

zones ranging from the Canadian border to northwest California. More than 26,000 first-generation parents have been evaluated based on more than 3 million progeny test trees (Lipow et al. 2003). This large number of breeding zones was recently reduced to eight second-generation zones, and the number of parents used in advanced-generation breeding is expected to be about 2,000 (Table 6.4). Today, the NWTIC serves as an umbrella organization that coordinates the breeding and testing activities for seven independent cooperatives that have breeding programs for eight second-generation breeding zones. The NWTIC is essentially a regional confederacy of autonomous, local cooperatives. NWTIC staff provides technical direction; program coordination; administrative support; and data management, analysis, and interpretation to the local cooperatives. In contrast, the costs of breeding and testing are shared among the members of these local, secondgeneration cooperatives. Seed orchards are managed by individual organizations or by seed orchard cooperatives that are mostly independent of the NWTIC. The main advantage of cooperatives is that expensive breeding programs are not duplicated among many organizations. The main drawbacks are the loss of competitive advantage (i.e., because improved genotypes are shared among cooperators) and the challenge of balancing different priorities, organizational cultures, and levels of funding (Prudham 2003). 2. British Columbia Ministry of Forests (BCMoF). Because 95% of the forestland in British Columbia is publicly owned (i.e., Crown land), private companies operate through various license agreements with strong oversight from the BCMoF. Historically, tree improvement programs were coordinated by cooperative tree improvement councils consisting of the BCMoF, forest companies, Canadian Forest Service (CFS), and universities. A Plus Tree Board was formed in the 1960s, followed by the Coastal Tree Improvement Council in 1979, and the Interior Tree Improvement Council in 1981. The Coastal and Interior Tree Improvement Councils were subsequently merged into the Forest Genetics Council of British Columbia (FGC) in 1998. The FGC is a multi-stakeholder group that coordinates and directs the operational tree improvement programs, management of seed orchards, gene conservation, and forest genetics extension activities. Within this framework, the responsibility for Douglas-fir breeding falls to the BCMoF, whereas the responsibility for producing improved materials for reforestation is shared between the forest industry and the BCMoF. First-generation selection and testing began around 1960 for coastal Douglas-fir and 1980 for interior Douglas-fir. The early improvement

294

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

efforts of Alan Orr-Ewing focused on making intensive plus-tree selections, developing inbred lines (for subsequent outcrossing), creating interracial hybrids, and testing provenances within coastal British Columbia (Orr-Ewing 1954; Orr-Ewing et al. 1972; Heaman 1977). After Gene Namkoong was hired as a consultant in the early 1970s, the focus changed to using structured mating designs (e.g., half-diallels) to create a pedigreed breeding population, estimate quantitative genetic parameters, determine the size and importance of genotype by environmental interactions, and provide family information for roguing seed orchards (Heaman 1977). In the coastal program, half-diallel, factorial, and openpollinated mating designs have been used to test about 660 parents in 130 field tests (Table 6.4). In the interior program, an open-pollinated mating design was used to test about 1,661 parents in 32 field tests (Table 6.4). 3. Inland Empire Tree Improvement Cooperative (IETIC). The IETIC was formed in 1968 to develop improved ponderosa pine (Pinus Ponderosa) for the Inland Empire (i.e., eastern Washington, eastern Oregon, northern Idaho, and western Montana). The IETIC is now pursuing genetic improvement of ponderosa pine, western larch (Larix occidentalis), western white pine (Pinus monticola), lodgepole pine (P. contorta), and Douglas-fir for planting in eastern Washington, northern Idaho, and western Montana. The IETIC is housed at the University of Idaho and consists of 19 organizations, including private industries, federal agencies, state agencies, tribal governments, and universities. The USFS is the largest landowner in the region. The Douglas-fir species group was formed in 1974, and cooperators began selecting first-generation parent trees shortly thereafter. In most of the 13 breeding zones, 200 to 300 trees were selected, and more than 2,500 first-generation parents have been field tested to date (Table 6.4). IETIC members have established four Douglas-fir seed orchards in the region. Compared to the Pacific Northwest, Douglas-fir is relatively less important in the Inland Empire and vies for attention with a number of other commercially important conifers. The number of hectares planted to Douglas-fir has declined in recent years because many sites have abundant natural reproduction and serious root diseases can be a problem in plantations. For these reasons, the Douglas-fir breeding program in the Inland Empire is much less intensive than it is in the Pacific Northwest, and no advanced-generation breeding is planned. 4. Weyerhaeuser. Weyerhaeuser Company has managed a large and sustained tree breeding program in Douglas-fir since 1963. One of their key

6. BREEDING DOUGLAS-FIR

295

assumptions was that rigorous phenotypic selection of superior trees in natural stands (i.e., accounting for competitive and environmental influences) would produce genetic gain in growth (Stonecypher et al. 1996). Therefore, the foundation of their program was an intensive plus-tree selection program in natural stands aged 25 to 80 years (Stonecypher et al. 1996). Some 3,500 parents were selected in six breeding zones (i.e., with high and low elevation splits) covering Weyerhaeuser lands in western Washington and western Oregon (Table 6.4). The primary objective of the first-generation program was to improve growth and stem quality (Woods 1993). Selection, breeding, and testing of a large secondgeneration population are almost complete, and the third generation of improvement is underway. Based on extensive field tests of adaptability and stability (i.e., low genotype by environment interaction), Weyerhaeuser significantly increased the size of their breeding zones in the U.S. There are now three low-elevation zones, one in Washington and two in Oregon. Although Weyerhaeuser typically produces improved reforestation stock from wind-pollinated orchard seed, some elite material was produced from rooted cuttings during the 1990s (Ritchie 1993). Nonetheless, the current long-term goal is to produce elite material via somatic embryogenesis and wind-pollinated seed orchards. B. Europe European interest in Douglas-fir tree improvement has grown in conjunction with its importance in European forestry. Many European countries and organizations are involved in provenance research on Douglas-fir through a comprehensive set of provenance trials established by the International Union of Forest Research Organizations (IUFRO) in the 1960s (Fletcher 2002). Seeds from 176 provenances collected throughout the range of Douglas-fir were distributed to 55 research organizations in 33 countries throughout the world (Breidenstein et al. 1990). These tests were used to identify the best provenances for Europe, and a second collection was made in the 1980s that concentrated on Washington provenances. The EUDIREC (European Douglas-fir Improvement Research Cooperative) project is a recent collaboration among research organizations in Belgium, France, Germany, Italy, Spain, and the United Kingdom (Héois 2000). Their objectives are to develop a database for European gene resources, test Douglas-fir provenances and families in the field and in controlled environments, build a common breeding population for Douglasfir, and improve methods of vegetative propagation and seed production.

296

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

C. New Zealand Douglas-fir is second only to radiata pine in importance as a plantation species in New Zealand. Large provenance trials were established in 1957 and 1959 by I. J. Thulin and results were summarized in the mid1960s and in 1974 (Sweet 1965; R. L. Knowles, pers. comm.). A Douglasfir breeding program was begun in 1970 by M. D. Wilcox of the New Zealand Forest Research Institute, and updated in 1987 by C. J. A. Shelbourne based on the most current provenance test results (R. L. Knowles, pers. comm.). In 1993, the Douglas-fir Cooperative was formed—a research-industry cooperative that is now managed by Forest Research, a Crown Research Institute (Douglas-fir Cooperative 2004). The cooperative has 12 regular members and 14 associate members, including seed producers, forest growers, manufacturers, and consultants. One objective of the cooperative is to advance technology in plantation-grown Douglas-fir by developing and deploying improved genetic materials. Although their traits of interest include growth, stem form, and needle cast resistance (Douglas-fir Cooperative 2004), the breeding program will focus on improving wood stiffness (R. L. Knowles, pers. comm.). Improved seed will be produced in clonal seed orchards, and they may use vegetative multiplication to produce operational quantities of elite material from control-pollinated seed.

IX. BREEDING AND TESTING METHODS A. Overall Breeding Strategies Each of the North American programs is using recurrent selection to improve breeding populations for specific breeding zones. Recurrent selection aims to progressively improve breeding values by increasing the frequency of desirable alleles. This approach is particularly compatible with the production of improved seed in wind-pollinated seed orchards, which is the main way that improved planting stock is produced. In the 1960s and 1970s, the BCMoF investigated other breeding methods, but these were eventually abandoned. These included a widecrossing project between “local” and “non-local” coastal parents and the development of inbred lines for outcrossing (Heaman 1977). Recurrent selection is also the foundation of the breeding programs in New Zealand and Europe.

6. BREEDING DOUGLAS-FIR

297

B. Breeding Zones—Theoretical Considerations Breeding zones and seed zones are used to manage the deployment of trees from breeding programs and wild seed collections, respectively. A breeding zone is a group of sites across which a breeding population can be planted and expected to perform well. Because long-term survival and growth are crucial for achieving all breeding objectives, breeding zones are generally delineated with these and other adaptive traits in mind. A seed zone is a group of sites from which wild seed can be collected and expected to perform well when deployed anywhere within the same seed zone. Breeding zones, not seed zones, are the main concern for organizations that have breeding programs. Some of the most important decisions in a breeding program are the appropriate number, sizes, and locations of the breeding zones. Although adaptability and genetic gain can be increased by decreasing the size and increasing the number of breeding zones, this is expensive because separate breeding programs are needed for each zone. Breeding zones and seed zones are delineated by building models that predict (1) the group of genotypes that are genetically suited to the environment at a specific planting site and (2) the group of sites that are environmentally suitable for a particular group of genotypes (Rehfeldt 1990). In seed zone delineation, the genotypes are derived from seed collected from indigenous trees that have specific locations. In breeding zone delineation, the genotypes are members of a breeding population, which do not have a specific location of origin after crossing begins in advanced generations. Within the optimal breeding zone, the genotype by environment interaction would be zero. That is, the genotypes would have the same relative performance everywhere. In the real world, however, breeding zones are delineated by balancing the size of the genotype by environment interaction against financial and other practical considerations. Discrete zones are usually circumscribed geographic areas (i.e., “circles on a map”), but could consist of a collection of non-contiguous sites that have similar environmental characteristics (Rehfeldt 1990; O’Neill and Aitken 2004). Seed transfer guidelines are alternatives to discrete seed zones that specify how far a seedlot can be safely moved from its native location to alternative planting sites, essentially defining a separate seed zone for each wild stand (i.e., indigenous breeding population; Campbell 1974a, 1974b, 1986). Furthermore, seed transfer guidelines are continuous because a certain level of risk can be associated with whatever transfer distance is contemplated. In contrast to seed zones (which are almost unlimited in number), the number of advanced-generation

298

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

breeding populations will always be modest. Therefore, discrete breeding zones are the rule. Nonetheless, the discrete nature of breeding zones is still a biological simplification because materials from multiple breeding populations may do equally well on some sites, performance and risk vary continuously both within and among breeding zones, and the relative performance of genotypes varies among sites within a zone (i.e., genotype by environment interaction is not zero). An objective way to delineate breeding zones is to (1) estimate the genotype by environment interaction among all sites within the region of interest, (2) decide on the maximum interaction to be allowed within a breeding zone (or the desired number of zones), and then (3) group the sites according to this criterion. Two vastly different approaches have been used to delineate breeding zones: direct approaches based on longterm field tests of breeding materials, and indirect approaches based on seedling tests of natural populations (i.e., genecological tests). The most direct way to delineate breeding zones is to plant long-term field tests over a large number of diverse sites, and then use this information to identify groups of genotypes and sites that have an acceptably low genotype by environment interaction. Using this approach, breeding zones and breeding populations are determined simultaneously, and there are no a priori assumptions about which genotypes constitute an appropriate breeding population, or which sites constitute an appropriate breeding zone. Two things are needed: (1) a measure of genotype by environment interaction to judge the suitability of the breeding zones, and (2) a method for grouping the sites. Direct measures of genotype by environment interaction commonly include interaction variances from analyses of variance (Stonecypher et al. 1996) and among-site genetic correlations (Johnson 1997). Sites with a low interaction variance or high among-site genetic correlation would be preferentially included within the same breeding zone. Methods used to group the sites include subjective approaches, value maximization, and various clustering techniques (Roberds and Namkoong 1989; Roberds et al. 1990; O’Neill and Aitken 2004). This direct approach is empirical and involves few assumptions. The main disadvantages are the high cost of field tests, difficulty in allocating untested sites to breeding zones, and the time it takes to obtain the long-term data needed to delineate zones with sufficient confidence. Furthermore, because the trees only experience a small sample of yearly environments, the results describe past growth, and may not predict future performance. The best information for delineating breeding zones in Washington and Oregon comes from a set of widely distributed genetic tests planted by Weyerhaeuser (Stonecypher et al. 1996). Data

6. BREEDING DOUGLAS-FIR

299

from the first-generation Progressive Program have also been used to address breeding zone questions (Silen and Mandel 1983; Johnson 1997; Balduman et al. 1999), but these analyses only included families that were planted within their first-generation breeding zones, so conclusions are limited. In the indirect approach, seedling genecological studies are used to delineate breeding zones (or seed zones). Open-pollinated families are collected from wild trees, tested in common garden studies, and the magnitude of genetic differences among populations is used to delineate breeding zones (e.g., O’Neill and Aitken 2004). In essence, the “genetic distance” between two populations in a seedling test is used as a surrogate for the genotype by environment interaction measured in long-term field tests. If two populations perform very differently in a common garden test (i.e., if they have a large genetic distance), then we assume that there would be a large genotype by environment interaction between genetic tests planted at the locations from which the seeds were collected. Differences in multivariate population means for quantitative traits (O’Neill and Aitken 2004) and allozyme markers (Westfall and Conkle 1992) have been used in indirect tests to delineate breeding zones. The indirect approach is based on some important assumptions. The main assumptions are that the (1) adaptability and productivity of local populations are nearly optimal, (2) genetic variation in the traits that are measured results from natural selection, (3) observed patterns of genetic variation reflect underlying differences in the environment, and (4) traits measured are relevant to breeding programs (Campbell 1986; Rehfeldt 1990; Morgenstern 1996). If each of these is true, then one might expect that the genetic differences measured in genecological tests would be highly correlated with the level of genotype by environment interaction that would be measured in field tests (the key criterion for delineating breeding zones). This final relationship also requires that trees respond to plantation environments and natural environments in the same way, levels of genetic resolution in seedling common-garden trials are relevant to field conditions, and select breeding populations have the same patterns of genotype by environment interaction as do natural populations. This last assumption is unlikely because stability probably increases when genotypes are chosen based on their performance across many sites (Stonecypher et al. 1996). Furthermore, breeding for stability has been proposed as a good way to increase breeding zone size (Rehfeldt 1990). Using the indirect approach, the biggest challenge is how to determine the appropriate sizes of the breeding zones (or seed zones). Campbell

300

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

developed a measure called “relative risk” that can be used to estimate the relative (but not absolute) risk of seed transfer and delineate seed zones. Relative risk is the degree of non-overlap between the frequency distributions of additive genotypes for two populations (Campbell 1986). Rehfeldt (1990) used least-significant-differences (LSDs) among population means, which he suggested should be less than 0.20 (i.e., an 80% chance of detecting a genetic difference). O’Neill and Aitken (2004) used cluster analysis to minimize the sum of squared differences among population means, but breeding zone size was ultimately determined by deciding how many zones to use. Other approaches used to delineate seed zones could also be used (e.g., differential systematics coefficient; DSC) (Parker 2000). Despite the many assumptions involved, the indirect approach has four main advantages. First, the data needed to delineate breeding zones can be acquired quickly. Second, the resulting breeding zones are likely to be conservative. Third, the number of sites (i.e., parent tree locations) that can be evaluated is enormous. A recent genecological study of Douglas-fir in Oregon and Washington, for example, included more than 1,300 families (J. B. St.Clair, pers. comm.). Fourth, the long-term adaptability of local populations is presumably guaranteed, although optimal productivity may not be. The biggest disadvantage of the indirect approach is that the relationship between the genetic distance measured in indirect tests and the genotype by environment interactions measured in long-term field tests is unknown. With field data, we can directly explore how genetic gains and financial returns change under different breeding zone scenarios, but this is not possible using indirect approaches. Instead, breeding zone size will be determined using strictly theoretical or arbitrary criteria. Other disadvantages include the fact that mature phenotypes are not measured and trees are never tested in the environments where they will be planted, so traits such as disease resistance may be ignored. Questions surrounding the appropriate size of breeding zones persist. Although some information is available from widely planted tests (Stonecypher et al. 1996), the data needed to construct optimal breeding zones via direct methods does not exist for most programs. Therefore, breeders must balance information from imperfect field tests with indirect tests that sample many parts of the region intensively. Genecological approaches are excellent for identifying the most important environmental gradients to consider when breeding zones are defined. Elevational and east-west distances, for example, should usually be given greater consideration then north-south distances. Genecological approaches are also valuable for identifying which sites are more or less

6. BREEDING DOUGLAS-FIR

301

similar to one another. Similar environments recur across the landscape and can be identified using genecological modeling (Rehfeldt 1990). Therefore, it is now feasible to develop fine-scaled, non-contiguous breeding zones using genecological analyses, geographic information systems (GIS), and advanced climate models (O’Neill and Aitken 2004). Optimally, sets of similar environments would be grouped into breeding zones based on data from long-term field tests. Genecological approaches are poor for predicting how genotypes will perform in longterm field tests and for deciding how large breeding zones should be. Indirect approaches would be more valuable if we knew the relationship between genetic distances measured in genecological tests and genotype by environment interactions measured in long-term field tests. Finally, no matter how breeding zones are delineated, they should be validated and refined by planting genotypes well beyond current breeding zone boundaries. There are no alternatives to long-term tests that intensively sample environmental heterogeneity (Rehfeldt 1990). Without this, breeding zones are likely to be overly conservative and breeding programs are likely to be much more costly than they need to be. C. Tree Breeding Populations Different types of populations are used in tree breeding, including gene resource, inter-situ, breeding, and production populations (Fig. 6.3; Burdon 1988; Johnson 1998a; Yanchuk 2001a). The size and makeup of these populations are important elements of all Douglas-fir breeding programs. The gene resource population, which is the population of individuals that could be included in current or future breeding populations, includes trees in the production, breeding, and inter-situ populations, as well as trees in the wild (i.e., in situ populations; Fig. 6.3). Because the entire species (and even closely related species) can be considered the gene resource population, this population is shared among all breeding populations. To maintain adaptability, however, most of the alleles for a particular breeding program will come from a limited geographical or ecological region. Furthermore, the use of trees from wild stands will be rare in advanced generations, and most infusions of genetic material will come from the inter-situ populations described below. Because the gene resource population is the foundation of all genetic improvement, its conservation is critical (Yanchuk 2001a; Lipow et al. 2004). The gene resource population should be large enough to maintain low-frequency alleles that could be valuable for meeting current and future breeding objectives. Because breeding programs are large, and

302

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

because Douglas-fir is a major component of western forests, Douglasfir has a vast gene resource population that contains large stores of genetic variation. The inter-situ population is the population of individuals in progeny tests, archives, and clone banks that are not included in the breeding or production populations (Fig. 6.3; Yanchuk 2001a). For the most part, each breeding population can be considered to have its own inter-situ population, which is usually the population from which the breeding population was derived. The trees in the inter-situ population provide a link between the gene resource and breeding populations—one or both parents are known and the trees will often have an intermediate level of genetic improvement. Because of the large first-generation selection and testing programs, most inter-situ populations have tens of thousands of individuals. The breeding population is the group of parents that will be used to produce offspring for the next cycle of selection (Fig. 6.3). The breeding population is associated with a particular breeding zone, which is a geographical or ecological area in which the individuals of the breeding population are expected to perform well for the traits of interest. The breeding population is usually evaluated by establishing field tests within its corresponding breeding zone. The size of the breeding population is an important factor to consider in long-term breeding. In longterm, recurrent selection programs, breeders often choose a population size that they intend to maintain over many generations. In Douglas-fir breeding, these stable population sizes have not yet been met in most programs (Table 6.4). High selection intensities are being used on very large first-generation populations to dramatically reduce the size of the first-generation breeding populations. In advanced generations, the sizes of the breeding populations are likely to be even smaller than they are in the second generation. If the breeding population is too small, genetic variation will be lost to genetic drift, and inbreeding will become a problem, but if it is too large, the breeding and testing program will be prohibitively expensive. Based on these considerations, Yanchuk (2001a) concluded that an effective population size (Ne) of only 20 to 80 is sufficient to maintain genetic variation, avoid inbreeding, and allow for multiple-trait breeding. White (1992) and Johnson et al. (2001) reviewed the literature and concluded that an Ne of 20 to 50 can sustain several generations of breeding. These are effective population sizes, however, not the actual number of individuals. For Douglas-fir, Johnson (1998a) recommended that the census number of individuals (N) should be 150–200 to maintain genetic diversity and gain for 10 generations (if

6. BREEDING DOUGLAS-FIR

303

inbreeding is managed effectively). More individuals may be needed if new traits become important in the future, especially if those traits are rare (Johnson et al. 2001). In contrast, Yanchuk (2001a) suggested that a breeding population of about 80 individuals would contain adequate amounts of quantitative genetic variation. Over many generations, the Ne of the breeding population should be monitored and unrelated genotypes should be infused as necessary (Yanchuk 2001a). Despite these theoretical considerations, operational programs tend to have breeding populations of 300–400 individuals (White 1992; Johnson et al. 2001). Within the NWTIC program, for example, most second-generation breeding populations have about 300 individuals to minimize inbreeding in sublines and seed orchards, and to allow the option of reorganizing them into multiple breeding populations in the future. Breeding populations may be divided into subpopulations to manage inbreeding, conserve genetic variation, and increase genetic gains by focusing breeding efforts on the very best selections. The advantages and disadvantages of various subpopulation approaches were reviewed by Johnson (1998a). Multiple breeding populations with different breeding objectives may be used to maintain genetic variation for a diversity of traits, thereby maintaining the flexibility to change breeding objectives in the future (Namkoong 1976). Viewed as a whole, Douglas-fir has multiple breeding populations that are being bred for optimal performance in different environments (i.e., breeding zones). In stratified breeding populations, or nucleus breeding, the breeding population is stratified into an elite population and a main population (Cotterill et al. 1988; Cotterill 1989). Most of the breeding effort is focused on the elite population (nucleus) to maximize genetic gain, and less effort is devoted to the main population, which mainly serves as an inter-situ source of genetic variation. Sublines are unrelated subsets of the breeding population that have a common breeding objective (i.e., unlike multiple populations). Sublines are used to manage inbreeding—breeding occurs within, but not between sublines, and the trees that are deployed are derived from crossing between sublines. Although inbreeding will develop in the sublines, it does not occur in the trees that are deployed. Sublines with only a small number of individuals are problematic because inbreeding will develop quickly and it is difficult to use very high selection intensities to achieve high genetic gains. If improved seed were produced in wind-pollinated seed orchards, the number of sublines should be as large as the number of orchard clones (e.g., at least 20 to 25). This will add substantial complexity to the breeding program. In contrast, if trees are deployed via control-pollinated seed orchards or vegetative

304

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

propagation, then only two sublines would be needed to avoid inbreeding. The alternative approach for avoiding inbreeding is to carefully monitor and restrict the relatedness of crosses made in the breeding population. The production population is a subset of the best individuals from the breeding population that is used to produce propagules for reforestation (Fig. 6.3). In Douglas-fir, the production population usually consists of individuals that have been grafted into seed orchards and managed to produce high quantities of open-pollinated or control-crossed seed. In other cases, the production population consists of young seedlings from elite crosses that are used to produce rooted cuttings for outplanting. In the future, the production population may consist of tissue cultures that are used to produce somatic embryos for clonal deployment. Propagules from the production population may be outplanted to the entire breeding zone, or to restricted deployment zones. Although most orchard seed is currently planted throughout its breeding zone, controlled crosses are sometimes made between the very best parents, and the resulting trees may be outplanted on only the very best sites (e.g., “high-site” deployment zone). In most cases, however, the breeding and deployment zones are the same. The size of the production population is an important consideration. Wind-pollinated seed orchards should probably contain at least 20 to 25 clones to ensure that the effective population size of the seed crop is 10 or greater (Johnson 1998a). D. Mating Designs—Theoretical Considerations Mating designs have two main functions. The first function is to provide information (e.g., breeding values, genotypic values, genetic and environmental variances) on which to select the best parents or families, and to predict genetic gains. The second function is to create new genotypes for the next round of testing and selection. Stonecypher et al. (1996) referred to the first function as the “progeny test function,” and the second as the “recurrent selection function.” Half-sib mating designs, which include open-pollinated and polycross (polymix) designs, are generally good for fulfilling the progeny test function, whereas full-sib designs, including single-pair, diallel, and factorial designs, are generally better for meeting the recurrent selection function. Because no mating design is optimal for both functions, half-sib and full-sib designs are often used together in a complementary fashion (Burdon and Shelbourne 1971). Five main types of mating designs have been used in Douglas-fir breeding: (1) open-pollinated families, (2) polycross families, (3) unstructured full-sibs (e.g., single-pair or double-pair matings), (4) struc-

6. BREEDING DOUGLAS-FIR

305

tured full-sibs (e.g., diallels and factorials), and (5) complementary designs. 1. Open-Pollinated Mating Design. Open-pollinated, or wind-pollinated, families are often used to estimate breeding values of trees in wild stands or seed orchards. Although we refer to open-pollinated families as “halfsibs,” they actually consist of a mixture of half-sibs, full-sibs, and a small percentage of selfs. The main advantage of using open-pollinated families is that they do not require controlled crosses, which are particularly difficult and expensive to do on large, widely distributed trees in the wild. Open-pollinated families are good for estimating breeding values, although differences in the makeup of the pollen pool results in some bias. Furthermore, because trees vary greatly in the proportion of selfed seed they produce, differences in selfing can affect family performance in open-pollinated genetic tests (Sorensen 1973; Sorensen and White 1988). Open-pollinated families are not good for generating new genotypes for recurrent selection because the male parent is unknown. 2. Polycross Mating Design. Polycross families are usually created by crossing a seed (assigned female) parent with a mix of pollen from a number of pollen (assigned male) “testers.” Although a single male can be crossed with a number of female testers, this approach is more complex and time consuming. Polycross families are better than openpollinated families for estimating breeding values because each female is crossed with the same pollen pool. They are similar to open-pollinated families in that the male parents of the resulting progeny are usually unknown. For this reason, they are rarely used to generate new genotypes for recurrent selection. This situation may change, however, now that highly variable genetic markers are available for Douglas-fir and other tree species (Slavov et al. 2004). Using SSR (simple sequence repeat) markers, for example, it may be possible to test polycross families in the field, make provisional forward selections based on the combined performance of the maternal family and individual tree, determine the male parent of the provisional selections using genetic markers, then make final selections by considering the performance of both the male and female parent (Lambeth et al. 2001). Using this approach, the number of parental combinations that are tested in the field is much larger than it is using full-sib crosses. Therefore, one should have a greater chance of finding offspring that result from crosses between the very best parents. Nonetheless, this approach remains untested and may not work well for subline breeding in advanced generations (i.e., once

306

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

the genotypes in the subline become related). In subline breeding, the testers should come from outside of the subline so that the parents are evaluated using outcrossed progeny, rather than closely related testers. If the testers are not in the subline, however, the polycross progeny will not be suitable for recurrent selection because the selections will not conform to the subline structure. Furthermore, relatedness among the trees in the subline may eventually hinder one’s ability to determine the paternal parent using genetic markers (Lambeth et al. 2001). In addition to providing good estimates of parental breeding values, polycross designs can be used to select which full-sib families to use for making forward selections (i.e., by calculating mid-parent values). Even when dominance is present, polycrosses are nearly as effective as fullsib families for estimating the performance of full-sib crosses (Johnson 1998a). 3. Unstructured Full-Sib Mating Designs. Unstructured full-sib families result from crosses among parents that are not organized according to a structured mating design such as a factorial or diallel (discussed below). In forest trees, most unstructured full-sib designs include each parent in only one or a few crosses (e.g., single-pair or double-pair matings in which each parent is used in one or two crosses). Single-pair or doublepair mating designs are often used because they generate new, pedigreed genotypes for recurrent selection, and are both easy and flexible. The choice of male and female parents is often opportunistic (e.g., based on which trees are flowering), rather than being pre-planned. Their main advantage is that crossing programs can be completed quickly, thereby maximizing genetic gain per year. In particular, the limited availability of male or female flowers is much less of a constraint than it is for structured full-sib designs. The parents can be mated randomly, or based on either assortative or nucleus mating, in which the better parents are used in more crosses (Johnson 1998a). The main drawback to single-pair and double-pair matings is that they are not as good for estimating the general combining ability (GCA) of the parents or the specific combining ability (SCA) of the cross (because each parent is used in only a few crosses). Nonetheless, Johnson (1998b) used simulations to conclude that gains from backwards selection for GCA increase very little when more than two or three crosses are used per parent. Even with substantial non-additive genetic variation (e.g., simulations in which the dominance and additive variation were assumed to be equal), three crosses per parent were enough to estimate breeding values and select parents reasonably well (Johnson 1998b).

6. BREEDING DOUGLAS-FIR

307

4. Structured Full-Sib Mating Designs. Structured full-sib designs, such as diallels or factorials, can also be used to produce new pedigreed genotypes for selection, but are better than the unstructured designs for estimating GCA and SCA. Instead of complete designs, partial disconnected diallels and disconnected factorials are typically used in forest trees because large, complete designs are usually not practical. Using these designs, breeding values can be estimated that are mostly free of non-additive genetic effects and new genotypes can be selected that have known male and female parents (Stonecypher et al. 1996). Although estimates of GCA and SCA may be biased by epistasis, dominance, and linkage disequilibrium, these effects should be small and not a serious problem in applied breeding programs (Stonecypher et al. 1996; Yanchuk 1996). The main disadvantages of full-sib designs are their complexity and the time they take to complete. Insufficient flowering, for example, usually makes it impossible to complete all crosses in a single year. Furthermore, for designs with many crosses per parent, many full-sib families will be produced, but relatively few unrelated progeny. 5. Complementary Mating Designs. Because single mating designs are not optimal for meeting all breeding objectives, a combination of complementary designs may be desirable (Burdon and Shelbourne 1971). The most common complementary approach is to combine openpollinated or polycross families with unstructured full-sib families (Cotterill and Jackson 1989). The main advantage of this approach is that multiple objectives can be met without using the time-consuming and complex factorial or diallel designs. Complementary approaches combine the advantages of each mating design—the ability to estimate breeding values using open-pollinated or polycross families, and the production of new pedigreed genotypes using unstructured full-sibs. The main disadvantage (i.e., compared to using unstructured full-sib families alone) is the added cost of creating and testing the openpollinated or polycross families. Assuming that open-pollinated seed orchards are used, Cotterill and Jackson (1989) predicted that gains would be greatest using half-diallels, a little less using a complementary combination of open-pollinated families and single-pair matings, and lowest using single-pair matings alone. They emphasized that the complementary design would probably produce the greatest gains per year. Furthermore, the choice of design depends on financial considerations and economic assumptions, as well as how the mating design affects field testing (discussed in Johnson 1998a).

308

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

E. Field Test Designs—Theoretical Considerations Field tests are used to meet many objectives. Most importantly, they are used to estimate the genetic worth of individual trees and families. Depending on the method of deployment, field tests should focus on accurately estimating breeding values, genotypic values, or the mean performance of specific crosses. When field tests are used to identify the best families (e.g., to select the best parents based on progeny tests), they should be designed to accurately assess family performance across a range of sites. In contrast, when field tests are used for within-family selection, they should be designed to facilitate the direct comparison of trees from the same family. This can be done using large, unreplicated family blocks, or by clonally replicating each genotype and testing them in a replicated field design. The mating design used may influence the choice of field design, particularly in advanced generations. When a complementary design is used, for example, it may be unnecessary to test the full-sib families in replicated field tests unless the goal is to capitalize on SCA. By planting the families in unreplicated family blocks, within-family selection may be easier and more effective than when members of a family are scattered among plantations and blocks. A detailed discussion of the interactions between mating designs and field tests was given by Johnson (1998a). Field tests are also used to estimate genetic and environmental variances, genetic correlations, heritabilities, and genetic gains. These objectives are usually secondary, however, because this information already exists from many past studies. Finally, field tests are used to characterize genotype by environment interactions and provide information that can be used to refine breeding zone boundaries. Field designs for forest trees must address three main challenges: (1) substantial environmental heterogeneity within and among planting sites, (2) large numbers of genotypes to be tested, and (3) mortality. These challenges are usually met by establishing tests across a range of sites, keeping the blocks (replications) small, and planting enough trees per genotype that reasonable amounts of mortality can be tolerated. Field tests must be established on multiple sites that represent the range of environments in the deployment zone. This is important for evaluating the extent of genotype by environment interactions and the stability of specific genotypes. It is also helpful to establish tests beyond the expected deployment zones so that breeding zone boundaries can be refined. By comparing predicted gains under different test scenarios, Johnson (1998a) concluded that four sites should be adequate for evaluating the height growth of families in an existing breeding zone. If

6. BREEDING DOUGLAS-FIR

309

only three sites provide usable data (i.e., if one site is lost), the expected gain is still 85% of what is expected from using six sites. The number of sites needed would increase when new breeding zone boundaries are being evaluated. Based on empirical data and theoretical considerations, Cotterill and James (1984) concluded that 10 to 20 trees should be tested per family, depending on the heritability of the trait. This number is a balance between providing enough trees to estimate family means, and enough families to practice among-family selection. In programs that seek to characterize genotype by environment interactions and evaluate breeding zone boundaries, each site must be able to stand alone. Therefore, 10 to 20 trees are needed per family on each site. Because of mortality, at least 20 trees should be planted per site to ensure that 10 to 20 healthy trees are available when the tests are measured in later years. F. First-Generation Strategies 1. First-Generation Breeding Zones. Because western North America is mountainous and environmentally heterogeneous, the first-generation breeding zones were small. In the southeastern U.S., for example, breeding zones are much larger because of greater environmental homogeneity and less pronounced local adaptation. Most breeding zones in Oregon and Washington were established as part of the Progressive Program based on the assumption that local populations are best, and genotypes are adapted to a narrow range of environmental conditions. Compared to other programs, the number of first-generation breeding zones (109) was huge. Because information on patterns of genetic variation was “fragmentary,” breeding zones were based on geographical, ecological, climatic, and land ownership patterns, rather than on genetic data (Silen 1966b). Although small breeding zones ensured adaptability, they also ensured that regional Douglas-fir breeding would be very expensive. First-generation zones were defined as ecologically similar units having a size less than 60,700 ha (150,000 acres) and an elevational range less than 304 m (1,000 ft), breeding populations were selected from wild stands within these zones, and the breeding populations were tested within these same zones (Silen and Wheat 1979). Because the breeding zones and populations were chosen a priori, different breeding populations were not systematically tested in the same experiment or outside their zone of origin, although over 1,400 families were included in tests in other (mostly adjacent) breeding zones. The Weyerhaeuser program covers a much smaller and less heterogeneous land base than the other programs—about 0.9 million ha (2.2

310

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

million ac.) in western Washington and Oregon in 1996 (Stonecypher et al. 1996). Compared to the Progressive Program, Weyerhaeuser used larger breeding zones, resulting in only 12 first-generation breeding zones—six geographical areas with two elevational zones in each area. Their advanced-generation breeding program concentrates on the six low-elevation zones (≤ 610 m) (Stonecypher et al. 1996; Table 6.4). In contrast to the Progressive Program, Weyerhaeuser undertook a wide testing program to test their first-generation breeding zones. Breeding zones for Rocky Mountain Douglas-fir were delineated based on genecological studies (Rehfeldt 1979a, 1982, 1983a). The IETIC established 13 first-generation breeding zones consisting of seven large regional zones in eastern Washington, northern Idaho, and western Montana, with elevational bands within each regional zone. The BCMoF interior program used results from Rehfeldt’s genecological studies and biogeoclimatic classifications to establish eight first-generation breeding zones, with separate first-generation testing programs within each zone (Woods 1993). The original breeding zones for the coastal BCMoF program were based on existing seed zones, but were later merged into two large first-generation zones—a low-elevation maritime zone and a zone covering the transition area between the coast and interior (Woods 1993). The decision to expand the breeding zones was based on results from provenance tests indicating that genetic differences among coastal provenances were not strong and maladaptation was not a problem (Woods 1993). 2. Selection of First-Generation Parents. Once first-generation breeding zones were delineated, parents were selected for the first-generation breeding populations. Each North American program decided to focus on selecting trees from wild stands within their respective breeding zones. That is, no program relied solely on exotic seed sources, although parents from western Washington were tested in the coastal BCMoF program. The methods used to select the parents ranged from non-intensive, mostly random selection of trees, to intensive methods of phenotypic selection in the field. In all cases, the selected parents were widely separated to avoid selecting related trees, and efforts were made to sample the entire breeding zone (i.e., all potential reforestation environments). The Progressive Program (now NWTIC) emphasized low-intensity, “roadside” selections and immediately began testing the parents using open-pollinated seed collected from the trees in the wild. This approach was taken because intensive phenotypic (mass) selection was judged to be ineffective and perhaps risky (Silen 1966b), and rapid progress was important. This approach was designed to be a quick, low-cost way of

6. BREEDING DOUGLAS-FIR

311

initiating tree improvement on a large scale. Rather than relying on phenotypic selection in the field, the main objective was to quickly select about 3 trees per 405 ha (1,000 acres) of forestland, then test the parents using open-pollinated progeny tests. Phenotypic selection was not used because the trees were selected from natural stands that had a great deal of variability in microsite and stand density. The only selection criteria were that dominant, well-formed trees with cones were chosen. A novel aspect of the Progressive Program was that crossing was done on field selections to produce full-sib families for seedling seed orchards. In contrast to the Progressive approach, other North American programs used intensive phenotypic selection in the field. Selections were typically chosen based on the growth rate of the tree relative to its neighbors, and scions (stem cuttings) were grafted into seed orchards or clone banks where full-sib families were produced for progeny testing. The latter approach delayed progeny testing, but was designed to achieve quick gains from mass selection in the field. Based on the assumption that heritabilities would be relatively high in even-aged stands of Douglas-fir, the BCMoF coastal program (Alan Orr-Ewing) began making intensive phenotypic selections in 1957 (Heaman 1977; Yeh and Heaman 1987). These plus-trees were grafted into clonal seed orchards and breeding arboreta, and became available for crossing in the early 1970s (J. C. Heaman, pers. comm.). The IETIC used a comparison-tree approach based on relative diameter growth and seed availability to make their field selections (Fins 1983). Weyerhaeuser used a similar comparison-tree approach that emphasized diameter growth per unit of growing space, but also considered stem form and crown quality (Stonecypher et al. 1996). The selected plus-trees were grafted into seed orchards and clone banks in the 1960s, and a backup orchard was established in 1973. Stonecypher et al. (1996) reported that plus-tree selection by Weyerhaeuser resulted in a 4.5% improvement in juvenile height compared to the unselected controls, which could translate into a 9% gain in height for seedlings derived from a clonal seed orchard of the selected parents (i.e., in the progeny of superior × superior parents). The first phase of the European and New Zealand programs was to find the best provenance via provenance testing. These programs are now in their second phase—selecting the best parents from the best provenances. In New Zealand, analyses of 13-year-old provenance tests demonstrated that coastal seed sources from southern Oregon and northern California grow the fastest (R. L. Knowles, pers. comm.). Earlier breeding efforts that focused on Washington seed sources were de-emphasized, and trees from the fog-belt of southern Oregon and northern California were

312

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

chosen to form most of their breeding population. Focusing on these provenances, about 400 trees were recently selected from (1) commercial plantations of known origin, (2) older genetic tests, (3) clonal archives, and (4) recent open-pollinated progeny tests derived from new seed collections in North America (R. L. Knowles, pers. comm.). As in New Zealand, forward selections from provenance-progeny tests will play a major role in Douglas-fir breeding in Europe. The large network of IUFRO provenance-progeny trials in Europe will provide a diverse genetic base for advanced-generation breeding. 3. Mating Designs in the First Generation. Each of the North American programs included some open-pollinated families in their testing program (Table 6.4). The less intensive IETIC and interior B.C. programs relied solely on open-pollinated tests, but the larger breeding programs (i.e., NWTIC, BCMoF coastal program, and Weyerhaeuser) used additional mating designs to meet some of their objectives. In addition to open-pollinated families, the Progressive Program used single-pair matings to generate full-sib families to plant in seedling seed orchards. Their plan was to quickly make roadside selections, test the parents using open-pollinated progeny tests, cross the selected parents using single-pair matings, establish full-sib seedling seed orchards, then periodically rogue the orchards as increasingly reliable information became available from the open-pollinated tests (Silen and Wheat 1979; Silen and Wanek 1986). Seedling seed orchards were used to circumvent the problem of graft incompatibility, which was a serious problem in the clonal orchards that were about eight years old at the time. In the early 1970s, the BCMoF began making crosses in their breeding arboreta. Although they began using a factorial (NC II) design and single-pair matings, they quickly switched to half-diallels because this design can be used to meet both primary objectives (J. C. Heaman, pers. comm.). The current foundation of the BCMoF coastal program is a large and ambitious series of half-diallels that were planted over many years and planting sites. Their basic design is a 6-tree half-diallel that consists of 15 crosses (i.e., no selfs and no reciprocal crosses; Heaman 1977; Woods 1993). From 1976 to 1986, 62 disconnected, half-diallels were planted on 88 test sites. The diallels were grouped into eight series, and each series was tested on 11 sites (Heaman 1977; Yanchuk 1996). Some diallels were included in more than one series to facilitate among-series comparisons. This core program was augmented by other designs. Openpollinated families and polycrosses were used to test other parents from western Washington and British Columbia, factorial designs were used to evaluate wide crosses, and selfing was used in an attempt to develop

6. BREEDING DOUGLAS-FIR

313

highly inbred lines for outcrossing (Heaman 1977; Woods 1993). In contrast to the coastal program, the less intensive interior program relied solely on open-pollinated families to estimate breeding values. Weyerhaeuser began making crosses in their grafted seed orchards in the late 1960s (C. A. Dean, pers. comm.). They began testing openpollinated families, polycrosses, and single-pair matings, but switched to 6-parent diallels in 1980 (Woods 1993; Stonecypher et al. 1996). Their first-generation field tests were planted from 1969 to 1985 (C. A. Dean, pers. comm.; Stonecypher et al. 1996). The New Zealand program began using a polycross design to estimate breeding values and full-sib crosses to produce new genotypes for the next round of selection. This approach, however, was unsuccessful because not enough seed could be produced in their clonal archives. Therefore, a revised breeding plan was developed that uses open-pollinated seed to meet both of these objectives (R. L. Knowles, pers. comm.). 4. Field Designs in the First Generation. Virtually all first-generation tests were established as randomized, complete block designs that were planted across multiple sites. To reduce block size in the NWTIC tests, large numbers of families were grouped into smaller “sets” of about 30 families each, and the sets were planted in balanced, complete block designs. The families were assigned to these sets randomly, or by grouping them according to geography (i.e., parent tree location). The most common design in the NWTIC program was the “reps-in-sets” design in which the blocks (or replications) belonging to each set were grouped together. Because the blocks were planted adjacent to one another on a site, each set was essentially a separate experiment. In the less common “sets-in-reps” design, the sets were randomized within blocks, which makes it easier to compare families in different sets. For both designs, a typical NWTIC test might consist of 150 to 300 families planted in 3 to 4 blocks of 4-tree noncontiguous plots at each of 8 to 10 sites. On average, 12 to 16 trees per family were planted at each site, and a total of 96 to 120 trees per family were tested in each experiment. The BCMoF coastal program tested a large number of half-diallels by allocating them to eight series that were planted over a 10-year period. In most cases, each series had about 150 families planted on 11 sites using 4 blocks of 4-tree row-plots, and the families were fully randomized within each block (i.e., the families within a diallel were not planted adjacent to one another; Fu et al. 1999). In some of the early tests, 2 blocks of 9-tree row plots were used (Yeh and Heaman 1987). The IETIC planted their first progeny tests in 1982, usually using a “sets-in-reps” design. Families were randomly assigned to sets, and

314

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

three test sites were established per elevation band in most zones. At each site, each family was planted in 4 blocks using a randomized, 9tree, non-contiguous plot design. In some tests, the fourth block was planted using 9-tree family block-plots, rather than in non-contiguous plots. The basic Weyerhaeuser field design consisted of a randomized complete block design with 8 blocks of 4-tree, non-contiguous plots. The main exception was that 8-tree plots were used for a common set of nonselected “control” seedlots that were planted at each site within a region. In 1977, they switched to using an interlocking block design that allows trees to be thinned while still maintaining an equal representation of trees per family. Other field designs that included as many as 151 families and 20 test sites were used to study genotype by environment interactions (Stonecypher et al. 1996). The intensity of site preparation for field tests varies from complete stump removal and intensive weed control to less intensive management. Fencing is essential to protect seedlings from the deer, elk, rabbits, hares, porcupines, and mountain beaver. Fencing costs are a major financial burden for field testing programs. Measurements in long-term field tests are typically made every three to five years, and often include stem height and diameter, ramicorn branches, forks, and stem sinuosity. The presence of root rot and needle cast disease has been measured in some tests where these diseases are prevalent. Short-term, “farm-field” tests and adaptability tests have been used to a limited extent. The IETIC established short-term trials on mild agricultural sites. These tests were designed for a final measurement three to four years after planting. In some tests, height was measured periodically during the growing season to determine the timing of height growth cessation (i.e., date of bud set). These data were used as an indirect measure of fall frost hardiness. G. Second-Generation Strategies Methods of advanced-generation breeding and testing are usually different from those used in the first generation. Breeding zone boundaries have been changed, open-pollinated mating designs are rarely used, selection at the family level is lower, within-family selection becomes more important, and field test designs evolve as data and experience accumulate. 1. Second-Generation Breeding Zones. Most of the second-generation programs are increasing the size of their breeding zones because infor-

6. BREEDING DOUGLAS-FIR

315

mation now suggests that the first-generation zones were unnecessarily small. Changes include everything from a dramatic reduction (NWTIC) to no decrease in the number of breeding zones from the first to second generation (coastal program, BCMoF). Based on a review of the literature, Randall (1996) and Randall and Berrang (2002) concluded that seed zones could be expanded in a northsouth direction, which suggested that the sizes of the breeding zones could be increased as well. At the same time, Weyerhaeuser demonstrated that a substantial number of their families perform well on mild, low-elevation sites across large areas (Stonecypher et al. 1996). Based on tests designed to measure genotype by environment interactions, Stonecypher et al. (1996) concluded that it is doubtful whether separate breeding zones are needed for Weyerhaeuser lands within Oregon and Washington. Instead, they suggested that improved families could be deployed on their low-elevation lands based on parental performance and stability, rather than on their first-generation breeding zones. NWTIC field tests were also used to examine first-generation breeding zones. Johnson (1997) analyzed genetic correlations among field tests within six breeding zones, and concluded that the first-generation zones were not too large. Balduman et al. (1999) analyzed the relationship between parent tree location and cold hardiness traits in two breeding zones and came to the same conclusion. Because different first-generation breeding populations were not systematically tested in the same experiment or outside their zones of origin, these studies are mostly suitable for determining how much smaller the first-generation zones should be, rather than how much larger. Nonetheless, because some families were planted outside their zone of origin, there is some potential for examining the consequences of larger breeding zones in the future. In contrast to the results discussed above, Campbell (1992) detected significant genotype by environment interactions among field tests within several breeding zones in Oregon. Stonecypher et al. (1996) also found significant genotype by environment interactions within Weyerhaeuser breeding zones, but less than 20% of the families contributed appreciably to the interaction, and there was no relationship between family height rank and stability. Although most results from these direct field tests suggest that larger breeding zones are appropriate, these larger zones are provisional and should be validated because none of these studies used rotation-aged trees. In contrast to the results from direct field tests, much of the genecological research that was completed from the mid-1970s until the early 1990s emphasized the importance of local adaptation. Although these indirect studies were sometimes used to justify small breeding

316

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

zones, the results from long-term field tests that suggested otherwise could not be ignored. The NWTIC merged about half of their first-generation zones (109) into eight new zones for the second generation (Table 6.4), and will focus on identifying broadly adapted genotypes in future generations. The remaining, mostly high-elevation zones and zones comprising federal lands in southwestern Oregon, were omitted because no secondgeneration breeding is planned. The lands now covered by the secondgeneration programs do not extend above 900 m (about 3,000 feet). Because environmental and genetic gradients are less pronounced latitudinally, the new breeding zones are much longer in a north-south direction. Furthermore, because they tend to follow bands of similar elevation, there are no elevational divisions within each zone. Today, there are seven “metacooperatives” that manage breeding programs for these eight zones (a metacooperative is an amalgamation of multiple firstgeneration cooperatives). Selections from multiple first-generation breeding programs have been combined to form the second-generation breeding population for each new zone (about 300 selections per secondgeneration breeding population). Furthermore, the field tests for these selections are designed to test the validity of these larger secondgeneration zones. Based on their first-generation wide-testing program, Weyerhaeuser reduced their number of low-elevation breeding zones from six to three (one in Washington and two in Oregon; Table 6.4). In Washington, four first-generation breeding zones were consolidated into a single secondgeneration zone, and the original first-generation breeding populations were organized into four sublines. In Oregon, the two low-elevation breeding zones were retained. The BCMoF is retaining two breeding zones for their coastal program, and the interior program has reduced their breeding zones from eight to five. Because the IETIC has no plans for advanced-generation breeding in Douglas-fir, they will continue to use their first-generation breeding zones, which now stand at thirteen (Table 6.4). 2. Mating Designs in the Second Generation. In the NWTIC program, the parents of the second-generation breeding population include firstgeneration parents (i.e., “backward selections”) and “forward selections” from wind-pollinated progeny tests and full-sib seedling seed orchards. At the parent or family level, a 10% selection intensity is being used, mostly based on age-15 height. Wood density and stem quality are also being considered when this information is available.

6. BREEDING DOUGLAS-FIR

317

Each second-generation breeding population of about 300 parents is subdivided into about 10 sublines of 20 to 40 individuals to manage inbreeding and conserve genetic variation. These sublines are designed to contain individuals that are genetically similar with respect to adaptive traits (e.g., from the same first-generation breeding zone). Other “elite” sublines are also being created that contain the top 10 to 20% of all parents in the second-generation breeding population, but without regard to geographic origin. Within each subline, an unstructured, fullsib family mating design is being used for both the progeny test and recurrent selection functions. Each parent is being crossed with two or three other parents to estimate parental GCA and to permit selection at the family level. If information is needed to rogue a wind-pollinated seed orchard, then three crosses are being used per parent because no complementary GCA tests are planned. Across all second-generation programs, about 2,600 crosses and 95 test plantations are planned. A complementary mating design is not being used because breeding values can be estimated from a small number of full-sib crosses when the parents are unrelated (Johnson 1998a). When inbreeding becomes a problem in advanced generations, a complementary mating design of full-sib crosses and polycrosses may be used. After this strategy was developed, Lambeth et al. (2001) suggested that polycross breeding with paternity analysis would be a good alternative to a complementary mating design. A pilot study of this approach would be valuable for Douglasfir. Weyerhaeuser used a somewhat similar approach in their secondgeneration program. Forward selections from their first-generation progeny tests (i.e., full-sib and half-sib tests) were allocated to sublines within each of three breeding zones, then crossed using a complementary mating design involving test crosses, positive assortative matings, and correctively-mated full-sib families. These mating designs were used to (1) estimate genetic parameters and develop efficient breeding strategies; (2) estimate the breeding values of the parents, full-sib families, and individual progeny; (3) produce new genotypes for the next cycle of selection; and (4) predict genetic gain. The number of crosses for each parent was determined by its relative performance and by genetic diversity targets. The controlled crosses began in 1991, the first field tests were planted in 1994, and the last mainline tests were established in 2003 (C. A. Dean, pers. comm.). The BCMoF coastal program is using a complementary mating design in their second generation. Forward selections were made in their 6parent diallel tests and the resulting breeding population was divided

318

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

into sublines. Crossing began in the late 1990s using about 15 full-sib crosses per subline. In addition, the breeding values of the forward selections are being estimated using polycrosses (10 parents per pollen mix). Field tests for 16 sublines have already been established in the field. 3. Field Designs in the Second Generation. The NWTIC is using two types of field designs in their second-generation breeding program—a design for ranking families and making selections, and another for testing long-term stability. The family-ranking and selection tests are designed to give precise estimates of family means and genotype by environment interactions. Most of these tests use an “alpha” design, which is a type of incomplete block design (Patterson and Williams 1976). A smaller number of these tests were established using a sets-inreps design. The boundaries of both the first-generation and secondgeneration breeding zones were considered when the test sites were chosen. Overall, five to six test plantations are being established in each new, second-generation breeding zone, with 20 trees per family per site planted in single-tree plots. For the tests that are already in the ground, the number of full-sib crosses ranges from 143 to 283. The trees will probably be measured twice—about seven and 12 years from seed, or when the trees are about 15 and 30 feet tall. Measurements for all NWTIC tests will include height, diameter, stem form, and (perhaps) cold hardiness and the timing of bud flush. The NWTIC long-term stability tests are designed to test adaptability and stability of families over half a rotation (i.e., > 25 years). These tests are being established using a sets-in-reps design (with sets consisting of families from the same subline, or first-generation breeding zone), and interlocking blocks that will leave a balanced number of trees in each family after thinning. In addition to testing long-term stability, these tests will be used for ranking families and making selections. The field tests used in Weyerhaeuser’s second-generation program were designed to test a total of 96 individuals from each parent or fullsib cross across three to four field sites. All sites were intensively prepared, fenced, and maintained to obtain high-quality genetic data. The BCMoF is using two field designs, one for each component of their complementary mating design. The full-sib families are being planted in 25tree blocks on each of two, intensively prepared sites (i.e., including removal of stumps). The polycrosses are being established on standard forest sites. The full-sib family blocks and GCA tests are being established at the same time.

6. BREEDING DOUGLAS-FIR

319

X. PRODUCTION OF IMPROVED MATERIALS FOR REFORESTATION A. Introduction Breeding strategies must mesh with the methods used to produce genetically improved trees for reforestation. Virtually all Douglas-fir breeding programs focus on recurrent selection for general combining ability. This approach is largely dictated by the fact that improved Douglas-firs are almost exclusively produced via seed from wind-pollinated seed orchards. Breeding strategies are likely to change if other production approaches gain favor. Production via rooted cuttings is the preferred method in species that are easy to root, such as Populus or redwood (Sequoia sempervirens), but rooting is inefficient and costly in Douglasfir. Other production approaches are biologically possible and used in a few species (e.g., loblolly pine, Pinus taeda), but are uncommon in Douglas-fir and most other species. These include controlled mass pollination (CMP), supplemental mass pollination (SMP), and somatic embryogenesis.

B. Seed Orchards 1. Conventional Seed Orchards. Improved seedlings of Douglas-fir are typically grown from seed produced in conventional, wind-pollinated seed orchards. Conventional orchards consist of large (≥ 15 m tall), widely spaced trees that are intensively managed to produce large amounts of high-quality seed. These orchards can remain productive for decades, and yields of 28 kg of seed per hectare (25 pounds/acre) are not uncommon (Cress and Daniels 1990). In 1990, seed orchards of coastal Douglas-fir covered nearly 1,012 hectares (2,500 acres) in the Pacific Northwest, with a predicted cumulative reforestation potential of 4.45 million hectares (11 million acres) by the year 2000 (Cress and Daniels 1990). Seed orchards are also common in other regions of the world. At least 13 clonal orchards covering 120 hectares have been established in France, and several have been established in Belgium, Denmark, and Germany (Chollet 1986; Héois et al. 1995). Although other production approaches may gain favor in the future (e.g., miniaturized seed orchards, rooted cuttings, and somatic embryogenesis), conventional seed orchards will remain the dominant approach for many years to come. In the Pacific Northwest, production of orchard seed began in the Dennie Ahl orchard in 1962 (Wheat 1966), and by 1986, about 50% of

320

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

the seed requirements for the Progressive Program and 30% of the requirements for British Columbia were met with genetically improved seed (Cafferata 1986; Arnott 1986). Seed production has increased over time, with thousands of pounds of orchard seed currently produced each year. Although data are unavailable for Oregon and Washington, the production of orchard seed in British Columbia was 344 kg (for coastal Douglas-fir) and 6 kg (for the interior variety) in 2003 (Anon. 2004). Weyerhaeuser has been deploying genetically improved orchard seed since 1980, and all of their seedlings have been grown from improved seed since 1986. Today, the same is true for many other organizations as well. On productive orchard sites, wind-pollinated seed can be produced for about 0.7 cents each (S. Lipow, pers. comm.), which translates to only 1–2% of the cost of planting Douglas-fir, including the costs of site preparation, seedling production, and planting. Thus, windpollinated orchards are a highly cost-effective way to boost the productivity of Douglas-fir plantations. Seed orchards can be classified as either clonal or seedling seed orchards, and may contain either tested or untested genotypes. Most first-generation orchards of Douglas-fir were established as clonal orchards by grafting scions from field selections onto seedling rootstock that were planted at wide spacings (e.g., 6 × 6 m). In clonal orchards, multiple ramets (copies) of each clone are separated from one another to maximize outcrossing. Orchard layouts vary widely in the number and arrangement of clones and ramets. In Douglas-fir, these layouts include random designs, randomized complete blocks, systematic designs, and arrangements based on the geographic origin of the parents (Cress and Daniels 1990). Detailed design considerations for windpollinated orchards have been discussed by Giertych (1975) and Hodge and White (1993). John Duffield pioneered the use of Douglas-fir clonal orchards in the 1950s, but these early orchards ran into problems with graft incompatibility, a type of graft failure caused by the tree’s defense response (Silen and Copes 1972; Copes 1989). To avoid graft incompatibility, the Progressive Program made crosses among their better field selections and established full-sib seedling seed orchards (Silen and Wanek 1986), but this approach was abandoned when graft-compatible rootstock became available. Donald Copes began studying graft incompatibility in 1964. He found that graft rejection is highly heritable (Copes 1974), and by the mid-1970s, seeds from graft-compatible families were widely available. Today, seedling seed orchards are rare, and virtually all scions are grafted onto graft-compatible rootstock (Copes 1999). Other cultural

6. BREEDING DOUGLAS-FIR

321

treatments, such as bark scoring, are also valuable for alleviating graft incompatibility problems in established orchards (Copes 1989). Tested orchards contain parents that have been chosen based on progeny test results, whereas untested orchards (which may be established when a breeding program is initiated) contain trees that have been chosen based on their phenotype alone. Tested orchards may be established using tested parents, or by roguing (i.e., removing) undesirable clones in untested orchards based on progeny tests results. In either case, orchards that contain only the best first-generation selections are typically referred to as 1.5 generation orchards. In the first generation, the IFA, BCMoF, and Weyerhaeuser programs grafted scions from field selections directly into clonal seed orchards. As information from progeny tests became available, these orchards were rogued based on progeny performance. This approach was desirable because it minimized the time until the orchards began producing seed, but also led to uneven spacing and gaps in the orchards because of the large number of clones that had to be removed. Most second-generation orchards are established as tested orchards. Compared to rogued orchards, these orchards will have more ramets per clone when they are established, and a better distribution of trees when they are in full production. Now that graft incompatibility is no longer a problem, virtually all Douglas-fir orchards are established by grafting scions from select trees onto young seedlings, either directly in the field, or onto trees temporarily growing in pots. Because it takes eight or more years for the grafts to develop large crowns that can carry large cone crops, some orchard managers have begun to accelerate cone production by grafting multiple scions onto a single rootstock, or by substantially increasing the density of the trees (see Miniaturized Seed Orchards). The number of clones in an orchard is a balance between the need to keep the number of clones small (i.e., to increase selection intensity and genetic gain) and the need to have enough clones to avoid inbreeding and maintain genetic diversity. If the clones are unrelated, inbreeding is rarely a concern because selfing (i.e., individual or clonal selfing) is low at the mature seed stage, and selfs are unlikely to survive culling during seedling production. Therefore, the main effect of selfing is to lower seed production. The inclusion of related clones in an orchard is more of a concern because inbred seedlings derived from matings between close relatives may not be culled in the nursery and could adversely affect the performance of plantations. Therefore, related individuals are usually excluded from the same orchard and the number of clones in the orchard is usually determined by balancing genetic gain

322

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

with genetic diversity. From a diversity standpoint, the effective population size (Ne) of orchard seed is more relevant than the census number of clones (N). The province of British Columbia, for example, stipulates that all orchard crops must have an Ne of 10 or more (British Columbia Ministry of Forests 2004). Clonal differences in male or female fertility and non-random mating tend to reduce Ne, whereas pollen contamination tends to increase Ne (i.e., compared to N). Furthermore, nonrandom mating, fertility differences, and pollen contamination are all common in Douglas-fir orchards (discussed below). Based on data from a wide variety of conifer seed orchards, 0.5 is a conservative estimate for the ratio of Ne to N (Johnson 1998a; Kang et al. 2001). Based on this estimate, Johnson (1998a) used genetic theory to compare genetic gains in relation to genetic diversity, and concluded that a final orchard population of 20 to 25 clones would be reasonable for Douglas-fir (see Fig. 1 in Johnson 1998a). Johnson and Lipow (2002) came to the same conclusion after reviewing studies that either examined genetic diversity in seed orchards (based on allozymes) or risks associated with clonal deployment. In contrast, many of the operational Douglas-fir orchards have had 50 to 150 or more clones. This dramatically limits genetic gains and is well beyond what is needed to maintain adequate genetic diversity. In contrast, if the number of clones were at the lower end of the spectrum, it would be wise for seed orchard managers to closely monitor the Ne of their orchard seed using molecular genetic markers. Unfortunately, the amount of genetic diversity that should be retained is mostly based on theoretical considerations because empirical data on the relationship between genetic diversity and long-term performance are unavailable. Roguing is an important part of orchard management. Many firstgeneration orchards were established with a large number of untested clones, or clones for which only sibling (i.e., not progeny) information is available. Roguing then takes place after progeny test results become available. When orchards are established using tested clones, some managers skew the clonal makeup of the orchard towards the best parents (e.g., 75% of the ramets may come from the top 50% of the clones). In either case, the final number of clones in the orchard must be low enough to achieve adequate genetic gain, but high enough to maintain an adequate Ne. Orchard managers seek to maintain orchard vigor, promote early seed production, and produce large quantities of genetically improved seed at regular intervals over the life of the orchard. This is done via grafting, flower stimulation, roguing, pollen management, weed control, irriga-

6. BREEDING DOUGLAS-FIR

323

tion, fertilization, frost protection, and pest control. Because these activities are costly, a rapid return on investment is important. One of the key management activities in Douglas-fir orchards is flower stimulation. Flower stimulation generally refers to the enhanced production of male and female strobili on trees that are already competent to flower. Flowering can be increased using stem injections of gibberellic acid (i.e., a mixture of GA4 and GA7), wounding the trees with stem girdles or root pruning, and/or fertilizing with calcium nitrate (Ebell 1972; Ross et al. 1985; Wheeler et al. 1985; Woods 1989). These treatments are applied in the spring, and mature seeds are available two growing seasons later. Therefore, flower-stimulating treatments are usually applied every other year. Flowering can be enhanced as early as two years after grafting using GA4/7 and wounding (Cherry et al., unpubl. data), whereas calcium nitrate is generally used on older trees. Using these techniques, either alone or in combination, cone and pollen production can be increased more than 10-fold, but work is still needed to reliably promote flowering on juvenile seedlings and recalcitrant genotypes. The mechanisms by which these treatments stimulate flowering have been the source of much speculation, but little focused research. The availability of advanced approaches in molecular biology and genomics should now make this a more tractable research question. A great deal of operational flowering research has been done in North America, and cooperative research is also underway in France, Belgium, Spain, and Italy (Héois 2000; Philippe et al. 2004). Pollen management is a particularly important aspect of seed orchard management that affects both genetic gain and genetic diversity. Douglas-fir pollen management was reviewed in detail by Webber (1995) and Webber and Painter (1996). Pollen represents half the potential genetic gain in an orchard, but is difficult to control. In conventional orchards, the optimal situation is to have all seeds fathered by other orchard trees, and all parents contributing equally to the seed crop. Unfortunately, neither goal is achieved in wind-pollinated orchards. Pollen contamination (which is measured as the proportion of orchard seeds that are fathered by non-orchard trees) is typically high because most Douglas-fir orchards are located near other Douglas-fir trees— either in native stands, or in other orchards containing trees from other breeding zones. Pollen contamination in conventional orchards often exceeds 40% (Smith and Adams 1983; Wheeler and Jech 1986a; Adams et al. 1997; Slavov 2004), which could reduce genetic gains by 20% or more (depending on the genetics of the surrounding stands), and potentially increase maladaptation, if the contamination comes from trees

324

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

from other breeding zones (Webber and Painter 1996). As genetic gains increase in advanced generations, the adverse impact of pollen contamination will increase as well. Several techniques are used to reduce pollen contamination. One way is to increase the amount of orchard pollen reaching the trees. Flower stimulation can be used to create large quantities of pollen in the orchard, and orchard blowers (e.g., helicopters or large fans pulled behind tractors) can be used to release large clouds of pollen, hopefully before peak pollen shed occurs in the surrounding stands (Sorensen and Webber 1997). In addition, some orchard managers graft new, secondgeneration selections into rogued, first-generation orchards that contain mature trees that are already producing large amounts of pollen. The first-generation trees are then removed after the second-generation trees mature and begin producing enough pollen of their own. Finally, methods have been developed for the collection, storage, and delivery of pollen, so that orchard pollen can be applied to unprotected female strobili via SMP (Copes 1985; Copes et al. 1991, 1995). SMP success is measured as the percentage of seed that are fathered by the pollen that is applied via SMP. Although the success of SMP can be greater than 50% (Wheeler and Jech 1986b, 1988, 1992), its success depends on many factors (Webber and Painter 1996). In operational settings, Webber and Painter (1996) suggest that a long-term success rate of only 20 to 30% is likely to be achieved. In addition to reducing pollen contamination, SMP can be used to increase seed yields when pollen is limiting, balance paternal contributions, and create elite full-sib crosses (Webber 1995). Controlled mass pollination can also be used to exclude foreign pollen, but this is more expensive. A second way to reduce pollen contamination is to reduce the amount of background pollen reaching the trees. Overhead irrigation can be used to cool the orchard and delay flowering relative to non-orchard trees, potentially reducing the amount of outside pollen that is present during peak orchard receptivity (Silen and Kean 1969). Cooling also shortens the duration of female receptivity and pollen shed, thereby increasing the number of clones with the potential to mate with one another (Fashler and El-Kassaby 1987). Webber and Painter (1996) conclude that bloom delay is the most consistently effective way to reduce contamination and improve parental balance. Condensing the time of female receptivity and pollen flight also alleviates some of the problems associated with unequal mating within the orchard. In wind-pollinated orchards, male and female parents contribute very unequally to the seed crop (El-Kassaby and Askew 1991; Stoehr et al. 1998; Slavov 2004). Furthermore, mating is not random.

6. BREEDING DOUGLAS-FIR

325

Trees with early pollen-shed preferentially mate with trees that are receptive early, and trees with late pollen-shed mate more often with trees that are receptive late (Slavov 2004). In addition to the reduction in Ne caused by non-random mating, these early- and late-receptive clones are more prone to pollen contamination (El-Kassaby and Ritland 1986; Slavov 2004). Therefore, it may be possible to reduce pollen contamination by harvesting seed selectively, but this will reduce pollen contamination only slightly because only a small proportion of the clones will fall into unusually early or late phenology classes (Slavov 2004). Although each of these approaches can reduce pollen contamination, their effectiveness has not been fully evaluated. Now that highly variable SSR genetic markers have been developed for Douglas-fir (Slavov et al. 2004), it is possible to precisely measure pollen contamination and test the effectiveness of these treatments. SSR markers are also valuable for measuring clonal selfing, determining the genetic contributions of each clone to the orchard crop, characterizing deviations from random mating, and uncovering seed contamination and mislabeled ramets (Slavov 2004). During the early life of the orchard, it is important to promote rapid vegetative growth and large crowns, but later, it may be more important to limit vegetative growth to promote flowering. Most orchard managers monitor the nutritional status of the soil and tree foliage to guide fertilizer treatments. It is also necessary to control seed and cone insects with pesticides. Therefore, pesticide testing and registration is an important ongoing effort. Because late spring frosts cause serious damage during peak receptivity of the strobili, frost protection is frequently needed. Some frost protection is possible by mixing the air with fans and helicopters, or by warming the air with smudge pots, gas-fired blowers, or under-crown watering, but these methods are often partially effective at best. The main advantage of conventional orchards is that large amounts of seed can be produced on relatively few trees with minimal crown management. Despite their success, conventional, wind-pollinated orchards have a few key limitations (Sweet 1995). First, because they are windpollinated, pollen contamination is a problem and intra-orchard mating is uneven. Second, because the trees are large, it is difficult and costly to harvest cones, apply pesticides, prevent frost damage, use overhead irrigation for bloom delay, and make elite crosses using CMP or SMP. Finally, because the trees are widely spaced, they require a great deal of land and per-hectare seed yields are low during the early life of the orchard. Because of these limitations, various miniaturized seed orchard

326

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

designs are being used or considered in Douglas-fir and other species (Sweet 1995; Webber and Painter 1996; Héois 2000). 2. Miniaturized Seed Orchards. Miniaturized seed orchards (MSOs) are promising alternatives to conventional seed orchards. We use MSOs to denote outdoor orchards in which the trees are planted at close spacings, and then maintained at a height of only 2 to 4m. Using this approach, seeds are produced close to the ground on many small trees, rather than on a few, much larger trees. Other terms have been used to describe particular types of MSOs, including HAPSOs (hedged artificially-pollinated seed orchards), monoclonal orchards, and micro-orchards (Sweet 1995). The goal of most MSOs is to facilitate controlled pollination via CMP or SMP. This is accomplished by keeping the trees small and planting them in clonal rows so that controlled pollination and other management techniques can be easily applied from the ground to individual clones (Sweet and Krugman 1977). Control-pollinated orchards allow breeders to eliminate pollen contamination and produce both elite families and specialty breeds (e.g., cold hardy families for deploying on particularly harsh microsites). Furthermore, a wind-pollinated orchard with 25 clones would require a long-term breeding program with 25 sublines to be able to establish 25 unrelated clones in the orchard. A controlpollinated orchard, in contrast, would only require two sublines because controlled crosses can be restricted to parents in different sublines. This would dramatically reduce the complexity of long-term breeding. Although the driving force behind many MSOs is the ability to use controlled-pollination, they have other benefits as well. In some cases, seed orchard managers are moving toward miniaturized seed orchards because land is scarce for conventional orchards, but it is not yet clear whether per-hectare seed yields will be higher for MSOs or not. Second, once they are established, the costs of MSO management should be lower because the crowns are closer to the ground, thereby facilitating seed collection, pest management, frost protection, and bloom delay. Third, because there are more trees per hectare, a given quantity of seed can be produced earlier in the life of the orchard (i.e., assuming that orchard acreage is held constant). In conventional orchards, genetic gains are delayed (and financial returns are reduced) because of the long time lag between seed orchard establishment and the production of commercial quantities of genetically improved seed. Many first-generation orchards took 10 to 15 years to produce useful amounts of seed (Cress and Daniels 1990), whereas seed production in MSOs is possible as early as three years after grafting because of their high densities (Cherry et al., unpubl.).

6. BREEDING DOUGLAS-FIR

327

MSOs are becoming increasingly popular for the production of horticultural tree crops and forest tree seed (Jackson 1989). In New Zealand, for example, radiata pine orchards were changed from mostly openpollinated orchards to control-pollinated MSOs in 1984 (Sweet 1995). At least five Douglas-fir MSOs have been established in recent years in the Pacific Northwest. Although seed is now being produced from the oldest of these orchards, much remains to be learned about the management, costs, and potential yields of Douglas-fir MSOs. Despite their potential advantages, MSOs remain unproven in Douglas-fir. MSOs will cost more to establish than conventional orchards because more trees are needed to produce equal amounts of seed. This should be offset, however, by earlier production of large seed crops. Some management costs will increase, such as the costs of extra labor needed to keep the trees small. Finally, there are many biological questions that need to be answered before they can be considered fully operational. C. Vegetative Propagation 1. Advantages of Vegetative Propagation. Compared to sexual propagation via seed, vegetative propagation has four main advantages. First, the problem of pollen contamination, which is inherent to wind-pollinated seed orchards, is eliminated. Second, by avoiding sexual reproduction, non-additive genetic variation can be captured to obtain greater genetic gains. Third, because vegetative propagation is possible from young seedlings and even immature embryos, improved materials can be tested almost immediately via clonal tests, rather than having to wait until the trees flower to conduct progeny tests. This means that deployment of improved, tested genotypes can occur earlier using vegetative propagation. Fourth, the genetic uniformity of clones should facilitate reforestation, stand management, harvesting, and processing, and permit smaller tests to be used for research and breeding. Vegetative propagation can also be used in conjunction with seed-based strategies to multiply or “bulk up” small numbers of seedlings produced by crossing elite parents. Although current methods of vegetative propagation are too expensive to be used operationally, vegetative propagation is an active area of research and costs should decrease. Douglas-fir can be vegetatively propagated via rooted stem cuttings, somatic embryogenesis, organogenesis, and grafting, but only rooted cuttings and somatic embryogenesis have the potential to produce large numbers of trees for outplanting, and only rooted cuttings have been deployed operationally.

328

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

2. Methods of Vegetative Propagation. Rooted cuttings are widely used to propagate forest tree species, including angiosperms and conifers. Compared to other methods of vegetative propagation, rooted cuttings have three main advantages—the technology is low-tech, does not require much labor, and once the cuttings have formed roots, they can be handled like seedlings throughout the rest of the reforestation process. The main disadvantages are that only modest numbers of cuttings can be produced from a single tree (i.e., without serial propagation), and once they are established in the field, they may have problems with plagiotropic (i.e., non-upright) growth. Propagation of Douglas-fir from rooted cuttings has been studied for more than 50 years (Griffith 1940), and has been actively pursued since the 1970s (Copes 1977; Ritchie 1993). In general, stem cuttings from young trees are treated with plant growth regulators (e.g., IBA, NAA), then cultured under mist, often with bottom heat and aggressive treatment with fungicides (Copes and Mandel 2000). Rooting success is often high, but varies by genotype and declines rapidly with the age of the ortet (Copes and Mandel 2000). Furthermore, the trees are expensive and often show plagiotropic growth. During the 1990s, Weyerhaeuser produced and deployed millions of Douglas-fir rooted cuttings (see Genotype multiplication below). They produced their rooted cuttings from rapidly growing, 1-year-old seedlings by rooting them in the greenhouse, and then transferring them to containers or bareroot nurseries for another year of growth before being outplanted. Somatic embryogenesis is the process whereby embryos are formed from vegetative (somatic) tissues without sexual reproduction. In Douglas-fir, embryos from immature seed can be induced to form embryogenic tissue cultures that can be repeatedly subcultured and multiplied to produce millions of clonal replicates. Although the starting embryos are usually derived from full-sib crosses between elite parents, the full genetic potential of the embryo is unknown when the cultures are initiated. Therefore, it is important that embryogenic cultures can be cryogenically stored, and then used to produce trees for outplanting once the embryogenic lines have been tested in the field. Because of the vast numbers of trees that can be produced, somatic embryogenesis is the most promising approach to large-scale clonal forestry (Farnum et al. 1983). At least two companies (Weyerhaeuser and CellFor) have large research programs aimed at developing commercially viable somatic embryogenesis for Douglas-fir (Sutton 2002; Gupta et al. 2003). Although large field tests with hundreds of Douglas-fir clones have already been established, large-scale operational plantations are still many years away.

6. BREEDING DOUGLAS-FIR

329

Unfortunately, somatic embryogenesis is labor intensive, difficult to mechanize, and biologically challenging. Labor may be as much as 60% of the total production costs, and efficient methods for storing and converting the embryos into somatic seedlings are still being developed (Sutton 2002; Gupta et al. 2003). Current areas of research, including the development of production systems based on liquid media and deployment via manufactured seed, would allow somatic embryos to be handled just like orchard seed is handled today (Sutton 2002; Gupta et al. 2003). Organogenesis is the de novo production of plant organs. The induction of roots on stem cuttings is one example of ex vitro organogenesis. Whole trees can also be recovered using in vitro organogenesis—i.e., by inducing shoot meristems on tissues cultured in vitro, and then inducing these shoots to form roots. Organogenesis can be used to produce Douglas-fir trees from seedling cotyledons and other tissues (Goldfarb et al. 1991a), but this system will never be able to compete with rooted cuttings or somatic embryogenesis because it is too costly and slow (i.e., shoot and root induction are separate processes). Furthermore, because only juvenile tissues can be used, the clones are untested and the number of ramets that can be produced is limited by the amount of cotyledonary tissue available (i.e., unless the primary ramets are subsequently propagated by rooted cuttings or some other method). Although grafting is an extremely valuable technique for establishing clonal seed orchards, it cannot be used to establish production plantations because it is too costly. 3. Genotype Multiplication. The production of elite full-sib families can be enhanced by using vegetative propagation to “bulk-up” a small number of full-sib seedlings. Throughout much of the 1990s, Weyerhaeuser used this approach to deploy large numbers of rooted cuttings (Ritchie 1993). Elite trees were crossed to generate full-sib families, and the progeny were used as motherstock (ortets) from which the cuttings were taken. Forty or more cuttings were rooted from each seedling, and these were combined to create large half-sib or full-sib families (i.e., containing mixtures of siblings and clones). These families were subsequently mixed, or deployed in pure family blocks. Although the trees performed well, this program is being discontinued because the costs are too high. Instead, Weyerhaeuser’s long-term goal is to capture genetic gain by deploying trees (“somatic seedlings”) derived from somatic embryos. 4. Clonal Forestry. Clonal forestry is the production, testing, and deployment of clones. It is distinguished from other types of forestry that use

330

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

vegetative propagation (such as genotype multiplication) by its use of clonal testing. Clonal forestry is widely used in species that are easy to propagate from stem cuttings (e.g., Populus, Eucalyptus, Sequoia, Cryptomeria), but is rarely practiced in most conifers. One advantage of true clonal forestry is the ability to capture non-additive genetic variation. In Douglas-fir, the ratio of non-additive to additive genetic variation seems to be about 0.5 or less for most commercially important traits (see Quantitative genetics). Therefore, it should be possible to achieve some additional gains by testing and deploying clones, rather than full-sib families or wind-pollinated seedlots. A second advantage is the ability to begin field tests immediately after the clones are available, rather than having to wait until seed production begins, which is the case for normal progeny tests designed to measure GCA or SCA. The major hurdle to clonal forestry is its cost. Producing reforestation stock via rooted cuttings or somatic embryos can be orders of magnitude more expensive than seedlings. Another concern is whether a sufficient number of clones can be developed given the economic and biological constraints of the production technologies. Although virtually all trees can be efficiently propagated via seed, many genotypes do not perform well in rooted cutting or somatic embryogenic systems. The challenge for large-scale clonal forestry systems in Douglas-fir will be to produce a sufficient number of genotypes for each of the breeding zones that exist in the region. 5. Genetic Testing. Vegetative propagation can also be used to efficiently test genotypes in the field. By estimating breeding values or genotypic values from clonal replicates, rather than from siblings, field testing is more precise and effective. This approach has been used to map quantitative trait loci in Douglas-fir (Wheeler et al. 2005) and to improve genetic gains in tree breeding programs using clonal replication. In particular, the effectiveness of within-family selection is greatly improved by clonally replicating and testing multiple genotypes within each fullsib family (Isik et al. 2004). D. Deployment There is a close link between the production of improved materials and their deployment. On the one hand, options for deployment are constrained by the methods used to produce the improved materials, but deployment goals can also drive changes in production technology. Because improved seed can be efficiently produced in wind-pollinated

6. BREEDING DOUGLAS-FIR

331

seed orchards, Douglas-fir is mostly deployed as heterogeneous collections of seedlings from a single orchard. Nonetheless, there has been a trend toward planting seedlots of greater uniformity and genetic gain. Because many of the Pacific Northwest orchards are now producing surplus seed, some companies are collecting and deploying seed from only the very best parents. The seedlings being deployed may come from a bulk collection of seed from the best parents in the orchard, or may consist of half-sibs from a single orchard clone. Other organizations are producing high-gain, control-pollinated seed to be deployed as seedlings (i.e., without genotype multiplication). In mature, highly productive orchards, this can be a surprisingly cost-effective approach. In addition, Weyerhaeuser used rooted cuttings to multiply seedlings from elite crosses, and was able to deploy full-sib family blocks on a modest scale (Ritchie 1993). The long-term goal of many forestry organizations is to practice true clonal forestry, at least on their best sites (Farnum et al. 1983). This deployment goal has the potential to drive major changes in seedling production strategies—from seed orchards to somatic embryogenesis, and perhaps manufactured seed. This deployment goal may also force changes in breeding strategy—from strategies that focus exclusively on additive genetic variation, to ones that focus more heavily on both additive and non-additive variation. Because of the high costs of clonal forestry, a shift in this direction will provide pressures to reduce the genetic diversity of the trees being outplanted and keep the number of breeding zones small (Sutton 2002), although this is not a necessary outcome. Clones could be deployed as small blocks of single clones or as clonal mixtures. Nonetheless, it will still be many years before a large number of clonal test plantations are established across the region. Therefore, there is little to suggest that that clonal deployment will compete with seed-based deployment of Douglas-fir within the next few decades.

XI. BIOTECHNOLOGY In the context of Douglas-fir breeding, biotechnology is the application of science and engineering technology to the genetic improvement and deployment of Douglas-fir trees. Four areas of biotechnology have current or potential application to forest tree breeding and deployment: vegetative propagation, transgenics, molecular genetic markers, and genomics (Yanchuk 2001b).

332

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

A. Vegetative Propagation Uses of rooted cutting technology, somatic embryogenesis, and manufactured seed for vegetative propagation have already been discussed. By facilitating clonal forestry, these biotechnologies have the potential to dramatically change tree breeding and deployment, but are not yet economically viable, at least for Douglas-fir. Furthermore, somatic embryogenesis is the most promising mode for producing transgenic trees (van Frankenhuyzen and Beardmore 2004), although operational deployment of transgenic Douglas-fir is highly unlikely in the foreseeable future (discussed below).

B. Transgenics Plant transformation is the process whereby genes are introduced, become stably integrated, and expressed in a plant. We refer to plants that have been transformed as “transgenic,” regardless of the origin of the genes. At least 33 species of forest trees have been genetically transformed (van Frankenhuyzen and Beardmore 2004). Although transgenic Douglas-firs have not been documented in the literature, at least one transgenic tree has been produced (D. D. Ellis, pers. comm.). The betaglucuronidase (GUS) gene was inserted into Douglas-fir via biolistic transformation of somatic embryos. The GUS gene was stably incorporated and expressed in one tree, which was grown to age three. Transformation was verified via the presence of GUS enzyme activity and PCR-based detection of GUS gene fragments in Douglas-fir DNA (D. D. Ellis, pers. comm.). This report is consistent with the fact that other species in the pine family (Pinaceae) have been transformed (van Frankenhuyzen and Beardmore 2004). Douglas-fir trees can be recovered from somatic embryogenic tissue cultures (Gupta et al. 2003), and foreign genes are expressed in Douglas-fir tissues (Goldfarb et al. 1991b). Despite this report, there has been no long-term, concerted effort to develop efficient transformation protocols for Douglas-fir. In some species, transgenic trees have already been tested in the field, but it may be 15 to 20 years before any are released for commercial use (at least in Canada; Hadley et al. 2001), and transgenic Douglas-fir will lag behind, if they are deployed at all. Why is there so little interest in transgenic Douglas-fir? First, we know of no single-gene or oligogenic traits that are valuable enough to warrant the large investments needed to develop transgenic Douglas-fir. Growth and adaptability are highly polygenic and no single phenotype would do well over a very large area; there is only a modest economic incen-

6. BREEDING DOUGLAS-FIR

333

tive to improve wood quality; no insect or disease pests are important enough to warrant the development of large resistance breeding programs, let alone transgene technology; and traits such as herbicide resistance are not as valuable for long-rotation trees as they are for annual agronomic crops. Furthermore, there is ample genetic variation to make progress via traditional breeding, so only novel traits are likely to attract real interest. The most compelling reason to have transgene technologies on hand would be to counteract the potentially serious problems caused by exotic insect or disease pests. Both gypsy moth (Lymantria dispar) and the sudden oak death pathogen (Phytophthora ramorum) are potential problems for Douglas-fir in the future. Furthermore, given that most programs seek to maintain substantial genetic diversity (even if clonal deployment becomes a reality), the development and long-term testing of large numbers of transgenic genotypes is an expensive proposition. Finally, public perceptions and regulatory hurdles may be insurmountable. Despite the technical and economic hurdles, environmental concerns have become the main obstacle to public acceptance and regulatory approval of transgenic trees (van Frankenhuyzen and Beardmore 2004). A decision to plant genetically transformed trees among native forests will receive close scrutiny, mainly because of concerns about transgenes introgressing into native populations. To prevent this, robust mechanisms for imparting sterility may be required for these trees to be deployed (Strauss et al. 1995). We must also pay careful attention to the effects of transgenic trees on other organisms and ecosystem processes (van Frankenhuyzen and Beardmore 2004), particularly for traits such as insect resistance. Use of the insect toxin gene from Bacillus thuringiensis to confer insect resistance is one of the more realistic applications of genetic transformation in forest trees (van Frankenhuyzen and Beardmore 2004). C. Molecular Genetic Markers Molecular genetic markers (mostly allozymes) have been widely used in Douglas-fir tree improvement and gene conservation programs since the early 1980s. DNA-based markers are beginning to supplant allozymes for most traditional uses, and have also made new uses possible, including the construction of genetic maps, analyses of quantitative trait loci (QTL), and association genetics. Allozymes were used throughout the 1980s and early 1990s to study Douglas-fir pollen contamination, seed orchard management, genotype mislabeling, population structure, genetic diversity, and mating systems (El-Kassaby et al. 1981; Neale and Adams 1985; Merkle and Adams

334

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

1987; Li and Adams 1989; Wheeler and Jech 1992; El-Kassaby and Ritland 1996; Adams et al. 1997). Since the mid-1980s, the use of DNAbased markers has steadily increased, including RFLPs (Neale et al. 1986; Jermstad et al. 1994), RAPDs (Krutovskii et al. 1998), microsatellites (Amarasinghe and Carlson 2002; Slavov et al. 2004), and single nucleotide polymorphisms (SNPs; Krutovskii et al. 2004). RFLPs were first used to confirm the paternal inheritance of chloroplast DNA in Douglas-fir (Neale et al. 1986) and to study the evolutionary relationships between Douglas-fir and its relatives (Strauss et al. 1990). RFLPs and RAPDs were subsequently used to construct linkage maps of Douglasfir and comparative maps between Douglas-fir and pine (Jermstad et al. 1998; Krutovskii et al. 1998; Krutovsky et al. 2004). These linkage maps were the foundation of a long-term project to map QTL for important growth and adaptive traits. Scientists at the U.S. Forest Service Institute of Forest Genetics and Weyerhaeuser collaborated to create two large, clonally replicated full-sib families for pedigree-based QTL mapping and verification. Phenotypic measurements in nursery, greenhouse, and field tests were used to map QTL for height growth, second flushing, fall cold hardiness, spring bud flush, and spring cold hardiness (Jermstad et al. 2001a, 2001b, 2003; Wheeler et al. 2005). Six to 10 QTL regions have been repeatedly detected for these traits. Although the percentages of phenotypic variance explained were usually less than 10%, some QTL explained as much as 50% or more of the genetic variation in some experiments (Jermstad et al. 2001a). These results, and results from other species, support an oligogenic or polygenic model of inheritance for adaptive traits—presumably involving tens to hundreds of loci (Howe et al. 2003). Similar conclusions have been drawn for other traits in forest trees (Sewell and Neale 2000). Pedigree-based QTL studies reveal much about the genetic architecture of growth and adaptation, and provide a foundation for markeraided-selection (MAS), which seems to be economically feasible in Douglas-fir (Johnson et al. 2000; Wu et al. 2000). Nonetheless, MAS will be challenging because of the polygenic control of important quantitative traits in forest trees (Howe et al. 2003) and substantial linkage disequilibrium (Strauss et al. 1992). Therefore, markers for MAS will probably be needed within the genes of interest, rather than in distant genomic regions, but these markers are not uncovered using pedigreebased QTL studies because of their low resolution. Therefore, another approach called association genetics may be needed to identify the actual genes responsible for variation in quantitative traits (discussed below).

6. BREEDING DOUGLAS-FIR

335

A newly developed set of SSR markers has been used to measure pollen contamination, selfing, relative paternal contributions, and positive assortative mating with respect to flowering phenology in a Douglas-fir seed orchard (Slavov 2004; Slavov et al. 2004). These SSRs have also been used to fingerprint Douglas-fir trees to uncover mislabeled parental selections in the field and mislabeled ramets in a seed orchard (Slavov 2004). This same approach would be valuable for confirming genotypes that are deployed in clonal forestry programs. Given their high variability, these markers should also be perfect for combining polycross breeding with paternity analysis (Lambeth et al. 2001). D. Genomics Genomics is the integrated study of the structure, function, and interactions of all the genes in an organism. It typically involves the use of high-throughput techniques to identify genes, determine their DNA sequences, map their locations on chromosomes, and understand their functions. Because the genome is the basis of all biology, genomics research will facilitate the integration of tree physiology, genecology, gene conservation, and applied tree breeding. Advanced genomics programs exist in a small number of forest trees, including loblolly pine, radiata pine, Eucalyptus, and Populus. We and others recently organized the Douglas-fir Genome Project to facilitate the development of genomics resources in Douglas-fir (http://dendrome.ucdavis.edu/dfgp/). Ongoing activities include the development of expressed sequence tag (EST) libraries, DNA microarrays, SNP markers, and association genetics. The main objective of this work is to identify the genes responsible for genetic variation in adaptive traits and wood quality. Association genetics is an emerging population-based approach for identifying genes that are responsible for genetic variation in natural populations (Neale and Savolainen 2004). Genetic association studies involve searching for statistical associations between phenotypes and marker alleles in populations of unrelated individuals (Howe et al. 2003; Neale and Savolainen 2004). Unlike pedigree analyses (used in typical QTL experiments), association studies should be useful for finding markers in the genes of interest, which are more likely to be useful for MAS in many different families. Therefore, the markers developed from association studies might be useful for MAS in typical breeding programs and for studying adaptation and genetic structure in natural populations. Nonetheless, this will only happen after we can reliably explain a reasonable proportion of the phenotypic variation among individuals and

336

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

populations—perhaps 20% or more. Otherwise, traditional breeding will probably be more efficient. Because we expect that individual markers will only reliably explain a few percent of the phenotypic variation in natural and breeding populations, our long-term goal is to develop robust, multilocus approaches for explaining substantial proportions of genetic variation (Howe et al. 2003). To conduct association studies in forest trees, we will need to (1) identify candidate genes, (2) find useful polymorphisms in candidate genes, (3) develop study populations, (4) phenotype and genotype individuals, and (5) verify associations using independent populations (Howe et al. 2003). Because complete genome scans using SNP markers are not yet possible in Douglas-fir, it is necessary to focus our efforts on candidate genes—genes believed to have important functional roles in the trait of interest based on indirect or circumstantial evidence. Candidate genes can be identified based on their similarity to genes of known function in other species, co-location with QTL on genetic maps, and patterns of gene expression. Once candidate genes are identified, molecular markers are developed in, or very near, these genes to use in association studies. The markers of choice for association studies are SNPs. Nearly 50 candidate genes have already been placed on a Douglas-fir genetic map, and nucleotide diversity and linkage disequilibrium have been measured (Krutovskii et al. 2004). Using these candidate genes, association studies will begin within the next year (D. B. Neale, pers. comm.). Although genomic analyses may involve determining the complete DNA sequence of an organism, the cost of doing this is prohibitive, and this is likely to be done for only a handful of model organisms in the near future. The main alternative is to sequence only those genes that are expressed in an individual at any given time. These DNA sequences are called expressed sequence tags, or ESTs. Because the expressed parts of the genome are most easily tied to organism function, an EST sequencing project is a critically important building block of any genomics program. We are developing large EST libraries in Douglas-fir to facilitate candidate gene discovery. Candidates will be identified using all three methods described above: EST sequence similarity to genes in other species, co-location with known QTL on genetic maps, and patterns of gene expression measured using DNA microarrays. Because they are transferable across species, EST markers also facilitate comparative genome analysis (Krutovsky et al. 2004). Microarrays can be used to monitor collective changes in gene expression in response to multiple factors, such as drought, temperature extremes, nutrition, mycorrhizae, insects, disease, and intrinsic developmental programs such as flowering.

6. BREEDING DOUGLAS-FIR

337

XII. GENE CONSERVATION The impacts of tree breeding and other silvicultural practices on genetic diversity must be carefully considered for ecological, economic, social, and ethical reasons. Except for small, isolated populations, population extirpation is not a serious threat in Douglas-fir. Therefore, tree breeders are mostly concerned with maintaining sufficient genetic diversity for populations to evolve and adapt to changing environmental conditions such as those imposed by exotic pests, air pollution, and climate change. In general, different types and amounts of genetic diversity will be maintained in the gene resource, inter-situ, breeding, and production populations (Fig. 6.3) to (1) permit continued gain in future generations, (2) maintain adaptability, and (3) conserve rare alleles that may be important in the future (Lipow et al. 2003). In situ and ex situ methods are used to conserve genetic diversity in Douglas-fir. Gene resource populations conserve genetic resources in their native habitats (in situ), whereas other genetic resources are conserved ex situ—in seed banks, clone banks, progeny tests, provenance tests, and seed orchards. Not only does coastal Douglas-fir have extensive in situ genetic resources, but also the largest and most complete collection of ex situ resources of any forest tree (Lipow et al. 2003). More than 4 million progeny from nearly 34,000 parents have been planted on nearly 1,000 sites in western Oregon, western Washington, British Columbia, and northern California (Lipow et al. 2003). Furthermore, seed stores contain seedlots from more than 20,000 parents, and additional materials are planted in Europe, New Zealand, and Chile (Lipow et al. 2004). Despite these huge numbers, the ability of tree breeding populations to meet gene conservation goals has been carefully considered (Yanchuk and Lester 1996; Yanchuk 2001a; Lipow et al. 2003, 2004). Tree breeders seek to conserve quantitative genetic variation because it is associated with adaptability and the ability to obtain genetic gains in economic traits. Breeders may also seek to conserve rare alleles based on the hypothesis that they will be important in the future. Although most low-frequency alleles are probably rare because they are deleterious (at least under current conditions), they might confer important phenotypes (such as disease resistance) if conditions change. Alternatively, they might confer novel traits that are not subject to strong positive selection in natural populations. A rare, mutated allele for a gene that normally encodes cinnamyl alcohol dehydrogenase (CAD), for example, may confer desirable wood characteristics in loblolly pine (Gill et al. 2003).

338

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

For long-term conservation of quantitative genetic variation, a population of 5,000 trees (i.e., Ne ≥ 1,000) should be more than adequate (Yanchuk 2001a). Based on this target, Lipow et al. (2004) performed a “gap analysis” of Douglas-fir gene resources in western Oregon and western Washington. In situ resources were considered adequate if at least 5,000 mature trees were estimated to occur in protected areas within each genetically distinct area (i.e., ecoregion, breeding zone, or seed zone). Based on these analyses, Lipow et al. (2004) concluded that Douglas-fir genetic resources are well protected in situ, except for a possible gap in the southern Puget lowlands, which has few protected areas. However, genetic resources from this area are well conserved in ex situ reserves. Not only is Douglas-fir well conserved in situ, but the ex situ populations that are associated with breeding programs (e.g., progeny test “inter-situ” populations) are large enough for most gene conservation purposes (Lipow et al. 2003). Douglas-fir genetic resources are also well conserved in British Columbia (Yanchuk and Lester 1996). What population sizes are needed to conserve rare alleles? Yanchuk (2001a) studied this question based on the assumption that (1) rare alleles must be captured in multiple individuals (e.g., 5 or 20) to be useful in a breeding program, and (2) only alleles that impart phenotypic effects can be captured in a form that is useful (i.e., can be located and purposefully used in breeding). Therefore, it is much easier to “capture” (i.e., conserve and use) dominant alleles that can be identified in heterozygotes than it is to capture recessive alleles that can only be identified in trees that are homozygous for the desired allele. Based on these analyses, Yanchuk (2001a) concluded that a population of 554 trees would have a good (95%) chance of capturing dominant alleles in 20 trees if the allele frequency is 0.025. For recessive alleles, a population of 554 trees will only capture alleles in 20 trees if the allele frequency is almost 10 times higher (i.e., 0.224). In contrast, a population of 55,755 trees would be needed to capture recessive alleles at a frequency of 0.022 (i.e., about the same frequency as for dominant alleles in a population of 554). In short, a population of a few thousand trees (i.e., equivalent to a localized inter-situ or gene resource population) is very effective for capturing rare dominant alleles (i.e., frequencies < 0.005), but hundreds of thousands of trees are needed to capture recessive alleles that are only moderately rare (e.g., < 0.01). Therefore, more geographically extensive inter-situ or gene resource populations are needed to capture these alleles. Nonetheless, given the huge number of trees conserved in all protected areas and ex situ populations, the conservation of these alleles is virtually assured, even if they may not occur in the most desirable genetic backgrounds (Yanchuk 2001a).

6. BREEDING DOUGLAS-FIR

339

XIII. ACKNOWLEDGMENTS We thank the following for reviewing the manuscript and providing unpublished information: Christine Dean, Barry Jaquish, Leith Knowles, Marc Rust, J. Bradley St.Clair, Michael Stoehr, and Jack Woods. We also thank Frank Sorensen, Charles Tauer, and one anonymous reviewer for providing helpful comments on an earlier draft of the manuscript, and Dana Howe for editorial help.

LITERATURE CITED Aagaard, J. E., S. S. Vollmer, F. C. Sorensen, and S. H. Strauss. 1995. Mitochondrial DNA products among RAPD profiles are frequent and strongly differentiated between races of Douglas-fir. Mol. Ecol. 4:441–447. Adams, W. T. 1992. Gene dispersal within forest tree populations. New Forests 6:217–240. Adams, W. T., and R. K. Campbell. 1981. Genetic adaptation and seed source specificity. p. 78–85. In: S. D. Hobbs, and O. T. Helgerson (eds.), Reforestation of skeletal soils, Proc. Reforestation Workshop, Medford, OR, 17–19 Nov. 1981, Forest Research Laboratory, Oregon State Univ., Corvallis. Adams, W. T., V. D. Hipkins, J. Burczyk, and W. K. Randall. 1997. Pollen contamination trends in a maturing Douglas-fir seed orchard. Can. J. For. Res. 27:131–134. Adams, W. T., S. N. Aitken, D. G. Joyce, G. T. Howe, and J. Vargas-Hernandez. 2001. Evaluating efficacy of early testing for stem growth in coastal Douglas-fir. Silvae Genet. 50:167–175. Aitken, S. N., and W. T. Adams. 1995. Impacts of genetic selection for height growth on annual developmental cycle traits in coastal Douglas-fir. p. AD1: 1–8. In: Evolution of breeding strategies for conifers from the Pacific Northwest. Joint meeting of the IUFRO Working Parties S2.02.05 (Douglas-fir), S2.02.06 (Pinus contorta), S2.02.12 (Sitka spruce) and S2.02.14 (Abies), Limoges, France, 1–4 Aug. 1995, INRA, Station d’Amélioration des Arbres Forestiers, Ardon, France. Aitken, S. N., and M. Hannerz. 2001. Genecology and gene resource management strategies for conifer cold hardiness. p. 23–53. In: F. J. Bigras, and S. J. Columbo (eds.), Conifer cold hardiness. Kluwer, Dordrecht, The Netherlands. Aitken, S. N., W. T. Adams, N. Schermann, and L. H. Fuchigami. 1996. Family variation for fall cold hardiness in two Washington populations of coastal Douglas-fir (Pseudotsuga menziesii var. menziesii (Mirb.) Franco). For. Ecol. Managem. 80:187–195. Amarasinghe, V., and J. E. Carlson. 2002. The development of microsatellite markers for genetic analysis in Douglas-fir. Can. J. For. Res. 32:1904–1915. Anekonda, T. S., W. T. Adams, and S. N. Aitken. 1998. Influence of second flushing on genetic assessment of cold hardiness in coastal Douglas-fir (Pseudotsuga menziesii var. menziesii (Mirb.) Franco). For. Ecol. Managem. 111:119–126. Anekonda, T. S., W. T. Adams, and S. N. Aitken. 2000. Cold hardiness testing for Douglasfir tree improvement programs: Guidelines for a simple, robust and inexpensive screening method. West. J. Appl. For. 15:129–136. Anonymous. 1998. Best seed origins for Douglas-fir. Making Headway with Forest Research, Coillte Research and Development Infor. Note No. 17, Newtownmountkennedy, County Wicklow, Ireland.

340

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Anonymous. 2003. Lumber production and mill stocks: 2002. U.S. Census Bureau, Current Industrial Reports MA321T(02)-1 (http://www.census.gov/industry/1/ ma321t02.pdf). Anonymous. 2004. Forest Genetics Council of British Columbia 2003/2004 Annu. Rep. Arnott, J. T. 1986. Douglas-fir stand establishment overview: Coastal British Columbia. p. 219–229. In: C. D. Oliver, D. P. Hanley, and J. A. Johnson (eds.), Douglas-fir: Stand management for the future. College of Forest Resources, Univ. Washington, Seattle. Balduman, L. M., S. N. Aitken, M. Harmon, and W. T. Adams. 1999. Genetic variation in cold hardiness of Douglas-fir in relation to parent tree environment. Can. J. For. Res. 29:62–72. Bastien, J.-C., B. Roman-Amat, and G. Vonnet. 1985. Natural variability of some wood quality traits of coastal Douglas-fir in a French progeny test: Implications on breeding strategy. In: Proc. IUFRO Working Party S2.02.05, Vienna, Austria, June 1985. Bever, D. N. 1954. Evaluation of factors affecting natural reproduction of forest trees in central western Oregon. Oregon State Board of Forestry, Res. Bul. #3. Birot, Y., and C. Christophe. 1983. Genetic structures and expected genetic gains from multitrait selection in wild populations of Douglas-fir and Sitka spruce. Silvae Genet. 32:141–151. Breidenstein, J., J.-C. Bastien, and B. Roman-Amat. 1990. Douglas-fir range-wide variation results from the IUFRO data base. p. 2.13–2.26. In: Proc. joint meeting Western Forest Genetics Assoc. and IUFRO Working Parties S2.02-05, 06, 12, and 14 Douglas-fir, contorta pine, Sitka spruce, and Abies breeding and genetic resources, Olympia, WA, 20–24 Aug. 1990, Weyerhaeuser Co., Federal Way, WA. British Columbia Ministry of Forests. 2004. Chief Forester’s standards for seed use. British Columbia Ministry of Forests, November 20, 2004, Victoria. Burdon, R. D. 1988. Recruitment for breeding populations: Objectives, genetics, and implementation. p. 555–572. In: B. S. Weir, E. J. Eisen, M. M. Goodman, and G. Namkoong (eds.), Proc. Second Intl. Conf. Quantitative Genetics, Raleigh, NC, 31 May–5 June 1987, Sinauer, Inc. Burdon, R. D., and C. J. A. Shelbourne. 1971. Breeding populations for recurrent selection: Conflicts and possible solutions. New Zealand J. For. Sci. 1:174–193. Cafferata, S. L. 1986. Douglas-fir stand establishment overview: Western Oregon and Washington. p. 211–218. In: C. D. Oliver, D. P. Hanley, and J. A. Johnson (eds.), Douglasfir: Stand management for the future. College of Forest Resources, Univ. Washington, Seattle. Campbell, R. K. 1964. Recommended traits to be improved in a breeding program for Douglas-fir. Weyerhaeuser Res. Note. No. 57, Forestry Research Center, Timberland Division, Weyerhaeuser Company, Centralia, WA. Campbell, R. K. 1965. Phenotypic variation and repeatability of stem sinuosity in Douglasfir. Northwest Sci. 39:47–59. Campbell, R. K. 1972. Genetic variability in juvenile height-growth of Douglas-fir. Silvae Genet. 21:126–129. Campbell, R. K. 1974a. A provenance-transfer model for boreal regions. Rep. Norwegian For. Res. Inst. 31:545–566. Campbell, R. K. 1974b. Use of phenology for examining provenance transfers in reforestation of Douglas-fir. J. Appl. Ecol. 11:1069–1080. Campbell, R. K. 1979. Genecology of Douglas-fir in a watershed in the Oregon Cascades. Ecology 60:1036–1050. Campbell, R. K. 1986. Mapped genetic variation of Douglas-fir to guide seed transfer in southwest Oregon. Silvae Genet. 35:85–96.

6. BREEDING DOUGLAS-FIR

341

Campbell, R. K. 1991. Soils, seed-zone maps, and physiography: Guidelines for seed transfer of Douglas-fir in southwestern Oregon. For. Sci. 37:973–986. Campbell, R. K. 1992. Genotype × environment interaction: A case study for Douglas-fir in western Oregon. USDA For. Serv. Res. Pap. PNW-RP-455, Portland, OR. Campbell, R. K., and F. C. Sorenson. 1973. Cold acclimation in seedling Douglas-fir related to phenology and provenance. Ecology 54:1148–1151. Campbell, R. K., and F. C. Sorensen. 1978. Effect of test environment on expression of clines and on delineation of seed zones in Douglas-fir. Theor. Appl. Genet. 51: 223–246. Campbell, R. K., and A. I. Sugano. 1979. Genecology of bud-burst phenology in Douglasfir: Response to flushing temperature and chilling. Bot. Gaz. 140:223–231. Cherry, M. L., and G. T. Howe (eds.). 2004. Proc. Genetics and Growth Modeling Workshop. Vancouver, WA, 4–6 Nov. 2003, Pacific Northwest Research Cooperative Report 21, Oregon State Univ., Corvallis. Ching, K. K. 1959. Hybridization between Douglas-fir and bigcone Douglas-fir. For. Sci. 5:246–254. Ching, K. K. 1965. Early growth of Douglas-fir in a reciprocal planting. Forest Research Laboratory Res. Pap. 3, Oregon State Univ., Corvallis. Ching, K. K., and P. N. Hinz. 1978. Provenance study of Douglas-fir in the Pacific Northwest region. Silvae Genet. 27:229–233. Chollet, F. 1986. Les vergers à graines forestiers. p. 74–80. In: Amélioration génétique des arbres forestiers. Revue Forestière Française, Numéro Spécial 38. Christophe, C., and Y. Birot. 1979. Genetic variation within and between populations of Douglas-fir. Silvae Genet. 28:197–206. Copes, D. L. 1974. Genetics of graft rejection in Douglas-fir. Can. J. For. Res. 4:186–192. Copes, D. L. 1977. Influence of rooting media on root structure and rooting percentage of Douglas-fir cuttings. Silvae Genet. 26:102–106. Copes, D. L. 1985. Fertility of Douglas-fir pollen after one year of storage in liquid nitrogen. For. Sci. 31:569–574. Copes, D. L. 1989. Bark scoring problem grafts in five Douglas-fir seed orchards: A case history. USDA For. Serv. Res. Note PNW-RN-487, Portland, OR. Copes, D. L. 1999. Breeding graft-compatible Douglas-fir rootstocks (Pseudotsuga menziesii (Mirb.) Franco). Silvae Genet. 48:188–193. Copes, D. L., and N. L. Mandel. 2000. Effects of IBA and NAA treatments on rooting Douglas-fir stem cuttings. New Forests 20:249–257. Copes, D. L., N. C. Vance, W. K. Randall, A. Jasumback, and R. Hallman. 1991. Vacuum collection of Douglas-fir pollen for supplemental mass pollinations. USDA For. Serv. Res. Note PNW-RN-503, Portland, OR. Copes, D. L., D. O. O’Rourke, and W. K. Randall. 1995. Field testing a modified duster for supplemental mass pollination in Douglas-fir seed orchards. Tree Planters’ Notes 46:118–125. Cornelius, J. 1994. Heritabilities and additive genetic coefficients of variation in forest trees. Can. J. For. Res. 24:372–379. Cotterill, P. P. 1989. The nucleus breeding system. p. 36–42. In: Proc. 20th Southern Forest Tree Improvement Conf., Charleston, SC, 26–30 June 1989, Southern Forest Tree Improvement Committee, Springfield, VA. Cotterill, P. P., and N. Jackson. 1989. Gains expected from clonal orchards under alternative breeding strategies. For. Sci. 35:183–196. Cotterill, P. P., and J. W. James. 1984. Number of offspring and plot sizes required for progeny testing. Silvae Genet. 33:203–209.

342

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Cotterill, P., C. Dean, J. Cameron, and M. Brindbergs. 1988. Nucleus breeding: A new strategy for rapid improvement under clonal forestry. p. 39–51. In: G. I. Gibson, A. R. Griffin, and A. C. Matheson, Breeding tropical trees: Population structure and genetic improvement strategies in clonal and seedling forestry. Proc. IUFRO Meeting on Breeding Tropical Trees, Pattaya, Thailand, 28 Nov.–2 Dec. 1988, Oxford For. Inst., Oxford, UK. Cress, D. W., and J. D. Daniels (eds.). 1990. Forest seed orchards of western Oregon, western Washington and northern California. Daniels and Associates Consultants, Centralia, WA. Doerksen, A. H., and K. K. Ching. 1972. Karyotypes in the genus Pseudotsuga. For. Sci. 18:66–69. Douglas-fir Cooperative 2004. http://www.forestresearch.co.nz/topic.asp?topic=Douglas% 20Fir%20Cooperative&title=Douglas%20Fir%20Cooperative Duffield, J. W. 1956. Damage to western Washington forests from November 1955 cold wave. USDA For. Serv. Pac. Northwest For. Range Exp. Sta. Res. Note 129, Portland, OR. Ebell, L. F. 1972. Cone-induction response of Douglas-fir to form of nitrogen fertilizer and time of treatment. Can. J. For. Res. 2:317–326. El-Kassaby, Y. A., and G. R. Askew. 1991. The relation between reproductive phenology and reproductive output in determining the gametic pool profile in a Douglas-fir seed orchard. For. Sci. 37:827–835. El-Kassaby, Y. A., and K. Ritland. 1986. The relation of outcrossing and contamination to reproductive phenology and supplemental mass pollination in a Douglas-fir seed orchard. Silvae Genet. 35:240–244. El-Kassaby, Y. A., and K. Ritland. 1996. Impact of selection and breeding on the genetic diversity in Douglas-fir. Biodivers. Conserv. 5:795–813. El-Kassaby, Y. A., F. C. Yeh, and O. Sziklai. 1981. Estimation of the outcrossing rate of Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco] using allozyme polymorphisms. Silvae Genet. 30:182–184. EUFORGEN. 2003. Summary of the meeting, Fourth EUFORGEN Conifers Network Meeting, Pitlochry, Scotland, 18–20 Oct. 2003. Falconer, D. S., and T. F. C. Mackay. 1996. Introduction to quantitative genetics, 4th ed. Longman Group Ltd., Essex, England. Farnum, P., R. Timmis, and J. L. Kulp. 1983. Biotechnology of forest yield. Science 219:694–702. Farrar, J. L. 1995. Trees in Canada. Fitzhenry and Whiteside Ltd., Nat. Res. Canada, Canadian Forest Service, Markham/Ottawa, Ont. Fashler, A. M. K., and Y. A. El-Kassaby. 1987. The effect of water spray cooling treatment on reproductive phenology in a Douglas-fir seed orchard. Silvae Genet. 36:245–249. Ferre, Y. de. 1955. Les Pseudotsuga del l’Arboretum de Joueou et leur resistance au Phaeocryptus. Trav. Lab. for. Toulouse 6:art 1 [Forestry-Abstracts 1956 017-02982]. Ferrell, W. K., and E. S. Woodard. 1966. Effects of seed origin on drought resistance of Douglas-fir (Pseudotsuga menziesii) (Mirb.) Franco. Ecology 47:499–503. Fins, L. 1983. Douglas-fir tree improvement plan. Inland Empire Tree Improvement Cooperative, Univ. Idaho, Moscow. Fletcher, A. M. 2002. Introduction of western North American species to Europe and the IUFRO seed collections. In: G. Philippe, and G. Bergevin (eds.), Pacific temperate conifers as native and introduced species: Genetics and sexual reproduction, Proc. IUFRO Working Parties S2.02.5 and S2.01.16, Orleans, France, 27–30 May 2002, Cemagref.

6. BREEDING DOUGLAS-FIR

343

Forest Products Laboratory. 1999. Wood handbook—Wood as an engineering material. USDA For. Serv. Forest Products Laboratory Gen. Tech. Rep. FPL-GTR-113, Madison, WI. Fu, Y. B., A. D. Yanchuk, and G. Namkoong. 1999. Spatial patterns of tree height variations in a Douglas-fir progeny trial: Implications for genetic testing. Can. J. For. Res. 29:714–723. Gamble, W. E., W. T. Adams, and J. Tappeiner II. 1996. Survival and growth of Douglasfir seed sources in the Hospital Tract rangewide source archive plantation. Forest Research Laboratory Res. Contrib. 12, Oregon State Univ., Corvallis. Gernandt, D. S., and A. Liston. 1999. Internal transcribed spacer region evolution in Larix and Pseudotsuga (Pinaceae). Am. J. Bot. 86:711–723. Giertych, M. 1975. Seed orchard designs. p. 25–37. In: R. Faulkner (ed.), Seed orchards. Forestry Comm. Bul. 54, London. Gill, G. P., G. R. Brown, and D. B. Neale. 2003. A sequence mutation in the cinnamyl alcohol dehydrogenase gene associated with altered lignification in loblolly pine. Plant Biotech. J. 1:253–258. Goldfarb, B., G. T. Howe, L. M. Bailey, S. H. Strauss, and J. B. Zaerr. 1991a. A liquid cytokinin pulse induces adventitious shoot formation from Douglas-fir cotyledons. Plant Cell Rep. 10:156–160. Goldfarb, B., S. H. Strauss, G. T. Howe, and J. B. Zaerr. 1991b. Transient gene expression of microprojectile-introduced DNA in Douglas-fir cotyledons. Plant Cell Rep. 10:517–521. Griffin, A. R. 1974. Geographic variation in juvenile growth characteristics of Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) from the coastal ranges of California. Ph.D. Diss., Oregon State Univ., Corvallis. Griffin, A. R., and K. K. Ching. 1977. Geographic variation in Douglas-fir from the coastal ranges of California. I. Seed, seedling growth and hardiness characteristics. Silvae Genet. 26:149–157. Griffith, B. G. 1940. Effect of indolebutyric acid, indoleacetic acid, and alpha naphthaleneacetic acid on rooting of cuttings of Douglas-fir and Sitka spruce. J. For. 38:496–501. Griffin, J. R. 1964. A new Douglas-fir locality in southern California. For. Sci. 10:317–319. Gupta, P. K., R. Timmis, and W. Carlson. 2003. Somatic embryo development in liquid medium for large-scale propagation of conifer trees. Abstract P-28. In: Proc. Congress on In Vitro Biology, Portland, OR, 31 May–4 June 2003. In Vitro Cell. Dev. Biol. - Plant 39:P-28. Hadley, M. J., J. S. Tanz, and J. Fraser. 2001. Biotechnology: Potential applications in tree improvement. Forest Genetics Council of British Columbia Extension Note 2. Hagenstein, W. D. 1986. Douglas-fir forestry from Allen to Mendel. p. 32–38. In: C. D. Oliver, D. P. Hanley, and J. A. Johnson (eds.), Douglas-fir: Stand management for the future. College of Forest Resources, Univ. Washington, Seattle. Heaman, J. C. 1977. Choosing strategies for a breeding program in Douglas fir (Pseudotsuga menziesii (Mirb.) Franco) for coastal British Columbia—A case study. p. 1205–1214. In: A. G. Brown, and C. M. Palmberg (eds.), Proc. FAO/IUFRO Third World Consultation Forest Tree Breeding, Vol. 2, Canberra, Australia, 21–26 March 1977, CSIRO, Canberra, Australia. Heiner, T. D., and D. P. Lavender. 1972. Early growth and drought avoidance in Douglasfir seedlings. Forest Research Laboratory Res. Pap.14, Oregon State Univ., Corvallis. Héois, B. (coordinator). 2000. EUDIREC: Towards a sustainable productive Douglas-fir forest of high quality raw material in Europe. Final report: European Union Project FAIRCT95-0909.

344

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Héois, B., M. Guibert, G. Philippe, N. Schermann, and J.-C. Bastien. 1995. Les vergers à graines de Douglas français: Une composante majeure du programme d’amélioration vers une organization européenne cooperative. In: Evolution of breeding strategies for conifers from the Pacific Northwest, Proc. IUFRO Working Parties S2.02.05, 06, 12 and 14, Limoges, France, 1–4 Aug. 1995. Hermann, R. K. 1982. The genus Pseudotsuga: Historical records and nomenclature. Forest Research Laboratory, Special publication 2a, Oregon State Univ., Corvallis. Hermann, R. K., and D. P. Lavender. 1990. Pseudotsuga menziesii (Mirb.) Franco. Douglasfir. p.527–540. In: R. M. Burns, and B. H. Honkala (eds.), Silvics of North America, Vol. 1, Conifers. USDA For. Serv. Agr. Handb. 654. Hermann, R. K., and D. P. Lavender. 1999. Douglas-fir planted forests. New Forests 17:53–70. Hernandez, G. T., G. V. Alonso, G. P. Arribas, and J. L. Jenkinson. 1993. Screening Douglasfir for rapid growth in common-garden tests in Spain. USDA For. Serv. Gen. Tech. Rep. PSW-GTR-146, Albany, CA. Hodge, G. R., and T. L. White. 1993. Advanced generation wind-pollinated seed orchard design. New Forests 7:213–236. Hoff, R. J. 1987. Susceptibility of inland Douglas-fir to Rhabdocline needle cast. USDA For. Serv. Res. Note INT-375, Ogden, UT. Howard, J. L. 2001. U.S. timber production, trade consumption, and price statistics 1965 to 1999. USDA For. Serv. Res. Pap. FPL-RP-595, Forest Products Laboratory, Madison, WI. Howe, G. T., and K. J. Jayawickrama. 2002. Genetics of stem quality in coastal Douglasfir. p. 67–71. In: K. J. Jayawickrama, and G. Bracher (eds.), Proc. Genetic Improvement of Wood Quality in Coastal Douglas-fir and Western Hemlock. Corvallis, OR, 27 June 2002. Oregon State Univ., Corvallis (http://www.fsl.orst.edu/pnwtirc/publicatons/ pnwtirc_pubs_date.htm). Howe, G. T., J. Davis, Z. Jekni´c, T. H. H. Chen, B. Frewen, H. D. Bradshaw, Jr., and P. Saruul. 1999. Physiological and genetic approaches to studying endodormancy-related traits in Populus. HortScience 34:1174–1184. Howe, G. T., S. N. Aitken, D. B. Neale, K. D. Jermstad, N. C. Wheeler, and T. H. H. Chen. 2003. From genotype to phenotype: Unraveling the complexities of cold adaptation in forest trees. Can. J. Bot. 81:1247–1266. Howe, G. T., B. Shindler, B. Cashore, E. Hansen, D. Lach, and W. Armstrong. 2005. Public influences on plantation forestry. J. For. (in press). Illingworth, K. 1978. Douglas-fir provenance trials in coastal British Columbia: Results to six years after planting. p. 411–425. In: Proc. IUFRO joint meeting of Working Parties. Vancouver, B.C., 21 Aug.–9 Sept. 1978. Irgens-Moller, H. 1962. Genotypic variation in photoperiodic response of Douglas-fir seedlings. For. Sci. 8:360–362. Irgens-Moller, H. 1967a. Geographical variation in growth patterns of Douglas-fir. Silvae Genet. 17:106–110. Irgens-Moller, H. 1967b. Patterns of height growth initiation and cessation in Douglas-fir. Silvae Genet. 16:56–58. Isaac, L. A. 1930. Seed flight in the Douglas-fir region. J. For. 28:492–499. Isik, F., B. Li, J. Frampton, and B. Goldfarb. 2004. Efficiency of seedlings and rooted cuttings for testing and selection in Pinus taeda. For. Sci. 50:44–53. Jackson, J. E. 1989. Worldwide development of high density planting in research and practice. Acta Hort. 243:17–28.

6. BREEDING DOUGLAS-FIR

345

Jaquish, B. C. 1990. Geographic variation in ten-year height growth of interior Douglas-fir in British Columbia. p. 2.144–2.155. In: Proc. joint meeting Western Forest Genetics Assoc. and IUFRO Working Parties S2.02-05, 06, 12, and 14 Douglas-fir, contorta pine, Sitka spruce, and Abies breeding and genetic resources, Olympia, WA, 20–24 Aug. 1990, Weyerhaeuser Co., Federal Way, WA. Jermstad, K. D., A. M. Reem, J. R. Henifin, N. C. Wheeler, and D. B. Neale. 1994. Inheritance of restriction fragment length polymorphisms and random amplified polymorphic DNAs in coastal Douglas-fir. Theor. Appl. Genet. 89:758–766. Jermstad, K. D., D. L. Bassoni, N. C. Wheeler, and D. B. Neale. 1998. A sex-averaged genetic linkage map in coastal Douglas-fir (Pseudotsuga menziesii [Mirb] Franco var. menziesii) based on RFLP and RAPD markers. Theor. Appl. Genet. 97:762–770. Jermstad, K. D., D. L. Bassoni, K. S. Jech, N. C., Wheeler, and D. B. Neale. 2001a. Mapping of quantitative trait loci controlling adaptive traits in coastal Douglas-fir: I. Timing of vegetative bud flush. Theor. Appl. Genet. 102:1142–1151. Jermstad, K. D., D. L. Bassoni, N. C. Wheeler, T. S. Anekonda, S. N. Aitken, W. T. Adams, and D. B. Neale. 2001b. Mapping of quantitative trait loci controlling adaptive traits in coastal Douglas-fir: II. Spring and fall cold-hardiness. Theor. Appl. Genet. 102: 1152–1158. Jermstad, K. D., D. L. Bassoni, K. S. Jech, G. A. Ritchie, N. C. Wheeler, and D. B. Neale. 2003. Mapping of quantitative trait loci controlling adaptive traits in coastal Douglasfir. III. Quantitative trait loci by environment interactions. Genetics 165:1489–1506. Johnsen, Ø. 1988. After-effects on progenies from orchard clones moved to non-native environments. p. 1–11. In: J. Worrall, J. Loo-Dinkins, and D. P. Lester (eds.), Physiology and genetics of reforestation, Proc. 10th North American Forest Biology Workshop, Vancouver, B.C., 20–22 July 1988, Univ. British Columbia Press, Vancouver. Johnson, G. R. 1997. Site-to-site genetic correlations and their implications on breeding zone size and optimum number of progeny test sites for coastal Douglas-fir. Silvae Genet. 46:280–285. Johnson, G. R. 1998a. Breeding design considerations for coastal Douglas-fir. USDA For. Serv. Gen. Tech. Rep. PNW-GTR-411, Portland, OR. Johnson, G. R. 1998b. Parental GCA testing: How many crosses per parent? Can. J. For. Res. 28:540–545. Johnson, G. R. 2002. Genetic variation in tolerance of Douglas-fir to Swiss needle cast as assessed by symptom expression. Silvae Genet. 51:80–86. Johnson, G. R., and S. Lipow. 2002. Compatibility of breeding for increased wood production and long-term sustainability: The genetic variation of seed orchard seed and associated risks. p. 169–179. In: Congruent management of multiple resources, Proc. Wood Compatibility Initiative Workshop, number 18. USDA For. Serv. PNW-GTR-563, Portland, OR. Johnson, G. R., R. A. Sniezko, and N. L. Mandel. 1997. Age trends in Douglas-fir genetic parameters and implications for optimum selection age. Silvae Genet. 46:349–358. Johnson, G. R., F. Temel, and K. Jayawickrama. 2002. Genetic studies involving Swiss needle cast. p.38–43. In: G. Filip (ed.), Swiss Needle Cast Cooperative Annual Report, 2002. Forest Research Laboratory, College of Forestry, Oregon State Univ., Corvallis. Johnson, G. R., N. C. Wheeler, and S. H. Strauss. 2000. Economic feasibility of marker-aided selection in Douglas-fir. Can. J. For. Res. 30:1942–1952. Johnson, G. R., J. B. St.Clair, and S. Lipow. 2001. Genetic conservation in applied tree breeding programs. p. 215–230. In: B. A. Thielges, S. Sastrapradja, and A. Rimbawanto (eds.), Proc. Intl. Conf. on Ex Situ and In Situ Conservation of Commercial Tropical

346

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Trees, Yogyakarta, Indonesia, 11–13 June 2001, International Tropical Timber Organization (ITTO), Yokohama, Japan. Joly, R. J., W. T. Adams, and S. G. Stafford. 1989. Phenological and morphological responses of mesic and dry site sources of coast Douglas-fir to water deficit. For. Sci. 35:987–1005. Kang, K. S., A. M. Harju, D. Lindgren, T. Nikkanen, C. Almqvist, and G. U. Suh. 2001. Variation in effective number of clones in seed orchards. New Forests 21:17–33. King, J. N., F. C. Yeh, and J. C. H. Heaman. 1988a. Selection of growth and yield traits in controlled crosses of coastal Douglas fir. Silvae Genet. 37:158–164. King, J. N., F. C. Yeh, J. C. H. Heaman, and B. P. Dancik. 1988b. Selection of wood density and diameter in controlled crosses of coastal Douglas-fir. Silvae Genet. 37:152–157. King, J. N., F. C. Yeh, J. C. Heaman, and B. P. Dancik. 1992. Selection of crown form traits in controlled crosses of coastal Douglas-fir. Silvae Genet. 41:362–370. Krutovskii, K. V., S. S. Vollmer, F. C. Sorensen, W. T. Adams, S. J. Knapp, and S. H. Strauss. 1998. RAPD genome maps of Douglas-fir. J. Hered. 89:197–205. Krutovskii, K. V., K. D. Jermstad, N. C. Wheeler, G. T. Howe, J. B. St.Clair, and D. B. Neale. 2004. Gene discovery for adaptive traits and candidate gene variation in Douglas-fir. Abstract P622. In: Plant and Animal Genomes XII Conference, San Diego, CA, 10–14 January 2004. http://www.intl-pag.org/12/abstracts/P5i_PAG12_622.html. Krutovsky, K. V., M. Troggio, G. R. Brown, K. D. Jermstad, and D. B. Neale. 2004. Comparative mapping in the Pinaceae. Genetics 168:447–461. Kung, F. H., and J. W. Wright. 1972. Parallel and divergent evolution in Rocky Mountain trees. Silvae Genet. 21:77–85. Lambeth, C. C., R. W. Stonecypher, and B. J. Zobel. 1982. Early testing of Douglas-fir in phytotron environments—The effect of selection trait and genotype-environment interaction. p. 137–148. In: B. A. Thielges (ed.), Proc. 7th North Am. For. Biol. Workshop, Univ. Kentucky, Lexington. Lambeth, C., B.-C. Lee, D. O’Malley, and N. Wheeler. 2001. Polymix breeding with parental analysis of progeny: An alternative to full-sib breeding and testing. Theor. Appl. Genet. 103:930–943. Larsen, J. B. 1981. Geographic variation in winter drought resistance of Douglas-fir (Pseudotsuga menziesii Mirb. Franco). Silvae Genet. 30:109–114. Lavender, D. P., and W. S. Overton. 1972. Thermoperiods and soil temperatures as they affect growth and dormancy of Douglas-fir seedlings of different geographic origins. Forest Research Laboratory Research Paper 13, Oregon State Univ., Corvallis. Lavender, D. P., K. K. Ching, and R. K. Hermann. 1968. Effect of environment on the development of dormancy and growth of Douglas-fir seedlings. Bot. Gaz. 129:70–83. Lester, D. T. 1986. Genetic considerations for Douglas-fir. p. 151–154. In: C. D. Oliver, D. P. Hanley, and J. A. Johnson (eds.), Douglas-fir: Stand management for the future. College of Forest Resources, Univ. Washington, Seattle. Li, P., and W. T Adams. 1989. Range-wide patterns of allozyme variation in Douglas-fir (Pseudotsuga menziesii). Can. J. For. Res. 19:149–161. Li, P., and W. T. Adams. 1993. Genetic control of bud phenology in pole-size trees and seedlings of coastal Douglas-fir. Can. J. For. Res. 23:1043–1051. Li, P., and W. T. Adams. 1994. Genetic variation in cambial phenology of coastal Douglasfir. Can. J. For. Res. 24:1864–1870. Lipow, S. R., G. R. Johnson, J. B. St.Clair, and K. J. Jayawickrama. 2003. The role of tree improvement programs for ex situ gene conservation of coastal Douglas-fir in the Pacific Northwest. For. Gen. 10:111–120.

6. BREEDING DOUGLAS-FIR

347

Lipow, S. R., K. Vance-Borland, J. B. St.Clair, J. Henderson, and C. McCain. 2004. Gap analysis of conserved genetic resources for forest trees. Conserv. Biol. 18:412–423. Magnussen, S., and A. D. Yanchuk. 1993. Selection age and risk: Finding the compromise. Silvae Genet. 42:25–40. Magnussen, S., and A. D. Yanchuk. 1994. Time trends of predicted breeding values in selected crosses of coastal Douglas-fir in British Columbia: A methodological study. For. Sci. 40:663–685. Marshall, K. A., and D. B. Neale. 1992. Inheritance of mitochondrial DNA in Douglas-fir. Can. J. For. Res. 22:73–75. McDonald, G. I. 1979. Resistance of Douglas-fir to western spruce budworm: Results of a preliminary progeny test. USDA For. Serv. Res. Note INT-264, Ogden, UT. MeGraw, R. 2002. Wood quality overview. p. 5–12. In: K. J. Jayawickrama, and G. Bracher (eds.), Proc. Genetic Improvement of Wood Quality in Coastal Douglas-fir and Western Hemlock. Corvallis, OR, 27 June 2002. Oregon State Univ., Corvallis (http://www.fsl .orst.edu/pnwtirc/publicatons/pnwtirc_pubs_date.htm). Merkle, S. A., and W. T. Adams. 1987. Patterns of allozyme variation within and among Douglas-fir breeding zones in southwest Oregon. Can. J. For. Res. 17:402–407. Monserud, R. A., and G. E. Rehfeldt. 1990. Genetic and environmental components of variation of site index in inland Douglas-fir. For. Sci. 36:1–9. Morgenstern, E. K. 1996. Geographic variation in forest trees: Genetic basis and application of knowledge in silviculture. UBC Press, Univ. British Columbia, Vancouver. Munger, T. T., and W. G. Morris. 1936. Growth of Douglas-fir trees of known seed source. U.S. Dept. Agric. Tech. Bul. 537. Namkoong, G. 1969. Nonoptimality of local races. p. 149–153. In: Proc. 10th Southern Conf. Forest Tree Improvement, Houston, Texas, 17–19 June 1969, Texas Forest Service and Texas A&M Univ., College Station. Namkoong, G. 1976. A multiple-index selection strategy. Silvae Genet. 25:199–201. Namkoong, G., and H. H. Kang. 1990. Quantitative genetics of forest trees. Plant Breed. Rev. 8:139–188. Neale, D. B., and W. T. Adams. 1985. The mating system in natural and shelterwood stands of Douglas-fir. Theor. Appl. Genet. 71:201–207. Neale, D. B., and O. Savolainen. 2004. Association genetics of complex traits in conifers. Trends Plant Sci. 9:325–330. Neale, D. B., N. C. Wheeler, and R. W. Allard. 1986. Paternal inheritance of chloroplast DNA in Douglas-fir. Can. J. For. Res. 16:1152–1154. O’Brien, I. E. W., D. R. Smith, R. C. Gardner, and B. G. Murray. 1996. Flow cytometric determination of genome size in Pinus. Plant Sci. 115:91–99. O’Neill, G. A., and S. N. Aitken. 2004. Area-based breeding zones to minimize maladaptation. Can. J. For. Res. 34:695–704. O’Neill, G. A., S. N. Aitken, and W. T. Adams. 2000. Genetic selection for cold hardiness in coastal Douglas-fir seedlings and saplings. Can. J. For. Res. 30:1799–1807. O’Neill, G. A., W. T. Adams, and S. N. Aitken. 2001. Quantitative genetics of spring and fall cold hardiness in seedlings from two Oregon populations of coastal Douglas-fir. For. Ecol. Manage. 149:305–318. Orr-Ewing, A. L. 1954. Inbreeding experiment with the Douglas-fir. For. Chron. 30:7–16. Orr-Ewing, A. L., A. R. Fraser, and I. Karlsson. 1972. Interracial crosses with Douglas-fir: Early field results. B.C. For. Serv. Res. Note 55. Parker, W. H. 2000. Rates of change of adaptive variation in Picea mariana visualized by GIS using a differential systematic coefficient. New Forests 20:259–276.

348

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Patterson, H. D., and E. R. Williams. 1976. A new class of resolvable incomplete block designs. Biometrika 63:83–92. Pharis, R. P., and W. K. Ferrell. 1966. Differences in drought resistance between coastal and inland sources of Douglas-fir. Can. J. Bot. 44:1651–1659. Philippe, G., S. J. Lee, G. Schüte, and B. Héois. 2004. Flower stimulation is cost-effective in Douglas-fir seed orchards. Forestry 77:279–286. Prudham, W. S. 2003. Taming trees: Capital, science, and nature in Pacific Slope tree improvement. Ann. Assoc. Am. Geograph. 93:636–656. Randall, W. K. 1996. Forest tree seed zones for western Oregon. State of Oregon Dept. Forestry, Portland, OR. Randall, W. K., and P. Berrang. 2002. Washington tree seed transfer zones. Washington Dept. Natural Resources, Olympia. Rehfeldt, G. E. 1977. Growth and cold hardiness of intervarietal hybrids of Douglas-fir. Theor. Appl. Genet. 50:3–15. Rehfeldt, G. E. 1978. Genetic differentiation of Douglas-fir populations from the northern Rocky Mountains. Ecology 59:1264–1270. Rehfeldt, G. E. 1979a. Ecological adaptations in Douglas-fir (Pseudotsuga menziesii var. glauca) populations. I. Idaho and northeast Washington. Heredity 43:383–397. Rehfeldt, G. E. 1979b. Variation in cold hardiness among populations of Pseudotsuga menziesii var. glauca. USDA For. Serv. Res. Paper INT-233, Ogden, UT. Rehfeldt, G. E. 1982. Ecological adaptations in Douglas-fir populations. II. Western Montana. USDA For. Serv. Res. Pap. INT-295, Ogden, UT. Rehfeldt, G. E. 1983a. Ecological adaptations in Douglas-fir (Pseudotsuga menziesii var. glauca) populations. III. Central Idaho. Can. J. For. Res. 13:626–632. Rehfeldt, G. E. 1983b. Genetic variability within Douglas-fir populations: Implications for tree improvement. Silvae Genet. 32:9–14. Rehfeldt, G. E. 1989. Ecological adaptations in Douglas-fir (Pseudotsuga menziesii var. glauca): A synthesis. For. Ecol. Managem. 28:203–215. Rehfeldt, G. E. 1990. Adaptability versus zone size: Continuous zones for the Rocky Mountains (USA). p. 2.265–2.275. In: Proc. joint meeting Western Forest Genetics Assoc. and IUFRO Working Parties S2.02-05, 06, 12, and 14 Douglas-fir, contorta pine, Sitka spruce, and Abies breeding and genetic resources, Olympia, WA, 20–24 Aug. 1990, Weyerhaeuser Co., Federal Way, WA. Riitters, K. H., and D. A. Perry. 1987. Early genetic evaluation of open-pollinated Douglasfir families. For. Sci. 33:577–582. Ritchie, G. A. 1993. Production of Douglas-fir, Pseudotsuga menziesii (Mirb.) Franco rooted cuttings for reforestation by Weyerhaeuser Company. p. 68–72. In: Comb. Proc. Intl. Plant Prop. Soc. 43:68–72. Ritchie, G. A., and J. W. Keeley. 1994. Maturation in Douglas-fir: I. Changes in stem, branch, and foliage characteristics associated with ontogenetic aging. Tree Physiol. 14:1245–1259. Roberds, J. H., and G. Namkoong. 1989. Population selection to maximize value in an environmental gradient. Theor. Appl. Genet. 77:128–134. Roberds, J. H., J. O. Hyun, G. Namkoong, and G. Rink. 1990. Height response functions for white ash provenances grown at different latitudes. Silvae Genet. 39:121–129. Rohmeder, E. 1956. Professor MUNCHS Anbauversuch mit Douglasien verschiedener Herkunft und anderen Nadelbaumarten im Forstamt Kaiserslautern-Ost 1912 bis 1954. Z. Forstgenet. 5:142–156 [Forestry Abstracts 1957, 018-01483].

6. BREEDING DOUGLAS-FIR

349

Ross, S. D., J. E. Webber, R. P. Pharis, and J. N. Owens. 1985. Interaction between gibberellin A4/7 and root-pruning on the reproductive and vegetative process in Douglasfir. I. Effects on flowering. Can. J. For. Res. 15:341–347. Sewell, M. M., and D. B. Neale. 2000. Mapping quantitative traits in forest trees. p. 407–423. In: S. M. Jain, and S. C. Minocha (eds.), Molecular biology of woody plants. Kluwer Academic Publishers, Dordrecht, The Netherlands. Silen, R. R. 1965. Regeneration aspects of the 50-year-old Douglas-fir heredity study. p. 35–39. In: Proc. 1964 Annual Meet. West. Refor. Coord. Comm., West. For. Conserve. Assoc., Spokane, WA. Silen, R. R. 1966a. A 50-year racial study of Douglas-fir in western Oregon and Washington. p. 6–7 (Abstract). In: Proc. 1965 Annual. Meet. West. For. Genetics Assoc. Silen, R. R. 1966b. A simple, progressive, tree improvement program for Douglas-fir. USDA For. Serv. Res. Note PNW-45, Portland, OR. Silen, R. R. 1978. Genetics of Douglas-fir. USDA For. Serv. Res. Pap. WO-35. Silen, R. R., and D. L. Copes. 1972. Douglas-fir seed orchard problems—A progress report. J. For. 70:145–147. Silen, R. R., and G. Kean. 1969. Cooling a Douglas-fir seed orchard to avoid pollen contamination. USDA For. Serv. Res. Note PNW-101, Portland, OR. Silen, R. R., and N. L. Mandel. 1983. Clinal genetic growth variation within two Douglasfir breeding zones. J. For. 81:216–220. Silen, R. R., and J. Wanek. 1986. The J.E. Schroeder forest tree seed orchard in Oregon. J. For. 84:31–37. Silen, R. R., and J. G. Wheat. 1979. Progressive tree improvement program in Coastal Douglas-fir. J. For. 77:78–83. Sinclair, T. R. 1998. Historical changes in harvest index and crop nitrogen accumulation. Crop Sci. 38:638–643. Slavov, G. T. 2004. Development and application of SSR markers for measuring gene flow in Douglas-fir. Ph.D. Dissertation, Oregon State Univ., Corvallis. Slavov, G. T., G. T. Howe, I. Yakovlev, K. J. Edwards, K. V. Krutovskii, G. A. Tuskan, J. E. Carlson, S. H. Strauss, and W. T. Adams. 2004. Highly variable SSR markers in Douglasfir: Mendelian inheritance and map locations. Theor. Appl. Genet. 108:873–880. Smith, D. B., and W. T. Adams. 1983. Measuring pollen contamination in clonal seed orchards with the aid of genetic markers. p. 69–77. In: Proc. 17th Southern Forest Tree Improve. Conf., Athens, Georgia, 6–9 June 1983, Univ. Georgia, Athens. Smith, W. B., J. S. Vissage, D. R. Darr, and R. M. Sheffield. 2001. Forest resources of the United States, 1997. USDA For. Serv. Gen. Tech. Report NC-219, North Central Research Station, St. Paul, MN. Sorensen, F. C. 1967. Two-year results of a west-east transect-provenance test of Douglasfir in Oregon. USDA For. Serv. Res. Note PNW-72, Portland, OR. Sorensen, F. C. 1973. Frequency of seedlings from natural self-fertilization in coastal Douglas-fir. Silvae Genet. 22:20–24. Sorensen, F. C. 1979. Provenance variation in Pseudotsuga menziesii seedlings from the var. menziesii—var. glauca transition zone in Oregon. Silvae Genet. 28:96–103. Sorensen, F. C. 1983. Geographic variation in seedling Douglas-fir (Pseudotsuga menziesii) from the western Siskiyou Mountains of Oregon. Ecology 64:696–702. Sorensen, F. C. 1997. Effects of sib mating and wind pollination on nursery seedling size, growth components, and phenology of Douglas-fir seed-orchard progenies. Can. J. For. Res. 27:557–566.

350

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Sorensen, F. C. 1999. Relationship between self-fertility, allocation of growth, and inbreeding depression in three coniferous species. Evolution 53:417–425. Sorensen, F. C., and J. E. Webber. 1997. On the relationship between pollen capture and seed set in conifers. Can. J. For. Res. 27:63–68. Sorensen, F. C., and T. L. White. 1988. Effect of natural inbreeding on variance structure in tests of wind-pollination Douglas-fir progenies. For. Sci. 34:102–118. Spicer, R., B. L. Gartner, and R. L. Darbyshire. 2000. Sinuous stem growth in a Douglasfir (Pseudotsuga menziesii) plantation: Growth patterns and wood-quality effects. Can. J. For. Res. 30:761–768. St.Clair, J. B. 1994. Genetic variation in tree structure and its relation to size in Douglasfir. I. Biomass partitioning, foliage efficiency, stem form, and wood density. Can. J. For. Res. 24:1226–1235. St.Clair, J. B., N. L. Mandel, and K. J. S. Jayawickrama. 2004. Early realized genetic gains for coastal Douglas-fir in the northern Oregon Cascades. West. J. Appl. For. 19: 195–201. Stein, W. I., and P. W. Owston. 2002. Pseudotsuga Carr. (Douglas-fir). p. 1–38. In: F. T. Bonner, and R. G. Nisley (eds.), Woody plant seed manual. USDA For. Serv., Washington, DC. Available at http://ntsl.fs.fed.us/wpsm/Pseudotsuga.pdf (accessed 7/23/04). Stephan, B. R. 1973. Über anfälligkeit und resistenz von Douglasien-Herkünften gegenüber Rhabdocline pseudotsugae (Susceptibility and resistance of Douglas-fir provenances to Rhabdocline pseudotsugae). Silvae Genet. 22:149–153. Stephan, B. R. 1987. Differences in the resistance of Douglas-fir provenances to the wooly aphid Gilletteella cooleyi. Silvae Genet. 36:76–79. Stoehr, M., and K. Bird. 2003. Coastal Douglas-fir. p. 7. In: B. Crown, and R. Painter (eds.), Forest Genetics Council of B.C. Tree Improvement Program Project Report 2002/2003. B.C. Ministry Forests, Victoria. Stoehr, M. U., B. L. Orvar, T. M. Vo, J. R. Gawley, J. E. Webber, and C. H. Newton. 1998. Application of a chloroplast DNA marker in seed orchard management evaluations of Douglas-fir. Can. J. For. Res. 28:187–195. Stoehr, M. U., J. E. Webber, C. C. Hollefreund, and R. A. Painter. 2004. Potential pollen contamination effects on progeny from an off-site Douglas-fir seed orchard: 9 year field results. Can. J. For. Res. 34:981–984. Stonecypher, R. W., R. F. Piesch, G. G. Helland, J. G. Chapman, and H. J. Reno. 1996. Results from genetic tests of selected parents of Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) in an applied tree improvement program. For. Sci. 42, Monograph 32:1–35. Strauss, S. H., A. H. Doerksen, and J. R. Byrne. 1990. Evolutionary relationships of Douglasfir and its relatives (genus Pseudotsuga) from DNA restriction fragment analysis. Can. J. Bot. 68:1502–1510. Strauss, S. H., R. Lande, and G. Namkoong. 1992. Limitations of molecular-marker-aided selection in forest tree breeding. Can. J. For. Res. 22:1050–1061. Strauss, S. H., W. H. Rottmann, A. M. Brunner, and L. A. Sheppard. 1995. Genetic engineering of reproductive sterility in forest trees. Mol. Breed. 1:5–26. Sutton, B. 2002. Commercial delivery of genetic improvement to conifer plantations using somatic embryogenesis. Ann. For. Sci. 59:657–661. Sweet, G. B. 1965. Provenance differences in Pacific Coast Douglas-fir. 1. Seed and seedling characteristics. Silvae Genet. 14:46–56. Sweet, G. B. 1995. Seed orchards in development. Tree Physiol. 15:527–530. Sweet, G. B., and S. L. Krugman. 1977. Flowering and seed production problems, and a new concept of seed orchards. p. 749–759. In: A. G. Brown, and C. M. Palmberg (eds.),

6. BREEDING DOUGLAS-FIR

351

Proc. FAO/IUFRO Third World Consultation Forest Tree Breeding, Vol. 2, Canberra, Australia, 21–26 March 1977, CSIRO, Canberra, Australia. Sziklai, O. 1990. Douglas-fir provenance-progeny test at Haney B.C., Canada. p. 2.313–2.321. In: Proc. joint meeting Western Forest Genetics Assoc. and IUFRO Working Parties S2.02-05, 06, 12, and 14 Douglas-fir, contorta pine, Sitka spruce, and Abies breeding and genetic resources, Olympia, WA, 20–24 Aug. 1990, Weyerhaeuser Co., Federal Way, WA. Temel, F., and W. T. Adams. 2000. Persistence and age-age genetic correlations of stem defects in coastal Douglas-fir (Pseudotsuga menziesii var. menziesii (Mirb.) Franco). For. Genet. 7:145–153. Temel, F., and G. R. Johnson. 2001. Early testing of Douglas-fir for Swiss needle cast tolerance. p. 36–40. In. G. Filip (ed.), Swiss Needle Cast Research Cooperative Annual Report, 2001, Forest Research Laboratory, Oregon State Univ., Corvallis. Timmis, R., J. Flewelling, and C. Talbert. 1994. Frost injury prediction model for Douglasfir seedlings in the Pacific Northwest. Tree Physiol. 14:855–869. van Frankenhuyzen, K., and T. Beardmore. 2004. Current status and environmental impact of transgenic forest trees. Can. J. For. Res. 34:1163–1180. Vargas-Hernandez, J., and W. T. Adams. 1991. Genetic variation of wood density components in young coastal Douglas-fir: Implications for tree breeding. Can. J. For. Res. 21:1801–1807. Vargas-Hernandez, J. J, W. T. Adams, and D. G. Joyce. 2003. Quantitative genetic structure of stem form and branching traits in Douglas-fir seedlings and implications for early selection. Silvae Genet. 52:36–44. von Rudloff, E. 1972. Chemosystematics studies in the genus Pseudotsuga. I. Leaf oil analysis of the coastal and Rocky Mountain varieties of the Douglas-fir. Can. J. Bot. 50:1025–1040. Webber, J. E. 1995. Pollen management for intensive seed orchard production. Tree Physiol. 15:507–514. Webber, J. E., and R. A. Painter. 1996. Douglas-fir pollen management manual, 2nd ed. Research Branch, B.C. Ministry Forests Working Paper 02/1996, Victoria. Westfall, R. D., and M. T. Conkle. 1992. Allozyme markers in breeding zone designation. New Forests 6:279–309. Wheat, J. 1966. IFA Tree Improvement Newsletter No. 3, Tree Improvement Laboratory, Rt. 2, Box 3226, Olympia, WA. Wheat, J., and R. Silen. 1977. Cooperative tree improvement in the Douglas-fir region. p. 1215–1220. In: A. G. Brown, and C. M. Palmberg (eds.), Proc. FAO/IUFRO Third World Consultation Forest Tree Breeding, Vol. 2, Canberra, Australia, 21–26 March 1977, CSIRO, Canberra, Australia. Wheeler, N. C., and K. S. Jech. 1986a. Pollen contamination in a mature Douglas-fir seed orchard. p. 160–171. In: R. J. Weir (ed.), Proc. IUFRO Conf. on breeding theory, progeny testing and seed orchards, Williamsburg, VA, 13–17 Oct. 1986, North Carolina State Univ., Raleigh. Wheeler, N. C., and K. S. Jech. 1986b. Estimating supplemental mass pollination (SMP) success electrophoretically. p. 111–120. In: G. E. Caron, A. G. Corriveau, and T. J. B. Boyle (eds.), New ways in forest genetics, Proc. 20th Canadian Tree Improvement Assoc. Meeting, Québec City, Québec, 19–22 Aug. 1985, Canadian Forestry Service, Ottawa. Wheeler, N. C., and K. S. Jech. 1988. Supplemental mass pollination (SMP) in Douglas-fir seed orchards: Biological efficacy and economic evaluation. Weyerhaeuser Tech. Rep. 050-3210/25.

352

G. HOWE, K. JAYAWICKRAMA, M. CHERRY, G. JOHNSON, AND N. WHEELER

Wheeler, N. C., and K. S. Jech. 1992. The use of electrophoretic markers in seed orchard research. New Forests 6:311–328. Wheeler, N. C., C. J. Masters, S. C. Cade, S. D. Ross, J. W. Keeley, and L. Y. Hsin. 1985. Girdling: An effective and practical treatment for enhancing seed yields in Douglas-fir seed orchards. Can. J. For. Res. 15:505–510. Wheeler, N. C., R. W. Stonecypher, and K. S. Jech. 1990. Physiological characterization of select families to aid in source movement decisions: Supplementing long-term performance trials. p. 2.362–2.377. In: Proc. joint meeting Western Forest Genetics Assoc. and IUFRO Working Parties S2.02-05, 06, 12, and 14, Douglas-fir, contorta pine, Sitka spruce, and Abies breeding and genetic resources, Olympia, WA, 20–24 Aug. 1990, Weyerhaeuser Co., Federal Way, WA. Wheeler, N. C., K. D. Jermstad, K. Krutovsky, S. N. Aitken, G. T. Howe, J. Krakowski, and D. B. Neale. 2005. Mapping of quantitative trait loci controlling adaptive traits in coastal Douglas-fir. IV. Cold-hardiness QTL verification and candidate gene mapping. Mol. Breed. (in press). White, T. L. 1987. Drought tolerance of southwestern Oregon Douglas-fir. For. Sci. 33:283–293. White, T. 1992. Advanced-generation breeding populations: Size and structure. p. 208–222. In: C. C. Lambeth, and W. Dvorak (eds.), Proc. IUFRO Conf. S2.02-08, breeding tropical trees—resolving tropical forest resource concerns through tree improvement, gene conservation, and domestication of new species, Cartagena-Cali, Colombia, 12–16 October 1992, CAMCORE, North Carolina State Univ., Raleigh. White, T. L., and K. K. Ching. 1985. Provenance study of Douglas-fir in the Pacific Northwest region. IV. Field performance at age 25 years. Silvae Genet. 34:84–90. Woods, J. H. 1989. Stem girdling to increase seed and pollen production of coast Douglasfir. B.C. Ministry Forests Res. Note. 103, Victoria. Woods, J. H. 1993. Breeding programs and strategies for Douglas-fir in North America. p. 1–11. In: Y. S. Park and G. W. Adams (eds.), Breeding strategies of important tree species in Canada. Natural Resources Canada, Fredericton, N.B. Inform. Rep. M-X186E. Woods, J. H., and M. Stoehr. 1993. Inbreeding effects on the seed production, nursery production, and early field growth of Douglas-fir. In: Proc. Western Forest Genetics Assoc. Annual Meeting, Oahu, HI, 4–8 October 1993. Woods, J., J. King, and C. Cartwright. 1995a. Early genetic gains verified in realized-gain trials of coastal Douglas-fir and western hemlock. B.C. Ministry Forests Extension Note 01, Victoria. Woods, J. H., D. Kolotelo, and A. D. Yanchuk. 1995b. Early selection of coastal Douglasfir in a farm-field test environment. Silvae Genet. 44:178–186. Woods, J. H., T. Wang, and S. N. Aitken. 2002. Effects of inbreeding on coastal Douglasfir: Nursery performance. Silvae Genet. 51:163–170. Wright, J. W. 1976. Introduction to forest genetics. Academic Press, New York. Wright, J. W., F. H. Kung, R. A. Read, W. A. Lemmien, and J. N. Bright. 1971. Genetic variation in Rocky Mountain Douglas-fir. Silvae Genet. 20:54–60. Wu, R. L., Z.-B. Zeng, S. E. McKeand, and D. M. O’Malley. 2000. The case for molecular mapping in forest tree breeding. Plant Breeding Rev. 19:41–68. Yanchuk, A. D. 1996. General and specific combining ability from disconnected partial diallels of coastal Douglas-fir. Silvae Genet. 45:37–45. Yanchuk, A. D. 2001a. A quantitative framework for breeding and conservation of forest tree genetic resources in British Columbia. Can. J. For. Res. 31:566–576.

6. BREEDING DOUGLAS-FIR

353

Yanchuk, A. D. 2001b. The role and implications of biotechnological tools in forestry. Unasylva 52:53–61. Yanchuk, A. D., and D. T. Lester. 1996. Setting priorities for conservation of the conifer genetic resources of British Columbia. For. Chron. 72:406–415. Yeh, F. C., and C. Heaman. 1982. Heritabilities and genetic and phenotypic correlations for height and diameter in coastal Douglas-fir. Can. J. For. Res. 12:181–185. Yeh, F. C., and J. C. Heaman. 1987. Estimating genetic parameters of height growth in seven-year-old coastal Douglas-fir from disconnected diallels. For. Sci. 33:946–957. Zobel, B., and J. Talbert. 1984. Applied forest tree improvement. Wiley, New York.

Subject Index Volume 27 B

Fungal diseases, maize breeding, 125–142

Bacterial disease, maize breeding, 156–159 Biography, Frederick A. Bliss, 1–14 Biotechnology: Cucurbitaceae, 213–244 Douglas-fir, 331–336 Rosaceae, 175–211 sugarcane, 15–118 Bliss, Frederick A. (biography), 1–14 Breeding: Cucurbitaceae, 213–24 Curcurbits, 213–244 Douglas-fir, 245–353 maize, foliar pathogens, 119–173 Rosaceae, 175–211 sugarcane: 15–158

G

C Cucurbits, mapping, 213–244 Cytogenetics, sugarcane, 74–78

Genetic engineering, sugarcane, 86–97 Genetics: Cucurbitaceae, 213–244 Maize foliar diseases, 118–173 Rosaceae, 175–211 Grain breeding, maize, 119–173 I Industrial crop breeding, sugarcane, 15–118 M Maize breeding, 119–173 Mapping: Cucubitaceae, 213–244 Rosaceae, 175–211 R Rosaceae, synteny, 175–211 S

D Disease and pest resistance, maize foliar pathogens, 119–173 Douglas-fir breeding, 245–353

Sugarcane breeding, 15–118 Synteny, Rosaceae, 175–211 T Transformation, sugarcane, 86–97 V

F

Virus diseases, maize breeding, 142–156

Forest crop breeding, Douglas-fir, 245–353

Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 354

Cumulative Subject Index (Volumes 1–27) A Adaptation: blueberry, rabbiteye, 5:351–352 durum wheat, 5:29–31 genetics, 3:21–167 testing, 12:271–297 Aglaonema breeding, 23:267–269 Alexander, Denton, E. (biography), 22:1–7 Alfalfa: honeycomb breeding, 18:230–232 inbreeding, 13:209–233 in vitro culture, 2:229–234 somaclonal variation, 4:123–152 unreduced gametes, 3:277 Allard, Robert W. (biography), 12:1–17 Allium cepa, see Onion Almond: breeding self-compatible, 8:313–338 domestication, 25:290–291 transformation, 16:103 Alocasia breeding, 23:269 Alstroemaria, mutation breeding, 6:75 Amaranth: breeding, 19:227–285 cytoplasm, 23:191 genetic resources, 19:227–285 Animals, long term selection 24(2):169–210, 211–234 Aneuploidy: alfalfa, 10:175–176 alfalfa tissue culture, 4:128–130 petunia, 1:19–21 wheat, 10:5–9 Anther culture: cereals, 15:141–186 maize, 11:199–224 Anthocyanin maize aleurone, 8:91–137 pigmentation, 25:89–114 Anthurium breeding, 23:269–271

Antifungal proteins, 14:39–88 Antimetabolite resistance, cell selection, 4:139–141, 159–160 Apomixis: breeding, 18:13–86 genetics, 18:13–86 reproductive barriers, 11:92–96 rice, 17:114–116 Apple: Domestication, 25:286–289 genetics, 9:333–366 rootstocks, 1:294–394 transformation, 16:101–102 Apricot: domestication, 25:291–292 transformation, 16:102 Arachis, see Peanut in vitro culture, 2:218–224 Artichoke breeding, 12:253–269 Avena sativa, see Oat Avocado domestication, 25:307 Azalea, mutation breeding, 6:75–76 B Bacillus thuringensis, 12:19–45 Bacteria, long-term selection, 24(2):225–265 Bacterial diseases: apple rootstocks, 1:362–365 cell selection, 4:163–164 cowpea, 15:238–239 maize, 27:156–159 potato, 19:113–122 raspberry, 6:281–282 soybean, 1:209–212 sweet potato, 4:333–336 transformation fruit crops, 16:110 Banana: breeding, 2:135–155 domestication, 25:298–299 transformation, 16:105–106

Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 355

356 Barley: anther culture, 15:141–186 breeding methods, 5:95–138 diversity, 21:234–235 doubled haploid breeding, 15:141–186 gametoclonal variation, 5:368–370 haploids in breeding, 3:219–252 molelcular markers, 21:181–220 photoperiodic response, 3:74, 89–92, 99 vernalization, 3:109 Bean (Phaseolus): breeding, 1:59–102; 10:199–269; 23:21–72 breeding mixtures, 4:245–272 breeding (tropics), 10:199–269 heat tolerance, 10:149 in vitro culture, 2:234–237 long-term selection, 24(2):69–74 photoperiodic response, 3:71–73, 86–92; 16:102–109 protein, 1:59–102 rhizobia interaction, 23:21–72 Beet (table) breeding, 22:357–388 Beta, see Beet Biochemical markers, 9:37–61 Biography: Alexander, Denton E., 22:1–7 Allard, Robert W., 12:1–17 Bliss, Frederick A., 27:1–14 Bringhurst, Royce S., 9:1–8 Burton, Glenn W., 3:1–19 Coyne, Dermot E., 23:1–19 Downey, Richard K., 18:1–12 Dudley, J.W., 24(1):1–10 Draper, Arlen D., 13:1–10 Duvick, Donald N., 14:1–11 Gabelman, Warren H., 6:1–9 Hallauer, Arnel R., 15:1–17 Harlan, Jack R., 8:1–17 Jones, Henry A., 1:1–10 Laughnan, John R. 19:1–14 Munger, Henry M., 4:1–8, Rédei, George, P., 26:1–33 Peloquin, Stanley J., 25:1–19 Ryder, Edward J., 16:1–14 Sears, Ernest Robert, 10:1–2 Simmonds, Norman W., 20:1–13 Sprague, George F., 2:1–11 Vogel, Orville A., 5:1–10 Vuylsteke, Dirk R., 21:1–25 Weinberger, John H., 11:1–10 Yuan, Longping, 17:1–13 Biotechnology: Cucurbitaceae, 27:213–244 Douglas-fir, 27:331–336 politics, 25:21–55 Rosaceae, 27:175–211

CUMULATIVE SUBJECT INDEX Birdsfoot trefoil, tissue culture, 2:228–229 Blackberry, 8:249–312 mutation breeding, 6:79 Black walnut, 1:236–266 Bliss, Frederick A. (biography), 27:1–14 Blueberry: breeding, 13:1–10 domestication, 25:304 rabbiteye, 5:307–357 Brachiaria, apomixis, 18:36–39, 49–51 Bramble: domestication, 25:303 transformation, 16:105 Brassica, see Cole crops Brassica: napus, see Canola, Rutabaga rapa, see Canola Brassicaceae: incompatibility, 15:23–27 molecular mapping, 14:19–23 Breeding: Aglaonema, 23:267–269 alfalfa via tissue culture, 4:123–152 almond, 8:313–338 Alocasia, 23:269 amaranth, 19:227–285 apomixis, 18:13–86 apple, 9:333–366 apple rootstocks, 1:294–394 banana, 2:135–155 barley, 3:219–252; 5:95–138; 26:125–169 bean, 1:59–102; 4:245–272; 23:21–72 beet (table), 22:357–388 biochemical markers, 9:37–61 blackberry, 8:249–312 black walnut, 1:236–266 blueberry, rabbiteye, 5:307–357 bromeliad, 23:275–276 cactus, 20:135–166 Calathea, 23:276 carbon isotope discrimination, 12:81–113 carrot, 19:157–190 cassava, 2:73–134 cell selection, 4:153–173 chestnut, 4:347–397 chimeras, 15:43–84 chrysanthemum, 14:321–361 citrus, 8:339–374 coffee, 2:157–193 coleus, 3:343–360 competitive ability, 14:89–138 cowpea, 15:215–274 cucumber, 6:323–359 Cucurbitaceae 27:213–244 cucurbits, 27:213–244 cytoplasmic DNA, 12:175–210

CUMULATIVE SUBJECT INDEX diallel analysis, 9:9–36 Dieffenbachia, 271–272 doubled haploids, 15:141–186; 25:57–88 Dougas-fir, 27:245–253 Dracaena, 23:277 drought tolerance, maize, 25:173–253 durum wheat, 5:11–40 Epepremnum, 23:272–2mn epistasis, 21:27–92 exotic maize, 14:165–187 fern, 23:276 fescue, 3:313–342 Ficus, 23:276 Flower color, 25:89–114 foliage plant, 23:245–290 forest tree, 8:139–188 fruit crops, 25:255–320 gene action 15:315–374 genotype × environment interaction, 16:135–178 grapefruit, 13:345–363 grasses, 11:251–274 guayule, 6:93–165 heat tolerance, 10:124–168 Hedera, 23:279–280 herbicide-resistant crops, 11:155–198 heritability, 22:9–111 heterosis, 12:227–251 homeotic floral mutants, 9:63–99 honeycomb, 13:87–139; 18:177–249 hybrid, 17:225–257 hybrid wheat, 2:303–319; 3:169–191 induced mutations, 2:13–72 insect and mite resistance in cucurbits, 10:199–269 isozymes, 6:11–54 legumes, 26:171–357 lettuce, 16:1–14; 20:105–133 maize, 1:103–138, 139–161; 4:81–122; 9:181–216; 11:199–224; 14:139–163, 165–187, 189–236; 25:173–253; 119–173 mitochondrial genetics, 25:115–238 molecular markers, 9:37–61, 10:184– 190; 12:195–226; 13:11–86; 14:13–37, 17:113–114, 179, 212–215; 18:20–42; 19:31–68, 21:181–220, 23:73–174 mosaics, 15:43–84 mushroom, 8:189–215 negatively associated traits, 13:141–177 oat, 6:167–207 oil palm, 4:175–201; 22:165–219 onion, 20:67–103 papaya, 26:35–78 palms, 23:280–281

357 pasture legumes, 5:237–305 pea, snap, 212:93–138 peanut, 22:297–356 pearl millet, 1:162–182 perennial rye, 13:265–292 persimmon, 19:191–225 Philodendron, 23:273 plantain, 2:150–151; 14:267–320; 21:211–25 potato, 3:274–277; 9:217–332; 16:15–86; 19:59–155, 25:1–19 proteins in maize, 9:181–216 quality protein maize (QPM), 9:181–216 raspberry, 6:245–321 recurrent restricted phenotypic selection, 9:101–113 recurrent selection in maize, 9:115–179; 14:139–163 rice, 17:15–156; 23:73–174 rol genes, 26:79–103 Rosaceae, 27:175–211 rose, 17:159–189 rutabaga, 8:217–248 sesame, 16:179–228 snap pea, 21:93–138 somatic hybridization, 20:167–225 sorghum male sterility, 25:139–172 soybean, 1:183–235; 3:289–311; 4:203–243; 21:212–307 soybean hybrids, 21:212–307 soybean nodulation, 11:275–318 soybean recurrent selection, 15:275–313 spelt, 15:187–213 statistics, 17:296–300 strawberry, 2:195–214 sugarcane, 16:272–273; 27:15–158 supersweet sweet corn, 14:189–236 sweet cherry, 9:367–388 sweet corn, 1:139–161; 14:189–236 sweet potato, 4:313–345 Syngonium, 23:274 tomato, 4:273–311 transgene technology, 25:105–108 triticale, 5:41–93; 8:43–90 Vigna, 8:19–42 virus resistance, 12:47–79 wheat, 2:303–319; 3:169–191; 5:11–40; 11:225–234; 13:293–343 wheat for rust resistance, 13:293–343 white clover, 17:191–223 wild rice, 14:237–265 Bringhurst, Royce S. (biography), 9:1–8 Broadbean, in vitro culture, 2:244–245 Bromeliad breeding, 23:275–276 Burton, Glenn W. (biography), 3:1–19

358

C Cactus: breeding, 20:135–166 domestication, 20:135–166 Cajanus, in vitro culture, 2:224 Calathea breeding, 23:276 Canola, R.K. Downey, designer, 18:1–12 Carbohydrates, 1:144–148 Carbon isotope discrimination, 12:81–113 Carica papaya, see Papaya Carnation, mutation breeding, 6:73–74 Carrot breeding, 19:157–190 Cassava: breeding, 2:73–134 long-term selection, 24(2):74–79 Castanea, see Chestnut Cell selection, 4:139–145, 153–173 Cereal breeding, see Grain breeding Cereal diversity, 21:221–261 Cherry, see Sweet cherry domestication, 25:202–293 Chestnut breeding, 4:347–397 Chickpea, in vitro culture, 2:224–225 Chimeras and mosaics, 15:43–84 Chinese cabbage, heat tolerance, 10:152 Chromosome, petunia, 1:13–21, 31–33 Chrysanthemum: breeding, 14:321–361 mutation breeding, 6:74 Cicer, see Chickpea Citrus: domestication, 25:296–298 protoplast fusion, 8:339–374 Clonal repositories, see National Clonal Germplasm Repository Clover: in vitro culture, 2:240–244 molecular genetics, 17:191–223 Coffea arabica, see Coffee Coffee, 2:157–193 Cold hardiness: breeding nectarines and peaches, 10:271–308 wheat adaptation, 12:124–135 Cole crops: Chinese cabbage, heat tolerance, 10:152 gametoclonal variation, 5:371–372 rutabaga, 8:217–248 Coleus, 3:343–360 Competition, 13:158–165 Competitive ability breeding, 14:89–138 Controlling elements, see Transposable elements Corn, see Maize; Sweet corn Cotton, heat tolerance 10:151

CUMULATIVE SUBJECT INDEX Cowpea: breeding, 15:215–274 heat tolerance, 10:147–149 in vitro culture, 2:245–246 photoperiodic response, 3:99 Coyne, Dermot E. (biography), 23:1–19 Cranberry domestication, 25:304–305 Crop domestication and selection, 24(2):1–44 Cryopreservation, 7:125–126, 148–151, 167 buds, 7:168–169 genetic stability, 7:125–126 meristems, 7:168–169 pollen, 7:171–172 seed, 7:148–151, 168 Cucumber, breeding, 6:323–359 Cucumis sativa, see Cucumber Cucurbitaceae: insect and mite resistance, 10:309–360 mapping, 213–244 Cucurbits mapping, 213–244 Cybrids. 3:205–210; 20:206–209 Cytogenetics: alfalfa, 10:171–184 blueberry, 5:325–326 cassava, 2:94 citrus, 8:366–370 coleus, 3:347–348 durum wheat, 5:12–14 fescue, 3:316–319 Glycine, 16:288–317 guayule, 6:99–103 maize mobile elements, 4:81–122 maize-tripsacum hybrids, 20:15–66 oat, 6:173–174 pearl millet, 1:167 perennial rye, 13:265–292 petunia, 1:13–21, 31–32 polyploidy terminology, 26:105–124 potato, 25:1–19 rose, 17:169–171 rye, 13:265–292 Saccharum complex, 16:273–275 sesame, 16:185–189 sugarcane, 27:74–78 triticale, 5:41–93; 8:54 wheat, 5:12–14; 10:5–15; 11:225–234 Cytoplasm: breeding, 23:175–210; 25:115–138 cybrids, 3:205–210; 20:206–209 incompatibility, 25:115–138 male sterility, 25:115–138, 139–172 molecular biology of male sterility, 10:23–51 organelles, 2:283–302; 6:361–393 pearl millet, 1:166

CUMULATIVE SUBJECT INDEX petunia, 1:43–45 sorghum male sterility, 25:139–172 wheat, 2:308–319

359 Dudley, J.W. (biography), 24(1):1–10 Durum wheat, 5:11–40 Duvick, Donald N. (biography), 14:1–11

D

E

Dahlia, mutation breeding, 6:75 Date palm domestication, 25:272–277 Daucus, see Carrot Diallel cross, 9:9–36 Dieffenbachia breeding, 23:271–272 Diospyros, see Persimmon Disease and pest resistance: antifungal proteins, 14:39–88 apple rootstocks, 1:358–373 banana, 2:143–147 barley, 26:135–169 blackberry, 8:291–295 black walnut, 1:251 blueberry, rabbiteye, 5:348–350 cassava, 2:105–114 cell selection, 4:143–145, 163–165 citrus, 8:347–349 coffee, 2:176–181 coleus, 3:353 cowpea, 15:237–247 durum wheat, 5:23–28 fescue, 3:334–336 herbicide-resistance, 11:155–198 host-parasite genetics, 5:393–433 induced mutants, 2:25–30 lettuce, 1:286–287 maize, 119–173 papaya, 26:161–357 potato, 9:264–285, 19:69–155 raspberry, 6:245–321 rutabaga, 8:236–240 soybean, 1:183–235 spelt, 15:195–198 strawberry, 2:195–214 virus resistance, 12:47–79 wheat rust, 13:293–343 Diversity: land races, 21:221–261 legumes, 26:171–357 DNA methylation, 18:87–176 Doubled haploid breeding, 15:141–186; 25:57–88 Douglas-fir breeding, 27:245–353 Downey, Richard K. (biography), 18:1–12 Dracaena breeding, 23:277 Draper, Arlen D. (biography), 13:1–10 Drought resistance: durum wheat, 5:30–31 maize, 25:173–253 soybean breeding, 4:203–243 wheat adaptation, 12:135–146

Elaeis, see Oil palm Embryo culture: in crop improvement, 5:181–236 oil palm, 4:186–187 pasture legume hybrids, 5:249–275 Endosperm: balance number, 25:6–7 maize, 1:139–161 sweet corn, 1:139–161 Endothia parasitica, 4:355–357 Epepremnum breeding, 23:272–273 Epistasis, 21:27–92. Escherichia coli, long-term selection, 24(2):225–224 Evolution: coffee, 2:157–193 fruit, 25:255–320 grapefruit, 13:345–363 maize, 20:15–66 sesame, 16:189 Exploration, 7:9–11, 26–28, 67–94 F Fabaceae, molecular mapping, 14:24–25 Fern breeding, 23:276 Fescue, 3:313–342 Festuca, see Fescue Fig domestication, 25:281–285 Flavonoid chemistry, 25:91–94 Floral biology: almond, 8:314–320 blackberry, 8:267–269 black walnut, 1:238–244 cassava, 2:78–82 chestnut, 4:352–353 coffee, 2:163–164 coleus, 3:348–349 color, 25:89–114 fescue, 3:315–316 garlic, 23:211–244 guayule, 6:103–105 homeotic mutants, 9:63–99 induced mutants, 2:46–50 pearl millet, 1:165–166 pistil in reproduction, 4:9–79 pollen in reproduction, 4:9–79 reproductive barriers, 11:11–154 rutabaga, 8:222–226 sesame, 16:184–185 sweet potato, 4:323–325 Flower color, 25:89–114

360 Forage breeding: alfalfa inbreeding, 13:209–233 diversity, 21:221–261 fescue, 3:313–342 perennials, 11:251–274 white clover, 17:191–223 Foliage plant breeding, 23:245–290 Forest crop breeding: black walnut, 1:236–266 chestnut, 4:347–397 Douglas-fir, 27:245–353 ideotype concept, 12:177–187 molecular markers, 19:31–68 quantitative genetics, 8:139–188 Fragaria, see Strawberry Fruit, nut, and beverage crop breeding: almond, 8:313–338 apple, 9:333–366 apple rootstocks, 1:294–394 banana, 2:135–155 blackberry, 8:249–312 blueberry, 13:1–10 blueberry, rabbiteye, 5:307–357 breeding, 25:255–320 cactus, 20:135–166 cherry, 9:367–388 citrus, 8:339–374 coffee, 2:157–193 domestication, 25:255–320 ideotype concept, 12:175–177 genetic transformation, 16:87–134 grapefruit, 13:345–363 mutation breeding, 6:78–79 nectarine (cold hardy), 10:271–308 origins, 25:255–320 papaya, 26:35–78 peach (cold hardy), 10:271–308 persimmon, 19:191–225 plantain, 2:135–155 raspberry, 6:245–321 strawberry, 2:195–214 sweet cherry, 9:367–388 Fungal diseases: apple rootstocks, 1:365–368 banana and plantain, 2:143–145, 147 barley, Fusarium head blight, 26:125–169 cassava, 2:110–114 cell selection, 4:163–165 chestnut, 4:355–397 coffee, 2:176–179 cowpea, 15:237–238 durum wheat, 5:23–27 Fusarium head blight (barley), 26:125–169 host-parasite genetics, 5:393–433 lettuce, 1:286–287 maize foliar, 125–142

CUMULATIVE SUBJECT INDEX potato, 19:69–155 raspberry, 6:245–281 soybean, 1:188–209 spelt, 15:196–198 strawberry, 2:195–214 sweet potato, 4:333–336 transformation, fruit crops, 16:111–112 wheat rust, 13:293–343 Fusarium head blight (barley), 26:125–169 G Gabelman, Warren H. (biography), 6:1–9 Gametes: almond, self compatibility, 7:322–330 blackberry, 7:249–312 competition, 11:42–46 forest trees, 7:139–188 maize aleurone, 7:91–137 maize anthocynanin, 7:91–137 mushroom, 7:189–216 polyploid, 3:253–288 rutabaga, 7:217–248 transposable elements, 7:91–137 unreduced, 3:253–288 Gametoclonal variation, 5:359–391 barley, 5:368–370 brassica, 5:371–372 potato, 5:376–377 rice, 5:362–364 rye, 5:370–371 tobacco, 5:372–376 wheat, 5:364–368 Garlic breeding, 6:81, 23:211–244 Genes: action, 15:315–374 apple, 9:337–356 Bacillus thuringensis, 12:19–45 incompatibility, 15:19–42 incompatibility in sweet cherry, 9:367–388 induced mutants, 2:13–71 lettuce, 1:267–293 maize endosperm, 1:142–144 maize protein, 1:110–120, 148–149 petunia, 1:21–30 quality protein in maize, 9:183–184 Rhizobium, 23:39–47 rol in breeding, 26:79–103 rye perenniality, 13:261–288 soybean, 1:183–235 soybean nodulation, 11:275–318 sweet corn, 1:142–144 wheat rust resistance, 13:293–343 Genetic engineering: bean, 1:89–91 DNA methylation, 18:87–176 fruit crops, 16:87–134

CUMULATIVE SUBJECT INDEX host-parasite genetics, 5:415–428 legumes, 26:171–357 maize mobile elements, 4:81–122 papaya, 26:35–78. rol genes, 26:79–103 salt resistance, 22:389–425 sugarcane, 27:86–97 transformation by particle bombardment, 13:231–260 transgene technology, 25:105–108 virus resistance, 12:47–79 Genetic load and lethal equivalents, 10:93–127 Genetics: adaptation, 3:21–167 almond, self compatibility, 8:322–330 amaranth, 19:243–248 Amaranthus, see Amaranth apomixis, 18:13–86 apple, 9:333–366 Bacillus thuringensis, 12:19–45 bean seed protein, 1:59–102 beet, 22:357–376 blackberry, 8:249–312 black walnut, 1:247–251 blueberry, 13:1–10 blueberry, rabbiteye, 5:323–325 carrot, 19:164–171 chestnut blight, 4:357–389 chimeras, 15:43–84 chrysanthemums, 14:321 clover, white, 17:191–223 coffee, 2:165–170 coleus, 3:3–53 cowpea, 15:215–274 Cucurbitaceae, 27:213–344 cytoplasm, 23:175–210 DNA methylation, 18:87–176 domestication, 25:255–320 durum wheat, 5:11–40 flower color, 25:89–114 forest trees, 8:139–188 fruit crop transformation, 16:87–134 gene action, 15:315–374 history, 24(1):11–40 host-parasite, 5:393–433 incompatibility, 15:19–42 incompatibility in sweet cherry, 9:367–388 induced mutants, 2:51–54 insect and mite resistance in Cucurbitaceae, 10:309–360 isozymes, 6:11–54 lettuce, 1:267–293 maize aleurone, 8:91–137 maize anther culture, 11:199–224 maize anthocynanin, 8:91–137

361 Maize, foliar diseases, 27:118–173 maize endosperm, 1:142–144 maize male sterility, 10:23–51 maize mobile elements, 4:81–122 maize mutation, 5:139–180 maize seed protein, 1:110–120, 148–149 male sterility, maize, 10:23–51 mapping, 14:13–37 markers to manage germplasm, 13:11–86 maturity, 3:21–167 metabolism and heterosis, 10:53–59 mitochondrial, 25:115–138. molecular mapping, 14:13–37 mosaics, 15:43–84 mushroom, 8:189–216 oat, 6:168–174 organelle transfer, 6:361–393 overdominance, 17:225–257 pea, 21:110–120 pearl millet, 1:166, 172–180 perennial rye, 13:261–288 petunia, 1:1–58 photoperiod, 3:21–167 plantain, 14:264–320 polyploidy terminology, 26:105–124 potato disease resistance, 19:69–165 potato ploidy manipulation, 3:274–277; 16:15–86 quality protein in maize, 9:183–184 quantitative trait loci, 15:85–139 quantitative trait loci in animals selection, 24(2):169–210, 211–224 reproductive barriers, 11:11–154 rhizobia, 21–72 rice, hybrid, 17:15–156, 23:73–174 Rosaceae, 27:175–211 rose, 17:171–172 rutabaga, 8:217–248 salt resistance, 22:389–425 selection, 24(1):111–131, 143–151, 269–290 sesame, 16:189–195 snap pea, 21:110–120 soybean, 1:183–235 soybean nodulation, 11:275–318 spelt, 15:187–213 supersweet sweet corn, 14:189–236 sweet corn, 1:139–161; 14:189–236 sweet potato, 4:327–330 temperature, 3:21–167 tomato fruit quality, 4:273–311 transposable elements, 8:91–137 triticale, 5:41–93 virus resistance, 12:47–79 wheat gene manipulation, 11:225–234 wheat male sterility, 2:307–308 wheat molecular biology, 11:235–250

362 Genetics (cont.) wheat rust, 13:293–343 white clover, 17:191–223 yield, 3:21–167 Genome: Glycine, 16:289–317 Poaceae, 16:276–281 Genomics, grain legumes, 26:171–357 Genotype × environment, interaction, 16:135–178 Germplasm, see also National Clonal Germplasm Repositories; National Plant Germplasm System acquisition and collection, 7:160–161 apple rootstocks, 1:296–299 banana, 2:140–141 blackberry, 8:265–267 black walnut, 1:244–247 cactus, 20:141–145 cassava, 2:83–94, 117–119 chestnut, 4:351–352 coffee, 2:165–172 distribution, 7:161–164 enhancement, 7:98–202 evaluation, 7:183–198 exploration and introduction, 7:9–18, 64–94 genetic markers, 13:11–86 guayule, 6:112–125 isozyme, 6:18–21 grain legumes, 26:171–357 legumes, 26:171–357 maintenance and storage, 7:95–110, 111–128, 129–158, 159–182; 13:11–86 maize, 14:165–187 management, 13:11–86 oat, 6:174–176 peanut, 22:297–356 pearl millet, 1:167–170 plantain, 14:267–320 potato, 9:219–223 preservation, 2:265–282; 23:291–344 rights, 25:21–55 rutabaga, 8:226–227 sesame, 16:201–204 spelt, 15:204–205 sweet potato, 4:320–323 triticale, 8:55–61 wheat, 2:307–313 Gesneriaceae, mutation breeding, 6:73 Gladiolus, mutation breeding, 6:77 Glycine, genomes, 16:289–317 Glycine max, see Soybean Grain breeding: amaranth, 19:227–285 barley, 3:219–252, 5:95–138; 26:125–169

CUMULATIVE SUBJECT INDEX diversity, 21:221–261 doubled haploid breeding, 15:141–186 ideotype concept, 12:173–175 maize, 1:103–138, 139–161; 5:139–180; 9:115–179, 181–216; 11:199–224; 14:165–187; 22:3–4; 24(1):11–40, 41–59, 61–78; 24(2):53–64, 109–151; 25:173–253; 27:119–173 maize history, 24(2):31–59, 41–59, 61–78 oat, 6:167–207 pearl millet, 1:162–182 rice, 17:15–156; 24(2):64–67 sorghum, 25:139–172 spelt, 15:187–213 transformation, 13:231–260 triticale, 5:41–93; 8:43–90 wheat, 2:303–319; 5:11–40; 11:225–234, 235–250; 13:293–343; 22:221–297; 24(2):67–69 wild rice, 14:237–265 Grape: domestication, 25:279–281 transformation, 16:103–104 Grapefruit: breeding, 13:345–363 evolution, 13:345–363 Grass breeding: breeding, 11:251–274 mutation breeding, 6:82 recurrent selection, 9:101–113 transformation, 13:231–260 Growth habit, induced mutants, 2:14–25 Guayule, 6:93–165 H Hallauer, Arnel R. (biography), 15:1–17 Haploidy, see also unreduced and polyploid gametes apple, 1:376 barley, 3:219–252 cereals, 15:141–186 doubled, 15:141–186; 25:57–88 maize, 11:199–224 petunia, 1:16–18, 44–45 potato, 3:274–277; 16:15–86 Harlan, Jack R. (biography), 8:1–17 Heat tolerance breeding, 10:129–168 Herbicide resistance: breeding needs, 11:155–198 cell selection, 4:160–161 decision trees, 18:251–303 risk assessment, 18:251–303 transforming fruit crops, 16:114 Heritability estimation, 22:9–111 Heterosis: gene action, 15:315–374

CUMULATIVE SUBJECT INDEX overdominance, 17:225–257 plant breeding, 12:227–251 plant metabolism, 10:53–90 rice, 17:24–33 soybean, 21:263–320 Honeycomb: breeding, 18:177–249 selection, 13:87–139, 18:177–249 Hordeum, see Barley Host-parasite genetics, 5:393–433 Hyacinth, mutation breeding, 6:76–77 Hybrid and hybridization, see also Heterosis barley, 5:127–129 blueberry, 5:329–341 chemical, 3:169–191 interspecific, 5:237–305 maize high oil selection, 24(1):153–175 maize history, 24(1):31–59, 41–59, 61–78 maize long-term selection, 24(2):43–64, 109–151 overdominance, 17:225–257 rice, 17:15–156 soybean, 21:263;-320 wheat, 2:303–319 I Ideotype concept, 12:163–193 In vitro culture: alfalfa, 2:229–234; 4:123–152 barley, 3:225–226 bean, 2:234–237 birdsfoot trefoil, 2:228–229 blackberry, 8:274–275 broadbean, 2:244–245 cassava, 2:121–122 cell selection, 4:153–173 chickpea, 2:224–225 citrus, 8:339–374 clover, 2:240–244 coffee, 2:185–187 cowpea, 2:245–246 embryo culture, 5:181–236, 249–275 germplasm preservation, 7:125, 162–167 introduction, quarantines, 3:411–414 legumes, 2:215–264 mungbean, 2:245–246 oil palm, 4:175–201 pea, 2:236–237 peanut, 2:218–224 petunia, 1:44–48 pigeon pea, 2:224 pollen, 4:59–61 potato, 9:286–288 sesame, 16:218 soybean, 2:225–228 Stylosanthes, 2:238–240

363 wheat, 12:115–162 wingbean, 2:237–238 zein, 1:110–111 Inbreeding depression, 11:84–92 alfalfa, 13:209–233 cross pollinated crops, 13:209–233 Incompatibility: almond, 8:313–338 molecular biology, 15:19–42 pollen, 4:39–48 reproductive barrier, 11:47–70 sweet cherry, 9:367–388 Incongruity, 11:71–83 Industrial crop breeding: guayule, 6:93–165 sugarcane, 27:5–118 Insect and mite resistance: apple rootstock, 1:370–372 black walnut, 1:251 cassava, 2:107–110 clover, white, 17:209–210 coffee, 2:179–180 cowpea, 15:240–244 Cucurbitaceae, 10:309–360 durum wheat, 5:28 maize, 6:209–243 raspberry, 6:282–300 rutabaga, 8:240–241 sweet potato, 4:336–337 transformation fruit crops, 16:113 wheat, 22:221–297 white clover, 17:209–210 Intergeneric hybridization, papaya, 26:35–78 Interspecific hybridization: blackberry, 8:284–289 blueberry, 5:333–341 citrus, 8:266–270 pasture legume, 5:237–305 rose, 17:176–177 rutabaga, 8:228–229 Vigna, 8:24–30 Intersubspecific hybridization, rice, 17:88–98 Introduction, 3:361–434; 7:9–11, 21–25 Ipomoea, see Sweet potato Isozymes, in plant breeding, 6:11–54 J Jones, Henry A. (biography), 1:1–10 Juglans nigra, see Black walnut K Karyogram, petunia, 1:13 Kiwifruit: domestication, 25:300–301 transformation, 16:104

364 L Lactuca sativa, see Lettuce Landraces, diversity, 21:221–263 Laughnan, Jack R. (bibliography), 19:1–14 Legume breeding, see also Oilseed, Soybean: cowpea, 15:215–274 genomics, 26:171–357 pasture legumes, 5:237–305 Vigna, 8:19–42 Legume tissue culture, 2:215–264 Lethal equivalents and genetic load, 10:93–127 Lettuce: breeding, 16:1–14; 20:105–133 genes, 1:267–293 Lingonberry domestication, 25:300–301 Linkage: bean, 1:76–77 isozymes, 6:37–38 lettuce, 1:288–290 maps, molecular markers, 9:37–61 petunia, 1:31–34 Lotus: hybrids, 5:284–285 in vitro culture, 2:228–229 Lycopersicon, see Tomato M Maize: anther culture, 11:199–224; 15:141–186 anthocyanin, 8:91–137 apomixis, 18:56–64 breeding, 1:103–138, 139–161; 27:119–173 carbohydrates, 1:144–148 cytoplasm, 23:189 doubled haploid breeding, 15:141–186 drought tolerance, 25:173–253 exotic germplasm utilization, 14:165–187 foliar diseases, 27:119–173 high oil, 22:3–4; 24(1):153–175 history of hybrids, 23(1):11–40, 41–59, 61–78 honeycomb breeding, 18:226–227 hybrid breeding, 17:249–251 insect resistance, 6:209–243 long-term selection 24(2):53–64, 109–151 male sterility, 10:23–51 marker-assisted selection. 24(1):293–309 mobile elements, 4:81–122 mutations, 5:139–180 origins, 20:15–66 origins of hybrids, 24(1):31–50, 41–59, 61–78 overdominance, 17:225–257 physiological changes with selection, 24(1):143–151

CUMULATIVE SUBJECT INDEX protein, 1:103–138 quality protein, 9:181–216 recurrent selection, 9:115–179; 14:139–163 RFLF changes with selection, 24(1):111–131 selection for oil and protein, 24(1):79–110, 153–175 supersweet sweet corn, 14:189–236 transformation, 13:235–264 transposable elements, 8:91–137 unreduced gametes, 3:277 Male sterility: chemical induction, 3:169–191 coleus, 3:352–353 genetics, 25:115–138, 139–172 lettuce, 1:284–285 molecular biology, 10:23–51 pearl millet, 1:166 petunia, 1:43–44 rice, 17:33–72 sesame, 16:191–192 sorghum, 139–172 soybean, 21:277–291 wheat, 2:303–319 Malus spp, see Apple Malus ×domestica, see Apple Malvaceae, molecular mapping, 14:25–27 Mango: domestication, 25:277–279 transformation, 16:107 Manihot esculenta, see Cassava Mapping: Cucurbitaceae, 27:213–244 Rosaceae, 27:175–211 Medicago, see also Alfalfa in vitro culture, 2:229–234 Meiosis, petunia, 1:14–16 Metabolism and heterosis, 10:53–90 Microprojectile bombardment, transformation, 13:231–260 Mitochondrial genetics, 6:377–380; 25:115–138 Mixed plantings, bean breeding, 4:245–272 Mobile elements, see also transposable elements: maize, 4:81–122; 5:146–147 Molecular biology: apomixis, 18:65–73 comparative mapping, 14:13–37 cytoplasmic male sterility, 10:23–51 DNA methylation, 18:87–176 herbicide-resistant crops, 11:155–198 incompatibility, 15:19–42 legumes, 26:171–357 molecular mapping, 14:13–37; 19:31–68

CUMULATIVE SUBJECT INDEX molecular markers, 9:37–61, 10:184–190; 12:195–226; 13:11–86; 14:13–37; 17:113–114, 179, 212–215; 18:20–42; 19:31–68, 21:181–220; 23:73–174; 26:292–299 papaya, 26:35–78 quantitative trait loci, 15:85–139 rol genes, 26:79–103 salt resistance, 22:389–425 somaclonal variation, 16:229–268 somatic hybridization, 20:167–225 soybean nodulation, 11:275–318 strawberry, 21:139–180 transposable (mobile) elements, 4:81–122; 8:91–137 virus resistance, 12:47–79 wheat improvement, 11:235–250 Molecular markers, 9:37–61, 10:184–190; 12:195–226; 13:11–86; 14:13–37; 17:113–114, 179, 212–215; 18:20–42; 19:31–68, 21:181–220, 23:73–174 alfalfa, 10:184–190 apomixis, 18:40–42 barley, 21:181–220 clover, white, 17:212–215 forest crops, 19:31–68 fruit crops, 12:195–226 maize selection, 24(1):293–309 mapping, 14:13–37 plant genetic resource mangement, 13:11–86 rice, 17:113–114, 23:73–124 rose, 17:179 somaclonal variation, 16:238–243 wheat, 21:181–220 white clover, 17:212–215 Monosomy, petunia, 1:19 Mosaics and chimeras, 15:43–84 Mungbean, 8:32–35 in vitro culture, 2:245–246 photoperiodic response, 3:74, 89–92 Munger, Henry M. (biography), 4:1–8 Musa, see Banana, Plantain Mushroom, breeding and genetics, 8:189–215 Mutants and mutation: alfalfa tissue culture, 4:130–139 apple rootstocks, 1:374–375 banana, 2:148–149 barley, 5:124–126 blackberry, 8:283–284 cassava, 2:120–121 cell selection, 4:154–157 chimeras, 15:43–84 coleus, 3:355 cytoplasmic, 2:293–295

365 gametoclonal variation, 5:359–391 homeotic floral, 9:63–99 induced, 2:13–72 long term selection variation, 24(1):227–247 maize, 1:139–161, 4:81–122; 5:139–180 mobile elements, see Transposable elements mosaics, 15:43–84 petunia, 1:34–40 sesame, 16:213–217 somaclonal variation, 4:123–152; 5:147–149 sweet corn, 1:139–161 sweet potato, 4:371 transposable elements, 4:181–122; 8:91–137 tree fruits, 6:78–79 vegetatively-propagated crops, 6:55–91 zein synthesis, 1:111–118 Mycoplasma diseases, raspberry, 6:253–254 N National Clonal Germplasm Repository (NCGR), 7:40–43 cryopreservation, 7:125–126 genetic considerations, 7:126–127 germplasm maintenance and storage, 7:111–128 identification and label verification, 7:122–123 in vitro culture and storage, 7:125 operations guidelines, 7:113–125 preservation techniques, 7:120–121 virus indexing and plant health, 7:123–125 National Plant Germplasm System (NPGS), see also Germplasm history, 7:5–18 information systems, 7:57–65 operations, 7:19–56 preservation of genetic resources, 23:291–34 National Seed Storage Laboratory (NSSL), 7:13–14, 37–38, 152–153 Nectarines, cold hardiness breeding, 10:271–308 Nematode resistance: apple rootstocks, 1:368 banana and plantain, 2:145–146 coffee, 2:180–181 cowpea, 15:245–247 soybean, 1:217–221 sweet potato, 4:336 transformation fruit crops, 16:112–113

366 Nicotiana, see Tobacco Nodulation, soybean, 11:275–318 O Oat, breeding, 6:167–207 Oil palm: breeding, 4:175–201, 22:165–219 in vitro culture, 4:175–201 Oilseed breeding: canola, 18:1–20 oil palm, 4:175–201; 22:165–219 sesame, 16:179–228 soybean, 1:183–235; 3:289–311; 4:203–245; 11:275–318; 15:275–313 Olive domestication, 25:277–279 Onion, breeding history, 20:57–103 Opuntia, see Cactus Organelle transfer, 2:283–302; 3:205–210; 6:361–393 Ornamentals breeding: chrysanthemum, 14:321–361 coleus, 3:343–360 petunia, 1:1–58 rose, 17:159–189 Ornithopus, hybrids, 5:285–287 Orzya, see Rice Overdominance, 17:225–257 Ovule culture, 5:181–236 P Palm (Arecaceae): foliage breeding, 23:280–281 oil palm breeding, 4:175–201; 22:165–219. Panicum maximum, apomixis, 18:34–36, 47–49 Papaya: Breeding, 26:35–78 domestication, 25:307–308 transformation, 16:105–106 Parthenium argentatum, see Guayule Paspalum, apomixis, 18:51–52 Paspalum notatum, see Pensacola bahiagrass Passionfruit transformation, 16:105 Pasture legumes, interspecific hybridization, 5:237–305 Pea: breeding, 21:93–138 flowering, 3:81–86, 89–92 in vitro culture, 2:236–237 Peach: cold hardiness breeding, 10:271–308 domestication, 25:294–296 transformation, 16:102

CUMULATIVE SUBJECT INDEX Peanut: breeding, 22:297–356 in vitro culture, 2:218–224 Pear: domestication, 25:289–290 transformation, 16:102 Pearl millet: apomixis, 18:55–56 breeding, 1:162–182 Pecan transformation, 16:103 Peloquin, Stanley, J. (biography), 25:1–19 Pennisetum americanum, see Pearl millet Pensacola bahiagrass, 9:101–113 apomixis, 18:51–52 selection, 9:101–113 Pepino transformation, 16:107 Peppermint, mutation breeding, 6:81–82 Perennial grasses, breeding, 11:251–274 Perennial rye breeding, 13:261–288 Persimmon: breeding, 19:191–225 domestication, 25:299–300 Petunia spp., genetics, 1:1–58 Phaseolin, 1:59–102 Phaseolus vulgaris, see Bean Philodendrum breeding, 23:273 Phytophthora fragariae, 2:195–214 Pigeon pea, in vitro culture, 2:224 Pineapple domestication, 25:305–307 Pistil, reproductive function, 4:9–79 Pisum, see Pea Plant breeders; rights, 25:21–55 Plant breeding: politics, 25:21–55 prediction, 15–40 Plant introduction, 3:361–434; 7:9–11, 21–25 Plant exploration, 7:9–11, 26–28, 67–94 Plantain: breeding, 2:135–155; 14:267–320; 21:1–25 domestication, 25:298 Plastid genetics, 6:364–376, see also Organelle Plum: domestication, 25:293–294 transformation, 16:103–140 Poaceae: molecular mapping, 14:23–24 Saccharum complex, 16:269–288 Pollen: reproductive function, 4:9–79 storage, 13:179–207 Polyploidy, see also Haploidy alfalfa, 10:171–184 alfalfa tissue culture, 4:125–128

CUMULATIVE SUBJECT INDEX apple rootstocks, 1:375–376 banana, 2:147–148 barley, 5:126–127 blueberry, 13:1–10 gametes, 3:253–288 isozymes, 6:33–34 petunia, 1:18–19 potato, 16:15–86; 25:1–19 reproductive barriers, 11:98–105 sweet potato, 4:371 terminology, 26:105–124 triticale, 5:11–40 Pomegranate domestication, 25:285–286 Population genetics, see Quantitative Genetics Potato: breeding, 9:217–332, 19:69–165 cytoplasm, 23:187–189 disease resistance breeding, 19:69–165 gametoclonal variation, 5:376–377 heat tolerance, 10:152 honeycomb breeding, 18:227–230 mutation breeding, 6:79–80 photoperiodic response, 3:75–76, 89–92 ploidy manipulation, 16:15–86 unreduced gametes, 3:274–277 Protein: antifungal, 14:39–88 bean, 1:59–102 induced mutants, 2:38–46 maize, 1:103–138, 148–149; 9:181–216 Protoplast fusion, 3:193–218; 20:167–225 citrus, 8:339–374 mushroom, 8:206–208 Prunus: amygdalus, see Almond avium, see Sweet cherry Pseudograin breeding, amaranth, 19:227–285 Psophocarpus, in vitro culture, 2:237–238 Q Quantitative genetics: epistasis, 21:27–92 forest trees, 8:139–188 gene interaction, 24(1):269–290 genotype × environment interaction, 16:135–178 heritability, 22:9–111 maize RFLP changes with selection, 24(1):111–131 mutation variation, 24(1):227–247 overdominance, 17:225–257 population size & selection, 24(1):249–268 selection limits, 24(1):177–225

367 statistics, 17:296–300 trait loci (QTL), 15:85–139; 19:31–68 variance, 22:113–163 Quantitative trait loci (QTL), 15:85–138; 19:31–68 animal selection, 24(2):169–210, 211–224 selection limits:24(1):177–225 Quarantines, 3:361–434; 7:12, 35 R Rabbiteye blueberry, 5:307–357 Raspberry, breeding, 6:245–321 Recurrent restricted phenotypic selection, 9:101–113 Recurrent selection, 9:101–113, 115–179; 14:139–163 soybean, 15:275–313 Red stele disease, 2:195–214 Rédei, George P. (bibliography), 26:1–33. Regional trial testing, 12:271–297 Reproduction: barriers and circumvention, 11:11–154 foliage plants, 23:255–259 garlic, 23:211–244 Rhizobia, 23:21–72 Rhododendron, mutation breeding, 6:75–76 Rice, see also Wild rice: anther culture, 15:141–186 apomixis, 18:65 cytoplasm, 23:189 doubled haploid breeding, 15:141–186 gametoclonal variation, 5:362–364 heat tolerance, 10:151–152 honeycomb breeding, 18:224–226 hybrid breeding, 17:1–15, 15–156; 23:73–174 long-term selection 24(2):64–67 molecular markers, 73–174 photoperiodic response, 3:74, 89–92 Rosa, see Rose Rosaceae, synteny, 27:175–211 Rose breeding, 17:159–189 Rubus, see Blackberry, Raspberry Rust, wheat, 13:293–343 Rutabaga, 8:217–248 Ryder, Edward J. (biography), 16:1–14 Rye: gametoclonal variation, 5:370–371 perennial breeding, 13:261–288 triticale, 5:41–93 S Saccharum complex, 16:269–288 Salt resistance: cell selection, 4:141–143

368 Salt resistance (cont.) durum wheat, 5:31 yeast systems, 22:389–425 Sears, Ernest R. (biography), 10:1–22 Secale, see Rye Seed: apple rootstocks, 1:373–374 banks, 7:13–14, 37–40, 152–153 bean, 1:59–102 garlic, 23:211–244 lettuce, 1:285–286 maintenance and storage, 7:95–110, 129–158, 159–182 maize, 1:103–138, 139–161, 4:81–86 pearl millet, 1:162–182 protein, 1:59–138, 148–149 rice production, 17:98–111, 118–119, 23:73–174 soybean, 1:183–235, 3:289–311 synthetic, 7:173–174 variegation, 4:81–86 wheat (hybrid), 2:313–317 Selection, see also Breeding bacteria, 24(2):225–265 bean, 24(2):69–74 cell, 4:139–145, 153–173 crops of the developing world, 24(2):45–88 divergent selection for maize ear length, 24(2):153– 168 domestication, 24(2):1–44 Escherichia coli, 24(2):225–265 gene interaction, 24(1):269–290 genetic models, 24(1):177–225 honeycomb design, 13:87–139; 18:177–249 limits, 24(1):177–225 maize high oil, 24(1):153–175 maize history, 24(1):11–40, 41–59, 61–78 maize long term, 24(1):79–110, 111– 131, 133–151; 24(2):53–64, 109–151 maize oil & protein, 24(1):79–110, 153–175 maize physiological changes, 24(1):133–151 maize RFLP changes, 24(1):111–131 marker assisted, 9:37–61, 10:184–190; 12:195–226; 13:11–86; 14:13–37; 17:113–114, 179, 212–215; 18:20–42; 19:31–68, 21:181–220, 23:73–174, 24(1):293–309; 26:292–299 mutation variation, 24(1):227–268 population size, 24(1):249–268 prediction, 19:15–40 productivity gains in US crops, 24(2):89–106

CUMULATIVE SUBJECT INDEX quantitative trait loci, 24(1):311–335 recurrent restricted phenotypic, 9:101–113 recurrent selection in maize, 9:115–179; 14:139–163 rice, 24(2):64–67 wheat, 24(2):67–69 Sesame breeding, 16:179–228 Sesamum indicum, see Sesame Simmonds, N.W. (biography), 21:1–13 Snap pea breeding, 21:93–138 Solanaceae: incompatibility, 15:27–34 molecular mapping, 14:27–28 Solanum tuberosum, see Potato Somaclonal variation, see also Gametoclonal variation alfalfa, 4:123–152 isozymes, 6:30–31 maize, 5:147–149 molecular analysis, 16:229–268 mutation breeding, 6:68–70 rose, 17:178–179 transformation interaction, 16:229–268 utilization, 16:229–268 Somatic embryogenesis, 5:205–212; 7:173–174 oil palm, 4:189–190 Somatic genetics, see also Gametoclonal variation; Somaclonal variation: alfalfa, 4:123–152 legumes, 2:246–248 maize, 5:147–149 organelle transfer, 2:283–302 pearl millet, 1:166 petunia, 1:43–46 protoplast fusion, 3:193–218 wheat, 2:303–319 Somatic hybridization, see also Protoplast fusion, 20:167–225 Sorghum: male sterility, 25:139–172 photoperiodic response, 3:69–71, 97–99 transformation, 13:235–264 Southern pea, see Cowpea Soybean: cytogenetics, 16:289–317 disease resistance, 1:183–235 drought resistance, 4:203–243 hybrid breeding, 21:263–307 in vitro culture, 2:225–228 nodulation, 11:275–318 photoperiodic response, 3:73–74 recurrent selection, 15:275–313 semidwarf breeding, 3:289–311

CUMULATIVE SUBJECT INDEX Spelt, agronomy, genetics, breeding, 15:187–213 Sprague, George F. (biography), 2:1–11 Starch, maize, 1:114–118 Statistics: advanced methods, 22:113–163 history, 17:259–316 Sterility, see also Male sterility, 11:30–41 Strawberry: biotechnology, 21:139–180 domestication, 25:302–303 red stele resistance breeding, 2:195–214 transformation, 16:104 Stress resistance: cell selection, 4:141–143, 161–163 transformation fruit crops, 16:115 Stylosanthes, in vitro culture, 2:238–240 Sugarcane: Breeding, 27:15–118 mutation breeding, 6:82–84 Saccharum complex, 16:269–288 Synteny, Rosaceae, 27:175–211 Sweet cherry: Domestication, 25:202–293 pollen-incompatibility and self-fertility, 9:367–388 transformation, 16:102 Sweet corn, see also Maize: endosperm, 1:139–161 supersweet (shrunken2), 14:189–236 Sweet potato breeding, 4:313–345; 6:80–81 T Tamarillo transformation, 16:107 Taxonomy: amaranth, 19:233–237 apple, 1:296–299 banana, 2:136–138 blackberry, 8:249–253 cassava, 2:83–89 chestnut, 4:351–352 chrysanthemum, 14:321–361 clover, white, 17:193–211 coffee, 2:161–163 coleus, 3:345–347 fescue, 3:314 garlic, 23:211–244 Glycine, 16:289–317 guayule, 6:112–115 oat, 6:171–173 pearl millet, 1:163–164 petunia, 1:13 plantain, 2:136; 14:271–272 rose, 17:162–169 rutabaga, 8:221–222 Saccharum complex, 16:270–272

369 sweet potato, 4:320–323 triticale, 8:49–54 Vigna, 8:19–42 white clover, 17:193–211 wild rice, 14:240–241 Testing: adaptation, 12:271–297 honeycomb design, 13:87–139 Tissue culture, see In vitro culture Tobacco, gametoclonal variation, 5:372–376 Tomato: breeding for quality, 4:273–311 heat tolerance, 10:150–151 Toxin resistance, cell selection, 4:163–165 Transformation and transgenesis alfalfa, 10:190–192 barley, 26:155–157 cereals, 13:231–260 fruit crops, 16:87–134 mushroom, 8:206 papaya, 26:35–78 rice, 17:179–180 somaclonal variation, 16:229–268 sugarcane, 86–97 maize breeding, 142–156 white clover, 17:193–211 Transpiration efficiency, 12:81–113 Transposable elements, 4:81–122; 5:146–147; 8:91–137 Tree crops, ideotype concept, 12:163–193 Tree fruits, see Fruit, nut, and beverage crop breeding Trifolium, see Clover, White Clover Trifolium hybrids, 5:275–284 in vitro culture, 2:240–244 Trilobium, long-term selection, 24(2):211–224 Tripsacum: apomixis, 18:51 maize ancestry, 20:15–66 Trisomy, petunia, 1:19–20 Triticale, 5:41–93; 8:43–90 Triticosecale, see Triticale Triticum: Aestivum, see Wheat Turgidum, see Durum wheat Tulip, mutation breeding, 6:76 U United States National Plant Germplasm System, see National Plant Germplasm System Unreduced and polyploid gametes, 3:253–288; 16:15–86 Urd bean, 8:32–35

370 V Vaccinium, see Blueberry Variance estimation, 22:113–163 Vegetable breeding: artichoke, 12:253–269 bean, 1:59–102; 4:245–272, 24(2):69–74 bean (tropics), 10:199–269 beet (table), 22:257–388 carrot 19:157–190 cassava, 2:73–134; 24(2):74–79 cucumber, 6:323–359 cucurbit insect and mite resistance, 10:309–360 lettuce, 1:267–293; 16:1–14; 20:105:-133 mushroom, 8:189–215 onion, 20:67–103 pea, 21:93–138 peanut, 22:297–356 potato, 9:217–232; 16:15–86l; 19:69–165 rutabaga, 8:217–248 snap pea, 21:93–138 tomato, 4:273–311 sweet corn, 1:139–161; 14:189–236 sweet potato, 4:313–345 Vicia, in vitro culture, 2:244–245 Vigna, see Cowpea, Mungbean in vitro culture, 2:245–246; 8:19–42 Virus diseases: apple rootstocks, 1:358–359 clover, white, 17:201–209 coleus, 3:353 cowpea, 15:239–240 indexing, 3:386–408, 410–411, 423–425 in vitro elimination, 2:265–282 lettuce, 1:286 maize, 142–156 papaya, 26:35–78 potato, 19:122–134 raspberry, 6:247–254 resistance, 12:47–79 soybean, 1:212–217 sweet potato, 4:336 transformation fruit crops, 16:108–110 white clover, 17:201–209

CUMULATIVE SUBJECT INDEX Vogel, Orville A. (biography), 5:1–10 Vuylsteke, Dirk R. (biography), 21:1–25 W Walnut (black), 1:236–266 Walnut transformation, 16:103 Weinberger, John A. (biography), 11:1–10 Wheat: anther culture, 15:141–186 apomixis, 18:64–65 chemical hybridization, 3:169–191 cold hardiness adaptation, 12:124–135 cytogenetics, 10:5–15 cytoplasm, 23:189–190 diversity, 21:236–237 doubled haploid breeding, 15:141–186 drought tolerance, 12:135–146 durum, 5:11–40 gametoclonal variation, 5:364–368 gene manipulation, 11:225–234 heat tolerance, 10:152 hybrid, 2:303–319; 3:185–186 insect resistance, 22:221–297 in vitro adaptation, 12:115–162 long-term selection, 24(2):67–69 molecular biology, 11:235–250 molecular markers, 21:191–220 photoperiodic response, 3:74 rust interaction, 13:293–343 triticale, 5:41–93 vernalization, 3:109 White clover, molecular genetics, 17:191–223 Wild rice, breeding, 14:237–265 Winged bean, in vitro culture, 2:237–238 Y Yeast, salt resistance, 22:389–425 Yuan, Longping (biography), 17:1–13. Z Zea mays, see Maize, Sweet corn Zein, 1:103–138 Zizania palustris, see Wild rice

Cumulative Contributor Index (Volumes 1–27) Abbott, A.G., 27:175 Abdalla, O.S., 8:43 Acquaah, G., 9:63 Aldwinckle, H.S., 1:294 Alexander, D.E., 24(1):53 Anderson, N.O., 10:93; 11:11 Aronson, A.I., 12:19 Arús, P., 27:175 Ascher, P.D., 10:93 Ashri, A., 16:179 Baggett, J.R. 21:93 Balaji, J., 26:171 Baltensperger, D.D., 19:227 Barker, T., 25:173 Basnizki, J., 12:253 Beck, D.L., 17:191 Beebe, S., .23:21–72 Beineke, W.F., 1:236 Bell, A.E., 24(2):211 Below, F.E., 24(1):133 Berzonsky, W.A., 22:221 Bingham, E.T., 4:123; 13:209 Binns, M.R., 12:271 Bird, R. McK., 5:139 Bjarnason, M., 9:181 Blair, M.W., 26:171 Bliss, F.A., 1:59; 6:1 Boase, M.R., 14:321 Borlaug, N.E., 5:1 Boyer, C.D., 1:139 Bravo, J.E., 3:193 Brenner, D.M., 19:227 Bressan , R.A., 13:235; 14:39; 22:389 Bretting, P.K., 13:11 Broertjes, C., 6:55 Brown, A.H.D., 21:221 Brown, J.W.S., 1:59 Brown, S.K., 9:333, 367

Buhariwalla, H.K., 26:171 Bünger, L., 24(2):169 Burnham, C.R., 4:347 Burton, G.W., 1:162; 9:101 Burton, J.W., 21:263 Byrne, D., 2:73 Camadro, E.L., 26:105 Campbell, K.G., 15:187 Campos, H., 25:173 Cantrell, R.G., 5:11 Carputo, D., 25:1; 26:105 Carvalho, A., 2:157 Casas, A.M., 13:235 Cervantes-Martinez, C.T., 22:9 Chen, J., 23:245 Cherry, M., 27:245. Chew, P.S., 22:165 Choo, T.M., 3:219; 26:125 Christenson, G.M., 7:67 Christie, B.R., 9:9 Clark, R.L., 7:95 Clarke, A.E., 15:19 Clegg, M.T., 12:1 Comstock, J.G., 27:15 Condon, A.G., 12:81 Cooper, M, 24(2):109; 25:173 Cooper, R.L., 3:289 Cornu, A., 1:11 Costa, W.M., 2:157 Cregan, P., 12:195 Crouch, J.H., 14:267; 26:171 Crow, J.F., 17:225 Cummins, J.N., 1:294 Dana, S., 8:19 Dean, R.A., 27:213 De Jong, H., 9:217

D’Hont, A., 27:15 Dekkers, J.C.M., 24(1):311 Deroles, S.C., 14:321 Dhillon, B.S., 14:139 Dickmann, D.I., 12:163 Ding, H., 22:221 Dirlewanger, E., 27:175 Dodds, P.N., 15:19 Dolan, D., 25:175 Donini, P., 21:181 Draper, A.D., 2:195 Drew, R., 26:35 Dudley, J.W. 24(1):79 Dumas, C., 4:9 Duncan, D.R., 4:153 Duvick, D.N., 24(2):109 Dwivedi, S.L., 26:171 Echt, C.S., 10:169 Edmeades, G., 25:173 Ehlers, J.D., 15:215 England, F., 20:1 Eubanks, M.W., 20:15 Evans, D.A., 3:193; 5:359 Everett, L.A., 14:237 Ewart, L.C., 9:63 Farquhar, G.D., 12:81 Fasoula, D.A., 14:89; 15:315; 18:177 Fasoula, V.A., 13:87; 14:89; 15:315; 18:177 Fasoulas, A.C., 13:87 Fazuoli, L.C., 2:157 Fear, C.D., 11:1 Ferris, R.S.B., 14:267 Flore, J.A., 12:163 Forsberg, R.A., 6:167 Forster, B.P., 25:57 Forster, R.L.S., 17:191

Plant Breeding Reviews, Volume 27, Edited by Jules Janick ISBN 0-471-73213-3 © 2006 John Wiley & Sons, Inc. 371

372

CUMULATIVE CONTRIBUTOR INDEX

Fowler, C., 25:21 Frusciante, L., 35:1 Frei, U., 23:175 French, D.W., 4:347

Howe, G.T., 27:245 Hunt, L.A., 16:135 Hutchinson, J.R., 5:181 Hymowitz, T., 8:1; 16:289

Gai, J., 21:263 Galiba, G., 12:115 Galletta, G.J., 2:195 Gepts, P., 24(2):1 Glaszmann, J.G., 27:15 Gmitter, F.G., Jr., 8:339; 13:345 Gold, M.A., 12:163 Goldman, I.L. 19:15; 20:67; 22:357; 24(1):61; 24(2):89 Gonsalves, D., 26:35 Goodnight, C.J, 24(1):269 Gordon, S.G., 27:119 Gradziel, T.M., 15:43 Gressel, J., 11:155; 18:251 Gresshof, P.M., 11:275 Griesbach, R.J., 25:89 Grombacher, A.W., 14:237 Grosser, J.W., 8:339 Grumet, R., 12:47 Gudin, S., 17:159 Guimarães, C.T., 16:269 Gustafson, J.P., 5:41; 11:225 Guthrie, W.D., 6:209

Janick, J., 1:xi; 23:1; 25:255 Jansky, S., 19:77 Jayaram, Ch., 8:91 Jayawickrama, K., 27:245 Jenderek, M.M., 23:211 Jifon, J., 27:15 Joobeur, T., 27:213 Johnson, A.A.T., 16:229; 20:167 Johnson, G.R., 27:245 Johnson, R., 24(1):293 Jones, A., 4:313 Jones, J.S., 13:209 Ju, G.C., 10:53

Habben, J., 25:173 Haley, S.D., 22:221 Hall, A.E., 10:129; 12:81; 15:215 Hall, H.K., 8:249 Hallauer, A.R., 9:115; 14:1, 165; 24(2):153 Hamblin, J., 4:245 Hancock, J.F., 13:1 Hancock, J.R., 9:1 Hanna, W.W., 13:179 Harlan, J.R., 3:1 Harris, M.O., 22:221 Hasegawa, P.M. 13:235; 14:39:22:389 Havey, M.J., 20:67 Henny, R.J., 23:245 Hill, W.G., 24(2):169 Hillel, J., 12:195 Hodgkin, T., 21:221 Hokanson, S.C., 21:139 Holbrook, C.C., 22:297 Holland, J.B., 21:27; 22:9 Hor, T.Y., 22:165

Kang, H., 8:139 Kann, R.P., 4:175 Karmakar, P.G., 8:19 Kartha, K.K., 2:215, 265 Kasha, K.J., 3:219 Keep, E., 6:245 Kleinhofs, A., 2:13 Keightley, P.D., 24(1):227 Knox, R.B., 4:9 Koebner, R.M.D., 21:181 Kollipara, K.P., 16:289 Koncz, C., 26:1 Kononowicz, A.K., 13:235 Konzak, C.F., 2:13 Krikorian, A.D., 4:175 Krishnamani, M.R.S., 4:203 Kronstad, W.E., 5:1 Kulakow, P.A., 19:227 Lamb, R.J., 22:221 Lambert, R.J., 22:1; 24(1):79, 153 Lamborn, C., 21:93 Lamkey, K.R., 15:1; 24(1):xi; 24(2):xi Lavi, U., 12:195 Layne, R.E.C., 10:271 Lebowitz, R.J., 3:343 Lee, M., 24(2):153 Lehmann, J.W., 19:227 Lenski, R.E., 24(2):225 Levings, III, C.S., 10:23 Lewers, K.R., 15:275 Li, J., 17:1, 15 Liedl, B.E., 11:11

Lin, C.S., 12:271 Lovell, G.R., 7:5 Lower, R.L., 25:21 Lukaszewski, A.J., 5:41 Lyrene, P.M., 5:307 Maas, J. L., 21:139 Mackenzie, S.A., 25:115 Maheswaran, G., 5:181 Maizonnier, D., 1:11 Marcotrigiano, M., 15:43 Martin, F.W., 4:313 Matsumoto, T.K. 22:389 McCoy, T.J., 4:123; 10:169 McCreight, J.D., 1:267; 16:1 McDaniel, R.G., 2:283 McKeand, S.E., 19:41 McKenzie, R.I.H., 22:221 McRae, D.H., 3:169 Medina-Filho, H.P., 2:157 Mejaya, I.J., 24(1):53 Mikkilineni, V., 24(1):111 Miles, D., 24(2):211 Miles, J.W., 24(2):45 Miller, R., 14:321 Ming, R., 27:15 Mirkov, T.E., 27:15 Mondragon Jacobo, C., 20:135 Moose, S.P., 24(1):133 Moore, P.H., 27:15 Morrison, R.A., 5:359 Mowder, J.D., 7:57 Mroginski, L.A., 2:215 Muir, W.M., 24(2):211 Mumm, R.H., 24(1):1 Murphy, A.M., 9:217 Mutschler, M.A., 4:1 Myers, J.R., 21:93 Myers, O., Jr., 4:203 Myers, R.L., 19:227. Namkoong, G., 8:139 Navazio, J., 22:357 Neuffer, M.G., 5:139 Newbigin, E., 15:19 Nyquist, W.E., 22:9 Ohm, H.W., 22:221 O’Malley, D.M., 19:41 Ortiz, R., 14:267; 16:15; 21:1; 25:1, 139; 26:171 Osborn, T.C., 27:1 Palmer, R.G., 15:275, 21:263 Pandy, S., 14:139; 24(2):45

CUMULATIVE CONTRIBUTOR INDEX Pardo, J. M., 22:389 Parliman, B.J., 3:361 Paterson, A.H., 14:13; 26:15 Patterson, F.L., 22:221 Peairs, F.B., 22:221 Pedersen, J.F., 11:251 Peiretti, E.G., 23:175 Peloquin, S.J., 26:105 Perdue, R.E., Jr., 7:67 Peterson, P.A., 4:81; 8:91 Polidorus, A.N., 18:87 Porter, D.A., 22:221 Porter, R.A., 14:237 Powell, W., 21:181 Prasartsee, V., 26:35 Pratt, R.C., 27:119 Proudfoot, K.G., 8:217 Rackow, G., 18:1 Rai, M., 27:15 Raina, S.K., 15:141 Ramage, R.T., 5:95 Ramesh, S., 25:139 Ramming, D.W., 11:1 Ratcliffe, R.H., 22:221 Ray, D.T., 6:93 Reddy, B.V.S., 25:139 Redei, G.P., 10:1; 24(1):11 Reimann-Phillipp, R., 13:265 Reinbergs, E., 3:219 Rhodes, D., 10:53 Richards, R.A., 12:81 Roath, W.W., 7:183 Robinson, R.W., 1:267; 10:309 Rochefored.T.R., 24(1):111 Ron Parra, J., 14:165 Roos, E.E., 7:129 Ross, A.J., 24(2):153 Rotteveel, T., 18:251 Rowe, P., 2:135 Russell, W.A., 2:1 Rutter, P.A., 4:347 Ryder, E.J., 1:267; 20:105 Samaras, Y., 10:53 Sansavini, S., 16:87 Saunders, J.W., 9:63 Savidan, Y., 18:13 Sawhney, R.N., 13:293 Schaap, T., 12:195 Schaber, M.A. 24(2):89 Schnell, R.J., 27:15 Schussler, J., 25:173

Schneerman, M.C. 24(1):133 Schroeck, G., 20:67 Scott, D.H., 2:195 Seabrook, J.E.A., 9:217 Sears, E.R., 11:225 Seebauer, J.R., 24(1):133 Serraj, R., 26:171 Shands, Hazel L., 6:167 Shands, Henry L., 7:1, 5 Shannon, J.C., 1:139 Shanower, T.G., 22:221 Shattuck, V.I., 8:217; 9:9 Shaun, R., 14:267 Sidhu, G.S., 5:393 Silva, da, J., 27:15 Simmonds, N.W., 17:259 Simon, P.W., 19:157; 23:211 Singh, B.B., 15:215 Singh, R.J., 16:289 Singh, S.P., 10:199 Singh, Z., 16:87 Slabbert, M.M., 19:227 Sleper, D.A., 3:313 Sleugh, B.B., 19:227 Smith, J.S.C., 24(2):109 Smith, S.E., 6:361 Snoeck, C., 23:21 Sobral, B.W.S., 16:269 Socias i Company, R., 8:313 Soh, A.C., 22:165 Sondahl, M.R., 2:157 Spoor, W., 20:1 Steadman, J.R., 23:1 Stalker, H.T., 22:297 Steffensen, D. M., 19:1 Stevens, M.A., 4:273 Stoner, A.K., 7:57 Stuber, C.W., 9:37; 12:227 Sugiura, A., 19:191 Sun, H. 21:263 Suzaki, J.Y., 26:35 Tai, G.C.C., 9:217 Talbert, L.E., 11:235 Tan, C.C., 22:165 Tarn, T.R., 9:217 Tehrani, G., 9:367 Teshome, A., 21:221 Tew, T.L., 27:15 Thomas, W.T.B., 25:57 Thompson, A.E., 6:93 Tiefenthaler, A.E. 24(2):89 Towill, L.E., 7:159, 13:179 Tracy, W.F., 14:189; 24(2):89 Tripathi, S., 26:35

373 Troyer, A.F., 24(1):41 Tsaftaris, A.S., 18:87 Tsai, C.Y., 1:103 Ullrich, S.E., 2:13 Upadhyaya, H.D., 26:171 Uribelarrea, M., 24(1):133 Vanderleyden, J., .23:21 Van Harten, A.M., 6:55 Varughese, G., 8:43 Vasal, S.K., 9:181; 14:139 Vegas, A., 26:35 Veilleux, R., 3:253; 16:229; 20:167 Villareal, R.L., 8:43 Vogel, K.P., 11:251 Volk, G.M., 23:291 Vuylsteke, D., 14:267 Wallace, D.H., 3:21; 13:141 Walsh, B. 24(1):177 Wan, Y., 11:199 Wang, Y.-H., 27:213 Waters, C., 23:291 Weber, K., 24(1):249 Weeden, N.F., 6:11 Wehner, T.C., 6:323 Welander, M., 26:79 Wenzel, G. 23:175 Westwood, M.N., 7:111 Wheeler, N.C., 27:245 Whitaker, T.W., 1:1 White, D.W.R., 17:191 White, G.A., 3:361; 7:5 Widholm, J.M., 4:153; 11:199 Widmer, R.E., 10:93 Widrlechner, M.P., 13:11 Wilcox, J.R., 1:183 Williams, E.G., 4:9; 5:181, 237 Williams, M.E., 10:23 Wilson, J.A., 2:303 Wong, G., 22:165 Woodfield, D.R., 17:191 Wright, D., 25:173 Wright, G.C., 12:81 Wu, K.-K., 27:15 Wu, L., 8:189 Wu, R., 19:41 Xin, Y., 17:1, 15 Xu, S., 22:113 Xu, Y., 15:85; 23:73

374 Yamada, M., 19:191 Yamamoto, T., 27:175 Yan, W., 13:141 Yang, W.-J., 10:53 Yonemori, K., 19:191 Yopp, J.H., 4:203 Yun, D.-J., 14:39

CUMULATIVE CONTRIBUTOR INDEX Zeng, Z.-B., 19:41 Zhu, L.-H., 26:79 Zimmerman, M.J.O., 4:245 Zinselmeier, C., 25:173 Zohary, D., 12:253

a

b

c

Plate 2.1. Chromosome preparation stained with DAPI after FISH of the 18-5.8S-26S rDNA (pTA71) detected with FITC (green) and the 5S rDNA detected with Texas red (red). (a) S. officinarum (2n=80); (b) S. spontaneum (2n=64); (c) Narenga (2n=30). Scale bars = 10 µm.

a

b

c

Plate 2.2. (a) Chromosome preparation of a S. barberi clone with 2n=82 after GISH using S. officinarum genomic DNA detected with FITC (green), S. spontaneum genomic DNA detected with Texas red (red) and the 18S-26S rDNA detected in blue (coumarin) (from D’Hont et al. 2001); (b) Chromosome preparation of the sugarcane cultivar Nco376 (2n=ca 112) after GISH using S. officinarum genomic DNA detected with FITC (green), S. spontaneum genomic DNA detected with Texas red (red) and the 18S-26S rDNA detected in blue (coumarin); (c) Chromosome preparation of an S. officinarum × E. arundinaceus hybrid after GISH using S. officinarum genomic DNA detected with FITC (green) and E. arundinaceus genomic DNA detected with Texas red (red)(from D’Hont et al. 1995). Scale bars = 10 µm.