Plant Breeding Reviews (Volume 23)

PLANT BREEDING REVIEWS Volume 23 Plant Breeding Reviews is sponsored by: American Society for Horticultural Science C...

Author: Jules Janick

350 downloads 2660 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

PLANT BREEDING REVIEWS Volume 23

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

Editorial Board, Volume 23 M. Gilbert I. L. Goldman C. H. Michler

PLANT BREEDING REVIEWS Volume 23

edited by

Jules Janick Purdue University

John Wiley & Sons, Inc.

This book is printed on acid-free paper. Copyright © 2003 by John Wiley & Sons. 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: 0-471-35421-X ISSN: 0730-2207 Printed in the United States of America 10

9

8

7

6

5

4

3

2

1

Contents List of Contributors 1. Dedication: Dermot P. Coyne Bean Breeder, Geneticist, Humanitarian

vii 1

James R. Steadman and Jules Janick

2. Strategies for Genetic Improvement of Common Bean and Rhizobia Towards Efficient Interactions

21

Carla Snoeck, Jos Vanderleyden, and Stephen Beebe I. Rhizobium-Common Bean Symbiosis II. Bean Breeding III. Selection of Optimized Rhizobium Strains for Bean Inoculation IV. Conclusions and Future Prospects Literature Cited

3. Developing Marker-Assisted Selection Strategies for Breeding Hybrid Rice

22 26 36 53 58

73

Yunbi Xu I. II. III. IV. V. VI. VII. VIII. IX. X. XI.

Introduction Features of Hybrid Breeding Components of Marker-Assisted Selection Germplasm Evaluation Traits Requiring Testcrossing or Progeny Testing Environment-Dependent Traits Quality Traits Gene Introgression and Whole Genome Selection Prediction of Hybrid Performance and Heterosis Seed Quality Assurance General Discussions Literature Cited

75 77 81 90 99 107 112 117 125 147 151 156 v

vi

CONTENTS

4. Significance of Cytoplasmic DNA in Plant Breeding

175

Ursula Frei, Edmundo G. Peiretti, and Gerhard Wenzel I. Introduction II. Some Basic Information on DNA in the Cytoplasm of Plants III. Agronomic Traits Influenced by Cytoplasmic Factors IV. Breeding Using Cytoplasmic Factors V. Conclusion Literature Cited

5. Flowering, Seed Production, and the Genesis of Garlic Breeding

176 177 179 187 201 203

211

Philipp W. Simon and Maria M. Jenderek I. II. III. IV. V. VI. VII.

Introduction Garlic Production Trends Garlic Taxonomy and Genetic Variation Garlic Growth and Reproductive Biology Garlic Seed Production Progress in Garlic Breeding and Future Prospects Conclusions Literature Cited

6. Cultivar Development of Ornamental Foliage Plants

211 212 215 218 227 235 239 240

245

Richard J. Henny and Jianjun Chen I. II. III. IV. V. VI.

Introduction Origin of New Cultivars Breeding Techniques Breeding Objectives Foliage Examples Future Prospects Literature Cited

246 248 255 260 267 281 283

CONTENTS

7. Preservation of Genetic Resources in the National Plant Germplasm Clonal Collections

vii

291

Gayle M. Volk and Christina Walters I. Introduction II. Clones as Genetic Resources III. Maintenance of Genetic Diversity in Clonal Collections IV. Clonal Collections in the NPGS V. Cryopreservation Principles VI. Cryopreservation: Variables to Consider VII. Application of Cryopreservation Technologies to Vegetative Materials VIII. Conclusions Literature Cited

292 294 302 308 314 319 328 332 333

Subject Index

345

Cumulative Subject Index

347

Cumulative Contributor Index

363

List of Contributors Beebe, Stephen, Centro International de Agricultura Tropical (CIAT), A. A. 6713, Cali, Colombia Chen, Jianjun, University of Florida, IFAS, Mid-Florida Research and Education Center (MREC), Apopka, FL 32703, [email protected] Frei, Ursula, Department of Agronomy and Plant Breeding, Center for Life and Food Sciences, Technical University of Munich, D-85350 FreisingWeihenstephan, Germany Henny, Richard J., University of Florida, IFAS, Mid-Florida Research and Education Center (MREC), Apopka, FL 32703, [email protected] Janick, Jules, Department of Horticulture and Landscape Architecture, Purdue University, West Lafayette, IN 47904-2010, [email protected] Jenderek, Maria M., National Arid Land Plant Genetic Resource Unit, United States Department of Agriculture—Agricultural Research Service, San Joaquin Valley Agricultural Science Center, Parlier, CA 93648 Peiretti, Edmundo G., Department of Agronomy and Plant Breeding, Center for Life and Food Sciences, Technical University of Munich, D-85350 FreisingWeihenstephan, Germany Simon, Philipp W., Vegetable Crops Research Unit, United States Department of Agriculture—Agricultural Research Service, Department of Horticulture, University of Wisconsin, Madison, WI, 53706 Snoeck, Carla, Centre of Microbial and Plant Genetics (CMPG), Department of Applied Plant Sciences, Katholieke Universiteit Leuven, Kasteelpark Arenberg 20, B-3001 Heverlee, Belgium, [email protected] Steadman, James R., Department of Plant Pathology, University of Nebraska– Lincoln, Lincoln, NE 68583-0722, [email protected] Vanderleyden, Jos, Centre of Microbial and Plant Genetics (CMPG), Department of Applied Plant Sciences, Katholieke Universiteit Leuven, Kasteelpark Arenberg 20, B-3001 Heverlee, Belgium, [email protected] Volk, Gayle M., National Center for Genetic Resources Preservation, United States Department of Agriculture—Agricultural Research Service, 1111 S. Mason Street, Fort Collins, CO 80521, [email protected] Walters, Christina, National Center for Genetic Resources Preservation, United States Department of Agriculture—Agricultural Research Service, 1111 S. Mason Street, Fort Collins, CO 80521 Wenzel, Gerhard, Department of Agronomy and Plant Breeding, Center for Life and Food Sciences, Technical University of Munich, D-85350 FreisingWeihenstephan, Germany, [email protected] Xu, Yunbi, RiceTec, Inc., P.O. Box 1305, Austin, TX 77512. Present address: Department of Plant Breeding, Cornell University, Ithaca, NY 14853-1901. [email protected] ix

Dermot P. Coyne

1 Dedication: Dermot P. Coyne Bean Breeder, Geneticist, Humanitarian James R. Steadman Department of Plant Pathology University of Nebraska–Lincoln Lincoln, Nebraska 68583-0722 Jules Janick Department of Horticulture and Landscape Architecture Purdue University West Lafayette, IN 47907-1165 Volume 23 of Plant Breeding Reviews is dedicated to the brilliant career of Dermot Patrick Coyne, who tragically passed away on April 12, 2002, following complications from treatment of hepatitis while this review was being prepared. It is some consolation for us that Dermot was aware of this dedication and, in his quiet way, expressed pleasure with this recognition. Dermot was an inspired and talented plant breeder who devoted his life to the improvement of beans for the farmers of Nebraska and the world. He was intensely concerned with the task of making progress to achieve genetic improvement of crops. Dermot was a kind and gentle man who loved life. His loss was unexpected and he will be sorely missed by family, friends, students, colleagues, and by the many farmers in the Americas and Africa whom he tried to help and the billions of poor people of the world whom he strove to assist.

Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 1

2

J. STEADMAN AND J. JANICK

BIOGRAPHICAL SKETCH Dermot was born on July 4, 1929, in Dublin, Ireland. A defining event in his life was his mother’s death when he was nine and his subsequent entry into foster care in a rural part of Ireland. A scared and grieving little boy was introduced to plants under the gentle care and guidance of Mrs. Margaret Edwards. After his father remarried when he was 11, he returned to Dublin but never forgot all that Mrs. Edwards had taught him. He took pleasure in growing vegetables to feed his family during the food shortages of World War II. He identified and spent time visiting with the neighborhood’s best gardeners, read all he could about vegetable production, and put to practice in his own garden what he learned. He always knew what it was like to be poor and to live on a farm. He held farmers in high esteem and knew that they had wisdom even when they lacked science. During 1947–1948, he completed a oneyear course in general horticulture at Johnstown Agricultural College in Wexford, and in 1948–1949 he received training at the National Botanic Gardens in Dublin. He received a Senior Certificate in Horticulture from the Royal Horticulture Society in England, and dreamed of heading a botanical garden. In 1949, he attended University College Dublin on a scholarship, and received a Bachelor in Agricultural Science in 1953 with first class honors. He captained the university field hockey team and was selected for the Irish international university team. Graduating first in his class won him the only graduate scholarship available, and he went on to receive a Masters in Agricultural Science in horticulture in 1954. In that year he obtained a research assistantship at Cornell under the tutelage of Dr. Henry M. Munger in the Department of Plant Breeding and Vegetable Crops and received the Ph.D. degree in 1958. It was there that Dermot discovered the passion for his life’s work. His first job after graduation was with the International Division of the Campbell Soup Co., in Kings Lynn, Norfolk, England, where he was assistant manager of agriculture research from 1958–1960. In late 1960, he returned to the United States, where he accepted a position at the University of Nebraska–Lincoln, in the Department of Horticulture and Forestry. He served as acting head in the Horticulture Department from 1974–1975 and in 1986 was awarded the George Holmes Regents Professorship, which he retained until his retirement in June 2001. He continued to work one-quarter time until his tragic illness. Dermot married Ann Gaffey, who was his beloved wife for nearly 45 years. Ann remains active in social change as professor in the School of Social Work at the University of Nebraska at Omaha. Together they raised six amazing children.

1. DEDICATION: D. P. COYNE

3

RESEARCH ACHIEVEMENT Bean Program Dermot made his first discovery in 1961 when he arrived in Scottsbluff, Nebraska, to observe the fields of great northern bean (Phaseolus vulgaris L.). He noticed late-maturing rogue plants with resistance to common bacterial blight in the susceptible cultivar Nebraska #1 and used one of the selections, coded as 27, to incorporate bacterial blight resistance into his breeding lines. These lines were used widely in the bean-growing regions of the United States and throughout the Americas as a source of common blight resistance. Originally he and other bean breeders believed the resistance came from the tepary bean (Phaseolus acutifolius) that had been crossed with ‘Montana #5’ (Phaseolus vulgaris) to produce Nebraska #1. In the last few years, the origin of this resistance was in doubt. To address that uncertainty, Dermot was part of a team that used molecular mapping data to demonstrate that the origin of common blight resistance was ‘Montana #5’, not the tepary bean. Dermot’s career studies into the genetics of resistance to bacterial pathogens in common bean culminated in major impacts on bean production in Nebraska as well as Africa and the Americas. At one time, the cultivars he developed occupied 60% of the bean crop area of Nebraska. Dermot was a keen observer and spent an enormous amount of time in his plots. He demonstrated the quantitative inheritance and low heritability of common blight resistance. He found that resistance was associated with late flowering, a problem in temperate regions. These revelations kept many bean breeders on the sideline for incorporating blight resistance into their breeding lines as they concentrated on more highly heritable traits. But Dermot persisted on the problem of improving resistance to blight. He made the crucial discovery that bean pods and leaves react differently to both the common blight and halo blight pathogens, which meant that selection for resistance must be made on both organs. Dermot published pioneering work on the effect of photoperiodism and temperature responses on the reaction to common blight. The late flowering of Nebraska #1 selection 27 was due to photoperiodic response under long days and was influenced by temperature. Genetic control was by major genes. Don Wallace later conducted more detailed analysis on these same photoperiod/temperature responses on bean yields. In collaborative work with scientists at CIAT (Centro Internacional Agricultura Tropical), common blight resistance sources from temperate regions such as Selection 27 were found to be susceptible in the tropics. Dermot showed that daylength affects reaction to common blight with both

4

J. STEADMAN AND J. JANICK

tropical and temperate lines and that resistance that functioned under long days did not function under short days. Dermot had a number of productive collaborations in plant pathology. Working with M. L. Schuster and later A. K. Vidaver, he began combining resistance to four bacterial diseases of bean [common blight (Xanthomonas campestris pv. phaseoli), halo blight (Pseudomonas syringae pv. phaseolicola), brown spot (Pseudomonas syringae pv. syringae), and wilt (Curtobacterium flaccumfaciens pv. flaccumfaciens)] even though the genetic mechanisms for resistance were complex. Dermot’s first releases in the late 1960s and early 1970s (Table 1.1) had high levels of common blight resistance and high yield potential, but were late. In a Plant Introduction he found resistance to three strains of the bacterial wilt pathogen and through transgressive segregation for earliness and backcrossing with Nebraska #1 selection 27 combined resistance to wilt, common blight, halo blight, and bean common mosaic to produce the great northern cultivar ‘Star’. Dermot also developed the wilt and common blight resistant ‘Emerson’, which is still used for specialty markets desiring a large, bright, white-seeded bean. In the middle phase of his career, Dermot addressed two increasing disease problems on the U.S. high plains: white mold (Sclerotinia scleTable 1.1. Release

Bean releases of Dermot Coyne and cooperators. Year

Attributesz

Dry Beans—Nebraska and Great Plains Great Northern (GN) Tara 1969 First GN with resistance to CBB, HB, BS, BCM GN Jules 1970 Resistant to CBB, HB, BS GN Emerson 1971 Moderate resistance to CBB and BW and large attractive seed GN Valley 1974 Resistant to CBB, HB, BS GN Star 1976 Resistant to CBB, BW, BCM, HB, BS GN Harris (PVP) 1980 Resistant to CBB, BCM Small White Monument 1985 Early, upright plant, resistant to BCM GN Starlight (PVP) 1990 Resistant to CBB, rust and with good avoidance of WM; bright white seed GN BelNeb-1, -2 1989 Released in cooperation with J. R. Stavely (USDA); resistant to CBB, HB, BS, rust (3 genes) and BCM GN-BELMINEB-1, -2 1993 Released in cooperation with J. R. Stavely (USDA); J. Kelly (MI); resistant to rust (2 genes) and BCM Pinto Chase 1993 First pinto with resistance to the bacterial diseases CBB, BS, HB, rust and with good avoidance of WM. Also high yielding. GN Weihing 1998 First GN bean with resistance to CBB, HB, BS, rust (2 genes), and BCM

1. DEDICATION: D. P. COYNE Table 1.1.

5

(continued)

Release

Year

Attributesz

GN BELMINEB-RMR-3

1996

GN BELMINEB-4, -5

1998

Resistant to all rust races and BCM strains in U.S. (released in cooperation with J. R. Stavely, USDA and J. Kelly, MI). Resistant to all rust races (3 genes) and BCM strains in U.S. (released in cooperation with J. R. Stavely, USDA and J. Kelly, MI).

Dry Beans—USAID-Bean/Cowpea CRSPy Arroyo Loro #1 1983 Resistant to rust in DR/PR, white seed PC-50 1987 Resistant to rust (adult plant and specific resistance); red mottled seed Anacaona 1993 Partial resistance to WB and resistance to rust; white seed Negro Sureno 1987 Developed with J. Kelly, MI; DR adapted black seed JB-178 1998 Resistant to some rust races in DR; good yielding; red mottled seed Arroyo Loro Negro 1998 Partial resistance to WB; black seed CIAS-95 1998 Resistant to some rust races in DR; red mottled seed Saladin-97 1998 Resistant to some rust races in DR; red mottled seed Winter Squash Butternut Ponca 1976 Stable fruit shape Butternut Patriot 1976 Stable fruit shape Near-Hubbard Lakota 1992 Novel decorative type Butter Bowl 1997 First novel near-oblate butternut squash for microwaving z CBB = common bacterial blight disease; BW = bacterial wilt; HB = halo blight; BS = brown spot; BCM = bean common mosaic; WB = web blight; WM = white mold y Cooperative releases: Universities of Nebraska & Puerto Rico (PR) and Secretary of Agriculture Dominican Republic (DR)

rotiorum) and rust (Uromyces appendiculatus). A Venezuelan graduate student of Dermot’s wanted to use a local black bean cultivar, ‘Tacaragua’, in his research project. When this cultivar was grown in the Scottsbluff breeding nursery, Dermot noticed that the ‘Tacaragua’ bean had low levels of white mold compared to the local pinto and great northern bean lines and cultivars. He demonstrated that hybrids with great northern and pinto lines resulted in germplasm that avoided severe white mold under conditions that favored the disease. Collaborations with J. R. Steadman and with K. G. Hubbard and a graduate student made it possible to define white mold disease avoidance microclimatology. He established that ‘Tacaragua’ contained a gene for rust resistance that is

6

J. STEADMAN AND J. JANICK

still effective on the high plains and north central U.S. bean-growing areas. As a result of a long-term, non-structured, recurrent-selection program, Dermot accomplished what few, if any, bean breeders had achieved: He combined resistance to six major diseases plus avoidance to another disease, with high yield and high seed quality in pinto and great northern cultivars and lines. He further collaborated with USDA scientists and combined resistance to four diseases and pyramided three separate rust resistance genes in elite great northern germplasm containing high seed quality. Bean seed quality was an issue that attracted Dermot’s interest throughout his career. As he had done while tackling disease resistance, Dermot performed basic genetic research on seed coat color and integrity. Although the black seeded ‘Tacaragua’ provided rust and white mold resistance, it brought undesirable seed coat colors to derived pinto lines. Recently, Dermot was able to break the linkage with seed coat yellowing to produce pinto lines with multiple disease resistance and good seed quality. The Nebraska great northern lines and cultivars that he developed (‘Starlight’, ‘Emerson’, and ‘Weihing’) were renowned for their high seed quality. As breeding and genetics moved to the molecular age, Dermot responded by upgrading his laboratory to incorporate gene mapping and molecular markers. Together with his graduate students and postdoctorals, he developed the first RAPD (random amplified polymorphic DNA) molecular map of bean that included quantitative trait loci for resistance to five bean diseases (Table 1.2). Using these maps, he confirmed the genetic relationships established earlier in his career through traditional crossing and statistical analyses. He also was able to demonstrate that adult plant rust resistance and abaxial leaf pubescence, once thought to be linked, were on different linkage groups. During the last two years in one of his final projects, he tackled a longtime goal to improve levels of common bacterial blight resistance so that seed transmission would be negligible and seed production would be feasible for Nebraska bean growers. With his last graduate student, Dermot used classical backcross breeding and SCAR (sequence characterized amplified region) markers for confirming resistance in the process of pyramiding common blight resistance from different bean sources. In 2002, lines in the field with superior common blight resistance were incorporated in agronomically elite pinto or great northern background, a realization of Dermot’s lifelong dream. Dermot Coyne relished the intellectual side of science, but his bottom line was to have a useful product for the poor and malnourished throughout the world. It was appropriate that Dermot chose to spend

Table 1.2. Mapping of quantitative trait loci and genes for disease resistance with RAPD molecular markers in recombinant inbred lines (RIL) of several bean crosses.

Pathogen

Linkage groups (LGs)

Disease

Cross for RIL

Xanthomonas campestris pv. phaseoli (bacterium) Thanatephorus cucumeris (fungus) Pseudomonas syringae pv. phaseolicola (bacterium) Uromyces appendiculatus (fungus)

Common blight

BAC6(R)* HT7719(S)*

B7, B10

Jung et al. 1996

Web blight

Xan-159(R)/PC-50(S)

B6, B8

Jung et al. 1996

Halo blight

Belneb RR-1 (R)/A55 (S)

B2, B3, B4, B5, B9, B10

Ariyarathne et al. 1999

Rust

PC-50 (R)/Xan-159 (S)

Jung et al. 1996

Sclerotinia sclerotiorum (fungus)

White mold

PC-50 (R)/Xan-159 (S)

B9—specific race resistance B2—adult plant resistance B10—abaxial leaf trichomes B4, B7, B8

*R = resistant; S = susceptible

Reference

Park et al. 2001

7

8

J. STEADMAN AND J. JANICK

most of his career working with the common bean, the most important food legume in the world from a production, consumption, nutrition, and socioeconomic standpoint. Because of its dual origins in the Andes of South America and Mexico, the common bean has been a staple in the diets of people in the Americas. Beans are a cash crop for all of the Americas. In the United States, the annual farm gate value of dry beans is $700 million, with a commercial value of $900 million for the canning and freezing sector. R. Perrin, an agricultural economist, published a unique study that estimated $5 million for the value of Dermot’s rust-resistant pinto ‘Chase’ during its five-year commercial life in the United States. Dermot’s work directly impacted bean growers in the Midwest and in the Dominican Republic and his legacy will live on through the germplasm that he created. Squash Program Dermot also had a successful squash breeding program. He produced two butternut types that provided stable fruit shape of a straight neck while retaining desirable horticultural qualities. He recently released a novel round butternut squash that could be cooked in a microwave. Always the observant breeder and history buff, Dermot grew out some ‘Hubbard’ squash seeds he obtained from descendants of pioneers in northwestern Nebraska near Fort Robinson. These seeds, traced to Native Americans nearly 200 years ago, provided soldiers with fruits whose flesh contained a rich source of Vitamin C to prevent scurvy. While the fruits no longer resembled the original descriptions, probably as a result of outcrossing, some of the fruits had decorative markings and through breeding and “reverse” selection, developed and released a one-of-a-kind cultivar with excellent cooking quality and decorative fruit surface features (orange, green, and combinations of both colors), resembling the original type. He also produced a large, elongated squash with unique color markings that is very similar to the squash grown by native Americans. This line should be of interest to the increasingly popular heritage seed industry.

CAREER Dermot’s career combined scientific curiosity and practical breeding that resulted in outstanding contributions to bean production in Nebraska and the world. A prolific writer, he authored or coauthored 160 journal articles, 8 book chapters, 350 abstracts/research notes, and 75

1. DEDICATION: D. P. COYNE

9

other publications. He was a caring nurturer to his graduate students. He believed in the generational connections in science. In his office was a picture of the famous geneticist R. A. Emerson, who happened to be a dry bean breeder at Nebraska from 1899 to 1914. Emerson moved to Cornell University and had among his graduate students Nobel Laureates George Beadle, a Nebraska native, and Barbara McClintock. Emerson’s last graduate student was H. M. Munger, who was Dermot’s mentor during his doctoral studies at Cornell. Dermot felt strongly that the research philosophy that he attempted to pass on to his students had been shaped by Munger and Emerson. Dermot took a great interest in the 43 graduate students he guided during his 42-year career. “There’s a great multiplier effect with graduate students,” he told an interviewer, “and great pleasure in seeing them develop, mature and accomplish great things in their careers.” His students came from Africa, Asia, Europe, the Middle East, and throughout the Americas. Dermot enjoyed learning something of the culture and background of each student. His students won many awards, including the Asgrow and Marion Meadow Publication Awards of the America Society for Horticultural Science (ASHS) and the Sigma Xi Distinguished Graduate Student Award. Many achieved eminence in their own right. In recognition of his dedication and success, Dermot received the ASHS Graduate Educator Award in 1998. Dermot’s commitment to graduate students was also reflected in his long-term membership in and chairing of the Horticulture Department’s Graduate Program Committee and his excellent course in Horticultural Plant Breeding. Dermot took his membership in the academic community seriously and was a quiet intervener. On one of his early committee assignments, he was asked to render a decision on an academic freedom issue resulting from on campus strikes and protests related to the Vietnam War. Although this was a difficult task, he continued to accept similar assignments dealing with academic freedom and citizenship. He was also frequently called upon for inspiration and insight into the future of a department, college, or university. Although these duties were somewhat disruptive to his research, he felt it was an important part of academic citizenship. In fact, he served in such a conscientious, courageous, and committed manner that the University of Nebraska–Lincoln Academic Senate bestowed upon him the James Lake Academic Freedom Award for “acts which support, defend, explain and apply in practice the principles of academic freedom.” Dermot’s citizenship carried over to ASHS, the professional society in which he served as president in 1985. He was president of the Nebraska Chapter of Sigma Xi, president of the Nebraska Chapter of the American

10

J. STEADMAN AND J. JANICK

Association of University Professors (AAUP), and president of the International Bean Improvement Cooperative. Dermot was a believer in the land-grant university system and told his ASHS colleagues in his ASHS Presidential Address that Our U.S. foreign aid and university administrators would do well where they can to propagate the merits of the land-grant system and education for all. The United States rose from an underdeveloped to the most developed country on Earth through this process! The European system of separate higher education, research and extension systems in agriculture, which have been adopted in many developing countries, are not satisfactory organizational models to deal with the current needs for broad-based agricultural improvements.

In an interview given after retiring from full-time to quarter-time status, Dermot noted that university research and public plant breeding face enormous change. With less federal and public funds, scientists increasingly compete for larger basic research grants. “I think we need to maintain a better balance between fundamental science and the type of (applied) research that seeks to enhance rural communities and is compatible with good management of our natural resources,” he said. He told the interviewer he expected future research success to be measured almost exclusively by grants and publications, but cautioned, “There’s much more to being a land-grant scientist than that.” Awards and Honors During his long, productive career, Dermot Coyne received many awards, including Fellow of the American Society of Agronomy, Crop Science Society of America, and ASHS. The outstanding researcher/scientist awards from ASHS, Gamma Sigma Delta, Sigma Xi, the Nebraska Legislature, and the University of Nebraska attest to the impact of his research career. As a result of over 20 years as Principal Investigator of a USAID Bean/Cowpea Collaborative Research Support Program, Dermot was presented with the End of Hunger Award by US Mayors on World Food Day. He was proud of his role in the Dominican Republic’s transformation from a net bean importer to self-sufficiency in bean production in the late 1990s. For this role, he and his colleagues received an award from the Agricultural Producers Association in the Dominican Republic for contributions to the development of agriculture and scientific collaboration contributing to bean production progress. He was the inaugural recipient of the Frazier-Zaumeyer Distinguished Lectureship Award from his bean colleagues.

1. DEDICATION: D. P. COYNE

11

Dermot Coyne was a thoughtful, caring human being with a delightful personality, a charming presence, and a wry sense of humor. His life and career remains an inspiration for plant breeders, horticulturists, pathologists, and crop scientists.

PUBLICATIONS OF DERMOT P. COYNE Journal Articles Coyne, D. P. 1962. Chemical and physiological changes in celery in relation to pithiness. Proc. Am. Soc. Hort. Sci. 8:341–346. Coyne, D. P., M. L. Schuster and S. Al-Yasiri. 1963. Reaction studies of bean species and varieties to common blight and bacterial wilt. Plant Dis. Rptr. 47:534–537. Coyne, D. P., and J. L. Serrano. 1963. Diurnal variations of soluble solids, carbohydrates and respiration rate of drought tolerant and susceptible bean species and varieties. Proc. Am. Soc. Hort. Sci. 83:453–460. Coyne, D. P. 1964. Species hybridization in Phaseolus. J. Hered. 55:5–6. Al-Yasiri, S., and D. P. Coyne. 1964. Effect of growth regulators in delaying pod abscission and embryo abortion in the interspecific cross Phaseolus vulgaris × P. acutifolius. Crop Sci. 4:433–435. Schuster, M. L., D. P. Coyne, and Kamla Singh. 1964. Population trends and movement of Corynebacterium flaccumfaciens var. aurantiacum in tolerant and susceptible beans. Plant Dis. Rptr. 48:823–827. Coyne, D. P., and R. G. Mattson. 1964. Inheritance of time of flowering and length of blooming period in Phaseolus vulgaris. Proc. Am. Soc. Hort. Sci. 85:366–373. Coyne, D. P. 1965. Component interaction in relation to heterosis for plant height in Phaseolus vulgaris variety crosses. Crop Sci. 5:17–18. Coyne, D. P., M. L. Schuster, and L. Harris. 1965. Inheritance, heritability and response to selection for common blight tolerance in Phaseolus vulgaris field bean crosses. Proc. Am. Soc. Hort. Sci. 86:373–379. Coyne, D. P., M. L. Schuster, and J. O. Young. 1965. A genetic study of bacterial wilt (Corynebacterium flaccumfaciens var. aurantiacum) tolerance in Phaseolus vulgaris crosses and the development of tolerance to two bacterial diseases in beans. Proc. Am. Soc. Hort. Sci. 87:279–285. Coyne, D. P. 1965. A genetic study of crippled morphology resembling virus symptoms in Phaseolus vulgaris L. J. Hered. 56:164. Coyne, D. P., M. L. Schuster, and L. W. Estes. 1966. Effect of maturity and environment on the genetic control of reaction to wilt bacterium in Phaseolus vulgaris L. crosses. Proc. Am. Soc. Hort. Sci. 88:393–399. Al-Yasiri, Salih Aziz, and D. P. Coyne. 1966. Interspecific hybridization in the genus Phaseolus. Crop Sci. 6:59–60. Coyne, D. P., M. L. Schuster, and Lyle Shaughnessey. 1966. Inheritance of reaction to halo blight and common blight bacteria in a Phaseolus vulgaris variety cross. Plant Dis. Rptr. 50:29–32. Coyne, D. P. 1966. A mutable gene system in Phaseolus vulgaris L. Crop Sci. 6:307–310. Coyne, D. P. 1966. The genetics of photoperiodism and the effect of temperature on the photoperiodic response for time of flowering in Phaseolus vulgaris L. varieties. Proc. Am. Soc. Hort. Sci. 89:350–360.

12

J. STEADMAN AND J. JANICK

Coyne, D. P., M. L. Schuster, and Robert Fast. 1967. Sources of tolerance and reaction of beans to races and strains of halo blight bacteria. Plant Dis. Rptr. 51:20–24. Coyne, D. P., and M. L. Schuster. 1967. A source of tolerance and reaction of tomato species and varieties to bacterial spot pathogen. Plant Dis. Rptr. 51:25–28. Coyne, D. P., and R. Mattson. 1967. Inheritance of pod maturity in a Phaseolus vulgaris L. variety cross. Crop Sci. 7:398–399. Coyne, D. P. 1967. A test to detect a mutator or unstable gene in Phaseolus vulgaris L. J. Hered. 58:146–147. Coyne, D. P. 1967. Some correlation studies in asparagus as related to cumulative season yield. Hort. Res. 7:105–112. Coyne, D. P. 1967. Photoperiodism. Inheritance and linkage studies in Phaseolus vulgaris L. J. Hered. 58:313–314. Coyne, D. P. 1968. Differential effect of soil moisture levels on style elongation in some tomato varieties. HortScience 3:39. Coyne, D. P. 1968. Correlation, heritability and selection of yield components in field beans (Phaseolus vulgaris L.). Proc. Am. Soc. Hort. Sci. 93:388–396. Coyne, D. P. 1969. Effect of growth regulators on time of flowering of a photoperiodic sensitive field bean (Phaseolus vulgaris L.). HortScience 4:100–101. Coyne, D. P. 1969. Breeding behavior and effects of temperature on expression of a variegated rogue in green beans. J. Am. Soc. Hort. Sci. 94:488–491. Coyne, D. P., and M. L. Schuster. 1969. Moderate tolerance of bean varieties to brown spot bacterium (Pseudomonas syringae). Plant Dis. Rptr. 53:677–680. Coyne, D. P. 1970. The genetic control of a photoperiod–temperature response for time of flowering in beans. Crop Sci. 10:246–248. Coyne, D. P., and M. L. Schuster. 1970. ‘Jules’, a Great Northern dry bean variety tolerant to common blight bacterium (Xanthomonas phaseoli). Plant Dis. Rptr. 54:557–559. Coyne, D. P. 1970. Inheritance of mottle-leaf in Cucurbita moschata Poir. HortScience 5:226–227. Coyne, D. P. 1970. Effect of 2-chlorethylphosphonic acid on sex expression and yield in Butternut squash and its usefulness in producing hybrid squash. HortScience 5:227–228. Arp, Gregory, D. P. Coyne, and M. L. Schuster. 1971. Disease reaction of bean varieties to Xanthomonas phaseoli and Xanthomonas phaseoli var. fuscans using two inoculation methods. Plant Dis. Rptr. 55:577–579. Coyne, D. P., M. L. Schuster, and Cesar C. Gallegos B. 1971. Inheritance and linkage of the halo blight systemic chlorosis and leaf water-soaked reaction in Phaseolus vulgaris crosses. Plant Dis. Rptr. 55:203–206. Schuster, M. L., and D. P. Coyne. 1971. New virulent strains of Xanthomonas phaseoli. Plant Dis. Rptr. 55:505–506. Coyne, D. P. 1971. A new procedure to develop hybrid Butternut squash relatively stable for fruit shape. Hort. Res. 11:183–187. Hill, K., D. P. Coyne, and M. L. Schuster. 1972. Leaf, pod, and systemic chlorosis reactions in Phaseolus vulgaris to halo blight controlled by different genes. J. Am. Soc. Hort. Sci. 97:494–498. Augustin, E., D. P. Coyne, and M. L. Schuster. 1972. Inheritance of resistance in Phaseolus vulgaris to Uromyces phaseoli typica Brazilian rust race B11 and of plant habit. J. Am. Soc. Hort. Sci. 96:526–529. Coyne, D. P., M. L. Schuster, and K. Hill. 1973. Genetic control of reaction to common blight bacterium in bean (Phaseolus vulgaris) as influenced by plant age and bacterial multiplication. J. Am. Soc. Hort. Sci. 98:94–99.

1. DEDICATION: D. P. COYNE

13

Schuster, M. L., D. P. Coyne, and B. Hoff. 1973. Comparative virulence of Xanthomonas phaseoli strains from Uganda, Colombia and Nebraska. Plant Dis. Rptr. 57:74–75. Coyne, D. P., and M. L. Schuster. 1973. Phaseolus germ plasm tolerant to Xanthomonas phaseoli bacterium. Plant Dis. Rptr. 57:111–114. Ibrahim, A. M., D. P. Coyne, R. C. Lommasson, and E. Davies. 1973. Orientation, anatomical and breeding behavior studies of the Crookneck rogue fruit in Butternut squash. J. Am. Soc. Hort. Sci. 98:576–580. Steadman, J. R., D. P. Coyne, and G. E. Cook. 1973. Reduction of severity of white mold disease on Great Northern beans by inter-row spacing and determinate plant growth habit. Plant Dis. Rptr. 57:1070–1071. Schuster, M. L., and D. P. Coyne. 1974. Survival mechanisms of phytopathogenic bacteria. Annu. Rev. Phytopath. 12:199–221. Coyne, D. P., and M. L. Schuster. 1974. Breeding and genetic studies of tolerance to several bean (Phaseolus vulgaris L.) bacterial pathogens. Euphytica 23:651–656. Coyne, D. P., and M. L. Schuster. 1974. Inheritance and linkage relations of reaction to Xanthomonas phaseoli (common blight), stage of plant development and plant habit in Phaseolus vulgaris L. Euphytica 23:195–204. Coyne, D. P., M. L. Schuster, and S. Magnuson. 1974. Differential reaction of pods and foliage of beans (Phaseolus vulgaris) to Xanthomonas phaseoli. Plant Dis. Rptr. 58:278–282. Coyne, D. P., J. R. Steadman, and F. N. Anderson. 1974. Effect of modified plant architecture on dry beans (Phaseolus vulgaris) on white mold severity and yield. Plant Dis. Rptr. 58:379–382. Coyne, D. P., and M. L. Schuster. 1974. ‘Great Northern Valley’ dry bean. HortScience 9:482. Anderson, F. M., J. R. Steadman, D. P. Coyne, and H. F. Schwartz. 1974. Tolerance to white mold in Phaseolus vulgaris dry bean edible types. Plant Dis. Rptr. 58:782–784. Schuster, M. L., and D. P. Coyne. 1975. Genetic variation in bean bacterial pathogens. Euphytica 24:143–147. Coyne, D. P., and M. L. Schuster. 1975. Genetic and breeding strategy for resistance to rust (Uromyces phaseoli (Reben) Wint) in beans (Phaseolus vulgaris). Euphytica 24:795–803. Ibrahim, A. M., and D. P. Coyne. 1975. Genetics of stigma shape, cotyledon position, and flower color in reciprocal crosses between Phaseolus vulgaris L. and Phaseolus coccineus (Lam.) and implications in breeding. J. Am. Soc. Hort. Sci. 100:622–626. Coyne, D. P., and M. L. Schuster. 1976. ‘Great Northern Star’ dry bean tolerant to bacterial diseases. HortScience 11:621. Coyne, D. P. 1976. ‘Butternut Ponca’ squash. HortScience 11:617. Coyne, D. P., and Robert M. Hill. 1976. ‘Butternut Patriot’ squash. HortScience 11:617–618. Coyne, D. P., and J. R. Steadman, and H. F. Schwartz. 1977. Reaction of Phaseolus dry bean germ plasm to Sclerotinia sclerotiorum. Plant Dis. Rptr. 61:226–230. Coyne, D. P., and J. R. Steadman. 1977. Inheritance and association of some traits in a Phaseolus vulgaris L. cross. J. Hered. 68:60–62. Coyne, D. P., J. R. Steadman, and H. F. Schwartz. 1977. Effect of genetic blends of dry beans (Phaseolus vulgaris) of different plant architecture on apothecia production of Sclerotinia sclerotiorum and white mold infection. Euphytica 27:225–231. Schwartz, H. F., J. R. Steadman, and D. P. Coyne. 1977. Influence of Phaseolus vulgaris blossoming characteristics and canopy structure upon reaction to Sclerotinia sclerotiorum. Phytopathology 68:465–470. Schuster, M. L., and D. P. Coyne. 1977. Supervivencia de patogenos bacterioles de plantas en el tropico con enfasis en frijol (Phaseolus vulgaris). Fitopath. Colombiana 6(2):101–111.

14

J. STEADMAN AND J. JANICK

Schuster, M. L., and D. P. Coyne. 1977. Characterization and variation of Xanthomonas and Corynebacterium incited diseases of beans (Phaseolus vulgaris L.). Fitopath. Brasileira 2:199–209. Schuster, M. L., and D. P. Coyne. 1977. Survival of plant parasitic bacteria of plants grown in the tropics with emphasis on beans (Phaseolus vulgaris L.). Fitopath. Brasileira 2:117–130. Coyne, D. P. 1978. Genetics of flowering in dry beans (Phaseolus vulgaris L.). J. Am. Soc. Hort. Sci. 103:606–608. Hulluka, M., M. L. Schuster, J. L. Weihing, and D. P. Coyne. 1978. Population trends of Corynebacterium flaccumfaciens strains in leaves of Phaseolus species. Fitopath. Brasileira 3:13–26. Schuster, M. L., D. P. Coyne, M. Hulluka, Lisa Brezina, and E. D. Kerr. 1978. Characterization of bean bacterial diseases and implications in control by breeding for resistance. Fitopath. Brasileira. III:149–161. Valladares-Sanchez, N. E., D. P. Coyne, and M. L. Schuster. 1979. Differential reaction of leaves and pods of Phaseolus germ plasm to strains of Xanthomonas phaseoli and transgressive segregation for tolerance from crosses of susceptible germ plasm. J. Am. Soc. Hort. Sci. 104:648–654. Coyne, D. P. 1979. Plant breeding and the public. Intl. Torch Mag. Spring p. 30–32. Schuster, M. L., D. P. Coyne, D. S. Nuland, and C. Christine Smith. 1980. Transmission of pathogenic Xanthomonas phaseoli in seeds of tolerant bean (Phaseolus vulgaris) cultivars. Plant Dis. Rptr. 63:955–959. Coyne, D. P. 1980. The role of genetics in vegetable improvement. Scientia Hort. 31:74–88. Coyne, D. P. 1980. Modification of plant architecture and crop yield by breeding. HortScience 15:244–247. Coyne, D. P. 1980. Horticulture and interdisciplinary research. HortScience 14:686. Coyne, D. P., David S. Nuland, M. L. Schuster, and F. N. Anderson. 1980. ‘Great Northern Harris’ dry bean. HortScience 15:531. Schuster, M. L., and D. P. Coyne. 1980. Biology, epidemiology, genetics and breeding for bacterial pathogens of Phaseolus vulgaris. Hort. Rev. 3:28–58. Al-Mukhtar, Faisal A., and D. P. Coyne. 1981. Inheritance and association of flower, ovule, seed, pod and maturity characters in dry edible beans (Phaseolus vulgaris L.). J. Am. Soc. Hort. Sci. 106:713–719. Korban, S. S., D. P. Coyne, and J. L. Weihing. 1981. Evaluation, variation and genetic control of seed-coat whiteness in dry beans (Phaseolus vulgaris L.). J. Am. Soc. Hort. Sci. 106:575–579. Korban, S. S., D. P. Coyne, and J. L. Weihing. 1981. Rate of water uptake and sites of water uptake in seeds of different cultivars of dry beans. HortScience 16:545–546. Korban, S. S., D. P. Coyne, J. L. Weihing, and M. A. Hanna. 1981. Testing methods, variation, morphological and genetic studies of seed-coat cracking in dry beans (Phaseolus vulgaris L.). J. Am. Soc. Hort Sci. 106:821–828. Coyne, D. P., and M. L. Schuster. 1982. Genetics and breeding for resistance to bacterial pathogens in vegetable crops. HortScience 18:30–36. Coyne, D. P., S. S. Korban, D. Knudsen, and R. B. Clark. 1982. Inheritance of iron deficiency in crosses of dry beans (Phaseolus vulgaris L.). Proc. Intl. Symp. Iron Nutrition and Interactions in Plants 5(4–7):575–585. Adeniji, A. A., and D. P. Coyne. 1982. Inheritance of resistance to trifluralin toxicity in Cucurbita moschata Poir. HortScience 16:774–775. Leyna, H. K., S. Korban, and D. P. Coyne. 1982. Changes in patterns of inheritance of flowering time of dry beans (Phaseolus vulgaris L.) in different environments. J. Hered. 73:306–308.

1. DEDICATION: D. P. COYNE

15

Valladares, N. E. Sanchez, D. P. Coyne, and R. F. Mumm. 1983. Inheritance and associations of leaf, external, and internal pod reactions to common blight bacterium in Phaseolus vulgaris L. J. Am. Soc. Hort. Sci. 108:272–278. Adeniji, A. A., and D. P. Coyne. 1983. Genetics and nature of resistance to powdery mildew in crosses of Butternut with Calabaza squash and ‘Seminole Pumpkin’. J. Am. Soc. Hort. Sci. 108(3):360–368. Schuster, M. L., D. P. Coyne, T. Behre, and H. Leyna. 1983. Sources of Phaseolus species resistance and leaf and pod differential reaction to common blight. HortScience 18:901–903. Zaiter, H. Z., and D. P. Coyne. 1984. Testing inoculation methods and sources of resistance to the halo blight bacteria (Pseudomonas syringae pv. phaseolicola) in Phaseolus vulgaris. Euphytica 33:133–141. Fuller, P. A., D. P. Coyne, and J. R. Steadman. 1984. Inter and intra-row inter-genotypic competition influences selection for avoidance to white mold disease in dry edible beans. J. Am. Soc. Hort. Sci. 109:567–572. Fuller, P. A., J. R. Steadman, and D. P. Coyne. 1984. Enhancement of white mold avoidance and yield in dry beans by canopy elevation. HortScience 19:78–79. Fuller, P. A., D. P. Coyne, and J. R. Steadman. 1984. Inheritance of resistance to white mold disease in a diallel cross of dry beans (Phaseolus vulgaris) in the field and greenhouse. Crop Sci. 24:929–933. Pierson, E. E., R. B. Clark, J. W. Maranville, and D. P. Coyne. 1984. Plant genotype differences to ferrous and total iron in emerging leaves. I. Sorghum and maize. J. Plant Nutr. 7:371–387. Pierson, E. E., R. B. Clark, D. P. Coyne, and J. W. Maranville. 1984. Plant genotype differences to ferrous and total iron in emerging leaves. II. Dry beans and soybeans. J. Plant Nutr. 7:355–369. Coyne, D. P. 1985. Tackling world hunger and malnutrition through horticultural research, graduate education, extension and management in cooperation with U.S. universities. HortScience 20:805–808. Beaver, J., C. Paniagua, D. P. Coyne, and G. Freytag. 1985. Yield stability of dry bean genotypes in the Dominican Republic. Crop Sci. 25:923–926. Zaiter, Haytham, D. P. Coyne, R. Clark, and David Nuland. 1986. Field, nutrient solution and temperature effect upon iron leaf chlorosis of dry beans (Phaseolus vulgaris). J. Plant Nutr. 9:397–415. Coyne, D. P., J. R. Steadman, D. S. Nuland, and C. L. Campbell. 1986. ‘Monument’ small white dry bean. HortScience 21:542. Pierson, E. E., R. B. Clark, D. P. Coyne, and J. W. Maranville. 1986. Iron deficiency stress effects on total iron in various leaves and nutrient solution pH in sorghum and beans. J. Plant Nutr. 9:893–907. Finke, M. Luann, D. P. Coyne, and J. R. Steadman. 1986. The inheritance and association of resistance to rust, common bacterial blight, plant habit and foliar abnormalities in Phaseolus vulgaris L. Euphytica 35:969–982. Zaiter, H. Z., D. P. Coyne, and R. B. Clark. 1987. Genetic variation and inheritance of resistance of leaf iron-deficiency chlorosis in dry beans. J. Am. Soc. Hort. Sci. 112:1019– 1022. Zaiter, H. Z., D. P. Coyne, and R. B. Clark. 1987. Temperature, grafting method, and rootstock influence on iron-deficiency chlorosis of bean (Phaseolus vulgaris L.). J. Am. Soc. Hort. Sci. 112:1023–1026. Zaiter, H. Z., D. P. Coyne, and R. B. Clark. 1988. Genetic variation, heritability, and selection response to iron-deficiency chlorosis in dry beans (Phaseolus vulgaris L.). J. Plant Nutr. 11:739–746.

16

J. STEADMAN AND J. JANICK

Poe, R. R., D. P. Coyne, B. A. Swisher, and M. D. Clegg. 1988. Differential Cucurbita spp. tolerance to the herbicide trifluralin. J. Am. Soc. Hort. Sci. 113:31–35. Jimenez, J., D. P. Coyne, and F. Saladin. 1989. Imbibition, germination and cooking time of seeds of dry beans (Phaseolus vulgaris L.) stored in different containers. J. Agr. Univ. Puerto Rico 73:327–338. Jimenez, J., D. P. Coyne, and F. Anderson. 1989. Imbibition of seed of dry beans (Phaseolus vulgaris L.) stored under high and low temperature and relative humidity conditions. Scientia Hort. 40:91–98. Ahmed, A., and D. P. Coyne. 1989. Heritability, phenotypic correlations and associations of the common blight disease reactions in beans. J. Am. Soc. Hort. Sci. 114:828–833. Zaiter, H. Z., D. P. Coyne, A. K. Vidaver, and J. R. Steadman. 1989. Differential reaction of tepary (Phaseolus acutifolius Gray) bean lines to Xanthomonas campestris pv. phaseoli. HortScience 24:134–137. Stavely, J. R., J. R. Steadman, D. P. Coyne, and D. T. Lindgren. 1989. Belneb rust resistant –1 and –2 Great Northern dry bean germplasm. HortScience 24:400–401. Ahmed, A., D. P. Coyne, A. K. Vidaver, and K. M. Eskridge. 1989. Transmission of the common blight pathogen in bean seed. J. Am. Soc. Hort. Sci. 114:1002–1008. A. Aggour, D. P. Coyne, and A. K. Vidaver. 1989. Comparison of leaf and pod disease reactions of beans (Phaseolus vulgaris L.) inoculated by different methods with strains of Xanthomonas campestris pv. phaseoli (Smith) Dye. Euphytica 43:143–152. Coyne, D. P. 1989. Research and extension’s role in the future of dry bean production. HortScience 24:542–546. Clark, R. B., D. P. Coyne, W. M. Ross, and B. E. Johnson. 1990. Genetic aspects of plant resistance to iron deficiency. Intern. Plant Physiol. Congr. 2:1096–1115. Zaiter, H. Z., D. P. Coyne, and J. R. Steadman. 1990. Rust reaction and pubescence in Alubia beans. HortScience 25:664–665. Zaiter, H. Z., D. P. Coyne, and J. R. Steadman. 1990. Inheritance of abaxial leaf pubescence in beans. J. Am. Soc. Hort. Sci. 115:158–160. Zaiter, H. Z., D. P. Coyne, and J. R. Steadman. 1990. Coinoculation effects of the pathogens causing common bacterial blight, rust and bean common mosaic in Phaseolus vulgaris. J. Am. Soc. Hort. Sci. 115:319–323. Zaiter, H. Z., D. P. Coyne, R. B. Clark, and J. R. Steadman. 1991. Medium, pH and leaf nutrient concentration influence on rust pustule diameter on leaves of dry beans. HortScience 26:412–414. Mohamed, M. F., P. E. Read, and D. P. Coyne. 1991. In vitro response of bean (Phaseolus vulgaris L.) cotyledonary explants to benzyladenine in the medium. Quart. Plant Growth Regulator Soc. Am. 19:19–26. Arnaud-Santana, E., E. Pena-Matos, D. P. Coyne, and A. K. Vidaver. 1991. Longevity of Xanthomonas campestris pv. phaseoli in naturally infested dry bean (Phaseolus vulgaris) debris. Plant Dis. 75:952–953. Coyne, D. P., J. R. Steadman, D. T. Lindgren, and D. S. Nuland. 1991. ‘Starlight’ Great Northern dry bean. HortScience 26:441–442. Beaver, J. S., J. R. Steadman, and D. P. Coyne. 1992. Field reaction of landrace components of red mottled beans to common bacterial blight. HortScience 27:50–51. Zaiter, H. Z., D. P. Coyne, R. B. Clark, D. T. Lindgren, D. T. Nordquist, W. W. Stroup, and L. A. Pavlish. 1992. Leaf chlorosis and seed yield of dry beans grown on high-pH calcareous soil following foliar iron sprays. HortScience 27:983–985. Mmbaga, M. T., E. Arnaud-Santana, J. R. Steadman, and D. P. Coyne. 1992. New sources of nonspecific resistance to rust and common bacterial blight in the dry bean landrace Pompadour. Euphytica 61:135–144.

1. DEDICATION: D. P. COYNE

17

Mohamed, M. F., P. E. Read, and D. P. Coyne. 1992. Plant regeneration from in vitro culture of embryonic axis explants in common and tepary beans. J. Am. Soc. Hort. Sci. 117:332–336. Mohamed, M. F., P. E. Read, and D. P. Coyne. 1992. Dark preconditioning, CPPU, and thidiazuron promote shoot organogenesis on seedling node explants of common and faba beans. J. Am. Soc. Hort. Sci. 117:668–672. Beaver, J. S., J. R. Steadman, and D. P. Coyne. 1992. Field reaction of landrace components of red mottled beans to common bacterial blight. HortScience 27:50–51. Arnaud-Santana, E., D. P. Coyne, J. S. Beaver, and H. Z. Zaiter. 1993. Effect of photoperiod and temperature on common blight disease of common beans (Phaseolus vulgaris L.). Euphytica 66:211–216. Arnaud-Santana, E., M. T. Mmbaga, D. P. Coyne, and J. R. Steadman. 1993. Sources of resistance to common bacterial blight and rust in elite Phaseolus vulgaris L. germplasm. HortScience 28:644–646. Mohamed, M. F., E. Arnaud-Santana, and D. P. Coyne. 1993. Rooting of bean leaves and use in germplasm evaluation for common bacterial blight resistance. Euphytica 65:161–166. Mohamed, M. F., D. P. Coyne, and P. E. Read. 1993. Shoot organogenesis in callus induced from pedicel explants of common bean (Phaseolus vulgaris L.). J. Am. Soc. Hort. Sci. 118:158–162. Wallace, D. H., J. P. Bandoin, J. Beaver, D. P. Coyne, D. E. Halseth, P. N. Masaya, H. M. Munger, J. R. Myers, M. Silbernagel, K. S. Yourstone, and R. W. Zobel. 1993. Improving efficiency of breeding for higher crop yield. Theor. Appl. Genet. 86:27–40. Arnaud-Santana, E., D. P. Coyne, K. M. Eskridge, and A. K. Vidaver. 1994. Inheritance, low correlations of leaf, pod, and seed reactions to common blight disease in common beans, and implications for selection. J. Am. Soc. Hort. Sci. 119:116–121. Coyne, D. P., D. S. Nuland, D. T. Lindgren, and J. R. Steadman. 1994. ‘Chase’ Pinto dry bean. HortScience 29:44–45. Deshpande, R. Y., K. G. Hubbard, D. P. Coyne, J. R. Steadman, and A. M. Parkhurst. 1995. Estimating leaf wetness in dry bean canopies as a prerequisite to evaluating white mold disease. Agron. J. 87:613–619. Mohamed, M. F., and D. P. Coyne. 1995. Photoperiod sometimes influences common bacterial blight disease of common beans. HortScience 30:551–553. Mohamed, M. F., D. P. Coyne, and P. E. Read. 1995. A radiation-induced mutant with resistance to common bacterial blights in common beans. HortScience 30:577–578. Lindgren, D. T., and D. P. Coyne. 1995. Injury and yield of leaf hopper infested dry beans. J. Am. Soc. Hort. Sci. 120:839–842. Coyne, D. P. 1995. Classical and molecular approaches to breeding horticultural plants for disease resistance: Introduction to the Colloquium. HortScience 30:448–449. Coyne, D. P., J. M. Reiser, and L. Sutton. 1995. ‘Lakota’ winter squash, a cultivar derived from Native American sources in Nebraska. HortScience 30:1106–1107. Arnaud-Santana, E., and D. P. Coyne. 1996. Herencia y relacion de la reaccion a la bacteriosis comun y dias a floracion en habichuelas (Phaseolus vulgaris L.). J. Agr. Univ. Puerto Rico 80:95–109. Beaver, J. S., E. Arnaud Santana, and D. P. Coyne. 1996. Yield stability of determinate and indeterminate red mottled beans. J. Agr. Univ. Puerto Rico 80:187–189. Eskridge, K. M., and D. P. Coyne. 1996. Estimating the number of genes affecting a trait using inbred-backcross data and modified minimum chi-square. J. Hered. 87:410–412. Mohamed, F. M., D. P. Coyne, and P. E. Read. 1996. Enhancement effect of CPPU on differentiation of somatic embryoids in callus cultures of common bean (Phaseolus vulgaris L.). Quart. Plant Growth Regulator Soc. Am. 24:97–103.

18

J. STEADMAN AND J. JANICK

Jung, G., D. P. Coyne, P. W. Skroch, J. Nienhuis, E. Arnaud-Santana, J. Bokosi, H. M. Ariyarathne, J. R. Steadman, J. Beaver, and S. Kaeppler. 1996. Molecular markers associated with plant architecture and resistance to common blight, web blight, and rust in common beans. J. Am. Soc. Hort. Sci. 121:794–803. Zhang, Z., D. P. Coyne, and A. Mitra. 1997. Factors affecting Agrobacterium-mediated transformation of common bean. J. Am. Soc. Hort. Sci. 122:300–305. Jung, G., P. W. Skroch, D. P. Coyne, J. Nienhuis, E. Arnaud-Santana, H. M. Ariyarathne, S. M. Kaeppler, and M. J. Bassett. 1997. Molecular marker-based genetic analysis of Tepary bean-derived common bacterial blight resistance in different developmental stages of common bean. J. Am. Soc. Hort. Sci. 122:329–337. Park, S. O., D. P. Coyne, A. Dursun, and G. Jung. 1998. Identifying randomly amplified polymorphic DNA (RAPD) markers linked to major genes for common bacterial blight resistance in tepary bean. J. Am. Soc. Hort. Sci. 123:278–282. Jung, G., D. P. Coyne, J. Bokosi, J. R. Steadman, and J. Nienhuis. 1998. Mapping genes from specific and adult plant resistance to rust and abaxial leaf pubescence and their genetic relationships using randomly amplified polymorphic DNA (RAPD) markers in common bean. J. Am. Soc. Hort. Sci. 123:859–863. Ariyarathne, H. M., D. P. Coyne, A. K. Vidaver, and K. M. Eskridge. 1998. Selecting for common bacterial blight resistance in common bean: Effects of multiple leaf inoculation and detached pod inoculation test. J. Am. Soc. Hort. Sci. 123:864–867. Zhang, Z., D. P. Coyne, A. K. Vidaver, and A. Mitra. 1998. Expression of human lactoferrin cDNA confers resistance to Ralstonia solanacearun in transgenic tobacco plants. Phytopathology 88:730–734. Park, S. O., D. P. Coyne, J. M. Bokosi, and J. R. Steadman. 1999. Molecular markers linked to genes for specific rust resistance and indeterminate growth habit in common bean. Euphytica 105:133–141. Sandlin, C. M., J. R. Steadman, C. M. Araya, and D. P. Coyne. 1999. Isolates of Uromyces appendiculatus with specific virulence to landraces of Phaseolus vulgaris of Andean origin. Plant Dis. 83:108–113. Jung, G., P. W. Skroch, D. P. Coyne, E. Arnaud Santana, and H. M. Ariyarathne. 1999. Confirmation of QTL associated with common bacterial blight resistance in four different genetic backgrounds in common bean. Crop Sci. 39:1448–1455. Ariyarathne, H. M., D. P. Coyne, G. Jung, P. N. Miklas, M. J. Bassett, and P. Skroch. 1999. Molecular mapping of disease resistance genes for halo blight, common bacterial blight, and bean common mosaic virus in a segregating population of common bean. J. Am. Soc. Hort. Sci. 124:654–662. Muharrem, E., E. T. Paparozzi, D. P. Coyne, D. Smith, S. Kachman, and D. S. Nuland. 2001. Testing the effects of moisture on seedcoat color of pinto dry beans. HortScience 36:302–304. Fall, A. L., P. F. Byrne, G. Jung, D. P. Coyne, M. A. Brick, and H. F. Schwartz. 2001. Detection and mapping locus for fusarium wilt resistance in common bean. Crop Sci. 41:1494–1498. Park, Soon O., Dermot P. Coyne, James R. Steadman, and Paul W. Skroch. 2001. Mapping of Qtl for resistance to white mold disease in common bean. Crop Sci. 41:1253–1262. Sutton, L. A., and D. P. Coyne. 2002. Bean-dry. Cultivar list. HortScience 37:16–18. Coyne, D. P., J. R. Steadman, G. Godoy-Lutz, R. Gilbertson, E. Arnaud Santana, J. S. Beaver, and J. R. Myers. Contributions of the Bean/Cowpea CRSP to Management of Bean Diseases. Field Crops Res. (In press).

1. DEDICATION: D. P. COYNE

19

Book Chapters Coyne, D. P. 1977. Dry edible beans. p. 70–71. In: Agricultural atlas of Nebraska. Univ. Nebraska Press, Lincoln. Coyne, D. P., and R. B. O’Keefe. 1977. Horticultural crops. p. 73–74. In: Agricultural atlas of Nebraska. Univ. of Nebraska Press, Lincoln. Coyne, D. P., and M. L. Schuster. 1979. Bacterial diseases of legumes: Breeding and resistance. p. 225–233. In: Advances in legume science. HMSO, London. Adams, W., D. P. Coyne, J. H. C. Davis, C. A. Francis, and P. H. Graham. 1985. The common bean. In: R. Summerfield (ed.), Grain legumes. Longman’s, UK. Coyne, D. P. 1989. Introductory essay on J. C. Walker and Rose Bloom’s classical paper “Effect of environmental factors upon the resistance of cabbage to yellows.” J. Agr. Res. 41:1–15 (1930). p. 526–528. In: J. Janick (ed.), Classical papers in horticultural science. Prentice Hall, Englewood Cliffs, NJ. Coyne, D. P. 1994. Tackling world hunger and malnutrition through horticultural research, graduate education, extension and management in cooperation with U.S. universities. p. 383–394. In: J. Janick (ed.), Presidential addresses. Am. Soc. Hort. Sci., Alexandria, VA. Wallace, D. H., K. S. Yourstone, J. P. Baudoin, J. Beaver, D. P. Coyne, J. W. White, and R. W. Zobel. 1995. Photoperiod × temperature interaction effects on the days of flowering of bean (Phaseolus vulgaris L.) p. 863–891. In: M. Pessarakli (ed.), Handbook of plant and crop physiology. Marcel Dekker, New York. Jahn, M. J., and D. P. Coyne. 1998. Dedication: Henry M. Munger. Cucurbitaceae. Am. Soc. Hort. Sci., Alexandria, VA.

2 Strategies for Genetic Improvement of Common Bean and Rhizobia Towards Efficient Interactions* Carla Snoeck and Jos Vanderleyden Centre of Microbial and Plant Genetics, Department of Applied Plant Sciences, Katholieke Universiteit Leuven Kasteelpark Arenberg 20, B-3001 Heverlee, Belgium Stephen Beebe Centro Internacional de Agricultura Tropical A.A. 6713, Cali, Colombia

I. RHIZOBIUM-COMMON BEAN SYMBIOSIS II. BEAN BREEDING A. Breeding Strategies B. Phosphorus Efficiency C. Enhanced Nitrogen Fixation III. SELECTION OF OPTIMIZED RHIZOBIUM STRAINS FOR BEAN INOCULATION A. Strains with Enhanced Nitrogen Fixation Capacity B. Nodulation Competitiveness and Persistence in the Soil C. Common Bean Rhizobia Adapted to Different Environmental Factors IV. CONCLUSIONS AND FUTURE PROSPECTS LITERATURE CITED * The authors thank E. Luyten, J. Michiels, and R. De Mot at CMPG for useful suggestions during the preparation of the manuscript and P. Gepts, E. Martínez-Romero, R. Merckx, and W. Broughton for critically reading the manuscript. Special gratitude goes to P. Gepts for providing the bean linkage map. C. Snoeck acknowledges the receipt of a predoctoral fellowship from the “Vlaams Instituut voor de Bevordering van het Wetenschappelijk Technologisch Onderzoek in de Industrie” and the receipt of a postdoctoral fellowship from the K.U. Leuven Research Council. The authors acknowledge financial support from a VLIR-DGOS grant from the Belgian government and a VIS grant from the K.U. Leuven Research Council. Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 21

22

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

I. RHIZOBIUM-COMMON BEAN SYMBIOSIS The wild ancestor of the cultivated common bean (Phaseolus vulgaris L., Fabaceae) is a climbing herbaceous annual that grows naturally at mid to high altitudes from Northern Mexico to Northwest Argentina (Gepts and Debouck 1991). Archeological findings indicate that domestication took place in both Mesoamerica and the Andean zone at least 4000 years ago (Kaplan and Lynch 1999). The pre-Colombian tribes domesticated four other Phaseolus species covering distinct ecological niches (Table 2.1): P. lunatus L., P. coccineus L., P. acutifolius A. Gray, and P. polyanthus Greenman (Toro et al. 1990). All of these species are of 2n=22. Nevertheless, common bean gained wider acceptance and was selected more intensively. Four centers of diversity can be distinguished for wild populations of P. vulgaris: the Mesoamerican, the Andean, the Colombian, and Ecuador-Northern Peru (Fig. 2.1) (Tohme et al. 1996). Geographically isolated domestications in the Andean and Mesoamerican gene pools have led to at least two major gene pools of domesticated bean and one minor, incipient gene pool (Gepts et al. 1986; Singh et al. 1991a; Islam et al. 2002). In addition, there are important secondary centers of diversity in Brazil, North America, Europe, the Middle East, and Africa. The world collection of cultivated and wild P. vulgaris is held in the germplasm bank of The International Center of Tropical Agriculture (CIAT) in Cali, Colombia. The large biodiversity of wild Phaseolus species is well recognized. Moreover, the available wild Phaseolus species give access to an excellent source for traits such as resistance/tolerance to biotic and abiotic stress, yield under agronomically relevant conditions, and symbiotic Table 2.1. Ecological and growth characteristics of selected cultivated species of Phaseolus in tropical areas.z

Phaseolus

Altitude (m)

Daytime temperature (°C)

Mean annual precipitation (mm/year)

Growth cycley (days)

Yield potentialx (kg/ha)

coccineus acutifolius lunatus polyanthus vulgaris

1400–2800 50–1900 50–2800 800–2600 50–3000

12–22 20–32 16–26 14–24 14–26

400–2600 200–400 0–2800 1000–2600 400–1600

90–365 60–110 90–365 110–365 70–330

400–4000 400–2000 400–5000 300–3500 400–5000

z

D. G. Debouck (http://www.hort.purdue.edu/newcrop/1492/beans.html) Duration of growth cycle from start to end of harvest x Yield potential in tropical areas y

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

23

A

Colombian Mesoamerican Andean Ecuador Northern Peru

B Guatemala Durango

Jalisco

Mesoamerica

Nueva Granada

?

Peru

Chile

?

?

Wild Middle American

Wild Andean

Wild Ecuador Northern Peru

Fig. 2.1. A. Distribution of wild P. vulgaris L. in Latin America. B. Evolution of P. vulgaris L.: Domestication of the Middle American and the Andean gene pools led to four races in the Middle American gene pool and three races in the Andean pool. Introgression between genotypes of races from different gene pools is shown. Modified from Gepts (1998b) and Beebe et al. (2000).

24

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

interactions with rhizobia and mycorrhizae that occurred because of coevolution under different selection pressures. Wild bean germplasm is also useful as a source of geographic markers in evolutionary studies (Gepts 1998a), allowing a more systematic search for a trait of interest in the cultivated germplasm (Kipe-Nolt et al. 1992). To date, this potential of wild ancestral Phaseolus germplasm has only been exploited for common bean (Debouck 1994). Common bean is the most important food legume for direct human consumption worldwide and especially in Latin America and Africa (Table 2.2). Beans, like other legumes, are superior to cereals as sources of proteins and micronutrients. Total worldwide production of common bean is difficult to quantify (estimates range from 12 to 19 million tonnes). According to the Food and Agricultural Organization, more than 50% of bean is produced in developing countries (FAO 2001). Beans are considered as a relatively profitable crop by local farmers in Latin America (CIAT 2000), and smallholder bean production, on farms ranging from 1 to 10 ha in size, is expanding rapidly. Nevertheless, yield of common bean on these farms is low (600–900 kg/ha) because cultivation is done on sloping land of low fertility. Under fertile conditions, a yield potential of 2500–5000 kg/ha can be obtained (Table 2.1). However, use of chemical fertilizer is difficult to implement because of economic constraints, while inoculation with N2-fixing rhizobia (exploiting the potential of symbiotic N2 fixation of legumes) is presently limited because of the low N2-fixing capacity of the commercial bean genotypes currently in use, combined with unfavorable edaphic constraints. Interest in common bean genetics is increasing with the recent identification of new sources of germplasm and the development of gene mapping techniques (Nodari et al. 1993). The challenge of increasing

Table 2.2.

Bean production in Latin America (CIAT 2000). Area (1000 ha)

Production (1000 tonnes)

Brazil Mexico Central America (Guatemala, Honduras, El Salvador, Nicaragua, Costa Rica, Panama) Southern cone (Chile, Argentina, Paraguay) Andean zone (Venezuela, Colombia, Ecuador, Peru, Bolivia) Caribbean (Cuba, Haiti, Dominican Republic)

5 092 2 259

3 055 1 300

526 357

337 398

299 157

265 141

TOTAL

8 690

5 496

Country/Region

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

25

bean yield through selection for disease resistance, improved nodulation following infection by efficient Rhizobium strains with consequently higher N2 fixation levels, and more efficient use of nutrients is high but technically feasible to tackle. Moreover, the selection/construction of superior Rhizobium strains can help to promote the development of low-input cropping systems. Promiscuity of bean as a host plant for rhizobia became obvious when native isolates recovered from bean nodules from different geographical regions (although with emphasis on Mesoamerica) were shown to have considerable genetic diversity (Martínez et al. 1985; Piñero et al. 1988; Laguerre et al. 1993; Eardly et al. 1995; Aguilar et al. 1998; VasquezArroyo et al. 1998; Herrera-Cervera et al. 1999; Rodriguez-Navarro et al. 2000). P. vulgaris is highly promiscuous for both fast-growing and slowgrowing rhizobia, but efficient (N2 fixing) symbiosis is only obtained with fast-growing rhizobia (Michiels et al. 1998a; Bala and Giller 2001). Among the fast-growing rhizobia, R. etli type I strains (formerly R. leguminosarum bv. phaseoli) (Segovía et al. 1993; van Berkum et al. 1996) were found to be predominantly present in the nodules of aboriginal wild bean varieties both in Mesoamerica and the Andean region. This led to the speculation that P. vulgaris co-evolved in the symbiosis with compatible Rhizobium spp. (Aguilar et al. 1998; Bernal and Graham 2001). R. tropici type II A and B strains (Martínez-Romero et al. 1991) are particularly associated with bean plants cultivated in South American acid soils. R. tropici, such as CIAT899, seemed a promising source for inoculant strains because of its tolerance to high temperatures and its higher genetic stability—being less subject to genomic rearrangements and plasmid loss—under stress conditions (Martínez-Romero et al. 1991; Michiels et al. 1994a; Soberón-Chavez et al. 1986). Other strains effectively nodulating beans are R. leguminosarum bv. phaseoli (Jordan 1984), Rhizobium sp. NGR234 (van Rhijn et al. 1994a), the recently described R. gallicum bv. phaseoli and R. giardinii bv. phaseoli (Amarger et al. 1997), and Sinorhizobium sp. BR816, retained from a Brazilian selection program for tolerance to high temperature (Hungria et al. 1993; Hernández-Lucas et al. 1995b). Former use of inoculant strains for application in Brazilian acid soils with low N content (Cerrados region) has been a discouraging experience, since these strains, none of them phylogenetically classified as R. tropici, had lost their ability to fix N2 in the field. Today, the demand of Brazilian farmers for bean inoculants is increasing. Bean yield is generally increased by inoculation with R. tropici CIAT899 to up to 2500 kg/ha, five times higher than the average Brazilian yield. Since 1998, a new isolate PRF 81 with R. tropici characteristics has been officially recommended for use in Brazilian commercial inoculants together with

26

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

CIAT899 (Hungria and Vargas 2000). PRF 81 is tolerant to high temperatures and acidity in vitro and in vivo, effective in N2 fixation with common bean, and highly competitive against indigenous bean rhizobia (Hungria et al. 2000). In this review, innovative research strategies and current trends to improve bean yield are highlighted. It emphasizes the need of coordinated screening for genes involved in disease resistance, abiotic stress and yield potential, through exploitation of the large genetic potential of available bean germplasm. In this respect, N2 fixation (selection of plant host, preferential nodulation by the host) and efficient P uptake (plant architecture, carbon/nitrogen metabolism, root activity) are indirect traits of high priority to select for higher bean yield and are discussed here. This study also deals with the selection and subsequent inoculation with superior N2-fixing rhizobia that are optimally interacting with bean cultivars, hereby exploiting the potential of symbiotic N2 fixation. An overview of the bacterial determinants involved in the Rhizobium-Phaseolus vulgaris interaction is given. Recent developments in the selection of optimized bean nodulating Rhizobium strains (enhanced N2 fixation capacity, nodulation competitiveness, abiotic stress resistance) through advances in molecular genetics are discussed. For further reading, we refer to a recent book covering various state-ofthe-art aspects and future directions of common bean improvement (Singh 1999a) as well as to a review on inoculant Rhizobium strains (Vlassak and Vanderleyden 1997).

II. BEAN BREEDING Although common bean is a staple food crop in Latin America and the international bean market is still quite small, the onset of globalization will certainly promote bean trade. Not only countries like China and Myanmar, but also the United States, Argentina, and Canada, are major grain legume exporters (FAO 2001). Since common bean consumption in most Latin American countries exceeds local production, there is increasing activity in the international market. Although Latin America has its own typical bean cultivars and undoubtedly cultural aspects are very important in this respect, yield will have to increase to compete on a world scale (CIAT 2000). Clear goals for yield potential must be seen in the context of a given region, production system, and grain type. Associated bean cropping is the most common traditional system in both Latin America (often maizebean association) and Africa (often bean-banana and bean-root crops association) and is focused on low-input farming. Environmental factors

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

27

are frequently far from optimal and soils can be critically limited by abiotic stresses, such as drought, toxicities, and nutrient deficiencies. Thus, breeding for improved productivity within these farming systems is best addressed through a combination of genetic and phytotechnical management solutions that are accessible to poor farmers. Moreover, increased productivity may be emerging from work on edaphic resistance. Lack of money to purchase pesticides and the dangers of pesticide toxicity also make breeding for disease resistance a desirable goal. Improvement of the nutritional quality of beans is another important aspect that would be beneficial for all bean-consuming populations but the subject will not be covered here. A. Breeding Strategies Wild germplasm, weedy forms, and landraces of common bean have been (and still are) the major source of genes for disease resistance, abiotic stress tolerance, and yield potential (Beebe et al. 1997). Recombination between the two major gene pools offers opportunities for bean breeders to broaden the genetic diversity of the cultivated races provided that viable and fertile progeny can be obtained and that appropriate breeding methodology is adopted to capture the complex gene interactions underlying performance (Gepts 1998a; Johnson and Gepts 2002). Most traits are still selected by conventional means in field sites where diseases and edaphic constraints can occur. However, the advent of biotechnology has provided new tools for optimally making use of genetic resources, encouraging integrated and coordinated screening programs. By analyzing patterns of molecular markers such as restriction fragment length polymorphism (RFLP), random amplification of polymorphic DNA (RAPD), sequence characterized amplified regions (SCAR), amplified fragment length polymorphism (AFLP), or single sequence repeats (SSR), genetic diversity in germplasm can be evaluated and genetic control of traits elucidated. These polymorphic markers are assembled to construct linkage groups covering the genome. RFLP and SSR have the advantages of being co-dominant and identifying specific sites in the genome and are preferred for developing maps among contrasting parental genotypes that in turn can be compared with the reference map (Freyre et al. 1998). If map saturation with RFLP and SSR is inadequate, AFLP and RAPD can be generated at relatively low cost and can serve to saturate a framework map, especially if the map is based on Recombinant Inbred Lines (RIL), where dominant expression of these markers is not an issue. Wide crosses among gene pools offer a high degree of polymorphism with all types of markers but suffer possible segregation distortion,

28

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

which complicates map development. Furthermore, wide crosses might present certain physiological dysfunctions that make phenotypic analysis doubtful. This is an important consideration for analysis of complex physiological traits. Narrow crosses (for example, within bean races) will likely perform more normally both genetically and physiologically, but close genetic relationship has the disadvantage of low genetic polymorphism, which makes mapping more difficult. Furthermore, these techniques should prove useful for selecting particular traits between different sources, and hopefully, loci affecting quantitatively inherited traits (QTL) can be detected and monitored during introgression. The parameters of QTL detection (necessary population size, desired degree of map saturation) depend on the precision that the researcher seeks for his particular purposes. Precise localization of QTL within one or two mapping units would require a population of several hundred individuals and a highly saturated map. But for practical breeding purposes of identifying markers that can serve to recover important regions of the genome in a breeding program, flanking markers at a distance of 5–10 cM might be adequate. This degree of precision requires rather more modest population sizes, and much can be learned with populations of 80–100 individuals. Furthermore, if QTL are sought for physiological traits, the capacity for detailed and reliable physiological analysis on large populations may be even more limiting than the genetic analysis. For more detailed information, other strategies such as functional genomics that focus on genes of known sequence and function might be a more effective avenue than QTL analysis (for example, to tease out knowledge of multigene families). Another option to obtain molecular markers is to sequence messenger RNAs extracted from tissues of interest. These so-called expressed sequence tags (ESTs) represent sequences of genes that are very useful for positional cloning. High-throughput molecular mapping making use of these ESTs as genetic landmarks will speed up modern plant breeding programs. Markers have been used to position agronomically important genes on the linkage map that several groups have developed for common bean (Vallejos et al. 1992; Nodari et al. 1993; Adam-Blondon et al. 1994; Beebe et al. 1998; Freyre et al. 1998). Markers for disease-resistance genes have been identified, including those for resistance to anthracnose (Adam-Blondon et al. 1994; Alzate-Marin et al. 1997; Melotto et al. 1997; Young and Kelly 1997; Mendoza et al. 2001), rust (Haley et al. 1993; Miklas et al. 1993), bean common mosaic virus (BCMV) (Haley et al. 1994; Johnson et al. 1997; Melotto et al. 1996), bean golden mosaic virus (BGMV) (Miklas et al. 1996; Urrea et al. 1996), common bacterial blight (Vallejos et al. 1992; Nodari et al. 1993; Jung et al. 1996; Miklas et al.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

29

1996; Jung et al. 1997), and ashy stem blight (Olaya 1995). A few genes for physiological traits have also been tagged, including photoperiod sensitivity (Gu et al. 1993; Gu et al. 1994), drought (Schneider et al. 1997), biological N2 fixation (CIAT 1998; Tsai et al. 1998) and root structure affecting P uptake (Beebe et al. 1998), as well as genes for the domestication syndrome (Koinange et al. 1996). A core map version of a linkage map of common bean, consisting of 11 linkage groups, is schematically drawn in Fig. 2.2. The reader is referred to Gepts (1999), Kelly et al. (2002), and references therein for a detailed description. Reliable PCR-based markers of high priority genes allow highthroughput screening to marker-assisted selection (MAS). The application of markers in MAS is most effective if markers are linked in coupling to desirable genes, thus permitting positive selection of the trait, including in the case of dominant markers like RAPD. However, most breeding programs are not in a position to select for more than a few genes, given the cost involved. Thus, for MAS to be cost effective, the genes under selection must have a relatively large effect on the trait of interest. In the case of QTL, these probably need to be what have been referred to as oligogenes. Although the implementation of MAS for QTL is still not common, markers have contributed greatly to understanding the genetics and inheritance of complex traits, such as biological N2 fixation and root structure for nutrient uptake or discovering more precisely physiological relationships. For this, RIL are a useful genetic tool because they are homozygous at most loci, thus permitting observations about cosegregation of traits of interest, and about relationships among those traits, in genotypes with a common genetic background. In addition, they can be used in replicated trials, thus allowing an analysis of traits subject to environmental influence. We are studying a population of 100 RIL derived from a cross of two small seeded genotypes of Central American race Mesoamerica (Singh et al. 1991c), DOR 364 × BAT 477. The latter has presented excellent nitrogen fixation capacity in both optimal and suboptimal conditions with drought and phosphorus stress. The recently formed international consortium PHASEOMICS (March 2001), which represents different bean research centers, conceived the plan to establish the necessary framework of knowledge and materials that will result in disease-resistant, stress-tolerant, high-quality protein, and high-yielding beans, especially for deprived regions in Latin America and Africa, and finally aims at sequencing the genome of Phaseolus vulgaris L. For a detailed description of the objectives of the project “Beans (Phaseolus spp.)—Model Food Legumes” by the Phaseomics consortium, we refer to Broughton et al. (2002). Efforts on EST sequencing and transcriptome analysis will complement current studies on

30

B1 SCN, Co-1, Co-x, Co-w Ur-9 Ppd

Pal-1 fin

B2 ROG19.1490 G19.1500

Bng171a

Q01.650 D1662 D1327 AD17.350 V12.550 R20.1250 T01.300 D1032

NM,DF,DM, PD,HI,SW MG

[I-B] F06.350 T07.550 AG08.700

CBBBA SWBJ NM,DF, DM,NP, WMAG L5

ROD3.930 AN16.700

V12.1050

Co-u, R3

107

St

O12.900 D1287 P07.700

D0166 V12.700 G05.850 G06.1100 U12.500 Bng174 Cel O12.1600 W20.1400 GRP1.8-1 I06.550 AM02.1500 O19.800 Vpe-2 AM10.800 O15.1800

PvPR-2 ChS-2 Pgip

ChS W02.1100 J01.2000 D1595

AN08.800 Q09.1200 R20.400 AD04.1000 AP07.1300 V20.700 AA03.650 D1020 O10.350 HBBA Bng165 F10.900 CBBBJ GRP1.8-2 D08.800 D1009 FRRMF U20.700 AM18.1050 DO,PL Z AM10.525 CBBBH AH05.650 O20.550 Q18.600 SS W06.700 PvPR-1 S08.500 G17.800 I18.1700 Y11.350 CBBPX D1377 X11.1300 WBBH , HTBH Bng12 WMBN Pu-a G03.1150 AO11.1000 2 bc-1 D1151

V10.900 AM07.300 F13.300 AN08.900

DO,SWBJ

CBBBA

175

rk, Ane, Ms-8

[alphaAlArl- Lec]

CBBS95

NN

G

Cab-1 RVI

O20.1000 Bng224 X07.250 Z04.1200 D1325 Rbcs P09.800 G08.1100 X01.950 F08.1600 Bng71 D1298

AM04.950 Co-9, Co-y, - Co-z, D1174 Ur-5, Ur-Dorado, Ur-Ouro Negro Y17.1100

CBBBH NN FRRMF WMBN

132

bc-u

Bng119b

B5 G19.1800

Me

SGou D1132

ROS3.380

B4

D1066-2

W07.1450 Bng122 O13.700 V20.400

B3

B6 CBBBJ

Bng104

D1080 Diap-1 ROD20b F07.1550 SWBJ&PX,WMPX U12.950 D1198 K10.1300 H13.300 Aco-2 S18.500 D1157 Lox-1 D1301 AL08.900 D1251 AH17.500 Bng162

ANTBJ,ASBDX, BGMVDX 95

Hsp70

S18.1500 U10.2000 U19.500 AD09.950

bc-3

Bng94 P2062 ROF7.540

Ur-4 Cdc-2 72

HBBA CBBS95

S18.650 AD09.700

WM HTB60 LDG

D1086 ROD3.560 D03.600 W13.350 T08.500 AD12.350 O10.650 V D0096 U19.350 G05.1150 AM06.1000

SW CBB 113

ChI Bng199 J09.950 Y04.1050 Bng060

Co-v

Phs

P Co-6 Ef

Pal-2 LegH Per Uri-2 Lec-2 Lec-3

y

Bng191

Asp

I16.900

NN, WM AG&PX CBBBJ&BH&DX, BGMVDX ,ASBDX

A10.1100 T08.650

SW BJ&MG&PX

DJ1kscar

Gy [C-R] Co-4

Bip, Ana Y04.1600

CBBXC

GS-c -

109

WMBN

U20.1150 D1468 D1505 Bng205 O16.600 I08.1500 AJ13.1350 E19.1500 I03.1550 H11.400 L04.950 X15.450 W01.650 E19.1200

WMPX

Y08.1350 AA03.1100 AD08.450

AM07.600 Bng73 D1055

WMB60 NM,DF,DM, NP,HI

CBBPX

W02.750 Bng228 Bng102 Cad, Ch G18.400 P07.350 P07.550 AI07.600 D1096-2 2 W13.550 T AM13.1500 O19.750 D1831 Uri-1 D1338-2

D1476 I07.700 AM14.1000 Bng200 AJ14.250 AG18.550 K19.450b U01.700 X11.700 Bng68 D1580

Bng112

L

AL15.1300 Q11.1000 AE07.800 U10.1000 AD12.450 D1308 AD17.950

CBBBH&BA&DX WBBH

L04.500 AN11.300

Ur-BAC6

AN03.350

CBBBJ ROF1.1180

HBBA

105

Bng218

89

CBBXC

D1228

HBBA,FWBA

Vpe-3

Fr

93

133

CBBBA

P06.600

AH05.1000

SW MG

rRNA-1

AO11.1200

PL

H12.1050b

G04.900 U10.900

B11

B10

Gluc

CBBXC

D1861 W20.550 P1090 AM13.450 Z04.600 Bng204 H14.550 D0190 Y04.1300

W16.550 AN03.650 I07.1200 AN08.650 D1107

B9

B8

B7

Ur-12 Hrgp36

Co-2

AJ16.250 G03.850 D1291 Bng145 P2027 W02.950 K10.700 DH20sT D1512 G08.1200

NN

WBBH CBBXC PD CBBBA

100

PL PepC

Ur-11, Ur-3, Ur-Dorado

CBBPX, WMPX

Fig. 2.2. Current core linkage map of common bean (modified from Kelly et al. 2002), representing the distribution of genes with a biochemical function, major genes coding for phenotypic traits, and QTLs. To the left of each linkage group are the framework molecular markers (smaller font), the biochemical genes (larger font), and major genes for phenotypic trait (shaded boxes). To the right are QTLs. For the meaning of the symbols, the reader is referred to Kelly et al. (2002). The stippled boxes surrounding gene symbols are putative disease or pest resistance gene clusters.

31

32

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

other legumes such as Lotus japonicus and Medicago truncatula as well as the EST projects in soybean by providing a framework for comparative genomics between legumes or within Phaseolus species. This will be of particular interest since Phaseolus, in comparison to Glycine max, L. japonicus, and M. truncatula, is a tropical legume species. To date, public databases hold relatively few entries (<500 nuclear-encoded genes) for Phaseolus. Useful data for breeding programs and functional genomics will be made freely accessible on the Phaseomic Information Resource (http://www.phaseolus.net). B. Phosphorus Efficiency Estimates from the CIAT database suggest that over 50% of bean-growing areas in Latin America and 65% to 80% of bean-growing areas in Africa are critically deficient in phosphorus (CIAT 2000). P deficiency is especially a nutritional limiting factor under N2-fixing conditions, as compared to growth in the presence of mineral nitrogen. Nodules are a strong sink for P since they require much ATP for nitrogenase functioning (AlNiemi et al. 1997; Gniazdowska et al. 1998). It is reported that their P concentration is three times higher than that of other organs (Vadez et al. 1997). There is growing interest in developing cultivars of common bean with acceptable yield at low P levels in many tropical systems. Bean plants exhibit numerous physiological and morphological responses to growth at suboptimal phosphorus supply (Lynch and Beebe 1995). These plant responses offer the possibility of studying P availability and have proven to be useful in screening of bean genotypes for low P tolerance. The effectiveness of selection for physiological traits greatly depends on factors such as the genetic variability within populations and specific stage of plant ontogeny (Araujo and Teixeira 2000). Genetic differences in P efficiency exist among common bean genotypes (Araujo et al. 1998; Yan et al. 1995) and seem to be related to geographic origin (Beebe et al. 1997). The amplitude of genotypic variability within wild accessions for certain biochemical traits suggests a genetic diversity superior to cultivated genotypes, although wild accessions did not seem more tolerant to low P conditions (Lynch and Beebe 1995). Variability of P utilization traits such as shoot biomass, shoot P concentration, and nodule dry mass was more expressive at low P levels, confirming the potential for detecting lines with high N2 fixation under low P availability, and is strongly associated with photosynthetic activity. Vadez et al. (1997) also observed genotypic variability in effects of P deficiency on nodule metabolism such as the permeability of nodules to oxygen diffusion and proton efflux, nodule nitrogenase activity, and

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

33

nodule number and mass. Thus, these traits could be a key issue in the selection of symbioses tolerant to low P, at least between genotypes of similar phenology. Low-P adapted root growth and architecture such as a vigorous branched root system (Lynch and Van Beem 1993), reduced gravitropic response of basal roots (Bonser et al. 1996; Liao et al. 2001), and “lowcost” roots of small diameter enable a plant to explore more efficiently the soil in time and space with a lower degree of spatial competition between adjacent roots within a root system (Nielsen et al. 1994). Fractal geometry is a potential new approach for a very sensitive measurement of root branching activity. With this method, differences between highly efficient genotypes and inefficient genotypes could be observed (Nielsen et al. 1999). Caution should be taken with the utility of certain morphological characteristics and evaluation criteria for selection of adapted genotypes, since this depends on other environmental constraints encountered in Phaseolus habitats, such as drought. Interestingly, lines that were selected under moderate aluminum and phosphorus stress are also performing well under optimal soil conditions, yielding as much as 40% more than the standard high yielding controls. It is possible that selection has led to improved root systems that perform well under any conditions (CIAT 2000). A well-recognized symptom of a limiting inorganic phosphate (Pi) concentration in plant tissues is a reduction in shoot growth, while root growth is unaffected or slightly stimulated (Yan et al. 1995). Increase of the root to shoot biomass ratio on the other hand is correlated with a higher carbohydrate concentration in the roots (Rychter and Randall 1994) due to higher translocation of photosynthates to the roots (Ciereszko et al. 1996). Increased sucrose synthase activities are responsible for the enhanced sucrose hydrolysis in the growing parts of roots under phosphorus deprivation (Ciereszko et al. 1998). P deficiency stress in root meristematic tissue also provokes increased size of the vacuolar compartment, ultrastructural changes of mitochondria in cortical cells, and changes in factors controlling respiration rate (Wanke et al. 1998). Carbohydrate mobilization and ultrastructure of bean root cells are traits worth considering in order to elucidate their role in acclimation of plants to P deficiency. Internal and exuded acid phosphatase activity (APA) and release of organic acids are generally believed to be important in the regulation of P nutrition inside the plant (Helal et al. 1990; Tadano et al. 1993; Ryan et al. 2001). However, their role in plant adaptation to low P availability remains controversial and their use as a diagnostic criterion for P deficiency needs to be critically evaluated. A positive relation was reported

34

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

between root APA and phosphorus uptake in bean (Helal et al. 1990). However, results obtained by Yan et al. (2001) do not support a major role for leaf APA induction in regulating common bean adaptation to phosphorus efficiency. On the other hand, Shen et al. (2002) demonstrated that Andean genotypes performed a higher P solubilizing activity than Mesoamerican genotypes, which they attribute to the higher exudation of organic acids, particularly citrate. Genomic research will certainly set the stage for the dissection of the P starvation signal transduction pathway in plants and the identification of key molecular determinants of P efficiency leading to genetically altered plants that are enhanced in acquiring P from the soil (Raghothama 2000). Beans that grow in P-deficient soils can particularly benefit from their association with arbuscular mycorrhizae (AM) through the acquisition of greater amounts of P from the soil. It is interesting to note that both AM and Rhizobium symbioses rely on partially overlapping genetic programs, suggesting a bacterial recognition cascade that is derived from an evolutionary ancient mycorrhizal recognition pathway (Duc et al. 1989; Stracke et al. 2002). Two symbiotic mutants were described for common bean with respectively a Nod+/Fix–/Myc– phenotype and a Nod–/Myc+phenotype (Shirtliffe and Vessey 1996). Given the promiscuous behavior of beans, not only referring to the generally low fungal specificity but also toward rhizobial strains, it will be of particular interest to decipher the unique components of the signaling pathway leading to nodule development and mycorrhizal interaction by using a comparative approach with other model legumes. Besides, understanding the mechanisms of nutrient exchange, such as P, in bean-AM symbiosis is an important future research goal. C. Enhanced Nitrogen Fixation Common bean is often grown on marginal lands limited in available soil N, and with minimal N fertilization. Improving the potential of biological N2 fixation would reduce N depletion in such soils (at least for crops with a small harvest index), improve grain yield, and possibly enhance protein content. The capacity to fix N2 is variable among genotypes of common bean, ranging from 4 to 59% of the required N2 derived from the atmosphere (Hardarson et al. 1993). Genetic improvement of nodulation and N2 fixation in common bean was discussed by Bliss (1985, 1993), emphasizing the evaluation of direct selection traits such as nodule mass, nitrogenase activity, and xylem ureide content in segregating populations grown under low-N soil conditions. Over a period of 13 years, Bliss and his colleagues selected

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

35

a high N2-fixing progeny from crosses of commercial and vigorous high N2-fixing parents under low-N soils (Bliss et al. 1989; Bliss 1993). Recently, Barron et al. (1999) also isolated elite populations with high N2-fixing ability through recurrent selection for high seed and biomass under low-N conditions. A few studies indicate that breeding for preferential infection/ nodulation has potential, and emphasize the genetic contribution of the plant to the nodulation competitiveness phenomenon. Identification of genotypes that do not nodulate with an undesirable strain and then introduce this qualitative trait into other genetic backgrounds is a possible approach, as demonstrated with selected soybean genotypes and isolates of the dominant Bradyrhizobium serogroup 123 (Cregan et al. 1989). With this strategy for a “restricted” symbiosis, the developed cultivars nodulate only with highly effective indigenous or inoculant strains. A converse approach is to breed the host plant for preferential infection/nodulation with one or more desirable, high N2-fixing strains (Rosas et al. 1998). This will provide more insight into the quantitative trait of strain preference. In order to identify germplasm and to follow the preferential nodulation trait in segregating populations, Rosas and coworkers (1998) used a rapid visual screen for plant lines that preferentially nodulate with a desirable strain, R. etli KIM5s, in the presence of indigenous R. etli strains. Consistent with the concept of co-evolution of the Rhizobium and the wild germplasm of beans, the preference for the indigenous R. etli strains appeared to be more common among Mesoamerican accessions, as compared to the Andean accessions. KipeNolt et al. (1992) found a tendency for Middle American bean materials to prefer CIAT632, a R. etli strain isolated from the same geographic region, whereas certain Andean accessions preferred R. tropici CIAT899 from Colombia. Similarly, Montealegre et al. (1995) demonstrated that the bean cultivar RAB39 is preferentially nodulated by R. tropici CIAT899 instead of R. etli strains when applied together, independently of the temperature or pH regime. A more detailed analysis showed that this host preference was extended even when the R. tropici strain was applied with a delay of 8 h compared to the competitor strains of R. etli (Montealegre and Graham 1996). A more straightforward approach will certainly be the dissection of the signal cascade leading to nodule development by determining the molecular basis of mutations in genes, partly based on existing genetic resources coming from other legume genome sequencing initiatives (Endre et al. 2002). Genes that were thought to have unrelated functions could then be seen as part of a complex network of interacting genes and their products, and unveil previously unrecognized relationships.

36

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

It would also be of interest to start breeding programs of beans for enhanced symbiotic N2 fixation in high-N soils, since there is evidence for such a trait in different members of the Phaseoleae tribe (Dakora 1998; Park and Buttery 1988). Soybean mutants resulting in increased nodule number (called hypernodulation or supernodulation) simultaneously led to an alleviation of nitrate inhibition of the symbiosis (Carroll et al. 1985). These mutants [termed nts (nitrate-tolerant symbiotic)] appeared defective in the autoregulation of nodulation, which is controlled by a single recessive gene operating through the shoot (Delves et al. 1986; reviewed in Gresshoff 1993), encoding a receptor-like protein kinase GmNARK (G. max nodule autoregulation recepter kinase) similar to Arabidopsis CLAVATA1 (CLV1) (Searle et al. 2002). GmNARK has a major role in long-distance signaling mechanisms. A characteristic reduction of total plant growth and restricted root growth, eventually resulting in reduced grain yield, was observed for these mutants as compared to their parental genotypes (Day et al. 1986). Similar mutant phenotypes have been generated from a supernodulation mutant of Phaseolus vulgaris (Buttery and Park 1990; Buttery et al. 1990). Field evaluation of supernodulating soybean genotypes revealed that the increase in N2 fixation could not compensate for whole plant yields that were commonly 20–30 % below those of the wild type cultivars (Herridge and Rose 2000). The major advantage of the supernodulators is to provide a source of N for succeeding non-leguminous crops leading to higher yields and to increased soil Rhizobium levels. Characterization and molecular mapping of the genes controlling nitrate tolerance is a potential approach for selection and breeding of beans for enhanced N2 fixation. In this way, breeding is much more focused on increasing the percentage of crop N derived from N2 fixation rather then genetically based improvements in yield. The reader is advised to consult the above mentioned comprehensive review (Herridge and Rose 2000) on the selection and breeding of soybean for enhanced N2 fixation. III. SELECTION OF OPTIMIZED RHIZOBIUM STRAINS FOR BEAN INOCULATION Rhizobia that is to be employed as inoculum must meet important criteria before release into the field (Date 2000). An inoculant strain must be able to form highly effective nodules on the host for which it is recommended, preferably under a wide range of field conditions. Furthermore, successful strains must be competitive in nodule formation and persist in the soil in the presence of the host plant. Nodulation com-

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

37

petitiveness is still poorly understood and a major hurdle in transferring knowledge from the laboratory to the field. In areas where beans have been planted for many years, often only a small percentage of the nodules contain the applied inoculant strain. Consequently, because indigenous saprophytic Rhizobium species are often less effective N2 fixers, significant yield increases were often not attained (Ramos and Boddey 1987; Thies et al. 1991; Wolff et al. 1991). The use of mixed strains rather than single strains of rhizobia is suggested since this strategy helps to overcome the competitiveness of the ineffective indigenous rhizobial strains and achieve good nodule occupancy by the inoculant (Daba and Haile 2002). Moreover, tropical environmental conditions can severely affect inoculant persistence in the soil and rhizosphere of plants as well as efficiency of nodulation. Both indigenous soil bacteria and the inoculant strain are competing for limited carbon sources and are exposed to abiotic stresses such as elevated temperature, acidity, high osmolarity, or oxidative stress. Thus, the success of an inoculant strain depends on its capacity to adapt to adverse conditions. Much effort is being directed toward identifying superior strains occupying bean nodules under geographically different agronomic field conditions. Other criteria include the need for easy growth of the inoculum strains in broth culture, growth and survival in peat culture, and ability to survive on the seed prior to germination (Date 2000). Using peat as a carrier for rhizobial inoculants is known to have an advantage in that it has a very high water holding and buffering capacity that protect rhizobia from unfavorable soil conditions. Products with an extended shelf life are high priority for inoculant manufacturers (Catroux et al. 2001). It has to be mentioned that seed inoculation, which today is still the most common way to introduce effective rhizobia into the soil, often results in a high density of bacteria near the seed, and nodulation that is restricted to the upper taproot (Hardarson et al. 1989). Therefore, proper inoculant placement away from the seed and near the lateral roots is a key factor for good strain performance (Date 2000; Wadisirisuk et al. 1989; Danso and Bowen 1989). Lateral root nodules play an important role in providing fixed nitrogen, especially during later stages of the growing season (Hardarson et al. 1989). Alternative inoculation strategies, aiming at increased movement and contact of the inoculant rhizobia with the legume roots, for post-emergence lateral root application include pivot irrigation delivery of the inoculant, either as a liquid or a peat powder, and post-emergence side-dress banding into the furrow (S. Smith, person. commun.). Based on the diversity of its microsymbionts, bean is an undiscriminating host. Information about infection and nodulation signaling in

38

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Rhizobium-bean symbiosis was reviewed recently by Laeremans and Vanderleyden (1998). Extensive research on the molecular basis of symbiotic promiscuity, the N2 fixation capacity, and regulatory mechanisms, among other traits, has revealed many molecular determinants needed for an effective and efficient Rhizobium-bean interaction (as summarized in Table 2.3). Recent developments and practical applications regarding the selecting of elite bean-nodulating Rhizobium strains under stress conditions are discussed below. Interestingly, not only α-Proteobacteria but also β-Proteobacteria are now described that are capable of nodulating leguminous plants, for which the genetic control of nodulation also involves nod genes (e.g., the so-called β-rhizobia belonging to the Burkholderia genus) (Moulin et al. 2001). This points to yet unexplored and overlooked classes of rhizobial bacteria that might be encountered in nodules of a wide range of leguminous plants. A. Strains with Enhanced Nitrogen Fixation Capacity During the past two decades, molecular analysis of many genetic loci in Rhizobium that have an effect on bean symbiosis has led to a better understanding of symbiotic N2 fixation (Table 2.3). Genetically engineered Rhizobium strains with high SNF capacity, at least under controlled environmental conditions, were constructed. With few exceptions, genetically modified bean-nodulating rhizobia have yet to result in improved N2 fixation under field conditions. Nevertheless, several approaches are promising. The need for very precise regulation of oxygen concentration during the symbiotic N2 fixation process was taken as a strategy for constructing improved inoculant strains. A genetically engineered R. etli strain expressing the vhb gene, encoding a bacterial hemoglobin from Vitreoscilla sp., shows increased respiratory activity, compared to the wild type strain, probably through a better O2 transporting capacity of leghemoglobin at the bacteroid level. A higher nitrogenase activity provoked a higher total nitrogen content of inoculated bean plants (Ramírez et al. 1999). Nitrogenase activity was strongly enhanced when the fixNOQP genes, encoding the symbiotic high-affinity cbb3 type terminal oxidase, were constitutively expressed from a strong promoter in a R. etli ntrC mutant background (Soberón et al. 1999). Altered bacterial carbon and/or nitrogen metabolism were also shown to be powerful tools for a higher N2 fixation capacity. Lowering the basal nitrogen metabolism in the bacteroids by obstructing the glutamine synthetase-glutamate synthase (GS-GOGAT) pathway positively affected nitrogenase activity in R. etli and increased nitrogen supply to the plant

Table 2.3. Effect of mutated genetic loci of Rhizobium spp. on Phaseolus vulgaris L. symbiosis. Abbreviations used are R. (Rhizobium), R. l. (Rhizobium leguminosarum), S. (Sinorhizobium), LPS (lipopolysaccharides), and EPS (exopolysaccharides).

Similarity Group

Phenotype on Phaseolus vulgarisx

Rhizobial strainz

Function or homologyy

R. tropici CIAT899 S. sp. BR816 S. sp. BR816 S. sp. BR816 S. sp. BR816 R. l. bv. phaseoli 8002 R. l. bv. phaseoli 8002 R. l. bv. phaseoli 8002 R. sp. NGR234 R. etli CNPAF512 R. tropici CIAT899

transcriptional activator transcriptional activator transcriptional activator transcriptional activator transcriptional activator transcriptional activator transcriptional activator transcriptional activator transcriptional activator deacetylase methyltransferase

Nod+/– Nod+ Nod+ Nod+/– Nod+/– Nod+v Nod+v Nod+v Nod– Nod– Nod+/– Fix–u

nodS

Rhizobium sp. NGR234

methyltransferase

Nod+Fix–

nodPQ

R. tropici CIAT899

Nod+

nodPQ

S. sp. BR816

Nod+

Laeremans et al. 1997

nodPQ

R. tropici CFN299

Nod++ or Nod+/–t

Laeremans et al. 1996

nodH nodH nolK nolL

R. tropici CFN299 R. tropici CIAT899 R. etli CNPAF512 R. etli CE3

ATP-sulphurylase and APS kinase ATP-sulphurylase and APS kinase ATP-sulphurylase and APS kinase sulphotransferase sulphotransferase fucose synthetase acetyltransferase

van Rhijn et al. 1994a van Rhijn et al. 1994a van Rhijn et al. 1994a van Rhijn et al. 1994a van Rhijn et al. 1994b Davis and Johnston 1990 Davis and Johnston 1990 Davis and Johnston 1990 van Rhijn et al. 1994a Xi et al. 2001 K. Vlassak and J. Vanderleyden, unpublished K. Vlassak and J. Vanderleyden, unpublished Folch-Mallol et al. 1996

Nod++ Nod+ Ndv–fix– Nod+ or Nod+/–t

Laeremans et al. 1996 Folch-Mallol et al. 1996 Xi et al. 2001 Corvera et al. 1999

Genes/orfs

Nodulation genes nodD1w nodD1 nodD2 nodD3 nodD4 nodD1 nodD2 nodD3 nodD1 nodB nodS

Reference

39

(continued)

40

Table 2.3.

Similarity Group

Continued

Genes/orfs

Nodulation genes (cont.) noeJ

Rhizobial strainz

R. etli CNPAF512

noeJs

R. tropici CIAT899

nodIJ nodO

R. etli CE3 R. sp. BR816

Infection genes LPS ABCDEF GHIKLq

Function or homologyy

mannose-1-P-GDPtransferase mannose-1-P-GDPtransferase Nod factor export facilitated Nod factor uptake ??r; altering cation flow through plant membrane

Phenotype on Phaseolus vulgarisx

Reference

Ndv – Inf – Fix–

Xi et al. 2001

Nod – Fix –

Nogales et al. 2002

Nod+/– Nodd

Cárdenas et al. 1995 Vlassak et al. 1998

Noel et al. 1986; Brink et al. 1990; Cava et al. 1990; Noel et al. 1996; Noel and Duelli 2000; Noel et al. 2000; Lerouge et al. 2001 Noel et al. 1996; Noel et al. 2000 Lerouge et al. 2001 Cava et al. 1989; Cava et al. 1990; Garcia de los Santos and Brom 1997 Garcia de los Santos and Brom 1997

R. etli CE3

ND

Ndv – Inf – Fix–

Jq

R. etli CE3

Ndv – Inf+Fix+

wzm/wzt lpsb1

R. etli CE3 R. etli CE3

?? subtle changes in decoration of LPSIr ABC-transporter galacturonic acid transferase

lpsb2

R. etli CE3

glucose epimerase or dehydratase

Ndv – Inf – Fix–

Ndv – Inf – Fix– Ndv – Inf – Fix–

EPS

lps-166

R. etli CFN42

lpcA

R. l. bv. phaseoli

lpcB exoBp orf2

R. l. bv. phaseoli R. etli CE3 R. etli KIM5s

orf3 orf4 lpeA psiAB

R. etli KIM5s R. etli KIM5s R. etli CE3 R. l. bv. phaseoli 8002

pssA

R. l. bv. phaseoli 8002

psrA

R. l. bv. phaseoli 8002

Nitrogen fixation genes iscN nifHan nifHb nifHc nifA nifA

R. etli CNPAF512 R. etli CE3 R. etli CE3 R. etli CE3 R. etli CNPAF512 R. l. bv. phaseoli 8002

synthesis of UDPNdv – Inf – Fix – QuiNAc (UDP-GlcNAc→ UDP-QuiNAc) galactosyl transferase ND

Cava et al. 1989; Noel et al. 2000 Poole et al. 1994; Allaway et al. 1996 Allaway et al. 1996 Diebold and Noel 1989 Vinuesa et al. 1999

41

Kdo transferase UDP-glucose 4’ epimerase sugar nucleotide epimerase or dehydratase α-glycosyltransferase β-glycosyltransferase methyltransferase exopolysaccharide production repressor, similar to Rhizobium ExoX similar to Rhizobium ExoY transcriptional repressor

ND ND ND

iron-sulfur cofactor metabolism nitrogenase reductase nitrogenase reductase nitrogenase reductase transcriptional regulator transcriptional regulator

Nod+Fix–o

Dombrecht et al. 2002

Nod+Fix+/–m Nod+Fix+/–m Nod+Fix+/–m Nod+Fix– Nod+Fix–

Romero et al. 1988 Romero et al. 1988 Romero et al. 1988 Michiels et al. 1994b Borthakur et al. 1987; Hawkins and Johnston 1988

Ndv– Fix– ND Ndv+ Fix+ Ndv – Inf – Fix–

Vinuesa et al. 1999 Vinuesa et al. 1999 Duelli et al. 2001 Borthakur et al. 1985; Latchford et al. 1991

Nod+Fix+

Borthakur et al. 1986; Latchford et al. 1991 Borthakur and Johnston 1987; Mimmack et al. 1994

Nod+Fix+

(continued)

42

Table 2.3.

Similarity Group

Continued

Genes/orfs

Rhizobial strainz

Nitrogen fixation genes (cont.) orf180 R. etli CNPAF512 fixLJ R. etli CNPAF512

Function or homologyy

Phenotype on Phaseolus vulgarisx

Reference

Nod+Fix+/– Nod+Fix+/–

Dombrecht 2001 D’Hooghe et al. 1995

Fix–

Xi et al. 2001

Nod+Fix+

Girard et al. 2000

Nod+Fix+

fixB

R. etli CNPAF512

fixL f l

R. etli CE3

fixKd l

R. etli CE3

peroxidase two-component regulatory system electron transfer fixAB complex two-component regulatory system transcriptional activator

fixKf l nifS s

R. etli CE3 R. tropici CIAT899

transcriptional activator cysteine desulfurylase

Nod+Fix+ Nod+/–Fix+/–

Girard et al. 2000; Soberón et al. 1999 Girard et al. 2000 Nogales et al. 2002

R. etli CNPAF512 R. etli CNPAF512 R. etli CNPAF512

calsymin transcriptional repressor LuxI type of autoinducer synthase LuxR type of transcriptional regulator

Nod+Fix+/– Nod+Fix+ Nod+Fix+/–

Xi et al. 2000 Xi et al. 2000 Daniels et al. 2002

Nod+Fix+/–

Daniels et al. 2002

Michiels et al. 1998b; Michiels et al. 1998c Michiels et al. 1998c J. Michiels, unpublished

Bacteroid development casA casR cinI cinR

R. etli CNPAF512

Global regulation rpoN1

R. etli CNPAF512

σ54 factor

Nod+Fix+

rpoN2 fnrN

R. etli CNPAF512 R. etli CNPAF512

σ54 factor transcriptional regulator

Nod+Fix–o Nod+Fix+/–k

lrp cyaC Nitrogen regulation ntrBC

R. etli CNPAF512 R. etli CE3

global regulator adenylate cyclase

Nod+Fix+ Nod+Fix+

Rosemeyer et al. 1998 Téllez-Sosa et al. 2002

R. etli CE3

two-component regulatory system glutamine synthetase I ammonium transporter glutamate synthase subunit (GOGAT) glutamate synthase subunit (GOGAT) arginine deiminase pathway sensor of two-component regulatory system

Nod+Fix+ Nod+Fix+/– ND j Nod+Fix++

Moreno et al. 1992; Patriarca et al. 1993 Moreno et al. 1991 Taté et al. 1998 Castillo et al. 2000

Nod+/– Fix+/–

Ferraioli et al. 2002

Nod+Fix+/–

D’Hooghe et al. 1997

Nod–Fix–

Nogales et al. 2002

Nod+/– Fix+ Nod+/– Fix+

Pardo et al. 1994; Hernández-Lucas et al. 1995a Pardo et al. 1994

Nod+Fix++

Cevallos et al. 1996

Nod++Fix++i

Marroqui et al. 2001

Inf – Fix–h

Soberón et al. 1993; Tabche et al. 1998

Nod+Fix–

Soberón et al. 1999

glnA amtB gltB

R. etli CE3 R. etli CE3 R. etli CE3

gltD

R. etli CE3

arcABC

R. etli CNPAF512

ntrYs

R. tropici CIAT899

Carbon regulation ccsA

R. tropici CFN299

pscA

R. tropici CFN299

phaC

R. etli CE3

glgA

R. tropici CIAT899

Respiration complexes cycHJKL

fixNOQPdl

R. etli CE3

R. etli CE3

chromosomal citrate synthase plasmid-encoded citrate synthase poly-β-hydroxybutyrate synthase glycogen synthase

43

subunits of haem lyase enzyme/cytochrome biogenesis synthesis of symbiotic cytochrome terminal oxidase cbb3

(continued)

44

Table 2.3.

Similarity Group

Continued

Genes/orfs

Rhizobial strainz

Respiration complexes (cont.) fixNOQPf l R. etli CE3

Function or homologyy

Phenotype on Phaseolus vulgarisx

Nod+Fix+

Girard et al. 2000

Nod+Fix+

Xi et al. 2001

Nod+Fix+

Xi et al. 2001

ND

Aguilar and Soberón 1996

Nod+/– Fix–

Soberón et al. 1989; Aguilar and Soberón 1996 Soberón et al. 1990; MirandaRíos et al. 1997

fixG

R. etli CNPAF512

fixI

R. etli CNPAF512

ccmA

R. etli CE3

ccmB

R. etli CE3

synthesis of symbiotic cytochrome terminal oxidase cbb3 assembly and/or stability of the cbb3-type heme-copper oxidase assembly and/or stability of the cbb3-type hemecopper oxidase ATP binding protein of ABC type transporter system hydrophobic protein

thiCOGE

R. etli CE3

thiamin synthesis

Nod+Fix++g

R. etli CE3 R. tropici CIAT899 R. tropici CIAT899 R. tropici CFN299 R. etli CNPAF512

transcriptional repressor glutathione synthetase ribose transport proteins ribose transport proteins host-inducible gene S. fredii

Nod+Fix+Cmp– Nod+Fix+Cmp– Nod+Fix+Cmp– Nod+Fix+Cmp+ NoddFix+Cmp–

Nodulation competitivity rosR gshB teu teu orf3

Reference

Bittinger et al. 1997 Riccillo et al. 2000 Rosenblueth et al. 1998 Rosenblueth et al. 1998 Michiels et al. 1995

nfeD

R. etli TAL182

slp

R. etli TAL182

Amino acid auxotrophs cysG metZ

R. etli CE3 R. etli CE3

argC

R. etli CE3

trpB serA hisA

R. etli CE3 R. etli CE3 R. etli CE3

leuC

R. etli CE3

pheA aroK lysA

R. etli CE3 R. etli CE3 R. etli CE3

Miscellaneous melA melC raiI raiR nadA csaA dnaJ

R. l. bv. phaseoli R. l. bv. phaseoli R. etli CNPAF512 R. etli CNPAF512 R. etli CNPAF512 R. etli CNPAF512 R. l. bv. phaseoli

ornithine cyclodeaminase stomatin-like protein siroheme synthetase O-succinylhomoserine sulfhydrylase N-acetylglutamylphosphate reductase tryptophan synthase L-serine dehydratase phosphoribosylformimino-5 aminoimidazole carboxamide ribotide isomerase isopropylmalate dehydratase chorismate mutase shikimate synthase diaminopimelate decarboxylase

45

tyrosinase similarity to ntrA/rpoN autoinducer synthase transcriptional regulator quinolate synthetase molecular chaperonin heat shock protein; involved in glutamate metabolism

pr. Cmp–

Borthakur and Gao 1996b

Nod+Fix+Cmp–

You et al. 1998

Nod+Fix+ Nod–Fix –

Taté et al. 1997 Taté et al. 1999

Nod–Fix –

Ferraioli et al. 2001

Nod–Fix – Nod+/–Fix+/– Nod+/–Fix+/–

Thorne and Williams 1999 Ferraioli et al. 2002 Ferraioli et al. 2002

Nod+/–Fix+/–

Ferraioli et al. 2002

Nod+/–Fix – Nod+/–Fix – Nod+/–Fix –

Ferraioli et al. 2002 Ferraioli et al. 2002 Ferraioli et al. 2002

Nod+Fix+f Nod+Fix –f Nod++Fix+ Nod++Fix+ Nod+Fix – Nod+Fix –

Hawkins and Johnston 1988 Hawkins et al. 1991 Rosemeyer et al. 1998 Rosemeyer et al. 1998 Xi et al. 2001 Xi et al. 2001

Nod–Fix –

Labidi et al. 2000 (continued)

46 Table 2.3.

Similarity Group

Continued

Genes/orfs

Miscellaneous (cont.) dnaJs

Rhizobial strainz

R. tropici CIAT899

greAs

R. tropici CIAT899

alaSs groESL rmrA

R. tropici CIAT899 S. sp. BR816 R. etli CFN42

rmrB

R. etli CFN42

kdpE

R. tropici CIAT899

kups

R. tropici CIAT899

recA

R. etli CFN42

Function or homologyy

heat shock protein; involved in glutamate metabolism transcriptional elongation factor; stress responsive protein alanyl-tRNA synthetase chaperonin membrane fusion protein major facilitator protein i.p. drug-resistance protein response regulator of two-component regulatory system involved in K+ uptake system recombinase

Phenotype on Phaseolus vulgarisx

Reference

Nod–Fix–

Nogales et al. 2002

Nod+/– Fix+/–

Nogales et al. 2002

Nod+/– Fix+/– Nod+Fix+/– Nod+/– Fix+

Nogales et al. 2002 Luyten 1999 Gonzalez-Pasayo and Martínez-Romero 2000 Gonzalez-Pasayo and Martínez-Romero 2000

Nod+/–Fix+

Cmp+/–e

Botero et al. 2000

Nod+/– Fix+/–

Nogales et al. 2002

NoddFix+

Martínez-Salazar et al. 1991

z Only those rhizobial strains are included that are genetically studied in relation to their symbiotic reaction with P. vulgaris. For a more complete picture of rhizobia nodulating bean, we refer to Michiels et al. (1998a).

y

Functional description was based on sequence similarity as well as in silico analysis and biochemical characterization. Bean phenotype compared to the wild type level of inoculation was scored as follows: enhanced nodulation (Nod++), unaltered nodulation (Nod+), delayed nodulation (NodDd), reduced nodule number (Nod+/–), no nodules formed (Nod–), different nodule development pattern (Ndv–), unaltered infection (Inf+), no infection (Inf –), enhanced N2 fixation (Fix++), unaltered N2 fixation (Fix+), reduced N2 fixation (Fix+/–), no N2 fixation (Fix –), reduced nodulation competitiveness (Cmp–), unaltered nodulation competitiveness (Cmp+) and not determined (ND). w Numbered nodD copies represent multiple alleles. v The authors only mentioned that the formation of nodules was not abolished. u Waelkens et al. (1995) found that these nodS mutants were Nod– on beans. t Increased or decreased nodule numbers were found depending on the inoculated bean cultivar. s Mutation in this gene (or downstream genes) plays a role in adaptation to saline and osmotic stress (Nogales et al. 2002). r ?? = postulated function. q Complementation groups. p The exoB mutant was affected in both LPS and EPS production. o Nitrogenase activity per plant is reduced to 10% of the wild type N2 fixation level. n Numbered nifH copies represent multiple alleles. m Romero et al. (1988) found that the triple nifH mutant is Fix–. l d and f stand for plasmid d and f of the six plasmids present in R. etli. k Nitrogenase activity per plant is reduced to 30% of the wild type level. j Downregulation of the amtB gene is essential for an effective symbiotic interaction. i The authors found that glgA mutants grown in glucose have reduced high-molecular-weight EPS, as well as other pleiotropic effects like increased respiration. h Tabche et al. (1998) report that a mutation in cycH not affecting the downstream genes is Inf+Fix–. g Enhanced N2 fixation is due to mutation in promoter region of this cluster, resulting in constitutive expression of the operon. f No melanin production. e The reduced symbiotic competence is based on initial studies on the symbiotic properties of the kdpE mutant; this gene is involved in the in situ P acquisition of bacteroids from bean nodules. x

47

48

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

(Castillo et al. 2000). Disturbed bacterial carbon metabolism in R. tropici and R. etli mutants, blocked in respectively the synthesis of glycogen and polyhydroxybutyrate, also enhanced symbiotic N2 fixation (Cevallos et al. 1996; Marroqui et al. 2001). B. Nodulation Competitiveness and Persistence in the Soil Studying the genetic basis of Rhizobium nodulation competitiveness is challenging because of the high variability and the fact that it is a quantitative trait. Consequently, so far identified genes affecting competition for nodulation encode a diversity of functions. Various physiological attributes such as bacteriocin and exopolysaccharide production, motility, speed of infection, level of N2 fixation, growth rate, and colonization in the rhizosphere have been shown to be involved in the competitiveness of Rhizobium strains (Triplett and Sadowsky 1992; Vlassak and Vanderleyden 1997). Antimicrobial compounds secreted by rhizobia can detrimentally affect concurrent rhizosphere bacteria. The peptide antibiotic trifolitoxin (TFX), isolated from R. leguminosarum bv. trifolii T24, has been described as a possible means to prevent nodulation by TFX sensitive strains including most Rhizobium spp. but also strains of Agrobacterium and various taxonomically more distant α-proteobacteria (Triplett and Barta 1987). Genetic enhancement of nodulation competitiveness in R. etli expressing the tfx genes was successful both in sterile and non-sterile soil and under field conditions (Robleto et al. 1997; Robleto et al. 1998a). Additionally, Robleto et al. (1998b) clearly demonstrated a reduction in the diversity of α-proteobacteria in the bean rhizosphere after inoculation with the genetically engineered R. etli strain producing TFX as compared to the wild type strain. Rosemeyer et al. (1998) showed that R. etli CNPAF512 produces an autoinducer molecule, which is comparable to a growth inhibiting factor, called small (Schripsema et al. 1996). Small inhibits the growth of closely related strains (Hirsch 1979) and is involved in stationary-phase survival of R. leguminosarum bv. phaseoli following nutrient starvation (Thorne and Williams 1999). Although exopolysaccharides (EPS), in contrast to lipopolysaccharides (LPS), do not seem to be essential for Rhizobium invasion in determinate nodules, like those formed on bean plants (Gray and Rolfe 1990), they are indispensable in terms of competitiveness. In R. etli, the transcriptional regulator encoded by rosR was found to be a determinant of nodulation competitiveness (Araujo et al. 1994; Bittinger et al. 1997). Subsequently, Bittinger and Handelsman (2000) demonstrated that a subset of the genes regulated by RosR are involved in cell surface char-

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

49

acteristics as well as nodulation competitivity, strongly suggesting that the altered competitiveness of a rosR mutant is due to dramatic changes in cell surface properties. Also, reduced competitiveness was observed for mutants of R. tropici CIAT899 affected in EPS production (Milner et al. 1992). Rosenblueth et al. (1998) isolated from R. tropici the teu genes (tropici exudate uptake) which are related to sugar transport genes. Interestingly, the inducing compound, though not yet identified, appears to be characteristic for bean and closely related legumes (e.g., Macroptilium) but is not present in soybean, Leucaena, pea, alfalfa, or maize. The bean exudate compound might be advantageous for rhizosphere survival and efficiency of its symbiont. A R. tropici CIAT899 teu mutant strain appeared less competitive than its wild type counterpart. Other approaches to identify genes involved in nodulation competitiveness consisted of transferring a cosmid library of the highly competitive R. leguminosarum bv. phaseoli strain KIM5 into the inferior strain R. etli CE3. However, the enhanced competitiveness of the isolated strains was not due to the introduced DNA but probably arose because of spontaneous mutations elsewhere in the genome (Beattie and Handelsman 1993). On the contrary, Borthakur and Gao (1996a) reported the presence of a DNA region on a 150 MDa plasmid (different from the symbiotic plasmid) important for nodulation competitiveness in R. etli strain TAL182 and some other R. etli strains but absent in other Rhizobium spp. One of these genes, slp, encoding a stomatin-like protein, was characterized by You et al. (You et al. 1998). The authors suggest that slp may be of competitive advantage through better ion exchange and nutrient uptake capacity in the infection thread. Increased competitivity was also shown to be plasmid-borne in R. leguminosarum bv. phaseoli since four of the six naturally occurring plasmids appear to possess traits for competitiveness (Brom et al. 1992). It was often seen that highly efficient rhizobia introduced in soils with no history of bean production no longer occupy the nodules of their host after 2–3 years of successive bean cropping. Vlassak et al. (1996) could clearly demonstrate in subsequent field trials that R. tropici CIAT899 was the most competitive of the tested inoculant strains after the first year but, in subsequent seasons, planting without reinoculation resulted in a much lower percentage of nodule occupancy by R. tropici CIAT899 (type IIB). Increased bean nodule populations of R. etli (type I) and R. tropici CFN299 (type IIA) were found even in plots that had never been inoculated or in plots only inoculated with type IIB strains. Interestingly, strains isolated from the field with a 3-year bean-planting history were more competitive and totally different compared to the

50

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

population of a plot without bean history (before first planting). This indigenous soil population arose probably from a population present on the seed (Pérez-Ramírez et al. 1998) or from contamination, followed by multiplication. Another possibility is that these indigenous nodulating rhizobia could arise from transfer of symbiotic genes to rhizobia that lack this symbiotic information and by possible genomic rearrangements (Romero et al. 1995). The presence of non-symbiotic strains of R. leguminosarum in the soil has been reported before (Laguerre et al. 1993; Segovía et al. 1993; Soberón-Chavez and Najera 1989). These “new” soil rhizobia are particularly well adapted to the environment and better able to survive and compete than the introduced inoculant strains. In R. etli, symbiotic plasmid transfer is absolutely dependent on the presence of another plasmid (Brom et al. 2000). Interestingly, subsequent multiple plasmid curing of R. etli revealed an accumulating effect of plasmid loss on the nodulation competitiveness of the strain compared to the wild type competition levels. This strongly suggests that Rhizobium plasmids are functionally intimately related. Based on these results, other approaches, directed to the modification of global genomic features through the use of genomic rearrangements such as DNA amplifications, may open other perspectives to understand and manipulate the competitiveness of a given Rhizobium strain (Mavingui et al. 1997). The plant-host genotype is of critical importance for determining the outcome of competition experiments. Strain differences in competitiveness with a single or limited bean cultivar(s) have been welldocumented (Josephson et al. 1991; Vásquez-Arroyo et al. 1998; RodriguezNavarro et al. 2000). Nevertheless, studies in which several host cultivars are utilized to evaluate competitiveness of Rhizobium bacteria are needed to have a better idea of the usefulness of a given strain, since the host genotype strongly influences all of the tested symbiotic characteristics after inoculation with a particular inoculant strain. C. Common Bean Rhizobia Adapted to Different Environmental Factors Soil bacteria are constantly challenged by a variety of stresses in their natural environments. Among these, high temperature and soil acidity are severely affecting all stages of the symbiosis, including the stability of the symbiotic properties of the Rhizobium microsymbiont, and limit Rhizobium growth and survival in tropical soils (Hungria and Vargas 2000). Several studies on the characterization of Rhizobium bean isolates from different soil types clearly revealed that broad-host-range R. tropici-

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

51

type bacteria are predominantly present in acid soils, whereas narrowhost-range R. etli and R. leguminosarum bv. phaseoli types predominate in near-neutral soils (Martínez-Romero et al. 1991; Amarger et al. 1994; Anyango et al. 1995; Herrera-Cervera et al. 1999). Both in acid and near neutral soil, characterization of the indigenous isolates revealed a high genetic diversity among these strains. Exploiting this variability to perform screening procedures of large numbers of native strains probably still is the best strategy to overcome environmental factors affecting an optimal symbiotic interaction (Hungria et al. 1993; Hungria et al. 2000; Hungria and Vargas 2000). Therefore, screening programs for competitive and efficient Rhizobium strains performing well in tropical conditions should be encouraged. The analysis of such strains will contribute to unraveling the underlying genetic and biochemical basis of the adaptional ability of a given strain, which in turn can be used to define rational screening procedures. Nevertheless, care should be taken in evaluating strains, since examples of a lack of relationship between ecological origin and in vitro behavior do exist. For example, RodriguezNavarro et al. (2000) found isolates from neutral-alkaline soils showing acidic tolerance in vitro. Wolff et al. (1991) also identified variations in tolerance of the strains to different environmental conditions in vitro that were not related to their relative competitiveness in the field. On the other hand, it is almost impossible to select strains that are resistant to all adverse conditions. The highly competitive strain R. leguminosarum bv. phaseoli KIM5 proved to be tolerant to high temperatures and aluminum toxicity but appeared more sensitive than other strains to a low pH and to tannins and phytoalexins, two host defense compounds that inhibit nodulation (Wolff et al. 1993). The broad-host-range type strain R. tropici CIAT899 displays a high intrinsic tolerance to acidity and high temperatures (Martínez-Romero et al. 1991) and has been used for studying the molecular basis of rhizobial responses to survive acid stress and elevated temperatures. The acid pH tolerance of R. tropici does not appear to be an adaptive response, is not plasmid mediated, and the strain shows a limited ability to short-term regulation of cytoplasmic pH (Graham et al. 1994). Recently, Riccillo et al. (2000) characterized a R. tropici CIAT899 Tn5 induced acid-sensitive mutant that was affected in the glutathione synthetase gene gshB. It was demonstrated that glutathione is essential to protect R. tropici CIAT899 against not only environmental stresses such as acidity but also osmotic and oxidative shock, and that the glutathionedeficient mutant strain is affected in its ability to maintain cellular potassium levels. Previous work also led to the observation of higher potassium and glutamate levels in CIAT899 following exposure to low

52

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

pH (Aarons and Graham 1991). Furthermore, bacterial outer membrane composition and structure might play a role in acid tolerance (Graham et al. 1994). The ability of Rhizobium strains to take up glucose has been related to their ability to survive a pH shift (Clarke et al. 1993). Steele et al. (1999) could demonstrate that in R. tropici the glucose uptake mechanism was not sensitive to short- or long-term exposure to acidic conditions, whereas the acid-sensitive R. etli strains tested showed differences in their ability to transport glucose due to short-term and long-term exposure to acidity. Furthermore, the authors showed that under different pH regimes, the susceptibility of the strains to the isoflavone daidzein changes. Daidzein is one of the nod gene-inducing flavonoids in the Phaseolus vulgaris–R. tropici symbiosis (BolanosVásquez and Warner 1997). The idea of implementing a flavonoid nod gene-inducer based strategy for inoculant strains to withstand acid stress conditions is described by Hungria and Stacey (1997). The molecular basis of high temperature tolerance of R. tropici strains is still poorly understood. As R. tropici CIAT899 does not fix atmospheric N2 at elevated temperatures, its symbiotic performance under temperature stress is probably due to its capacity to survive the periods of thermal stress and to recover afterward (Michiels et al. 1994a). The requirement of heat-shock proteins for successful symbiosis was shown in several Rhizobium species (Krishnan and Pueppke 1991; Fischer et al. 1993; Ogawa and Long 1995). The involvement of a 41-kDa heatshock protein encoded by dnaJ in R. leguminosarum bv. phaseoli in glutamate metabolism and symbiosis was demonstrated by Labidi et al. (2000). In Sinorhizobium sp. BR816, the chaperonin GroEL was identified as being specifically important for symbiosis with bean but not with Leucaena (Luyten 1999). Stresses in acid soils are not only the direct effect of low pH but also include aluminium and manganese toxicity or deficiency of essential nutrients. Estimates suggest that around 40% of total bean area in Latin America is affected by Al toxicity (CIAT 2000). Aluminium resistance is a trait that varies markedly from strain to strain and does not necessarily correlate with acid tolerance (Wolff et al. 1991). The precise mechanisms involved in tolerance of R. tropici for survival at high aluminium concentrations is still not understood. Although EPS was suggested to play a role in Al protection via binding in several Rhizobium strains, a nonmucoid EPS– mutant of R. tropici CIAT899 was equally as tolerant to aluminium as the parental strains, indicating that EPS does not confer Al resistance in this strain (Kingsley and Bohlool 1992). Furthermore, possible determinants of Al resistance are not located on the pSym plasmid of CIAT899.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

53

Low available P concentrations resulting from complex formation with aluminium could be another cause of the aluminium toxicity observed in acid soils. Understanding the mechanisms by which bacteria acquire phosphorus should contribute in every respect to the identification/construction of strains with an efficient P uptake system at low external P levels. Also, mutants affected in genes for in situ P acquisition of bacteroids will provide useful information concerning phosphorus exchange between the two partners of the symbiosis and should be used in current breeding programs for P efficient crops under N2fixing conditions (see above). Recently, two phosphate (Pi) stressinducible phosphate transport systems were identified in R. tropici CIAT899 differing in their affinities for Pi. A mutant with reduced in situ Pi acquisition by its bacteroids was identified (Botero et al. 2000). Previously, Al-Niemi et al. (1997) showed that under free-living conditions, P transport and phosphatase activity increased in response to P stress in R. tropici CIAT899. Interestingly, high levels of alkaline phosphatase activity were present in bacteroids from nodules of bean plants grown in non-P limiting nutrient solution. The authors suggested that bean plants provide very low levels of Pi to the bacteroids in the nodules under normal growth conditions.

IV. CONCLUSIONS AND FUTURE PROSPECTS Phaseolus species distinguish themselves from other major grain legume species by their broad genetic base, which holds promises for achieving higher production needs, provided the germplasm is sufficiently evaluated and bred for performance. In spite of this diversity, it is well known that the genetic base of commercial bean cultivars is narrow. Systematic evaluations of wild accessions of common bean and other related species for their responses to diseases, insects, long photoperiod, drought, and low soil fertility are lagging behind as compared to cultivated species. The available genetic resources of Phaseolus should be fully exploited to develop new bean cultivars with enhanced biological N2 fixation, well-protected or adapted to adverse environmental conditions. Inputs of research centers focusing on the molecular aspects of bean and bean interactions with microorganisms should be brought together, whereas future bean breeding strategies should be directly based on integrated approaches proposed by breeders and microbiologists. Future bean breeding and genetic improvement of Rhizobium will benefit from the development of innovative techniques and new biotechnological tools. Molecular marker analysis helped in clarifying genetic

54

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

relationships and levels of genetic diversity in common bean germplasm (Gepts 1998b). This marker technology should be incorporated in bean breeding programs and integrated genetic maps will assist in gene discovery using map-based cloning. As an example, the recent cloning of a Phaseolus resistance gene cluster (R genes) of ancestral origin (comprising both Andean and Mesoamerican resistance specificities) against its fungal pathogen Colletotrichum lindemuthianum has already proven its use in understanding the Phaseolus vulgaris-pathogen coevolution phenomenon at the molecular level in the two independent gene pools (Geffroy et al. 1999). The existence of both major resistance genes and strain-specific QTL against C. lindemuthianum is challenging in terms of defining and selecting an ideotype capable of resisting the complete spectrum of pathogenic diversity (Geffroy et al. 2000). Implementation of a combination of both partial and complete resistance genes to provide durable resistance, especially in developing countries where farmers often cannot afford the purchase of pathogen-free seeds or fungicides, offers new perspectives for improving common bean yield. This strategy has been successfully applied in other crops such as rice (Wang et al. 1994). Although in recent years major progress has been made, a widely applicable transformation and regeneration protocol for Phaseolus is still awaited. Stable and heritable transformation of beans has been successful either via biolistic gene delivery in P. vulgaris (Russell et al. 1993; Aragão et al. 1998) or via Agrobacterium tumefaciens mediated integration in P. acutifolius (Dillen et al. 1997), albeit with a very low frequency of recovery. As the interspecies hybridization among the genus Phaseolus is possible (e.g., P. vulgaris × P. acutifolius) through congruity backcross and using embryo rescue techniques (Mejía-Jiménez et al. 1994), genetic improvement of common bean indirectly is among the possibilities. Good transformation technology will be employed for introducing novel genes to broaden the genetic base of common beans and perform functional analysis of agronomically important genes. The advent of molecular approaches has revolutionized the genetic fingerprinting based on stable (chromosomal) properties (MartínezRomero and Caballero-Mellado 1996; Thies et al. 2001). Phylogenetic studies play an important role in our knowledge of relationships on both sides of the plant-bacteria symbiosis. Multi-locus enzyme electrophoresis, for instance, can be considered as a highly robust tool for deriving estimates of strain relatedness and unraveling potential evolutionary pathways or for assessing bacterial diversity. Ancient lineages of native Rhizobium bacteria seem to have co-evolved with wild Phaseolus species, coinciding with their distribution (Aguilar et al. 1998; Bernal

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

55

and Graham 2001). The resulting broad geographical distribution of Rhizobium species is most probably related to their huge genetic diversity. The mechanisms of generating diversity, influenced by agricultural practices, were recently studied for the traditional mixed cropping of maizebean associations in Mexico, called “milpas” (bean climbing over maize plant) (Gutiérrez-Zamora and Martínez-Romero 2001). Interestingly, a common pool of R. etli strains was found in bean nodules and as naturally occurring maize endophytes. Below, the strategy is given that we are currently following in a program for the implementation of biofertilization in bean cultivation by optimizing the use of Rhizobium-bean symbiosis (Fig. 2.3). Landrace germplasm and new bean lines have been identified at CIAT that produce good yields and are less restricted to phosphate limitation in

Selection criteria for varietal selection based on physiological analysis and molecular markers established Populations developed

Optimization of rhizobial strains and preparation for field release

BNF performance, shoot dry weight, N content at variable N and P levels under greenhouse conditions QTL confirmed in greenhouse

SELECTION

Socio-Economics

Field trials of different Rhizobium-bean combinations Evaluation of yield and competition of rhizobial strains under specific environments (field N, P and climate) QTL confirmed in field over environments

Elucidation of new signals involved

Extrapolation of results obtained in field to other environments

Fig. 2.3. Flow chart of the strategy for integration of biofertilization in bean cultivation by optimizing the use of the Rhizobium-bean symbiosis.

56

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

comparison with the currently available commercial bean lines. From these sources, RIL have been generated for genetic studies and for eventual marking of genes for biological nitrogen fixation (BNF) potential, drought resistance, and low P tolerance with DNA markers. Since N and P are interrelated in plant nutrition, these newly developed lines, which are not commercially available yet, are the best possible starting material to test enhanced BNF under field conditions. Different combinations of selected bean lines and optimized Rhizobium strains will be examined in greenhouse and field trials in Mexico in order to define the optimal environment for selection and the appropriate selection criteria for an efficient bean cultivar and a rhizobial strain in respect to environmental settings. The influence of the applied strain on indigenous bacterial populations will be surveyed, as well as data about survival and competitivity of the inoculum will be obtained, in the long term leading to further elucidation of signals involved and giving input for optimization of inoculant strains. Simultaneously, the value of putative QTL that are potentially of agronomic utility and that have been tentatively identified under controlled conditions will be tested during the field trials to confirm their usefulness in a given set of environments and respective markers are to be converted to suitable PCR-based markers (SCAR or microsatellites). This activity is building up a set of data that will serve to analyze the effect of QTL over sites and therefore determine the range of environments for which the genes can be of use. These activities will lead to the identification of selection criteria for the bean breeding program (either physiological traits or molecular markers) that will permit breeders to recognize and select higher BNF in segregating cultivars and advanced lines. Once QTL are identified, marker assisted backcrossing can be used very effectively to introduce QTL into commercial varieties. In Participatory Breeding Programs, elite genotypes will be backcrossed to local cultivars, such as the small opaque black class, which is the most widely used in Mexico, and in the Flor de Mayo type, which guarantees high prices for farmers, by improving local cultivars such as Negro 8025 and FM Sol. For example, a number of these commercial cultivars from the Mexican plateau could benefit from BNF genes from BAT477, which has proven its potential for BNF in this environment. Geographical Information Systems (GIS) will allow extrapolating data generated during field trials to similar environments that can be expected to give a similar phenotypic response. Recently, a program called “FloraMap” was developed to map environments in which wild ancestors of crop plants can be expected to flourish, through a mathematical comparison of climatic data at known collection sites with that at all other sites (Beebe et al. 1997). Conceptually, it should be possible to perform a similar function with crop plants.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

57

Another important recent development in the field of rhizobia-bean interaction is the genome sequencing of R. etli CFN42 as well as its expression under symbiotic conditions (by transcriptomics and proteomics). So far, the complete nucleotide sequence of the symbiotic plasmid of this strain has been obtained and annotated, and can be consulted at http://www.cifn.unam.mx/retlidb/. Initiatives for comparative genomics with available genome sequencing data of different Rhizobium strains (like Sinorhizobium meliloti and Mesorhizobium loti) are awaited. The bean genome project will undoubtedly speed up integrated research. In Europe, scientific interest in bean breeding for human consumption is also high. A concerted action named PHASELIEU was formed in 1998 by an EU wide network of experts. The group focuses on the elaboration of an integrated model for improvement of Phaseolus production in Europe for human consumption through exchange and dissemination of the knowledge and expertise regarding bean research in Europe. Surprisingly, decisions of policymakers often do not stimulate ongoing bean research despite the current high demand for alternative protein sources in the diet. There is need for a rethinking of current production and consumption processes, moving toward more sustainable processes. Soybean, a major supply of protein in animal feed, is mainly imported in Europe since its production in these countries is limited by climatological factors and international agreements. Another major hurdle can be attributed to the strict EU legislation concerning the use of genetically modified crops in the food chain, which has an important impact on availability and price. Hence, opportunities may be offered for bean cropping for both animal feed as well as human consumption. Research that leads to new bean varieties with higher yields under temperate climate environments and increased nutritional value for potential markets would be valuable. Furthermore, existing molecular markers generated in international research programs can be integrated into European breeding programs. Besides technological factors, the success of bean research relies heavily on socio-economic factors. Yield improvement of sustainable lowinput bean cropping systems will only reach its goal when research centers and farmer unions reflect on the farmers’ needs and their preferences, and specificities of local farming systems. Making beans more competitive on a global scale is an important aspect, but the establishment of regional market niches with its typical, often highly priced bean cultivars adapted to regional environments, with biotic and abiotic constraints, should not be neglected. In addition to the different market classes of dry bean of major economic importance in the world, some dry

58

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

beans that are produced and consumed on a smaller scale are highly valued and appreciated locally (for example, Flor de Mayo and other closely related types or variants, produced under both irrigation and rainfed farming systems in the central highlands of Mexico) because of high yield potential and moderate levels of tolerance to low soil fertility and water stress. Other examples are Bolón Amarillo in Ecuador, the Rosinha type in Brazil, Cargamanto in Colombia, Kablankati type in Tanzania, and the ñuña or popping bean that only grows in the highlands of Peru and Bolivia (Singh 1999b). LITERATURE CITED Aarons, S. R., and P. H. Graham. 1991. Response of Rhizobium leguminosarum bv. phaseoli to acidity. Plant Soil 134:145–151. Adam-Blondon, A. F., M. Sevignac, M. Dron, and H. Bannerot. 1994. A genetic map of common bean to localize specific resistance genes against anthracnose. Genome 37:915–924. Aguilar, G. R., and M. Soberón. 1996. Cloning and sequence analysis of the Rhizobium etli ccmA and ccmB genes involved in c-type cytochrome biogenesis. Gene 182:129–135. Aguilar, O. M., M. V. Lopez, P. M. Riccillo, R. A. Gonzalez, M. Pagano, D. H. Grasso, A. Pühler, and G. Favelukes. 1998. Prevalence of the Rhizobium etli-like allele in genes coding for 16S rRNA among the indigenous rhizobial populations found associated with wild beans from the Southern Andes in Argentina. Appl. Environ. Microbiol. 64:3520– 3524. Allaway, D., B. Jeyaretnam, R. W. Carlson, and P. S. Poole. 1996. Genetic and chemical characterization of a mutant that disrupts synthesis of the lipopolysaccharide core tetrasaccharide in Rhizobium leguminosarum. J. Bacteriol. 178:6403–6406. Al-Niemi, T. S., M. L. Kahn, and T. R. McDermott. 1997. P metabolism in the bean Rhizobium tropici symbiosis. Plant Physiol. 113:1233–1242. Alzate-Marin, A. L., G. S. Baia, T. J. de Paula, G. A. de Carvalho, E. G. de Barros, and M. A. Moreira. 1997. Inheritance of anthracnose resistance in common bean differential cultivar AB 136. Plant Dis. 81:996–998. Amarger, N., M. Bours, F. Revoy, M. R. Allard, and G. Laguerre. 1994. Rhizobium tropici nodulates field-grown Phaseolus vulgaris in France. Plant Soil 161:147–156. Amarger, N., V. Macheret, and G. Laguerre. 1997. Rhizobium gallicum sp. nov. and Rhizobium giardinii sp. nov., from Phaseolus vulgaris nodules. Intl. J. Syst. Bacteriol. 47: 996–1006. Anyango, B., K. J. Wilson, J. L. Beynon, and K. E. Giller. 1995. Diversity of rhizobia nodulating Phaseolus vulgaris L. in 2 Kenyan soils with contrasting pHs. Appl. Environ. Microbiol. 61:4016–4021. Aragão, F. J. L., S. G. Ribeiro, L. M. G. Barros, A. C. M. Brasileiro, D. P. Maxwell, E. L. Rech, and J. C. Faria. 1998. Transgenic beans (Phaseolus vulgaris L.) engineered to express viral antisense RNAs show delayed and attenuated symptoms to bean golden mosaic geminivirus. Mol. Breed. 4:491–499. Araujo, A. P., and M. G. Teixeira. 2000. Ontogenetic variations on absorption and utilization of phosphorus in common bean cultivars under biological nitrogen fixation. Plant Soil 225:1–10.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

59

Araujo, A. P., M. G. Teixeira, and D. L. de Almeida. 1998. Variability of traits associated with phosphorus efficiency in wild and cultivated genotypes of common bean. Plant Soil 203:173–182. Araujo, R. S., E. A. Robleto, and J. Handelsman. 1994. A hydrophobic mutant of Rhizobium etli altered in nodulation competitiveness and growth in the rhizosphere. Appl. Environ. Microbiol. 60:1430–1436. Bala, A., and K. E. Giller. 2001. Symbiotic specificity of tropical tree rhizobia for host legumes. New Phytol. 149:495–507. Barron, J. E., R. J. Pasini, D. W. Davis, D. D. Stuthman, and P. H. Graham. 1999. Response to selection for seed yield and nitrogen (N2) fixation in common bean (Phaseolus vulgaris L.). Field Crops Res. 62:119–128. Beattie, G. A., and J. Handelsman. 1993. Evaluation of a strategy for identifying nodulation competitiveness genes in Rhizobium leguminosarum biovar phaseoli. J. Gen. Microbiol. 139:529–538. Beebe, S., J. Lynch, N. Galwey, J. Tohme, and I. Ochoa. 1997. A geographical approach to identify phosphorus-efficient genotypes among landraces and wild ancestors of common bean. Euphytica 95:325–336. Beebe, S., F. Pedraza, M. Rojas, J. Gutierrez, and J. Tohme. 1998. A genetic map of common bean combining RFLP, RAPD, SCAR and AFLP markers. Annu. Rep. Bean. Improv. Coop. 41:95–96. Beebe, S., P. W. Skroch, J. Tohme, M. C. Duque, F. Pedraza, and J. Nienhuis. 2000. Structure of genetic diversity among common bean landraces of Middle American origin based on correspondence analysis of RAPD. Crop Sci. 40:264–273. Bernal, G., and P. H. Graham. 2001. Diversity in the rhizobia associated with Phaseolus vulgaris L. in Ecuador, and comparisons with Mexican bean rhizobia. Can. J. Microbiol. 47:526–534. Bittinger, M. A., and J. Handelsman. 2000. Identification of genes in the RosR regulon of Rhizobium etli. J. Bacteriol. 182:1706–1713. Bittinger, M. A., J. L. Milner, B. J. Saville, and J. Handelsman. 1997. rosR, a determinant of nodulation competitiveness in Rhizobium etli. Mol. Plant-Microbe Interact. 10: 180–186. Bliss, F. A. 1985. Breeding for enhanced dinitrogen fixation potential of common bean (Phaseolus vulgaris L.). p. 303–310. In: P. W. Ludden and J. E. Burris (eds.), Nitrogen fixation and CO2 metabolism. Elsevier, Amsterdam. Bliss, F. A. 1993. Breeding common bean for improved biological nitrogen-fixation. Plant Soil 152:71–79. 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. Dasilva. 1989. Registration of 5 high nitrogen-fixing common bean germplasm lines. Crop Sci. 29:240–241. Bolanos-Vásquez, M. C., and D. Warner. 1997. Effects of Rhizobium tropici, R. etli, and R. leguminosarum bv. phaseoli on nod gene-inducing flavonoids in root exudates of Phaseolus vulgaris. Mol. Plant-Microbe Interact. 10:339–346. Bonser, A. M., J. Lynch, and S. Snapp. 1996. Effect of phosphorus deficiency on growth angle of basal roots in Phaseolus vulgaris. New Phytol. 132:281–288. Borthakur, D., and X. Gao. 1996b. The Rhizobium etli nfeD gene is required for nodulation competitiveness in Phaseolus vulgaris L. Genbank Accession #U41754. Borthakur, D., and X. F. Gao. 1996a. A 150-megadalton plasmid in Rhizobium etli strain TAL182 contains genes for nodulation competitiveness on Phaseolus vulgaris L. Can. J. Microbiol. 42:903–910.

60

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Borthakur, D., and A. W. B. Johnston. 1987. Sequence of psi, a gene on the symbiotic plasmid of Rhizobium phaseoli which inhibits exopolysaccharide synthesis and nodulation and demonstration that its transcription is inhibited by psr, another gene on the symbiotic plasmid. Mol. Gen. Genet. 207:149–154. Borthakur, D., C. E. Barber, J. W. Lamb, M. J. Daniels, J. A. Downie, and A. W. B. Johnston. 1986. A mutation that blocks exopolysaccharide synthesis prevents nodulation of peas by Rhizobium leguminosarum but not of beans by R. phaseoli and is corrected by cloned DNA from Rhizobium or the phytopathogen Xanthomonas. Mol. Gen. Genet. 203:320–323. Borthakur, D., J. A. Downie, A. W. B. Johnston, and J. W. Lamb. 1985. psi, a plasmid-linked Rhizobium phaseoli gene that inhibits exopolysaccharide production and which is required for symbiotic nitrogen-fixation. Mol. Gen. Genet. 200:278–282. Borthakur, D., J. W. Lamb, and A. W. B. Johnston. 1987. Identification of 2 classes of Rhizobium phaseoli genes required for melanin synthesis, one of which is required for nitrogen fixation and activates the transcription of the other. Mol. Gen. Genet. 207:155–160. Botero, L. M., T. S. Al Niemi, and T. R. McDermott. 2000. Characterization of two inducible phosphate transport systems in Rhizobium tropici. Appl. Environ. Microbiol. 66:15–22. Brink, B. A., J. Miller, R. W. Carlson, and K. D. Noel. 1990. Expression of Rhizobium leguminosarum CFN42 genes for lipopolysaccharide in strains derived from different R. leguminosarum soil isolates. J. Bacteriol. 172:548–555. Brom, S., A. G. Delossantos, T. Stepkowsky, M. Flores, G. Davila, D. Romero, and R. Palacios. 1992. Different plasmids of Rhizobium leguminosarum bv. phaseoli are required for optimal symbiotic performance. J. Bacteriol. 174:5183–5189. Brom, S., A. Garcia-de los Santos, L. Cervantes, R. Palacios, and D. Romero. 2000. In Rhizobium etli symbiotic plasmid transfer, nodulation competitivity and cellular growth require interaction among different replicons. Plasmid 44:34–43. Broughton, W. J., G. Hernández, M. Blair, S. Beebe, P. Gepts, and J. Vanderleyden. 2002. Beans (Phaseolus spp.)—model food legumes. Plant Soil 252:55–128. Buttery, B. R., and S. J. Park. 1990. Effect of nitrogen, inoculation and grafting on the expression of supernodulation in a mutant of Phaseolus vulgaris L. Can. J. Plant Sci. 70:375–381. Buttery, B. R., S. J. Park, and B. N. Dhauvantari. 1990. Effect of combined nitrogen, Rhizobium strain and substrate on a supernodulation mutant of Phaseolus vulgaris L. Can. J. Plant Sci. 70:935–963. Cárdenas, L., J. Domínguez, O. Santana, and C. Quinto. 1996. The role of the nodI and nodJ genes in the transport of Nod metabolites in Rhizobium etli. Gene 173:183–187. Carroll, B. J., D. L. McNeil, and P. M. Gresshoff. 1985. A supernodulation and nitratetolerant symbiotic (nts) soybean mutant. Plant Physiol. 78:34–40. Castillo, A., H. Taboada, A. Mendoza, B. Valderrama, S. Encarnación, and J. Mora. 2000. Role of GOGAT in carbon and nitrogen partitioning in Rhizobium etli. MicrobiologyUK 146:1627–1637. Catroux, G., A. Hartmann, and C. Revellin. 2001. Trends in rhizobial inoculant production and use. Plant Soil 230:21–30. Cava, J. R., P. M. Elias, D. A. Turowski, and K. D. Noel. 1989. Rhizobium leguminosarum CFN42 genetic regions encoding lipopolysaccharide structures essential for complete nodule development on bean-plants. J. Bacteriol. 171:8–15. Cava, J. R., H. Tao, and K. D. Noel. 1990. Mapping of complementation groups within a Rhizobium leguminosarum CFN42 chromosomal region required for lipopolysaccharide synthesis. Mol. Gen. Genet. 221:125–128. Cevallos, M. A., S. Encarnación, A. Leija, Y. Mora, and J. Mora. 1996. Genetic and physiological characterization of a Rhizobium etli mutant strain unable to synthesize polybeta-hydroxybutyrate. J. Bacteriol. 178:1646–1654.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

61

CIAT (Centro Internacional de Agricultura Tropical). 1998. Bean Program Annual Report. CIAT, Cali, Colombia. CIAT (Centro Internacional de Agricultura Tropical). 2000. Bean Program Annual Report. CIAT, Cali, Colombia. Ciereszko, I., A. Gniazdowska, M. Mikulska, and A. M. Rychter. 1996. Assimilate translocation in bean plants (Phaseolus vulgaris L.) during phosphate deficiency. J. Plant Physiol. 149:343–348. Ciereszko, I., A. Zambrzycka, and A. Rychter. 1998. Sucrose hydrolysis in bean roots (Phaseolus vulgaris L.) under phosphate deficiency. Plant Sci. 133:139–144. Clarke, L. M., M. J. Dilworth, and A. R. Glenn. 1993. Survival of Rhizobium meliloti WSM419 in laboratory culture—effect of combined pH shock and carbon substrate stress. Soil Biol. Biochem. 25:1289–1291. Corvera, A., D. Promé, J.-C. Promé, E. Martínez-Romero, and D. Romero. 1999. The nolL gene from Rhizobium etli determines nodulation efficiency by mediating the acetylation of the fucosyl residue in the nodulation factor. Mol. Plant-Microbe Interact. 12:236–246. Cregan, P. B., H. H. Keyser, and M. J. Sadowsky. 1989. Soybean genotype restricting nodulation of a previously unrestricted serocluster-123 Bradyrhizobia. Crop Sci. 29:307–312. D’Hooghe, I., J. Michiels, K. Vlassak, C. Verreth, F. Waelkens, and J. Vanderleyden. 1995. Structural and functional analysis of the fixLJ genes of Rhizobium leguminosarum biovar phaseoli CNPAF512. Mol. Gen. Genet. 249:117–126. D’Hooghe, I., C. VanderWauven, J. Michiels, C. Tricot, P. deWilde, J. Vanderleyden, and V. Stalon. 1997. The arginine deaminase pathway in Rhizobium etli: DNA sequence analysis and functional study of the arcABC genes. J. Bacteriol. 179:7403–7409. Daba, S., and M. Haile. 2002. Effects of rhizobial inoculant and nitrogen fertilizer on yield and nodulation of common bean under intercropped conditions. J. Plant Nutr. 25:1443– 1455. Dakora, F. D. 1998. Nodule function in symbiotic bambara groundnut (Vigna subterranea L.) and Kersting’s bean (Macrotyloma geocarpum L.) is tolerant of nitrate in the root medium. Ann. Bot. 82:687–690. Daniels, R., D. E. De Vos, J. Desair, G. Raedschelders, E. Luyten, V. Rosemeyer, C. Verreth, E. Schoeters, J. Vanderleyden, and J. Michiels. 2002. The cin quorum-sensing locus of Rhizobium etli CNPAF512 affects growth and symbiotic nitrogen fixation. J. Biol. Chem. 277:462–468. Danso, S. K. A., and G. D. Bowen. 1989. Methods of inoculation and how they influence nodulation pattern and nitrogen fixation using 2 contrasting strains of Bradyrhizobium japonicum. Soil Biol. Biochem. 21:1053–1058. Date, R. A. 2000. Inoculated legumes in cropping systems of the tropics. Field Crops Res. 65:123–136. Davis, E. O., and A. W. B. Johnston. 1990. Regulatory functions of the 3 nodD genes of Rhizobium leguminosarum biovar phaseoli. Mol. Microbiol. 4:933–941. Day, D. A., H. Lambers, J. Bateman, B. J. Carroll, and P. M. Gresshoff. 1986. Growth comparisons of a supernodulating soybean (Glycine max) mutant and its wildtype parent. Physiol. Plant. 68:375–382. Debouck D. G. 1994. Beans (Phaseolus spp.). p. 47–62. In: J. E. Hernández Bermejo and J. Léon, Neglected crops: 1492 from a different perspective. FAO, Rome. Delves, A. C., A. Mathews, D. A. Day, A. S. Carter, B. J. Carroll, and P. M. Gresshoff. 1986. Regulation of the soybean-Rhizobium nodule symbiosis by root and shoot factors. Plant Physiol. 82:588–590. Diebold, R., and K. D. Noel. 1989. Rhizobium leguminosarum exopolysaccharide mutants—biochemical and genetic analyses and symbiotic behavior on 3 hosts. J. Bacteriol. 171:4821–4830.

62

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Dillen, W., J. De Clercq, A. Goossens, M. Van Montagu, and G. Angenon. 1997. Agrobacterium-mediated transformation of Phaseolus acutifolius A. Gray. Theor. Appl. Genet. 94:151–158. Dombrecht, B. 2001. The complex regulation and role of the Rhizobium etli RpoN regulon in the symbiotic interaction with the common bean plant (Phaseolus vulgaris L.). Ph.D. thesis, K. U. Leuven, Belgium. Dombrecht, B., M. Z. Tesfay, C. Verreth, C. Heusdens, M. C. Nápoles, J. Vanderleyden, and J. Michiels. 2002. Rhizobium etli iscN is highly expressed in bacteroids and required for nitrogen fixation. Mol. Genet. Genom. 267:820–828. Duc, G., A. Trouvelot, V. Gianinazzi Pearson, and S. Gianinazzi. 1989. First report of nonmycorrhizal plant mutants (Myc–) obtained in pea (Pisum sativum L.) and Faba bean (Vicia faba L.). Plant Sci. 60:215–222. Duelli, D. M., A. Tobin, J. M. Box, V. S. Kolli, R. W. Carlson, and K. D. Noel. 2001. Genetic locus required for antigenic maturation of Rhizobium etli CE3 lipopolysaccharide. J. Bacteriol. 183:6054–6064. Eardly, B. D., F. S. Wang, T. S. Whittam, and R. K. Selander. 1995. Species limits in Rhizobium populations that nodulate the common bean (Phaseolus vulgaris). Appl. Environ. Microbiol. 61:507–512. Endre, G., A. Kereszt, Z. Kevei, S. Mihacea, P. Kaló, and G. B. Kiss. 2002. A receptor kinase gene regulating symbiotic nodule development. Nature 417:962–966. FAO (Food and Agricultural Organization). 2001. http://apps.fao.org/cgi-bin/nph-db.pl? subset=agriculture. Ferraioli, S., R. Taté, E. Caputo, A. Lamberti, A. Riccio, and E. J. Patriarca. 2001. The Rhizobium etli argC gene is essential for arginine biosynthesis and nodulation of Phaseolus vulgaris. Mol. Plant-Microbe Interact. 14:250–254. Ferraioli, S., R. Taté, M. Cermola, R. Favre, M. Iaccarino, and E. J. Patriarca. 2002. Auxotrophic mutant strains of Rhizobium etli reveal new nodule development phenotypes. Mol. Plant-Microbe Interact. 15:501–510. Fischer, H. M., M. Babst, T. Kaspar, G. Acuna, F. Arigoni, and H. Hennecke. 1993. One member of a groESL-like chaperonin multigene family in Bradyrhizobium japonicum is coregulated with symbiotic nitrogen-fixation genes. EMBO Journal 12:2901–2912. Folch-Mallol, J. L., S. Marroqui, C. Sousa, H. Manyani, I. M. Lopez-Lara, K. M. G. M. vanderDrift, J. Haverkamp, C. Quinto, A. GilSerrano, J. Thomas-Oates, H. P. Spaink, and M. Megias. 1996. Characterization of Rhizobium tropici CIAT899 nodulation factors: the role of nodH and nodPQ genes in their sulfation. Mol. Plant-Microbe Interact. 9:151–163. Freyre, R., P. W. Skroch, V. Geffroy, A. F. Adam-Blondon, A. Shirmohamadali, W. C. Johnson, V. Llaca, R. O. Nodari, P. A. Pereira, S. M. Tsai, J. Tohme, M. Dron, J. Nienhuis, C. E. Vallejos, and P. Gepts. 1998. Towards an integrated linkage map of common bean. 4. Development of a core linkage map and alignment of RFLP maps. Theor. Appl. Genet. 97:847–856. Garcia de los Santos, A., and S. Brom. 1997. Characterization of two plasmid-borne lps beta loci of Rhizobium etli required for lipopolysaccharide synthesis and for optimal interaction with plants. Mol. Plant-Microbe Interact. 10:891–902. Geffroy, V., M. Sévignac, J. C. F. de Oliveira, G. Fouilloux, P. Skroch, P. Thoquet, P. Gepts, T. Langin, and M. Dron. 2000. Inheritance of partial resistance against Colletotrichum lindemuthianum in Phaseolus vulgaris and co-localization of quantitative trait loci with genes involved in specific resistance. Mol. Plant-Microbe Interact. 13:287–296. Geffroy, V., D. Sicard, J. C. F. de Oliveira, M. Sévignac, S. Cohen, P. Gepts, C. Neema, T. Langin, and M. Dron. 1999. Identification of an ancestral resistance gene cluster

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

63

involved in the coevolution process between Phaseolus vulgaris and its fungal pathogen Colletotrichum lindemuthianum. Mol. Plant-Microbe Interact. 12:774–784. Gepts, P. 1998a. Origin and evolution of common bean: past events and recent trends. HortScience 33:1124–1130. Gepts, P. 1998b. What can molecular markers tell us about the process of domestication in common bean? p. 198–209. In: A. Damania et al. (eds.), The origins of agriculture and crop domestication. ICARDA, Aleppo, Syria. Gepts, P. 1999. Development of an integrated linkage map. p. 53–91. In: S. P. Singh (ed.), Common bean improvement in the twenty-first century. Kluwer Academic Publ., Dordrecht. Gepts, P., and D. Debouck. 1991. Origin, domestication, and evolution of the common bean (Phaseolus vulgaris L.). p. 7–53. In: A. van Schoonhoven and O. Voyest (eds.), Common beans: research for crop improvement. C.A.B. International, Wallingford, UK and CIAT, Colombia. 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. Girard, L., S. Brom, A. Davalos, O. Lopez, M. Soberón, and D. Romero. 2000. Differential regulation of fixN-reiterated genes in Rhizobium etli by a novel fixL-fixK-cascade. Mol. Plant-Microbe Interact. 13:1283–1292. Gniazdowska, A., M. Mikulska, and A. M. Rychter. 1998. Growth, nitrate uptake and respiration rate in bean roots under phosphate deficiency. Biologia Plantarum 41:217–226. Gonzalez-Pasayo, R., and E. Martínez-Romero. 2000. Multiresistance genes of Rhizobium etli CFN42. Mol. Plant-Microbe Interact. 13:572–577. Graham, P. H., K. J. Draeger, M. L. Ferrey, M. J. Conroy, B. E. Hammer, E. Martínez, S. R. Aarons, and C. Quinto. 1994. Acid pH tolerance in strains of Rhizobium and Bradyrhizobium, and initial studies on the basis for acid tolerance of Rhizobium tropici UMR1899. Can. J. Microbiol. 40:198–207. Gray, J. X., and B. G. Rolfe. 1990. Exopolysaccharide production in Rhizobium and its role in invasion. Mol. Microbiol. 4:1425–1431. Gresshoff, P. M. 1993. Molecular genetic analysis of nodulation genes in soybean. Plant Breed. Rev. 11:275–319. Gu, W. K., N. F. Weeden, D. H. Wallace, and S. Singh. 1993. A DNA marker for ppd, a gene conferring insensitivity to photoperiod in common bean. Annu. Rep. Bean Improv. Coop. 36:1–2. Gu, W., N. Weeden, J. Zhu, and D. Wallace. 1994. Identification of a DNA marker for hr, a gene that interacts with Ppd to confer extreme photoperiod sensitivity. Annu. Rep. Bean Improv. Coop. 37:125–126. Gutiérrez-Zamora, M. L., and E. Martínez-Romero. 2001. Natural endophytic association between Rhizobium etli and maize (Zea mays L.). J. Biotechnol. 91:117–126. Haley, S. D., L. Afanador, and J. D. Kelly. 1994. Identification and application of a random amplified polymorphic DNA marker for the I gene (potyvirus resistance) in common bean. Phytopathology 84:157–160. Haley, S. D., P. N. Miklas, J. R. Stavely, J. Byrum, and J. D. Kelly. 1993. Identification of RAPD markers linked to a major rust resistance gene block in common bean. Theor. Appl. Genet. 86:505–512. Hardarson, G., M. Golbs, and S. K. A. Danso. 1989. Nitrogen fixation in soybean (Glycine max L. Merrill) as affected by nodulation patterns. Soil Biol. Biochem. 21:783–787. Hardarson, G., F. A. Bliss, M. R. Cigalesrivero, R. A. Henson, J. A. Kipenolt, L. Longeri, A. Manrique, J. J. Penacabriales, P. A. A. Pereira, C. A. Sanabria, and S. M. Tsai. 1993.

64

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Genotypic variation in biological nitrogen-fixation by common bean. Plant Soil 152: 59–70. Hawkins, F. K. L., and A. W. B. Johnston. 1988. Transcription of a Rhizobium leguminosarum biovar phaseoli gene needed for melanin synthesis is activated by nifA of Rhizobium and Klebsiella pneumoniae. Mol. Microbiol. 2:331–337. Hawkins, F. K. L., C. Kennedy, and A. W. B. Johnston. 1991. A Rhizobium leguminosarum gene required for symbiotic nitrogen-fixation, melanin synthesis and normal growth on certain growth media. J. Gen. Microbiol. 137:1721–1728. Helal, H. M., H. Arisha, and E. Rietz. 1990. The uptake of trace-elements by spinach and bean varieties of different root parameters. Plant Soil 123:229–232. Hernández-Lucas, I., M. Pardo, L. Segovía, J. Miranda, and E. Martínez-Romero. 1995a. Rhizobium tropici chromosomal citrate synthase gene. Appl. Environ. Microbiol. 61:3992– 3997. Hernández-Lucas, I., L. Segovía, E. Martínez-Romero, and S. G. Pueppke. 1995b. Phylogenetic relationships and host range of Rhizobium spp. that nodulate Phaseolus vulgaris L. Appl. Environ. Microbiol. 61:2775–2779. Herrera-Cervera, J. A., J. Caballero-Mellado, G. Laguerre, H. V. Tichy, N. Requena, N. Amarger, E. Martínez-Romero, J. Olivares, and J. Sanjuan. 1999. At least five rhizobial species nodulate Phaseolus vulgaris in a Spanish soil. FEMS Microbiol. Ecol. 30:87–97. Herridge, D., and I. Rose. 2000. Breeding for enhanced nitrogen fixation in crop legumes. Field Crops Res. 65:229–248. Hirsch, P. R. 1979. Plasmid-determined bacteriocin production by Rhizobium leguminosarum. J. Gen. Microbiol. 113:219–228. Hungria, M., and G. Stacey. 1997. Molecular signals exchanged between host plants and rhizobia: basic aspects and potential application in agriculture. Soil Biol. Biochem. 29:819–830. Hungria, M., and M. A. T. Vargas. 2000. Environmental factors affecting N2 fixation in grain legumes in the tropics, with an emphasis on Brazil. Field Crops Res. 65:151–164. Hungria, M., D. D. Andrade, L. M. D. Chueire, A. Probanza, F. J. Guttierrez-Manero, and M. Megias. 2000. Isolation and characterization of new efficient and competitive bean (Phaseolus vulgaris L.) rhizobia from Brazil. Soil Biol. Biochem. 32:1515–1528. Hungria, M., A. A. Franco, and J. I. Sprent. 1993. New sources of high-temperature tolerant rhizobia for Phaseolus vulgaris L. Plant Soil 149:103–109. Islam, F. M. A., K. E. Basford, R. J. Redden, A. V. Gonzalez, P. M. Kroonenberg, and S. Beebe. 2002. Genetic variability in cultivated common bean beyond the two major gene pools. Gen. Res. Crop Evol. 49:271–283. Johnson, W., and P. Gepts. 2002. The role of epistasis in controlling seed yield and other agronomic traits in an Andean × Mesoamerican cross of common bean (Phaseolus vulgaris L.). Euphytica 125:69–79. Johnson, W. C., P. Guzman, D. Mandala, A. B. C. Mkandawire, S. Temple, R. L. Gilbertson, and P. Gepts. 1997. Molecular tagging of the bc-3 gene for introgression into Andean common bean. Crop Sci. 37:248–254. Jordan, D. C. 1984. Family III: Rhizobiaceae. p. 234–242. In: N. R. Krieg and J. G. Holt (eds.), Bergey’s manual of systematic bacteriology. Williams and Wilkins Co., Baltimore, MD. Josephson, K. L., D. P. Bourque, F. A. Bliss, and I. L. Pepper. 1991. Competitiveness of kim5 and viking-1 bean rhizobia–strain by cultivar interactions. Soil Biol. Biochem. 23:249–253. Jung, G. W., D. P. Coyne, P. W. Skroch, J. Nienhuis, E. Arnaud-Santana, J. Bokosi, H. M. Ariyarathne, J. R. Steadman, J. S. Beaver, and S. M. Kaeppler. 1996. Molecular markers associated with plant architecture and resistance to common blight, web blight, and rust in common beans. J. Am. Soc. Hort. Sci. 121:794–803.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

65

Jung, G., P. W. Skroch, D. P. Coyne, J. Nienhuis, E. Arnaud-Santana, H. M. Ariyarathne, S. M. Kaeppler, and M. J. Bassett. 1997. Molecular-marker-based genetic analysis of tepary bean-derived common bacterial blight resistance in different developmental stages of common bean. J. Am. Soc. Hort. Sci. 122:329–337. Kaplan, L., and T. Lynch. 1999. Phaseolus (Fabaceae) in archaeology: AMS radiocarbon dates and their significance for pre-Columbian agriculture. Econ. Bot. 53:261–272. Kelly, J., P. Gepts, P. Miklas, and D. Coyne. 2003. Tagging and mapping of genes and QTL and molecular marker-assisted selection for traits of economic importance in bean and cowpea. Field Crops Res. 82:135–154. Kingsley, M. T., and B. Bohlool. 1992. Extracellular polysaccharide is not responsible for aluminium tolerance of Rhizobium leguminosarum bv. phaseoli CIAT899. Appl. Environ. Microbiol. 58:1095–1101. Kipe-Nolt, J. A., C. M. Montealegre, and J. Tohme. 1992. Restriction of nodulation by the broad host range Rhizobium tropici strain CIAT899 in wild accessions of Phaseolus vulgaris L. New Phytol. 120:489–494. Koinange, E. M. K., S. P. Singh, and P. Gepts. 1996. Genetic control of the domestication syndrome in common-bean. Crop Sci. 36:1037–1045. Krishnan, H. B., and S. G. Pueppke. 1991. nolC, a Rhizobium fredii gene involved in cultivar-specific nodulation of soybean, shares homology with a heat-shock gene. Mol. Microbiol. 5:737–745. Labidi, M., S. Laberge, L. P. Vezina, and H. Antoun. 2000. The dnaJ (hsp40) locus in Rhizobium leguminosarum bv. phaseoli is required for the establishment of an effective symbiosis with Phaseolus vulgaris. Mol. Plant-Microbe Interact. 13:1271–1274. Laeremans, T., and J. Vanderleyden. 1998. Infection and nodulation signalling in RhizobiumPhaseolus vulgaris symbiosis. World J. Microbiol. Biotechnol. 14:787–808. Laeremans, T., I. Caluwaerts, C. Verreth, M. A. Rogel, J. Vanderleyden, and E. MartínezRomero. 1996. Isolation and characterization of the Rhizobium tropici Nod factor sulfation genes. Mol. Plant-Microbe Interact. 9:492–500. Laeremans, T., N. Coolsaet, C. Verreth, C. Snoeck, N. Hellings, J. Vanderleyden, and E. Martínez-Romero. 1997. Functional redundancy of genes for sulfate activation enzymes in Rhizobium sp. BR816. Microbiology 143:3933–3942. Laguerre, G., E. Geniaux, S. I. Mazurier, R. R. Casartelli, and N. Amarger. 1993. Conformity and diversity among field isolates of Rhizobium leguminosarum bv. viciae, bv. trifolii, and bv. phaseoli revealed by DNA hybridization using chromosome and plasmid probes. Can. J. Microbiol. 39:412–419. Latchford, J. W., D. Borthakur, and A. W. B. Johnston. 1991. The products of Rhizobium genes, psi and pss, which affect exopolysaccharide production, are associated with the bacterial cell surface. Mol. Microbiol. 5:2107–2114. Lerouge, I., T. Laeremans, C. Verreth, J. Vanderleyden, C. Van Soom, A. Tobin, and R. W. Carlson. 2001. Identification of an ATP-binding cassette transporter for export of the O-antigen across the inner membrane in Rhizobium etli based on the genetic, functional, and structural analysis of an lps mutant deficient in O-antigen. J. Biol. Chem. 276:17190–17198. Liao, H., G. Rubio, X. L. Yan, A. Q. Cao, K. M. Brown, and J. P. Lynch. 2001. Effect of phosphorus availability on basal root shallowness in common bean. Plant Soil 232:69–79. Luyten, E. 1999. Analysis of a Rhizobium sp. BR816 megaplasmid region that is involved in host range extension and symbiotic efficiency. Ph.D. thesis, K. U. Leuven, Belgium. Lynch, J. P., and S. E. Beebe. 1995. Adaptation of beans (Phaseolus vulgaris L.) to low phosphorus availability. HortScience 30:1165–1171. Lynch, J., and J. J. Vanbeem. 1993. Growth and architecture of seedling roots of common bean genotypes. Crop Sci. 33:1253–1257.

66

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Marroqui, S., A. Zorreguieta, C. Santamaria, F. Temprano, M. Soberón, M. Megias, and J. A. Downie. 2001. Enhanced symbiotic performance by Rhizobium tropici glycogen synthase mutants. J. Bacteriol. 183:854–864. Martínez, E., M. A. Pardo, R. Palacios, and M. A. Cevallos. 1985. Reiteration of nitrogen fixation gene sequences and specificity of Rhizobium in nodulation and nitrogenfixation in Phaseolus vulgaris. J. Gen. Microbiol. 131:1779–1786. Martínez-Romero, E., and J. Caballero-Mellado. 1996. Rhizobium phylogenies and bacterial genetic diversity. Crit. Rev. Plant Sci. 15:113–140. Martínez-Romero, E., L. Segovía, F. M. Mercante, A. A. Franco, P. Graham, and M. A. Pardo. 1991. Rhizobium tropici, a novel species nodulating Phaseolus vulgaris L. beans and Leucaena sp. trees. Intl. J. Syst. Bacteriol. 41:417–426. Martínez-Salazar, J. M., D. Romero, M. D. Girard, and G. Davila. 1991. Molecular cloning and characterization of the recA gene of Rhizobium phaseoli and construction of recA mutants. J. Bacteriol. 173:3035–3040. Mavingui, P., M. Flores, D. Romero, E. Martínez-Romero, and R. Palacios. 1997. Generation of Rhizobium strains with improved symbiotic properties by random DNA amplification (RDA). Nature Biotechnol. 15:564–569. Mejía-Jiménez, A., C. Munoz, H. J. Jacobsen, W. Roca, and S. P. Singh. 1994. Interspecific hybridization between common and tepary beans—increased hybrid embryo growth, fertility and efficiency of hybridization through recurrent and congruity backcrossing. Theor. Appl. Genet. 88:324–331. Melotto, M., L. Afanador, and J. D. Kelly. 1996. Development of a SCAR marker linked to the I gene in common bean. Genome 39:1216–1219. Melotto, M., L. Afanador, and J. Kelly. 1997. SCAR markers linked to major disease resistance genes in common bean. Bean Improv. Coop. Annual Meeting Abstr. Annapolis, MD. November 4–7. Mendoza, A., F. Hernandez, S. Hernandez, D. Ruiz, O. M. de la Vega, G. Mora, J. Acosta, and J. Simpson. 2001. Identification of Co-1 anthracnose resistance and linked molecular markers in common bean line A193. Plant Dis. 85:252–255. Michiels, J., I. Dhooghe, C. Verreth, H. Pelemans, and J. Vanderleyden. 1994b. Characterization of the Rhizobium leguminosarum biovar phaseoli nifA gene, a positive regulator of nif gene-expression. Arch. Microbiol. 161:404–408. Michiels, J., B. Dombrecht, N. Vermeiren, C. W. Xi, E. Luyten, and J. Vanderleyden. 1998a. Phaseolus vulgaris is a non-selective host for nodulation. FEMS Microbiol. Ecol. 26:193–205. Michiels, J., M. Moris, B. Dombrecht, C. Verreth, and J. Vanderleyden. 1998c. Differential regulation of Rhizobium etli rpoN2 gene expression during symbiosis and free-living growth. J. Bacteriol. 180:3620–3628. Michiels, J., H. Pelemans, K. Vlassak, C. Verreth, and J. Vanderleyden. 1995. Identification and characterization of a Rhizobium leguminosarum bv. phaseoli gene that is important for nodulation competitiveness and shows structural homology to a Rhizobium fredii host-inducible gene. Mol. Plant-Microbe Interact. 8:468–472. Michiels, J., T. Van Soom, I. D’Hooghe, B. Dombrecht, T. Benhassine, P. de Wilde, and J. Vanderleyden. 1998b. The Rhizobium etli rpoN locus: DNA sequence analysis and phenotypical characterization of rpoN, ptsN, and ptsA mutants. J. Bacteriol. 180:1729–1740. Michiels, J., C. Verreth, and J. Vanderleyden. 1994a. Effects of temperature stress on beannodulating Rhizobium strains. Appl. Environ. Microbiol. 60:1206–1212. Miklas, P. N., E. Johnson, V. Stone, J. S. Beaver, C. Montoya, and M. Zapata. 1996. Selective mapping of QTL conditioning disease resistance in common bean. Crop Sci. 36:1344–1351.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

67

Miklas, P. N., J. R. Stavely, and J. D. Kelly. 1993. Identification and potential use of a molecular marker for rust resistance in common bean. Theor. Appl. Genet. 85:745–749. Milner, J. L., R. S. Araujo, and J. Handelsman. 1992. Molecular and symbiotic characterization of exopolysaccharide-deficient mutants of Rhizobium tropici strain CIAT899. Mol. Microbiol. 6:3137–3147. Mimmack, M. L., D. Borthakur, M. A. Jones, J. A. Downie, and A. W. B. Johnston. 1994. The psi-operon of Rhizobium leguminosarum biovar phaseoli—identification of 2 genes whose products are located at the bacterial cell surface. Microbiology-UK 140:1223– 1229. Miranda-Ríos, J., C. Morera, H. Taboada, A. Davalos, S. Encarnación, J. Mora, and M. Soberón. 1997. Expression of thiamin biosynthetic genes (thiCOGE) and production of symbiotic terminal oxidase cbb3 in Rhizobium etli. J. Bacteriol. 179:6887–6893. Montealegre, C., and P. H. Graham. 1996. Preference in the nodulation of Phaseolus vulgaris cv. RAB39: Effect of delayed inoculation or low cell representation in the inoculant on nodule occupancy by Rhizobium tropici UMR1899. Can. J. Microbiol. 42:844–850. Montealegre, C., P. H. Graham, and J. A. Kipenolt. 1995. Preference in the nodulation of Phaseolus vulgaris cultivar RAB39. Can. J. Microbiol. 41:992–998. Moreno, S., R. Meza, J. Guzman, A. Carabez, and G. Espin. 1991. The glnA gene of Rhizobium leguminosarum bv. phaseoli and its role in symbiosis. Mol. Plant-Microbe Interact. 4:619–622. Moreno, S., E. J. Patriarca, M. Chiurazzi, R. Meza, R. Defez, A. Lamberti, A. Riccio, M. Iaccarino, and G. Espin. 1992. Phenotype of a Rhizobium leguminosarum ntrC mutant. Res. Microbiol. 143:161–171. Moulin, L., A. Munive, B. Dreyfus, and C. Boivin-Masson. 2001. Nodulation of legumes by members of the β-subclass of Proteobacteria. Nature 411:948–950. Nielsen, K. L., J. P. Lynch, A. G. Jablokow, and P. S. Curtis. 1994. Carbon cost of root systems—an architectural approach. Plant Soil 165:161–169. Nielsen, K. L., C. R. Miller, D. Beck, and J. P. Lynch. 1999. Fractal geometry of root systems: field observations of contrasting genotypes of common bean (Phaseolus vulgaris L.) grown under different phosphorus regimes. Plant Soil 206:181–190. Nodari, R. O., S. M. Tsai, R. L. Gilbertson, and P. Gepts. 1993. Towards an integrated linkage map of common bean: Development of an RFLP-based linkage map. Theor. Appl. Genet. 85:513–520. Noel, K. D., and D. M. Duelli. 2000. Rhizobium lipopolysaccharide and its role in symbiosis. p. 415–431. In: E. W. Triplett (ed.), Prokaryotic nitrogen fixation: a model system for analysis of a biological process. Horizon Scientific Press, Wymondham, U.K. Noel, K. D., D. M. Duelli, H. Tao, and N. J. Brewin. 1996. Antigenic change in the lipopolysaccharide of Rhizobium etli CFN42 induced by exudates of Phaseolus vulgaris. Mol. Plant-Microbe Interact. 9:180–186. Noel, K. D., L. S. Forsberg, and R. W. Carlson. 2000. Varying the abundance of O-antigen in Rhizobium etli and its effect on symbiosis with Phaseolus vulgaris. J. Bacteriol. 182:5317–5324. Noel, K. D., K. A. Vandenbosch, and B. Kulpaca. 1986. Mutations in Rhizobium phaseoli that lead to arrested development of infection threads. J. Bacteriol. 168:1392–1401. Nogales, J., R. Campos, H. BenAbdelkhalek, J. Olivares, C. Lluch, and J. Sanjuan. 2002. Rhizobium tropici genes involved in free-living salt tolerance are required for the establishment of efficient nitrogen-fixing symbiosis with Phaseolus vulgaris. Mol. PlantMicrobe Interact. 15:225–232. Ogawa, J., and S. R. Long. 1995. The Rhizobium meliloti groELC locus is required for regulation of early nod genes by the transcription activator nodD. Genes Developm. 9:714–729.

68

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Olaya, G. 1995. Genetics of resistance to Macrophomina phaseolina in beans and influence of water potential on the pathogen and on disease development. Ph.D. thesis, Cornell University, Ithaca, NY. Pardo, M. A., J. Lagunez, J. Miranda, and E. Martínez. 1994. Nodulating ability of Rhizobium tropici is conditioned by a plasmid-encoded citrate synthase. Mol. Microbiol. 11:315–321. Park, S. J., and B. R. Buttery. 1988. Nodulation mutants of white bean (Phaseolus vulgaris L.) induced by ethyl-methane sulfonate. Can. J. Plant Sci. 68:199–202. Patriarca, E. J., A. Riccio, R. Taté, S. Colonnaromano, M. Iaccarino, and R. Defez. 1993. The ntrBC genes of Rhizobium leguminosarum are part of a complex operon subject to negative regulation. Mol. Microbiol. 9:569–577. Pérez-Ramírez, N. O., M. A. Rogel, E. Wang, J. Z. Castellanos, and E. Martínez-Romero. 1998. Seeds of Phaseolus vulgaris bean carry Rhizobium etli. FEMS Microbiol. Ecol. 26:289–296. Piñero, D., E. Martínez, and R. K. Selander. 1988. Genetic diversity and relationships among isolates of Rhizobium leguminosarum biovar phaseoli. Appl. Environ. Microbiol. 54:2825–2832. Poole, P. S., N. A. Schofield, C. J. Reid, E. M. Drew, and D. L. Walshaw. 1994. Identification of chromosomal genes located downstream of dctD that affect the requirement for calcium and the lipopolysaccharide layer of Rhizobium leguminosarum. MicrobiologyUK 140:2797–2809. Raghothama, K. G. 2000. Phosphorus acquisition; plant in the driver’s seat! Trends Plant Sci. 5:412–413. Ramírez, M., B. Valderrama, R. Arredondo-Peter, M. Soberón, J. Mora, and G. Hernandez. 1999. Rhizobium etli genetically engineered for the heterologous expression of Vitreoscilla sp. hemoglobin: effects on free-living and symbiosis. Mol. Plant-Microbe Interact. 12:1008–1015. Ramos, M. L. G., and R. M. Boddey. 1987. Yield and nodulation of Phaseolus vulgaris and the competitivity of an introduced Rhizobium strain—effects of lime, mulch and repeated cropping. Soil Biol. Biochem. 19:171–177. Riccillo, P. M., M. M. Collavino, D. H. Grasso, R. England, F. J. de Bruijn, and O. M. Aguilar. 2000. A guaB mutant strain of Rhizobium tropici CIAT899 pleiotropically defective in thermal tolerance and symbiosis. Mol. Plant-Microbe Interact. 13:1228–1236. Robleto, E. A., K. Kmiecik, E. S. Oplinger, J. Nienhuis, and E. W. Triplett. 1998a. Trifolitoxin production increases nodulation competitiveness of Rhizobium etli CE3 under agricultural conditions. Appl. Environ. Microbiol. 64:2630–2633. Robleto, E. A., J. Borneman, and E. W. Triplett. 1998b. Effects of bacterial antibiotic production on rhizosphere microbial communities from a culture-independent perspective. Appl. Environ. Microbiol. 64:5020–5022. Robleto, E. A., A. J. Scupham, and E. W. Triplett. 1997. Trifolitoxin production in Rhizobium etli strain CE3 increases competitiveness for rhizosphere colonization and root nodulation of Phaseolus vulgaris in soil. Mol. Plant-Microbe Interact. 10:228–233. Rodriguez-Navarro, D. N., A. M. Buendia, M. Camacho, M. M. Lucas, and C. Santamaria. 2000. Characterization of Rhizobium spp. bean isolates from South-West Spain. Soil Biol. Biochem. 32:1601–1613. Romero, D., J. Martínez-Salazar, L. Girard, S. Brom, G. Davila, R. Palacios, M. Flores, and C. Rodriguez. 1995. Discrete amplifiable regions (amplicons) in the symbiotic plasmid of Rhizobium etli CFN42. J. Bacteriol. 177:973–980. Romero, D., P. W. Singleton, L. Segovía, E. Morett, B. Bohlool, R. Palacios, and G. Davila. 1988. Effect of naturally-occurring nif reiterations on symbiotic effectiveness in Rhizobium phaseoli. Appl. Environ. Microbiol. 54:848–850.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

69

Rosas, J. C., J. A. Castro, E. A. Robleto, and J. Handelsman. 1998. A method for screening Phaseolus vulgaris L. germplasm for preferential nodulation with a selected Rhizobium etli strain. Plant Soil 203:71–78. Rosemeyer, V., J. Michiels, C. Verreth, and J. Vanderleyden. 1998. luxI- and luxRhomologous genes of Rhizobium etli CNPAF512 contribute to synthesis of autoinducer molecules and nodulation of Phaseolus vulgaris. J. Bacteriol. 180:815–821. Rosenblueth, M., M. F. Hynes, and E. Martínez-Romero. 1998. Rhizobium tropici teu genes involved in specific uptake of Phaseolus vulgaris bean exudate compounds. Mol. Gen. Genet. 258:587–598. Russell, D. R., K. M. Wallace, J. H. Bathe, and B. J. Martinell. 1993. Stable transformation of Phaseolus vulgaris via electric-discharge mediated particle acceleration. Plant Cell Rep. 12:165–169. Ryan, P. R., E. Delhaize, and D. L. Jones. 2001. Function and mechanism of organic anion exudation from plant roots. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52: 527–560. Rychter, A. M., and D. D. Randall. 1994. The effect of phosphate deficiency on carbohydratemetabolism in bean roots. Physiol. Plant. 91:383–388. Schneider, K. A., M. E. Brothers, and J. D. Kelly. 1997. Marker-assisted selection to improve drought resistance in common bean. Crop Sci. 37:51–60. Schripsema, J., K. E. de Rudder, T. B. van Vliet, P. P. Lankhorst, E. de Vroom, J. W. Kijne, and A. A. van Brussel. 1996. Bacteriocin small of Rhizobium leguminosarum belongs to the class of N-acyl-L-homoserine lactone molecules, known as autoinducers and as quorum sensing co-transcription factors. J. Bacteriol. 178:366–371. Searle, I. R., A. E. Men, T. Laniya, D. M. Buzas, I. Iturbe-Ormaetxe, B. J. Caroll, and P. Gresshoff. 2002. Long-distance signaling in nodulation directed by a CLAVATA1-like receptor kinase. Science (in press). Segovía, L., J. P. Young, and E. Martínez-Romero. 1993. Reclassification of American Rhizobium leguminosarum biovar phaseoli type I strains as Rhizobium etli sp. nov. Intl. J. Syst. Bacteriol. 43:374–377. Shen, H., X. L. Yan, M. Zhao, S. L. Zheng, and X. R. Wang. 2002. Exudation of organic acids in common bean as related to mobilization of aluminum- and iron-bound phosphates. Environ. Expt. Bot. 48:1–9. Shirtliffe, S. J., and J. K. Vessey. 1996. A nodulation (Nod+/Fix–) mutant of Phaseolus vulgaris L. has nodule-like structures lacking peripheral vascular bundles (Pvb–) and is resistant to mycorrhizal infection (Myc–). Plant Sci. 118:209–220. Singh, S. P. 1999a. Common bean improvement in the twenty-first century. Kluwer Academic Publ., Dordrecht. Singh, S. P. 1999b. Production and utilization. p. 1–24. In: S. P. Singh (ed.), Common bean improvement in the twenty-first century. Kluwer Academic Publ., Dordrecht. Singh, S. P., J. A. Gutierrez, A. Molina, C. Urrea, and P. Gepts. 1991a. Genetic diversity in cultivated common bean: Marker-based analysis of morphological and agronomic traits. Crop Sci. 31:23–29. Singh, S. P., R. Nodari, and P. Gepts. 1991b. Genetic diversity in cultivated common bean allozymes. Crop Sci. 31:19–23. Singh, S. P., P. Gepts, and D. G. Debouck. 1991c. Races of common bean (Phaseolus vulgaris, Fabaceae). Econ. Bot. 45:379–396. Soberón, M., G. R. Aguilar, and F. Sanchez. 1993. Rhizobium phaseoli cytochrome C-deficient mutant induces empty nodules on Phaseolus vulgaris L. Mol. Microbiol. 8:159–166. Soberón, M., O. Lopez, C. Morera, M. L. Girard, M. L. Tabche, and J. Miranda. 1999. Enhanced nitrogen fixation in a Rhizobium etli ntrC mutant that overproduces the

70

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Bradyrhizobium japonicum symbiotic terminal oxidase cbb3. Appl. Environ. Microbiol. 65:2015–2019. Soberón, M., J. Membrillo-Hernandez, G. R. Aguilar, and F. Sanchez. 1990. Isolation of Rhizobium phaseoli Tn5-induced mutants with altered expression of cytochrome terminal oxidases o and aa3. J. Bacteriol. 172:1676–1680. Soberón, M., H. D. Williams, R. K. Poole, and E. Escamilla. 1989. Isolation of a Rhizobium phaseoli cytochrome mutant with enhanced respiration and symbiotic nitrogen fixation. J. Bacteriol. 171:465–472. Soberón-Chavez, G., and R. Najera. 1989. Isolation from soil of Rhizobium leguminosarum lacking symbiotic information. Can. J. Microbiol. 35:464–468. Soberón-Chavez, G., R. Najera, H. Olivera, and L. Segovía. 1986. Genetic rearrangements of a Rhizobium phaseoli symbiotic plasmid. J. Bacteriol. 167:487–491. Steele, H. L., B. Volker, G. F. Fuhrmann, and D. Werner. 1999. Strain specificities in Rhizobium tropici and Rhizobium etli using glucose transport inhibition by acidity and daidzein. Soil Biol. Biochem. 31:1059–1061. Stracke, S., C. Kistner, S. Yoshida, L. Mulder, S. Sato, T. Kaneko, S. Tabata, N. Sandal, J. Stougaard, K. Szczyglowski, and M. Parniske. 2002. A plant-receptor-like kinase required for both bacterial and fungal symbiosis. Nature 417:959–962. Tabche, M. L., E. G. Garcia, J. Miranda, J. E. Escamilla, and M. Soberón. 1998. Rhizobium etli cycHJKL gene locus involved in c-type cytochrome biogenesis: sequence analysis and characterization of two cycH mutants. Gene 208:215–219. Tadano, T., K. Ozawa, H. Sakai, M. Osaki, and H. Matsui. 1993. Secretion of acidphosphatase by the roots of crop plants under phosphorus-deficient conditions and some properties of the enzyme secreted by lupin roots. Plant Soil 156:95–98. Taté, R., A. Riccio, E. Caputo, M. Iaccarino, and E. J. Patriarca. 1999. The Rhizobium etli metZ gene is essential for methionine biosynthesis and nodulation of Phaseolus vulgaris. Mol. Plant-Microbe Interact. 12:24–34. Taté, R., A. Riccio, M. Iaccarino, and E. J. Patriarca. 1997. A cysG mutant strain of Rhizobium etli pleiotropically defective in sulfate and nitrate assimilation. J. Bacteriol. 179:7343–7350. Taté, R., A. Riccio, M. Merrick, and E. J. Patriarca. 1998. The Rhizobium etli amtB gene coding for an NH4+ transporter is down-regulated early during bacteroid differentiation. Mol. Plant-Microbe Interact. 11:188–198. Téllez-Sosa, J., M. Soberón, A. Vega-Segura, M. E. Torres-Marquez, and M. A. Cevallos. 2002. The Rhizobium etli cyaC product: characterization of a novel adenylate cyclase class. J. Bacteriol. 184:3560–3568. Thies, J. E., E. M. Holmes, and A. Vachot. 2001. Application of molecular techniques to studies in Rhizobium ecology: a review. Austral. J. Expt. Agr. 41:299–319. Thies, J. E., P. W. Singleton, and B. Bohlool. 1991. Influence of the size of indigenous rhizobial populations on establishment and symbiotic performance of introduced rhizobia on field-grown legumes. Appl. Environ. Microbiol. 57:19–28. Thorne, S. H., and H. D. Williams. 1999. Cell density-dependent starvation survival of Rhizobium leguminosarum bv. phaseoli: identification of the role of an N-acyl homoserine lactone in adaptation to stationary-phase survival. J. Bacteriol. 181:981–990. Tohme, J., D. O. González, S. Beebe, and M. C. Duque. 1996. AFLP analysis of gene pools of a wild bean core collection. Crop Sci. 36:1375–1384. Toro, O., J. Tohme, and D. G. Debouck. 1990. Wild bean (Phaseolus vulgaris L.): decription and distribution. International Board for Plant Genetic Resources (IBPGR) and Centro Internacional de Agricultura Tropical (CIAT). Cali, Colombia. Triplett, E. W., and T. M. Barta. 1987. Trifolitoxin production and nodulation are necessary for the expression of superior nodulation competitiveness by Rhizobium leguminosarum bv. trifolii strain T24 on clover. Plant Physiol. 85:335–342.

2. GENETIC IMPROVEMENT OF COMMON BEAN AND RHIZOBIA

71

Triplett, E. W., and M. J. Sadowsky. 1992. Genetics of competition for nodulation of legumes. Annu. Rev. Microbiol. 46:399–428. Tsai, S. M., R. O. Nodari, D. H. Moon, L. E. A. Camargo, R. Vencovsky, and P. Gepts. 1998. QTL mapping for nodule number and common bacterial blight in Phaseolus vulgaris L. Plant Soil 204:135–145. Urrea, C. A., P. N. Miklas, J. S. Beaver, and R. H. Riley. 1996. A codominant randomly amplified polymorphic DNA (RAPD) marker useful for indirect selection of bean golden mosaic virus resistance in common bean. J. Am. Soc. Hort. Sci. 121:1035–1039. Vadez, V., D. P. Beck, J. H. Lasso, and J. J. Drevon. 1997. Utilization of the acetylene reduction assay to screen for tolerance of symbiotic N-2 fixation to limiting P nutrition in common bean. Physiol. Plant. 99:227–232. Vallejos, C. E., N. S. Sakiyama, and C. D. Chase. 1992. A molecular marker-based linkage map of Phaseolus vulgaris L. Genetics 131:733–740. van Berkum, P., D. Beyene, and B. D. Eardly. 1996. Phylogenetic relationships among Rhizobium species nodulating the common bean (Phaseolus vulgaris L.). Intl. J. Syst. Bacteriol. 46:240–244. van Rhijn, P., J. Desair, K. Vlassak, and J. Vanderleyden. 1994a. Functional analysis of nodD genes of Rhizobium tropici CIAT899. Mol. Plant-Microbe Interact. 7:666–676. van Rhijn, P., J. Desair, K. Vlassak, and J. Vanderleyden. 1994b. The NodD proteins of Rhizobium sp. strain BR816 differ in their interactions with coinducers and in their activities for nodulation of different host plants. Appl. Environ. Microbiol. 60:3615–3623. Vasquez-Arroyo, J., A. Sessitsch, E. Martínez, and J. J. Pena-Cabriales. 1998. Nitrogen fixation and nodule occupancy by native strains of Rhizobium on different cultivars of common bean (Phaseolus vulgaris L.). Plant Soil 204:147–154. Vinuesa, P., B. L. Reuhs, C. Breton, and D. Werner. 1999. Identification of a plasmid borne locus in Rhizobium etli KIM5s involved in lipopolysaccharide O-chain biosynthesis and nodulation of Phaseolus vulgaris. J. Bacteriol. 181:5606–5614. Vlassak, K. M., E. Luyten, C. Verreth, P. van Rhijn, T. Bisseling, and J. Vanderleyden. 1998. The Rhizobium sp. BR816 nodO gene can function as a determinant for nodulation of Leucaena leucocephala, Phaseolus vulgaris and Trifolium repens by a diversity of Rhizobium spp. Mol. Plant-Microbe Interact. 5:383–392. Vlassak, K. M., and J. Vanderleyden. 1997. Factors influencing nodule occupancy by inoculant rhizobia. Crit. Rev. Plant Sci. 16:163–229. Vlassak, K. M., J. Vanderleyden, and A. Franco. 1996. Competition and persistence of Rhizobium tropici and Rhizobium etli in tropical soil during successive bean (Phaseolus vulgaris L) cultures. Biol. Fert. Soils 21:61–68. Wadisirisuk, P., S. K. A. Danso, G. Hardarson, and G. D. Bowen. 1989. Influence of Bradyrhizobium japonicum location and movement on nodulation and nitrogen fixation in soybeans. Appl. Environ. Microbiol. 55:1711–1716. Waelkens, F., T. Voets, K. Vlassak, J. Vanderleyden, and P. van Rhijn. 1995. The nodS gene of Rhizobium tropici strain CIAT899 is necessary for nodulation on Phaseolus vulgaris and on Leucaena leucocephala. Mol. Plant-Microbe Interact. 8:147–154. Wang, G. L, D. J. Mackill, J. M. Bonman, S. R. McCouch, M. C. Champoux, and R. J. Nelson. 1994. RFLP mapping of genes conferring complete or partial resistance to blast in a durably resistant rice cultivar. Genetics. 136:1421–1434. Wanke, M., I. Ciereszko, M. Podbielkowska, and A. M. Rychter. 1998. Response to phosphate deficiency in bean (Phaseolus vulgaris L.) roots. Respiratory metabolism, sugar localization and changes in ultrastructure of bean root cells. Ann. Bot. 82:809–819. Wolff, A. B., P. W. Singleton, M. Sidirelli, and B. B. Bohlool. 1993. Influence of acid soil on nodulation and interstrain competitiveness in relation to tannin concentrations in seeds and roots of Phaseolus vulgaris. Soil Biol. Biochem. 25:715–721.

72

C. SNOECK, J. VANDERLEYDEN, AND S. BEEBE

Wolff, A. B., W. Streit, J. A. Kipenholt, H. Vargas, and D. Werner. 1991. Competitiveness of Rhizobium leguminosarum bv. phaseoli strains in relation to environmental stress and plant defense mechanisms. Biol. Fert. Soils 12:170–176. Xi, C., G. Dirix, J. Hofkens, F. C. De Schryver, J. Vanderleyden, and J. Michiels. 2001. Use of dual marker transposons to identify new symbiosis genes in Rhizobium. Microb. Ecol. 41:325–332. Xi, C., E. Schoeters, J. Vanderleyden, and J. Michiels. 2000. Symbiosis-specific expression of Rhizobium etli casA encoding a secreted calmodulin related protein. Proc. Natl. Acad. Sci. (U.S.A.) 97:11114–11119. Yan, X. L., H. Liao, M. C. Trull, S. E. Beebe, and J. P. Lynch. 2001. Induction of a major leaf acid phosphatase does not confer adaptation to low phosphorus availability in common bean. Plant Physiol. 125:1901–1911. Yan, X. L., J. P. Lynch, and S. E. Beebe. 1995. Genetic variation for phosphorus efficiency of common bean in contrasting soil types: Vegetative response. Crop Sci. 35:1086–1093. You, Z. R., X. F. Gao, M. M. Ho, and D. Borthakur. 1998. A stomatin-like protein encoded by the sip gene of Rhizobium etli is required for nodulation competitiveness on the common bean. Microbiology-UK 144:2619–2627. Young, R. A., and J. D. Kelly. 1997. RAPD markers linked to three major anthracnose resistance genes in common bean. Crop Sci. 37:940–946.

3 Developing Marker-Assisted Selection Strategies for Breeding Hybrid Rice Yunbi Xu* RiceTec, Inc., P.O. Box 1305, Alvin, Texas 77512. Present address: Department of Plant Breeding, Cornell University Ithaca, New York 14853-1901 I. INTRODUCTION II. FEATURES OF HYBRID BREEDING A. Hybrid Prediction B. Selection for Hybrid Performance C. Seed Production and Commercialization D. Grain Production III. COMPONENTS OF MARKER-ASSISTED SELECTION A. Genetic Markers and Maps B. Marker Characterization C. Marker-Trait Associations D. Genotyping and High-throughput Genotyping Systems E. Data Management and Delivery IV. GERMPLASM EVALUATION A. Assessing Collection Redundancies and Gaps B. Monitoring Genetic Shifts C. Identifying Unique Germplasm D. Construction of Core Collection E. Germplasm Genotyping Database V. TRAITS REQUIRING TESTCROSSING OR PROGENY TESTING A. Fertility Restoration B. Outcrossing C. Wide Compatibility *The author gratefully acknowledges Drs. Susan R. McCouch, Barry Tillman, Mark Walton, Jules Janick, and three anonymous reviewers for their critical reading and invaluable comments and suggestions on the manuscript, and Dr. Junjian Ni for his help in collecting references. This article is dedicated to Emeritus Professor Zongtan Shen at Zhejiang University, Hangzhou, China, for his contributions to agriculture education, rice genetics, and plant breeding. Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 73

74

Y. XU

VI. ENVIRONMENT-DEPENDENT TRAITS A. Photoperiod/Temperature Sensitivity B. Environment-induced Genic Male Sterility C. Biotic and Abiotic Stresses VII. QUALITY TRAITS A. Seed Traits B. Hybrid Seed Traits C. Selection of Quality Traits VIII. GENE INTROGRESSION AND WHOLE GENOME SELECTION A. Gene Introgression B. Whole Genome Selection C. Selection for Multiple Genes/Traits D. Integrated Genetic Mapping and MAS E. Response to Selection IX. PREDICTION OF HYBRID PERFORMANCE AND HETEROSIS A. Combining Ability and Heterosis B. Genetic Basis of Heterosis C. Construction of Heterotic Groups D. Hybrid Prediction X. SEED QUALITY ASSURANCE A. Off-type B. False Hybrids XI. GENERAL DISCUSSIONS A. Economic Consideration B. Bioinformatics and Breeding Database C. Opportunities and Challenges LITERATURE CITED

LIST OF ABBREVIATIONS AFLP ASI BLUP BSSS CHA CMS DH DTF DTH EGMS FNP GCA GD IBPGR IHO

Amplified fragment length polymorphism Anthesis-silking interval Best linear unbiased prediction Iowa (B) Stiff Stalk Synthetic Chemical hybridization agent Cytoplasmic male sterility Doubled haploid Days-to-flowering Days-to-heading Environment-induced genic male sterility Functional nucleotide polymorphism General combining ability Genetic distance International Board for Plant Germplasm Resources Illinois high oil

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

IHP ILO ILP IRMI IRRI LSC MAS MPH NIE NIL NPGS OPC PGMS PIC PTS QTL QTN RAPD RFLP RIL RYD SAF SCA SCAR SNP SSR STS TD TG-BLUP TGMS

75

Illinois high protein Illinois low oil Illinois low protein International Rice Microsatellite Initiative International Rice Research Institute Lancaster Sure Crop Marker-assisted selection Mid-parent heterosis Near iso-environment Near-isogenic line National Plant Germplasm System (USA) Open pollinate cultivar Photoperiod-sensitive genic male sterility Polymorphic information content Photo-thermo sensitivity Quantitative trait locus/loci Quantitative trait nucleotide Random amplified polymorphic DNA Restriction fragment length polymorphism Recombinant inbred line Reid Yellow Dent Shared allele frequency Specific combining ability Sequence characterized amplified region Single nucleotide polymorphism Simple sequence repeat Sequence tagged site Transposon display Trait and gene best linear unbiased prediction Thermo-sensitive genic male sterility

I. INTRODUCTION Exploitation of heterosis or hybrid vigor to increase crop yields started early in the twentieth century with maize. From inbreeding a number of crop plants including maize, George H. Shull developed a perspective on heterosis that he outlined in a 1908 publication entitled “Composition of a Field of Maize.” He argued persuasively that by inbreeding isolated homozygous lines, these lines could be crossed to capitalize on heterosis. East (1908) discussed the danger of inbreeding. Commercially feasible F1 maize hybrids were developed following Jones’s 1918

76

Y. XU

proposition of the doubled cross. Average yields increased twofold after double crosses replaced the open-pollinate cultivars (OPCs) during 1935–1965 and increased another twofold after single crosses replaced the double crosses during 1966–1996 (Hallauer 1999). The T cytoplasmic male sterility (CMS) had been used to produce hybrid seeds during 1950–1970. Wide use of this single cytoplasm resource resulted in the epidemic of Southern leaf blight in 1970 caused by a new virulent Race T of Helminthosporium maydis, and an estimated 15% of the maize crop in the United States was destroyed. Hybrid maize is now based on hand detasseling, which avoids the use of CMS female parents. The success of hybrid maize is due to the ability of major seed companies such as Pioneer Hi-Bred International, Inc., to select for yield stability based on extensive testing and the efficiency of seed production that insures profitability. Now over 90% of maize in the United States is hybrid (Duvick 1999). Hybrid rice was first commercialized in the early 1970s in China. Unlike maize, rice is self-pollinated and hermaphroditic. Hybrid rice breeding has been based on using CMS or environment-induced genic male sterility (EGMS). A breeding system using three-lines (CMS line, CMS maintainer, and CMS restorer) was established by using a male sterile plant discovered from a wild rice species (Oryza rufipogon Griff. or O. f. spontanea) in 1970 by Prof. Longping Yuan and his assistant (Li and Yuan 2000). A two-line hybrid rice system using EGMS was established by using a photoperiod-sensitive genic male sterility (PGMS) mutant discovered in 1973 by Mingsong Shi from a japonica cultivar ‘Nongken 58’ (Shi 1981). Now over 50% of rice in China is hybrid (Li and Yuan 2000). Hybrid rice is also commercialized in other Asian countries and hybrid breeding programs have been established in several South American and North American countries. Hybrids provide many advantages in a crop production system. The principal benefit is increased yield. In open-pollinating species, one of the most often overlooked benefits is uniformity, an element which has allowed for the rapid expansion of production in many crop plants such as the vegetables. Additional benefits may include stress tolerance and pest resistance and other performance characteristics. Breeders of hybrid crops can react faster and with more options to meet changing markets, customer needs, and production demands. Other advantages of hybrids include the ability to combine useful dominant genes available in different inbred lines, to optimize the expression of genes in the heterozygous state, and to produce unique traits. Yield increases accounted for about 92% of increased cereal production in the developing world between 1961 and 1990. The yield advantage of hybrids ranged from 15% (maize) to 50% (sunflower), compared

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

77

to non-hybrids (Duvick 1999). However, yield growth rates are stagnating in some areas and, in a few cases, falling. A slowdown in the rate of yield increase of major cereals raises concern because increased yields are expected to be the source of increased food production in the future (Reeves et al. 1999). Utilization of heterosis and the improvement of the efficiency of hybrid breeding is one of the ways by which we can attempt to lift the yield ceiling. Producing F1 hybrids depends on the development of an efficient system for hybridization. There are several approaches to production of hybrids: CMS, EGMS including PGMS and thermo-sensitive genic male sterility (TGMS), protogyny, self incompatibility, chemical hybridization agent (CHA), and hand-emasculation systems. No matter which system is used, the central task is to develop and maintain two parents that have desirable traits. One serves as a pollen receptor and the other as a pollen donor for hybrid seed production. All desirable traits for hybrid seed production and heterosis itself, which are not required in pure line breeding, are difficult to measure, compared to other agronomic traits. Genomics in cereals, especially in rice, has created a substantial information base for their improvement. Marker-assisted selection (MAS), the first benefit that breeders can obtain from genomics, has been receiving a great deal of attention and will play an important role in hybrid breeding. The objective of this paper is to discuss the development of molecular marker strategies in breeding hybrid rice by drawing upon information obtained from maize and other crops. Information on the types of marker systems available and their relative merits is included, as is the use of markers in (1) evaluation and characterization of germplasm resources, (2) selection for different types of traits, (3) gene introgression, (4) prediction of hybrid performance, and (5) monitoring of seed quality in the seed production process. II. FEATURES OF HYBRID BREEDING A. Hybrid Prediction The process of plant breeding has developed through several key phases, including unconscious selection in Neolithic times, empirical art during the development and expansion of agriculture, and a predictive science-based approach practiced today. In general, prediction in plant breeding started from the evaluation of progeny performance. Pedigree methods and use of statistical tools in assessing progeny performance brought about a way for plant breeders to quantitatively separate the heritable portion of variation from the non-heritable and thus make parental choices based on heritability (Goldman 2000). As a result, exploitation

78

Y. XU

of additive genetic variance during the inbreeding process, and dominance variance during the testcrossing phase when assessing performance across multiple testers, has become a standard for cross-pollinated crops. There are several breeding practices that also make plant breeding more predictive. The breeding procedure known as backcrossing makes the progeny highly predictive. Identification of general and specific combining ability in maize helps identify superior inbred lines that can be used more efficiently in breeding new inbreds and hybrids. Wide area testing of hybrids makes it possible to develop widely adapted hybrid maize and improves the efficiency of maize breeding operations. Hybrid rice breeding has a very different story. Breeders of selfpollinating crops have been highly successful in breeding inbred cultivars. Although inbred breeding procedures were established and used for many years, hybrid rice breeding attempted a completely different approach. However, hybrid breeding procedures established for openpollinated crops were unsuitable for rice. Testcrossing and wide area testing, which is not a problem for hybrid maize breeding, has been a bottleneck for hybrid rice. Rice breeders have been searching for more reliable prediction methods in their breeding programs. High performance of hybrid plants results from the complementary action of both parents. Thus, parents with excellent performance per se may not produce desirable hybrids; superior hybrids may come from low-yielding parental lines. The evidence from maize indicates that inbred grain yield is not highly correlated with hybrid grain yield (Hallauer and Miranda 1988). Correlations between midparent values and hybrid means for the 91 crosses studied by Dudley et al. (1992) ranged from 0.46 for ear height to 0.71 for plant height. For grain yield, the correlation was 0.56. These correlations are too small to be of practical use in a breeding program. Despite their low values, the inbred-hybrid yield correlations were positive. They indicated a tendency for high-yielding inbreds to produce high-yielding hybrids. For a specific breeding program, however, the tendency does not make any kind of prediction meaningful. The second reason for the unpredictability is the lack of a full understanding of the genetic basis of heterosis, which affects all aspects of hybrid performance. As noted in the detailed discussion in Section IX, hybrid performance is positively related to the genetic distance between the parental lines when comparison is based on crosses derived from the intra-heterotic group. This “prediction” does not work, however, when comparison is based on the crosses derived from parental lines belonging to different heterotic groups, ecotypes, or subspecies. Although many hypotheses, such as physiological complementary, overdominance, dominance, and partial dominance, have been proposed as the

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

79

genetic explanation of heterosis, “isolated” research conclusions cannot be extended to other populations from the same varietal group, let alone the hybrids derived from very different crosses or species. As a result, limited and less-reliable information often provides contradictory results. As will be discussed later, prediction of hybrid performance depends on a thorough understanding of heterosis from different aspects, including genetics, physiology, and developmental biology. B. Selection for Hybrid Performance Hybrid performance depends on genes, and their interactions and combinations. Selection for hybrid performance in breeding programs is based on testcrossing and progeny testing. That is, we breed hybrids through selection of parental lines with desirable agronomic traits. To associate the parental phenotype with hybrid performance, breeders have to cross their candidate breeding lines with several testers, and from the hybrid progeny, determine if the candidates contain the genes required for hybrids and whether the parental combinations produce useful hybrids. This indirect selection, based on testcrossing and progeny testing, is time-consuming and very expensive. Furthermore, the association between the parental line and hybrid from one cross cannot be used to make a prediction about other associations. A cross of two extremely low-yielding inbreds can give a hybrid with good mid-parent or high-parent heterosis but poor performance, whereas a cross of two high-yielding inbreds might exhibit less mid-parent or high-parent heterosis but nevertheless produce a hybrid with good performance. High-yielding hybrids owe their yield not only to heterosis but also to other heritable factors that are not necessarily influenced by heterosis. For effective selection, one needs to know the relative importance of each genetic contribution—of heterosis and non-heterosis—in individual hybrids (Duvick 1999). Furthermore, when examining yield trends in a time series of successively released hybrids, breeders need to know what portion of the genetic yield gains (if gains are made) is due to increase in heterosis, and what portion to increases in nonheterosis (Duvick 1999). C. Seed Production and Commercialization In cultivar breeding, once an outstanding line is identified to be superior to the commercial control, it can be registered as a new commercial cultivar for production. In hybrid breeding, however, a superior hybrid requires not only a high yield potential and desirable agronomic traits such as disease resistance and good quality, but also an economically

80

Y. XU

viable seed production and maternal production system for each specific hybrid. Simply to exhibit good hybrid performance is insufficient. To be commercially viable, potential lines must produce excellent hybrids and have many traits desirable for seed production, such as high outcrossing rate and good flowering habits. This allows the cost of seed production to be low enough so that the seed producer benefits from the production and sale of hybrid seed. Requirements for activities such as roguing, detasseling/emasculation, pollination assistance, and other special management systems for seed production and parent multiplication can affect the cost and quality of the final product. This is very important to crops such as rice and wheat where reducing seed cost becomes critical for successful commercialization. Thus, a good seed production system will determine the cost of seed production and therefore the final commercialization of the hybrid. D. Grain Production In hybrids, commercial grain is produced from hybrid plants, but hybrid seed is produced from inbred parents. The agronomic practice used for the two production systems could be very different. For example, hybrid rice may have a very different nitrogen response than the parental lines and thus different kinds of nitrogen management are required, depending on the environments in which the hybrid is growing. Low potassium can be a problem for hybrid rice in many areas, although this element may be sufficient enough for inbred lines (Yuan and Chen 1988). Modern maize hybrids differ from open-pollinated and earlier hybrids primarily in their response to stress. New hybrids have improved water stress performance, are much less prone to silk delay, have significantly lower respiration rates during silking, have longer periods of grain fill, and are higher-yielding under both low- and high-input environments (Tomes 1998). Severe droughts in the American Cornbelt during 1934 and 1936 resulted in poor maize crops; however, hybrids often outperformed their open-pollinated counterparts under these conditions (Goldman 1999). This is due to the strong root system and lodging resistance of hybrids. These characteristics helped hybrid maize quickly replace OPCs. Contrary to maize, the first rice hybrids developed in China and the United States were tall and very prone to lodging if the same nitrogen rate as for inbred cultivars was used. As a result, a very different kind of nitrogen management is required for hybrids to maintain a good plant stand. In conclusion, hybrid breeding consists of all the components of inbred line breeding plus the additional complexity of recombining

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

81

inbreds. Plant breeders have continually looked to genetics for assistance (Duvick 1996), including quantitative genetics, cytogenetics, tissue culture, and mutation breeding. While utility has come from all new innovations, practical plant breeding based on hybridization and selection continues with little change in its basic structure. Now, a powerful new plant breeding technique, MAS, has appeared that promises to vastly increase selection efficiency. This review will explore the development and use of this MAS strategy in improving hybrid cereals, with an emphasis on rice.

III. COMPONENTS OF MARKER-ASSISTED SELECTION Key components that are required for an efficient MAS system include (1) suitable genetic markers and their characterization, (2) high-density molecular maps, (3) established marker-trait associations for traits of interest, (4) high-throughput genotyping systems, and (5) functional data analysis and delivery systems. A. Genetic Markers and Maps Genetic polymorphism at morphological, cytological, and molecular levels can be used as markers to tag traits, chromosomes, or DNA fragments. Genetic markers have been used for several decades in linkage analysis, gene mapping, gene transfer, and as aids to selection. Many different types of markers have been developed, including morphological variants, protein polymorphisms such as variation at isozyme loci, and DNA polymorphisms. DNA-based markers received the most attention since the first genetic map was established using restriction fragment length polymorphism (RFLP) (Botstein et al. 1980). While DNA markers have many advantages over other types of markers, morphological and protein polymorphisms are still very useful, especially when marker-trait associations have been well established or when desirable agronomic traits are used as markers. There are examples using morphological markers for heterosis-related applications. In this review, the term “markers” includes all types. DNA-based markers reflect genetic polymorphism at the DNA level, which result from any possible differences existing in nucleotides. Compared with other types of genetic markers, DNA markers have almost no practical limitation in numbers, often have no direct phenotypic effect, and are unaffected by environment. DNA markers can be classified into four different types based on the method used for polymorphism detection: (1) DNA-DNA hybridization such as RFLPs; (2) PCR-based markers

82

Y. XU

such as random amplified polymorphic DNA (RAPD) using arbitrary primers, or simple sequence repeat (SSR) and sequence tagged site (STS), both of which use specific primers; (3) combining use of PCR and restriction enzyme technique such as amplified fragment length polymorphism (AFLP) or transposon display (TD); and (4) single nucleotide differences that are referred to as single nucleotide polymorphisms (SNPs). Desirable DNA markers for MAS should meet the following requirements: detection of high frequency of polymorphism, codominance, abundance, whole genome coverage, high duplicability, suitability for high-throughput analysis and multiplexing, technical simplicity, cost effectiveness, requirement of small amount of DNA, and user-friendly (such as suitability for different genotyping systems and facilities). Among all these requirements, codominance is the most important for characterization of F1 hybrids. The two parental inbreds and hybrid combinations can be distinguished unambiguously. For all types of DNA markers mentioned above, SSR satisfies all the requirements. As estimated from a draft rice sequence, the density of SSRs in the genome is approximately one SSR per gene. These markers can be shared internationally through Internet-distributed primer sequences. SSR markers can be genotyped manually using agarose or polyarylamide gels and ethidium bromide or silver staining, or in highly automated facilities using ABI3700 Sequencers. SSR markers can be multiplexed doing PCR and for multiple-sample loading on gels using fluorescent labeling (Coburn et al. 2002). DNA extracted from a small piece of leaf or from single dry seeds will be enough to run several hundreds of markers (Xu et al. 2002). SSR is a more mature technique than SNP, although the latter has great potential for super high-throughput analysis using chip technology. As more and more genes are cloned, it will be possible to develop molecular markers based on sequences with the gene of interest. Intragenic markers provide several advantages over gene-linked markers. First, there is no recombination between the marker and the gene, or intergenic recombination. Second, multiple alleles can be tagged and distinguished. An example in rice is the SSR marker, RM190, derived from a microsatellite sequence with a splice site in the waxy gene, which is responsible for amylose synthesis, an important grain quality trait for hybrid rice. Selection of target chromosomal regions based on associated markers (foreground selection) and selection of genetic background for one of the parental genomes (background selection) may need different markers. Markers for foreground selection must be genetically mapped and associated with agronomic traits. Genetic markers revealing multiple bands or representing multiple loci are usually difficult to trace back to the specific allele/locus known to be associated with the trait, particularly when

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

83

the population used for MAS is different from the mapping population. These types of markers include RAPDs and AFLPs, which as such are not good for foreground selection. To use marker-trait associations based on these markers, it is best to convert them into types of markers that are more locus-specific, such as STS or SSR markers. For background selection, any type of markers that detect a high rate of polymorphism is useful. Background selection does not require the use of mapped markers as long as they can reveal genome-wide polymorphism. The efficiency of MAS largely depends on how well markers are linked to the target trait. Construction of a high-density genetic map using highthroughput molecular markers is the first step to a large-scale MAS program. A reference map is required for each crop based on a permanent segregating population that can be shared internationally such that it has the potential for continuous placement of additional markers by the entire community on the map later on. This map should be constructed using markers that are friendly to users. There are two reasons why we need a high-density molecular map. First, a minimum requirement for MAS based on marker-trait association includes a three-marker system: one marker cosegregating with the trait for foreground selection and the other two in each side of it for background selection. Since the target gene can be located in any region of the genome, a dense map is required for identifying this triplet at any position in the genome. Second, markers identified using mapping populations may not be polymorphic in breeding populations derived from other parental lines. To guarantee that the three-marker system will work for other breeding populations, many markers have to be identified around the target region. To saturate the rice SSR map, an international effort has been initiated through the International Rice Microsatellite Initiative (IRMI) (McCouch et al. 2002), with a goal of characterizing and mapping 2000 microsatellite markers developed from the Monsanto rice sequence database. As a member of the IRMI group, RiceTec, Inc. has undertaken a genetic mapping effort to place IRMI markers onto the existing SSR map, which was constructed using a subset of doubled haploid lines derived from an indica × japonica cross, ‘IR64’/‘Azucena’, and having a framework map consisting of 432 RFLP and SSR markers. A reference genetic map consisting of a thousand SSR markers is now available to rice molecular breeding programs. B. Marker Characterization It is not enough to just have thousands of genetic markers in hand. To use molecular markers efficiently, they have to be characterized for many features, including number of alleles; polymorphic information content

84

Y. XU

(PIC), allele sizes and ranges; allele sizes in standard or control cultivars; signal strength under specific genotyping conditions; background or noise signal; PCR conditions; chromosome location (flanking markers and genetic distances); and information required for multiplexing. Characterization of molecular markers helps to identify markers close to the genes of importance to breeding programs and to evaluate germplasm and breeding materials. A core set of SSR markers evenly distributed on all rice chromosomes and suited for multiplexing has been established at RiceTec, Inc., and is currently used for evaluation of germplasm accessions, construction of heterotic pools, and MAS in hybrid rice breeding. In this section, marker characterization will be illustrated using an example taken from a rice research project at Cornell University (Xu et al. 1997a,b; Xu et al. 2004). A total of 236 cultivars were collected to represent world genetic diversity, which included two subsets, 125 U.S. rice cultivars and 111 collected from 22 other countries. All germplasm accessions were genotyped using 100 RFLP and 60 SSR markers. The results related to marker characterization are summarized below. 1. Allele Number. The number of alleles at a marker locus is related to the genetic diversity that can be revealed by that marker. The more alleles at a locus, the higher the degree of diversity that can be revealed and the more efficiently closely related lines can be distinguished. A total of 274 alleles were detected at the 100 RFLP loci, with an average of 2.7 alleles per locus. SSR markers detected a total of 714 alleles, with an average of 11.9 alleles per locus. The world collection embodied 99.3% of RFLP and 95.8% of SSR alleles, while the U.S. collection embodied 82% of RFLPs and only 56% of SSR alleles. 2. PIC Value. The relative informativeness of each marker can be evaluated based on its PIC value, which reflects the amount of polymorphism and is a function of the number of alleles and allele frequencies at any given locus. The average PIC value was almost twice as high for SSRs (0.66) as for RFLPs (0.36). Average PIC values for the 100 RFLP markers were 0.40 for the world collection, and 0.21 for the U.S. collection. Average PIC values for the 60 SSR markers were 0.74 for the world collection, and 0.50 for the U.S. cultivars. 3. Informative Markers. Based on PIC values and the number of alleles detected, a set of highly informative markers can be selected such that the same amount of genotyping information can be obtained by surveying fewer molecular markers. Selected markers should be evenly distributed throughout the genome. Based on the 236 × 160 rice dataset, a group of 24 RFLP/SSR markers were selected to represent all 12 chromosomes and

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

85

are recommended as a set of highly informative markers for preliminary fingerprinting of rice germplasm and breeding populations. Molecular markers that are suited for MAS in hybrid breeding should have the same requirements as for other breeding projects. In addition to the requirements discussed in Section IIA for a marker system, there are some other requirements for a marker and a core set of markers. A useful marker should have many alleles per locus (>10 for SSRs), high PIC value (>0.8), suitable difference in allele sizes (4 to 10 bp between any two alleles for SSRs), strong signal for detection, less background or noise signal, and high replicability or reliability. A useful set of markers should provide whole genome coverage, even distribution on each chromosome, and high potentiality for multiplexing. C. Marker-Trait Associations Establishment of highly significant marker-trait associations is one of the prerequisites for MAS. Demonstrated linkages between target traits/genes and molecular markers are traditionally based on genetic mapping experiments, and it is important to confirm that these associations are consistent in mapping populations and breeding populations. For efficient MAS, marker(s) should co-segregate or be closely linked with the target trait, with a distance of 2 cM or less. Markers associated with major genes or quantitative trait loci (QTL) in one population may be used directly for MAS in other materials. For genes with relatively small effects, however, cross-population comparison of genes, alleles, and gene effects are required because of multilocus and multiallelic features that characterize most quantitative traits. To find tight marker-trait associations, a two-step process could be involved (especially for quantitative traits). The first step is based on a primary mapping population derived from very diverse parents, often with complicated genetic backgrounds. The second step is based on near isogenic lines that share a common genetic background and differ only at the target locus. There are many factors that are related to the detection of marker-trait associations and the efficiency of MAS. Marker-trait association or trait mapping has been discussed in detail elsewhere (Xu 1992; Tanksley 1993; Xu 1997, 2002; Liu 1998; Lynch and Walsh 1998; Paterson 1998; Flint and Mott 2001). 1. Genetic Backgrounds. Gene mapping requires the ability to extract a genetic signal from background “noise.” Sources of “noise” include variability in the external environments in which plant phenotypes are evaluated and variability due to differences in the internal genetic backgrounds of the individuals in a population. For accurate gene mapping,

86

Y. XU

the “noise” must be minimized or eliminated. “Controlled” environmental and/or genetic backgrounds are created to help filter the “noise.” Creation of homogeneous genetic backgrounds will help define markertrait associations. Xu (2002) described five approaches for creation of homogeneous or isogenic backgrounds: backcross-derived near-isogenic lines (NILs), selfing-derived NILs, whole genome selection of permanent populations, mutation, and chromosome substitution. Genetic materials, such as NILs with homogeneous backgrounds, have been used in many different investigations. If NILs are used, interaction between the target gene and other major genes/QTL can be eliminated and only epistasis between multiple target genetic loci needs to be considered. With removal of noise from heterogeneous backgrounds, the proportion of variance explained by the target loci will increase and minor genes (genes with smaller effects) can be identified. By minimizing the disturbance from the genetic background, multiple loci in a single chromosomal region can be separated and their effect on the phenotype can be partitioned. When all genotypic variation comes from the target loci, environmental effects can be estimated. Heterogeneous genetic backgrounds can also come from populations with different structures, such as F2, doubled haploids (DH), and recombinant inbred lines (RILs), but derived from the same cross, or come from various crosses derived from different cultivars, subspecies, species, and families. Genetic materials with heterogeneous genetic backgrounds can be used to estimate epistasis, detect non-allelic genes, discover multiple alleles, and identify paralogous and orthologous genes. As a contribution to complicated genetic backgrounds, many quantitative traits per se are a complex consisting of several components or subtraits. For example, polygenic sterility in rice can be partitioned into several components, including male and female sterility, or ovary and pollen abortion, so that polygenes can be divided into several components with different functions and, thus, can be handled more easily (Shen and Xu 1992). Genetic backgrounds in a population can also be complicated by the contribution of other related traits. 2. Alleles at Multiple Loci. When multiple loci control a trait, their alleles of positive or negative effect (increasing or decreasing trait value) tend to be dispersed between parents, each with positive alleles at one or some loci but negative alleles at others (Xu and Shen 1992). These dispersed alleles can be cryptic transgressive, which has been found in the parents with similar phenotypes (Xu et al. 1998). In genetic mapping, phenotypic difference between parents is unnecessary for detection of QTL. In many QTL studies, mapping populations were developed without consideration of phenotypic differences between the two parents and

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

87

QTL mapped without a statistical test of the parental difference. In most cases where no parental difference is found, QTL are still detected, which could be due to the complementary distributions of positive and negative allelic effects in the parents. As observed in QTL mapping, on average, about four QTL are identified for each trait in rice (Xu 2002), the same as the average obtained for 176 trial-trait combinations as reviewed by Kearsey and Farquhar (1998). When QTL identified for the same trait are summarized over different projects/populations, this number becomes much larger. For example, plant height has been mapped using 13 populations with 63 QTL reported. Some of the QTL are allelic to each other, that is, they were mapped to the same chromosomal region or intervals of less than 15 cM. After elimination of possible allelic QTL, the total number of QTL for plant height is reduced to 29, with up to five QTL existing on one chromosome (Xu 2002). In contrast, over 50 independent single gene mutations have been identified so far for plant height in rice, as summarized by Kinoshita (1998). Some plant height QTL were co-located with major plant height loci, suggesting that the gene controlling quantitative variation may be the same as those associated with macromutations. This has been demonstrated based on high-resolution mapping and cloning of a QTL for plant height in rice (E. M. Septiningsih and S. R. McCouch at Cornell, pers. commun.). To date, QTL alleles that have been cloned in rice all correspond to previously identified single gene mutants. For example, a photoperiod sensitivity QTL, Hd1, is allelic to the major gene, Se1 (Yano et al. 2000). QTL allelism tests and determination of major-gene and QTL correspondences are facilitated by the availability of high-density molecular maps with a common set of markers shared among researchers. 3. Multiple Alleles at a Locus. Two-parent derived populations in diploid crops have only two alleles segregating at each locus. Identification of multiple alleles requires comparison of populations derived from different crosses. To distinguish alleles identified in one cross from those in another, all alleles must be accurately sized and documented. Rice amylose content, mainly controlled by the wx gene, is a good example for multiple alleles at a locus. A polymorphic microsatellite was identified in the wx gene (Bligh et al. 1995) located 55 bp upstream of the putative 5’-leader intron splice site. A total of 16 wx microsatellite alleles were identified in worldwide rice germplasm (Ayres et al. 1997; Zeng et al. 2000). Now the question is whether the multiple alleles identified at the waxy locus can be associated with specific quantitative effects, including developing an understanding of how each allele interacts with other genes/alleles in the genetic background

88

Y. XU

and in response to environment. Using the best-characterized examples, such as wx, the challenge will be to extend this kind of analysis to other traits or genetic loci. Using gene-based molecular markers that have multiple alleles in gene mapping could help identify multiple alleles at a locus. Genetic mapping studies using different populations have identified some common major genes and QTL. It is necessary, however, to further clarify whether common or different functional alleles were identified at those loci. Reporting the sizes of associated (closely linked) alleles and using allele-rich markers in marker-trait association studies will provide a baseline of information required for this clarification, with the assumption that each marker has a corresponding allele at the trait locus. There are many reasons why close marker-trait associations are required: (1) chromosomal location associated with the trait must be reduced to a manageable piece of DNA if cloning of specific genes is necessary; (2) to identify all the related genes for a specific trait, a highdensity genetic map is required because the fewer markers are used, the smaller proportion of genetic factors contributing to that trait will be sampled; (3) large genetic distances between markers and target traits will contribute to the rapid decrease of MAS efficiency after several successive cycles of selection; and (4) to minimize linkage drag involved in gene introgression, closely linked markers around the target region are needed. QTL mapping presumes accurate phenotypic scoring methods, something that can be difficult to optimize and even more difficult to keep consistent for months or years. Just a few misscored individuals can totally confound QTL discovery and placement (Young 1999). This is also true for fine mapping of major genes for map-based cloning, where misscoring of several plants in a population with thousands of individuals will result in a large error (up to one cM) in estimating genetic distances. High levels of accuracy are required to dissect a chromosomal region associated with a given trait and narrow down the candidate region to a single contig, that is, a set of clones that can be assembled into a linear order. D. Genotyping and High-throughput Genotyping Systems To make marker-based technology practical for breeding applications, an automated genotyping system is required. Such an automated system using SSRs to genotype rice germplasm and breeding populations was developed, through improved DNA extraction and loading with multiple-PCR products. This system brings the cost per data point down to as low as $0.30 with daily data output up to 4608 data points per ABI

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

89

3100 Sequencer (Xu et al. 2002). This level of efficiency makes it possible to genotype thousands of individual plants with a panel of eight SSR markers in a week using two sequencers. SNPs have gained wide acceptance as genetic markers for use in linkage and association studies, especially for human genetics. Highthroughput SNP genotyping has great potential for many applications, including MAS on the basis of whole genome approaches. This has led to a requirement for high-throughput SNP genotyping platforms. Development of such a platform depends on coupling reliable chemical assays with an appropriate detection system to maximize efficiency with respect to accuracy, speed, and cost. Current technology platforms are able to deliver throughputs in excess of 100,000 genotypes per day, with an accuracy of >99%, at a cost of 20–30 cents per genotype (Jenkins and Gibson 2002). In order to meet the demands of the coming years, however, genotyping platforms need to deliver throughputs in the order of one million genotypes per day at a cost of only a few cents per genotype. In addition, DNA template requirements must be minimized such that hundreds of thousands of SNPs can be interrogated using a relatively small amount of genomic DNA. Jenkins and Gibson (2002) predicted that the next generation of high-throughput genotyping platforms would exploit large-scale multiplex reactions and solid phase assay detection systems. Released genomic sequences of rice and Arabidopsis can be used to develop gene-based SNPs for other related species. E. Data Management and Delivery To handle the daily data flow from the lab to the breeder and integrate information from molecular markers, genetic mapping, and phenotyping, many informatics tools are needed. For efficient data management and delivery, it is important for all researchers to follow general rules through all these procedures. A standard reporting system is also critical for comparative genomics, QTL allelism tests, data sharing and mining, and the association between major genes and QTL. As discussed by Xu (2002), a standard system for marker-trait association should include associated alleles and allele characterization such as allele sizes, gene effects, variation explained by each gene or all genes in the model, gene interaction if more than one gene is identified, and genotype × environment interaction if more than one environment is involved. Genetic information should be shared and combined with data generated in plant breeding, for example, germplasm diversity, mapping populations, pedigrees, graphical genotypes, mutants, and other genetic stocks. With several thousand data points flowing out of the laboratory every day, timely scoring and delivery of the results to breeders are basic

90

Y. XU

requirements for a high-efficiency breeding system. Well-trained assistants for genotyping and scoring, coupled with research scientists who can analyze data in meaningful ways, are the key components for a data management and delivery system. A laboratory with well-equipped facilities has to be also well equipped with qualified personnel and software required for data integration, manipulation, analysis, and mining. Timely delivery of data to the breeder is also equally important, because in many cases the time window the breeder can use for selection is very limited. With the high-throughput genotyping and data management systems currently available, it takes about a week to generate and analyze data for a breeding-related population consisting of several hundred individuals. This includes activities ranging from leaf tissue harvesting to DNA extraction, genotyping, data scoring, analyzing, summarizing, and reporting.

IV. GERMPLASM EVALUATION Germplasm resources represent the genetic variability required for continuous improvement of crop plants. The old paradigm for evaluation and utilization of germplasm involves looking for a clearly defined character by screening entries from a genebank. This approach works well when the trait of interest is controlled by one or few genes. For traits such as yield, genetically controlled by many genes, it is impossible to identify all these genes phenotypically because each gene has a relatively small but similar effect. As a result, exotic germplasm, which is perceived to be a poor bet for the improvement of most traits based on phenotypic examination, may contain some favorable genes (alleles) that lie buried amidst the thousands of accessions maintained in genebanks (de Vicente and Tanksley 1993; Tanksley and McCouch 1997; Xiao et al. 1998). The new paradigm involves looking for genes using molecular markers and/or the integrative power of QTL analysis, which can be used to extract superior genes (alleles) from the inferior germplasm accessions. Molecular marker-assisted germplasm evaluation aims to complement phenotypic evaluation by helping define the genetic architecture of germplasm resources and by identifying alleles that are associated with key phenotypic traits. Molecular markers may allow for characterization based on gene, genotype, and genome, which provide more accurate and detailed information than classical phenotypic or passport data. Many features revealed by molecular markers, such as unique alleles, allele frequency, and heterozygosity at marker loci, mirror the genetic loci for heterosis and the traits of agronomic importance. On a more fundamental level, molecular marker information may lead

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

91

to the identification of useful genes contained in collections and transferal of those genes into well-adapted cultivars. Bretting and Widrlechner (1995) comprehensively reviewed genetic markers and their application in plant genetic resource management, including procedures related to acquisition/distribution, maintenance, and utilization. In this section, discussion will be focused on the aspects more related to breeding applications. A. Assessing Collection Redundancies and Gaps With a large number of germplasm accessions available for each cultivated plant, it is likely that many represent duplicate or nearly identical samples of the same cultivar, while others embody rare alleles or highly unusual allele combinations, with many genes or alleles still missing in current collections. According to the International Board for Plant Genetic Resources (IBPGR), over 3.6 million germplasm accessions for different crop species are conserved at international and national genebanks, which include 90,000 for rice, 120,000 for wheat, and 25,000 for maize (Iwanaga 1993). Evaluation of genetic diversity will help in the understanding of genetic structure of existing collections and design acquisition strategies. In particular, calculation of genetic distance (GD) can be used to identify particularly divergent subpopulations that might harbor valuable genetic variation that is underrepresented in current holdings. Redundancy germplasm accessions exist in many germplasm collections because of different names for the same cultivars or duplicate samplings of the same accessions. Pedigree-related cultivars, siblines, and NILs may represent another type of redundancy because they are genotypically duplicated at most of the genetic loci. For example, U.S. rice cultivars ‘M5’, ‘M301’, ‘M103’, ‘S201’, ‘Calrose’, ‘Calrose 76’, ‘CS-M3’, and ‘Calmochi-202’ shared the same panel of alleles at all 100 RFLP loci tested by Xu et al. (2004). All these cultivars can be traced back to a common ancestor, ‘Caloro’. No genetic polymorphism could be detected either at any of the 60 SSR loci between ‘Calrose’ and ‘Calrose 76’ (Xu et al. 2004), because they are isolines, with ‘Calrose 76’ representing a variant derived from ‘Calrose’ via chemical mutagenesis (Rutger et al. 1977). Using 15 SSR markers, Dean et al. (1999) assayed 19 sorghum [Sorghum bicolor (L.) Moench] accessions identified as “Orange” presently maintained by the U.S. National Plant Germplasm System (NPGS). They found most accessions are genetically distinct, but two redundant groups were found. The variance analysis also indicated that it should be possible to reduce the number of Orange accessions held by NPGS by almost half without seriously jeopardizing the overall amount

92

Y. XU

of genetic variation contained in these holdings. Chavarriaga-Aquirre et al. (1999) evaluated genetic diversity and redundancy in a cassava core collection. The core collection (630 accessions) was selected from the base collection (over 5500 accessions) on the basis of diversity of origin (country and geographic), morphology, isozyme patterns, and specific agronomic criteria. A small number (1.34%) of potential duplicates were identified from the core collection through isozyme and AFLP profiles. Different germplasm collections can be compared for the frequencies of alleles at all marker loci so that distinctive alleles, allele combinations, and allele frequency patterns can be identified. Chromosomal regions containing loci that show the greatest changes in allele frequency between the collections can be located. The rationale for this analysis is to define the genomic regions where selection under the environment had given rise to allele combinations or allele frequency patterns that distinguished a group of accessions with less diversity from more diverse accession groups. In the rice example (Xu et al. 2004), alleles at two RFLP loci (a 6.5 kb allele at CDO686/HindIII and a 6.0 kb allele at BCD808/XbaI) and alleles at six SSR loci were represented at frequencies of 17.1% to 33.6% in the world collection, but had been completely lost in the U.S. cultivars. When low-frequency or underrepresented alleles are defined as those that occur in four or fewer U.S. cultivars but are very frequent (i.e., >17% for SSR and >30% for RFLP) in the world collection, alleles underrepresented in the U.S. collection were found at 19 RFLP and 18 SSR loci. Three U.S. cultivars, ‘Della’, ‘Rexmont’, and ‘Caloro’, retained 34 of the 37 low-frequency alleles. ‘Della’ alone retained 24 (64.9%) of them, which could be traced back to two of its ancestral cultivars, ‘Rexoro’ and ‘Delitus’. The U.S. rice cultivars that were developed from a small set of germplasm introductions help explain why these cultivars retained an unusually large number of alleles that show a decline in frequency among later developed U.S. cultivars. B. Monitoring Genetic Shifts Maintaining genetic diversity and preventing genetic drifts is one of the most important objectives for germplasm conservation. In openpollinated species, deviations from random mating, primarily in the form of assortative or consanguineous matings, needs to be monitored during germplasm regeneration. In maize, deviations from random mating were widely studied with emphasis on detailed multilocus isozyme analyses of one or two synthetic or open-pollinated maize cultivars (Kahler et al. 1984; Pollak et al. 1984; Bijlsma et al. 1986). In general, levels of selfing did not exceed those expected under random-mating mod-

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

93

els, but significant deviations were caused by temporal variation in the pollen pool or by gametophytic selection. The genetic profiles of germplasm accessions can change during the course of medium- or long-term storage. Storage effects fall into three broad categories: (1) the occurrence of mutations, (2) the occurrence of chromosomal aberrations, and (3) shifts in gene frequencies resulting from differential genotypic viability in heterogeneous populations (Roos 1988). After a comprehensive review of storage effects on seeds, Roos (1988) found little evidence for heritable changes in germplasm attributable to storage-induced chromosomal aberrations, and noted “little need for concern about mutation as a significant factor in altering the composition of germplasm collection.” However, differential seed longevity can markedly reduce genetic variability over time (Bretting and Widrlechner 1995). This is well documented by experiments involving mixtures of eight bean lines (Roos 1984) and four seed storage protein genotypes within a cultivar of wheat (Stoyanova 1991). Genetic shifts can be caused by in vitro culture. The genetic stability of germplasm maintained in tissue culture (in vitro) has generally been monitored with karyotypic markers such as chromosome number and morphology (D’Amato 1975), because cytological variability has been considered a primary cause of somaclonal variation. Lassner and Orton (1983) reported that isozymatically identical in vitro cultures of celery were markedly variable cytologically. This finding should reinforce the concept that the genetic stability of in vitro cultures should be monitored with a battery of different genetic markers, particularly those DNA markers that collectively span the whole genome (Bretting and Widrlechner 1995). A certain level of heterogeneity could exist in germplasm accessions that are mainly self-pollinated, which provides a buffer for maintaining genetic diversity and preventing genetic drifts. Monitoring heterogeneous accessions will help develop strategies for regeneration of germplasm samples without loss of the allelic diversity provided by heterogeneity. In general, traditional cultivars had a higher level of heterozygosity, as reported in rice by Olufowote et al. (1997). Genetic diversity resulting from the heterozygosity was also found within inbred lines from different sources in rice (Olufowote et al. 1997) and maize (Gethi et al. 2002). As reported by Xu et al. (2004), a total of 120 (50.6%) of the 236 rice accessions was found heterozygous at one or more RFLP or SSR loci, and the number of heterozygous loci detected in a single rice accession ranged from 0 to 39 (25.3% of the 160 loci). These heterozygous allele patterns could indicate either seed mixtures or true heterozygosity remaining in these cultivars despite the fact that all accessions had been purified.

94

Y. XU

C. Identifying Unique Germplasm Progress in hybrid breeding is demonstrated by the development of new parental lines and hybrids that are superior for one or more characteristics once a commercialized hybrid system has already been established. Progress depends on (1) discovery and generation of genetic variation for heterosis and agronomic traits and (2) accurate selection of rare genotypes that possess new or improved attributes due to superior combinations of alleles at multiple loci. Over the past century, the development and successful application of modern breeding methodologies has produced the high-yielding cultivars and hybrids on which modern farming is based. As the demand for uniform performance and grain quality has increased, new cultivars and hybrids are increasingly derived from adapted, genetically related, and elite modern cultivars/hybrids, while genetically more variable, but less productive, primitive ancestors are excluded from most breeding programs (Tanksley and McCouch 1997). In a study of pedigree relationships among 140 U.S. rice accessions, Dilday (1990) concluded that all parental germplasm in public cultivars used in the southern United States today could be traced back to 22 plant introductions in the early 1900s, and those used in California could be traced to 23 introductions. The same situation is true for soybean and wheat. Virtually all modern U.S. soybean cultivars can be traced back to a dozen strains from a small area in northeastern China, and the majority of hard red winter wheat cultivars in the United States originated from just two lines imported from Poland and Russia (Tanksley and McCouch 1997). To broaden the genetic base of specific cultivated species, the genetic diversity within collections must be assessed in the context of the total available genetic diversity for each species. With the use of DNA profiles, the genetic uniqueness of each accession in a germplasm collection or in a population can be determined, and the identity and frequency of individual alleles can be clearly described and characterized (Brown and Kresovich 1996; Smith and Helentjaris 1996). The sampling of exotic germplasm should emphasize the genetic composition rather than the appearance of exotic accessions. Accessions with DNA profiles most distinct from that of modern germplasm are likely to contain the greatest number of novel alleles. Assuming that most marker alleles having the same molecular weight are likely to be common by descent in a specific varietal group, we are able to trace alleles that are frequent in one specific collection but existed in low frequency or not at all in the other. Examination of the chromosomal distribution of the loci harboring underrepresented alleles indicated that underrepresented alleles in the U.S. rice cultivars as discussed in previous sections tended to cluster on 11 chromosomal fragments (Xu

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

95

et al. 2004). This raises the question about what genes are located in these regions and whether U.S. breeders have consciously or unconsciously narrowed the range of genetic variation in these regions. It also suggests that molecular marker analysis could be used to identify parents harboring rare or novel alleles in these regions so that the functional significance of the resident genes could be determined using both traditional crossing and sequence-based genomics approaches. Considering the allele frequency profiles across all cultivars will give us some idea of which cultivars may retain or contain the rare genes/alleles and whether these alleles may be important to our future breeding programs. The germplasm that holds unique alleles may contain unique genetic variation required for trait improvement. In the rice example (Xu et al. 2004), 15 (6.4%) of the rice accessions in the 236 × 160 data set had unique alleles for at least one RFLP locus, with a total of 21 unique RFLP alleles were found. Eighty-one (34.3%) rice accessions had unique SSR alleles with a total of 153 unique SSR alleles identified. The germplasm accessions identified with unique alleles have unique geographical origins with high genetic diversity and could have potential use in the exploitation of heterosis. Genetic similarity between any two cultivars can be calculated as the proportion of shared alleles. Theoretically, shared allele frequency (SAF) is positively related to the number of cultivars in the analysis and negatively related to the informativeness of the markers. The most similar accessions should share alleles at almost all marker loci, while the least similar accessions should have none of the alleles in common. When evaluating genetic similarity, SAFs are averaged over all possible pairs of cultivars. A smaller average similarity indicates a greater genetic difference with respect to the rest of the cultivars in the collection. Based on the averaged SAF, the most diverse accessions can be selected to represent cultivars that host the least-frequent alleles and are genotypically most different from other accessions. From 236 rice cultivars, Xu et al. (2004) selected the 16 most diverse accessions (with SAF < 50%) based on RFLP markers and 49 accessions based on SSR markers. Most of these selections, such as ‘Caloro’, ‘Cina’, ‘Badkalamkati’, ‘DGWG’, and ‘TN1’, were ancestral cultivars that had been used as parents in breeding programs over 40 years ago, and none of the selections is from the U.S. collection that has a much narrower genetic basis. D. Construction of Core Collection According to its original definition (Frankel 1984), a core subset of a germplasm collection contains, with minimal redundancy, most of the entire collection’s genetic diversity. Several different methods have

96

Y. XU

been used to construct core collections (Crossa et al. 1995; Hamon et al. 1995; Schoen and Brown 1995; van Hintum 1995). Construction of a core collection has a target of selecting approximately 10% of the germplasm accessions to represent at least 70% of the genetic variability (e.g., Brown 1989a,b). In addition to phenotypic evaluation, molecular marker technology provides us with a new tool to construct a core collection that can represent most of the genetic diversity at molecular level. Combining the use of different types of markers that reveal different levels of genetic diversity will help select a core collection to better represent genome-wide diversity. Shared allele frequency and frequency for hosting rare alleles are two important criteria that can be used to construct the core collection. In the rice example given in Section IIIB, accessions were selected based on the frequency of unique RFLP and SSR alleles and shared allele frequency. Subsets of various sizes were selected (representing 5% to 50% of each of the U.S. and world collections), using random selection as a control. For each sample size, 200 replications were analyzed through a resampling technique and the number of alleles in each subgroup was compared to the total number of alleles identified in the larger collection from which the subsets were sampled. The following conclusions were obtained (Xu et al. 2004): (1) more samples are needed to represent a diverse germplasm collection (the world collection) than a germplasm collection containing more pedigree-related cultivars (the U.S. collection), (2) combining use of shared allele frequency and unique alleles improves representativeness of a core collection, (3) core collections selected on the basis of shared allele frequency require much fewer samples than random selection for the same level of representativeness, and (4) more samples are needed to represent genetic diversity at marker loci that reveal a higher level of polymorphism. A 31-cultivar subset (13% of the entire collection), selected on the basis of both shared allele frequencies and number of unique alleles detected, represented 94.9% of RFLP alleles and 74.4% of SSR alleles. It can be expected that selection criteria based on additional sources of information will further improve the value and representativeness of core collections. When molecular markers are developed from DNA sequences with unknown function, identical marker alleles in two collections may not mean that the two collections share identical genes that are linked to the marker locus. There is a potential risk that genetic variation for important phenotypic traits could be lost if core collections are based only on those markers. As the genome sequence is deciphered and the function of many genes is determined, intragenic markers will become available for many genes. Core collections of germplasm constructed using intra-

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

97

genic markers and functional alleles will represent a “core collection” of genes. As gene structure-function relationships are clarified with greater precision, it will be possible to focus attention on genetic diversity within the active sites of a structural gene or within key promoter regions. This will make it productive to screen large germplasm collections for functional nucleotide polymorphism (FNP), targeting the search for alleles that are likely to be phenotypically relevant at specific loci. From a primary collection, a user who had identified an accession or accessions of interest would move to the next level of information, where clusters of germplasm known to represent a broader spectrum of diversity within a specific gene pool or a specific trait could be defined. The second level of investigation could be conducted using carefully designed sets of molecular markers known to target specific traits or regions of the genome. The construction of core collections may help establish heterotic groups and choose parents for establishing base populations. In the age of genomics, the context of germplasm resources should be expanded to include whole plants, seeds, plant parts, tissue, and clones, from distinct species and synthetic germplasm and all types of mutants. An extreme example is DNA or protein sequence. The ultimate goal of germplasm conservation is to maintain the genes and gene combinations. With a suitable vehicle, germplasm could be maintained in the form of a gene sequence. Genetic manipulation allows gene flow across the reproductive barriers existing among distant species. So germplasm evaluation is not necessarily defined for each species or crop (Xu and Luo 2002). The issues discussed in this section, currently suitable for each crop, can be extended to all vehicles of genetic information such as tissues, clones, genes, and sequences. E. Germplasm Genotyping Database Plant geneticists and breeders may use the data from a germplasm evaluation project as a guide in choosing the most efficient crosses for genetic studies and breeding. For instance, it provides preliminary polymorphism data for many pairwise combinations of parents. Theoretically, the dataset with 236 rice cultivars provides polymorphism surveys for 236 × 235/2 = 27,730 possible cross combinations, including thousands of indica/indica, indica/japonica, and japonica/japonica crosses. With an increasing number of markers surveyed on a variety of germplasm accessions and as more data flows into the database from multiple sources, it is increasingly possible to determine the genetic constitution and genetic relationships among a wide range of parental lines, cultivars,

98

Y. XU

and wild relatives. This also provides the foundation for developing hypotheses based on association genetics to relate agronomically important phenotypes to the presence or absence of specific molecular marker alleles. A comprehensive DNA fingerprinting of crop gene pools, including as many cultivars, hybrid parents, and progeny as possible, is the first step for using MAS in hybrid breeding. These data can be integrated with both phenotypic information and pedigree information. A database of DNA marker alleles for the elite gene pool of a crop provides information on specific DNA polymorphisms that is needed to design, execute, and analyze genetic mapping experiments, targeted at specific traits or specific crosses. The same database serves as a classification tool, describing the overall levels and patterns of variation within the crop gene pool and illustrating subdivisions within a gene pool such as heterotic groups. Such information is useful in making predictions about the performance of new cultivars and hybrids, or selecting parents for crosses that are likely to yield new gene combinations, or afford an optimal degree of heterosis. An efficient approach to the screening of germplasm involves the ability to rapidly create a nested series of core collections, based on information about geographical, phenotypic, and genotypic diversity stored in a database. The construction of such a system would require a largescale effort to provide genotypic information using a standard set of markers that could serve as a reference point. As new markers and marker systems were developed, they could be overlaid onto the essential framework of diversity established previously. An increasingly powerful information system could be developed if data models were made explicit and the data structures were modular so that new types of genetic information could be readily incorporated as they became available. As we have seen from the rice example (Xu et al. 2004), RFLP and SSR markers provide complementary information about overall genetic diversity, but one marker system may have specific advantages that recommend it for a particular type of study. By accumulating historical information in a systematic way, germplasm collections would rapidly gain in value because they could be screened computationally for essential molecular and phenotypic characteristics of interest. Databases for whole genomic sequences for several important species, both dicots and monocots, are available, with more being added, allowing directed discovery of genes in higher plants and classification of alleles present in a wide range of breeding germplasm. As indicated by Sorrells and Wilson (1997), identification of the genes controlling a trait and knowledge of their DNA sequence would facilitate classification of variation in a germplasm pool based on gene fingerprinting or charac-

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

99

terization of variation in key DNA sequences. Classification of sequence variants within genes such as SNPs at a large number of targeted loci would substantially reduce the amount of work required to determine their relative breeding value and lead to the identification of superior alleles.

V. TRAITS REQUIRING TESTCROSSING OR PROGENY TESTING Improvements in inbreds per se will play an increasingly important role in breeding hybrid crops. Bringing dominant genes for non-heterotic traits such as disease resistance into inbreds will enhance the overall performance of hybrids. MAS for non-heterotic traits should be performed the same as those in breeding cultivars. Although there is every reason to believe that plant breeding in the 21st century will still depend, to a great extent, on conventional methods for phenotypic selection, molecular biology could help identify and manipulate favorable alleles and select the traits that are not measurable under normal environments with conventional methods. Using molecular markers in plant breeding has been discussed elsewhere (Beckmann and Soller 1986; Paterson et al. 1991; Dudley 1993; Stuber 1994; Xu 1994; Xu and Zhu 1994; Lee 1995; Paterson 1996; Hospital and Charcosset 1997; Mackill and Ni 2001; Xu 2002). Xu (2002) described six situations that are suitable for MAS with the current knowledge available. These include selection without testcrossing or a progeny test; selection independent of environments; selection without laborious fieldwork or intensive laboratory work; selection at an earlier breeding stage; selection for multiple genes and/or multiple traits; and whole genome selection. Selection for traits requiring testcrossing and/or a progeny test will be discussed in this section. In breeding hybrids, many traits need testcrossing and progeny testing for unambiguous identification. Typical examples for all hybrid crops include testcrossing for screening of heterosis and combining ability. For hybrid crops based on a cytoplasmic male sterility system, malesterility restoration is at the top of the testcross list. For hybrid crops that use distant crosses with hybridization barriers, for example, partial sterility in indica × japonica crosses of rice, a testcross is required to find genes related to the hybridization barrier. In testcrossing, each candidate plant or family will be crossed to testers and then its genotype will be inferred from a progeny test in the next season. Each candidate must be harvested and maintained separately and only the plants/families with the target trait will advance to

100

Y. XU

the next procedure. Testcrossing may continue for several generations until the selected plants reach a certain level of homozygosity. A successful breeding program would take more than 50% of the breeders’ effort for testcrossing and progeny testing. Using MAS, one might reduce or eliminate testcrossing and/or progeny testing for traits controlled by major genes. A. Fertility Restoration Many important crop species, including rice, sorghum, and sunflower, depend on CMS and its fertility restoration for hybrid seed production. As indicated above, a large amount of testcrossing and progeny testing is involved in breeding CMS lines and their restorers. Testcrossing can start as early as with the F2 generation. F2 plants will be selected first for other agronomic traits, and selected plants are testcrossed to maintainer lines for CMS maintaining ability or to restorer lines for restorability. The testcross progeny will be planted the following season for fertility observation. Only the plants with complete sterile testcross progeny (for CMS) or completely fertile testcross progeny (for restorability) will be moved to the next breeding procedure. MAS can be used to replace testcross and progeny testing if markers closely linked to fertility restorability are identified. Table 3.1 lists crops and their restorability genes that have been associated with molecular markers. A total of 12 crop species were reported with molecular markers identified for fertility restoration, including maize, rice, sorghum, wheat, barley, rye, sunflower, oilseed rape, sugar beet, and onion. Genes for fertility restoration in rice have been found each on chromosomes 1, 2, 3, 4, and 5, and two on chromosome 10. Many markers are closely linked with fertility restoration and can be directly used for MAS. In rice, an RFLP marker RG140 with PvuII digestion linked with the Rf3 on chromosome 1 was useful for increasing screening efficiency for restorers (Virmani 2002). Although many genes for fertility restoration have been reported as QTL without genetic distances available, associated markers still provide useful information for MAS. In several cases, RAPD markers have been converted to STS or sequence characterized amplified region (SCAR) markers that are more suitable for MAS. B. Outcrossing Evolutionary change in plant mating systems from outcrossing (crosspollination) to inbreeding (self-pollination) has occurred frequently throughout the history of flowering plants and has been described as the

Table 3.1.

Molecular markers associated with fertility restoration in crops. R gene

Species

CMS

Allium cepa Beta vulgaris

S Owen H

X R1H

Owen Owen Ogu

2QTL QTL Rfo

Ogura

R

pol

Rfp

PET1

Rf1

Brassica napus

Helianthus annuus

Raphanus sativus Oryza sativa

PEF1 PEF1 ogu

101

BT BT HL WA WA WA WA WA WA WA

Rf1 Rf1

Rf3

Marker type RFLP RFLP RAPD, RFLP RFLP RFLP RAPD, RFLP RAPD RAPD, RFLP RAPD RFLP RAPD RAPD, SCAR RFLP SSR SSR RFLP RFLP RFLP RFLP RFLP RFLP RAPD, RFLP

Linked markers AOB210, API65 pKP1238 K11-1000, pKP753

OPC02-1150, OPD02-1000 OPK12-750, F04-500 4ND7b, 5E12b

Chromosome or (linkage group) (B) (3) 4

Distance (cM) 14, 15 9.6 5.2, 1.7

3 4 (DY15)

(18)

King et al. 1998 Pillen et al. 1993 Laporte et al. 1998 Hjerdin-Panagopoulos et al. 2002 Hjerdin-Panagopoulos et al. 2002 Delourme at al. 1994, 1998

1.2, 7.7

Hansen et al. 1997

10.8, 5.4

Jean et al. 1997

OPC07-900, OPD10-750

Ji et al. 1996 (6) (1)

OPH11-410 G2155, C1361 OSRRf RM258 RG69a, RG413 C22, RG4449d RZ404c-RG241B RG69A-RG413 C22-RG449D RG435-RG172A RG532, RG140/RG458

Reference

1.2 10L 10 3 4 2 3 4 5 1

3.5, 3.9 3.7 7.8

1.4, 1.9

Gentzbittel et al. 1995 Quillet et al. 1995 Murayana et al. 1999 Kurata et al. 1994 Akagi et al. 1996 Huang et al. 1999 Li et al. 1996 Li et al. 1996 Zhu et al. 1996 Zhu et al. 1996 Zhu et al. 1996 Zhu et al. 1996 Zhang et al. 1997 (continued)

102

Table 3.1.

Continued R gene

Species

CMS

Oryza sativa (cont.)

WA WA WA WA WA WA WA Ci

Rf4 Rf6(t) Rf5 Fr

Ci

Fr2

Phaseolus vulgaris

Secale cereale

Sorghum bicolor Triticum aestivum Zea mays

Hordeum vulgare

G P P A1

rf1

A3 T T T T T T S C msm1

rf4 Rf3 Rf3 Rf6 Rf1 Rf2 Rf8 Rf3 Rf4 Rfm1

Marker type RFLP RFLP RFLP RFLP SSR SSR RFLP RAPD, RFLP RAPD, RFLP RAPD, RFLP RFLP RFLP AFLP/ SSR ST/CAPS RFLP RFLP RFLP RFLP RFLP AFLP RFLP RFLP RAPD/STS

Linked markers RG532, R173 G403, C234 C1361, S11019 R2309, RG257 RM171, RM228 RM244 RG374, RG394 Bng228, R335F/UBC487 Bng228, Bng102

PSR596-PSR634 PSR899-MWG573 Xtxa2582, Xtxp18, Xtxp250 LW7, LW8 Xabc249, Xcdo442 Xbcd156, Xcdo388 Xksug48 umc97, umc92 umc153, sus1 Arf-8 whp1, bnl17.14 NP1114A OPI-18/900, MWG2218

Chromosome or (linkage group)

Distance (cM)

1 10 10L 10S 10L 10S 1 (K)

4.5, 0

Yao et al. 1997 Tan et al. 1998 Tan et al. 1998 Tan et al. 1998 Jing et al. 2001 Jing et al. 2001 Shen et al. 1998 He et al. 1995

(K)

0, 0.7

Jia et al. 1997

4RL 1RS 4RL (H) (E) 1BS 1BS 6BS 3S 9L 2L 8S 6H

6.0, 18.4 3.3, 19.1

Reference

3.7, 3.4

4.6 5 2.4, 12, 10.8 5.3, 3.2

1.2, 9.5 3.8, 5.8 4.5 4.3, 6.4 1.5 5.2, 5.6

Börner et al. 1998 Miedaner et al. 2000 Miedaner et al. 2000 Klein et al. 2001 Wen et al. 2002 Ma and Sorrells 1995 Kojima et al. 1997 Ma et al. 1995 Wise and Schnable 1994 Wise and Schnable 1994 Wise et al. 1999 Kamps and Chase 1997 Sisco 1991 Matsui et al. 2001

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

103

most common evolutionary trend in angiosperm reproduction (Stebbins 1957, 1970). For example, wild rice is frequently cross-pollinated, while cultivated rice is self-pollinated. Many characters involved in mating system evolution, such as sizes of floral organs or amount of pollen produced, are quantitative in nature. Hybrid seed production depends on the improvement of outcrossing-related traits and for selfpollinated crops, it might involve a reconstruction (or recovery) of the outcrossing mating system. Various techniques to produce hybrids have been developed depending on the crop, including hand emasculation, roguing of staminate plants in dioecious lines, use of gynoecious or highly female lines, CMS and genetic male sterility, protogyny, or self-incompatibility (Janick 1998). The rate of outcrossing is often the limiting factor determining whether a hybrid has potential for commercialization: Seed cost and price are both largely dependent on how easy it is to produce highquality hybrid seed that both seed providers and farmers accept. Maize was particularly suitable for hybrid breeding because of monoecism and the simple emasculation techniques practiced in breeding that allowed for easy inbreeding and outbreeding (Simmonds 1979). The necessity of high seeding rates in highly self-pollinated crops such as rice and wheat introduces an economic problem: Seed production costs must be low enough and yield of hybrids in the farmers’ fields must be high enough that farmers can profit from purchase and use of hybrid seed and companies can profit from their production and sale (Goldman 1999). Demands for low-cost seed dictates that seed yields be increased. As with sorghum and maize, some of the best new parents in sunflower are clearly more vigorous and high yielding than their predecessors. Sunflower is unique among the hybrid crops in that females are single headed but males have multiple heads, a recessive trait. Presence of multiple heads in the male ensures a long period of pollen availability, and better seed yield on the female, but it also hinders visual estimates of yield of the line per se (Duvick 1999). Yield of hybrid seed is determined by many variables, both genetic and environmental. In productive, favorable environments, seed yield from seed set through cross-pollination can approach those of conventional self-pollinated cultivars in wheat (Lucken 1986) or might be up to 80% of inbred lines in rice (Yuan and Chen 1988; Lu et al. 2001). The breeder’s approach to high, stable seed production is, first, to identify those plant and flower features that affect cross-pollination; second, to find variation for these traits; and third, to incorporate genes for favorable expression of traits into parental lines (Lucken 1986). Considering all hybrid cereal crops with the CMS system, measurements for

104

Y. XU

increased outcrossing rate will include choice of favorite climate conditions for seed production; ensuring flowering synchronization of the two parents; providing a suitable pollen source; developing male sterile lines with desirable outcrossing traits; supplementary pollination; and adjustment of flowering habit and stigma characteristics using growth regulators such as gibberellic acid. Many cereals are naturally self-pollinated. Their floral structure is adapted for inbreeding. Breeding parental lines may need to completely convert the floral structure and make them suitable for outcrossing. Outcrossing in rice depends on the capacity of stigmas to receive alien pollen and the capacity of anthers to emit much pollen to pollinate other plants in the proximity (Oka 1988). Genetic variation of floral traits in wild rice provides opportunities to modify floral structure for hybrid rice parents through the breeding process (Xu and Shen 1987, 1988a,c; Xu et al. 1988). For example, a wild rice, Oryza longistaminata, has big and exerted stigmas with a long life for outcrossing, which can be used for improving the female parent. Although genetic difference of stigma exertion between cultivated rice and a wild species is controlled by a major gene (Xu and Shen 1987), traditional breeding using Oryza longistaminata as an exerted stigma source has not been successful. Linkage between long exerted stigma and undesirable agronomic traits in wild rice species is quite strong and needs to be broken to incorporate these traits into selected genotypes. Using the gene eui (elongated upmost internode) to correct the panicle enclosure associated with CMS has been used in China for high-yielding seed production with the minimized gibberellic acid application, but has not been terribly successful. In addition, many cultivation practices have been used to facilitate outcrossing in rice by clipping flagleaf, applying gibberellic acid, supplementary pollination, and adjustment of flowering dates (Yuan and Chen 1988). The situation described for rice is also applicable to wheat and other self-pollinated species, especially those for which hand-emasculation is impractical. Although Wilson (1968) considered that the floral structure of wheat is oriented toward cross-pollination, a close examination of its floral traits clearly indicated that wheat is less suited, in its present form, to cross-pollination than crops such as maize, sorghum, and rye (Wilson and Driscoll 1983). A spike of cereal rye has ten times as many pollen grains as wheat, its flowers open considerably wider, and it has the ability to cross-pollinate under adverse weather conditions (De Vries 1971). While a great deal of variability for floral characteristics exists in wheat, the possibility of introducing additional characters favoring cross-pollination from wheat’s ancestral relatives should be explored.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

105

Miller and Lucken (1976) reported that environmental variation resulted in a sixfold difference in grain yield on a male-sterile wheat line grown at five locations in North Dakota over three years. After review of the status of hybrid wheat, Lucken and Johnson (1988) indicated the need for acquiring more knowledge about genetic variation of floral biology, including (1) spike and flower morphology; (2) pollen dispersal, buoyancy, durability, and vigor; (3) stigma accessibility, receptivity, and durability; and (4) development of selection screens for these traits. Genetic mapping of genes for restoration, pollen shedding, anther extrusion, seed set, female receptivity, combining ability, yield, quality, and disease resistance could revolutionize hybrid wheat breeding (Jordaan et al. 1999). Openness of the wheat flower and longevity of the stamen have been found to be under genetic control and can be improved through selection. Selection for better pollen quality and greater quantities thereof produced by the male will help improve the outcrossing. Technology is also available to transfer these characteristics from rye or even triticale (×Triticosecale) to wheat. Many factors affecting outcrossing provide opportunities for MAS. However, there are very few investigations on genetic mapping of traits related to outcrossing. Grandillo and Tanksley (1996) examined anther length in a backcross between Lycopersicon esculentum and L. pimpinellifolium. They found two QTL affecting this trait, on chromosomes 2 and 7, which accounted for only 24% of the phenotypic variation. Georgiady et al. (2002) investigated traits that distinguish outcrossing and selfpollinating forms of currant tomato, L. pimpinellifolium. Five QTL total were found involving four traits: total anther length, anther sterile length, style length, and flowers per inflorescence. Each of these four traits had a QTL of major (>25%) effect on phenotypic variance. In rice, some genetic mapping projects have been undertaken that target outcrossing. It is anticipated that MAS will provide a powerful tool to help fix the outcrossing-related issues in crops that are naturally selfpollinated but have great potential in hybrid breeding. Linkage drag associated with the introgression of outcrossing-related genes from wild species or distant cultivars may be overcome with marker-assisted background selection. Testcrossing-required traits, such as stigma longevity and receptivity, and labor-intensive traits, such as pollen load, can be selected much more easily through linked markers. C. Wide Compatibility Hybridization barriers exist in distant crosses of many crop species to some extent. Because the parents are not genetically compatible, hybrids

106

Y. XU

derived from intersubspecific crosses such as indica × japonica in rice are partially or completely sterile with seed set less than 30%, in addition to other unfavorable agronomic traits such as tall plant height and long days-to-heading (DTH) (Wang et al. 1991). Some intermediate cultivars have little or no hybrid barrier with either indica or japonica. The “intermediate type” was studied as early as in the 1930s, but Ikehashi (1982) first proposed the “wide compatibility” trait, which can be defined as the ability to make intersubspecific hybrids fertile. Discovery of wide compatibility in rice offers an opportunity for overcoming the reproductive barrier exhibited in hybrids between indica and japonica, and thereby for using the strong heterosis derived from intersubspecific crosses. To identify wide compatibility and transfer the related genes to other genetic backgrounds, testcrossing and progeny testing are required, as for fertility restoration. Several sets of testers were carefully selected for this purpose. IRRI evaluated wide compatibility using ‘Akihikari’, ‘Toyonishki’, and ‘Taichung 65’ as japonica testers, and ‘IR36’, ‘IR50’, and ‘IR64’ as indica testers. The China National Two-line Hybrid Rice Research Cooperative Group selected ‘Youmangzaoshajing’, ‘Banilla’, and ‘Akihikari’ as japonica testers and ‘Nantehao’, ‘Nanjing 11’, and ‘IR36’ as indica testers. Cultivars or individuals in breeding populations are considered wide compatible if the pollen fertility and seed set of the hybrids between them and all six testers are over 70%. A nation-wide effort in China for screening wide compatibility cultivars identified 51 cultivars, 49 of which had an average F1 fertility of over 80% when crossed with the six testers (Gu and Tang 2001). A lot of work is involved in testcrossing and progeny test to find out the cultivars or plants with wide compatibility. Molecular marker-assisted identification of wide compatibility genes will accelerate and facilitate the breeding process by eliminating or minimizing testcrossing and progeny testing. A genetic model was proposed by Ikehashi and Araki (1986) to account for wide compatibility. According this model, there are three alleles at the S5 locus: a neutral allele, S5n, an indica allele, S5i, and a japonica allele, S5j. A zygote from S5n allele with either of the other two alleles, S5nS5i and S5nS5j, would be fully fertile, while a zygote genotypically S5iS5j would be partially sterile. Using morphological markers, S5 was found to be closely linked with genes C (chromogen for apiculus pigmentation) and wx (waxy endosperm) on chromosome 6 (Ikehashi and Araki 1987). This chromosomal location was confirmed in several studies using isozymes (Li et al. 1991) and RFLP markers (Liu et al. 1992; Zheng et al. 1992; Yanagihara et al. 1995). A more precise

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

107

location of the S5 locus was provided by K. Liu et al. (1997), which was about 1 cM from an RFLP marker R2349. Many other wide compatibility cultivars have been identified (Xu et al. 1989; Gu and Tang 2001) and five new loci, S7, S8, S9, S15, and S16, were located on chromosomes 4, 6, 7, 12, and 1, respectively, by Yanagihara et al. (1992) and Wan et al. (1993). Allelic interaction at these loci can cause hybrid sterility, independently of each other, and neutral alleles to overcome this problem have been identified in different rice cultivars (Ikehashi and Wan 1998). These neutral alleles at different loci plus S5n are extremely important for enhancing the level of heterosis in rice. Except for these major genes, wide compatibility can be attributed to multiple loci with small effects (Shen and Xu 1992). QTL analysis of a wide compatibility cultivar ‘Dular’ identified five loci, located on chromosomes 1, 3, 5, 6, and 8, which jointly explained 55% of the variation for fertility (Wang et al. 1998). Wide compatibility has been selected using associated SSR markers in our breeding program.

VI. ENVIRONMENT-DEPENDENT TRAITS Plant populations used for gene analysis can be evaluated in either natural or controlled environments or both. Controlled environments can be compared with each other or with natural environments. If two environments mainly differ in one macro-environmental factor, they are considered contrasting or near iso-environments (NIEs), and the standard plot-to-plot variation and other residual micro-environmental effects can be neglected (Xu 2002). If the two environments are from experiments in different years or locations, it is assumed that location and year effects do not confound the effect of the macro-environmental factor. Some traits need to be measured under NIEs, where plants respond differently. In such cases, one environment imposes much less stress on plants than the other, for example, two environments with normal and high temperatures, respectively. The effect of the stress environment can be measured by comparing it to a much-less-stress or non-stress environment. A relative trait value is then derived from two direct trait values measured in each environment to ascertain the sensitivity of plants to the stress (Ni et al. 1998). If different plants have an identical phenotype under the much-less-stress environment (this is not true for a segregating population in most cases), the direct trait value in the stress environment can be used to measure sensitivity. When both environments impose little stress on plants and the plants respond differently, however, one should use relative trait values.

108

Y. XU

A. Photoperiod/Temperature Sensitivity A typical example for environment-dependent traits is photoperiod sensitivity that can only be measured in NIEs, one with short daylength and the other with long daylength. Plants start to flower when specific photoperiod and/or temperature conditions are met. Because of the complementary action of genes from two parents, hybrids could have very different photoperiod/temperature responses so that it is difficult to predict the flowering date of hybrids from those of their parents without a complete understanding of all related genes in the parents. Flowering synchronization of two parents is one of the factors influencing hybrid seed production and thus the economic advantage over the inbred lines/cultivars. To understand photoperiod and temperature responses, hybrids and their parents must be planted in a variety of environments or NIEs. Genetic study of these responses will finally characterize the parental photoperiod-thermo response pattern and its effect on their hybrids and thus make hybrid photoperiod-thermo response predictive. Once molecular markers have been associated with photoperiodand thermo-sensitivity, MAS will help minimize the requirements for multi-environment/location tests, which will reduce the breeding cost and shorten the breeding cycle. Using a rice DH population between ‘Zhaiyeqing 8’ and ‘Jingxi 17’, DTH and photo-thermo sensitivity (PTS) were investigated in two environments (Beijing and Hangzhou, China) that differ mainly in daylength and temperature (Xu et al. 1997c, Xu 2002). Four chromosomal regions were significantly associated with DTH in either or both locations, whereas a different locus on chromosome 7 (G397A-RM248) was significantly associated with PTS, indicating that the PTS QTL was independent of the QTL for DTH. By evaluating days-to-flowering (DTF) of individual ‘CO39’/ ‘Moroberekan’ RILs under 10 h and 14 h daylengths and greenhouse conditions, Maheswaran et al. (2000) identified 15 QTL for DTF. Only four of them were also identified as influencing response to photoperiod. Different QTL have been identified using direct and relative trait values, and in rice, DTH and photoperiod are often controlled by different QTL. On the other hand, direct and relative traits could share some QTL. That means DTH and photoperiod sensitivity are genetically related to some extent because both traits are related to the basic vegetative growth that rice plants must achieve to flower. There are QTL mapping studies undertaken in NIEs, but QTL were mapped using trait values scored in each environment rather than using relative measures. The traits themselves were mapped rather than the relative response

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

109

measured under the NIEs. In rice, numerous QTL for days to heading or flowering have been mapped using molecular markers but very few of them have been tested under both long- and short-day conditions. Using an F2 between japonica ‘Nipponbare’ and indica ‘Kasalath’, Yano et al. (1997) identified two major and three minor QTL for heading date. Three of them (Hd1, Hd2, and Hd3) were identified later as photoperiod sensitivity genes by test of the QTL-NILs under different daylengths (Lin et al. 2000), and one of them (Hd1) was cloned (Yano et al. 2000). In hybrid breeding, it is very important to understand what genes are involved in each parent for flowering and how they are working together in their hybrids. To synchronize flowering time of both parents for seed production, both parents should have the genes (of the same type) related to heading date and photoperiod and temperature sensitivity. Otherwise, two parents have to be planted separately to make them flower about the same time. Using MAS, days to flowering for the hybrid could be predicted from the parental genotypes and specifically designed from different combinations of parental genes. B. Environment-induced Genic Male Sterility Male sterility can be induced by specific environmental factors. An EGMS was first discovered in rice by Shi (1981) from ‘Nongken 58’, a japonica cultivar. The mutant ‘Nongken 58S’ is sterile when the days are long (>13.5 h) but becomes fertile when days are short (<13.5 h). Thus, fertility conversion is triggered by the length of photoperiod. This EGMS is called PGMS. Investigations on some derivatives from ‘Nongken 58S’ and other EGMS lines indicated that male sterility was affected by both photoperiod and temperature, and some EGMS lines were influenced more by temperature (and thus can be called TGMS), which reverted to partial or full fertility under certain temperature regimes (Zhou et al. 1988; Yang et al. 1989; Wu et al. 1991; Virmani and Voc 1991; Lu et al. 1994). Based on the EGMS system, Yuan (1987) proposed a strategy of hybrid rice breeding that did not involve a maintainer line and hence was called the two-line method. So far, more than ten EGMS lines have been used in China to breed two-line hybrid rice (Chen 2001; Lu et al. 2001; Lu et al. 2002), and over 20 two-line hybrids have been released to farmers; the area of two-line hybrid rice increased to 2.67 million ha in 2001 (Ma and Yuan 2002). EGMS has also been reported in pepper, tomato, wheat, barley, sesame, pea, rape, and soybean (Li and Yuan 2000). Compared with the CMS-based three-line method, the two-line method has several advantages. First, the EGMS parent can be used to produce hybrids with any rice cultivar, which removes the restriction of restorer

110

Y. XU

genes. Second, it is easier to introduce wide compatibility gene(s) into EGMS lines than into CMS lines, which makes intersubspecific hybrid breeding more feasible. Third, negative effects inherited from the male sterile cytoplasm can be overcome. However, the dependency of male sterility on temperature or photoperiod-temperature interaction requires two different environments in the breeding and selection process. Breeding populations have to be planted in one environment where the plants will be sterile to make sure of the presence of sterility genes and in another where the plants will be fertile to confirm the fertility conversion and produce seeds. Using associated molecular markers, confirmation of fertility conversion involving two environments can be avoided. So far three genes, pms1, pms2, and pms3, conferring PGMS in ‘Nongken 58S’ and its derivative ‘32001S’, were associated with RFLP markers on chromosomes 3, 7, and 12, respectively (Table 3.2). Although most associated markers are RFLPs, it is easy to convert them into PCR-based markers. Molecular markers on five chromosomes (2, 6, 7, 8, and 9) were associated with TGMS, most of which were microsatellite markers. A gene for reverse TGMS, which is fertile under high temperature but sterile under low temperature (contrary to the original TGMS), was associated with molecular markers on chromosome 10 (Table 3.2). These studies have laid a foundation for MAS in breeding EGMS lines. To facilitate incorporation of the tms2 gene, a SSR marker, RM11, located on chromosome 7, was identified and found to be useful in identifying heterozygous fertile plants in F2 populations and F3–F4 progenies for selection of progenies in advance (Lu et al. 2002). Lang et al. (1999) reported that PCR-based markers were 85% accurate in identifying tms3 in the juvenile stage. C. Biotic and Abiotic Stresses Breeding of insect and disease resistance and tolerance to abiotic stresses has become a worldwide issue. To identify insect/disease resistance, plants must be inoculated artificially or naturally, or in specific environments where the stress exists. Artificial inoculation is impractical when the insects/diseases are under quarantine control. On the other hand, evaluation of plant response to different insects/diseases or different biotypes/strains/races of the same stress agents is very difficult if not impossible using traditional screening methods. Using molecular markers associated with the stress response will help select for resistance without inoculation or creating or finding specific environments. Plant response to multiple biotic stresses can be screened simultaneously using molecular markers associated with these stresses. In traditional breeding programs, selection for tolerance to abiotic stress such as salinity, drought

Table 3.2. Marker-trait associations identified in rice for photoperiod-sensitive genic male sterility (PGMS), thermo-sensitive genic male sterility (TGMS), and reverse TGMS (RTGMS).

Trait PGMS

TGMS

RTGMS

Gene

Source

pms1

32001S

pms2 pms3

32001S Nongken 58S

tms1 tms2 tms3 tms5 tms-vnl

5460S Lorin PL12 IR32364S Annong S-1 SA2 TGMS-VN1

rtms1

J207S

Chromosome 7 3 12 8 7 6 2 9 2 10

Linked markers

Distance (cM)

Reference

RG477, RG511/RZ272 RG477/R277, R1807 RG191, RG348 RZ261/C751, R2708 RZ261, M36

3.5, 15.0 0.1, 6 7.0, 10.6 5.5, 9.0 1.6, 1.6

Zhang et al. 1994b Liu et al. 2001 Zhang et al. 1994b Mei et al. 1999 X. Li et al. 2001

RZ562, RG978 (R643A, R1440) OPAC3-640, OPAA7-550 RM174, R394 RM257/EAA/MCAG E5/M12-600, E3M16-4003

1.7 7.7, 10.0 0, 2.5 6.2, 5.3 3.3, 28.8

Wang et al. 1995 Yamaguchi et al. 1997 Subudhi et al. 1997 Jia et al. 2000 Reddy et al. 2000 Dong et al. 2000

RM222, RG257

11.8, 4.6

Jia et al. 2001

111

112

Y. XU

and submergence tolerance, and lodging resistance can only be done at specific environments that are either present at specific locations or created at well-controlled environments. Selection for these traits is considered most difficult in breeding programs. To measure responses to agrochemicals, such as herbicides and plant growth regulators, such chemicals must be applied to plants at the right stage under suitable environments. MAS has made it possible to perform indirect selection for all these traits using tightly linked or intragenic markers. There are several successful examples using MAS for biotic stress in rice. Hittalmani et al. (2000) used MAS to combine three rice blast resistance genes (Pi1 on chromosome 11, Piz-5 on chromosome 5, and Pi-ta on chromosome 12) into a single genotype. For Piz-5, a single marker was used, whereas flanking markers were used for the other two. MAS was efficient in developing gene pyramids and the line containing all three resistance genes had a broader resistance spectrum than lines with only one of them. Huang et al. (1997) pyramided four bacterial blight resistance genes, Xa4, xa5, xa13, and Xa21, using PCR-based markers. Sanchez et al. (2000) transferred three bacterial blight resistance genes into susceptible rice lines possessing desirable agronomic characteristics. For efficient MAS, sequences from RFLP or genomic clones, linked to the resistance genes (xa5, xa13, and Xa21), were converted to STS markers. This work showed the effectiveness of using markers linked to recessive genes in a backcrossing program, particularly in the presence of a dominant resistant gene. In an F2 population, selection efficiency was as high as 95% for xa5 and 96% for xa13. Breeding for drought tolerance is becoming a top priority in many countries, as less and less water is available for agriculture. Current knowledge on physiology suggests that drought tolerance in rice depends on one or more of the following components: (1) the ability of roots to exploit deep soil water to provide for evapotranspirational demand, (2) the capacity for osmotic adjustment that allows plants to retain turgor and protect meristems from extreme desiccation, and (3) control over nonstomatal water loss from leaves (Nguyen et al. 1997). These components are generally applicable to other cereal crops. There is a large body of literature published in China indicating that hybrid rice has a strong root system characterized by more and longer roots, with better distribution in the soil as compared to its parents and inbred lines (Yuan and Chen 1988). In 1999, Zhang et al. (1999) summarized the QTL identified for drought-tolerance components in rice. More than 100 QTL had been identified for osmotic adjustment, dehydration tolerance, abscisic acid accumulation, stomatal behavior, root penetration index, root thickness, total root number, root length, total dry root weight, deep root dry weight, and root pulling force.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

113

In maize, grain yield under drought stress is negatively correlated with the anthesis-silking interval (ASI), the difference in days between pollen shedding and silk emergence. A short ASI means rapid silk extrusion because time to anthesis is little affected by drought. Five QTL that were stable over stressed environments were identified under several levels of drought. A backcross MAS scheme to improve the drought tolerance of an elite but drought-susceptible inbred line, CML 247, has been successfully completed, using PCR-based markers as a preselection tool (Ribaut et al. 1999). A second MAS experiment used molecular markers to select changes in the frequency of alleles at loci having a known association with drought tolerance as a result of recurrent selection in an open-pollinated population. Plant selection based on the presence/ absence of those alleles whose frequency changed could give increased drought tolerance in less time than recurrent selection. Ribaut et al. (1999) concluded that new breeding schemes involving optimal combinations of MAS and conventional selection to improve drought tolerance in maize hold considerable promise for the future. In rice, the International Rice Research Institute (IRRI) has several drought-tolerance breeding programs using identified QTL and MAS. QTL affecting root parameters were identified using a rice DH population derived from the cross ‘IR64’/‘Azucena’. A marker-assisted backcross program was started to transfer the alleles of ‘Azucena’ (upland rice) at four QTL for deeper roots on chromosomes 1, 2, 7, and 9 from selected DH lines into ‘IR64’ (elite rice cultivar) (Shen et al. 2001). The backcross progeny were selected strictly on the basis of their genotype at the marker loci in the target regions up to the BC3F2, from which BC3F3 near-isogenic lines (NILs) were developed and compared to ‘IR64’ for the target root traits. Of the three tested NILs carrying target 1 (QTL on chromosome 1), one had significantly improved root traits over ‘IR64’. Three of the seven NILs carrying target 7 (QTL on chromosome 7) alone, as well as three of the eight NILs carrying both targets 1 and 7, showed significantly improved root mass. Four of the six NILs carrying target 9 (QTL on chromosome 9) had significantly improved maximum root length. The results indicate that MAS can be used to select the traits requiring specific environments for phenotyping.

VII. QUALITY TRAITS Many important traits in crops are phenotypically invisible or unscorable and must be measured in the laboratory using sophisticated equipment or facilities. Some analyses require a large amount of seeds so that they cannot be measured until late generations when a relatively

114

Y. XU

large amount of seed becomes available for each selection entry. Grain chemical properties such as starch and protein content, and physical properties such as milling quality, are important quality traits for most cereal crops that fall under this category. A. Seed Traits As a major storage organ of cereal seeds such as rice, endosperms provide humans with proteins, essential amino acids, and oils. An understanding of the inheritance of endosperm traits is critical for the improvement of seed quality. Genetic behavior in triploid endosperms is much different from that of the maternal plants that supply assimilates for grain growth and development. Thus, methods suited for genetic analysis of traits in maternal plants (diploids for most cereal crops) cannot directly be used for endosperm traits (Xu 1997). Any genetic analytical method for endosperm traits needs to combine a genetic method developed for diploid maternal plants with a triploid model proposed for conventional genetic analysis. The genetic system controlling endosperm traits may be much more complicated than that controlling traits of the plant per se. Because the plant provides seeds with a portion of their genetic material and almost all the nutrients required for growth and development, seed traits are genetically affected by both the seed nuclear genes and maternal nuclear genes. In addition, cytoplasmic genes may also affect some seed traits through their indirect effects on the biosynthetic processes of chloroplasts and mitochondria. To understand endosperm traits with biological accuracy, one should take into consideration maternal genetic effects and cytoplasmic effects along with the direct genetic effects of seeds. As seeds initiate a new generation that differs from their maternal plants, some seed traits should be considered as one generation advanced over their maternal plants. Genetic analysis of endosperm traits should be based on the DNA extracted from both maternal plants and endosperm tissues in order to understand the relative contribution of the different genetic factors to the variation of endosperm traits (Xu 1997). Currently all endosperm traits have been treated the same as other traits of the plant, with few reports (Tan et al. 1999) that considered the generation advancement issue. B. Hybrid Seed Traits Although F1 plants are uniform, seeds borne on them represent the F2 seed generation and are expected to segregate for some grain characteristics. Major determinants of grain quality in cereal crops are milling;

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

115

grain size, shape and appearance; and cooking and eating characteristics. Some grain tissues are of maternal origin and some result from fertilization and union of genetically diverse gametes. For example, the lemma and palea of the rice hull are maternal tissues. Seed size and shape are determined by the shape and size of hulls and the latter is determined by the genotype of F1 plants. As a result, all F2 seeds borne on F1 plants have nearly identical dimensions even though the parents could have very different seed sizes. Endosperm is triploid tissue resulting from the union of one male nucleus with two female nuclei. If the parents differ in endosperm traits, the F2 grains on F1 plants show clearcut segregation (Kumar and Khush 1986; Tan et al. 1999). Single seed analysis of a hybrid rice, ‘Shanyou 63’, indicated that the amylose content for seeds on a F1 plant could range from 8% to 32% when two parents had 15.8% and 27.2% amylose content, respectively. A similar situation was reported for barley. If the parents differ significantly in malting quality characters, the grain produced by barley hybrids will be heterogeneous and heterozygous for characters critical to the malting process (Ramage 1983). The cooking and eating quality of milled rice is related to starch properties such as amylose content, gelatinization temperature, and gel consistency. These three traits are highly correlated and the latter two are controlled by a locus located in the same chromosomal region as amylose content (Tan et al. 1999). When both parents are non-glutinous but differ widely in amylose content, the raw F2 grains are distinctly classifiable into different amylose categories. But, when cooked in bulk, the grains do not vary in cooking and eating characteristics (Khush et al. 1988). That might explain why the hybrid rice ‘Shangyou 63’, which contains 8% to 32% amylose contents among different grains, has been acceptable for many years in China. Although genetic heterozygosity of hybrids may not impair grain quality in terms of physical and chemical characteristics as long as one of the parents is not glutinous or poor in grain quality (Khush et al. 1988), amylose difference among grains may contribute to unstable milling quality and uneven endosperm traits such as chalk. For hybrid rice breeding programs in countries demanding high-quality standards, both parents should be equally good if hybrids are to meet the standard quality of inbred cultivars. To develop glutinous rice hybrids, both parents must have glutinous grains. To develop hybrid rice possessing premier grain quality like that of basmati, both parents must possess basmati quality (Virmani and Zaman 1998). The same situation may be true for other cereal hybrids. It is strongly suggested that to develop hybrid crops with grain quality as good as the best inbred lines/cultivars, both parents must have the best quality. Use of parental lines with equally good quality will help minimize this effect, but might

116

Y. XU

result in low heterosis because of less genetic diversity for grain traits between the parents. Genetic contribution to quality comes from both parents, but one of them could be more important than another in some specific situations. Endosperm properties might be affected more by female parents due to the maternal effect, or more by male parents due to the xenia effect (Xu and Shen 1988b). The composition and development of the kernels can be changed by the nature of pollen. This was first shown by Kiesselbach (1926) as the change of a sweet corn endosperm into a starchy endosperm after pollination of a sweet corn female by a flint endosperm male. Large xenia effects were observed for sorghum malt quality in the F1 but this was entirely lost in the F2 generation (Wenzel and Pretorius 2000). Curtis et al. (1956) observed that the germ is markedly influenced in weight, oil and protein content by both the seed parent and the pollen parent of corn, with a pronounced maternal effect. C. Selection of Quality Traits Many quality traits are genetically controlled by multigenic loci, or by multiple alleles at a locus because of the triploidy of the endosperm. As a result, the same phenotypes may come from different genetic factors or different alleles from the same locus. Phenotypic selection for the same trait values may not result in the same alleles or genes fixed in parents. MAS will help distinguish different genetic loci that contribute to the same quality traits to avoid the problem that the two parents used for hybrid production have different alleles or alleles from different loci. Almost all quality traits are only measurable at or after the reproductive stage and they are good candidates for MAS. MAS can be made at any stage and in any generation so that breeders do not need to maintain a large number of candidate plants generation after generation. Methods for non-destructive extraction of DNA from single dry seeds provide an opportunity for selection of seed traits so that selection can be processed before planting. Early-stage selection also provides more opportunities for selection of traits with relatively low heritability. MAS could be used for early-stage quality tests or DNA-based quality tests, whereas such tests would be delayed in a conventional breeding program because a relatively large amount of seeds is required. Malt quality traits of barley, which are influenced by many loci, are suitable candidates for MAS. The genetic make-up of modern malting barley cultivars has been achieved by carefully balancing these traits through lengthy breeding processes. In a MAS project conducted among the progeny from the same cross for which the QTL had been mapped,

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

117

Igartua et al. (2000) confirmed the presence of QTL on chromosome 7 affecting all traits. MAS was effective in selecting phenotypically superior lines and the magnitudes of the combined effects for these regions were close to the estimates calculated in the mapping experiment. The use of marker information in the selection did not eliminate the need to gather reliable phenotypic data but it should permit breeders to allocate resources to the evaluation of progeny that are more likely to carry favorable alleles. Rice amylose content, mainly controlled by the wx gene, is a good example of MAS. Ayres et al. (1997) determined the relationship between polymorphism at that locus and variation in amylose content. Eight wx microsatellite alleles were identified from 92 long-, medium-, and shortgrain U.S. rice cultivars. When used as predictors of amylose content, these eight alleles explained an average of 85.9% of the variation. The amplified products ranged from 103 bp to 127 bp in length and contained (CT)n repeats, where n ranged from 8 to 20. Average amylose content in cultivars with different alleles varied from 14.9% to 25.2%. Using more diverse rice germplasm accessions (n=243), Zeng et al. (2000) identified 15 alleles at the wx locus using microsatellite class and G-T polymorphism, resulting in a total of 16 alleles identified so far. Although the mircrosatellite marker was located in the intron of the waxy gene, a complete association between marker alleles and amylose contents still depends on fully understanding other genes involved in the starch synthesis.

VIII. GENE INTROGRESSION AND WHOLE GENOME SELECTION In previous sections, several types of special traits that are important to hybrid breeding and most suitable for MAS were discussed. Except for quality traits discussed in Section VII, there are many other traits that need laborious fieldwork or intensive laboratory work. Xu (2002) fully described all the trait categories most suitable for MAS, and Dekkers and Hospital (2002) also listed the opportunities for the use of molecular data. In this section, some general considerations for all trait categories will be discussed. A. Gene Introgression Emerging MAS technology should provide the vehicles for using markers to expedite the acquisition of important genes from exotic populations or from wild species. Gene introgression involves the introduction

118

Y. XU

of a target gene into a productive, recipient line or cultivar. Gene introgression can be used in both backcrossing and intercrossing programs. By using DNA markers to identify recombinants, introgressed chromosome segments might be “trimmed” to minimal size, reducing the extent to which the recurrent genotype is disrupted by undesirable alleles closely linked to the target trait (Tanksley and Rick 1980). It is critical in plant breeding that allelic substitution be precise so that only the target gene and the shortest possible segment of the linked chromosome are transferred from the donor parent to the recipient parent, the latter of which is usually a cultivar or inbred line with very good combining ability. To reduce false positives in MAS, markers must be tightly linked to the target trait, and flanking markers or multiple markers around the region could be used simultaneously. A three-marker system, with three markers located on a chromosome block of a few (<5 cM), will be desirable in this case (Zhang and Huang 1998). The marker in the middle, preferably intragenic or cosegregating with the gene, will be used to indicate the presence of the target gene in the selection process. The marker on each side will be used to indicate the absence of the chromosome segment from the donor parent (negative selection), that is, selection for recombination between the target gene locus and the marker locus. As more and more genes have been cloned, the marker in the middle could be developed from the cloned gene or gene sequence, as discussed in Section IIIA. This system will be very useful when the target gene is only available in a wild species and linkage drag is proven to be associated with the chromosome segment to be transgressed. In rice, a series of advanced backcrossing populations have been developed through collaborations between Cornell University and breeders around the world to identify and introgress trait-enhancing alleles from wild species into high-yielding elite cultivars. The first such study employed a cross between the wild rice relative Oryza rufipogon and the Chinese indica hybrid ‘V20’/‘Ce64’ (Xiao et al. 1998). Although the O. rufipogon accession was phenotypically inferior for all 12 traits studied, transgressive segregation was observed for all traits, and 51% of the QTL detected had beneficial alleles from O. rifupogon. By MAS and field selection, an excellent CMS restorer line (‘Q661’) carrying one of the QTL for yield components has been developed. Its hybrid, ‘J23A’/‘Q661’, outyielded the check hybrid by 35% in a replicated trial for the second rice crop in 2001 (Yuan 2002). A second QTL study used an advanced backcross population between the same O. rufipogon accession and the upland japonica rice cultivar ‘Caiapo’, and identified beneficial QTL alleles from O. rufipogon for 56% of the traitenhancing QTL detected (Moncada et al. 2001). A third study employed the O. rufipogon in a cross with the long-grain ‘Jefferson’, a U.S. tropi-

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

119

cal japonica cultivar, and the O. rufipogon allele was favorable for 53% of the yield and yield component QTL (Thomson et al. 2003). There are several ongoing projects to introgress these favorable alleles from the O. rufipogon accession into cultivated rice. Information from the mapping studies reported may not be used directly in designing MAS experiments in breeding programs because different genetic backgrounds and environments have been involved. Other than epistasis and G × E interaction, many important questions regarding genetic regions tagged by molecular markers remain unknown. For instance, the number of polymorphic genes and the number of functional alleles at each of these loci in the gene pool are generally unknown for most traits. Cross-population comparison of genetic mapping studies using biparental materials can only provide limited information. As a result, many questions remain to be answered before any MAS experiment for major gene/QTL transfer can be used: (1) information on genetic parameters of a locus such as location, effect, and linked markers; (2) number of loci and number of alleles at each locus and variation explained by each locus and by all loci; (3) background effect on genes and gene interaction; and (4) linkage drag and its genetic basis. A reasonable strategy for molecular breeding is to do simultaneous gene identification and introgression such as advanced backcross QTL analysis as suggested by Tanksley and Nelson (1996). A comprehensive backcrossing breeding program through international breeding activities has been initiated for integration of DNA markers with phenotypic selection (Li 2001), which focuses on improvement of inbred cultivars and the parental lines for hybrid rice. B. Whole Genome Selection MAS can also be practiced at the whole genome level. DNA markerbased whole genome selection or background selection can be used to accelerate recovery of a recurrent genotype in the backcrossing process for breeding parental lines. Compared to a backcross program that usually takes five to seven generations to recover most recurrent parental backgrounds, MAS may save two to four backcross generations in the transfer of a single target allele (Tanksley et al. 1989; Hospital et al. 1992; Fisch et al. 1999). Combined with selection for target traits, whole genome selection allows the breeder to simultaneously transfer multiple traits through backcrossing. When single chromosomes are distinguishable, partial genome selection or whole chromosome selection are alternatives to whole genome selection so that the other chromosomes remain unchanged. MAS could be focused on a chromosomal region/arm if it is separable from the rest

120

Y. XU

of the genome. Genes controlling the same traits or trait category may cluster in some specific chromosomal regions, which are called gene blocks. Regional mapping strategies (Xu 1997; Monna et al. 2002), combined with a high-density genetic map, can help construct high-density regional maps that target gene blocks for separation of closely linked genes. As genetic mapping information accumulates from different mapping populations, it will be possible to establish a complete profile for all the genes associated with a specific trait or trait category. Whole genome selection can be used to select the best trait/gene combinations based on selection for each of the target loci whose position in the genome is known. It is possible to select the best cassette for any traits and/or trait combinations. To transfer the bacterial blight resistance gene Xa21, 128 RFLP markers, evenly distributed on the 12 rice chromosomes, were used to recover the genetic background of ‘Minghui 63’, a widely used parent. Plants containing Xa21 in the BC3F1 generation were screened with markers covering the genome, and those homozygous for ‘Minghui 63’ alleles were saved. The improved version of ‘Minghui 63’ and its hybrid with ‘Zhenshan 97’ showed the same resistance spectrum as the resistant donor. Field examination of a number of agronomic traits showed that the improved ‘Minghui 63’ and its hybrid were identical to the originals except for their resistance to bacterial blight (Chen et al. 2000). MAS was also used by the same group to improve ‘6078’, an elite restorer line with high yield potential, by transferring Xa21 from IRBB21 (Chen et al. 2001). Background selection in this study took place in the BC1F1 and also in the BC2F1 population using AFLP markers with unknown positions. It is not clear exactly how much of the genetic background of the recurrent parent was recovered in the individual obtained in the final selection (because of unknown chromosomal distribution of the 129 polymorphic bands), but the agronomic performance and combining ability of the selected line was very similar to the original. C. Selection for Multiple Genes/Traits MAS provides opportunities for simultaneous selection of multiple traits/genes. In some cases, multiple pathogen races or insect biotypes must be used to identify plants for multiple resistances, but in practice this may be difficult or impossible because different genes may produce similar phenotypes that cannot be distinguished from each other. Marker-trait association can be used to simultaneously select multiple resistances from different disease races and/or insect biotypes, and pyramid them into a single line through MAS as discussed in Section VIC.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

121

To find a CMS restorer in rice through testcrossing and progeny test, a candidate male plant has to be testcrossed with a CMS line to find out if it has fertility restorability based on the fertility of testcross progeny. However, sterility in testcross progeny could result from the absence of either restorability genes or wide compatibility genes or both when an intersubspecific cross is involved. MAS could be used to distinguish the two different types of sterility. Consider phenotypic selection for multiple traits in rice, such as TGMS, amylose, and wide compatibility. Candidate plants must be tested in two different environments where TGMS can be identified. Each plant must be testcrossed with wide compatibility testers, following up with a progeny test in the next season. At the same time, a relatively large amount of seed must be harvested for amylose measurement. While conventional selection methods require a delay until a large number of seeds are available and a reasonable level of homozygosity is reached, in MAS only a leaf harvested at any growth stage in any segregating population is required. Hybrid rice provides an advantage over inbred cultivars because dominant genes and/or QTL with additive effects from both parents can be integrated into one hybrid. An integrated breeding program including MAS was initiated in China to improve an elite hybrid rice, ‘Shanyou 63’, a cross between ‘Zhenshan 97’ and ‘Minghui 63’. Xa21, a widespectrum bacterial blight resistance gene, was introduced into the restorer ‘Minghui 63’ by MAS and a Bt gene that is toxic to stem borer was introduced into ‘Minghui 63’ through transformation. An allele at the Wx locus from ‘Minghui 63’ was transferred by MAS to ‘Zhenshan 97’ to improve cooking and eating quality of the hybrid, resulting in a new version of ‘Zhenshan 97’ with medium amylose content, soft gel consistency, and high gelatinization temperature. The pyramiding of Bt, Xa21, and wx genes created an improved ‘Shanyou 63’ (He et al. 2002). D. Integrated Genetic Mapping and MAS MAS programs involve genetic mapping to identify genes of interest followed by transfer to another genetic background. This procedure was used to enhance ‘B73’ and ‘Mo17’ inbreds in maize (Stuber et al. 1999). The procedure per se is inefficient because it requires identification of the targeted segments (containing the putative genetic loci) prior to transfer to the recipient line. In many cases, genetic mapping results obtained from specific crosses cannot be used for MAS for the same traits in different crosses. There are three reasons for this phenomenon. First, quantitative traits are usually controlled by many genes. Genes are only segregating at the loci where two parents are genetically different and

122

Y. XU

thus can be mapped using the population derived from these two parents. For a randomly selected mapping population, the parents will have a strong chance to share identical alleles at some of the genetic loci. There is a high probability that segregating genes in any breeding population could be different from the genes already mapped. Second, multiple alleles at a locus work in the same way to complicate MAS, because mapping parents could have alleles that are different from those of breeding populations. Interaction among these multiple alleles will modify marker-trait associations when different allele combinations are considered. Third, G × E interaction could make the establishment of marker-trait association depend on specific environments. One of the best ways to avoid these limitations is to integrate genetic mapping with MAS, that is, marker-trait associations identified from a breeding population will be used for MAS of the same population. This is critical for quantitative traits that are genetically controlled by many genes and interact with environments. Advanced backcrossing QTL analysis, proposed by Tanksley and Nelson (1996) to accelerate the process of molecular breeding, is one of the approaches that can be used for this purpose. Stuber et al. (1999) discussed their effort to test a marker-based breeding scheme for systematically generating superior lines without any prior identification of genes in the donor sources. The identification of and mapping of genes in the donor is a bonus obtained when the derived NILs are evaluated. This method is somewhat similar to advanced backcross-QTL analysis. Other approaches include using associations identified in F2 populations to select the subsequent selfpollinated populations. E. Response to Selection MAS for traits controlled by major genes will receive a strong response. However, the response to selection for quantitative traits using associated molecular markers will depend on several factors: linkage between markers and genes, trait heritability, gene effects, gene interactions, population size, the number of plants selected, and the breeding scheme. In this section, response to MAS of QTL will be discussed. An example in Lande and Thompson (1990) demonstrated that on a single trait the potential selection efficiency by using a combination of molecular and phenotypic information, compared to standard methods of phenotypic selection, depends on the heritability of the trait, the proportion of additive genetic variance associated with marker loci, and the selection scheme. The relative efficiency of MAS is greatest for traits with low heritability if a large fraction of the additive genetic variance is associated with marker loci. Limitations that may affect the potential

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

123

utility of MAS in applied breeding programs include (1) the level of linkage disequilibrium in the populations, which affects the number of marker loci needed; (2) sample size needed to detect trait loci with low heritability; and (3) sample errors in the estimation of relative weights in the selection indices. In theory, MAS is proposed to be more efficient than phenotypic selection when the heritability of a trait is low, where there is tight linkage between QTL and DNA markers (Dudley 1993; Knapp 1998), with larger population sizes (Moreau et al. 1998), and in earlier generations of selection before recombinational erosion of marker-trait associations (Lee 1995). Edwards and Page (1994) proposed that the distance between markers and QTL was the factor that most limited gains from MAS. Yousef and Juvik (2001) reported an empirical experiment that provided equivocal results regarding the relative efficiency of MAS and phenotypic selection in enhancing economically important quantitative traits in sweet corn. MAS and phenotypic selection were applied to three F2:3 base populations with either the sugary 1 (su1), sugary enhancer 1 (se1), or shrunken 2 (sh2) endosperm mutations. One cycle of selection was applied to both single and multiple traits such as seedling emergence. Selection efficiencies were evaluated on the basis of gains over one cycle. Among 52 paired comparisons between MAS and phenotypic selection composite populations, MAS resulted in significantly higher gain than phenotypic selection for 38% of the comparisons, while phenotypic selection was significantly greater in only 4% of the cases. The average MAS and phenotypic selection gain, calculated as percent increase or decrease from the randomly selected controls, was 10.9% and 6.1%, respectively. Recognizing that small mapping populations are not adequate for QTL mapping is the first and most important realization needed in the research community (Young 1999). Scientists must understand that simply demonstrating that a complex trait can be dissected into QTL and mapped to approximate genomic regions using DNA markers is not enough. Projects need to utilize better scoring methods, larger population sizes, multiple replications and environments, appropriate quantitative genetic analysis, various genetic backgrounds, and, whenever possible, independent verification through advanced generations or parallel populations. Only then will sufficient experimental evidence be in place for a successful MAS program. “What if we knew all the genes for a quantitative trait in hybrid crops?” This was asked by Bernardo (2001), who has been working on the prediction of hybrid performance through computer simulation. With maize as a model species, he found through trait and gene best linear unbiased prediction (TG-BLUP) that gene information is most useful in selection

124

Y. XU

when few loci (e.g., 10) control the trait. With many loci (50), the least square estimates of gene effects become imprecise. Gene information consequently improves selection efficiency among hybrids by only 10% or less, and actually becomes detrimental to selection, as more loci become known. Bernardo further indicated that increasing the population size and trait heritability to improve the estimates of gene effects also improves phenotypic selection, leaving little room for improvement of selection efficiency via gene information. He thought genomics is of limited value in selection for quantitative traits in hybrid crops. Epistatic interactions, which were assumed absent in his study, would make the estimation of gene effects even more difficult. It is unknown whether methods other than TG-BLUP or multiple regression would substantially enhance the usefulness of gene information in selection. Response to phenotypic selection can be evaluated using molecular markers. The Illinois long-term selection experiment on maize oil and protein contents (Dudley and Lambert 1992; Dudley 2004) and markerassisted evaluation (Goldman et al. 1993) is an example of markerassisted evaluation of the response to selection. This selection experiment was initiated in 1896, and by 1989 had experienced 90 generations of selection, increasing oil content from 4.7% in the original population to 19.3% in the Illinois High Oil (IHO) strains. In contrast, 87 generations of selection for low oil concentration reduced oil content from 4.7% to <1% in the Illinois Low Oil (ILO) strains. Mean protein concentration for 76 generations were 25% for the Illinois High Protein (IHP) strains and 4% for the Illinois Low Protein (ILP) strains (Dudley and Lambert 1992). This long-term divergent selection response can be attributed to the accumulative action of alleles with similar effect that had been dispersed among the individuals of the original population (Xu 1997), although de novo mutations may be an alternative explanation for this divergence, as indicated by selection for bristle number in Drosophila (Mackay 1995). The selection strains offer a unique opportunity to investigate the genetic basis of kernel chemical traits, and have been used to produce maize populations to map the QTL responsible for the selection response (Goldman et al. 1993). By using 90 genomic and cDNA clones distributed throughout the maize genome to detect RFLPs between IHP and ILP strains, 22 loci distributed on 10 chromosome arms were significantly associated with protein concentration, and clusters of three or more significant loci were detected on chromosome arms 3L, 5C, and 7L, suggesting the presence of QTL with large effects at these locations. A multiple linear regression model consisting of six significant loci on different chromosomes explained over 64% of the total variation (Goldman et al. 1993). These significant QTL associations can be used to account for the long-term selection response and the protein

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

125

content difference between the IHP and ILP strains. It can be expected that the longer the selection proceeds, the bigger the difference of protein content will be in the resulting selection strains, and thus the potential to detect additional QTL, as long as the populations continue to respond to selection. This expectation can be tested by QTL mapping using the crosses from the IHP and ILP strains derived from different cycles of selection. We would expect genetic fixation with long-term selection programs. However, selection experiments discussed above for maize for high and low protein or oil and in Drosophila for bristle numbers (Yoo 1980) show no indication of genetic fixation from long-term selection resulting in remarkable changes in phenotype. Frequent identification of large-effect QTL, as reviewed by Tanksley (1993), Kearsey and Farquhar (1998), and Xu (2002), makes steady and sustained selection response puzzling: Alleles of large effects should be fixed rapidly, after which no further response would be seen. Barton and Keightley (2002) named two factors that might explain this apparent paradox. First, QTL-mapping experiments underestimate the number of QTL and overestimate their effects. Second, mutation generates alleles of large effect, which can be picked up quickly enough by selection to sustain a continuing selection response. Several mechanisms have been described that can create de novo variation, including intragenic recombination, unequal crossing over among repeated elements, transposon activity, DNA methylation, and paramutation (R. L. Phillips, pers. commun.). Barton and Keightley (2002) listed several factors that make it difficult to estimate the true numbers and effects of loci influencing a quantitative trait. Hyne and Kearsey (1995) pointed out that in a typical experiment (heritability ~40%, ~300 F2 individuals), no more than ~12 QTL are ever likely to be detected, which is supported by empirical data on the numbers of QTL detected in plants (Tanksley 1993; Kearsey and Farquhar 1998; Xu 2002). Beavis (1994) indicated that unless samples are large (>500, for example), the effects of statistically significant QTL are substantially overestimated. IX. PREDICTION OF HYBRID PERFORMANCE AND HETEROSIS A. Combining Ability and Heterosis There are two types of combining ability, general and specific. General combining ability (GCA) is defined as an attribute of an inbred line and is measured as the average performance of all hybrids made with that inbred line as a parent. The higher the GCA of an inbred, the higher the average performance of its hybrids. Specific combining ability (SCA) is

126

Y. XU

defined for specific combinations of parents and is measured by the deviation of the hybrid performance from the expected performance as estimated from the GCA of the parents. As a result, hybrid performance is determined by its parents’ GCA and the cross’s SCA. The widely grown rice hybrid in China, ‘Shanyou 63’, is an example of this. The female parent (CMS line), ‘Zhenshan 97’, and the male parent, ‘Minghui 63’, have been used to produce many successful hybrids with various fertility restorers and CMS lines. Heterosis is an important cause of the increasingly high yields of rice, maize, grain sorghum, and oil sunflower, although it is not the only cause. Improvements in GCA, in additive genes as well as in dominant, overdominant, or epistatic gene combinations, have been crucial to the improvement of hybrids in all four crops. Heterosis will continue to be a very important cause of hybrid superiority in yield and yield stability. SCA—specific combinations of inbred lines with good GCA—will remain the essential requirement for production of superior new hybrids (Duvick 1999). The use of marker-aided prediction of advanced generation combining ability on the basis of data from early generation testcrossing was evaluated by Johnson and Mumm (1996). Using a total marker score generated from regressing F3 line testcross yield on the corresponding F3 line RFLP genotypes, F5 testcross yields were predicted with more accuracy than they would have been with F3 testcross yields alone as predictors. It was concluded that marker-assisted prediction of advanced generation performance from early generation testcrosses was effective. These results were in agreement with those reported by Eathington et al. (1997). B. Genetic Basis of Heterosis Heterosis is a complex physiological phenomenon affected by many factors. Yield is the most important trait in crop-based heterosis analysis. In many investigations, genes for yield per se and genes for yieldrelated heterosis have been confounded with each other. Several different hypotheses have been proposed for the explanation of heterosis but none of them is widely accepted so far. Among these hypotheses, arguments have focused on the dominance hypothesis (Davenport 1908) and the overdominance hypothesis (East 1908; Shull 1908), both of which are based on describing the genetic effects of single loci. Recent studies, as will be discussed later, have indicated that epistasis plays an important role in genetic control of both quantitative traits and heterosis. The dominance hypothesis proposes that heterosis results from the cancellation of effects from deleterious recessive alleles, contributed by one parent, by dominant alleles contributed by the other parent in the heterozygous F1. This hypothesis emphasizes the contribution of the

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

127

dominance to heterosis. The overdominance hypothesis assumes that a specific heterozygous combination of alleles at a single locus is superior to either of the homozygous combinations of the parental alleles at that locus. Two alleles with different effects are presumed to interact physiologically such that heterozygotes show superiority in viability and adaptability compared to either of the homozygotes, and the degree of superiority increases with the degree of allelic heterozygosity. Supporters of the overdominance hypothesis offer two main objections to the dominance hypothesis. First, it should be possible to accumulate, by selection, all the favorable dominant alleles into one homozygous strain and obtain inbreds that are as vigorous as hybrids. Second, F2 distributions should be skewed because of the 3⁄4 dominants to 1⁄4 recessive segregation (Lamkey and Edwards 1999). But in 1916, Jones noted that with linkage the two hypotheses became indistinguishable and Collins pointed out in 1921 that with a large population size the F2 distribution is essentially symmetrical. From the results of Gardner (1963) and Moll et al. (1963), the clear conclusion was that statistical overdominance in the early generations was a consequence of linkage disequilibrium, that is, favorable dominants linked to deleterious recessives. For selfpollinated species, the genetic basis for heterosis may be different because of purging of deleterious recessive alleles by long-time inbreeding (Crow 1999). As a complex trait, heterosis is likely to be conditioned by numerous genetic factors functioning and responding to a wide variety of interacting situations, and epistasis must also play an important role in the genetic control of heterosis. Crow (1999, 2000) provided a historical review on the dominance and overdominance hypotheses. 1. Evidences from Genetic Mapping Using Molecular Markers. Xiao et al. (1995) investigated the genetic basis of heterosis in rice by backcrossing an indica/japonica RIL population to its two parental cultivars. For 12 agronomic traits, a total of 27 QTL could be detected in either backcross population. At about 82% of these QTL, heterozygotes had higher phenotypic values than their corresponding homozygotes. Ten QTL influencing grain yield components detected in both backcross populations were completely or partially dominant. RILs having phenotypic values superior to the F1 hybrid between the parental lines were found for all traits evaluated. They concluded that dominance complementation is the major genetic basis of heterosis in rice. This conclusion was strengthened by the finding that there was no correlation between most traits and overall genome heterozygosity and that there were some RILs in the F8 population having phenotypic values superior to the F1 for all of the traits evaluated—a result not expected if overdominance were a major contributor to heterosis. Digenic interaction was not evident in this study.

128

Y. XU

Stuber et al. (1992) analyzed the genetic mechanism of heterosis in a maize single-cross hybrid, ‘Mo17’ × ‘B37’, using RFLP markers. They found that most yield-QTL heterozygotes had higher phenotypic values than either homozygote class and they concluded that overdominance was the major genetic basis of heterosis in maize. This overdominance hypothesis was also supported by the results from wheat (Sun et al. 1997), Arabidopsis (Mitchell-Olds 1995), and juvenile aspen (Li and Wu 1996). However, maize researchers were careful to point out that estimates of dominance variance exceeding that for straight dominance could be due to either overdominance or linkage disequilibrium of linkage loci with favorable alleles in repulsion phase (pseudo-overdominance). Subsequently, fine mapping experiments demonstrated that the two major QTL in maize that had been identified with overdominance effects by Stuber et al. (1992) acted in a dominant manner with alleles in repulsion phase linkage by fine mapping (Graham et al. 1997). Thus, the effects at these two QTL strongly support the dominance theory of heterosis in this chromosomal region of maize. Epistasis represents non-allelic interaction, including additive-byadditive interaction, additive-by-dominance interaction, and dominanceby-dominance interaction. Because almost all heterotic traits are genetically controlled by the combined effect of many QTL, epistasis is expected to play an important role in the genetic control of these complex traits, as it does for many other quantitative traits. The first evidence for epistasis in rice is provided by Yu et al. (1997). They located QTL for yield and its components using F3 families derived from a rice hybrid, ‘Shanyou 63’. A total of 32 QTL were detected for four traits; 12 were observed in both years and the remaining 20 were detected in only one year. Overdominance was observed for most of the yield QTL and also for a few yield-component QTL. Correlations between marker heterozygosity and trait expression were low, indicating that the overall heterozygosity made little contribution to heterosis (yield). Digenic interactions, including additive-by-additive, additive-by-dominance, and dominance-by-dominance, were frequent and widespread in this population. The interactions involved large numbers of marker loci, most of which individually were not detectable on a single-locus basis; many interactions among loci were detected in both years. The authors concluded that epistasis plays a major role as the genetic basis of heterosis. The three reports discussed above for dominance, overdominance, and epistasis hypotheses were all based on the use of yield and yield components per se to measure hybrid performance without use of parental lines as a control to derive values for the midparent or betterparent heterosis. The method of measurement will identify genes for yield and yield components rather than genes for heterosis. For open-

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

129

pollinated species like maize, which has severe inbreeding depression, it is very difficult (if not impossible) to do side-by-side comparisons of the F1 hybrids with their parents. But, theoretically, this comparison is absolutely necessary if heterosis rather hybrid performance needs to be measured. Several recent investigations have been reported for genetic analysis of heterosis per se in rice. Zhang et al. (2001) designed a mating scheme that generated a fixed or “immortalized” F2 population, using a population of 240 RILs derived from the ‘Zhenshan 97’/‘Minghui 63’ cross. In this design, crosses were made between the RILs chosen by random permutations of the 240 RILs. In each round of permutation, the 240 RILs were randomly divided into two groups and lines in the two groups were paired at random without replacement to provide parents for 120 crosses. The procedure can be repeated as many times as desired and each round of permutation will pair parents for 120 crosses. This population provides opportunities for genetic mapping of heterosis per se rather than analyses based on measurements of trait performance, as long as the hybrids and the parents for each cross are planted side by side in the field, allowing measurement of heterosis for each cross. Three rounds of such random permutations, including 360 crosses, resulted in two conclusions (Zhang and Li 2002). First, all kinds of genetic effects, including single-locus heterotic effects caused mostly by overdominance, and all three forms of digenic interactions (additive by additive, additive by dominance, and dominance by dominance) appeared to play a role in the genetic basis of heterosis in the “immortalized F2” population. However, the QTL were not fine mapped, leaving open the possibility that, as in maize, the single-locus effects were due to pseudooverdominance, rather than true overdominance. Second, single-locus heterotic effects and dominance-by-dominance interaction could, together, adequately account for the genetic basis of heterosis in the F1 hybrid. Z.-K. Li et al. (2001) investigated the genetic basis of heterosis in rice using 254 RILs derived from a cross between ‘Lemont’ (japonica) and ‘Teqing’ (indica) and two BC and two testcross populations derived from crosses between the RILs and their parents plus two testers (‘Zhong 413’ and ‘IR64’). For both grain yield per plant and biomass per plant, there were significant negative effects associated with homozygous loci detected in the RIL population and a high level of heterosis in each of the BC and testcross hybrid populations. Epistasis was found for a large number of QTL pairs and a few QTL with significant main-effects were identified. Together, these QTL were responsible for over 65% of the phenotypic variation of biomass and grain yield in each of the populations, with epistatic effects explaining a much greater portion of the

130

Y. XU

variation than the main effects. As a result, most QTL associated with decreased grain yield and biomass, or with heterosis in rice appeared to be involved in epistasis, and about 90% of the QTL contributing to heterosis appeared to be overdominant. These results and the results from the genetic analysis of yield components (Luo et al. 2001) indicate that epistasis and overdominance, rather than dominance, were the major genetic basis of heterosis in rice. 2. Conclusions. As a complex character involving yield and yield components, heterosis should be genetically controlled by many genes. Although genetic study of quantitative traits has identified a limited number of QTL, each explaining a relatively large proportion of genetic variation, much more QTL could be found when multiple populations are considered. For example, a total of 63 QTL for plant height have been identified from 13 rice mapping populations and these QTL distributed on 29 chromosomal regions (Xu 2002). For a specific hybrid, heterosis is more likely genetically controlled by a relatively small number of genes; for explanation of heterosis involved in all hybrids derived from a species, a large number of QTL will be needed. Heterozygosity and its related gene interactions are the primary genetic basis for explanation of heterosis because the hybrid is heterozygous across all genetic loci that differ between the parents. Thus, the degree of heterosis depends on which loci are heterozygous and how withinlocus alleles and inter-locus alleles interact with each other. Interaction of within-locus alleles results in dominance, partial dominance, or overdominance, with a theoretical range of dominance degree from zero (no dominance) to larger than 1 (overdominance). Interaction of inter-locus alleles results in epistasis. Genetic mapping results have indicated that most QTL involved in heterosis and other quantitative traits had a dominance effect. As statistical methods that can estimate epistasis more efficiently become available, epistasis has been found more frequently and proven to be a common phenomenon in the genetic control of quantitative traits including heterosis. With so many genetic loci involved, it is unlikely that there is no interaction at all between any pair of them. When a trait is controlled by multiple QTL, their alleles of positive or negative effects (increasing or decreasing trait value) tend to be dispersed among cultivars that are used as parents for developing mapping populations (Xu and Shen 1992). Considering all genetic loci controlling a specific trait, a specific genotype (cultivar) usually has alleles of increasing effect at some loci but alleles of decreasing effect at the others, which is called allele dispersion or repulsion-phase linkage if the related loci link to each other. Identification of numerous QTL from

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

131

crosses derived from two parents with the same phenotype strongly support the hypothesis of allele dispersion. Accumulation of alleles with similar effects from allele-dispersed cultivars will result in transgressive segregation, which has been proven in two rice crosses where transgressive segregation for tiller angle has been fixed in the inbred lines selected for extreme phenotypes (large tiller angle). When similar alleles dispersed in four rice cultivars were accumulated, the fixed transgressive selections showed tiller angle three times larger than that of the original parents (Xu et al. 1998). The accumulation of dispersed alleles should add to the list of the factors that can be used to explain the response of long-term selection for oil and protein contents in maize (Dudley 2004; Walsh 2004). Allele dispersion or repulsive linkage can result in pseudo-overdominance when two such loci closely linked to each other and they are not separable by linked markers. As an example, the overdominance identified in the maize cross Mo17 × B37 by Stuber et al. (1992) has been proven to be pseudo-overdominant later (Graham et al. 1997). As reported by Z.-K. Li et al. (2001), most (~90%) QTL contributing to heterosis in rice appeared to be overdominant. It is very unlikely that each of these overdominant QTL is due to pseudooverdominance from the repulsive linkage of two completely or partially dominant QTL, although the genetic map used for heterosis mapping was less saturated. It can be concluded that two different types of allele interaction, both within-locus and inter-locus, each should play an important role in the genetic control of heterosis. Contribution of a specific locus to heterosis could be due to any single type of these interactions. When multiple loci are involved that were not taken into account in the early 1900s, various combinations of within-locus and inter-locus interactions (especially dominance-by-dominance interaction) could contribute to the genetic control of heterosis. For a specific cross and specific trait, heterosis might be explainable by any single type of these interactions. For different crosses, species, or traits, however, their heterosis has to be explained by the dominance of different degrees in combination with all possible inter-locus interactions, as indicated by Goldman (1999). Currently available information on heterosis has been provided by individual studies using specific crosses, most of which were designed for hybrid performance rather than heterosis per se. As a result, none of the mechanisms discussed above contributing to heterosis can be completely excluded. A full understanding of heterosis will depend on cloning and functional analysis of all genes that are related to heterosis. This process would be very similar to that for understanding disease resistance genes that functionally appear much simpler than heterosis.

132

Y. XU

C. Construction of Heterotic Groups 1. Germplasm Classification. Understanding the genetic structure and diversity in the gene pools with which breeders work is critical for the choice of parents, construction of heterotic groups and, thus, for the development of hybrids. Germplasm accessions can be classified based on morphological traits, geographic distribution, evolutionary and breeding history, pedigree, and/or genotypic diversity at molecular marker loci. Geographic distribution and breeding process are straightforward criteria for germplasm classification. Rice is an example. Cultivated rice, O. sativa, originated in South and Southeast Asia between 10,000 and 15,000 years ago (Chang 1984) and was dispersed throughout the world by human migration and international trade. It is believed that at least two independent domestication events gave rise to the two major subspecies of O. sativa, namely the long-grained, tropical “indica” rice, and the short-grained, temperate “japonica” rice (Oka 1988). In addition, a third group of cultivars is recognized as “tropical japonica” or “javanica” rice, characterized by bold grains, adaptation to dry, or upland, growing conditions in the tropics (Chang 1976). Rice culture was successfully introduced into the United States around 1685, when a tropical japonica cultivar, ‘Carolina Gold’, was carried to Charleston Harbor on a ship that came from Madagascar (Wilson 1979). Tropical japonica rice became the mainstay of rice production along the Atlantic Coast for 200 years and later spread to the tidal wetlands and prairie land along the Gulf Coast (Wilson 1979). The short-grained, temperate japonica types grown in California are derived largely from introductions from Japan, Korea, and China (Wilson 1979; Rutger and Bollich 1991). Germplasm accessions can also be classified based on phenotypic and/or genotypic similarity. There are several genetic distance (GD) or similarity indices that have been used as criteria for classification. A similarity index can be constructed using various types of original data. Both categorical and quantitative data have been used for phenotype-based classification. However, genotype-based classification or clustering is usually based on binary (categorical) data by scoring the presence and absence of molecular marker alleles. This scoring method has two shortcomings. First, the score does not generally reflect differences in allele sizes. For example, a difference of 2 bp and a difference of 200 bp at a SSR locus are reflected simply as a polymorphism, losing all size-related information. Second, the binary scoring method expands the data size by as many times as the number of alleles. When a clustering program has a limitation on the size of the dataset, the data has to be cut to fit the program. A better scoring method retains information about allele sizes, and similarity is then calculated for each pair of samples based on allele

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

133

sizes per se or shared allele frequency. Allele sizes can be easily converted into allele presence-absence, whereas the latter is not convertible. In rice, a genetic similarity index was computed based on SAFs (S) between each pair of rice accessions, and then 1-S was used as the GD to construct the dendrograms depicting genetic relationships among these rice accessions (Y. Xu and S. R. McCouch, Cornell University, unpubl. data). Different types of molecular markers are suited for different levels of classification. RFLP marker-based analysis of the whole collection of 236 rice cultivars identified two major groups corresponding to indica and japonica, respectively, whereas most of the U.S. cultivars belonged to a different group. SSR markers that can distinguish closely related accessions were then used to subgroup the U.S. cultivar collection. Two subgroups were identified, representing two different types of grain shapes, long grain and medium- and short-grain, with average grain lengths 9.7 mm and 8.4 mm, respectively. The major longgrain cultivars released in Texas tended to cluster in the long-grain subgroup while the medium- and short-grain cultivars released in California formed a cluster in the short/medium subgroup. Within each subgroup, some cultivars were closely clustered. They were either closer in pedigree or more similar in morphology. Many rice accessions released in Texas clustered on the top of the long-grain subgroup. There is very limited heterosis in the hybrids derived from cultivars within this subgroup. Southern U.S. cultivars are mostly derived from tropical japonica germplasm that traces its ancestry to Indonesia. This ancestral source, coupled with selection during breeding, makes the U.S. cultivars distinct from other indica or japonica cultivars, though they share many genes from these groups. For example, these U.S. cultivars may have long grains as typical of indica rice, although many have intermediate or short grains. Also akin to the tropical japonica germplasm native to Indonesia, many of these cultivars were found compatible with both typical indica and typical japonica without much hybridization barrier (Xu et al. 1989; Gu and Tang 2001). The intermediate-type cultivars and the cultivars with wide-compatibility genes, as discussed in Section VC, have been frequently used to overcome the indica-japonica crossing barrier, resulting in the diffusion of genes and disappearance of reproductive isolation between the two subspecies. These wide-compatible cultivars belong to the third major group or subspecies of rice cultivars. Representation of this group will greatly affect germplasm classification and heterosis studies. Using male sterility in hybrid seed production provides the opportunity to produce large amounts of seed. Simultaneously, it raises concern that cytoplasm uniformity of CMS lines could result in genetic vulnerability, as happened in maize in 1970 because of the extensive use of

134

Y. XU

T-CMS. In rice, sorghum, and rye, only one or few CMS sources for each crop have been widely used for hybrid breeding. Genetic characterization and classification of alternative CMS sources will help establish diverse CMS pools and avoid CMS uniformity. To facilitate this, marker profiles could be used to classify genotypes. Molecular marker alleles specific to cultivar groups could be identified and then used to classify other germplasm accessions such as those containing different sources of CMS. Marker information could be exploited better for germplasm classification if it were integrated with phenotypic information, including information about CMS. 2. Concept of Heterotic Groups. The concept of heterotic groups or heterotic pools was first developed in maize based on the observation that inbreds selected out of certain populations tended to produce better performing hybrids when crossed to inbreds from other groups (Hallauer et al. 1988). This recognition resulted from the systematic crossing of thousands of inbred lines from different source populations and evaluation of the hybrids (Havey 1998). In the review of capturing heterosis in forage crop cultivar development, Brummer (1999) indicated that the key to successful semihybrid production is to keep heterotic groups separate, only intercrossing them for testing and release. Merging of heterotic groups into a larger breeding population results in the loss of interacting linkats (alleles), limiting potential yield advance. Breeding highly heterotic hybrids largely depends on selection of desirable parents as a prerequisite for most hybrid breeding programs, and thus depends on genetic diversity in the germplasm resources available to plant breeders. For example, commercial maize hybrids are typically made between inbreds from opposite, complementary heterotic groups. Therefore, construction or development of heterotic groups has been one of the key components in hybrid breeding for many crops. Introgressing exotic germplasm is often suggested as an approach to increase genetic differences between opposing heterotic populations, thereby potentially increasing heterotic response. An understanding of heterotic relationship between populations is needed to exploit exotic germplasm intelligently. Melchinger and Gumber (1998) reviewed the development of heterotic groups in five major crops with different pollination systems: allogamous—maize and rye, partially allogamous—faba bean and oilseed rape, and autogamous—rice. In this section, discussion will be devoted to construction of heterotic groups based on both conventional methods and molecular marker information. A possible explanation for heterotic groups is that populations of divergent genetic backgrounds have unique allelic diversity that may have arisen from founder effects, genetic drift, or the accumulation of

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

135

unique allelic diversity by mutation or selection. Significantly greater heterosis could result from this genetic diversity by specific interallelic interactions (overdominance), repulsion-phase linkage among loci showing dominance (pseudo-overdominance) (Havey 1998), and/or inter-locus interaction (epistasis). Apparently, the most obvious potential heterotic groups are either geographically separated populations or separate subspecies and ecotypes. Melchinger and Gumber (1998) recommended the following criteria for the identification of heterotic groups and patterns in descending order of importance: (1) high mean performance and large genetic variance in the hybrid population to ascertain future selection response; (2) high per se performance and good adaptation of both or at least one of the parental heterotic groups; (3) low inbreeding suppression in the source materials for the development of inbreds; and (4) a stable CMS system without deleterious side effects, as well as effective restorers and maintainers, if hybrid breeding is based on CMS. 3. Construction of Heterotic Groups Based on Hybrid Performance. With large numbers of inbred or open-pollinated lines or populations available, it is unfeasible in most crops to make diallel crosses and produce sufficient F1 seed for multi-environment field-testing. Therefore, Melchinger and Gumber (1998) suggested a multi-stage procedure to identify heterotic groups, which consists of the following steps: (1) grouping the germplasm based on genetic similarity; (2) selection of representative genotypes (e.g., two or four lines or one population) from each subgroup for producing diallel crosses; (3) evaluation of diallel crosses among the subgroups together with parents in replicated field trials; and (4) selection of the most promising cross combinations as potential heterotic patterns using the identification criteria. If established heterotic patterns are available, using selected elite genotypes from them as testers for the production and evaluation of the germplasm to be classified is recommended. Based on the testcross performance, populations or lines having similar combining ability and heterotic response could be merged to constitute a new independent heterotic group, if they behave differently from the existing heterotic groups; however, if their behavior is similar to an existing heterotic group, they could be merged with it to enlarge its genetic base. Heterotic patterns in many crop species have been established solely based on the large numbers of testcrosses and breeding experience, without the use of DNA-based markers. Ron Parra and Hallauer (1997) reviewed heterotic patterns used in the major maize production regions of the world. Some patterns have had importance in specific production regions. Others have been exploited on several continents, for example, the heterotic patterns based on ‘Reid Yellow Dent’ (RYD) and ‘Lancaster Sure Crop’ (LSC) from the

136

Y. XU

temperate United States, and ‘Tuxpeño’ and Estación Tulio Ospina’ from tropical Mexico and South America. Two heterotic groups from which inbreds commonly are selected and used to produce superior maize hybrids are Iowa (B) Stiff Stalk Synthetic (BSSS) and derivatives of LSC (Darrah and Zuber 1986; Gerdes and Tracy 1993). Although both populations are primarily comprised of southern dent germplasm, LSC has more northern flint germplasm than BSSS (Smith 1986; Gerdes and Tracy 1993). With genetically balanced sets of crosses, inter-group hybrids outyielded the respective intra-group hybrids by 21% in RYD × LSC crosses (Dudley et al. 1991) and by 16% in Flint × Dent crosses (Dhillon et al. 1993). In both studies, the percentage of increase in heterosis for yield of inter-group over intra-group crosses was about twice as large as for the hybrid yield itself. Intrapopulation maize hybrids have been developed and commercialized in tropical maize. Some of the earlier hybrids were intrapopulation hybrids. The first hybrid (‘Suwan 2301’), developed by Kasetsart University in Thailand, was an intrapopulation interline hybrid derived from ‘Suwan-1’. Data suggested that interpopulation interline hybrids are generally superior (Han et al. 1991). Even when the populations are not heterotic, the interpopulation interline hybrids give superior performance. However, populations such as ‘AED’ × ‘Tuxpeno’ (P44), ‘Tuxpen’ (P21), and ‘La Posta’ (P43) have produced outstanding intrapopulation interline hybrids. It is possible to produce good hybrids from the same population provided it has high per se performance and high general combining ability (Vasal et al. 1999). The systematic search for suitable heterotic patterns by Hepting (1978) laid the foundation for hybrid breeding in rye. Based on the hybrid performance and heterotic derivation in a complete diallel among seven open-pollinated populations, he found that cross combinations involving populations from the two most widely used germplasm groups, Petkus and Carsten, were most promising for grain yield. In fact, all rye hybrids released in Germany since 1985 are based on the Pampa CMS system and are of the Petkus × Carsten type. In oilseed rape, Grant and Beversdorf (1985) evaluated diallel crosses among six OPCs of European and Canadian spring rapeseed and suggested European spring × Canadian spring rapeseed as a promising heterotic pattern. Similarly, Lefrot-Buson et al. (1987) found crosses between winter rapeseed of European × Asian origin to be most productive. The results of diallel crosses among seven spring rapeseed cultivars of European, Canadian, and Asian origin confirmed that crosses between European × Asian and Canadian × Asian spring rapeseed exhibited higher heterosis than crosses between parents originating from the same region (Brandle and McVetty 1990).

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

137

Four distinct heterotic groups within sunflower are now being utilized by breeders throughout the world (Vear and Miller 1993). OPCs developed in Russia are used in deriving female maintainer inbred lines. The U.S. restorer group, derived by crossing wild annual species of sunflower with cultivated lines, is a distinct source of disease resistance and fertility restorer genes. Romanian female lines, along with their South African derivatives, are used throughout the industry. Also used are the Argentinean INTA OPCs for developing female lines (Miller 1999). Rice might be the only crop where hybrids are widely grown but very few studies on heterotic groupings have been reported. Heterosis in rice has been utilized largely through CMS. Fortunately, rice breeders in China identified the restorers for CMS from geographically distant rice cultivars from southeastern Asia and used them in hybrid rice breeding. This resulted in high levels of heterosis among intra subspecies (indica × indica) hybrids. A large-scale screening of diverse CMS maintainers and restorers provided some clue as to heterotic pattern. Three ecotypes from different subspecies, indica, japonica, and javanica, have different morphological and physiological characteristics and ecogeographical distribution and, therefore, serve as a basis for defining distinct heterotic groups. As summarized by Yuan (1992a), heterosis for grain yield in crosses among the three rice ecotypes has the following trend: indica × japonica > indica × javanica > javanica × japonica > indica × indica > japonica × japonica. This mirrors the current situation of heterotic pools in rice. It is well known to hybrid rice breeders that a high level of heterosis results from crosses between CMS lines bred in China and restorer lines derived from southeast Asian indica cultivars, which is the heterotic pattern for indica × indica hybrids. Wheat breeders lag much behind their colleagues in other crops in establishing heterotic pools (Jordaan et al. 1999). Heterotic groups have not been well described in vegetables either. 4. Construction of Heterotic Groups Using Molecular Marker Information. DNA-based markers may be used to classify parental lines into different heterotic groups, each with a high level of similarity in genetic backgrounds. This reveals genetic diversity at the whole genome level, and helps identify effects of selection, genetic drift, and mutation. Molecular markers have been playing an increasingly important role in the construction of heterotic groups since the 1990s. Most reports are focused on maize, wheat, barley, and canola. Because marker-based groupings reflect the genetic differences among parental lines, they can contribute to parental improvement and to effective selection for heterotic hybrids. In general, heterotic groups constructed on the basis of

138

Y. XU

marker information match up very well with pedigrees, but have the advantage that missing historical information, such as the incomplete pedigree information or ambiguous pedigree, will not affect the markerbased method. Using RFLP markers, Meng et al. (1996) grouped 46 cultivars of Brassica napus, originating from China, into six groups, and a very significant difference was found between the cultivars from China and a group of six cultivars from B. napus L. var. oleifera subvar. biennis originating from Europe. They suggested that the latter could be used to broaden the genetic diversity of breeding populations of oilseed rape. In maize, different types of molecular markers have been successfully used to differentiate heterotic groups with results that are consistent with pedigree-based grouping (Mumm and Dudley 1994; X. Liu et al. 1997; Peng et al. 1998; Wu et al. 2000). Based on heterosis and combining ability analyses using cultivars from different heterotic groups, Peng et al. (1998) proposed seven heterotic patterns for the utilization of maize heterosis. Divergence at molecular marker loci has been useful in assigning maize inbreds to known heterotic groups previously established in breeding programs and the molecular information agreed with pedigree information (Lee et al. 1989; Melchinger et al. 1991; Messmer et al. 1993). In rice, Zhang et al. (1992) analyzed 12 indica and 14 japonica cultivars using RFLP markers. The average GD measured by RFLPs between indica lines was three to four times higher than that between japonica, which confirmed the results based on the morphological studies. Using 160 RFLP markers and 21 wide-compatibility cultivars and three indica and three japonica cultivars, Zheng et al. (1994) constructed a dendrogram tree and discussed the potential of wide compatibility in hybrid breeding using indica/japonica crosses. Based on diallel crosses among eight indica lines representing the parents of the best-performing commercial rice hybrids grown in China, Zhang et al. (1995) studied molecular divergence and hybrid performance. Their results suggest the existence of two heterotic groups within indica, one comprised of rice strains from southern China and the other comprised of strains from Southeast Asia. Using two types of molecular markers, RFLPs and AFLPs, Mackill et al. (1996) obtained similar grouping results. Using RAPD and SSR markers, Xiao et al. (1996a) separated the ten parental lines into two major groups that correspond to indica and japonica subspecies. These researches indicated that molecular markers are useful tools in detection of genetic diversity between parental cultivars. The results from barley (Melchinger et al. 1994) and wheat (Sun et al. 1996; Ni et al. 1997) also supported the conclusion that DNA markers are very useful tools for construction of heterotic groups.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

139

5. Future Directions. Heterotic groups are the backbone of successful hybrid breeding. Decisions regarding the definition and utilization of heterotic groups are of fundamental importance and must be made at the beginning of a crop improvement program. In most cases, breeding for heterosis without knowledge of heterotic patterns has proven to be a hitor-miss approach (Jordaan et al. 1999). It is evident from the review of various studies that adapted populations, isolated either by time and/or space, are the most suitable candidates for promising heterotic patterns. Genetic diversity can be related to geographic origin of parental lines. The geographical variation can be related to ecological and environmental variations that, in turn, dictate survival fitness, created by spontaneous and induced genetic variation in natural and directed-selection situations. Consequently, the parental lines derived from different geographic origins are considered to have more genetic diversity than those derived from the same geographic origin. International breeding efforts through various collaborative breeding programs in rice, wheat, and maize have been very successful in breeding both hybrid and non-hybrid cultivars. However, internationalization of plant breeding efforts and massive exchange of unimproved and improved germplasm throughout the world have altered the genetic structure and adaptation of germplasm accessions with which breeders have been working. As a consequence, differences in geographic origin of the parental lines may not always reflect genetic diversity among them. On the other hand, extensive hybridization practiced in several international and national crop breeding programs has created new forms of genetic diversity, and one can expect to find substantial genetic diversity among parents from the same geographical origin (Virmani 1996). The negative effect of using distant crosses is the confusion of heterotic groups existing among cultivars of different geographic origins. For example, breeding wide-compatible inbred cultivars as a bridge for harnessing indica/japonica heterosis in rice has reduced heterosis compared to what would be expected from crosses between typical indica and japonica cultivars. Therefore, it is important to keep in mind that we are not disturbing the current heterotic groups, which have been established either naturally or creatively, when we use distant crosses in hybrid breeding programs. Heterotic groups should not be considered as closed populations, but should be broadened continuously by introgressing unique germplasm to warrant medium- and long-term gains from selection. Heterotic groups consisting of poorly utilized and unadapted germplasm should be enhanced through joint public-private breeding ventures. Different phenotypes may or may not reflect divergent genetic backgrounds.

140

Y. XU

Phenotypically different populations may possess the same genetic background, and divergent phenotypes may be conditioned by allelic differences at relatively few loci (Havey 1998). MAS can be useful in creating, maintaining, and improving heterotic groups. As discussed above, marker-based grouping of germplasm and breeding populations will help establish heterotic groups that hold maximum genetic diversity between groups but minimum diversity within groups. Identification of marker alleles that are specific to each heterotic group will help keep them genotypically separated. MAS can be used to improve the existing heterotic groups through introgressing target genes from one heterotic group or outsource germplasm to another with minimum linkage drag from the donor. As we discussed previously, MAS will help breeders realize their goals without linkage drag and unwanted genetic background. D. Hybrid Prediction 1. Reasons for Hybrid Prediction. Hybrid breeding includes two major procedures: breeding parental lines and selection for the best hybrids from the cross combinations of those parental lines. These procedures involve a large amount of work for field evaluation, testcrossing, and progeny tests. Breeders continually have to decide which experimental single crosses to test, which advanced hybrids to recommend for further testing or commercialization, and which inbred parents to cross to form new base populations for inbred/population development (Bernardo 1999). Suppose a breeder has 100 inbreds from heterotic group 1 and 100 inbreds from heterotic group 2. There are 10,000 possible (group 1 × group 2) single crosses. For developing new hybrids, there are 495,000 possible (group 1 F2 ) × (group 2 tester) combinations, and 495,000 possible (group 1 tester) × (group 2 F2) combinations, if testcrossing starts from the F2. Due to limited resources, breeders are unable to test all combinations in all environments of interest but may test a limited set of single crosses and F2 × tester combinations. Typically, <1% of the maize single crosses tested by a breeder eventually become commercial hybrids (Hallauer 1990). Therefore, predicting hybrid performance has always been a primary objective in all hybrid-breeding programs. Methods for predicting the performance of single crosses would greatly enhance the efficiency of hybrid breeding programs. Development of a reliable method for predicting hybrid performance and/or heterosis without generating and testing hundreds or thousands of single cross combinations has been the goal of numerous studies using marker data and combinations of marker and phenotypic data, particularly in maize and rice. It is reasonably believed that heterosis originates, in some way, from the genetic differences or heterozygosity between the parents. Theoret-

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

141

ically, hybrid performance is equal to the average parental performance plus heterosis. In the past several decades, hybrid prediction has been largely based on the evaluation of genetic diversity among parental lines. It has been expected that understanding the relationship between heterozygosity/parental difference and heterosis would help predict hybrids. The development of molecular marker techniques has provided new tools for hybrid prediction and DNA markers have been used extensively in investigating correlations between parental GD and hybrid performance. 2. Statistical Methods for Hybrid Prediction. The best linear unbiased prediction (BLUP) procedure has been used for decades for evaluating the genetic merit of animals, especially dairy cattle. Intrapopulation, additive genetic models have traditionally been used for BLUP in animal breeding (Henderson 1975). During the last several years, Bernardo has attempted to use BLUP in maize breeding with interpopulation genetic models that involve both GCA and SCA (Bernardo 1994, 1996). Results have indicated that BLUP is useful for routine prediction of single-cross performance. The predicted performance of single crosses may subsequently be used to predict the performance of F2 × tester combinations, three-way crosses, or double crosses. Along with the pedigree relationship, the BLUP method can use trait data, or both trait and marker data, for prediction. 3. Genome-wide Heterozygosity and Hybrid Prediction. In most cases, isozyme-based GD estimates are positively associated with hybrid performance for grain yield (for a review, see Stuber 1994); however, it remained unclear whether the low correlations between both measures observed in most cases were due to poor coverage of the genome or whether other causes were involved. The availability of a large number of DNA-based markers provided the opportunity for genome-wide surveys of heterozygosity and laid the foundation for a better understanding of the relationship between heterozygosity at marker loci and heterosis. The relationship between parental genetic divergence and hybrid performance was first studied in maize. Variability for molecular markers generally agreed with pedigree information and assignment (based on hybrid performance) to known heterotic groups (Smith et al. 1990; Dudley et al. 1991; Melchinger et al. 1991); however, variability at molecular marker loci was ineffective in predicting specific hybrid performance from crosses among maize inbreds (Lee et al. 1989; Melchinger et al. 1992). Some reports indicated high correlation between hybrid performance/heterosis and parental GDs or the degree of heterozygosity (Smith et al. 1990; Lee

142

Y. XU

et al. 1989; Stuber et al. 1992), while others revealed very weak correlations (Godshalk et al. 1990; Dudley et al. 1991). Considering all these and other studies in maize, the correlation depends largely on the plant materials and their origins and suggests that specific loci are involved in heterosis rather than heterozygosity per se. Correlations between single-cross performance and molecular marker diversity for unrelated parental inbreds have been too low to be of any predictive value (Godshalk et al. 1990; Melchinger et al. 1990; Dudley et al. 1991). Molecular-based GD estimates also failed to predict superior hybrid performance in oat (Moser and Lee 1994), soybean (Gizlice et al. 1993), and chickpea (Sant et al. 1999). 4. Hybrids Are More Predictable Within than Between Heterotic Groups. Let us first consider intra-heterotic crosses. Correlations between heterozygosity/GD and hybrid performance/heterosis varied for hybrids between lines that belong to the same heterotic group (within-group hybrids). In maize, correlations of GD with F1 performance and heterosis were significant and positive for all traits of within-group hybrids, flint × flint crosses (Boppenmaier et al. 1993). This was supported by Benchimol et al. (2000) using 18 tropical maize inbred lines where correlations of parental GDs with single crosses and their heterosis for grain yield were higher for line crosses from the same heterotic groups. The same conclusion was reported by Zhao et al. (1999) using diallel crosses derived from 11 elite rice lines where the correlations of marker heterozygosity with hybrid performance and heterosis were high in crosses within subspecies. In other cases, however, weak or no correlation was found for within-group hybrids. Examples include weak or no significant associations of GD with F1 performance and mid-parent heterosis in soybean (Cerna et al. 1997), wheat (Martin et al. 1995), and U.S. longgrain rice cultivars (Saghai Maroof et al. 1997). These results may be due to the low levels of heterosis in these varietal groups. Weak or no correlation was found for hybrids between lines that belong to different heterotic groups (between-group hybrids). In maize, correlations of GD with F1 performance and heterosis were not significant for the subsets of flint × dent and dent × dent crosses (Boppenmaier et al. 1993). Using 18 tropical maize inbred lines, Benchimol et al. (2000) found that correlations of parental GDs with single crosses and their heterosis for grain yield were low for line combinations from different heterotic groups. In rice, Xiao et al. (1996a) reported that yield potential and its heterosis showed significantly positive correlations with GD for indica × indica or japonica × japonica crosses, but the correlations were not significant for indica × japonica crosses. This was confirmed by Zhao et al. (1999) that very little correlation was detected in intersubspecific crosses using diallel crosses derived from 11 elite rice cultivars.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

143

A theoretical study conducted by Charcosset and Essioux (1994) supported the reports discussed above. The correlation between heterozygosity at marker loci and heterosis was investigated for (1) hybrids between lines that belong to the same heterotic group (within-group hybrids); (2) hybrids between lines that belong to different groups (between-group hybrids); and (3) all hybrids, both within- and betweengroups. Within a group, significant correlations may be detected because of linkage disequilibrium generated by drift. At the between-group level, no correlation is expected since linkage disequilibium should differ randomly from one group to another, which is consistent with the results of recent experiments. When all hybrids are considered simultaneously, divergence of allelic frequencies among groups for the markers and the QTL produces a correlation between heterosis and heterozygosity at marker loci. This correlation increases with the number of markers that are considered. Based on results from various studies in maize, Melchinger (1993) summarized the relationship between parental GD and mid-parent heterosis (MPH) in a schematic representation. For crosses among related lines, there exists a tight association between GD and MPH for yield characters because both measures are a linear function of coancestry, f, and, thus, decrease with increasing f. For intra-group crosses, the correlation r (GD,MPH) is generally positive, too. This can be explained by hidden relatedness between some parents considered to be unrelated based on their pedigree, and the presence of the same linkage phase between QTL and marker loci in the maternal and paternal gametic arrays of intra-group hybrids, which results in a positive covariance between GD and MPH (Charcosset et al. 1991). In contrast, no significant association between both measures exists for inter-group hybrids. In this case, the maternal and paternal gametic arrays may differ in the linkage phase for many QTL-marker pairs; as a consequence, positive and negative terms cancel each other in their net contribution to covariance (GD,MPH), resulting in a low or zero correlation (Charcosset and Essioux 1994). High estimates of r (GD,MPH) could be expected if correlations are calculated across different types of crosses due to group effects. This is because both GD and MPH are expected to increase from crosses among related lines to intra-group crosses and further to inter-group crosses (Melchinger 1999). 5. Heterosis-associated Markers and Hybrid Prediction. It has been common practice in most studies to determine GD or heterozygosity estimates from a set of DNA markers chosen for good coverage of the entire genome but not for linkage to genes influencing heterosis of the target trait. Theoretical investigations (Charcosset et al. 1991) and computer

144

Y. XU

modeling (Bernardo 1992) demonstrated that with intra- and inter-group crosses the correlation between GD and MPH is expected to decrease if genes influencing heterosis are not closely linked to markers used for calculation of genetic estimates and vice versa if markers employed for calculation of GDs are not linked to genes controlling the trait. Hence, increasing the marker density alone will not necessarily improve the ability to predict MPH by GD estimates; rather, markers must additionally be selected for tight linkage to genes affecting heterosis of the target trait in the germplasm under study. This is corroborated by comparison of results obtained with 209 AFLPs vs. 135 RFLPs (Ajmone Marsan et al. 1998) and a study by Dudley et al. (1991). Using these associative loci will help establish strong correlations between heterozygosity and heterosis. However, allelic differences at marker loci do not assure allelic differences at linked loci for heterosis. For a limited number of markers to be useful as predictors for hybrid performance, the effects of alleles at the loci linked to specific marker alleles must be ascertained (Stuber et al. 1999). Zhang et al. (1994a) proposed two statistical parameters: general and specific heterozygosity, to measure genotypic heterozygosity. The former is the heterozygosity calculated from the GDs between the parents using all possible markers, and the latter is that from using marker loci that are significantly associated with the traits of interest revealed by single factorial analysis of variance. The results from rice indicated that there was a weak correlation between general heterozygosity and heterosis but a significant correlation between specific heterozygosity and heterosis for yield and biomass. As a result, heterosis could be predictive to some extent (Zhang et al. 1994a, 1995). Joshi et al. (2001) studied the correlations between GD, hybrid performance, and heterosis in rice using the markers associated with grain yield and other traits reported by Xiao et al. (1996b) and McCouch et al. (1997) and 21 F1 hybrids derived from two CMS lines and 14 restorer lines. There was no significant correlation for grain yield. Two factors contributed to this result: loose marker-trait associations and different genetic systems for controlling heterosis and grain yield in the materials used by different researchers. As a result, markers identified for grain yield from one population cannot be used to predict heterosis for another. Successful heterosis prediction based on molecular markers will depend on how many genes have been identified from different genetic backgrounds and how these genes interact with each other to control heterosis. The complexity of these interactions explains why there are no successful examples of heterosis prediction. Attempts to identify QTL-marker associations by use of multiple regression techniques, in which observed phenotypic values for a fixed

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

145

set of hybrids are regressed on their coded marker genotypes, must be regarded with great caution, because the number of variables from which the regressors are selected is usually as large or even greater than the number of phenotypic observations. Moreover, comparison of QTL mapping results from different populations in maize (Stuber 1995; Lübberstedt et al. 1998) and rice (Xu 2002) suggested that QTL regions affecting a given trait are not necessarily consistent across different germplasm accessions. 6. Favorable Allele Combination and Hybrid Prediction. Although the importance of genetic diversity to heterosis has been known for several decades, heterogenic gene combinations may not always lead to heterosis and heterosis is ultimately dependent upon the balance between favorable and unfavorable interactions of genes. It is reasonably inferred that heterosis could be caused by specific gene combinations derived from the two parents. Those genes may simultaneously produce different genetic effects in different genetic backgrounds. So, for parental improvement and hybrid prediction, investigating the specific gene combinations that contribute to heterosis should be more important than studying any single gene or QTL. Using 99 half-diallel rice hybrids derived from nine CMS lines and 11 restorer lines, Liu and Wu (1998) found that four favorable alleles and six favorable heterotic patterns on the parental lines significantly contributed to the heterosis of their hybrids for grain yield, whereas six unfavorable alleles and six unfavorable heterotic patterns significantly reduced heterosis. They suggested that optimal hybrids with superior grain yield could be developed by assembling those favorable alleles into, and removing the unfavorable alleles from, their parental lines. 7. Conclusions. There are several conclusions that can be drawn from the numerous investigations on the relationships between heterozygosity and GD with hybrid performance and heterosis. First, the higher the heterozygosity between the parents, the stronger the heterosis. Second, using more markers alone will not improve the prediction. Third, prediction is possible using markers known to be associated with hybrid performance or heterosis if the association is used to predict performance of a hybrid derived from the same heterotic pattern. Fourth, genetic variation (the presence of heterosis) is a prerequisite for prediction. Fifth, the relationship of heterozygosity with heterosis and with hybrid performance will be different if the two involve different genes. The last conclusion was supported by results of Zhu et al. (2001) that heterosis was highly significant but hybrid performance was not when 57 rice accessions from six ecotypes and their hybrids were genotyped

146

Y. XU

by 48 SSR and 50 RFLP markers. In conclusion, heterozygosity at general marker loci can be used to construct heterotic groups. It is anticipated that prediction could be possible if heterozygosity is derived from specific marker loci that are associated with heterosis and hybrid performance and all possible associated loci have been identified and their effects and interactions clearly defined. Considering the fact that only heterotic crosses are of commercial importance and of interest to the breeder, the practical value of the genetic distance approach for prediction of heterosis and hybrid performance is limited (Vuylsteke et al. 2000). This is true for some crop species like maize. For rice, however, the reproductive barrier between the two subspecies, indica and japonica, has enforced a limitation on the utilization of indica/japonica heterosis, although the use of the widecompatibility gene(s) has had a great impact on the limitation. Hybrid breeding for indica rice has been based on crosses within the indica group. The strong relationship between the heterozygosity at marker loci and heterosis within the indica group as reported before (Xiao et al. 1996a) indicates that GD estimates based on molecular markers could be very useful in assigning indica cultivars into different subgroups for hybrid indica rice development. Screening for heterosis-related molecular markers as suggested by Melchinger et al. (1990), using specific heterozygosity proposed by Zhang et al. (1994a), and identifying favorable combinations of allele and heterotic patterns (Liu and Wu 1998) are among the approaches that could be exploited further to improve the prediction of hybrid performance/heterosis using molecular markers. Understanding genetic variation among cultivars to be tested and identifying markers associated with heterosis and heterosis-related traits are two important components in hybrid prediction. We should keep in mind that markerheterosis associations identified in one cross may not be suitable for selection in others because heterosis could be controlled by many genes and each cross has different genes and gene combinations in action. Despite their low values, the inbred-hybrid yield correlations were positive. They indicated a tendency for high-yielding inbreds to produce high-yielding hybrids. Hybrid breeding is always accompanied by the improvement of parental lines. Modern maize inbreds, grown at today’s high density, can yield nearly as much as hybrids of the 1930s (Duvick 1984; Meghi et al. 1984). Duvick (1999) has suggested that if as much effort had been put into improvement of OPCs as has been devoted to hybrid improvement over the years, the gap between the best hybrids and the best OPCs might be less than what it currently is. Some authors even argue that OPCs might be superior to hybrids (Lewontin and Berlan 1990), but their assumption is not backed up by data. No experiments

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

147

have been reported that test the theory that OPCs could be improved at the same rate as hybrids (or at a higher rate), if equal effort were expended on each kind of breeding (Duvick 1999). This assumption, if true, would greatly support the dominance theory of heterosis. In conclusion, the potential application of DNA markers in hybrid breeding depends very much upon whether divergent heterotic groups have been established or not and upon crop species. If well-established heterotic groups are unavailable, marker-based GD estimates can be used to avoid producing and testing crosses between closely related lines. Furthermore, crosses with inferior MPH could be discarded prior to field-testing based on prediction. Another potential application exists. If new lines of unknown heterotic pattern or inbreds developed from crosses between parents from different heterotic groups (e.g., commercial hybrids) are to be evaluated for testcross performance, GD estimates could assist the breeder in the choice of appropriate testers for evaluating the combining ability of the lines. However, with regard to the typical situation of hybrid breeding, in which crosses are produced between lines from genetically divergent heterotic groups, GD estimates based on an unselected set of DNA markers alone are not promising for predicting hybrid performance or any of its components.

X. SEED QUALITY ASSURANCE Production of high-quality hybrid seed is crucial to the success of a heterosis-breeding program. Two major factors influencing seed quality are purity and viability. Hybrid seed do not show large differences from non-hybrids in viability despite heterosis for some traits such as germinating ability or disadvantages such as short seed life for hybrid rice because of the partially enclosed hull. In this section, only seed purity will be discussed. Two major factors contribute to the purity of hybrid seed: off-types and false hybrids. Off-types includes non-hybrid seed resulting from non-parental plants in parental multiplication and hybrid seed production, and false hybrids includes non-hybrid seed resulting from the selfing of the female parent due to instability of male sterility if a male sterile line is involved in seed production. A. Off-type Off-type in hybrid seed and grain production is a potential quality problem. Traditionally, off-types are defined as individual plants that are phenotypically different from the plants developed by breeders. They may result from mechanical mixtures, outcrossing, mutation, or residual

148

Y. XU

genetic variation. Off-types may be found in a parental line or only visible in the hybrid. The presence of off-type plants will reduce the uniformity of the crop and thus reduce its productivity and quality. Phenotypic off-types can be easily rogued if there are not too many. Besides the offtypes that are visible phenotypically, many off-types are genetically different from the typical plants and they are hard to distinguish visually. Genotypic off-types may impose more severe effect on hybrids, which could be one of the major reasons for reduced performance of hybrid cultivars or inbred lines. They also impose an effect on the parent, which could be exaggerated by multiplication of the parents themselves and their progeny. Molecular marker technology provides a powerful way for distinguishing both the phenotypic and genotypic off-types from hybrid plants and their parents. Marker-trait associations and high-resolution molecular markers such as SSRs could be used to distinguish two plants with very similar genetic backgrounds. With ten or more SSR markers, rice breeders can identify distinct off-types from their breeding populations and hybrid seed bulks and obtain detailed genotypic information such as where the off-type genotypes come from and what the proportion of the off-types is to the typical plants. A selection and purification decision can be made to refine their breeding materials and hybrids. B. False Hybrids False hybrids come from the selfing of the female parent because a male sterile line becomes fertile due to instability of male sterility. This has happened to two-line hybrid rice when the EGMS female parent becomes fertile during seed production. As discussed previously, a TGMS line can serve as a sterile line under one environmental condition and can propagate itself under another. The ability to maintain sterility makes TGMS lines practicable as a female to cross with any other commercial rice cultivars. However, most TGMS lines require a specific temperature to maintain their sterility. Abnormal weather can bring the temperature down below the critical temperature that is required for conversion of TGMS lines from sterility to fertility (or simply called fertility conversion), which makes TGMS lines fertile or partially fertile in locations where they are sterile in normal years. This results in self- or sib-seed contamination, a potential problem for seed production of two-line hybrid rice. The mixture of real hybrids with selfed seeds from the EGMS line cannot be used in rice production, resulting in a great loss to seed producers or rice growers. As a way of insuring the seed production, marking the seeds from EGMS lines using genetic markers can help identify and remove the false hybrids from the mixture.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

149

1. Approaches to Eliminating False Hybrids. To eliminate pseudo hybrids from seed production, three strategies can be used: (1) breeding herbicide mutants and use of the herbicides to kill false hybrids, (2) use of morphological markers (such as purple leaf mutants) to distinguish the false hybrids and remove them by hand, and (3) use of mutants with faded green leaves to get rid of false hybrids. Morphological and chemical markers have been investigated as a way to identify specific seeds/plants from a mixture. In rice, several morphological markers, such as purple leaves (Mou et al. 1995) and pale green leaves (Dong et al. 1995), have been used for marking EGMS lines. These markers can be used to distinguish real F1 hybrids from selfed seeds (false hybrids) at the seedling stage. However, removing the purple false hybrid seedlings must be done by hand, which is laborintensive and there is no assurance that all false hybrids will be completely eliminated. Faded green mutants with chlorophyll deficiency during the seedling stage cannot compete with normal plants and will die in the densely planted seedling nursery but will grow well in a less dense planting. Selective responses of plants to herbicides can be used to identify a specific type of plant from a mixture and the response can be exploited as a “marker.” There are two different types of herbicides that can be used as selective chemicals to remove false hybrids in rice and other crops: herbicides that can selectively kill a crop and ones that are safe to the crop. Prerequisites for use of herbicide mutants in hybrid seed production include availability of herbicides (registered chemicals), low cost of herbicides, response to low concentration of herbicides, long sensitivity duration, minimal effect of environmental factors/weather, and easy identification of the herbicide response. 2. Herbicide Mutants in Rice. In rice, there are several herbicide mutants reported, including IMI mutants such as IMI-Cypress, resistant to imidiazone; Bar (phosphinothrin acetyltransferase), resistant to Basta (Liberty); EPSP synthase, resistant to Roundup; and Norin 8 mutant, recessively sensitive to bentazon at the flowering stage. Zhang et al. (1998) transferred the bar gene into a rice restorer line to make it resistant to the herbicide Basta, which can be used to selectively kill false hybrids. When a herbicide is safe to use with normal rice cultivars, recessive sensitivity to the herbicide can be used to selectively kill the seedlings from the selfed seeds of the female once it possesses the sensitivity. The herbicide-sensitive rice mutant obtained by Mori (1984) through radiation was lethal to bentazon, which was controlled by a recessive gene.

150

Y. XU

To prevent seed production from selfing contamination, Zhang et al. (2002) developed a system to secure seed purity using a herbicide sensitive TGMS mutant, M8077S, obtained by radiation. Genetic analysis using the F1, F2, F3 populations derived from this mutant and other normal cultivars revealed that bentazon lethality/sensitivity was controlled by a single recessive gene, bel. The mutant can be killed at the seedling stage by bentazon at 300 mg/L or higher, a dosage that is safe for its F1 hybrids and all other normal cultivars. This mutant is also sensitive to all the tested sulfonylurea herbicides. The response of segregating plants to these two types of herbicides indicated that the sulfonylurea sensitivity was also controlled by bel. By crossing this mutant with Pei-Ai 64S (a TGMS line), the bel locus was located on chromosome 3 with 7.1 cM from the closest microsatellite marker RM168. Phenotypic analysis indicated that the bel gene had no negative effect on agronomic traits in either homozygous or heterozygous status. Because the M8077S mutant is lethal in the response of bentazon, homozygous plants at the mutated locus will be killed by the herbicide and selection in breeding programs must be made without application of the chemical. Associated microsatellite markers can be used to select these homozygous plants without spraying the herbicide. These markers have also been used to identify heterozygous plants during continuous backcrossing procedures for gene transfer from one genetic background to another. After two cycles of MAS, a new version of PeiAi 64S (a widely used EGMS cultivar) with bentazon sensitivity was obtained and is ready for seed production (Zhang et al. 2002). The Bel gene can also be used to test for purity in seed production. Hybrid rice seeds must be tested for purity before release to rice producers. Traditionally seed samples were planted and evaluated for purity after flowering, when they could be distinguished from the false hybrid plants based on distinct agronomic traits. In order to obtain purity test results before the next planting season, seed samples are usually sent to a location such as Hainan, China, where rice can be planted in the winter. Nevertheless, this is labor-intensive and also very expensive. Using bentazon sensitivity, false hybrids can be detected by spraying at the 2- to 3-leaf stage. A seedling tray in a greenhouse or in a growth chamber will be enough for a purity test required for any sample of hybrid seeds, which functions the same way as molecular markers used to identify off-types and false hybrids. In seed production, sterile plants are currently planted with pollinators in alternative rows. This system requires a great amount of manpower for transplanting and harvesting if planting is not mechanized. It is very difficult to use manpower to produce F1 seeds in countries where labor is limited and/or very expensive. Mixed planting of sterile plants

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

151

with pollinators could make mechanized seed production easier. This could also ensure a higher ratio of seed set on male sterile plants because the average distance between the pollinators and male sterile flowers becomes closer than under the alternative-row planting (Zhang et al. 2002). As discussed by Maruyama et al. (1991), incorporating a herbicidesensitivity gene to a pollinator, that could be killed by spraying a specific herbicide right after pollination, could ensure that all the harvested seeds are F1 hybrids only. Using PGMS and TGMS lines transformed with the herbicide-resistant gene Bar, the seeds of the herbicide resistant PGMS and TGMS lines were mixed with the pollen parent at the ratio of 4:1. The herbicide Basta was sprayed after natural pollination ended. The pollen parent died and the hybrid seeds could be harvested from the herbicide-resistant PGMS and TGMS lines. The natural outcrossing rate without artificial pollination ranged from 10.6% to 24.5% (Suh et al. 2002). Combining the use of a herbicide-sensitive mutant with a herbicideresistant mutant could be helpful not only for mixed plantings in hybrid seed production but also for removal of selfings. When a pollinator has a sensitivity gene to herbicide A and a resistance gene to herbicide B, it can be killed after pollination by spraying the herbicide A, while the plants from false hybrids can be killed by spraying the herbicide B. M8077S reported by Zhang et al. (2002) can be used as a gene donor for breeding herbicide sensitivity, while the bar gene as reported by Zhang et al. (1998) can be used for breeding a herbicide-resistant pollinator. XI. GENERAL DISCUSSIONS A. Economic Consideration The value of MAS depends on several factors. Acceleration of breeding programs and shortening the breeding cycle would be the first advantage of MAS, because the benefits of releasing new hybrids more quickly can be substantial, particularly in competitive markets. The economic merit of MAS could include situations in which molecular costs are more than offset by savings in phenotypic evaluation. If molecular costs are in addition to, not in place of, phenotypic costs, the economic merit of MAS will become questionable and more difficult to evaluate. In other cases, the ability to select early offsets the extra costs that are associated with MAS. The economic story from DNA sequencing may tell us what we can expect in terms of cost reduction in marker genotyping. Sequencing cost per finished base was $10 in 1990, but was reduced to $1 in 1996 and $0.1 in 2002. With further technology development, it is anticipated that the cost will be $0.01 in 2006, which is a thousand times cheaper

152

Y. XU

than in 1990. The cost of genotyping using molecular markers depends on marker type and its capacity in high-throughput analysis. For example, the lowest cost of SNP analysis is now about 20 to 30 cents per genotype, and a cost of only a few cents per genotype is expected in the coming years (Jenkins and Gibson 2002). With well-established marker systems and sequencing facilities, genotyping with SSR markers costs about 30 to 80 cents per data point, depending on marker multiplexing and the number of markers genotyped for each sample (Xu et al. 2002). There are several ways to reduce the MAS cost. First, high-throughput analysis using automated genotyping and data scoring systems will help increase the daily data output. Second, using the same sample for selection of multiple traits will reduce the trait-based cost. Third, selection at an early stage of plant development and an early stage of the breeding process will minimize the number of plants that need to be retained so that the overall breeding cost will come down. Fourth, optimization of MAS systems, including facilities and personnel, will result in less cost per data point. However, there are very few experiments that have been done with the aim of cost reduction. A comparison between MAS and conventional greenhouse screening of common beans for resistance to common bacterial blight showed that the cost of MAS is about onethird less than that of the greenhouse test (Yu et al. 2000). Today, a database literature search from CAB Abstracts for “markerassisted selection” provides about a thousand hits, but, in most cases, MAS is mentioned only as a future perspective. Others have evaluated the potential of MAS using computer simulation. Overall, there are still few reports of successful MAS experiments. Apparently, almost all MAS projects in plant breeding that have been reported received special monies for demonstration of MAS applications rather than for pure breeding. Several private companies have been routinely using MAS in breeding programs, benefiting from their long-term basic research programs and the availability of all the components of MAS, as discussed in Section III. It is certainly a big investment for a breeding company/institution to start from scratch to run a MAS-based breeding program. Further work on the economic evaluation and optimization of strategies for the use of MAS in breeding programs is required. The economically optimal use of MAS will likely necessitate a complete re-designing or at least some modification of breeding schemes (Dekkers and Hospital 2002). B. Bioinformatics and Breeding Database Enabling infrastructure provides the raw materials and the data and the methods for collating and verifying information within the appropriate biological context. The infrastructure typically includes cDNA and

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

153

genomic sequence data, genetic maps of mutants, DNA markers and maps, candidate genes and quantitative trait loci, physical maps based on chromosome breakpoints, and libraries of large inserts of DNA such as bacterial artificial chromosomes and radiation hybrids. Information flow from molecular markers to genetic maps to sequences and to genes has been established. Apparently, however, there is a gap between the sequence-based information and breeding-related information such as germplasm, pedigree, and phenotype. We will depend on phenotyping as the basis for functional analysis of about 40% of genes even though a complete sequence is available. Integration of breeding-related information with a genomics database is required for genomics-based breeding programs. MAS can be considered the first benefit that breeders can obtain now from genomics. The explosion of interest in marker-trait association studies has led to numerous reports in plants, each based on its own experimental population(s). Each experiment is limited in size and usually restricted to a single population or a cross planted in a specific environment. As suggested by Xu (2002), it is important for all researchers to follow general rules for reporting about genes and traits. One direction for the use of this database is to combine information from several studies, for example, by meta-analysis of results of QTL studies (Goffinet and Gerber 2000) or joint analysis of the raw data (Haley 1999). Extension of current databases to include raw data from gene mapping projects will stimulate this effort. On the other hand, many permanent populations have been shared internationally for genetic mapping. The raw data should be shared, too. A rice RFLP map constructed by using IR64/Azucena DHs has been saturated with about a thousand SSR markers (Chen et al. 1997; Temnykh et al. 2000; McCouch et al. 2002; Y. Xu, RiceTec, Inc., unpubl. data). Researchers involved in QTL mapping, however, have been using the first version of the molecular map consisting of only 175 RFLP markers. Sharing marker and phenotype information through a well-established database will make all sources of data more valuable. C. Opportunities and Challenges Genetic improvement through artificial selection has contributed greatly to advances in productivity that have been achieved over the past century in crops. Plant breeding has generally accounted for one-half of the increases in productivity of the major crops and the future will continue to depend on its advances. Most of the traits that are selected are complex quantitative traits. So far, most selection has been on the basis of observable phenotypic variation, which represents the collective effect

154

Y. XU

of all genes and the environment. Advances in genomics and MAS provide breeders both opportunities and challenges for the improvement of crops through manipulation of both qualitative and quantitative traits. Well-established physical maps and the publicly available DNA sequence for a whole genome, as is currently available for rice, will help eliminate the many fine mapping steps that are required now to narrow down the candidate genome region to a kilobase resolution. Once identified, a target region could be associated with a contig that has been well located on the physical map and defined by molecular markers. On the other hand, high-throughput analysis combined with highly informative molecular markers enables us to manage populations with thousands of plants and thousands of markers in fine mapping. Recent advances in genomics have made it possible to map and determine the magnitude of the effect of individual loci controlling both qualitative and quantitative traits. Positional cloning has been successfully used to clone several QTL with relatively large effects (Frary et al. 2000; Fridman et al. 2000; Yano et al. 2000; Takahashi et al. 2001). Since heterosis and other quantitative traits are usually controlled by many genes, each with a relatively small effect, it will be a great challenge for molecular geneticists to verify the effects of minor genes (Xu 1997). If many genes are known, and favorable alleles are present in different lines or cultivars, MAS can be used to design new genotypes that combine favorable alleles at all loci. A large number of molecular polymorphisms such as SSRs or SNPs and small or large insertions/deletions, discovered with genome sequencing, provide an opportunity for identifying the nucleotide change associated with quantitative trait variation. The nucleotide change that contributes to quantitative variation has been referred to as quantitative trait nucleotide (QTN) (Lyman et al. 1999) or FNP. Fine mapping combined with sequence analysis can quickly narrow the chromosomal region associated with quantitative variation (QTL) down to a specific nucleotide change. Rice as a major food source for human beings has been moving ahead of other crop plants in terms of genome sequencing. Any success in this crop will benefit others in certain ways. Rice has become the model species for the cereals; chromosomes of other cereals such as maize, sorghum, sugarcane, millet, oats, wheat, and barley share much similarity to each other and to rice. The complete DNA sequence of rice will allow the tracking of genes/traits from rice to other grass species. It has become apparent that the differences between species of plants appear due to novel allelic specifications and interactions other than novel genes. Can molecular genetics improve our understanding and manipulation of heterosis? Plant complexity and numerous interactions with the envi-

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

155

ronment will make it difficult or impossible for molecular genetics to identify universal aspects of heterosis suitable for alternation. Elements of a crop’s heterotic response are interwoven with many aspects of plant metabolism and development. We may not find a biologically based unifying mechanism or pathway of heterosis. Certainly, molecular genetics alone will not accomplish this, especially if there are not adequate human resources to analyze, explore, and integrate new sources of information while maintaining a highly effective infrastructure of plant breeding (Lee 1999). Without a deeper understanding of the physiological determinants of yield potential, molecular approaches that seek to empirically concentrate “yield genes” or heterosis genes are likely to fail. Instead, molecular geneticists must actively collaborate with crop physiologists, agronomists, and plant breeders so that genetic differences in yield potential and heterotic loci can be properly measured and identified. Additionally, measurement of yield-determining traits should be made on a time-series of hybrids grown at yield-potential levels, to determine if the changes in hybrid characteristics in breeding history have contributed to an increase in yield potential (Duvick and Cassman 1999). As our awareness and sophistication grows, with the new information, it should be possible to develop more rational and informationdriven assessments and strategies. Molecular-marker technology has revolutionized our understanding of quantitative traits in different backgrounds at different levels of genomics. Considering interactions of all types that could happen among numerous genes with minor effects, and interaction with external environments that could change from time to time and from location to location, we need to develop much more efficient systems for separating, pyramiding, and packaging all the genes related to hybrid performance and heterosis and make them functional at full scale in specific environments. These systems would be too complicated to be practical. Increasing efficiency should be a major objective for future generations of geneticists and plant breeders. Highly informative markers, isogenic mutation libraries, high-throughput technology, availability of full sequences from several model plants and genomics software are key tools for genetic manipulation of genes for quantitative traits. With all these developments, MAS could be an important component in plant breeding, especially for heterosis and hybrid performance. The accepted dogma has been that recent advances in molecular genetics promises to revolutionize agricultural practices. However, as Lande and Thompson (1990) point out, there are several reasons why molecular genetics can never replace traditional methods of agricultural improvement, but instead should be integrated to obtain the maximum improvement in the economic value of domesticated populations. The

156

Y. XU

analytical results, as well as the more recent computer simulations and the limited empirical results (Stuber et al. 1999), are encouraging and support the use of genetic markers to achieve a substantial increase in the efficiency of hybrid breeding. LITERATURE CITED Ajmone Marson, P., P. Castiglioni, F. Fusari, M. Kuiper, and M. Motto. 1998. Genetic diversity and its relationship to hybrid performance in maize as revealed by RFLP and AFLP markers. Theor. Appl. Genet. 96:219–227. Akagi, H., Y. Yokozeki, A. Inagaki, A. Nakamura, and T. Fujimura. 1996. A codominant DNA marker closely linked to the rice nuclear restorer gene, Rf-1, identified with interSSR fingerprinting. Genome 39:1205–1209. Ayres, N. M., A. M. McClung, P. D. Larkin, H. F. J. Bligh, C. A. Jones, and W. D. Park. 1997. Microsatellites and a single-nucleotide polymorphism differentiate apparent amylose classes in an extended pedigree of US rice germ plasm. Theor. Appl. Genet. 94:773–781. Barton, N. H., and P. D. Keightley. 2002. Understanding quantitative genetic variation. Nat. Rev. Genet. 3:11–21. Beavis, W. D. 1994. The power and deceit of QTL experiments: lessons from comparative QTL studies. p. 250–266. In: Proceedings of the 49th annual corn and sorghum industry research conference. Am. Seed Trade Assoc., Washington, DC. Beckmann, J. S., and M. Soller. 1986. Restriction fragment length polymorphisms in plant genetic improvement. Oxford Surveys Plant Molec. Cell Biol. 3:196–250. Benchimol, L. L., C. L. de Souza, Jr., A. A. F. Garcia, P. M. S. Kono, C. A. Mangolin, A. M. M. Barbosa, A. S. G. Coelho, and A. P. de Souza. 2000. Genetic diversity in tropical maize inbred lines: heterotic group assignment and hybrid performance determined by RFLP markers. Plant Breed. 119:491–496. Bernardo, R. 1992. Relationship between single-cross performance and molecular marker heterozygosity. Theor. Appl. Genet. 83:628–634. Bernardo, R. 1994. Prediction of maize single-cross performance using RFLPs and information from related hybrids. Crop Sci. 34:20–25. Bernardo, R. 1996. Best linear unbiased prediction of maize single-cross performance. Crop Sci. 36:50–56. Bernardo, R. 1999. Best linear unbiased predictor analysis. p. 269–276. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Bernardo, R. 2001. What if we knew all the genes for a quantitative trait in hybrid crops? Crop Sci. 41:1–4. Bijlsma, R., R. W. Allard, and A. L. Kahler. 1986. Nonrandom mating in an open-pollinated maize population. Genetics 112:669–680. Bingham, E. T. 1998. Role of chromosome blocks in heterosis and estimates of dominance and overdominance. p. 71–87. In: K. R. Lamkey and J. E. Staub (eds.), Concepts and breeding of heterosis in crop plants. CSSA, Madison, WI. Bligh, H. F. J., R. I. Till, and C. A. Jones. 1995. A microsatellite sequence closely linked to the waxy gene of Oryza sativa. Euphytica 86:83–85. Boppenmaier, J., A. E. Melchinger, G. Seitz, H. H. Geiger, and R. G. Herrmann. 1993. Genetic diversity for RFLPs in European maize inbreds. III. Performance of crosses within versus between heterotic groups for grain traits. Plant Breed. 111:217–226.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

157

Börner, A., V. Korzun, A. Polley, S. Malyshev, and G. Melz. 1998. Genetics and molecular mapping of a male fertility restoration locus (Rfg1) in rye (Secale cereale L.). Theor. Appl. Genet. 97:99–102. Botstein, D., R. L. White, M. Skolnick, and R. W. Davis. 1980. Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am. J. Hum. Genet. 32:314–331. Brandle, J. E., and P. B. E. McVetty. 1990. Geographical diversity, parental selection and heterosis in oilseed rape. Can. J. Plant Sci. 70:935–940. Bretting, P. K., and M. P. Widrlechner. 1995. Genetic markers and plant genetic resource management. Plant Breed. Rev. 13:11–86. Brown, A. D. H. 1989a. The case for core collections. p. 136–156. In: A. D. H. Brown, O. H. Frankel, R. D. Marshall, and J. T. Williams (eds.), The use of plant genetic resources. Cambridge Univ. Press, Cambridge, UK. Brown, A. D. H. 1989b. Core collection: A practical approach to genetic resources management. Genome 31:818–824. Brown, S. M., and S. Kresovich. 1996. Molecular characterization for plant genetic resources conservation. p. 85–93. In: A. H. Paterson (ed.), Genome mapping in plants. R. G. Landes Co., Austin, TX. Brummer, E. C. 1999. Capturing heterosis in forage crop cultivar development. Crop Sci. 39:943–954. Cerna, F. J., S. R. Cianzio, A. Rafalski, S. Tingey, and D. Dyer. 1997. Relationship between seed yield heterosis and molecular marker heterozygosity in soybean. Theor. Appl. Genet. 95:460–467. Chambliss, C. E., and J. M. Jenkins. 1923. Some new varieties of rice. USDA Bul. 1127. Chang, T. T. 1976. The origin, evolution, cultivation, dissemination and diversification of Asian and African rices. Euphytica 25:425–441. Chang, T. T. 1984. Conservation of rice genetic resources: luxury or necessity? Science 224: 251–256. Charcosset, A., and L. Essioux. 1994. The effect of population structure on the relationship between heterosis and heterozygosity at marker loci. Theor. Appl. Genet. 89:336–343. Charcosset, A., M. Lefort-Buson, and A. Gallais. 1991. Relationship between heterosis and heterozygosity at marker loci: a theoretical computation. Theor. Appl. Genet. 81:571– 575. Chavarriaga-Aquirre, P., M. M. Maya, J. Tohme, M. C. Duque, C. Iglesias, M. W. Bonierbale, S. Kresovich, and G. Kochert. 1999. Using microsatellites, isozymes and AFLPs to evaluate genetic diversity and redundancy in the cassava core collection and to assess the usefulness of DNA-based markers to maintain germplasm collections. Mol. Breed. 5:263–273. Chen, L. (ed.). 2001. The principles and techniques of two-line hybrid rice. Shanghai Scientific and Technical Publ., Shanghai, China. Chen S., X. H. Lin, C. G. Xu, and Q. Zhang. 2000. Improvement of bacterial blight resistance of ‘Minghui 63’, an elite restorer line of hybrid rice, by molecular marker-assisted selection. Crop Sci. 40:239–244. Chen, S., C. G. Xu, X. H. Lin, and Q. Zhang. 2001. Improving bacterial blight resistance of ‘6078’, an elite restorer line of hybrid rice, by molecular marker-assisted selection. Plant Breed. 120:133–137. Chen, X., S. Temnykh, Y. Xu, Y. G. Cho, and S. R. McCouch. 1997. Development of microsatellite framework map providing genome-wide coverage in rice (Oryza sativa L.). Theor. Appl. Genet. 95:553–567.

158

Y. XU

Coburn, J. R., S. V. Temnykh, E. M. Paul, and S. R. McCouch. 2002. Design and application of microsatellite marker panels for semiautomated genotyping of rice (Oryza sativa L.). Crop Sci. 42:2092–2099. Crossa, J., I. H. DeLacy, and S. Taba. 1995. The use of multivariate methods in developing a core collection. p. 77–92. In: T. Hodgkin, A. H. D. Brown, Th. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Wiley, West Sussex, England. Crow, J. F. 1999. Dominance and overdominance. p. 49–58. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Crow, J. F. 2000. The rise and fall of overdominance. Plant Breed. Rev. 17:225–257. Curtis, J. J., A. M. Brunson, J. E. Hubbard, and F. R. Earle. 1956. Effect of the parent on oil content of the corn kernel. Agron. J. 48:551–555. D’Amato, F. 1975. The problem of genetic stability in plant tissues and cell cultures. p. 333–348. In: O. Frankel and J. G. Hawkes (eds.), Crop genetic resources for today and tomorrow. Cambridge Univ. Press, Cambridge, UK. Darrah, L. L., and M. S. Zuber. 1986. 1985 United States maize germplasm base and commercial breeding strategy. Crop Sci. 26:1109–1113. Davenport, C. B. 1908. Degeneration, albinism and inbreeding. Science 28:454–455. Dean, R. E., J. A. Dahlberg, M. S. Hopkins, and S. Kresovich. 1999. Genetic redundancy and diversity among ‘Orange’ accessions in the U.S. National sorghum collection as assessed with simple sequence repeat (SSR) markers. Crop Sci. 39:1215–1221. Dekkers, J. C. M., and F. Hospital. 2002. The use of molecular genetics in the improvement of agricultural populations. Nat. Rev. Genet. 3:22–32. Delourme, R., A. Bouchereau, N. Hubert, M. Renard, and B. S. Landry. 1994. Identification of RAPD markers linked to fertility restorer gene for the Ogura radish cytoplasmic male sterility of rapeseed (Brassic napus L.). Theor. Appl. Genet. 88:741–748. Delourme, R., N. Foisset, R. Horvais, P. Barret, G. Champagne, W. Y. Cheung, B. S. Landry, and M. Renard. 1998. Characterisation of the radish introgression carrying the Rfo restorer gene for the Ogu-INRA cytoplasmic male sterility in rapeseed (Brassica napus L.). Theor. Appl. Genet. 97:129–134. De Vicente, M. C., and S. D. Tanksley. 1993. QTL analysis of transgressive segregation in an interspecific tomato cross. Genetics 134:585–596. De Vries, A. Ph. 1971. Flowering biology of wheat, particularly in view of hybrid seed production: A review. Euphytica 20:152–170. Dhillon, B. S., J. Boppenmaier, W. G. Pollmer, R. G. Hermann, and A. E. Mechinger. 1993. Relationship of restriction fragment length polymorphisms among European maize inbreds with ear dry matter yield of their hybrids. Maydica 38:245–248. Dilday, R. H. 1990. Contribution of ancestral lines in the development of new cultivars of rice. Crop Sci. 30:905–911. Dong, F., X. Zhu, Z. Xiong, S. Cheng, Z. Sun, and S. Min. 1995. Breeding of photoperiodthermo sensitive genic male sterile indica rice with a pale-green-leaf marker. Chinese J. Rice Sci. 9:65–70. Dong, N. V., P. K. Subudhi, P. N. Luong, V. D. Quang, T. D. Quy, H. G. Zheng, B. Wang, and H. T. Nguyen. 2000. Molecular mapping of a rice gene conditioning thermosensitive genic male sterility using AFLP, RFLP and SSR techniques. Theor. Appl. Genet. 100:727–734. Dudley, J. W. 1993. Molecular markers in plant improvement: manipulation of genes affecting quantitative traits. Crop Sci. 33:660–668. Dudley, J. W. 2004. 100 generations of selection for oil and protein in corn. Plant Breed. Rev. 24 (in press).

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

159

Dudley, J. W., and R. J. Lambert. 1992. Ninety generations of selection for oil and protein in maize. Maydica 37:81–87. Dudley, D. N., M. A. Saghai Maroof, and G. K. Rufener. 1991. Molecular markers and grouping of parents in a maize breeding program. Crop Sci. 31:718–723. Dudley, D. N., M. A. Saghai Maroof, and G. K. Rufener. 1992. Molecular marker information and selection of parents in corn breeding program. Crop Sci. 32:301–304. Duvick, D. N. 1984. Genetic contribution to yield grains of U.S. hybrid maize, 1930–1980. p. 15–47. In: W. R. Fehr (ed.), Genetic contributions to yield grains of five major crop plants. CSSA Spec. Publ. 7. CSSA and ASA, Madison, WI. Duvick, D. N. 1996. Plant breeding: an evolutionary concept. Crop Sci. 36:539–549. Duvick, D. N. 1999. Heterosis: feeding people and protecting natural resources. p. 19–29. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASACSSA-SSSA, Madison, WI. 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. East, E. M. 1908. Inbreeding in corn. Rep. Connecticut Expt. Stat. Years 1907–1908. p. 419–428. Eathington, S. R., J. W. Dudley, and G. K. Rufener. 1997. Marker effects estimated from testcrosses of early and late generations of inbreeding in maize. Crop Sci. 37:1679–1685. Edwards, M. D., and N. J. Page. 1994. Evaluation of marker-assisted selection through computer simulation. Theor. Appl. Genet. 88:376–382. Fehr, W. R. (ed.). 1984. Genetic contribution to yield grains of five major crop plants, Special Publ. No. 7. CSSA, Madison, WI. Fisch, M., M. Bohn, and A. E. Melchinger. 1999. Comparison of selection strategies for marker-assisted backcrossing of a gene. Crop Sci. 39:1295–1301. Flint, J., and R. Mott. 2001. Finding the molecular basis of quantitative traits: successes and pitfalls. Nat. Rev. Genet. 2:437–445. Frankel, O. 1984. Genetic perspectives of germplasm conservation. p. 161–170. In: W. Arber, K. Limensee, W. J. Peacock, and P. Starlinger (eds.), Genetic manipulation: impact on man and society. Cambridge Univ. Press, Cambridge, UK. Frary, A., T. C. Nesbitt, A. Frary, S. Grandillo, E. van der Knaap, B. Cong, J. Liu, J. Meller, R. Elber, K. B. Alpert, and S. D. Tanksley. 2000. fw2.2: a quantitative trait locus to the evolution of tomato fruit size. Science 289:85–88. Fridman, E., T. Pleban, and D. Zamir. 2000. A recombination hotspot delimits a wildspecies quantitative trait locus for tomato sugar content to 484 bp within an invertase gene. Proc. Natl. Acad. Sci. (USA) 97:4718–4723. Gardner, C. O. 1963. Estimates of genetic parameters in cross fertilizing plants and their implications in plant breeding. p. 225–252. In: W. D. Hanson and H. F. Robinson (ed.), Statistical genetics and plant breeding. NAS-NRC, Washington, DC. Gentzbittel, L., F. Vear, Y. X. Zhang, A. Berville, and P. Nicolas. 1995. Development of a consensus linkage RFLP map of cultivated sunflower (Helianthus annuus L.). Theor. Appl. Genet. 90:1079–1086. Georgiady, M. S., R. W. Whitkus, and E. M. Lord. 2002. Genetic analysis of traits distinguishing outcrossing and self-pollinating forms of currant tomato, Lycopersicon pimpinellifolium (Jusl.) Mill. Genetics 161:333–344. Gerdes, J. T., and W. F. Tracy. 1993. Pedigree diversity within the Lancaster Surecrop heterotic group of maize. Crop Sci. 33:334–337. Gethi, J. G., J. A. Labate, K. R. Lamkey, M. E. Smith, and S. Kresovich. 2002. SSR variation in important U.S. maize inbred lines. Crop Sci. 42:951–957.

160

Y. XU

Gizlice, Z., T. E. Carter, Jr., and J. W. Burton. 1993. Genetic diversity in North American soybean: II. Prediction of heterosis in F2 populations of southern founding stock using genetic similarity measures. Crop Sci. 33:620–626. Godshalk, E. B., M. Lee, and K. R. Lamkey. 1990. Relationship of restriction fragment length polymorphisms to single-cross hybrid performance of maize. Theor. Appl. Genet. 80:273–280. Goffinet, B., and S. Gerber. 2000. Quantitative trait loci: A meta-analysis. Genetics 155:463–473. Goldman, I. L. 1999. Inbreeding and outbreeding in the development of a modern heterosis concept. p. 7–18. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Goldman, I. L. 2000. Prediction in plant breeding. Plant Breed. Rev. 19:15–40. Goldman, I. L., T. R. Rocheford, and J. W. Dudley. 1993. Quantitative trait loci influencing protein and starch concentration in the Illinois long term selection maize strains. Theor. Appl. Genet. 87:217–224. Graham, G. I., D. W. Wolff, and C. W. Stuber. 1997. Characterization of a yield quantitative locus on chromosome five of maize by fine mapping. Crop Sci. 37:1601–1610. Grandillo, S., and S. D. Tanksley. 1996. QTL analysis of horticultural traits differentiating the cultivated tomato from the closely related species Lycopersicon pimpinellifolium. Theor. Appl. Genet. 92:935–951. Grant, I., and W. D. Beversdorf. 1985. Heterosis and combining ability estimates in spring oilseed rape (Brassica napus L.). Can. J. Genet. Cytol. 27:472–478. Gu, M. H., and S. Z. Tang. 2001. Theory and practice of two-line intersubspecific hybrid rice. p. 100–132. In: X. G. Lu, M. H. Gu, and C. Q. Li (eds.), Theory and technology of two-line hybrid rice. China Science Press, Beijing. Haley, C. 1999. Advances in quantitative trait locus mapping. p. 47–59. In: J. C. M. Dekkers, S. J. Lamont, and M. F. Rothschild (eds.), From Jay Lush to genomics: Visions for animal breeding and genetics. Animal Breeding & Genetics Group, Dept. Animal Sci., Iowa State Univ., Ames. Hallauer, A. R. 1990. Methods used in developing maize inbreds. Maydica 35:1–16. Hallauer, A. R. 1999. Temperate maize and heterosis. p. 353–361. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Hallauer, A. R., and J. B. Miranda. 1988. Quantitative genetics and maize breeding. 2nd ed. Iowa State Univ. Press, Ames. Hallauer, A. R., W. A. Russell, and K. R. Lamkey. 1988. Corn breeding. p. 463–564. In: G. F. Sprague and J. W. Dudley (eds.), Corn and corn improvement. 3rd ed. Agron. Monogr. 18. ASA-CSSA-SSSA, Madison, WI. Hamon, S., M. Noirot, and F. Anthony. 1995. Developing a coffee core collection using the principal components score strategy with quantitative data. p. 117–126. In: T. Hodgkin, A. H. D. Brown, Th. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Wiley, West Sussex, England. Han, G. C., S. K. Vasal, D. L. Beck, and E. Elias. 1991. Combining ability analysis of inbred lines derived from CIMMYT maize (Zea mays L.) germplasm. Maydica 36:57–64. Hansen, M., C. Halldén, N.-O. Nilsson, and T. Säll. 1997. Marker-assisted selection of restored male-fertile Brassica napus plants using a set of dominant RAPD markers. Mol. Breed. 3:449–456. Havey, M. J. 1998. Molecular analyses and heterosis in the vegetables: can we breed them like maize? p. 109–116. In: K. R. Lamkey and J. E. Staub (eds.), Concepts and breeding of heterosis in crop plants. CSSA, Madison, WI.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

161

He, S., Z.-H. Yu, C. E. Vallejos, and S. A. Mackenzie. 1995. Pollen fertility restoration by nuclear gene Fr in common bean: an Fr linkage map and the mode of Fr action. Theor. Appl. Genet. 90:1056–1062. He, Y., C. Chen, J. Tu, P. Zhou, G. Jiang, Y. Tan, C. Xu, and Q. Zhang. 2002. Improvement of an elite rice hybrid, Shanyou 63, by transformation and marker-assisted selection. p. 43. In: Abstract of the Fourth Intl. Symp. Hybrid Rice, 14–17 May 2002, Hanoi, Vietnam. Henderson, C. R. 1975. Best linear unbiased estimation and prediction under a selection model. Biometrics 31:423–447. Hepting, L. 1978. Analyse eines 7×7—Sortendiallels zur Ermittlung geeigneten Ausgansmaterials für die Hybridzüchtung bei Roggen. Z. Pflanzenzücht. 80:188–197. 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. Hjerdin-Panagopoulos, A., T. Kraft, I. M. Rading, S. Tuvesson, and N.-O. Nilsson. 2002. Three QTL regions for restoration of Owen CMS in sugar beet. Crop Sci. 42:540–544. Hospital, F., and A. Charcosset. 1997. Marker-assisted introgression of quantitative trait loci. Genetics 147:1469–1485. Hospital, F., C. Chevalet, and P. Mulsant. 1992. Using markers in gene introgression breeding programs. Genetics 132:1199–1210. Huang, N., E. R. Angeles, J. Domingo, G. Magpantay, S. Singh, G. Zhang, N. Kumaravadivel, J. Bennett, and G. S. Khush. 1997. Pyramiding of bacterial blight resistance genes in rice: marker-assisted selection using RFLP and PCR. Theor. Appl. Genet. 95:313–320. Huang, Q. Y., Y. Q. He, R. C. Jing, R. S. Zhu, and Y. G. Zhu. 1999. Mapping of the nuclear fertility restorer gene for HL cytoplasmic male sterility in rice using microsatellite markers. Chinese Sci. Bul. 45:430–432. Hyne, V., and M. J. Kearsey. 1995. QTL analysis—further uses of marker regression. Theor. Appl. Genet. 91:471–476. Igartua, E., M. Edney, B. G. Rossnagel, D. Spaner, W. G. Legge, G. L. Scoles, P. E. Ecksteins, G. A. Penner, N. A. Tinker, K. G. Briggs, D. E. Falk, and D. E. Mather. 2000. Markerassisted selection of QTL affecting grain and malt quality in two-row barley. Crop Sci. 40:1426–1433. Ikehashi, H. 1982. Prospects for overcoming barriers in the utilization of indica-japonica crosses in rice breeding. Oryza 19:69–77. Ikehashi, H., and H. Araki. 1986. Genetics of F1 sterility in remote crosses of rice. p. 119– 130. In: Rice genetics. Intl. Rice Res. Inst., P. O. Box 933, Manila 1099, Philippines. Ikehashi, H., and H. Araki. 1987. Screening and genetic analysis of wide compatibility in F1 hybrids of distant crosses in rice, Oryza sativa. Tech. Bul. Trop. Agr. Res. Center No. 23. TARC, Japan. Ikehashi, H., and J. Wan. 1998. The wide compatibility system: current knowledge of its genetics and use for enhanced yield heterosis. p. 67–77. In: S. S. Virmani, E. A. Siddiq, and K. Muralidharan (eds.), Advances in hybrid rice technology. Proc. Third Intl. Symp. Hybrid Rice, 14–16 Nov. 1996, Hyderabad, India. Intl. Rice Res. Inst., Manila, Philippines. Iwanaga, M. 1993. Enhancing links between germplasm conservation and use in a changing world. p. 407–413. In: International crop science I. CSSA, Madison, WI. Janick, J. 1998. Hybrids in horticulture crops. p. 45–56. In: K. R. Lamkey and J. E. Staub (eds.), Concepts and breeding of heterosis in crop plants. CSSA, Madison, WI. Jean, M., G. G. Brown, and B. S. Landry. 1997. Genetic mapping of nuclear fertility restorer genes for the ‘Polima’ cytoplasmic male sterility in canola (Brassica napus L.) using DNA markers. Theor. Appl. Genet. 95:321–328.

162

Y. XU

Jenkins, S., and N. Gibson. 2002. High-throughput SNP genotyping. Comp. Funct. Genom. 3:57–66. Ji, Q., G. Wang, E. Belhassen, H. Serieys, and A. Berville. 1996. Molecular markers of nuclear restoration gene Rf1 in sunflower using bulked segregant analysis—RAPD. Science in China (Ser. C) 39:551–560. Jia, J. H., C. Y. Li, X. P. Qu, Q. Wang, Q. Y. Deng, M. L. Weng, and B. Wang. 2000. Construction of genetic linkage map and mapping of thermo-sensitive genic male sterile gene tms5 in rice. p. 37. In: Abstract for Plant Genomics in China I, July 2000, Dalian, China. Jia, J. H., D. S. Zhang, C. Y. Li, X. P. Qu, S. W. Wang, V. Chamarerk, H. T. Nguyen, and B. Wang. 2001. Molecular mapping of the reverse thermo-sensitive genic male sterile gene (rtms1) in rice. Theor. Appl. Genet. 103:607–612. Jia, M. H., S. He, W. Vanhouten, and S. Machenzie. 1997. Nuclear fertility restorer genes map to the same linkage group in cytoplasmic male-sterile bean. Theor. Appl. Genet. 95:205–210. Jing, R., X. Li, P. Yi, and Y. Zhu. 2001. Mapping fertility-restoring genes of rice WA cytoplasmic male sterility using SSLP markers. Bot. Bul. Acad. Sin. 42:167–171. Johnson, G. R., and R. H. Mumm. 1996. Marker-assisted maize breeding. Proc. 51st Annual Corn and Sorghum Industry Research Conf., American Seed Trade Assoc. 51:75–84. Johnston, T. H. 1958. Registration of rice varieties. Agron. J. 50:694–700. Jordaan, J. P., S. A. Engelbrecht, J. H. Malan, and H. A. Knobel. 1999. Wheat and heterosis. p. 411–421. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Joshi, S. P., S. G. Bhave, K. V. Chowdari, G. S. Apte, B. L. Dhonukshe, K. Lalitha, P. K. Rangekar, and V. S. Gupta. 2001. Use of DNA markers in prediction of hybrid performance and heterosis for a three-line hybrid system in rice. Biochem. Genet. 39(5/6):179–200. Kahler, A. L., C. O. Gardner, and R. W. Allard. 1984. Nonrandom mating in experimental populations of maize. Crop Sci. 24:350–354. Kamps, T. L., and C. D. Chase. 1997. RFLP mapping of the maize gametophytic restorerof-fertility locus (rf3) and aberrant pollen transmission of the nonrestoring rf3 allele. Theor. Appl. Genet. 95:525–531. Kearsey, M. J., and A. G. L. Farquhar. 1998. QTL analysis in plants: Where are we now? Heredity 80:137–142. Khush, G. S., I. Kumar, and S. S. Virmani. 1988. Grain quality of hybrid rice. p. 201–215. In: International Rice Research Institute (ed.), Hybrid rice. P.O. Box 933, Manila, Philippines. Kiesselbach, T. A. 1926. The immediate effect of gametic relationship and of parental type upon the kernel weight of corn. Neb. Agr. Expt. Sta. Bul. 33:1–69. King, J. J., J. M. Bradeen, O. Bark, J. A. McCallum, and M. J. Havey. 1998. A low-density genetic map of onion reveals a role for tandem duplication in the evolution of an extremely large diploid genome. Theor. Appl. Genet. 96:52–62. Kinoshita, T. 1998. Linkage mapping using mutant genes in rice. Rice Genet. Newsl. 15:13–74. Klein, R. R., P. E. Klein, A. K. Chhabra, J. Dong, S. Pammi, K. L. Childs, J. E. Mullet, W. L. Rooney, and K. F. Schertz. 2001. Molecular mapping of the rf1 gene for pollen fertility restoration in sorghum (Sorghum bicolor L.). Theor. Appl. Genet. 102:1206–1212. Knapp, S. J. 1998. Marker-assisted selection as a strategy for increasing the probability of selecting superior genotypes. Crop Sci. 38:1164–1174. Kojima, T., H. Tsujimoto, and Y. Ogihara. 1997. High-resolution RFLP mapping of the fertility restoration (Rf3) gene against Triticum timopheevi cytoplasm located on chromosome 1BS of common wheat. Genes Genet. Syst. 72:353–359.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

163

Kumar, I., and G. S. Khush. 1986. Genetics of amylose content in rice (Oryza sativa L.). J. Genet. 65:1–11. Kurata, N., Y. Nagamura, K. Yamamoto, Y. Harushima, N. Sue, J. Wu, B. A. Antonio, A. Shomura, T. Shimizu, S. Y. Lin, T. Inoue, A. Fukuda, T. Shimano, Y. Kuboki, T. Toyama, Y. Miyamoto, T. Kirihara, K. Hayasaka, A. Miyao, L. Monna, H. S. Zhong, T. Tamura, Z. X. Wang, T. Momma, Y. Umehara, M. Yano, T. Sasaki, and Y. Minobe. 1994. A 300 kilobase interval genetic map of rice including 883 expressed sequences. Nat. Genet. 8:365–372. Lamkey, K. R., and J. W. Edwards. 1999. Quantitative genetics of heterosis. p. 31–48. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASACSSA-SSSA, Madison, WI. Lande, R., and R. Thompson. 1990. Efficiency of marker-assisted selection in the improvement of quantitative traits. Genetics 124:743–756. Lang, N. T., P. K. Subudhi, S. S. Virmani, D. S. Brar, G. S. Khush, Z. Li, and N. Huang. 1999. Development of PCR-based markers for thermosensitive genetic male sterility gene tms3(t) in rice (Oryza sativa L.). Hereditas 131:121–127. Laporte, V., D. Merdinoglu, P. Saumitou-Laprade, G. Butterlin, P. Vernet, and J. Cuguen. 1998. Identification and mapping of RAPD and RFLP markers linked to a fertility restorer gene for a new source of cytoplasmic male sterility in Beta vulgaris ssp. maritima. Theor. Appl. Genet. 96:989–996. Lassner, M. W., and T. J. Orton. 1983. Detection of somatic variation. p.209–217. In: S. D. Tanksley and T. J. Orton (eds.), Isozymes in plant genetics and breeding. Vol. 1A. Developments in plant genetics and breeding, 1. Elsevier, Amsterdam. Lee, M. 1995. DNA markers and plant breeding programs. Adv. Agron. 55:265–344. Lee, M. 1999. Towards understanding and manipulating heterosis in crop plants: can molecular genetics help? p. 185–194. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Lee, M., E. B. Godshalk, K. R. Lamkey, and W. L. Woodman. 1989. Association of restriction length polymorphism among maize inbreds with agronomic performance of their crosses. Crop Sci. 29:1067–1071. Lefort-Buson, M., B. Guillot-Lemoine, and Y. Datte. 1987. Heterosis and genetic distance in rape seed (Brassica napus L.): Crosses between European and Asiatic selfed lines. Genome 29:413–418. Lewontin, R. C., and J.-P. Berlan. 1990. The political economy of agricultural research: The case of hybrid corn. p. 613–628. In: C. R. Carroll et al. (eds.), Agroecology. McGraw Hill, New York. Li, B., and R. Wu. 1996. Genetic causes of heterosis in juvenile aspen: a quantitative comparison across intra- and inter-specific hybrids. Theor. Appl. Genet. 93:380–391. Li, H., L. Tang, and J. Zou. 1991. Marker-based analysis of wide compatibility in rice (in Chinese). Hybrid Rice (4):22–24. Li, J., and L. Yuan. 2000. Hybrid rice: Genetics, breeding, and seed production. Plant Breed. Rev. 17:15–158. Li, P., K. D. Zhou, Y. Chen, P. He, R. R. Li, and L. H. Zhu. 1996. Molecular mapping of the restorer genes of wild type cytoplasmic male sterility in rice. Acta Genetica Sinica 23:357–362. Li, X., Q. Lu, F. Wang, C. Xu, and Q. Zhang. 2001. Separation of the two-locus inheritance of photoperiod sensitive genic male sterility in rice and precise mapping the pms3 locus. Euphytica 119:343–348. Li, Z. 2001. QTL mapping in rice. p. 153–171. In: G. S. Khush, D. S. Brar and B. Hardy (eds.), Rice genetics IV. Proc. Fourth Intl. Rice Genetics Symp., 22–27 Oct. 2000. Los

164

Y. XU

Baños, Philippines. Science Publ., Inc., New Delhi, India, and Intl. Rice Res. Inst., Los Baños, Philippines. Li, Z.-K., L. J. Luo, H. W. Mei, D. L. Wang, Q. Y. Shu, R. Tabien, D. B. Zhong, C. S. Ying, J. W. Stansel, G. S. Khush, and A. H. Paterson. 2001. Overdominance epistatic loci are the primary genetic basis of inbreeding depression and heterosis in rice: I. Biomass and grain yield. Genetics 158:1737–1753. Lin, H. X., T. Yamamoto , T. Sasaki, and M. Yano. 2000. Characterization and detection of epistatic interactions of 3 QTLs, Hd1, Hd2, and Hd3, controlling heading date in rice using nearly isogenic lines. Theor. Appl. Genet. 101:1021–1028. Liu, A., Q. Zhang, and H. Li. 1992. Location of a gene for wide-compatibility in the RFLP linkage map. Rice Genet. Newsl. 9:134–136. Liu, B. H. 1998. Statistical genomics: Linkage, mapping, and QTL analysis. CRC Press, Boca Raton, FL. Liu, K., J. Wang, H. B. Li, C. G. Xu, A. M. Li, X. H. Li, and Q. Zhang. 1997. A genome-wide analysis of wide compatibility in rice and the precise location of the S5 locus in the molecular map. Theor. Appl. Genet. 95:809–814. Liu, N., Y. Shan, F. P. Wang, C. G. Xu, K. M. Peng, X. H. Li, and Q. Zhang. 2001. Identification of an 85-kb DNA fragment containing pms1, a locus for photoperiod-sensitive genic male sterility in rice. Mol. Genet. Genomics 266:271–275. Liu, X. Z., Z. B. Peng, J. H. Fu, L. C. Li, and C. L. Huang. 1997. Application of RAPD in group classification studies. Scientia Agr. Sinica 30(3):44–51. Liu, X. C., and J. L. Wu. 1998. SSR heterotic patterns of parents for making and predicting heterosis in rice breeding. Mol. Breed. 4:263–268. Lu, X. G., M. H. Gu, and C. Q. Li (eds.). 2001. Theory and technology of two-line hybrid rice. China Science Press, Beijing. Lu, X. G., T. M. Mou, N. T. Hoan, and S. S. Virmani. 2002. Two-line hybrid rice breeding in and outside of China. p. 6. In: Abstract of the Fourth Intl. Symp. Hybrid Rice, 14–17 May 2002, Hanoi, Vietnam. Lu, X., Z. Zhang, S. S. Virmani, and K. Maruyama. 1994. Current status of two-line method of hybrid rice breeding. Chinese J. Rice Sci. 8:48–54. Lübberstedt, T., A. E. Melchinger, S. Fähr, D. Klein, A. Dally, and P. Westhoff. 1998. QTL mapping in testcrossing of flint lines of maize: II. Comparison across populations for forage traits. Crop Sci. 38:1278–1289. Lucken, K. A. 1986. The breeding and production of hybrid wheat. p. 87–107. In: E. L. Smith (ed.), Genetic improvement in yield of wheat. CSSA and ASA, Madison, WI. Lucken, K. A., and K. D. Johnson. 1988. Hybrid wheat status and outlook. p. 243–255. In: International Rice Research Institute (ed.), Hybrid rice. P.O. Box 933, Manila, Philippines. Luo, L. J., Z.-K. Li, H. W. Mei, Q. Y. Shu, R. Tabien, D. B. Zhong, C. S. Ying, J. W. Stansel, G. S. Khush, and A. H. Paterson. 2001. Overdominant epistatic loci are the primary genetic basis of inbreeding depression and heterosis in rice. II. Grain yield components. Genetics 158:1755–1771. Lyman, R. F., C. Lai, and T. F. Mackay. 1999. Linkage disequilibrium mapping of molecular polymorphisms at the scabrous locus associated with naturally occurring variation in bristle number in Drosophila melanogaster. Genet. Res. 74:303–311. Lynch, M., and B. Walsh. 1998. Genetics and analysis of quantitative traits. Sinauer Associates, Sunderland, MA. Ma, G., and L. P. Yuan. 2002. Achievements and development of hybrid rice in China. p. 22. In: Abstract of the Fourth International Symposium on Hybrid Rice, 14–17 May 2002, Hanoi, Vietnam.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

165

Ma, Z. Q., and M. E. Sorrells. 1995. Genetic analysis of fertility restoration in wheat using restriction fragment length polymorphisms. Crop Sci. 35:1137–1143. Ma, Z. Q., Y. H. Zhao, and M. E. Sorrells. 1995. Inheritance and chromosomal locations of male fertility restoring gene transferred from Aegilops unbellulata Zhuk. to Triticum aestivum L. Mol. Gen. Genet. 247:351–357. Mackay, T. F. C. 1995. The genetic basis of quantitative variation: number of sensory bristles of Drosophila melanogaster as a model system. Trends Genet. 11:464–470. Mackill, D. J., and J. Ni. 2001. Molecular mapping and marker-assisted selection for majorgene traits in rice. p. 137–151. In: G. S. Khush, D. S. Brar, and B. Hardy (eds.), Rice genetics IV, Proc. Fourth Intl. Rice Genetics Symp., 22–27 Oct. 2000, Los Baños, Philippines. Science Publ., Inc., New Delhi, India, and Intl. Rice Res. Inst., Los Baños, Philippines. Mackill, D. J., Z. Zhang, E. D. Redoña, and P. M. Colowit. 1996. Level of polymorphism and genetic mapping of AFLP markers in rice. Genome 39:969–977. Maheswaran, M., N. Huang, S. R. Sreerangasamy, and S. R. McCouch. 2000. Mapping quantitative trait loci associated with days to flowering and photoperiod sensitivity in rice (Oryza sativa L.). Mol. Breed. 6:145–155. Martin, J. M., L. E. Talbert, S. P. Lanning, and N. K. Blake. 1995. Hybrid performance in wheat as related to parental diversity. Crop Sci. 35:104–108. Maruyama, K., H. Kato, and H. Araki. 1991. Mechanized production of F1 seeds in rice by mixed planting. Japan Agr. Res. Quart. 4:243–252. Matsui, K., Y. Mano, S. Taketa, N. Kawada, and T. Komatsuda. 2001. Molecular mapping of a fertility restoration locus (Rfm1) for cytoplasmic male sterility in barley (Hordeum vulgare L.). Theor. Appl. Genet. 102:477–482. McCouch, S. R., X. Chen, O. Panaud, S. Temnykh, Y. Xu, Y. G. Cho, N. Huang, T. Ishii, and M. Blair. 1997. Microsatellite marker development, mapping and applications in rice genetics and breeding. Plant Mol. Biol. 35:89–99. McCouch, S. R., L. Teytelman, Y. Xu, K. B. Lobos, K. Clare, M. Walton, B. Fu, R. Magnirang, Z. Li, Y. Xing, Q. Zhang, I. Kono, M. Yano, R. Fjellstrom, G. DeClerck, D. Schneider, S. Cartinhour, D. Ware, and L. Stein. 2002. Development of 2240 new SSR markers for rice (Oryza sativa L.) DNA Res. 9:199–207. Meghi, M. R., J. W. Dudley, R. J. Lamkey, and G. F. Sprauge. 1984. Inbreeding depression, inbred and hybrid grain yields, and other traits of maize genotypes representing three eras. Crop Sci. 24:545–549. Mei, M. H., X. K. Da, C. G. Xu, and Q. Zhang. 1999. Mapping and genetic analysis of the genes for photoperiod-sensitive genic male sterility in rice using the original mutant Nongken 58S. Crop Sci. 39:1711–1715. Melchinger, A. E. 1993. Use of RFLP markers for analyses of genetic relationships among breeding materials and prediction of hybrid performance. p. 621–628. In: D. R. Duxton et al. (eds.), International crop science I. CSSA, Madison, WI. Melchinger, A. E. 1999. Genetic diversity and heterosis. p. 99–118. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Melchinger, A. E., and R. K. Gumber. 1998. Overview of heterosis and heterotic groups in agronomic crops. p. 29–44. In: K. R. Lamkey and J. E. Staub (eds.), Concepts and breeding of heterosis in crop plants. CSSA, Madison, WI. Melchinger, A. E., M. Lee, K. R. Lamkey, and W. L. Woodman. 1990. Genetic diversity for restriction fragment length polymorphisms: relation to estimated genetic effects in maize inbreds. Crop Sci. 30:1033–1040.

166

Y. XU

Melchinger, A. E., M. M. Messmer, M. Lee, W. L. Woodman, and K. R. Lamkey. 1991. Diversity and relationships among U.S. maize inbreds revealed by restriction fragment length polymorphism. Crop Sci. 31:669–678. Melchinger, A. E., J. Boppenmaier, B. S. Dhillon, W. G. Pollmer, and R. G. Herrmann. 1992. Genetic diversity for RFLPs in European maize inbreds. II. Relation to performance of hybrids within versus between heterotic groups for forage traits. Theor. Appl. Genet. 84:627–681. Melchinger, A. E., A. Graner, M. Singh, and M. M. Messmer. 1994. Relationships among European germplasm: I. Genetic diversity among winter and spring cultivars revealed by RFLPs. Crop Sci. 34:1191–1199. Meng, J. L., A. Sharpe, C. Bowman, Z. H. Tian, T. D. Fu, X. Z. Qian, and D. Lydiate. 1996. Genetic diversity of Brassica napus revealed by RFLP markers. Acta Genetica Sinica 23:293–306. Messmer, M. M., A. E. Melchinger, R. G. Herrmann, and J. Boppenmaier. 1993. Relationships among early European maize inbreds: II. Comparisons of pedigree and RFLP data. Crop Sci. 33:944–950. Miedaner, T., C. Glass, F. Dreyer, P. Wilde, H. Wortmann, and H. H. Geiger. 2000. Mapping of genes for male-fertility restoration in ‘Pampa’ CMS winter rye (Secale cereale L.). Theor. Appl. Genet. 101:1226–1233. Miller, J. F. 1999. Oilseeds and heterosis. p. 399–404. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Miller, J. F., and K. A. Lucken. 1976. Hybrid wheat seed production methods for North Dakota. Crop Sci. 16:217–221. Mitchell-Olds, T. 1995. Interval mapping of viability loci causing heterosis in Arabidopsis. Genetics 140:105–109. Moll, R. H., M. F. Lindsey, and H. F. Robinson. 1963. Estimates of genetic variances and level of dominance in maize. Genetics 49:411–423. Moncada, P., C. P. Martinez, J. Borrero, M. Chatel, H. Gauch, E. Guimaraes, J. Tohme, and S. R. McCouch. 2001. Quantitative trait loci for yield and yield components in an Oryza sativa × Oryza rufipogon BC2F2 population evaluated in an upland environment. Theor. Appl. Genet. 102:41–52. Monna, L., H. X. Lin, S. Kojima, T. Sasaki, and M. Yano. 2002. Genetic dissection of a genomic region for a quantitative trait locus, Hd3, into two loci, Hd3a and Hd3b, controlling heading date in rice. Theor. Appl. Genet. 104:772–778. Moreau, L., A. Charcosset, F. Hospital, and A. Gallais. 1998. Marker-assisted selection efficiency in population of finite size. Genetics 148:1353–1365. Mori, K. 1984. Inheritance of susceptible mutant in rice plant to herbicide bentazon. Japan. J. Breed. 34 (Suppl 1):340–341. Moser, H., and M. Lee. 1994. RFLP variation and genealogical distance, multivariate distance, heterosis, and genetic variance in oats. Theor. Appl. Genet. 87:947–956. Mou, T., C. Li, G. Yang, and X. Lu. 1995. Genetical studies on seedling leaf color in purple rice. Chinese J. Rice Sci. 9:45–48. Mumm, R. H., and J. W. Dudley. 1994. Classification of 148 U.S. maize inbreds. I. Cluster analysis based on RFLPs. Crop Sci. 34:842–851. Murayana, S., H. Yamagishi, and T. Terachi. 1999. Identification of RAPD and SCAR markers linked to a restorer gene for Ogura cytoplasmic male sterility in radish (Raphanus sativus L.) by bulked segregant analysis. Breed. Sci. 49:115–121. Nguyen, H. T., R. Chandra Babu, and A. Blum. 1997. Breeding for drought tolerance in rice: physiology and molecular genetics considerations. Crop Sci. 37:1426–1434.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

167

Ni, J. J., P. Wu, D. Senadhira, and N. Huang. 1998. Mapping QTLs for phosphorus deficiency tolerance in rice (Oryza sativa L.). Theor. Appl. Genet. 97:1361–1369. Ni, Z. F., Q. X. Sun, Z. Y. Liu, and T. C. Huang. 1997. Studies on heterotic grouping in wheat: II. Genetic diversity among common wheat, Tibet semi-wild wheat and spelt wheat. J. Agr. Biotech. (China) 5:103–111. Oka, H. I. 1988. Origin of cultivated rice. Japan Scientific Societies Press, Elsevier, Tokyo. Olufowote, J. O., Y. Xu, X. Chen, W. D. Park, H. M. Beachell, R. H. Dilday, M. Goto, and S. R. McCouch. 1997. Comparative evaluation of within-cultivar variation of rice (Oryza sativa L.) using microsatellite and RFLP markers. Genome 40:370–378. Paterson, A. H. 1996. DNA marker-assisted crop improvement. p. 71–83. In: A. H. Paterson (ed.), Genome mapping in plants. R. G. Landes Company, Austin, TX. Paterson, A. H. 1998. Molecular dissection of complex traits. CRC Press, Boca Raton, FL. Paterson, A. H., S. D. Tanksley, and M. E. Sorrells. 1991. DNA markers in plant improvement. Adv. Agron. 46:39–90. Peng, Z. B., X. Z. Liu, J. H. Fu, L. C. Li, and C. L. Huang. 1998. Preliminary studies on the superior inbred groups and construction of heterosis mode. Acta Agron. Sinica 24:711–717. Pillen, K., G. Steinrucken, R. G. Herrmann, and C. Jung. 1993. An extended linkage map of sugar beet (Beta vulgaris L.) including nine putative lethal genes and the restorer gene X. Plant Breed. 111:265–272. Pollak, L. M., C. O. Gardner, A. L. Kahler, and M. Thomas-Compton. 1984. Further analysis of the mating system in two mass selected populations of maize. Crop Sci. 24:793–796. Quillet, M. C., N. Madjidian, Y. Griveau, H. Serieys, M. Tersac, M. Lorieux, and A. Berville. 1995. Mapping genetic factors controlling pollen viability in an interspecific cross in Helianthus sect. Helianthus. Theor. Appl. Genet. 91:1195–1202. Ramage, R. T. 1983. Heterosis and hybrid seed production in barley. p. 71–93. In: R. Frankel (ed.), Monographs on theoretical and applied genetics, Vol. 6, Heterosis. Springer-Verlag, Berlin and Heidelberg. Reddy, O. U. K., E. A. Siddiq, N. P. Sarma, J. Ali, A. J. Hussain, P. Nimmakayala, P. Ramasamy, S. Pammi, and A. S. Reddy. 2000. Genetic analysis of temperaturesensitive male sterility in rice. Theor. Appl. Genet. 100:794–801. Reeves, T., P. Pinstrup-Anderson, and R. Randya-Lorch. 1999. Food security and role of agricultural research. p. 1–5. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Ribaut, J.-M., G. O. Edmeades, F. J. Betrán, C. Jiang, and M. Bänziger. 1999. Markerassisted selection for improving drought tolerance in tropical maize. p. 193–209. In: O. Ito, J. O’Toole, and B. Hardy (eds.), Genetic improvement of rice for water-limited environments. Intl. Rice Res. Inst., Manila, Philippines. Ron Parra, J., and A. R. Hallauer. 1997. Utilization of exotic maize germplasm. Plant Breed. Rev. 14:165–187. Roos, E. E. 1984. Genetic shifts in mixed bean populations. I. Storage effects. Crop Sci. 24:240–244. Roos, E. E. 1988. Genetic changes in a collection over time. HortScience 23:86–90. Rutger, J. N., and C. N. Bollich. 1991. Use of introduced germplasm in U.S. rice improvement. p. 1–13. In: Use of plant introductions in cultivar development, Part I. CSSA Special Publ. no. 17 Rutger, J. N., M. L. Peterson, and C. H. Hu. 1977. Registration of Calrose 76 rice. Crop Sci. 17:978.

168

Y. XU

Saghai Maroof, M. A., G. P. Yang, Q. Zhang, and K. A. Gravois. 1997. Correlation between molecular marker distance and hybrid performance in US southern long grain rice. Crop Sci. 37:145–150. Sanchez, A. C., D. S. Brar, N. Huang, Z. Li, and G. S. Khush. 2000. Sequence tagged site marker-assisted selection for three bacterial blight resistance genes in rice. Crop Sci. 40:792–797. Sant, V. J., A. G. Patankar, N. D. Sarode, L. B. Mhase, M. N. Sainani, R. B. Deshmukh, P. K. Ranjekar, and V. S. Gupta. 1999. Potential of DNA markers in detecting divergence and in analyzing heterosis in Indian elite chickpea cultivars. Theor. Appl. Genet. 98:1217–1225. Schoen, D. J., and A. H. D. Brown. 1995. Maximising genetic diversity in core collections of wild relatives of crop species. p. 55–76. In: T. Hodgkin, A. H. D. Brown, Th. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Wiley, West Sussex, England. Shen, L., B. Courtois, K. L. McNally, S. Robin, and Z. Li. 2001. Evaluation of near-isogenic lines of rice introgressed with QTLs for root depth through marker-aided selection. Theor. Appl. Genet. 103:75–83. Shen, Y. W., Z. Q. Guan, J. Lu, J. Y. Zhuang, K. L. Zheng, M. W. Gao, and X. M. Wang. 1998. Linkage analysis of a fertility restoring mutant generated from CMS rice. Theor. Appl. Genet. 97:261–266. Shen, Z., and Y. Xu. 1992. Genetic analysis for hybrid sterility of intersubspecific crosses in rice. Chinese J. Rice Sci. 6:105–112. Shi, M. S. 1981. Preliminary report of later japonica natural 2-lines and applications. Hubei Agr. Sci. (7):1–3. Shull, G. H. 1908. The composition of a field of maize. Am. Breed. Assoc. 4:296–301. Simmonds, N. W. 1979. Principles of crop improvement. Longman, London. Sisco, P. H. 1991. Duplications complicate genetic mapping of Rf4, a restorer gene for cmsC cytoplasmic male sterility in corn. Crop Sci. 31:1261–1266. Smith, J. S. C. 1986. Genetic diversity within the corn belt dent racial complex of maize (Zea mays L.). Maydica 21:349–367. Smith, O. S., J. S. C. Smith, S. L. Bowen, R. A. Tenborg, and S. J. Wall. 1990. Similarities among a group of elite maize inbreds as measured by pedigree, F1 grain yield, grain yield heterosis, and RFLPs. Theor. Appl. Genet. 80:833–840. Smith, S., and T. Helentjaris. 1996. DNA fingerprinting and plant variety protection. p. 95–110. In: A. H. Paterson (ed.), Genome mapping in plants. R. G. Landes Company, Austin, TX. Sorrells, M. E., and W. A. Wilson. 1997. Direct classification and selection of superior alleles for crop improvement. Crop Sci. 37:691–697. Stebbins, G. L. 1957. Self fertilization and population variability in the higher plants. Am. Nat. 91:337–354. Stebbins, G. L. 1970. Adaptive radiation of reproductive characteristics in angiosperms: I. Pollination mechanisms. Annu. Rev. Ecol. Syst. 1:307–326. Stoyanova, S. D. 1991. Genetic shifts and variations of gliadins induced by seed aging. Seed Sci. Technol. 19:363–371. Stuber, C. W. 1994. Heterosis in plant breeding. Plant Breed. Rev. 12:227–251. Stuber, C. W. 1995. Mapping and manipulating quantitative traits in maize. Trend. Genet. 11:477–481. Stuber, C. W., S. E. Lincoln, D. W. Wolff, T. Helentjaris, and E. S. Lander. 1992. Identification of genetic factors contributing to heterosis in a hybrid from two elite maize inbred lines using molecular markers. Genetics 132:823–839.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

169

Stuber, C. W., M. Polacco, and M. L. Senior. 1999. Synergy of empirical breeding, markerassisted selection, and genomics to increase crop yield potential. Crop Sci. 39:1571–1583. Subudhi, P. K., R. P. Borkakati, S. S. Virmani, and N. Huang. 1997. Molecular mapping of a thermosensitive genetic male sterility gene in rice using bulked segregant analysis. Genome 40:188–194. Suh, H. S., D. S. Lee, L. Chen, and S. J. Kim. 2002. Hybrid rice seed production using herbicide-resistance GMS lines. p. 48. In: Abstr. Fourth Intl. Symp. Hybrid Rice, 14–17 May 2002, Hanoi, Vietnam. Sun, Q. X., T. C. Huang, Z. F. Ni, and D. J. Procunier. 1996. Studies on heterotic grouping in wheat: I. Genetic diversity between varieties revealed by RAPD. J. Agr. Biotech. (China) 4:103–110. Sun, Q. X., Z. F. Ni, X. Y. Chen, Z. Y. Liu, and T. C. Huang. 1997. Partial genome heterosis in winter wheat. J. China Agr. Univ. 2:64, 116. Takahashi, Y., A. Shomura, T. Sasaki, and M. Yano. 2001. Hd6, a rice quantitative trait locus involved in photoperiod sensitivity, encodes the alpha subunit of protein kinase CK2. Proc. Natl. Acad. Sci. (USA) 98:7922–7927. Tan, X. L., A. Vanavichit, S. Amornsilpa, and S. Trangoonrung. 1998. Genetic analysis of rice CMS-WA fertility restoration based on QTL mapping. Theor. Appl. Genet. 96:994–999. Tan, Y. F., J. X. Li, S. B. Yu, Y. Z. Xing, C. G. Xu, and Q. Zhang. 1999. The three important traits for cooking and eating quality of rice grain are controlled by a single locus in an elite rice hybrid, Shanyou 63. Theor. Appl. Genet. 99:642–648. Tanksley, S. D. 1993. Mapping polygenes. Annu. Rev. Genet. 27:205–233. Tanksley, S. D., and S. R. McCouch. 1997. Seed banks and molecular maps: unlocking genetic potential from the wild. Science 277:1063–1066. Tanksley, S. D., and J. C. Nelson. 1996. Advanced backcross QTL analysis: a method for the simultaneous discovery and transfer of valuable QTLs from unadapted germplasm into elite breeding lines. Theor. Appl. Genet. 92:191–203. Tanksley, S. D., and C. M. Rick. 1980. Isozyme gene linkage map of the tomato: applications in genetics and breeding. Theor. Appl. Genet. 57:161–170. Tanksley, S. D., N. D. Young, A. H. Paterson, and M. W. Bonierbale. 1989. RFLP mapping in plant breeding: new tools for an old science. Bio/Technology 7:257–263. Temnykh, S., W. D. Park, N. Ayres, S. Cartinhour, N. Hauck, L. Lipovich, Y. G. Cho, T. Ishii, and S. R. McCouch. 2000. Mapping and genome organization of microsatellite sequences in rice (Oryza sativa L.). Theor. Appl. Genet. 100:697–712. Thomson, M. J., T. H. Tai, A. M. McClung, M. E. Hinga, K. B. Lobos, Y. Xu, C. Martinez, and S. R. McCouch. 2003. Mapping quantitative trait loci for yield, yield components, and morphological traits in an advanced backcross population between Oryza rufipogon and the Oryza sativa cultivar Jefferson. Theor. Appl. Genet. (in press). Tomes, D. T. 1998. Heterosis: performance stability, adaptability to changing technology, and the foundation of agriculture as a business. p. 13–27. In: K. R. Lamkey and J. E. Staub (eds.), Concepts and breeding of heterosis in crop plants. CSSA, Madison, WI. Tsaftaris, A. S., M. Kafka, A. Polidoros, and E. Tani. 1999. Epigenetic changes in maize DNA and heterosis. p. 195–203. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. van Hintum, Th. J. L. 1995. Hierarchical approaches to the analysis of genetic diversity in crop plants. p. 23–34. In: T. Hodgkin, A. H. D. Brown, Th. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Wiley, West Sussex, England.

170

Y. XU

Vasal, S. K., H. Cordova, S. Pandey, and S. Srinivasan. 1999. Tropical maize and heterosis. p. 363–373. In: J. G. Coors and S. Pandey (eds.), Genetics and exploitation of heterosis in crops. ASA-CSSA-SSSA, Madison, WI. Vear, F., and J. F. Miller. 1993. Sunflower. p. 95–111. In: Traditional crop breeding practices: an historical review to serve as a baseline for assessing the role of modern biotechnology. OEDC, Paris, France. Virmani, S. S. 1996. Hybrid rice. Adv. Agron. 57:377–462. Virmani, S. S. 2002. Advances in hybrid rice research and development in the tropics. p. 4. In: Abstr. Fourth Intl. Symp. Hybrid Rice, 14–17 May 2002, Hanoi, Vietnam. Virmani, S. S., and P. C. Voc. 1991. Induction of photo- and thermo-sensitive male sterility in indica rice. Agron. Abst. 119. Virmani, S. S., and F. U. Zaman. 1998. Improving grain quality of hybrid rice: challenges, strategies, and achievements. p. 177–186. In: S. S. Virmani, E. A. Siddiq, and K. Muralidharan (eds.), Advances in hybrid rice technology. Proc. Third Intl. Symp. Hybrid Rice, 14–16 Nov. 1996, Hyderabad, India. International Rice Research Institute, Manila, Philippines. Vuylsteke, M., M. Kuiper, and P. Stam. 2000. Chromosomal regions involved in hybrid performance and heterosis: their AFLP®-based identification and practical uses in prediction models. Heredity 85:208–218. Walsh, B. 2004. Population- and quantitative-genetic models of selection limits. Plant Breed. Rev. 24 (in press). Wan, J. M., S. Yanagihara, H. Kato, and H. Ikehashi. 1993. Multiple alleles at a new locus causing hybrid sterility between Korean indica variety and a javanica variety in rice (Oryza sativa L.). Japan. J. Breed. 43:507–516. Wang, B., W. W. Xu, J. Z. Wang, W. Wu, H. G. Zheng, Z. Y. Yang, J. D. Ray, and H. T. Nguyen. 1995. Tagging and mapping in the thermosensitive genic male sterility gene in rice (Oryza sativa) with molecular markers. Theor. Appl. Genet. 91:1111–1114. Wang, J., Y. Xu, and Z. Shen. 1991. Exploring the major restrains to the utilization of Indica-Japonica hybrids in rice. Scientia Agr. Sinica 24(1):27–33. Wang, J., K. D. Liu, C. G. Xu, X. H. Li, and Q. Zhang. 1998. The high level of widecompatibility of variety ‘Dular’ has a complex genetic basis. Theor. Appl. Genet. 97:407–412. Wen, L., H. V. Tang, W. Chen, R. Chang, D. R. Pring, P. E. Klein, K. L. Childs, and R. R. Klein. 2002. Development and mapping of AFLP markers linked to the sorghum fertility restorer gene rf4. Theor. Appl. Genet. 104:577–585. Wenzel, W. G., and A. J. Pretorius. 2000. Heterosis and xenia in sorghum malt quality. South-African J. Plant Soil 17(2):66–69. Wilson, J. A. 1968. Problems in hybrid wheat breeding. Euphytica 17 (Suppl. 1):13–33. Wilson, J. H. 1979. Rice in California. Butte County Rice Growers Association. Wilson, P., and C. J. Driscoll. 1983. Hybrid wheat. p. 94–123. In: R. Frankel (ed.), Monographs on theoretical and applied genetics, Vol. 6, Heterosis. Springer-Verlag, Berlin and Heidelberg. Wise, R. P., and P. S. Schnable. 1994. Mapping complementary genes in maize: positioning the rf1 and rf2 nuclear-fertility restorer loci of Texas (T) cytoplasm relative to RFLP and visible markers. Theor. Appl. Genet. 88:785–795. Wise, R. P., K. Gobelman-Werner, D. Pei, C. L. Dill, and P. S. Schnable. 1999. Mitochondrial transcript processing and restoration of male fertility in T-cytoplasm maize. J. Hered. 90:380–385. Wu, M. S., S. C. Wang, and J. R. Dai. 2000. Application of AFLP markers to heterotic grouping of elite maize inbred lines. Acta Agron. Sinica 26:9–13.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

171

Wu, X. J., H. Q. Yin, and H. Yin. 1991. Preliminary study of the temperature effect on Annong S-1 and W6154S. Crop Res. (China) 5(2):4–6. Xiao, J., J. Li, L. Yuan, and S. D. Tanksley. 1995. Dominance is the major genetic basis of heterosis in rice as revealed by QTL analysis using molecular markers. Genetics 140:745–754. Xiao, J., J. Li, L. Yuan, S. R. McCouch, and S. D. Tanksley. 1996a. Genetic diversity and its relationship to hybrid performance and heterosis in rice as revealed by PCR-based markers. Theor. Appl. Genet. 92:637–643. Xiao, J., J. Li, L. Yuan, and S. D. Tanksley. 1996b. Identification of QTLs affecting traits of agronomic importance in a recombinant inbred population derived from a subspecific rice cross. Theor. Appl. Genet. 92:230–244. Xiao, J., L. Li, S. Grandillo, L. Yuan, S. D. Tanksley, and S. R. McCouch. 1998. Identification of trait-improving quantitative trait loci alleles from a wild rice relative, Oryza rufipogon. Genetics 150:899–909. Xu, Y. 1992. Molecular markers in quantitative trait locus mapping. Hereditas (Beijing) 14(2):45–48. Xu, Y. 1994. Application of molecular markers in genetic improvement of quantitative traits in plants. p. 38–49. In: Proc. Third Young Scientists Symp. Crop Genetics and Breeding. Publ. House of Agricultural Science and Technology of China, Beijing, China. Xu, Y. 1997. Quantitative trait loci: separating, pyramiding, and cloning. Plant Breed. Rev. 15:85–139. Xu, Y. 2002. Global view of QTL: Rice as a model. p. 109–134. In: M. S. Kang (ed.), Quantitative genetics, genomics and plant breeding. CAB International, Wallingford, UK. Xu, Y., H. M. Beachell, R. Dilday, and S. R. McCouch. 1997a. A genome-wide view of rice germplasm. Plant and Animal Genome V, Poster 62. Xu, Y., H. M. Beachell, and S. R. McCouch. 1997b. Genetic diversity of 256 rice germplasm accessions revealed by microsatellite and RFLP markers. General Meeting Intl. Program on Rice Biotechnol., Sept. 15–19, 1997, Malacca, Malaysia. Xu, Y., H. Beachell, and S. R. McCouch. 2004. A marker-based approach to broadening the genetic base of rice (Oryza sativa L.) in the U.S. Crop Sci. (in press) Xu, Y., K. B. Lobos, and K. M. Clare. 2002. Development of SSR markers for rice molecular breeding. p. 49. In: Proc. Twenty-Ninth Rice Technical Working Group Meeting, February 24–27, 2002. Little Rock, AR. Xu, Y., and L. Luo. 2002. Biotechnology and germplasm resource management in rice. p. 229–250. In: L. Luo, C. Ying, and S. Tang (eds.), Rice germplasm resources. Hubei Sci. Tech. Publisher, Wuhan, China. Xu, Y., S. R. McCouch, and Z. Shen. 1998. Transgressive segregation of tiller angle in rice caused by complementary action of genes. Crop Sci. 38:12–19. Xu, Y., and Z. Shen. 1987. Inheritance of stigma exertion in rice. Rice Genet. Newsl. 4:76–78. Xu, Y., and Z. Shen. 1988a. Receptivity of exerted stigma. Intl. Rice Res. Newsl. 13(3):7–8. Xu, Y., and Z. Shen. 1988b. Maternal effect on chalkiness in rice kernels. Rice Genet. Newsl. 5:111–113. Xu, Y., and Z. Shen. 1988c. Study on improving percentage of cross-pollination in rice. II. Inheritance of stigma exertion and development of exerted stigma in Oryza sativa L. Acta Agr. Univ. Zhejiangensis 14:66–70. Xu, Y., and Z. Shen. 1992. Detection and genetic analyses of the gene dispersed crosses: Some theoretical considerations. Acta Agr. Univ. Zhejiangensis 18:109–117 (in English with Chinese abstr.). Xu, Y., Z. Shen, C. Ying, and Z. Yang. 1988. Variation in stigma exertion in rice. Intl. Rice Res. Newsl. 13(3):6–7.

172

Y. XU

Xu, Y., J. Wang, and Z. Shen. 1989. Screening indica and japonica varieties for wide compatibility. Intl. Rice Res. Newsl. 14(5):6–7. Xu, Y., and L. Zhu. 1994. Molecular quantitative genetics (in Chinese). China Agr. Press, Beijing. Xu, Y., L. Zhu, Y. Chen, C. Lu, L. Shen, P. He, and S. R. McCouch. 1997c. Tagging genes for photo-thermo sensitivity in rice using RFLP and microsatellite markers. Plant and Animal Genome V, Poster 149. Yamaguchi, Y., R. Ikeda, H. Hirasawa, M. Minami, and P. Ujihara. 1997. Linkage analysis of thermosensitive genic male sterility gene, tms-2 in rice (Oryza sativa L.). Breed. Sci. 47:371–373. Yanagihara, S., H. Kato, and H. Ikehashi. 1992. A new locus for multiple alleles causing hybrid sterility between an aus variety and javanica varieties in rice (Oryza sativa L.). Japan. J. Breed. 42:793–801. Yanagihara, S., S. R. McCouch, K. Ishikawa, Y. Ogi , K. Maruyama, and H. Ikehashi. 1995. Molecular analysis of the inheritance of the S-5 locus, conferring wide compatibility in indica/japonica hybrids of rice (Oryza sativa L.). Theor. Appl. Genet. 90:182–188. Yang, G. P., M. A. Saghai Maroof, C. G. Xu, Q. Zhang, and R. M. Biyashev. 1994. Comparative analysis of microsatellite DNA polymorphism in landraces and cultivars of rice. Mol. Gen. Genet. 245:187–194. Yang, R. C., W. M. Li, N. Y. Wang, K. J. Liang, and Q. H. Chen. 1989. Discovery and preliminary study on indica photosensitive genic male sterile germplasm 5460s. Chinese J. Rice Sci. 3:47–48. Yano, M., Y. Harushima, Y. Nagamura, N. Kurata, Y. Minobe, and T. Sasaki. 1997. Identification of quantitative trait loci controlling heading date in rice using a high-density linkage map. Theor. Appl. Genet. 95:1025–1032. Yano, M., Y. Katayose, M. Ashikari, U. Yamanouchi, L. Monna, T. Fuse, T. Baba, K. Yamamoto, Y. Umehara, Y. Nagamura, and T. Sasaki. 2000. Hd1, a major photoperiod sensitivity quantitative trait locus in rice, is closely related to the Arabidopsis flowering time gene CONSTANS. Plant Cell 12:2473–2483. Yao, F. Y., C. G. Xu, S. B. Yu, J. X. Li, Y. J. Gao, X. H. Li, and Q. Zhang. 1997. Mapping and genetic analysis of two fertility restorer loci in the wild-abortive cytoplasmic male sterility system of rice (Oryza sativa L.). Euphytica 98:183–187. Yoo, B. H. 1980. Long-term selection for a quantitative character in large replicate populations of Drosophila melanogaster. I. Response to selection. Genet. Res. 35:1–17. Young, N. D. 1999. A cautiously optimistic vision for marker-assisted breeding. Mol. Breed. 5:505–510. Yousef, G. G., and J. A. Juvik. 2001. Comparison of phenotypic and marker-assisted selection for quantitative traits in sweet corn. Crop Sci. 41:645–655. Yu, K., S. J. Park, and V. Poysa. 2000. Marker-assisted selection of common beans for resistance to common bacterial blight. Plant. Breed. 119:411–415. Yu, S. B., J. X. Li, C. G. Xu, Y. F. Tan, Y. J. Gao, X. H. Li, Q. F. Zhang, and M. A. Saghai Maroof. 1997. Importance of epistasis as the genetic basis of heterosis in an elite rice hybrid. Proc. Nat. Acad. Sci. (USA). 94:9226–9231. Yuan, L. P. 1987. Strategy conception of hybrid rice breeding. Hybrid Rice (1):1–4. Yuan, L. P. 1992a. Development and prospects of hybrid rice breeding. p. 97–105. In: C. B. You and Z. L. Chen (eds.), Agricultural biotechnology. Proc. Asian Pacific Conf. Agric. Biotech., China Agr. Press, Beijing. Yuan, L. P. 1992b. The strategy of breeding rice PGMS and TGMS lines. Hybrid Rice (1):1–4. Yuan, L. P. 2002. Future outlook on hybrid rice research and development. p. 3. In: Abstract of the Fourth International Symposium on Hybrid Rice, 14–17 May 2002, Hanoi, Vietnam.

3. MARKER-ASSISTED SELECTION STRATEGIES FOR BREEDING HYBRID RICE

173

Yuan, L. P., and H. X. Chen (eds.). 1988. Breeding and cultivation of hybrid rice. Hunan Science and Technology Press, Changsha, China. Zeng, R., Z. Zhang, and G. Zhang. 2000. Identification of multiple alleles at the Wx locus in rice using microsatellite class and G-T polymorphism. p. 202–205. In: X. Liu. (ed.) Theory and application of crop research. China Science and Technology Press, Beijing. Zhang, G., T. S. Bharaj, Y. Lu, S. S. Virmani, and N. Huang. 1997. Mapping of the Rf-3 nuclear fertility-restoring gene for WA cytoplasmic male sterility in rice using RAPD and RFLP markers. Theor. Appl. Genet. 94:27–33. Zhang, J., R. Chandra Babu, G. Pantuwan, A. Kamoshita, A. Blum, L. Wade, S. Sarkarung, J. C. O’Toole, and H. T. Nguyen. 1999. Molecular dissection of drought tolerance in rice: from physio-morphological traits to field performance. p. 331–343. In: O. Ito, J. O’Toole, and B. Hardy (eds.), Genetic improvement of rice for water-limited environments. Intl. Rice Res. Inst., Manila, Philippines. Zhang, J., Y. Xu, X. Wu, and L. Zhu. 2002. A bentazon and sulfonylurea sensitive mutant: breeding, genetics and potential application in seed production of hybrid rice. Theor. Appl. Genet. 105:16–22. Zhang, Q., Y. J. Gao, M. A. Saghai Maroof, S. H. Yang, and J. X. Li. 1995. Molecular divergence and hybrid performance in rice. Mol. Breed. 1:133–142. Zhang, Q., Y. J. Gao, S. H. Yang, R. A. Ragab, M. A. Saghai Maroof, and Z. B. Li. 1994a. A diallel analysis of heterosis in elite hybrid rice based on RFLPs and microsatellites. Theor. Appl. Genet. 89:185–192. Zhang, Q., J. Hua, S. Yu, L. Xiong, and C. Xu. 2001. Genetic and molecular basis of heterosis in rice. p. 173–185. In: G. S. Khush, D. S. Brar, and B. Hardy (eds.), Rice genetics IV. Proc. Fourth Intl. Rice Genetics Symp., 22–27 Oct. 2000, Los Baños, Philippines. Science Publishers, Inc., New Delhi, India, and Intl. Rice Res. Inst., Los Baños, Philippines. Zhang, Q., and N. Huang. 1998. Mapping and molecular marker-based genetic analysis for efficient hybrid rice breeding. p. 243–256. In: S. S. Virmani, E. A. Siddiq, and K. Muralidharan (eds.), Advances in hybrid rice technology. Proc. Third Intl. Symp. Hybrid Rice, 14–16 Nov. 1996, Hyderabad, India. Intl. Rice Res. Inst., Manila, Philippines. Zhang, Q., and Z. Li. 2002. Advances in understanding the genetic basis of heterosis in rice. p. 11. In: Abstr. Fourth Intl. Symp. Hybrid Rice, 14–17 May 2002, Hanoi, Vietnam. Zhang, Q., M. A. Saghai Maroof, T. Y. Lu, and B. Z. Shen. 1992. Genetic diversity and differentiation of indica and japonica rice detected by RFLP analysis. Theor. Appl. Genet. 83:495–499. Zhang, Q., B. Z. Shen, X. K. Dai, M. H. Mei, M. A. Saghai Maroof, and Z. B. Li. 1994b. Using bulked extremes and recessive class to map genes for photoperiod-sensitive genic male sterility in rice. Proc. Natl. Acad. Sci. (USA) 91:8675–8679. Zhang, S. Q., H. H. Tong, R. Xue, Z. H. Hua, X. L. Wang, D. L. Huang, and X. B. Xie. 1998. Tests on improving purity in hybrid rice seed production by introducing bar gene to restorers. Scientia Agr. Sinica 31(6):33–37. Zhao, M. F., X. H. Li, J. B. Yang, C. G. Xu, R. Y. Hu, D. J. Liu, and Q. Zhang. 1999. Relationships between molecular marker heterozygosity and hybrid performance in intraand inter-subspecific crosses of rice. Plant Breed. 118:139–144. Zheng, K., H. Qian, B. Shen, J. Zhuang, H. Liu, and J. Lu. 1994. RFLP-based phylogenetic analysis of wide compatibility varieties in Oryza sativa L. Theor. Appl. Genet. 88:65–69. Zheng, K., P. Shen, H. Qian, and J. Wang. 1992. Tagging genes for wide compatibility in rice via linkage to RFLP markers. Chinese J. Rice Sci. 6:145–150. Zhou, T. B., H. C. Xiao, D. Y. Lei, and Q. X. Duan. 1988. The breeding of indica photosensitive male sterile line. J. Hunan Agr. Sci. (6):16–18. Zhu, L., C. Lu, P. Li, L. Shen, Y. Xu, P. He, and Y. Chen. 1996. Using doubled haploid populations of rice for quantitative trait locus mapping. p. 631–636. In: Rice genetics III.

174

Y. XU

Proc. Third Intl. Rice Genetics Symp., 16–20 Oct. 1995, Manila, Philippines. Intl. Rice Res. Inst., Manila, Philippines. Zhu, Z. F., C. Q. Sun, T. B. Jiang, Q. Fu, and X. K. Wang. 2001. The comparison of genetic divergences and its relationships to heterosis revealed by SSR and RFLP markers in rice (Oryza sativa L.). Acta Genetica Sinica 28:738–745.

4 Significance of Cytoplasmic DNA in Plant Breeding Ursula Frei, Edmundo G. Peiretti, and Gerhard Wenzel Department of Agronomy and Plant Breeding, Center for Life and Food Sciences, Technical University of Munich D-85350 Freising-Weihenstephan, Germany

I. INTRODUCTION II. SOME BASIC INFORMATION ON DNA IN THE CYTOPLASM OF PLANTS III. AGRONOMIC TRAITS INFLUENCED BY CYTOPLASMIC FACTORS A. Cytoplasmic Male Sterility (CMS) B. Yield and Quality Parameters C. Disease Resistance D. Tissue Culture Responses and Regeneration E. Combining Ability IV. BREEDING USING CYTOPLASMIC FACTORS A. Characterization of Plant Material 1. Potato 2. Maize 3. Rice 4. Wheat 5. Amaranth B. Creating New Variability 1. Undirected Processes 2. Directed Alterations by Plasmone Transformations C. Selection 1. Correlation of Phenotype and Cytoplasm 2. Somatic Recombination V. CONCLUSION LITERATURE CITED

Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 175

176

U. FREI, E. PEIRETTI, AND G. WENZEL

I. INTRODUCTION Genomics, proteomics, and most recently, metabolomics are fashionable tools in plant genetics. Most of this research concentrates in plant breeding on genetic information of the cell nucleus, the genomic DNA (Lübberstedt et al. 2002). Although the nucleus is the main organelle where genetic information is active, additional DNA is found in the cytoplasm as plasmone. This non-nuclear, plasmone DNA is located as plastome in plastids (ptDNA) and as chondriome in mitochondria (mtDNA). The size of these organelle DNA is small compared to the genomic DNA: only about 200 kb per plastid and up to 2500 kb per mitochondrion. However, the small amount of organelle DNA has to be multiplied by the number of organelles per cell, up to 100 in the case of plastids and several hundred in the case of mitochondria, and thus the non-nuclear DNA cannot be neglected when estimating the inheritance of specific traits. Nevertheless, in plant breeding programs, the importance of genetic information about mitochondria and chloroplasts is often ignored or excluded in studies of DNA polymorphisms and their correlation to phenotypes. Influences of the genetic information of the cytoplasm are normally detected via reciprocal crosses, since, in general, cytoplasmic genes are only inherited maternally. This does not mean, however, that microspores do not contain plastids or mitochondria; the regeneration of functional plants from isolated microspores during androgenetic haploid production proves their presence. Thus, the situation of a uniparental inheritance, which is a strong indication of the presence of plasmone DNA, needs an additional explanation. It is assumed that the plastids and mitochondria are stripped off of the sperm nucleus during the process of fertilization, and consequently they are not transmitted into the zygote (Hagemann 2002). Maternal inheritance was described for variegated leaf color in Pelargonium zonale at the beginning of the last century (Baur 1909). The genetic basis of this phenomenon could not be explained at that time, but today we know that eucaryotic plant cells contain ptDNA and mtDNA that interact with the nuclear genetic information and interact with each other (Hedtke et al. 1999; Comai 2000; Herrmann et al. 2003). They contribute in cell metabolism to photosynthesis and respiration. Furthermore, cytoplasmic genes contribute to the operation of the sexual reproductive systems and to the differentiation and evolution of plant species, including maize and teosinte (Poggio et al. 1997), potato (Hosaka and Hanneman 1988), rubber tree (Luo et al. 1995), soybean (Xu et al. 2002), and wheat (Vedel et al. 1978; Wang et al. 2000). Thus, mitochondria and chloroplasts represent a source of genetic diversity, as they contain DNA and the machinery to express that information as RNA and,

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

177

after RNA editing and tRNA import, by their own translation system, as proteins (Stern et al. 1997; Steinhauser et al. 1999; Yamaguchi and Knoblauch 2000; Bock 2001). The expression of plastid genes may be enhanced under light and may show tissue specificity that leads to increased protein production in chloroplasts of photosynthetically active tissues (Bruik and Mayfield 1999). This review will summarize which agronomic traits are influenced by cytoplasmic factors or nuclear-cytoplasmic and/or organelle interaction and where knowledge about cytoplasmic DNA is ready to be considered in breeding programs. A first area in cytoplasm research was the production and use of alloplasmic lines by combination breeding, where by successive backcrossing of two parents the cytoplasm of one is changed. A second tool was provided by the field of somatic cell fusion, circumventing uniparental inheritance and permitting complete combination of cytoplasms (Hanson and Folkerts 1992; Lössl et al. 1994). The third application is opened up by direct transfer of genes into organelles, resulting in the case of plastids in transplastomic plants. Most present knowledge has been driven by the applied goals of hybrid breeding, and consequently much of the available information is based on trials to understand and subsequently to use the mitochondrial trait of cytoplasmic male sterility (CMS). Before discussing applied approaches, some basic information will be summarized. II. SOME BASIC INFORMATION ON DNA IN THE CYTOPLASM OF PLANTS Most basic information on the plasmone structure is published for mitochondria of microorganisms and animals, a substantial part of which can, however, be applied to plant organelles (Scheffler 1999). With the very first analyses of ptDNA, it became evident that plastids and eubacterial genomes are related (Schwarz and Kössel 1979). According to the endosymbiont hypothesis, plastids originated from cyanobacteria, whereas it has been suggested that the ancestors of mitochondria were eubacteria (Gray et al. 1999), and thus research on microorganisms offers a lot of information about higher plant plasmones. Substantial basic plant knowledge has been gained from model plants such as tobacco or Arabidopsis (Abdallah et al. 2000; Barkan and Goldschmidt-Clermont 2000; Lim et al. 2000); and transfer of these findings to crop plants has started. Reviews of the physical structure, information content, RNA editing, and interaction with the nucleus are given for plant mitochondria by Knoop and Brennicke (2002) and for plastids by Bock and Hippler (2002). Using molecular tools, research on non-nuclear DNA of higher plants resulted in sequencing of the DNA from maize mitochondria (Lonsdale

178

U. FREI, E. PEIRETTI, AND G. WENZEL

et al. 1984), from tobacco (Shinozaki et al. 1986), and recently from wheat chloroplasts (Ogihara et al. 2002). A first rather simple and descriptive use of information deduced from organelle sequencing shows the genetic diversity of pt and mtDNA. Polymorphisms of plasmone DNA could be used to elucidate phylogenetic relationships of crops such as Zea (Timothy et al. 1979), Hordeum (Neale et al. 1986), Triticum and Aegilops (Breiman 1987), Theobroma (Laurent et al. 1993), Hevea (Luo et al. 1995), pines (Powell et al. 1995), Coffea (OrozcoCastillo et al. 1996), Allium (Friesen et al. 1999), Solanum (Bryan et al. 1999), rice (Sun et al. 2002), or soybean (Xu et al. 2002). For the breeder, such information about the genetic diversity of the plasmone is a helpful addition to the characterization of nuclear genomes, and thus helps in parent selection. By comparing two fully sequenced ptDNAs, in this particular case Atropa and Nicotiana plastomes, Schmitz-Linneweber et al. (2002) found that promoters and translational and replicational signals are well conserved between the two species. The authors concluded that pt-genes are even more conserved, but RNA editotypes differ between species, which makes RNA editing a major trait in speciation. This may have a substantial effect on plant fitness via protein function. The mtDNA of higher plants has a large coding capacity and encodes, in addition to the standard set, for the five mitochondrial respiratory chain complexes, and for additional proteins like cytochrome c oxidase or the α subunit of ATPase. The model plant Arabidopsis has 57 mtgenes, and the gene number varies between species (Scheffler 1999). Mitochondrial DNA is constantly reorganized, probably by an own recombination system. In order to account for the structural rearrangements, the presence of repeated sequences in direct or inverted orientation has been noted. Since the DNA is circular, two direct repeats at some distance in the master circle will yield two subgenomic circles. Recombination between a subgenomic circle and the master circle can give rise by homologous recombination to larger circles with duplications, permuting the mt-genome without mutating the coding regions and ending up in a wide variability. In contrast to this variability, sequences of individual genes are remarkably constant. The mtDNA has captured genes encoded in the nucleus, suffers mutations that can effect growth and sterility, and yields transcripts that have to be edited, or trans-spliced, but lack a subset of tRNAs that have to be imported from the cytoplasm (Herrmann 1992; Scheffler 1999). The synthesis of mRNA from the ptDNA in tobacco and its in vitro translation provided the final proof for pt-protein synthesis, though one interdependent with the nucleus (Hirose and Sugiura 1997). Since it is possible to transform chloroplasts of higher plants (Svab et al. 1990), the ptDNA can be engineered to permit study of even those steps in gene

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

179

expression for which no in vitro systems are currently available. In addition, plastid transformation technology has allowed the functional characterization of ptDNA-encoded open reading frames (orf) by reverse genetics (Ruf et al. 1997), and by the use of plastome mutants (Bock and Hippler 2002). This research on the pt-genome is still of an experimental nature, mostly done on model plants, but, in the view of the potential environmental concerns of nuclear genetic engineering, pt- or mt-transformation can help to create a strong barrier against undesired vertical gene transfer via pollen (Daniell 2002). Consequently, this research will increasingly be applied to important crop species. Although mitochondria and plastids retain a semiautonomous status, they depend heavily for their biogenesis and function on interaction with the nucleus (Herrmann 1992; Gray et al. 1999). Interaction between the nucleus and the cytoplasm has been well studied in several crops, such as Beta (Samitou-Laparde et al. 1991), Brassica (Gourret et al. 1992), Helianthus (Moneger et al. 1994), Oryza (Ishii et al. 1993), Pennisetum (Virk and Brar 1993), Solanum (Hosaka 1995), Triticum (Terachi and Tsunewaki 1992), and Zea (Rocheford and Pring 1994). Cytoplasms of Triticum species and related genera differ with regard to the effects on the expression and interaction of nuclear genes. The cytoplasmic diversity detected within a series of distantly related varieties seems to correspond to their nuclear divergence. This predicts that nuclear-cytoplasmic interactions should not be neglected when interpreting a phenotypic segregation in a breeding program. The first experiments to test the nuclear-organelle interaction were carried out with microorganisms. In yeast, for example, a large number of nuclear genes have been identified that influence, for example, the respiratory chain or oxidative phosphorylation system of mitochondria (Scheffler 1999). The interaction is either by transcriptional regulation or post-transcriptional and affects mRNA processing and the stability of translation (Grivell 1995). In the meantime, these findings could be verified for higher plants, and it became quite clear that organelles may also control the nuclear gene expression, as shown for chloroplasts by Gray et al. (1999). III. AGRONOMIC TRAITS INFLUENCED BY CYTOPLASMIC FACTORS A principal problem in applied reports of cytoplasmic effects on agronomic traits is that many have been evaluated in populations carrying different sources of cytoplasms, or do not differentiate plastome and chondriome. In addition there is the problem of accurately assessing cytoplasmic variation with respect to quantitative agronomic traits

180

U. FREI, E. PEIRETTI, AND G. WENZEL

(Edwards and Coors 1996). Thus, a comparison of the phenotypes of alloplasmic isonuclear lines with different backgrounds in order to evaluate their use for hybrid breeding programs is difficult. Furthermore, a clear distinction of cytoplasmic effects and nuclear-cytoplasmic interactions is complicated, depending on the nuclear genome that is combined with the cytoplasm under evaluation. The differentiation between cytoplasmic types is primarily based on their sterile phenotype controlled by CMS of the mtDNA and the corresponding nuclear restorer system. Less important and obvious are variations of the ptDNA due to differences in chloroplasts or other plastid types. A. Cytoplasmic Male Sterility (CMS) In higher plants, CMS is a maternally inherited defect in pollen production that is thought to result from the expression of unusual or aberrant mt-genes interacting with the nuclear genome (Kaul 1988; Ducos et al. 2001). Various chimeric mitochondrial genes produced by homologous recombination or via protoplast fusion are believed to disturb the mitochondrial function at a critical stage of tapetum development, thus causing male sterility (Hanson and Folkerts 1992; Dragoeva et al. 2001). Laser and Lersten (1972) reviewed cytoplasmic pollen sterility in more than 140 species of 47 genera and 20 families. Although half of the CMS types described occurred naturally, 20% of them were the result of intraspecific crosses, and the rest originated from interspecific crosses (Levings 1990). Several nuclear genes are known to control this expression of CMS and, thus, different CMS types can be distinguished by the specific nuclear genes (rfs) that restore pollen fertility. This nuclear-cytoplasmic system is extensively used in hybrid breeding programs: Currently, great effort is ongoing with the outbreeders, rye and sunflower; the facultative outbreeder, rapeseed; and in the inbreeders, wheat, rice, triticale, barley, and oats. In maize, a specific S-plasmid of the mitochondria can integrate into the mtDNA and causes CMS (Schardl et al. 1985), while in sunflower the PET1 mitochondrial mutation is associated with premature programmed cell death and cytochrome c release (Balk and Leaver 2001). The threat of a disaster such as occurred in Texas (T) cytoplasm of maize indicated a need to broaden the cytoplasmic base (Levings 1990). If no CMS is available in wild relatives of a cultivar, it may be possible to create CMS via recombination of mtDNA. Besides the production of new nuclear-cytoplasmic interactions resulting from sexual production of alloplasmic lines, this can be achieved via cell fusion. After somatic fusion, recombination of mtDNA may take place, making cell fusion also a promising tool for sexually propagated plants such as in rapeseed (Pelletier et al. 1988) or in Nicotiana cybrids (Atanassov et al. 1998), yielding new CMS lines.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

181

Since the discovery of an induction of complete male sterility in wheat with the G-type cytoplasm of Triticum timopheevi by Wilson and Ross (1962), G-type cytoplasm has been used as a source of male sterility for developing high-yield hybrid cultivars (Wilson 1967). Murai and Tsunewaki (1993) and Ogihara et al. (1999) reported a new CMS system in wheat, namely, photoperiod-sensitive cytoplasmic male sterility. It is induced by interaction between the nuclear genome of common wheat, Triticum aestivum ‘Norin 26’, and the cytoplasm of Aegilops crassa. The alloplasmic line of ‘Norin 26’ with Ae. crassa cytoplasm shows almost complete male sterility, resulting from the pistillate nature of the stamens that develop under long day conditions (15 hr or more of light), but this line is highly male fertile under short-day conditions (<15 hr of light). Ogihara et al. (1999) described this action as a result of the production of a truncated mtDNA gene product (open reading frame, orf 48) under the influence of the euplasmsic mt gene of Ae. crassa and the rps7 gene of wheat. In hybrid triticale programs it appeared difficult to permanently express CMS irrespective of the wheat cytoplasm used. The molecular investigation of genetically well defined lines revealed a nuclear dependence on gene structures as well as transcriptional and posttranscriptional expression in wheat and triticale. Wheat mtDNA exhibits multiple DNA polymorphisms and rearrangements, creating chimeric or truncated open reading frames leading to novel gene copies, for example, of orf25-gene with altered expression profiles (Kueck et al. 1995). B. Yield and Quality Parameters Agronomic traits influenced by cytoplasmic factors, aside from CMS, are summarized in Table 4.1. In most cases, alloplasmic lines were developed in sexual hybrid breeding programs for the seed (CMS) parent. Ekiz et al. (1998) studied different quality traits in bread wheat (Triticum aestivum). They reported small but significant differences between reciprocal crosses of alloplasmic lines for thousand kernel weight (TKG) and protein percentage. The authors emphasized that the behavior of a given cytoplasm may be different in different years and with different nuclei. It was difficult to find cytoplasms that had stable positive effects with all nuclear genotypes in all locations and years. Symmetric somatic hybrids are available for the vegetatively propagated potato (Wenzel 1994). Clones regenerated after protoplast fusion of two parents may have the same fused nuclear genome but varying cytoplasm compositions. Such hybrids are ideal for studying influences on agronomic traits of a broad range of recombined mitochondrial types with one of the two parental chloroplasts and with the background of an unaltered nuclear DNA genome. The investigations of Lössl et al. (1994) concentrate on yield and quality with a strong emphasis on starch production.

Table 4.1.

Agronomic traits influenced by cytoplasmic factors.

Agronomic trait

Species

Regeneration ability

Helianthus annuus

Protoplast culture Phoma macdonaldii

Helianthus annuus Helianthus annuus

Fusarium culmorum Yield parameters Eating, cooking, processing quality GCA

Hordeum vulgare Oryza sativa Oryza sativa

Grain yield, plant height, days to flower Starch content

Pennisetum glaucum

Ustilago tritici resistance Septoria nodorum resistance Anther culture response Stripe rust resistance Tissue culture response Starch B-granule content Anther culture response Rust resistance

Analyzed in

References

Triticum aestivum Triticum aestivum

Alloplasmic lines (diff. CMS sources) Alloplasmic lines Alloplasmic lines (diff. CMS sources) Alloplasmic lines Alloplasmic lines Incomplete diallel (diff. CMS and restorer lines) Alloplasmic lines (diff. CMS sources) Alloplasmic lines (diff. CMS sources) Intraspecific somatic hybrids Alloplasmic lines Alloplasmic lines

Dhitaphichit et al. 1989 Keane and Jones 1990

Triticum aestivum Triticum aestivum Triticum aestivum Triticum aestivum Triticum aestivum Vicia faba

Alloplasmic lines Alloplasmic lines Alloplasmic lines F1, F2, and BC Tritordeum Diallel of F1 hybrids

Hasan and Konzak 1991; Orlov et al. 1997 Chen and Line 1995 Bebeli 1995 Stoddard 2000 Hernandez et al. 2001 Stoddard and Hearth 2001

Pennisetum typhoides

Solanum tuberosum

Nestares et al. 1998 Bolandi et al. 1999 Rousstaee and Barrault 2000 Adamski et al. 1997 Wang et al. 1998 Bao and Xia 1999 Virk and Brar 1993; Yadav 1994 Yadav 1994 Lössl et al. 2000

182

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

183

C. Disease Resistance Disease and pest resistance is the most important breeding goal, since it stabilizes yield and quality. One example for qualitative cytoplasmic resistance is maize: Hybrids with the Texas male sterile cytoplasm (cmsT) are rendered highly susceptible to leaf blight caused by the T-toxins of Bipolaris maydis and Phyllostica maydis. Intensive research is underway to understand this interaction, which at the same time might help to elucidate host pathogen interaction. Associated with cms-T is the unidentified reading frame, URF13 (Levings 1990). The mitochondrial gene T-urf13 codes for three trans-membrane protein helices acting as receptor for the host specific toxin produced by each of the pathogens. Binding of T-toxin to URF13 leads to the formation of hydrophilic pores that permeabilize the inner mitochondrial membrane (Rhoads et al. 1996). By transfer of urf13 into Escherichia coli, this system could be further elucidated. Several chemicals protect E. coli and cms-T mitochondria from the permeability caused by T-toxin. Also, light reduces the leaking of electrolytes and thus has an influence on the sensitivity of cms-T cytoplasm to T-toxin (Garraway et al. 1998). The increasing but still incomplete knowledge about the interaction of a known gene with a known toxin underlines the complexity of a host pathogene interaction. Further examples of quantitative influences of specific male sterile cytoplasms on resistance responses are given in Table 4.1. Stoddard and Hearth (2001) reported on the selection of Vicia faba beans with increased rust resistance within a complete diallel of F1 hybrids. Highly susceptible cytoplasms could be distinguished from resistant ones by the size of the pustule developing on detached leaves. Chen and Line (1995) suggested that there are maternal or cytoplasmic factors that affect stripe rust resistance in wheat, and Keane and Jones (1990) found an effect of alien cytoplasm substitution on the response of wheat cultivars to Septoria nodorum. In sunflower, alloplasmic lines showed significant cytoplasmic effects for improved partial resistance to Phoma macdonaldii (Rousstaee and Barrault 2000). The alien cytoplasm had a positive effect on the resistance in sunflower, while the substitution of the ‘Chinese Spring’ nucleus into the cytoplasm of Aegilops squarrosa, Ae. variabilis, or Ae. mutica resulted in increased susceptibility to Ustilago tritici (Dhitaphichit et al. 1989). It cannot be ruled out that the different reactions are due to an altered interaction between the new nucleus and the Aegilops cytoplasm. Similarly, the finding of Adamski et al. (1997) that, in barley, resistance against Fusarium culmorum decreased when Hordeum vulgare cytoplasm was replaced by H. bulbosum cytoplasm may be an unspecific effect of the new mixture. Fortunately, for haploid

184

U. FREI, E. PEIRETTI, AND G. WENZEL

induction by the parthenogenetic bulbosum method, where the zygote eliminates the paternal chromosomes and results in maternal haploids (Wenzel and Foroughi-Wehr 1994), the paternal cytoplasm is not transferred. Otherwise, increased susceptibility might accompany parthenogenetic haploid induction. D. Tissue Culture Responses and Regeneration In vitro technologies to increase the efficiency of specific breeding steps include rapid propagation, androgenetic haploid induction, cell fusion, and regeneration of cells after gene transfer (Altman 1998). Even under optimized protocols, differences in the success rate among cultivars are common. Regeneration success in microspore culture of potato was achieved by the selection of clones with a high culture response. Transferring this character to progenies by sexual crosses resulted in an increased tissue culture ability of several of the F1 clones (Wenzel and Uhrig 1981; Sonnino et al. 1989). The Mendelian segregation of this trait implicated nuclear origin, but an additional influence of the plasmone during regeneration processes was apparent. Similarly, Hernandez et al. (2001) report on the beneficial influence of specific Triticum cytoplasms on the anther culture response of tritordeum, and cytoplasms from Triticum dicoccum ssp. pseudomacrotherum, T. monococcum ssp. aegilopoides, T. turgidum v. persicum fuliginosum, T. dicoccum v. khapli, and T. turgidum v. turanicum notabile. This may offer useful cytoplasmic genetic variation for enhancing anther culture response in wheat (Hasan and Konzak 1991). Orlov et al. (1997) reported on the positive effects of an interspecific cytoplasm substitution between the five species T. aestivum, T. turgidum, T. durum, T. dicoccum, and T. aetiopicum on different variables evaluated during anther culture (number of responding anthers, and number of embryoids per 100 anthers, number of green and albino regenerants). The cytoplasms of Aegilops umbellulata, Ae. triuncialis, Ae. columnaris, Ae. kotschyi, and Ae. variabilis were found to induce haploid parthenogenesis in common wheat, ‘Salmon’, at relatively high frequencies (Tsunewaki et al. 1974). Cytoplasms also have an effect on immature embryo culture leading to embryogenic suspension cultures, a prerequisite for in vitro regeneration of cereals. Bebeli (1995) compared 21 alloplasmic lines and one euplasmic line of bread wheat having a common nuclear genotype (‘Chinese Spring’ of T. aestivum) with cytoplasm belonging to different Triticum and Aegilops. In comparison to the euplasmic line, the lowest induction rates were observed in the alloplasmic lines having Ae. speltoides, bicornis, and kotschyi cytoplasms, while the best performances were shown by the triuncialis and umbellulata lines.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

185

A different way to induce haploidy by using cytoplasmic effects was described by Hsam and Zeller (1993). They developed an alloplasmic haploid-inducer in the durum wheat cultivar ‘Cando’, which possesses the homologous wheat-rye translocation 1BL/1RS from the 6x-wheat ‘Veery’. The nucleus of 4x-‘Cando-Veery 1BL/1RS’ was introduced into Ae. kotschyi cytoplasm, and 6% haploids were produced by crossing the alloplasmic durum line with ‘Cando-Veery 1BL/1RS’. The general regeneration ability of sunflower was analyzed in malesterile and male-fertile forms of inbred lines using explants from cotyledons of mature seeds (Nestares et al. 1998). Significant differences were observed between male-sterile and male-fertile forms, indicative of the universal influence of cytoplasmic genes on the cell metabolism. Furthermore, protoplast regeneration is influenced by the plasmone. Bolandi et al. (1999) evaluated the effect of different sunflower cytoplasms on protoplast culture. The complex interactions between the genotypes tested indicated that protoplast culture responses are affected independently by nuclear-cytoplasm interactions. Some nuclear-cytoplasm combinations improved protoplast culture response. E. Combining Ability High general (GCA) and specific combining abilities (SCA) of parental lines are prerequisite to a successful hybrid breeding program. Alien male-sterile cytoplasm introduced in cultivars for efficient hybrid production can influence combining ability both positively and negatively. Yadav (1994) reported such a positive influence of the A1-cytoplasm, an important source of CMS, in pearl millet hybrid breeding on GCA estimated by plant height and grain yield, and a negative influence on days to flowering. Virk and Brar (1993) studied cytoplasmic differences for several agronomic traits in near-isonuclear male-sterile lines in pearl millet. A differential behavior of cytoplasm, in combination with both a common pollinator and a cross pollinator, was observed for yield, plant height, and ear length. Since most plants contain mitochondrial populations in a heteroplasmic state, sorting is a common phenomenon. An unexpected segregation during a selection may be the result of such a dissorting. Combining ability of cytoplasms is of special importance in somatic hybridization. The fusion process combined cytoplasms as well as different nuclei. The differentiation of cytoplasm types in dihaploid potatoes revealed five distinct mitochondrial types (α, β, γ, δ, ε) associated with three chloroplast types T, W, and S (Fig. 4.1). Analyzing symmetric protoplast fusion experiments performed over several years with dihaploid clones derived by chromosome reduction of tetraploid potato cultivars, Frei et al. (1998) calculated the mean hybrid efficiency

186

U. FREI, E. PEIRETTI, AND G. WENZEL

α 38% γ 8% W 48% δ 2% ε 2%

S 2%

β 50%

T 50%

Fig. 4.1. Distribution of cytoplasm types within a pool of dihaploid breeding clones of potato. Inner circle: chloroplast types nomenclature according to Hosaka (1995). Outer circle: mitochondrial types nomenclature according to Frei et al. (1998).

(percentage of hybrids among regenerates) of different cytoplasm combinations (Fig. 4.2). Clear differences in hybrid recovery were demonstrated for some combinations. Since this may be due to competition among mitochondria, to stoichiometric shifting, or even to a new mixture of mitochondria, the observed variability may not easily be reproduced in different genotypes. Hybrid efficiency (%) 40

30

20

10

0

(16)

(26)

(4)

(7)

(15)

(3)

(1)

(1)

(1)

α (+) α α (+) β α (+) γ β (+) β β (+) γ α (+) δ α (+) ε β (+) ε γ (+) δ Cytoplasmic Composition (mt type)

Fig. 4.2. Mean hybrid efficiency (percentage of hybrids among regenerated plants; total number analyzed in brackets) of fusion combinations with different cytoplasm types of the parental fusion clones (Frei et al. 1998).

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

187

IV. BREEDING USING CYTOPLASMIC FACTORS Cytoplasmic factors are obviously important in qualitative traits such as CMS or disease resistance/susceptibility. However, the contribution of cytoplasmic factors to quantitative agronomic traits is probably of minor significance. Nevertheless, in order to further enhance the performance of cultivars, plant breeders can exploit such minor effects. They should be considered in order (1) to characterize variability, (2) to use or generate variability by recombination, and (3) to select favorable genotypes. The following section emphasizes the degree of integration of cytoplasm research into breeding crop plants. A. Characterization of Plant Material There are several approaches that can be used to understand the role variability in cytoplasmic genomes plays in phylogenetic studies. Misleading results occur when the investigated species is crossed with an unidentified pollen parent, even if only evolutionary pathways derived from the seed parent are analyzed. The molecular basis of the genetic variability of cytoplasms has revealed greater variability of mitochondrial genomes compared to chloroplast genomes, and an interdependence between both organelles (Newton 1988; Samitou-Laparde et al. 1991). In all cases, the specific effects of the independent organelles and nuclei are difficult to estimate due to their interactions. Differentiation of cytoplasmic types, in addition to male sterility influences, is a prerequisite for efficient breeding work. Development of molecular marker techniques based on DNA polymorphisms in the organelle genomes may facilitate parent selection. However, it should always be taken into consideration that the mt-genome is very dynamic and consequently such polymorphisms may not be stable. If a breeding program is initiated with a neglected crop, such as amaranth, reproducible classification of the starting material is essential. This classification should not be restricted to the nuclear genome but include the plasmone, since the additional knowledge will help to optimize breeding strategies. 1. Potato. The origin of the cultivated tetraploid potato was traced via plasmones, predominantly ptDNA, information as well as by genetic relationships among cultivars (Hosaka and Hanneman 1988; Hosaka 1995). Lössl et al. (1999) characterized different mitochondrial types in 100 dihaploid and 144 tetraploid potato clones (Fig. 4.1). Based on reproducible RFLP patterns using 11 homologous mtDNA probes, five distinct classes (α, β, γ, δ, and ε) of mtDNA were separated. A dendrogram showing the relationship of wild potato species and primitive cultivars based on mitochondria descent is shown in Fig. 4.3. This

188

U. FREI, E. PEIRETTI, AND G. WENZEL

Rescaled Distance Cluster Combine CASE

0

5

10

15

20

25

Beta tuberosum berthaulthii poloytrichon verrucosum Gamma

gourlayi traijense stoloniferum Alpha demissum acaule sparsipilum

goniocalix andigena phureja maglia multidissectum curtilobum ajanhuiri chauca Epsilon megistacrolobum

capsibaccatum circaefolium brevidens gibberulosum spegazinii vernei microdontum kutzianum Delta chacoense bulbocastanum Fig. 4.3. Dendrogram with genetic distances between different Solanum mtDNAs (Lössl et al. 1999).

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

189

type of information can be used to select distantly related parents for sexual and somatic combinations. 2. Maize. Molecular probes are increasingly used for characterization of inbred lines of maize. Peiretti (2003) evaluated the plasmone of 30 dent and 32 flint lines using 23 mt-probes, and found clear polymorphisms. A genetic distance analysis revealed three clusters. The majority of lines of one cluster were flint types, and the majority of lines of another were dents. Only a smaller, third cluster, more distant from the other two, contained lines of both. RFLP analysis of nuclear genomes using the same genotypes gave a corresponding classification (Messmer et al. 1992, 1993). Since mt-probes result in simpler restriction patterns, they offer an easier system for the characterization of inbred line diversity for applied use. 3. Rice. Cytoplasmic variability of cultivated rice species, Oryza sativa and O. glaberrima (Ishii et al. 1993) indicates high conservation of pt- and mt-genomes. Lin (1992) performed pedigree analyses to quantify the ancestral contributions to 27 rice cultivars released at IRRI and to estimate the coefficients of parentage among them. One ancestor was the ultimate cytoplasmic parent for 22 of the 27 cultivars. These investigations are particularly helpful in current hybrid breeding programs. 4. Wheat. In wheat and its related genus Aegilops, genetic diversity of cytoplasmic genomes among species has been extensively studied by two different means. One was to compare, by repeated backcrosses, alloplasmic lines with common wheat nuclei and different cytoplasms introduced from numerous Triticum and Aegilops species. The second was restriction endonuclease analyses of mt and ptDNAs (Vedel et al. 1978; Terachi et al. 1984). Based on mitochondrial RFLP analyses with cloned genes as probes, Terachi and Tsunewaki (1992) generated a phylogenetic tree of the mitochondrial genomes of Triticum and Aegilops that differentiated into three major groups: (1) Triticum monococcum, (2) Aegilops speltoides and the polyploid wheats, and (3) all the other species. Today, it is possible to classify the cytoplasm from Triticum and Aegilops as belonging to 10 major plasma types, of which six are further subdivided into minor plasma types. Each plasma type can be characterized by the effects on common wheat (Table 4.2). Investigations of the plasmone of more than 100 genotypes of Triticum, Aegilops, and Agropyron using RFLP mt-probes combined with different restriction enzymes showed in 20% of the combinations

190

U. FREI, E. PEIRETTI, AND G. WENZEL

Table 4.2. Plasmone types of wheat modified from Mukai (1982), Panayotov (1983), Terachi and Tsunewaki (1992), Wang et al. (2000). Group origin

Type

Characteristics in the alloplasmic lines

1. T. monococcum

A

2. Ae. caudata, Ae. triuncialis

C1 C2 Cu1

No winter hardiness, growth inhibition, variegation in winter, CMS CMS, pistilloid, anther malformation

3. Ae. umbellulata, Ae. binuncialis Ae. triaristata Ae. columnaris 4. Ae. tauschii, Ae. cylindrica Ae. ventricosa, Ae juvenalis Ae. crassa, Ae vavilovii 5. Ae. comosa Ae. heldreichii Ae. ovata 6. Ae. uniaristata 7. Ae. mutica 8. T. speltoides, T. turgidum, T. aestivum Ae. bicornis, Ae. kotschyi, Ae. variabilis, Ae. nudiglumis 9. Ae. aucheri, T. timopheevi, T. zhukovskyi 10. Ae. longissima Ae. sharonensis

Growth depression, late heading variegation, partial CMS

Cu2 D1

Anther malformation No remarkable effects

D2 M1 M2 M3 Mu Mt B/S1

Pistilloid Late heading, CMS Late heading, CMS, growth depression Late heading, CMS, tall plant height Growth depression Late heading No remarkable effects

S2

Slightly reduced fertility

G

CMS, anther malformation

S11 S12

Slight reduction of plant vigor Growth depression, partial CMS

polymorphisms indicative of a substantial variability of the mitochondria. A phylogenetic analysis of the cytoplasms on the basis of such genetic distances resulted in a dendrogram differentiating eight clear plasmone types for Triticum and one type for Agropyron trichophorum. Lines clustering closely were homogeneous in the genetic structure of the plasmone and could be assigned to the groups given in Table 4.2. For groups 1, 3, 4, 6, 8, and 9, a precise ordering was possible. Two other groups expressed lower homogeneity but could still be differentiated. In one group the majority of the plasmone corresponded to type S2 but contained in addition S11, C2 Cu2, and D1; and one group showed similar amounts of M3, Mt Mu and S12, implying mt-genome recombination. This may be caused by exchange of nuclear and mitochondrial genetic information; so it is essential to document where the gene is encoded before assaying its expression in different nuclear backgrounds.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

191

5. Amaranth. Although rather neglected by plant breeders, amaranth has a good potential as a new crop (Peiretti and Gesumaria 1998). Due to the high protein quality of its seeds and flour, it could be valuable in baking bread (Schmidhuber and Zeller 1998). In an analysis of genetic diversity, the plasmone was analyzed using mt-probes in combination with three restriction enzymes in about 50 genotypes of 10 different species of Amaranthus. One third of the combinations revealed polymorphisms. The results clearly demonstrate wide genetic variability of the cytoplasm, which could be classified distinctly into five groups in agreement with the phylogenetic structure. In addition, two small clusters were found, consisting of two genotypes each. Phenotypic resemblances within the respective plasmone groups suggests that recombination did not result in new types. On this basis, Peiretti (2003) formulated five cytoplasmic types: (1) A. cruentus plasma; (2) A. caudatus plasma; (3) A. mantegazzianus with slight addition of A. caudatus plasma; (4) A. hipochondriacus with slight addition of A. hybridus and A. caudatus plasmone similar to A. hipochondriacus (also in the phenotype), producing one small subgroup with A. tricolor and A. quitensis plasma; and (5) A. retroflexus with addition of A. cruentus, A. hypochondriatus, producing one small subgroup with A. dubius plasmone. Based on this classification of the plasmones, the present systematic order of the Amaranthus species might require re-evaluation. Analysis of the plasmone may detect errors in taxonomic classification and can provide information in the direction of subspecies formation. B. Creating New Variability When natural variability is insufficient for direct selection of a desired plasmone, new variability needs to be created via mixing or recombination processes. The classical approaches combine the many genes of two parents that use meiotic recombination. This process works by chance. More precise means are available by gene transfer techniques in which at least the gene structure and function are known. 1. Undirected Processes. Different kinds of variability need to be considered: (1) the generation of new nuclear cytoplasmic combinations, (2) new combinations between unaltered chloroplasts and mitochondria, or (3) the creation of new mitochondrial or plastid genotypes. New Cytoplasmic Nuclear Combinations via Alloplasmic Crossing. Alloplasmic genotypes can be developed via a repeated backcross process,

192

U. FREI, E. PEIRETTI, AND G. WENZEL

always using the seed parent as donor for the cytoplasm and the pistillate parent as the donor for the nuclear information (Tsunewaki et al. 1996). Rare transfer of proplastids and/or mitochondria via the pollen tube may result in mixed zygotes when the parents have different cytoplasms. Since each cell has hundreds of mitochondria, and only a few paternal promitochondria will be transferred, the stabilization of a new cytoplasm takes normally at least six backcrossing generations (Linnert 1997). Alloplasmic effects were predominantly investigated in allopolyploids such as Brassica napus (Hallden et al. 1993), Triticum aestivum (Tsunewaki et al. 1996), or triticale (Geiger et al. 1993). For Triticum and its relatives, alloplasmic lines with substituted cytoplasms could be bred combining good starch (falling number and α-amylase activity) and protein quality (gluten ductility and swelling) with resistance against biotic and abiotic stresses (Busch and Rauch 1994; Öttler and Mares 1994; Ekiz et al. 1998). Wu et al. (1998) found a new range of nucleoplasmic hybrids having Aegilops crassa cytoplasm to be the best under practical field growing conditions. Wang et al. (2000) characterized 46 alloplasmic lines of common wheat carrying plasmones of 33 Triticum and Aegilops species. Using mtDNA probes, they described high mtDNA variability. Using a minimum set of three mtDNA probeenzyme combinations, they identified the plasmone type and screening of new plasmone types became possible. The cytoplasms of Triticum boeoticum, Aegilops umbellulata, Ae. triuncialis, Ae. biuncialis, Ae. columnaris, and Ae. triaristata 6x induce variegation in many common wheats (Mukai 1982). In general, alien cytoplasmic effects are more deleterious in tetraploid Triticum turgidum var. durum than in hexaploid T. aestivum (Sasakuma and Maan 1978). In the allopolyploid Brassica napus, the action of the cytoplasm has became of enormous interest since hybrid cultivars have been successful (in Germany since 1996). By means of restriction enzymes, pt and mtDNAs were characterized for alloplasmic effects (Hallden et al. 1993; Landgren et al. 1996). In B. rapa, Matsuzawa et al. (1999) found CMS in B. napus carrying the Eruca sativa plasmone. Since hybrid breeding is of increasing importance in rice (Virmani and Sharma 1993), the search for CMS has become essential. Cytoplasmic effects of alloplasmic lines varied significantly with the nuclear background. Due to such variation, it should be possible to minimize the present negative cytoplasmic effects so as to obtain improved hybrids by selection of the desired CMS lines (Wang et al. 1998; Li and Yuan 2000). Also, in hybrid breeding of sugar beet, the sexual reproductive system of the gynodioecious and hermaphroditic Beta vulgaris ssp. maritima on cytoplasmic differentiation was evaluated in populations of wild beet

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

193

using chloroplast DNA polymorphism and applied for parent selection (Samitou-Laparde et al. 1991; Forcioli et al. 1998). For hybrid production of Daucus, alloplasmic cultivars are intensively used for the utilization of CMS (Steinborn et al. 1995; Nothnagel et al. 2000). The authors evaluated a collection of alloplasmic carrot lines and found three CMS lines among 100 lines. The genetic analysis revealed that a nuclear-cytoplasmic interaction determined male sterility in these three sources. A clear separation of lines was possible by mtDNA markers. In a similar way, Vivek et al. (1999) proved the maternal inheritance of ptDNA in carrot by markers. New Cytoplasmic-nuclear Combinations in Cybrids. Interspecific and intergeneric protoplast fusions were developed to introduce novel germplasm from sexually incompatible species in order to broaden the genetic base of parents. Besides nuclear incompatibility, Gleba and Sytnik (1984) described the phenomenon of nuclear-cytoplasmic incompatibility. The multifusion of potato (Frei 1994) or wider plant groups can help to understand these phenomena (Binding et al. 1992). In all fusion events, however, organelle segregation and recombination takes place. Fusion provides the opportunity to optimize not only the genetic composition of the nuclear genome but also that of the cytoplasm and the interaction of both. While maternal inheritance operates in sexual recombination, this principle is switched off in somatic cell fusion where both parental cells contribute genomic and cytoplasmic information. The somatic fusion approach for cytoplasmic improvement was attempted for applied breeding. Trials were initiated to exchange cytoplasms completely, leading to alloplasmic lines. Thus, under this aspect, asymmetric hybrids with unequal contributions of the cytoplasmic and nuclear genetic material of each parent were of greater interest than symmetric fusions. Ideally, the asymmetry should go so far that one parent delivered the nucleus only and the other one the cytoplasm only, forming a cybrid. The problem of combining two nuclear genomes with the probability to incorporate undesired genes and the consequence of doubling the ploidy level could be overcome. In the simplest procedure, protoplasts of a donor species (normally a distantly related genotype) are irradiated, destroying the nuclear DNA before fusion with the recipient protoplast (normally a cultivar) in which the metabolism of the cytoplasm is blocked by rhodamin-6-G or iodoacetamide (Sidorov et al. 2000). As a result, the nuclear genome of the cultivar is combined with the plastome of, for example, a distantly related species. Since under these conditions nuclear-cytoplasmic incompatibilities appear, de novo CMS systems can be created. Cybridization can replace the cytoplasm

194

U. FREI, E. PEIRETTI, AND G. WENZEL

of a cultivar in one step, saving lengthy backcrossing programs classically applied for the introduction of alien cytoplasms (Melchers et al. 1992). The most advanced system occurs in the Brassicaceae, in which new vegetables (Earle and Dickson 1995) and rape seed hybrids have been produced. Cybridization provided a way to replace the Raphanus chloroplasts with Brassica ones, while retaining the CMS associated with the Raphanus mitochondria (Pelletier et al. 1988; Gourret et al. 1992). This CMS source was patented as the INRA/Ogura system. Sometimes, along with the desired CMS character undesired chlorosis is also induced. By a subsequent somatic fusion, such mistakes could be corrected (Arumugam et al. 2000). Several such hybrids of rapeseed have reached the market. The cultivar ‘Life’ was still registered in Germany in the year 2001, but due to problems of restoration, it needs a line producing pollen for fertilization. In the meantime, restored hybrids have been developed via classical breeding, resulting in the MSL system, replacing hybrid cultivars with the more complicated INRA/Ogura system (Manthey 2001). Several other cybrids were produced, particularly in tobacco and potato, but since they are not yet advanced enough to make predictions on their practical value, the reader is referred to the review of Johnson and Veilleux (2001). An approach to tackle defined nuclear-cytoplasmic combinations is offered by the microfusion of plasma-free caryoplasts with enucleated cytoplasts (Spangenberg et al. 1991). This process is, however, so laborious and difficult to reproduce that it has not reached the level of practical application. New Cytoplasmic Combinations by Symmetric Fusions. Most of the applied work using asymmetric cell fusion and cybridization has used the donor-recipient principle described above. The selection of desired types of organelles before cell fusion has a number of disadvantages and the effect of metabolic inhibitors is unclear. These inhibitors act differently in different species and probably also in different genotypes. Furthermore, segregation of the chloroplasts cannot be influenced (Walters and Earle 1993). Molecular analysis of somatic hybrids, especially of their cytoplasmic composition, has been performed for various organelles in Brassica. In plants regenerated after fusion between B. oleracea var. italica (broccoli) containing the petaloid B. nigra type of CMS and an atrazine-resistant biotype of B. campestris var. oleifera, plants were regenerated that were CMS and atrazine resistant up to the level of 25 µM (Christey et al. 1991). Molecular analysis showed that they contain chloroplasts from the atrazine-resistant B. campestris parent and mitochondria from the B. nigra parent. In that case, no recombination or rearrangement of the mt-genomes in the fusion products was detected.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

195

Symmetric fusion may result in CMS, as demonstrated by Bhattacharjee et al. (1999) for rice. Symmetric fusion with a subsequent selection using markers of pt and mtDNA is promising for de novo construction of plasmones by recombination. In potato, Lössl et al. (1994) characterized by RFLP analysis the nuclear and cytoplasmic composition of five different fusion combinations consisting of up to 50 hybrid regenerants each. The same hybrids were evaluated in repeated field experiments for their yield structure and starch content. Complete chloroplast segregation was found with a 1:1 ratio in all but one hybrid. This complete segregation of plastids may be an effect of the recently detected connections between plastids (stromules), highly dynamic structures with continuously and rapidly changing shape. Stromules have been shown to interconnect plastids and permit exchange of proteins between plastids (Gray et al. 2001). The mtDNA shows a much more complex picture; new RFLP bands were indicative of rearrangements in the mt-genome. The assignment of hybrids to the different chloroplast types in potato did not affect yield or starch content. However, mt-types could be distinguished. Their analysis with different probes homologous to coding regions revealed a relationship between the homogeneity of the mtgenome, the yield level, and the starch content of the potato. Further evaluations of the correlation between the cytoplasmic types showed that an interaction existed between starch content and mt-pt combinations. In general, the highest field performances were associated with those cytoplasmic configurations that appeared at a high frequency after completion of the segregation process in vitro. It is assumed that an advantage of clones with an optimized organelle composition exists by such a selection and that they regenerate in vitro preferentially (Lössl et al. 2000). Somatic hybridization is a standard breeding technique in potato for combining nuclear genomes (Hofferbert and Wenzel 1997). Since potato is vegetatively propagated, no meiosis will recombine the heterozygous hybrid structure. Somatic hybridization offers a chance for recombining mt-genomes, resulting in new mtDNA sequences that might be superior for a desired breeding aim. In sexually propagated crops, somatic hybridization is of minor importance, since during meiosis preceding the sexual process, the nicely combined nuclear alleles will recombine and segregate. If the construction of really combined and not just mixed mt or ptDNA can be maintained via meiosis, somatic hybrids are a much more efficient way to use in organelle genetics than the lengthy way of the production of alloplasmic lines via backcrossing. 2. Directed Alterations for Plasmone Transformations. While whole plasmones can be mixed or recombined, plasmone transformation

196

U. FREI, E. PEIRETTI, AND G. WENZEL

allows a more specific technique: adding one or a few genes to a plastome or chondriome, resulting in transgene organelles. Due to their better genetic stability, plastids are the major object in transferring organelle genes. Four discoveries have established them as a feasible platform for genetic modification of plants: (1) the stable transformation of cloned DNA into the Chlamydomonas chloroplasts (Boynton et al. 1988), (2) the successful extension of these approaches to higher plants (Svab et al. 1990), (3) the technical developments in plastid transformation (Heifetz 2000), and (4) advances in the understanding of the rules of plastid gene expression (Maliga 2001). The transformation of pt-genomes (Table 4.3) has a number of advantages compared to the nuclear genome. Plastid transformation usually results in a high level of transgene expression partly attributed to the presence of 50–60 chloroplasts per leaf cell and 60–100 chromosomes per plastid, compared to the usual two copies of each of many chromosomes in the nucleus. The chloroplasts of flowering plants encode 120–150 genes, many of which are grouped in operons. In contrast, the nuclear chromosome contains thousands of genes arranged separately and transcribed individually. As the vast majority of interesting agronomic traits are multigenic, their introduction into the nuclear genome has to be done one gene at a time with unpredictable expression level. The introduction of even short biosynthetic chains is laborious and time consuming. It required a 7-year effort by the group of Ingo Potrykus to induce 3 genes into rice to produce high provitamin A in ‘Golden Rice’ (Ye et al. 2000). The organization of plastid genes in operons allows the introduction of blocks of foreign genes in a single operon under the control of a single promotor. Even the site of insertion can be selected, and the integration via the homologous recombination process facilitates Table 4.3. Summary of plant species where a stable chloroplast transformation has been achieved.

Species

Transformation method

Arabidopsis thalinana Nicotiana plumbaginifolia Potato Tobacco

Biolistic PEG Biolistic Biolistic, protoplast

Tomato Rice

Protoplast Biolistic

References Sikdar et al. 1998 O’Neill et al. 1993 Sidorov et al. 2000 Svab et al. 1990; Golds et al. 1993; McBride et al. 1995; Koop et al. 1996; Guda et al. 2000 Ruf et al. 2001 Khan and Maliga 1999

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

197

targeted gene replacement and precise transgene control. And at the least, because chloroplasts of angiosperms are generally transmitted only by the maternal plant, trans-plastomic plants will not unintentionally release a trait introduced by genetic engineering to wild relatives. The principal methods to introduce DNA into chloroplasts are biolistic bombardment (Svab et al. 1990) and the use of polyethylene glycol (O’Neill et al. 1993; Koop et al. 1996). Alternatively, Knoblauch et al. (1999) have shown that DNA is expressed after its direct injection into individual chloroplasts in photosynthetic leaf cells of tobacco. The final integration of foreign genes into higher plant plastid genomes probably occurs through homologous recombination. Currently the marker genes aadA, codA, and nptII, and the reporter gene ui9dA are used for selection of transformed plastids (Serino and Maliga 1997). In chloroplasts, due to the presence of an Escherichia coli-like system for homologous recombination, transformation is a promising system (Bock and Hagemann 2000). An advantage of transplastomic plants is the accumulation of high levels of alien proteins in the cell organelles, sheltered from proteases of the cytoplasm or enclosed, if biosynthetic products would be harmful to the cell in higher amounts. The introduction of genes coding for insecticidal proteins are currently being investigated. Crops expressing the Bacillus thuringiensis (Bt) toxin under the control of the nucleus generally produce it only in sub-optimal amounts. McBride et al. (1995) introduced the Bt gene encoding the protoxin into the chloroplast chromosomes of tobacco. In the resulting plants, this protoxin constituted 2–3% of the total soluble leaf protein. Further efforts have focused on the engineering of herbicide-resistant transplastomic plants by EPSPS transfer (Daniell et al. 1998; Lee et al. 2000). Although transformation of plastids has been achieved for several plant species, homoplasmic fertile plants have been produced only in tobacco and tomato as yet (Ruf et al. 2001). There is great hope regarding the exploitation of the abundant expression capacity of the plastome for plastid mediated molecular farming, leading to the biofabrication of biopolymers and pharmaceutical proteins. A first step can be seen in the work of Staub et al. (2000), engineering tobacco plants able to produce massive amounts of the therapeutic protein, human somatropin, or of insecticidal crystals by over-expressing a Bt operon (De Cosa et al. 2001). Guda et al. (2000) obtained a 100-fold higher yield compared to the transgenes of bioelastic protein-based polymers with transplastomic tobacco plants carrying an artificial gene encoding a repetitive sequence observed in all sequenced mammalian elastin proteins. Mason et al. (2002) have already discussed the usefulness of edible plant vaccines for their application

198

U. FREI, E. PEIRETTI, AND G. WENZEL

in prophylactic and therapeutic molecular medicine. Although there might be still a long way to go, two major modifications of plant metabolism of high importance for agriculture will get renewed consideration: enabling plants to fix N2 and improving the photosynthetic CO2 fixation (Bogard 2000). C. Selection The procedures for directed transformation of ptDNA are not yet precise enough to expect reliable gene expression in any transformed plant. Therefore, powerful selection systems are needed allowing detection of the rather rare event. Molecular techniques allow sensitive differentiation between cytoplasm types. Alcala et al. (1997) succeeded in differentiating the plastomes of CMS onion cultivars from fertile lines using genetic bit analysis. Nevertheless, apart from the information about phylogeny and genetic variability, the molecular characterization of cytoplasms is only of value for breeding approaches if phenotypic traits can be correlated with the markers. When the modifications of the plastome are undirected, as in cell fusion experiments, an additional drawback is the lack of correlation between types of altered cytoplasms and phenotypic changes caused by them. Molecular selection steps may also help to revive the use of somaclonal and protoclonal variation. When it becomes possible to screen for specific mutations that are responsible for a specific step in the biochemical pathway, predictability may replace “luck” in the selection process. 1. Correlation of Phenotype and Cytoplasm. In most cases, the influence of the plasmone on agronomic traits is not distinguished from biparentally inherited traits. To judge the importance of the plasmone, Lössl et al. (2000) combined distinct parental cytoplasms in symmetric tetraploid somatic hybrids in potato. This made it possible, in the presence of nearly isogenic nuclear genomes, to estimate the contribution of mt and ptDNA (Fig. 4.1). The starch content was selected as a phenotypic character. Analysis of mt-pt configurations in a broad gene pool of German potato cultivars, in a reciprocal dihaploid population of 180 clones, in dihaploid fusion parents and in their somatic hybrids emphasized the effects of the plasmone. The results showed that in starch content, the wild type mt-pt cytoplasms have a significant advantage to other combinations of the cytoplasm. An interaction between starch content and different mt-pt combinations was found in hybrids. In general, the highest starch content and yield measured in the field was associated with

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

199

cytoplasmic configurations that appeared in a high frequency within a population after completion of the segregation process. This fact is explained by a selection advantage of clones with optimized organelle segregation during the in vitro phase. 2. Somatic Recombination. Chloroplast segregation has been reported for somatic hybrids by Maliga and Menczel (1986). Chloroplasts normally segregated after a few cell cycles. In potato and Brassica, however, different ratios of parental plastids were observed, perhaps due to different organelle replication rates (Glimelius et al. 1990; Sundberg and Glimelius 1991). When such segregations of hybrids were compared with the field data, no correlation was detectable between the type of plastome and yield characteristics. For potato mitochondria, Xu et al. (1993) reported about 75% recombination of the mtDNA. Such recombinations occur during normal protoclonal variation without any fusion event (Kemble and Shepard 1984). Similar results have been found in the genus Brassica, where Temple et al. (1992) located regeneration breakpoints in the mtDNA, and by Walters and Earle (1993), who found mitochondrial recombination in one third of somatic Brassica hybrids. Lössl et al. (2000) found a correlation between the mitochondrial composition and yield parameters of somatic potato hybrids. Different levels of starch content and tuber yield could be ascribed to different mt-genotypes. Especially for yield, mitochondria from one parent were superior to mitochondria of the other parent. However, most recombinations or new mixtures (new heteroplasmic forms) of mitochondria resulted in a decrease in vigor. It is suggested that plant vigor measured by tuber yield correlates positively with the homogeneity of the mtgenome. The fact that, in sexual hybridizations, only the maternal plasmone is inherited by the next generation suggests a very limited mt-recombination. Evolution resulted in uniform mitochondria with an optimal function for the fitness of the genotype. There is, however, a chance that, by inducing new heteroplasmic mixtures or even mitochondrial recombination, mitochondria can be combined or produced that fit better man’s breeding aims. If so, then protoplast fusion may provide a means for the construction of improved mt-genomes in other crops. As an explanation for the process of recombination among mtDNA resulting in new mtDNA arrangements, Lössl et al. (1999) proposed the following model for potato (Fig. 4.4). In view of the multiple copies and the variability of the wild type mt-genome with master, super, and subcircles coexisting, both fusion parents may contribute to the new mtgenome different mixtures of mtDNA. The sub-circles are the result of

200

U. FREI, E. PEIRETTI, AND G. WENZEL

α cob

β cob

cox I cob rps14 rpl 5

α

cox I

(+)

cob rps14 rpl 5

β

Parents

α cob cox I cob rps14 rpl 5

β

R1 β cob cox I

cob rps14 rpl 5

α

R2

Fig. 4.4. Model of parental mt-genomes and recombinant hybrid mt-genomes R1 and R2. The mt-genome is shown as consisting of a major (master circle) and minor component (sub-circle) for each plant. After protoplast fusion, parts of mt-genome α are replaced by sub-genomes from chondriome type β (Lössl et al. 1999).

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

201

homologous recombination between sequence repeats (Newton 1988). At least during the reorganization of the mt-genome after the fusion process, the sub-genomes seem to be excluded from the master circle so that they can be exchanged independently from the mt main chromosome. The various types of mtDNA within a fusion hybrid population are combined by chance in the somatic hybrids (Fig. 4.4). Such a form of organization could explain the mixtures between different mt-loci found in some hybrids. Somatic fusion should permit the formation of new mixtures of mitochondria that might be stably inherited. The evolution of the mitochondria up to now was little influenced by plant breeding. Now, recombinations after cell fusion open up an opportunity to produce de novo many new mt-genomes, some of which might provide superior traits.

V. CONCLUSION Information on genetic variability and genetic distance is most useful for optimal selection of parents, especially for hybrid breeding. Thus, increasingly molecular markers based on genome polymorphisms are mapped on the chromosomes and used for marker aided selection (MAS) (Wenzel et al. 2001). This nuclear genomic picture should be backed by an understanding of plasmone variability. Different reports on the influence of the plasmone emphasize the importance of the non-nuclear DNA for breeding processes, in particular for hybrid production using CMS, but also, as shown for potato, for yield and/or quality. In contrast to the complex nuclear genome, mt- or pt-probes may allow a simpler but faster classification due to less complex polymorphisms. Most of this research is underway in the major cereals (wheat, rice, and maize) but also in vegetables such as carrot. If their usefulness can be demonstrated, the transfer of these techniques will pay off in speeding up the work with minor crops such as legumes, or neglected crops such as amaranth. The approach to obtain new nuclear-cytoplasmic combinations through sexual combination is reliable but time consuming. The biotechnological (asexual) methods of somatic hybridization and transformation offer faster but laborious alternatives. However, both procedures are not yet efficient enough to be used for plant breeding without proper selection of the superior recombinants. Programmed combination breeding of the nuclear genomes could be successfully speeded up by the extended use of MAS, thanks to extensive efforts to correlate phenotypic and molecular data. Consequently,

202

U. FREI, E. PEIRETTI, AND G. WENZEL

combination and/or recombination should also be a powerful tool in breeding for improved plasmones as soon as the newly obtained cytoplasm types can be correlated with phenotypic variation. This can be achieved, for example, in hybrid regenerants after somatic hybridization. As their nuclear genomes reveal only minor variability, as proven by RFLP analysis, the new phenotypic variation can be ascribed to the cytoplasm and predominantly to the chondriome. This variation, a result of plasmone recombination, allows the characterization and selection of improved mt-genomes. Breeding work using the plasmone thus enhances the opportunity to select improved cultivars. The somatic cell fusion approach will be useful in breeding vegetatively propagated crops. In particular, symmetric protoplast fusion is an elegant method to create new cytoplasms. Cell fusion will also be applicable in sexually propagated crops for the formation of improved mtgenomes. Modifying ptDNA by somatic cell fusion is more difficult since their number per cell is lower and they are separated during the first cell divisions. Chloroplasts have, however, become an object for transformation since under the aspect of safeguarding transgenes, plants carrying the transformation in the pt-genome minimize the ecological risk of vertical and horizontal gene transfer (Bergelson et al. 1998). Integration of foreign genes into the plastid genome appears to occur via homologous recombination, facilitating the targeting of foreign genes to a specific location in the ptDNA (Bock and Hagemann 2000). Consequently, all transformants from the same construct should have the same expression level. Revealing structure-function relationships for pt and mtDNA results in increased understanding of the action of cytoplasmic multiprotein complexes. Such basic knowledge is creating the hope of exploiting the abundant expression capacity of the plastome for plastid-mediated biosynthesis. Molecular farming in cell organelles will lead to the biofabrication of biopolymers and pharmaceutical proteins. In summary, work on the genetic basis of the plasmone is still difficult, but it could have enormous benefits for genetic improvement of feed, food, and non-food crops. The importance of the mtDNA for CMS in hybrid breeding highlights the potential of non-nuclear genes. In addition, there are always new surprises coming from basic research such as the detection of stromules permitting exchange of proteins between plastids (Gray et al. 2001). Chondriomes, perhaps optimized by the use of stromules or transgenic plastomes, may become biotechnological tools with sufficient reproducibility and reliability to achieve known and new breeding goals.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

203

LITERATURE CITED Abdallah, F., F. Salamini, and D. Leister. 2000. A prediction of the size and evolutionary origin of the proteome of chloroplasts of Arabidopsis. Trends Plant Sci. 5:141–142. Adamski, T., J. Chelkowski, Z. Kaczmarek, M. Surma, and H. Wisniewska. 1997. Evaluation of susceptibility of auto- and alloplasmic barley DH lines to Fusarium seedling blight. Euphytica 93:169–172. Alcala, J., J. J. Giovannoni, L. M. Pike, and A. S. Reddy. 1997. Application of genetic bit analysis for allelic selection in plant breeding. Mol. Breed. 3:495–502. Altman, A. (ed.). 1998. Agricultural biotechnology. Dekker, New York, Basel, Hong Kong. Arumugam, N., A. Mukhopachyay, V. Gupta, Y. S. Sodhi, J. K. Verma, D. Pental, and A. K. Pradhan. 2000. Somatic cell hybridization of oxy CMS Brassica juncea (AABB) with B. oleracea (CC) for correction of chlorosis and transfer of novel organelle combinations to allotetraploid brassicas. Theor. Appl. Genet. 100:1043–1049. Atanassov, I. I., S. A. Atanassova, A. I. Dragoeva, and A. I. Atanassov. 1998. A new CMS source in Nicotiana developed via somatic hybridisation between N. tabacum and N. alata. Theor. Appl. Genet. 97:982–985. Balk, J., and C. J. Leaver. 2001. The PET1-CMS mitochondrial mutation in sunflower is associated with premature programmed cell death and cytochrome release. Plant Cell 13:1803–1818. Bao, J. S., and Y. W. Xia. 1999. Genetic control of paste viscosity characteristics in indica rice (Oryza sativa L.). Theor. Appl. Genet. 98:1120–1124. Barkan A., and M. Goldschmidt-Clermont. 2000. Participation of nuclear genes in chloroplast gene expression. Biochimie 82:559–572. Baur, E. 1909. Das Wesen und die Erblichkeitsverhältnisse der “Varietates albomarginatae” hort. von Pelargonium zonale. Z. Ind. Abst. U. Vererbungsl. 1:330–351. Bebeli, P. J. 1995. Cytoplasmic effects on tissue culture response in wheat. J. Genet. Breed. 49:201–208. Bergelson, J., C. B. Purrington, and G. Wichmann. 1998. Promiscuity in transgenic plants. Nature 395:25. Bhattacharjee, B., A. P. Sane, and H. S. Gupta. 1999. Transfer of wild abortive cytoplasmic male sterility through protoplast fusion in rice. Mol. Breed. 5:319–327. Binding, H., L. D. Won, L. H. Qing, G. Buch, G. Tank, J. Rudnick, and W. G. Rong. 1992. Interspecific interactions in the regeneration from protoplasts. p. 27–34. In: K. Oono et al. (eds.), Plant tissue culture and gene manipulation for breeding and formation of phytochemicals. NIAR, Tsukuba. Bock, R. 2001. RNA editing in plant mitochondria and chloroplasts. In: B. L. Bass (ed.), RNA editing. Frontiers in molecular biology. 38–60. Oxford Univ. Press, Oxford. Bock, R., and R. Hagemann. 2000. Extranuclear inheritance: plastid genetics: manipulation of plastid genomes and biotechnological applications. Prog. Bot. 61:76–90. Bock, R., and M. Hippler. 2002. Extranuclear inheritance: Functional genomics in chloroplasts. Prog. Bot. 63:106–131. Bogard, L. 2000. Engineering chloroplasts: an alternative site for foreign genes, proteins, reactions and products. TIBTECH 18:257–263. Bolandi, A. R., M. Branchard, G. Alibert, H. Serieys, and A. Sarrafi. 1999. Cytoplasmicnuclear interaction and medium effect for protoplast culture in sunflower. Plant Cell Tissue Organ Culture 57:189–193. Boynton, J. E., N. W. Gilham, and E. H. Harris. 1988. Chloroplast transformation in Clamydomonas with high velocity microprojectiles. Science 240:1534–1538.

204

U. FREI, E. PEIRETTI, AND G. WENZEL

Breiman, A. 1987. Mitochondrial DNA diversity in the genera of Triticum and Aegilops revealed by Southern-blot hybridisation. Theor. Appl. Genet. 73:563–570. Bruik, R., and S. P. Mayfield. 1999. Light-activated translation of chloroplast mRNAs. Trends Plant Sci. 4:190–194. Bryan, G. J., J. McNicoll, G. Rfamssay, and R. C. Meyer. 1999. Polymorphic simple sequence repeat markers in chloroplast genomes of Solanaceous plants. Theor. Appl. Genet. 99:859–867. Busch, R. H., and T. L. Rauch. 1994. Alien cytoplasmic effects on agronomic performance of two high yielding semidwarf spring wheat cultivars. Crop Sci. 34:389–391. Chen, X., and R. F. Line. 1995. Gene action in wheat cultivars for durable, hightemperature, adult-plant resistance and interaction with race-specific, seedling resistance to Puccinia striiformis. Genetics 85:567–572. Christey, M. C., C. A. Makaroff, and E. D. Earle. 1991. Atrazine-resistant cytoplasmic male-sterile-nigra broccoli obtained by protoplast fusion between cytoplasmic malesterile Brassica oleracea and atrazine-resistant Brassica campestris. Theor. Appl. Genet. 83:201–208. Comai, L. 2000. Genetic and epigenetic interactions in allopoliploid plants. Plant Mol. Biol. 43:387–399. Daniell, H. 2002. Engineering the chloroplast genome to confer stress tolerance and production of pharmaceutical proteins. p. 427–451. In: S. M. Jain et al. (eds.), Molecular techniques in crop improvement. Kluwer Acad. Publ., Dordrecht. Daniell, H., R. Datta, S. Varma, S. Fray, and S. B. Lee. 1998. Containment of herbicide resistance through genetic engineering of the chloroplast genome. Nat. Biotechnol. 16: 345–348. De Cosa, B., W. Moar, S. B. Lee, M. Miller, and H. Daniell. 2001. Overexpression of the Bt cry2Aa2 operon in chloroplasts leads to formation of insecticidal crystals. Nat. Biotechnol. 19:71–74. Dhitaphichit, P., P. Jones, and E. M. Keane. 1989. Nuclear and cytoplasmic gene control of resistance to loose smut (Ustilago titici) in wheat (Triticum aestivum L.). Theor. Appl. Genet. 78:897–903. Dragoeva, A., I. Atanassov, and A. Atanassov. 2001. CMS due to tapetal failure in cybrids between Nicotiana tabacum and Petunia hybrida. Plant Cell Tissue Organ Culture 55:67–70. Ducos, E., P. Touzet, P. Saumitou-Laprade, P. Vernet, and J. Cuguen. 2001. Nuclear effect on mitochondrial protein expression of the CMS Owsen cytoplasm in sugar beet. Theor. Appl. Genet. 102:1299–1304. Earle, E. D., and M. H. Dickson. 1995. Brassica oleracea cybrids for hybrid vegetable production. p. 171–176. In: M. Terzi et al. (eds.), Current issues in plant molecular and cellular biology. Kluwer, Dordrecht. Edwards, J. W., and J. G. Coors. 1996. Teosinte cytoplasmic genomes: II. Performance of maize hybrids with teosinte cytoplasms. Crop Sci. 36:1092–1098. Ekiz, H., A. S. Kiral, A. Akcin, and L. Simsek. 1998. Cytoplasmic effects on quality traits of bread wheat (Triticum aestivum L.). Euphytica 100:189–196. Forcioli, D., P. Samitou-Laparade, M. Valero, and J. Cuguen. 1998. Distribution of chloroplast DNA diversity within and among populations in gynodioecious Beta vulgaris ssp. marritima (Chenopodiaceae). Mol. Evol. 7:1193–1204. Frei, U. 1994. Multifusions. Potato Res. 37:338. Frei, U., M. Stattmann, A. Lössl, and G. Wenzel. 1998. Aspects of fusion combining ability of dihaploid Solanum tuberosum L.: influence of the cytoplasm. Potato Res. 41:155–162. Friesen, N., S. Pollner, K. Bachmann, and F. R. Blattner. 1999. RAPDs and noncoding chloroplast DNA reveal a single origin of the cultivated Allium fistulosum from A. altaicum (Alliaceae). Am. J. Bot. 86:554–562.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

205

Garraway, M. O., D. S. Park, and J. D. Beltran. 1998. Role of light and malate in the decreased sensitivity of cms-T cytoplasm maize leaves to Bipolaris maydis race T toxin. Phytopathology 88:556–562. Geiger, H. H., G. Öttler, R. Marker, F. H. Utz, and F. Wehmann. 1993. Evaluating rye inbred lines per se and in euplasmic and alloplasmic crosses for triticale suitability. Crop Sci. 33:725–729. Gleba, Y. Y., and K. M. Sytnik. 1984. Protoplast fusion. Springer Verlag, Berlin. Glimelius, K., J. Fahlesson, M. Landgren, M. Sjödin, and C. Sundberg. 1990. Improvement of the Brassica crops by transfer of genes from alien species via somatic hybridisation. p. 299–312. In: H. Nijkamp et al. (eds.), Progress in plant cell and molecular biology. Kluwer, Dordrecht. Golds, T., P. Maliga, and H. U. Koop. 1993. Stable plastid transformation in PEG-treated protoplasts of Nicotiana tabaccum. Bio/Technology 11:95–97. Gourret, J. P., R. Delourme, and M. Renard. 1992. Expression of Ogu cytoplasmic male sterility in cybrids of Brassica napus. Theor. Appl. Genet. 83:549–556. Gray, J. C., J. A. Sullivan, J. M. Hibberd, and M. R. Hansen. 2001. Stromules: Mobile protrusions and interconnections between plastids. Plant Biol. 3:223–233. Gray, J. C., J. A. Sullivan, and R. Sornarajah. 1999. Chloroplast control of nuclear gene expression. J. Exp. Bot. 50 (Suppl.) 48. Gray, M. W. 1999. Evolution of organellar genomes. Curr. Opin. Genet. Dev. 9:678–687. Grivell, L. A. 1995. Nuclei-mitochondrial interactions in mitochondrial gene expression. Crit. Rev. Biochem. Mol. Biol. 30:121–164. Guda, C., S. B. Lee, and H. Daniell. 2000. Stable expression of a biodegradable proteinbased polymer in tobacco chloroplasts. Plant Cell Rep. 19:257–262. Hagemann, R. 2002. Milestones in plastid genetics of higher plants. Prog. Bot. 63:5–51 Hallden, C., B. Gertsson, T. Sall, and C. Lind-Hallden. 1993. Characterization of organellar DNA in alloplasmic lines of Brassica napus L. Plant Breed. 111:185–191. Hanson, M. R., and O. Folkerts. 1992. Structure and function of higher plant mitochondrial genome. Int. Rev. Cytol. 141:129–172. Hasan, E., and C. F. Konzak. 1991. Nuclear and cytoplasmic control of anther culture response in wheat: I. Analyses of alloplasmic lines. Crop Sci. 31:1421–1427. Hedtke, B., I. Wagner, T. Börner, W. R. Hesws. 1999. Inter-organellar crosstalk in higher plants: impaired chloroplast development affects mitochondrial gene and transcript levels. Plant J. 19:635–643. Heifetz, P. B. 2000. Genetic engineering of the chloroplast. Biochimie 82:655–666. Hernandez, P., P. Barcelo, A. Martin, and A. Cabrera. 2001. The effect of Hordeum chilense and Triticum cytoplasm on anther culture response of tritordeum. Plant Cell Rep. 20:542–546. Herrmann, R. G. 1992. Cell organelles. Springer Verlag, Wien. Herrmann, R. G., R. M. Maier, and C. Schmitz-Linneweber. 2003. Eukaryotic genome evolution: Rearrangement and co-evolution of compartmentalised genetic information. Phil. Trans. Roy. Soc. (London) (in press). Hirose, T., and M. Sugiura. 1997. Both RNA editing and RNA cleavage are required for translation of tobacco chloroplast psbA mRNAs: development of an in vitro translation system from tobacco chloroplasts. EMBO J 16:6804–6811. Hofferbert, H. R., and G. Wenzel. 1997. Alternative Hybridzüchtung—somatische Hybridisation bei der Kartoffel. Ber. Arbeitstagung Gumpenstein 48:111–114. Hosaka, K. 1995. Successive domestication and evolution of the Andean potatoes as revealed by chloroplast DNA restriction endonuclease analysis. Theor. Appl. Genet. 90:356–363. Hosaka, K., and R. E. Hanneman. 1988. The origin of the cultivated tetraploid potato based on chloroplast DNA. Theor. Appl. Genet. 76:172–176.

206

U. FREI, E. PEIRETTI, AND G. WENZEL

Hsam, S. L. K., and F. J. Zeller. 1993. Haploid production in durum wheat by the interaction of Aegilops kotschyi cytoplasm and 1BL/RS chromosomal interchange. Theor. Appl. Genet. 86:951–954. Ishii, T., T. Terachi, N. Mori, and K. Tsunewaki. 1993. Comparative study on the chloroplast, mitochondrial and nuclear genome differentiation in two cultivated rice species, Oryza sativa and Oryza glaberrima, by RFLP analyses. Theor. Appl. Genet. 86:88–96. Johnson, A. A. T., and R. E. Veilleux. 2001. Somatic hybridization and applications in plant breeding. Plant Breed. Rev. 20:167–225. Kaul, M. L. H. 1988. Male sterility in higher plants. Springer, Berlin, Heidelberg. Keane, E. M., and P. W. Jones. 1990. Effect of alien cytoplasm substitution on the response of wheat cultivars to Septoria nodorum. Ann. Appl. Biol. 117:299–312. Kemble, R. J., and J. F. Shepard. 1984. Cytoplasmic DNA variation in a potato protoclonal population, Theor. Appl. Genet. 69:211–216. Khan, M. S., and P. Maliga. 1999. Fluorescent antibiotic resistance marker for tracking plastid transformation in higher plants. Nat. Biotechnol. 17:910–915. Knoblauch, M., A. J. E. van Bel, J. M. Hibberd, and J. C. Gray. 1999. Micro-injection and expression of reporter genes in plastids. J. Expt. Bot. 50:48. Knoop, V., and A. Brennicke. 2002. Molecular biology of the plant mitochondrion. Crit. Rev. Plant Sci. 21:111–126. Koop, U., K. Steinmüller, H. Wagner, C. Roessler, C. Eibl, and L. Sacher. 1996. Integration of foreign sequences into the tobacco plastome via polyethylene glycol-mediated protoplast transformation. Planta 199:193–201. Kueck, U., B. Laser, S. Mohr, E. Schulte-Kappert, C. Nowak, G. Öttler, B. Pfeiffer-Schad, C. U. Hesemann, R. Weigel, U. Pfeil, M. Stierle, T. Stösser, H. van der Kuip, M. Wolf, and W. Odenbach. 1995. CMS in wheat and triticale. Fortschritte der Pflanzenzüchtg. 18:19–36. Landgren, M., M. Zetterstrand, E. Sundberg, and K. Glimelius. 1996. Alloplasmic malesterile Brassica lines containing BBB. Tournefortii mitochondria express on ORF 3 of the atp6 gene a 32 Kda protein. Plant Molec Biol. 32:879–890. Laser, K. D., and N. R. Lersten. 1972. Anatomy and cytology of microsporogenesis in cytoplasmic male sterile angiosperms. Bot. Rev. 38:425–454. Laurent, V., A. M. Risterucci, and C. Lanaud. 1993. Chloroplast and mitochondrial DNA diversity in Theobroma cacao. Theor. Appl. Genet. 87:81–88. Lee, H. J., S. B. Lee, J. S. Chung, S. U. Han, J. O. Guh, J. S. Jeon, G. An, and K. Back. 2000. Transgenic rice plants expressing the Bacillus subtilis protoporphyrinogen oxidase gene are resistant to diphenyl ether herbicide oxyfluorfen. Plant Cell Physiol. 41:743–749. Levings, C. S. 1990. The Texas cytoplasm of maize: cytoplasmic male sterility and disease susceptibility. Science 250:942–947. Li, J., and L. Yuan. 2000. Hybrid rice: Genetics, breeding, and seed production. Plant Breed. Rev. 11:15–158. Lim, K. Y., R. Matyasek, C. P. Lichtenstein, and A. R. Leitch. 2000. Molecular cytogenetic analyses and phylogenetic studies in the Nicotiana section Tomentosae. Chromosoma 109:245–258. Lin, M. S. 1992. Ancestral contribution to IRRI rice varieties. Japan. J. Breed. 42:437–442. Linnert, G. 1997. Plasmatische Vererbung. p. 202–210. In: W. Odenbach (ed.), Biologische Grundlagen der Pflanzenzüchtung. Parey, Berlin. Londsdale, D. M., T. P. Hodge, and C. M.-R. Fauron. 1984. The physical map and organisation of the mitochondrial genome from fertile genome of maize. Nuc. Acid. Res. 12:9249–9261. Lössl, A., U. Frei, and G. Wenzel. 1994. Interaction between cytoplasmic composition and yield parameters in somatic hybrids of S. tuberosum L. Theor. Appl. Genet. 89:873–878. Lössl, A., N. Adler, R. Horn, U. Frei, and G. Wenzel. 1999. Chondriome-type characterization of potato: mt α, β, γ, δ, ε and novel plastid-mitochondrial configurations in somatic hybrids. Theor. Appl. Genet. 99:1–10.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

207

Lössl, A., M. Götz, A. Braun, and G. Wenzel. 2000. Molecular markers for cytoplasm in potato: male sterility and contribution of different plastid-mitochondrial configurations to starch production. Euphytica 116:221–230. Lübberstedt, T., V. Mohler, and G. Wenzel. 2002. Function of genetic material: Genes involved in quantitative and qualitative resistance. Prog. Bot. 63:80–105. Luo, H., B. V. Coppenolle, M. Seguin, and M. Boutry. 1995. Mitochondrial DNA phylogenetic relationships in Hevea brasiliensis. Mol. Breed. 1:51–63. Maliga, P. 2001. Plastid engineering bears fruit. Nat. Biotechnol. 19:826–827. Maliga, P., and L. Menczel. 1986. Chloroplast transfer and recombination through protoplast fusion. p. 601–612. In: I. K. Vasil (ed.), Cell culture and somatic cell genetics. Academic Press, Orlando. Manthey, R. 2001. Stand der in Deutschland zugelassenen Rapssorten. Raps 19:152–155. Mason, H. S., H. Warzecha, T. Mor, and C. J. Arntzen. 2002. Edible plant vaccines: applications for prophylactic and therapeutic molecular medicine. Trends Mol. Med. 8:324–329. Matsuzawa, Y., S. Mekiyanon, Y. Kaneko, S. W. Bang, K. Wakui, and Y. Takahata. 1999. Male sterility in alloplasmic Brassica rapa L. carrying Eruca sativa cytoplasm. Plant Breed. 118:82–84. McBride, K. F., Z. Svab, D. J. Schaaf, P. S. Hogan, and D. M. Stalker. 1995. Amplification of a chimeric Bacillus gene in chloroplasts leads to an extraordinary level of insecticidal protein in tobacco. Bio/Technology 13:362–365. Melchers, G., Y. Mohri, K. Watanabe, S. Wakabayashi, and K. Harada. 1992. One-step generation of cytoplasmic male sterility by fusion of mitochondrial-inactivated tomato protoplasts with nuclear-inactivated Solanum protoplasts. Proc. Natl. Acad. Sci. (USA) 89:6832–6836. Messmer, M., A. E. Melchinger, J. Boppenmaier, E. Brunklaus, and R. Herrmann. 1992. Relationships among early European maize inbreds: I. Genetic diversity among flint and dent lines revealed by RFLPs. Crop Sci. 32:1301–1309. Messmer, M., A. E. Melchinger, R. Herrmann, and J. Boppenmaier. 1993. Relationships among early European maize inbreds: II. Comparison of pedigree and RFLP data. Crop Sci. 33:944–950. Moneger, F., C. J. Smart, and C. J. Leaver. 1994. Nuclear restoration of male sterility in sunflower is associated with the tissue-specific regulation of a novel mitochondrial gene. EMBO J. 13:8–17. Mukai, Y. 1982. Genetic characterization of the cytoplasms of tetraploid Aegilops species. Memoirs Osaka Univ. Ser. III 31:13–34. Murai, K., and K. Tsunewaki. 1993. Photoperiod-sensitive cytoplasmic male sterility in wheat with Aegilops crassa cytoplasm. Euphytica 67:41–48. Neale, D. B., M. A. Saghi-Maroof, R. W. Allard, Q. Zhang, and R. A. Jorgensen. 1986. Chloroplast DNA diversity in populations of wild and cultivated barley. Genetics 120: 1105–1110. Nestares, G., R. Zorzoli, L. Mroginski, and L. Picardi. 1998. Cytoplasmic effects on the regeneration ability of sunflower. Plant Breed. 117:188–190. Newton, K. J. 1988. Plant mitochondrial genomes: organization, expression and variation. Ann. Rev. Plant Physiol. Plant Mol. Biol. 39:503–532. Nothnagel, T., P. Straka, and B. Linke. 2000. Male sterility in populations of Daucus and the development of alloplasmic male-sterile lines of carrot. Plant Breed. 119: 145–152. O’Neill, C., G. V. Horvath, E. Horvath, P. J. Dix, and P. Medgysey. 1993. Chloroplast transformation in plants: polyethylene glycol (PEG) treatment of protoplasts is an alternative to biolistic delivery systems. Plant J. 3:729–738. Öttler, G., and D. J. Mares. 1994. Falling number and alpha-amylase activity in developing grain of alloplasmic hexaploid wheat. Plant Breed. 112:47–52.

208

U. FREI, E. PEIRETTI, AND G. WENZEL

Ogihara, Y., K. Isono, T. Kato et al. (16 co-authors). 2002. Structural features of a wheat plastome as revealed by complete sequencing of chloroplast DNA. Mol. Gen. Genomics 266:740–746. Ogihara, Y., Y. Kurihara, K. Futami, K. Tsuji, and K. Murai. 1999. Photoperiod-sensitive cytoplasmic male sterility in wheat: Nuclear-mitochondrial incompatibility results in differential processing of the mitochondrial orf25 gene. Current Genet. 36:354–362. Orlov, P. A., E. B. Mavrishcheva, and A. N. Palilova. 1997. Analysis of the effect of nucleus and cytoplasm interactions on the anther culture efficiency in wheat. Cereal Res. Commun. 25:883–890. Orozco-Castillo, C., K. J. Chalmers, W. Powell, and R. Waugh. 1996. RAPD and organelle specific PCR re-affirms taxonomic relationships within the genus Coffea. Plant Cell Rep. 15:337–341. Panayotov, I. 1983. The cytoplasm in triticinae. Proc. 6th Intern. Wheat Gen. Symp., Kyoto, p. 481–497. Peiretti, E. G. 2003. Molekulare Charakterisierung der Cytoplasmen von Amarant, Mais und Weizen. Diss. TU München. Peiretti, E. G., and J. J. Gesumaria. 1998. Influencia de la distancia entre lineas sobre el crecimiento y rendimiento de amaranto granifero (Amaranthus spp.). Invest. Agr. Prod. Prot. Veg. 13:139–151. Pelletier, G., C. Primard, M. Frerault, F. Vedel, P. Chetrit, M. Renard, and R. Delourme. 1988. Use of protoplasts in plant breeding: cytoplasmic aspects. Plant Cell Tissue Organ Cult. 12:173–180. Poggio, L., M. Rosato, and C. A. Naranjo. 1997. Meiotic behavior and DNA content in alloplasmic lines of maize. Genome 40:723–729. Powell, W., M. Morgante, R. McDevitt, G. G. Vendramin, and J. A. Rafalski. 1995. Polymorphic simple sequence repeat regions in chloroplast genomes: application to the population genetics of pines. Proc. Natl. Acad. Sci. (USA) 92:7759–7763. Rhoads, D. M., H. C. Griffin, B. B. Neuschwander, C. S. Levings, and J. N. Siedow. 1996. Assays for characterizing URF13, the pathotoxin and methomyl receptor of cms-T maize. Meth. Enzymology 264:566–581. Rocheford, T. R., and D. R. Pring. 1994. Interaction of nuclear and mitochondrial genomes in the alteration of maize mitochondrial transcripts. Theor. Appl. Genet. 89:951–958. Rousstaee, A., and G. Barrault. 2000. Inheritance of partial resistance to black stem (Phoma madonaldii) in sunflower. Plant Pathol. 49:396–401. Ruf, S., H. Kössl, and R. Bock. 1997. Targeted inactivation of a tobacco intron containing open reading frame reveals a novel chloroplast-encoded photosystem I-related gene. J. Cell Biol. 139:95–102. Ruf, S., M. Herrmann, I. J. Berger, H. Carrer, and R. Bock. 2001. Stable genetic transformation of tomato plastids and expression of a foreign protein in fruit. Nat. Biotechnol. 19:870–875. Samitou-Laparde, P., G. Pannenbecker, V. Boutin-Stadler, G. Michaelis, and P. Vernet. 1991. Plastid DNA diversity in natural populations of Beta maritima showing additional variation in sexual phenotype and mitochondrial DNA. Theor. Appl. Genet. 81:533–536. Sasakuma, T., and S. S. Maan. 1978. Male sterility-fertility restoration systems in Triticum durum Desf. Can. J. Genet. Cytol. 20:389–398. Schardl, C. L., D. M. Lonsdale, and D. R. Pring. 1985. Mitochondrial DNA structures in male-sterile maize Zea mays and its fertile revertants. J. Cell Biochem. Suppl. 9C:262. Scheffler, E. 1999. Mitochondria. Wiley-Liss, New York. Schmidhuber, M., and F. J. Zeller. 1998. Amarant: mehr Protein fürs Brot. Back J. 98:65–66. Schmitz-Linneweber, C., R. Regel, T. G. Du, H. Hupfer, R. G. Herrmann, and M. Maier. 2002. The plastid chromosome of Atropa belladonna and its comparison with that of Nicotina tabacum: The role of RNA editing in generating divergences in the process of plant speciation. Mol. Biol. Evol. 19:1602–1612.

4. SIGNIFICANCE OF CYTOPLASMIC DNA IN PLANT BREEDING

209

Schwarz, Z., and H. Kössel. 1979. Sequencing of the 3’-terminal region of a 16 S rRNA gene from Zea mays chloroplasts reveals homology with E. coli 16 S rRNA. Nature 279: 520–522. Serino, G., and P. Maliga. 1997. A negative selection scheme based on the expression of cytosine deaminase in plastids. Plant J. 12:697–701. Shinozaki, K., M. Ohme, M. Tanaka, T. Wakasuki, N. Hayashida, T. Matsubayasha, N. Zaita, J. Chungwongse, J. Obokarta, K. Yamaguchi-Shinozaki, C. Ogto, K. Torazawa, B. Y. Meng, M. Sugita, H. Deno, T. Kamogashira, K. Yamada, J. Kusuda, F. Takaiwa, A. Kata, N. Tohdo, H. Shimada, and M. Sugiura. 1986. The complete nucleotide sequence of the tobacco chloroplast genome: its gene organisation and expression. EMBO J. 5:2043–2049. Sidorov, V. A., D. Kasten, S. Z. Pang, P. Hajdukiewicz, J. M. Staub, and N. S. Nehra. 2000. Stable chloroplast transformation in potato: use of fluorescent protein as a plastid marker. Plant J. 19:209–216. Sikdar, S. R., G. Serino, S. Chaudhun, and P. Maliga. 1998. Plastid transformation in Arabidopsis thaliana. Plant Cell Rep. 18:20–24. Sonnino, A., S. Tanaka, M. Iwanaga, and L. Schilde-Rentschler. 1989. Genetic control of embryo formation in anther culture of diploid potatoes. Plant Cell Rep. 8:105–107. Spangenberg, G., E. Freydl, M. Osusky, J. Nagel, and I. Potrykus. 1991. Organelle transfer by microfusion of defined protoplast-cytoplast pairs. Theor. Appl. Genet. 81:477–486. Staub, J. M., B. Garcia, J. Graves, P:T. J. Hajdukiewicz, P. Hunter, N. Nehra, V. Paradkar, M. Schlittler, J. A. Carroll, and L. Spatola. 2000. High-yield production of a human therapeutic protein in tobacco chloroplasts. Nat. Biotechnol. 18:333–338. Steinborn, R., B. Linke, T. Nothnagel, and T. Börner. 1995. Inheritance of chloroplast and mitochondrial DNA in alloplasmic forms of the genus Daucus. Theor. Appl. Genet. 91:632–638. Steinhauser, S., S. Beckert, and I. Capesius. 1999. Plant mitochondrial RNA editing. J. Molec. Evol. 48:303–312. Stern, D. B., D. C. Higgs, and J. Yang. 1997. Transcription and translation in chloroplasts. Trends Plant Sci. 2:308–315. Stoddard, F. L. 2000. Genetics of wheat starch B-granule content. Euphytica 112:23–31. Stoddard, F. L., and I. H. M. H. B. Hearth. 2001. Genetic analysis of partial rust resistance in faba beans. Austral. J. Agr. Res. 52:73–84. Sun, C. Q., X. K. Wang, A. Yoshimura, and K. Doi. 2002. Genetic differentiation for nuclear, mitochondrial and chloroplast genomes in common wild rice (Oryza rufipogon Griff.) and cultivated rice (Oryza sativa L.). Theor. Appl. Genet. 104:1335–1345. Sundberg, E., and K. Glimelius. 1991. Effects of parental ploidy level and genetic divergence on chromosome elimination and chloroplast segregation in somatic hybrids within Brassicaceae. Theor. Appl. Genet. 83:81–88. Svab, Z., P. Hajdukiewicz, and P. Maliga. 1990. Stable transformation of plastids in higher plants. Proc. Natl. Acad. Sci. (USA) 87:8526–8530. Temple, M., C. A. Makaroff, M. A. Mutschler, and E. D. Earle. 1992. Novel mitochondrial genomes in Brassica napus somatic hybrids. Curr. Genet. 22:243–249. Terachi, T., Y. Ogihara, and K. Tsunewaki. 1984. The molecular basis of genetic diversity among cytoplasms of Triticum and Aegilops. 3. Chloroplast genomes of the M and modified M genome-carrying species. Genetics 108:681–695. Terachi, T., and K. Tsunewaki. 1992. The molecular basis of genetic diversity among cytoplasms of Triticum and Aegilops. VIII. Mitochondrial RFLP analyses using cloned gene probes. Mol. Biol. Evol. 9:917–931. Timothy, D. H., C. S. Levings, D. R. Pring, M. F. Cone, and J. L. Kermicle. 1979. Oranelle DNA variation and systematic relationships in the genus Zea: Teosinte. Proc. Natl. Acad. Sci. (USA) 76:4220–4224.

210

U. FREI, E. PEIRETTI, AND G. WENZEL

Tsunewaki, K., T. R. Endo, and Y. Mukai. 1974. Further discovery of alien cytoplasms inducing haploids and twins in common wheat. Theor. Appl. Genet. 45:104–109. Tsunewaki, K., G. Z. Wang, and Y. Matsuoka. 1996. Plasmone analysis of Triticum (wheat) and Aegilops. 1. Production of alloplasmic common wheats and their fertilities. Genes Genetic Syst. 71:293–311. Vedel, F., F. Quetier, F. Dosba, and G. Doussinauls. 1978. Study of wheat phylogeny by EcoRI analysis of chloroplastic and mitochondrial DNAs. Plant Sci. Lett. 13:97–102. Virk, D. S., and J. S. Brar. 1993. Assessment of cytoplasmic differences of near-isonuclear male-sterile lines in pearl millet. Theor. Appl. Genet. 87:106–112. Virmani, S. S., and H. L. Sharma. 1993. Manual for hybrid rice seed production. IRRI, Manila. Vivek, B. S., Q. A. Ngo, and P. W. Simon. 1999. Evidence for maternal inheritance of the chloroplast genome in cultivated carrot (Daucus carota L. ssp. sativus). Theor. Appl. Genet. 98:669–672. Walters, T. W., and E. D. Earle. 1993. Organellar segregation, rearrangement and recombination in protoplast fusion derived Brassica oleracea calli. Theor. Appl. Genet. 85:761–769. Wang, C., S. Tang, and Y. Tang. 1998. Effects of male sterile cytoplasm on yield and agronomic characters in Japonica hybrid rice, Oryza sativa L. Breed. Sci. 48:263–271. Wang, G. Z., Y. Matsuoka, and K. Tsunewaki. 2000. Evolutionary features of chondriome divergence in Triticum (wheat) and Aegilops shown by RFLP analysis of mitochondrial DNAs. Theor. Appl. Genet. 100:221–232. Wenzel, G. 1994. Tissue culture. p. 173–195. In: J. E. Bradshaw and G. R. Mackay (eds.), Potato genetics. CAB Intl., Wallingford. Wenzel, G., and B. Foroughi-Wehr. 1994. Production and use of isogenic lines. p. 153–172. In: I. K. Vasil and T. A. Thorpe (eds.), Plant cell and tissue culture. Kluwer, Dordrecht. Wenzel, G., V. Mohler, and F. Thuemmler. 2001. Marker techniques as tools for efficient plant breeding. Proc. Workshop Soybean Biotechnology, Bogor, p. 4–14. Wenzel, G., and H. Uhrig. 1981. Breeding for nematode and virus resistance in potato via anther culture. Theor. Appl. Genet. 80:359–365. Wilson, J. 1967. Problems in hybrid wheat breeding. Euphytica Suppl. Wilson, J. A., and W. M. Ross. 1962. Male sterility interaction of the Triticum aestivum nucleus and Triticum timopheevi cytoplasm. Wheat Inform. Service 14:29–30. Wu, Y., C. Zhang, L. Chunguang, R. Shuxoin, and Z. Yan. 1998. Breeding technology of alloplasmic wheat. Sci. China C 41:449–458. Xu, D. H., J. Abe, J. Y. Gai, and Y. Shimamoto. 2002. Diversity of chloroplast DNA SSRs in wild and cultivated soybeans: evidence for multiple origins of cultivated soybean. Theor. Appl. Genet. 105:645–653. Xu, Y. S., M. S. Clark, and E. Pehu. 1993. Use of RAPD markers to screen somatic hybrids between Solanum tuberosum and S. brevidens. Plant Cell Rep. 12:107–114. Yadav, O. P. 1994. Effect of the cytoplasmic source on the combining ability of agronomic traits in pearl millet. Plant Breed. 113:242–245. Yamaguchi, K., and K. V. Knoblauch. 2000. The plastid ribosomal proteins. Identification of all the proteins in the 30 S subunit of an organelle ribosome (chloroplast). J. Biol. Chem. 275:28455–28465. Ye, X., S. Al-Babili, A. Klöti, J. Zhang, P. Lucca, P. Beyer, and I. Potrykus. 2000. Engineering the provitamin A (β-carotene) biosynthetic pathway into (carotenoid-free) rice endosperm. Science 287:303–305.

5 Flowering, Seed Production, and the Genesis of Garlic Breeding Philipp W. Simon Vegetable Crops Research Unit, United States Department of Agriculture—Agricultural Research Service, Department of Horticulture, University of Wisconsin, Madison, WI 53706 Maria M. Jenderek National Arid Land Plant Genetic Resource Unit, United States Department of Agriculture—Agricultural Research Service, San Joaquin Valley Agricultural Science Center, Parlier, CA 93648 I. II. III. IV.

INTRODUCTION GARLIC PRODUCTION TRENDS GARLIC TAXONOMY AND GENETIC VARIATION GARLIC GROWTH AND REPRODUCTIVE BIOLOGY A. Morphology and Growth B. Reproductive Biology V. GARLIC SEED PRODUCTION A. Importance of Garlic Germplasm for Seed Production B. Processes and Procedures for Garlic Seed Production VI. PROGRESS IN GARLIC BREEDING AND FUTURE PROSPECTS A. Selection for Improved Floral Characteristics and Fertility B. Selection for Improved Seed Size and Vigor C. Garlic Breeding Goals D. Garlic Breeding Methods VII. CONCLUSIONS LITERATURE CITED

I. INTRODUCTION Garlic is one of the oldest horticultural crops. There are Egyptian and Indian references to garlic 5000 years ago, clear evidence of Babylonian usage 4500 years ago, and usage in China 2000 years ago, although some Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 211

212

P. SIMON AND M. JENDEREK

writings suggest that it was grown there as long as 4000 years ago (Hahn 1996). In spite of its long history, little is known about early garlic production or plant types used for cultivation. Garlic production today relies completely upon asexual propagation of the crop and, while flowering in garlic has long been known, no record of true seed production in garlic is known before 1950. While domestication implies selection for improved adaptation and production in cultivation by humans, it could be argued that despite its long cultivation, garlic still remains undomesticated. This statement presumes that domestication only proceeds when humans can tap the broad range of heritable variation resulting from sexual reproduction. However, it seems likely that sexual reproduction and selection among products of meiosis and syngamy may have been realized in cultivation from time to time during garlic’s long history. Furthermore, clonal selection has clearly generated new cultivars of non-sexual, clonally propagated crops such as seedless bananas, sweet potatoes, Caladium, Colocasia, potatoes, apples (Sharma 1956), and also garlic (Burba 1997). Nevertheless, flowering and seed production research in garlic over the last 50 years sets the scene for utilizing a system of classical plant breeding in this important crop that has been used as the mainstay of plant breeding today. True seed production in garlic provides a foundation for dramatically changing the production and reproduction technologies used by garlic producers in the future. Etoh (1985) reviewed the early research on garlic sterility and seed production, while Etoh and Simon (2002) recently reviewed garlic diversity, fertility, and seed production. Darlington (1939) maintained that, in nature, asexual reproduction provides only limited genetic options for continued adaptation, saying: “With loss of sexual recombination, the apomict . . . is cut off from ultimate survival. Apomixis is an escape from sterility but it is an escape into a blind alley of evolution.” Human intervention has apparently sustained and even favored asexual reproduction throughout the agricultural history of garlic. The prospects for true seed production in garlic now hold out the possibility of capturing the broader diversity that allogamy and recombination can bring.

II. GARLIC PRODUCTION TRENDS Garlic is widely grown, with current worldwide production at nearly 10 million tonnes on nearly a million hectares, and it is used as a spice, condiment, and vegetable (Table 5.1). China and India are the largest producers of garlic, collectively accounting for 60% of the production area and 69% of the world yield. World garlic production area has more than doubled since 1970, and its availability has almost doubled in that

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

213

Table 5.1. World garlic production and availability, 1961–2000. Source: FAO Yearbook Production Statistics; average of 3 years for 1970, 1980, 1990, and 2000 (i.e., 1970 = average of 1969–1971). Yield (t/ha)

Availability (kg/capita)

1 1 2 3 3

10.5 23.2 16.3 12.7 11.4

0.2 0.4 0.4 0.4 0.4

1961 1970 1980 1990 2000

86 74 74 77 76

5.7 7.4 7.3 9.0 10.7

2.2 1.2 1.3 1.7 2.2

Europe

1961 1970 1980 1990 2000

8 16 15 13 12

5.2 5.3 5.2 5.4 6.1

0.8 0.8 0.9 1.0 0.7

North and Central America

1961 1970 1980 1990 2000

1 2 2 3 3

4.3 6.8 9.7 13.6 15.2

0.1 0.2 0.4 0.6 0.8

Oceania

1961 1970 1980 1990 2000

<1 <1 <1 <1 <1

1.0 1.0 7.0 5.9 6.0

<0.1 <0.1 <0.1 <0.1 <0.1

South America

1961 1970 1980 1990 2000

4 6 6 5 6

3.1 3.7 4.5 5.1 7.4

0.6 0.6 0.6 0.6 0.9

World

1961 1970 1980 1990 2000

771 464 627 771 985

5.6 6.9 7.0 8.4 10.0

1.4 0.9 1.0 1.2 1.6

Continent

Year

Africa

1961 1970 1980 1990 2000

Asia

Area (% world)

period to a current annual consumption of 1.64 kg/capita. Asia is by far the major garlic production region, but increased area and availability have kept up with population in the rest of the world, with especially sharp rises between 1990 and 2000 in the Americas. In spite of the wide geographic range of garlic production today, little is known about the similarities and differences among major

214

P. SIMON AND M. JENDEREK

cultivars grown in different growing regions. In this paper, we use the term “clone” to indicate garlic cultivars, strains, or genotypes that are asexually propagated. The major distinction among garlic clones is the tendency to produce a flower stalk, or not. Those producing a flower stalk are referred to as “bolting,” “stalking,” or “hard neck” types, whereas those without a flower stalk are “non-bolting,” “non-stalking,” or “soft neck.” Sometimes a third category, “incomplete or partial bolting,” is used for clones in which the inflorescence begins to develop but stem elongation is incomplete and mature flowers do not develop (Takagi 1990). There is a presumption that this character is stable across environments but this is unproven. Likewise, it is assumed that, like onion, garlic bulb formation is a response to photoperiod and some reference is made to “shortday” cultivars (which produce bulbs as photoperiods lengthen slightly in the spring in regions near the Equator), and “long-day” cultivars (which produce bulbs as photoperiods lengthen dramatically in the spring in regions farther from the Equator). This has not been substantiated either, although both bolting and non-bolting garlic clones are produced in nearly all production regions. Grower and consumer preferences for garlic vary among different geographic regions and end uses. Intact or processed garlic bulbs constitute the usual form of the crop traded in commerce, but in parts of Asia and North Africa leaves are marketed. Preferred garlic cultivars for fresh bulb sale include those with red or pink cloves surrounded by white bulb scales and those with white cloves and bulb scales. Processors of prepared fresh garlic (e.g., minced, chopped, crushed, or sliced) use both types. Dehydrators require the latter type with little pigment and high dry matter content. The inclination of consumers toward unusual and pigmented vegetables has resulted in specialty markets for novel and colored garlic. The development of garlic for the nutraceutical industry has resulted in another specialty market with high levels of sulfur compounds such as allicin, that are often correlated with strong flavor. True seed production of garlic will likely provide a more diverse array of variation from which users can select. The strictly asexual propagation of garlic brings with it several viruses, nematodes, and other pests, which lower garlic yield. Consequently, an extensive research effort and very significant production resources are dedicated to reducing or eliminating these pathogens from “seed” garlic, plants used for crop production rather than consumption (van Dijk 1994; Verbeek et al. 1995; Salomon 2002). The prospects of true seed production in garlic could dramatically reduce the resources dedicated to this aspect of garlic production, especially if genetic resistance can be identified and incorporated into new cultivars.

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

215

III. GARLIC TAXONOMY AND GENETIC VARIATION Plant systematists have struggled with the taxonomy of garlic, Allium sativum L. It is a diploid species (2n = 2x = 16) in the subgenus Allium of the Alliaceae (formerly in the Liliaeceae, and then the Amaryllidaceae). The other cultivated plants in this subgenus are leek, usually tetraploid, or elephant garlic, usually hexaploid (both A. ampeloprasum L.). Leek and garlic have flat, folded leaves. Elephant garlic and garlic form a bulb, but leek does not. Elephant garlic bulbs consist of 2 to 6 large cloves and several small cloves, while garlic bulbs usually have more cloves of a relatively consistent size, especially for bolting types. Bolting garlic, leek, and elephant garlic all have a solid scape, unlike the hollow scape of the most economically important Allium, onion (A. cepa L.). Garlic flower color ranges from white to pink to purple and inflorescences almost always include not only flowers but also bulbils (small, undivided bulbs occurring in the inflorescence; also referred to as “topsets” or “air bulbs”), whereas A. ampeloprasum inflorescences form bulbils only sporadically. Early classifications considered A. sativum to be a species known only in cultivation, with A. longicuspis Rgl. as a closely related wild relative originating in Central Asia (Vvedensky 1935; Jones and Mann 1963). The diversity of Central Asian garlic supports the idea that this region is the primary center of origin of garlic, although this region was likely much larger in past history (reviewed by Engeland 1991; Etoh and Simon 2002). The distinctiveness of A. longicuspis as a separate species, primarily differing from garlic in having exserted anthers, has been brought into question. Several botanical varieties of garlic are described in the literature, including A. sativum var. sativum L., which rarely or never flowers; A. sativum var. ophioscorodon (Link) Doll (the varietal name meaning “serpent garlic”), which regularly flowers; and A. sativum var. pekinense (Prokh.) Makino, which rarely flowers, like A. sativum var. sativum, but has wider leaves (Helm 1956; Jones and Mann 1963). Several other garlic subspecies names have been used in the literature. For example Kuznetsov (1954) and Kommisarov (1964, 1965) referred to ssp. vulgare and ssp. sagittatum, while Kazakova (1971, 1978) referred to ssp. sativum or mediterraneum and ssp. asia-mediae Kaz., but the contrast of bolting versus non-bolting has been the primary distinction made among subspecies. Most recent considerations of A. sativum taxonomy include either two or three groups. Hanelt (1990) included a “common garlic group” (consisting of A. sativum var. sativum, A. sativum var. typicum Rgl., and A. pekinense Prokh.) and an “ophioscorodon group” (consisting of A. sativum var. ophioscorodon (Link) Doll, A. ophioscorodon Link, and A. sativum var. controversum (Schrad.)

216

P. SIMON AND M. JENDEREK

Moore jr.), while Fritsch and Friesen (2002) added a third “longicuspis group” (consisting of A. longicuspis Rgl.). More recently, A. tuncelianum has been proposed as the wild progenitor of garlic (Mathew 1996). This suggestion has generated discussion (Etoh and Simon 2002; Fritsch and Friesen 2002), including the point that garlic is regarded as a Central Asian species, while A. tuncelianum is currently found in Turkey. A wide range of morphological diversity has been observed in garlic, including flowering ability (and scape length of flowering clones), leaf traits (color and attitude), bulb traits (shape, outer leaf scale color, clove color, number, size, and organization of cloves), plant maturity, bulbing response to temperature and photoperiod, cold-hardiness, bulbil traits (number, size, color, and ease of removal), and flower traits (timing, number, size, color, petal closure, and stigma position) (McCollum 1976; Hwang 1993; Messian et al. 1993; Lallemand et al. 1997; Hong et al. 2000a; Senula and Keller 2000; Jenderek and Hannan 2002). Detailed classification of garlic germplasm collections based solely on morphological diversity has been considered difficult due to the subjective nature of categorization for several of these traits, although Hwang (1993) and Senula and Keller (2000) were able to develop detailed comparisons based upon principal component analysis of 10 traits and cluster analysis of 16 traits. Concomitant with this morphological variation, evaluations of molecular variation have noted significant levels of variation in isozymes, RAPDs, and AFLPs (Etoh and Ogura 1981; Siqueria et al. 1985; Etoh 1985; Tsuneyoshi et al. 1992; Pooler and Simon 1993a; Maaß and Klaas 1995; Bradley et al. 1996; Al-Zahim et al. 1997; Lallemand et al. 1997; Hong et al. 1997; Etoh et al. 2001; Garcia Lampasona et al. 2002; Ipek et al. 2003). In earlier studies utilizing isozymes, relatively few polymorphisms were detected and consequently unique “fingerprints” were unable to be determined for even clearly morphologically different clones, whereas with the utilization of RAPDs, and more recently AFLP molecular markers, hundreds of polymorphic bands can be detected to allow for clear demarcations of differences in diverse garlic germplasm collections. Generally the groupings ascertained by morphological variation are positively correlated with molecular variation, and clustering of clones in the ophioscorodon and “common” (sativum) botanical varieties is evident. Maaß and Klaas (1995) included subtropical and pekinense clones in their study, and suggested that subtropical clones were “clearly separated from all other types,” while the pekinense subgroup was relatively similar to the stalking types. Interestingly, most molecular “fingerprinting” studies which included A. longicuspis clones produced no basis for distinction between A. sativum and A. longicuspis as separate species. Field evaluation of a genetically—and

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

217

geographically—diverse collection of all A. sativum subspecies, A. longicuspis, and A. tuncelianum in very diverse climates, combined with molecular fingerprinting, may help further resolve garlic taxonomy. A combination of field evaluation and molecular fingerprinting will also be helpful in characterizing world garlic cultivars. Cultivars are clones, so any garlic production plot is expected to be genetically invariable. Molecular fingerprinting studies have demonstrated wide DNA variation concomitant with wide phenotypic variation observed among cultivars in a growing area. However, it is not known how cultivars from one region compare to those from another, since cultivar names are rarely constant over diverse growing areas. Knowledge of these relationships may shed light on crop history and adaptation. Variation among chromosomes in the garlic karyotype has also been noted for centromere location, chromosome length, and the number of chromosome satellites (Takenaka 1931; Krivenko 1938; Mensinkai 1939; Battaglia 1963; Gohil and Koul 1971; Konvicka and Levan 1972; Verma and Mittal 1976; Etoh and Ogura 1978; Etoh 1979; Etoh 1983b; Etoh 1985; Hong et al. 2000a). The extensive karyotypic evaluation by Hong et al. (2000a) demonstrated a preponderance of the “basic” karyotype (2 sets of chromosomes) in bolting Central Asian clones proven to produce seed, while a collection of partial-bolting clones from the Iberian Peninsula usually had “non-basic” karyotypes, including heteromorphic pairs and unusual chromosome variation in terms of centromere location and/or satellites. Assessment of morphological variation in garlic can be complicated by what is perceived as a gradual response by garlic to the environment. For example, an accurate assessment of yield, maturity, and even flowering ability for a garlic clone is often reserved until 3–5 years of propagation after arrival at a new location (Engeland 1991; Pooler 1991), especially if large changes in photoperiod or climate are involved. With this, garlic clones adapted to shorter summer photoperiods (i.e., at lower latitudes) will usually grow well the first year or two after being moved to a long summer photoperiod environment, but in subsequent years smaller bulbs are produced and eventually no bulbing occurs before senescence, and the clone is lost. This is thought to be a photoperiod response, but it is unproven. One variable that has a significant negative effect on garlic performance and a confounding effect on germplasm evaluation is virus infection. The potyviruses onion yellow dwarf virus and leek yellow stripe virus are most economically important. Because virus infection is so common in garlic, and resistance has not been identified, there has been extensive research and effort devoted to generating virus-free garlic for production areas around the world (van Dijk 1994; Verbeek et al. 1995;

218

P. SIMON AND M. JENDEREK

Salomon 2002). It is not uncommon for the virus-freeing process to increase garlic yield from 5 to 20% with a delay of plant senescence of up to 5 weeks. IV. GARLIC GROWTH AND REPRODUCTIVE BIOLOGY A. Morphology and Growth Factors influencing garlic morphology and growth have been discussed in many reports, including several comprehensive studies (Mann 1952; Mann and Minges 1958; Takagi 1990). The garlic bulb of commerce typically weighs 60 to 120 g, and consists of cloves surrounded by the dried basal sheaths of the foliage leaves. Bolting garlic bulbs usually consist of 4–12 cloves that tend to be relatively similar in size, while nonbolting garlic usually have more cloves that can vary greatly in size. Cloves are sessile lateral bulbs, that originate from axillary buds of inner (younger) foliage leaves. An outer, thin protective leaf and an inner, thickened, bladeless storage leaf are the most predominant parts of a garlic clove, with the storage leaf accounting for most of the volume and weight of the clove. The storage leaf subtends a central vegetative bud over a flattened basal plate, which is a modified very short stem. The central vegetative bud includes a predominant sprout leaf and several foliage leaf primordia that surround the apical meristem. The growth of a garlic plant typically begins from an individual clove that has been exposed to cool temperatures (15°C or less) and depleted or “broken” dormancy (Rahim and Fordham 1988; Takagi 1990). Depending upon how long the mature bulb was stored, the sprout leaf and foliage leaves may still be contained within the storage leaf, or may have elongated (sprouted) through the top of that leaf to protrude above it. Roots grow quickly around the perimeter of the basal plate and precede further leaf development. Leaves arise in an opposite and alternate orientation. Since a given leaf blade emanates from a broad meristematic arch, the bases of consecutive leaves are extensively overlapped. This overlap of leaf bases results in the production of a “pseudostem” that consists of closely appressed leaf sheaths surrounding young leaves. Above the pseudostem, leaf blades extend without overlap. The root system and flat leaves usually develop before clove initiation (bulbing) ensues from inner (younger) leaf axillary buds. Many variables, including storage, temperature, growing temperature, planting time, and photoperiod, interact with clove size and the genetic predisposition of a garlic clone to determine the number and size of leaves and cloves realized in a production cycle. Plant maturation and senescence usually ensues after garlic bulb production. As the main stem of the bulb dies,

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

219

intact and living basal plates remain on each clove to carry on future growth. Studies investigating broad ranges in photoperiod, air and soil temperature, and light intensity during production, and bulb storage time and temperature have noted relatively small effects on garlic leaf size and number, bulb and clove size, and bulb dormancy, compared to the effects of those environmental variables in other Alliaceae (Takagi 1990; Pooler and Simon 1993b), although extended cold storage increased the incidence of flowering plants for some clones, shading slowed maturity and reduced yield, and “short day” cultivars often produced small, undivided bulbs (“rounds”) in long photoperiods. B. Reproductive Biology Most reports of garlic morphology and growth focus on the plant and bulb, although several include evaluations of floral initiation and development (Vanin 1947; Aoba 1966; Takagi 1990; Pooler and Simon 1993b). Like other alliums, garlic flowers are perfect, with 6 petals, 6 anthers, and 3 locules consisting of 2 ovules each (Fig. 5.1). Garlic flowers are smaller

Fig. 5.1.

Garlic flowers and a bulbil from a male fertile clone.

220

P. SIMON AND M. JENDEREK

than onion. The number of flowers in an umbel varies from less than 10 to over 300, but the typical range is from 150 to 200. The inception of routine seed production in garlic has provided a wealth of information about its reproductive biology (Etoh 1985; Pooler and Simon 1994; Jenderek 1998). Garlic flowers are protandrous with 2 to 4 days between pollen shedding and stigma receptivity. The stigma is receptive for 1 to 2 days and styles are usually longer than petals. Across a garlic inflorescence, anthesis occurs over 5 to 20 days. The opposing petals of most mature garlic flowers form an angle of approximately 45° although some seed-derived plants have nearly closed petals. Petal color in bolting garlic clones is usually light purple, with or without darker coloration of pedicel and petal tip. In seed-derived plants, lilac and white petals also occur. Anthers of typical clonally-propagated garlic are yellow or purple, while in seed-derived plants light gray and lilac anthers also occur. Development of the floral apical meristem in bolting garlic requires exposure to low temperatures (< 5°C) before onset of the primary growth period. The details of garlic vernalization requirements are not well known. Most garlic is fall planted and if winter temperatures are lower than 5°C, the bolting phenotype is presumed to be induced. However, a systematic study of cold-treated non-bolting garlic clones typically grown in warmer climates could reveal bolting clones. A very significant genetic predisposition contributes to flowering in garlic, as some clones almost never flower with inductive conditions (non-bolting clones), while others flower readily (bolting clones) (Takagi 1990; Pooler and Simon 1993b). No treatment to routinely induce flowering in non-bolting garlic has yet been devised. As with other alliums, garlic clones that tend to flower must attain a minimum size or physiological maturity to realize that potential. Therefore, spring planting, which reduces bulb yield and does not allow enough time for plants to reach a certain maturity level, can significantly reduce or eliminate flowering. Low and, especially, high temperatures during inflorescence and scape elongation can also result in incomplete bolting (Takagi 1990). The garlic seed stalk develops from the tip of the main stem, precluding further vegetative growth (Mann 1952). Seed stalk formation represents the only significant internode elongation in the life of a garlic plant. Early scape elongation precedes any visible indication in the apical meristem that floral development is commencing. During scape elongation the meristematic region at its tip begins to differentiate floral initials and subdivides into identifiable flowers interspersed with bulbils (Kothari and Shah 1974a,b; Etoh and Ogura 1977; Etoh 1985; QuYing et al. 1994; Kamenetsky and Rabinowitch 2001, 2002). For incomplete bolting clones, the arrested scape development often results in a very short scape with flower and bulbil formation ranging from

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

221

slightly above the cloves within the bulb, to a more elevated position in the pseudostem, to a short stem protruding slightly above the pseudostem. Complete bolting in garlic involves floral induction, scape elongation, inflorescence development, and floral maturation (Takagi 1990; Kamenetsky and Rabinowitch 2001, 2002). Clove development tends to be initiated somewhat earlier in the development of non-bolting plants than in incomplete or complete bolting plants, suggesting that there is a balance or competition between bulb and inflorescence development, with bulb development taking precedence and floral development being optional. The extent to which a plant initiates floral development is strongly influenced by genotype and contingent upon exposure to low temperatures. During the growth of flowering garlic, there is a balance or competition between storage organs and flowers in terms of relative development and resource allocation (Pooler 1991; Pooler and Simon 1993b, 1994). The interplay between clove development in the bulb (bulb formation) and clove development in the inflorescence (bulbil formation) at the whole plant level seems to have a counterpart in the interplay between flowers and bulbils within the inflorescence. Bulbing without floral initiation and development is common, as seen in non-bolting garlic, but floral development without bulbing is not reported except for rare instances during in vitro cultivation. In most bolting cultivated garlic, bulbil formation takes precedence over floral development, although usually not to the complete exclusion of at least some floral development. Thus we can represent the relative strength of these three competing resource sinks as follows: 1. Bulb development excludes bulbil and flower development in non-bolting or unvernalized bolting plants; 2. Bulb development predominates over bulbil development and flower development in most cultivated bolting garlic plants; or 3. Bulb development, bulbil development, and flower development occur in some bolting cultivars, wild garlic, and some seedlings. Once the environmental conditions to stimulate floral development have been met, the rate and degree of floral and bulbil development varies widely among flowering garlic clones (Fig. 5.2). Without the development of mature flowers, the production of viable gametes and seed is not possible. Garlic fertility is primarily determined by genetic background, and fertile clones usually have numerous healthy flowers and small bulbils. Infertile, barren flowering clones vary widely in floral and bulbil development. Large bulbils usually indicate little potential for seed production. For some barren clones, mature healthy-appearing flowers develop, but many barren clones develop flowers that senesce early in the bud stage before buds even reach meiosis so mature flowers never

222

P. SIMON AND M. JENDEREK

Fig. 5.2. Inflorescences from six garlic clones with bulbils and senesced flowers late in the season.

develop (Weber 1929; Gvaladze 1965; Etoh 1985; Pooler and Simon 1994). The causes of this form of sterility have not been studied. Examination of developing flowers during scape elongation also noted abortion of some flowers in fertile clones (Qu-Ying et al. 1994). Floral development is no guarantee of viable gamete production. Studies of 154 diverse primarily Eastern garlic clones (Etoh 1985) and of 210 Central and Western clones (Pooler and Simon 1994) yielded complete bolting in 46% and 65% of the clones, respectively, but in the latter study about 1/2 to 2/3 of these clones failed to develop fully mature flowers,

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

223

since buds withered and died. This premature senescence takes several forms. In some clones, sporogenous tissue failed to enter meiosis, while in others microsporogenesis was complete, but then microspores degenerated at, or beyond, the tetrad stage. In the former study, one flowering clone developed fertile pollen, while in the latter study 27 male fertile clones were recorded, and in both studies these were Central Asian clones. Numerous other clones from the region of garlic’s center of diversity have also been collected and found to be male fertile (Etoh 1986; Etoh et al. 1992; Pooler and Simon 1994; Hong and Etoh 1996; Hong et al. 2000a). The causes of garlic floral abortion and sterility are not well understood. An accumulation of pathogens, chromosome aberrations, postmeiotic tapetal degeneration, and unsuccessful competition of floral development with bulbils have been implicated as contributing to the failure of flowers to develop (Etoh 1985). Konvicka and co-workers noted leaf necrosis in flowering plants and microscopic inclusions resembling rickettsia-like organisms, both of which became less apparent with antibiotic treatments (tetracycline and tylosine) of cut stems (Konvicka 1973, 1984; Konvicka et al. 1978). Some clones with complete synapsis, bivalent formation, and normal microsporogenesis are male sterile. In these clones, tapetal hypertrophy and other disturbances inhibit further microspore development (Novak 1972; Etoh 1982). These observations were interpreted by Konvicka to support a pathogenic cause of garlic floral abortion and sterility. However, Novak and Havranek (1975) and Pooler (1991) observed ambiguous results with antibiotic treatments, while Etoh (1980) observed phytotoxicity and arrested development of pollen with antibiotic treatment of cut scapes. Furthermore, the extensive production of virus-free garlic has not resulted in increased fertility of bolting clones, although it could be argued that the methods to produce virus-free garlic may not be expected to also remove rickettsia-like organisms. Until further substantiation is provided, a pathogenic basis for floral abortion and male sterility in garlic seems unlikely. While some karyotypic variation was noted in mitosis of diverse garlic clones, evaluations of meiosis indicate an interesting general difference between clones collected in Central Asia (including those from the wild), east of that region, or west of it (which we refer to as Central, Eastern, and Western clones, respectively) that relate to fertility and perhaps floral abortion (Table 5.2). Several authors have observed quadrivalents, hexavalents, octavalents, and decavalents in microsporogenesis of numerous diverse bolting Eastern garlic clones that were all male sterile, while only a small number of Eastern clones, notably those from India, had all bivalents in the first meiotic division. These Eastern clones with bivalents may represent the subtropical group noted by Maaß and Klaas (1995) to differ significantly from other Eastern garlic based on

224 Table 5.2.

Reports of garlic flowering, meiosis, and male fertility. No. clones evaluatedz

Reference

Meioticy configuration

No. clones flowering

Takenaka 1931 Katayama 1936

1E 3E

1 3

Krivenko 1938 Kononkov 1953 Takenaka 1953 Shimada and Shozaki 1954 Gvaladze 1961

1C 1C 1E 1E

No. malefertile clones

1 1 1 1

II and multivalents 8 II (1 clone); 1 VI & 5 II (2 clones) 8 II — IV, VI, VIII + II —

1 1 0 0

1C

1

variable

Partial

Gvaladze 1965

—

—

—

Koul and Gohil 1970 Gohil and Koul 1971 Konvicka and Levan 1972 Konvicka 1972 Novak 1972

3 E? (not specified) 1 E (India)

3

8 II

50–60% fertile pollen; 1–3% normal ovules 0

1

Desynapsis (I)

0

4W

3

0

1W 1W+1 A. longicuspis

1 2

8 II in most meiocytes; IV in one clone 8 II 8 II

Remarks

0 0

0 0

Climate limits seed production Seed produced Better flower retention with bulbil removal Better flower retention with bulbil removal; suggested aposporic origins of seed

Normal meiosis, but microspores degenerate

Post-meiotic tapetal degeneration

Konvicka 1973

14 W

14

8 II predominate

Novak and Havranek 1975

2 W+ 1 A. longicuspis

3

8 II

Konvicka et al. 1978

3W

3

8 II

Etoh and Ogura 1978 Koul et al. 1979

1E

1

5 E (India)

2 or more

VIII + 4 II usually; never 8 II 8 II

Etoh 1979 Cheng 1982 Etoh 1983a

43 (39 E + 4W) 1 28 (25 E + 2 W + 1 C)

43 1 28

Etoh 1983b

153

72

Etoh 1985

154

71 complete 57 partial

VI, VIII, or X + II II 8 II in 2 clones (1 W + 1 C); VI or VIII + II in 25 sterile 8 II in fertile clones; VI, VIII, or X + II in sterile clones 8 II in fertile clones; VI, VIII, or X + II in sterile clones

0 (with no antibiotics) to 14 (with antibiotics) 0

0 (with no antibiotics) to 3 (with antibiotics) 0 0

0 0 1C

1C

1C

Postmeiotic disturbances alleviated with application of tetracycline Binucleate pollen not viable; better flower retention with bulbil removal and cut stems Postmeiotic disturbances; better flower retention with bulbil removal

2, 4 - D and gamma-irradiation do not improve flower retention; better flower retention with bulbil removal Bulbils compete with flowers Male fertile clone from Moscow; fertile flowers have violet anthers Seed produced

Bulbil number correlated with flower number; partial bolting clones failed to retain flowers to meiosis in 50 of 57 clones

225 (continued)

226 Table 5.2.

Continued

Reference

No. clones evaluatedz

Meioticy configuration

No. clones flowering

No. malefertile clones

Etoh 1986

31C

31

8 II

14

Bozzini 1991 Etoh et al. 1992 Pooler and Simon 1994 Hong and Etoh 1996 Hong et al. 2000a

1 several 210 W and C

1

—

137

8 II in 29 examined

1 several 27

42 C

42

8 II in 39/42

31/35

30 W + 30 C

0 W (all partial bolting) + 29 C 38

Not reported

9 (all C)

8 II in 5 of 5

15 (14 C + 1 E)

Etoh et al. 2001 z

38 (33C + 4W + 1 E)

Remarks Fertile flowers have purple anthers; several male sterile, female fertile clones Tetraploid, not garlic?

Fertile flowers have purple anthers

Clones collected in Central Asia (C), East of center (E), and west of center (W). I, II, III, IV, VI, VIII, and X refer to univalents, bivalents, trivalents and associations of 4, 6, 8, and 10 chromosomes at meiosis I, respectively.

y

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

227

molecular “fingerprinting.” The multivalent associations in Eastern garlic north of India indicate a clear basis for their male sterility, since gametes that result will have unbalanced chromosome sets. In contrast, Central and Western garlic clones usually form only bivalents, with multivalents rarely noted, although desynapsis has been noted in some clones. Yet Western garlic clones are almost always male sterile, while Central clones are often fertile. Post-meiotic tapetal degeneration is observed in those Western clones closely analyzed (Novak 1972; Gori and Ferri 1982; Gori 1983), as well as in sterile Eastern clones (Etoh 1982), but the cause of degeneration is not known. Genetic and cytoplasmic male sterility in seed-propagated crops is only beginning to be understood after much effort, and a similar mechanism accounting for male sterility in Western garlic may be uncovered. A fourth contributor to impaired garlic floral development in some studies points to unsuccessful floral competition with bulbils. Katarzhin and Katarzhin (1982) and Etoh (1983b, 1997) produced garlic seed without bulbil removal, but Konvicka (1984), Pooler and Simon (1994), and Jenderek (1998) removed bulbils during floral development to improve development of flowers and seed, although it should be noted that this was only successful for clones with some male fertility without this procedure. It is reasoned that developing flowers are relatively poor competitors with bulbils for photosynthate and, at least in some cases, this competition results in floral abortion. Verification of this hypothesis has not yet been forthcoming. Konvicka (1984) and Pooler and Simon (1994) carried this notion a step further and decapitated garlic plants just above the pseudostem around the time when the scapes were nearly fully elongated. These cut stems were then placed in containers of water (with and without antibiotics, respectively) for pollination and seed development, with good success. The rationale for this treatment is that developing flowers also compete poorly with the developing bulb. Experiments demonstrating a clear, positive effect from this treatment (versus seed production on intact plants) have not been reported. V. GARLIC SEED PRODUCTION A. Importance of Garlic Germplasm for Seed Production One recurring observation made in all recent studies seeking garlic flowering and seed formation is the need to include a diverse range of garlic germplasm with an emphasis on Central Asian cultivated clones and Allium longicuspis (reviewed by Etoh and Simon 2002) (Table 5.3). The

228

Table 5.3.

Reports of garlic seed production. No. Clones

Reference Kononkov 1953 Gvaladze 1961 Katarzhin and Katarzhin 1978 Katarzhin and Katarzhin 1982

Etoh 1983b Konvicka 1984 Etoh 1985 Etoh et al. 1988 Pooler and Simon 1994 Inaba et al. 1995 Hong and Etoh 1996 Etoh 1997 Jenderek 1998 Hong et al. 2000c

No. seeds produced

Germination rate (%)

Evaluated

Flowering

Producing seed

Several Several 120

+ + +

1 1 >1

1 1 —

1 1 —

Several

—

4

—

—

25.90 41.50 36.80

153 7 2

72 7 2

1 7 2

3 to 83 9.60

17 210

17 137

17 11

80 .0

—

—

—

0.01

42

39

17

12.30 0 to 100 up to 93 0.02

17 10.0 421 49 original + 124 previously derived from seed

17 10 — —

17 10 64 —

27 41 200 harvested, 85 mature 22,850 655 52,508 346 3481 (1984) >7000 (1985) >2M >3500

Remarks Bulbil removal Two generations produced in field conditions The first seed generation was without bulbil removal; two generations of seeds were produced Bulbil removal Seed production from cross-pollination

Four cycles of selection, up to 248 seeds/plant All seed from selfpollination Multiple pollination Up to 656 seeds/plant Inclement weather limited seed production

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

229

first three reports of garlic seed production came from research in the Soviet Union apparently using germplasm from Central Asia and neighboring regions (Kononkov 1953; Gvaladze 1961; Katarzhin and Katarzhin 1978, 1982). Etoh and co-workers (Etoh 1983b, 1985, 1997; Etoh et al. 1988; Inaba et al. 1995; Hong and Etoh 1996; Hong et al. 2000c) included Central Asian germplasm in their materials, which resulted in successful production of seed. Konvicka (1984) used a diverse germplasm collection in his successful production of garlic seed. Pooler and Simon (1994) found especially good seed production with recently collected wild garlic from Central Asia. Jenderek (1998) also utilized a very diverse germplasm collection of garlic originating from 41 countries, including many from the former USSR, for highly successful seed production. As Central Asia is the center of garlic germplasm diversity and only area where wild garlic or Allium longicuspis still occurs, it seems likely that the broad genetic variation observed in visible characters as well as molecular assessments likely contributed significantly to successful seed production. In fact, we may infer that sexual reproduction in wild, and perhaps cultivated, Central Asian garlic has served as the reservoir of recombination to generate most of the new genetic diversity of this crop. As migrating hunters and gatherers, and more recently travelers followed major migration and trade routes through Central Asia, it is easy to envision them collecting and carrying wild garlic for cultivation far from that region. This process likely continues today. Based on molecular fingerprinting, it does not appear that newly collected garlic from Central Asia represents a different basic gene pool from clones long propagated in cultivation and collected around the world. Therefore, we may speculate that long-term cultivation and even selection has brought an accumulation of genetic changes that limit the ability of garlic to flower and produce seeds. Long-term asexual propagation is well known to accumulate structural and numerical chromosome aberrations (Sharma 1956), which appears to be the likely mechanism for male sterility in Eastern garlic. But other genomic changes that affect gene expression, such as altered patterns of DNA methylation, may account for male sterility in Western garlic. Experiments comparing clonally propagated garlic with true seedlings may lead to a better understanding of the mechanisms contributing to the success or failure of garlic flowering, fertility, and seed production. B. Processes and Procedures for Garlic Seed Production With the discovery of several sources of male fertile garlic, the prospects for seed production and breeding seemed likely. Yet while the

230

P. SIMON AND M. JENDEREK

development of fertile flowers in garlic is necessary for seed production and breeding, it is not sufficient. Once fertile clones are identified in diverse germplasm, flowers must be pollinated and seed produced; seed must be harvested, stored and germinated; and seedlings must be grown to plants to move forward in a breeding program. The extensive research contributions of T. Etoh in seeking male fertility and successful seed production in garlic are particularly notable in their breadth and depth. 1. Pollination and Seed Production. Optimal plant growth conditions contribute to successful garlic seed production. Fall planting in a climate that induces bolting is essential. Once bolting plants produce mature flowers in the spring or summer, they must be prepared for pollination. In more optimal climates, garlic seed production is most straightforward on plants growing in the field since this is where optimal growth occurs. This requires situating plants at planting time to accommodate pollinating cages, if they will be placed over flowering plants the following spring. Alternatively, in less optimal climates, field-grown plants can be used as a source of cut stems to perform seed production at a different site. This practice allows bulbil removal and pollination to be performed outside the field, for example, in a laboratory or air-conditioned greenhouse. Garlic cut stems are prepared by severing the scape just above the pseudostem when the scape is fully, or almost, straightened and usually after bulbil removal. Cutting the stem too early can impede full development of the inflorescence. Adequate photosynthate is apparently provided by the scape, as seed production capacity does not appear to be limited by this procedure. Production of garlic seed on cut stems is more labor-intensive than seed production on plants in the field, and consequently would not be practical for commercial large-scale production of seed, but the flexibility it provides will likely recommend its continued use in small-scale breeding programs, or in making initial crosses in larger programs. Bulbil removal is very conducive to prevention of early senescence of flowers (Fig. 5.3), but it is not required for all clones, especially in the second and subsequent generations of seed production. Bulbil removal is a time-intensive and tedious exercise that involves opening the spathe of inflorescences while bulbils are no larger than 2 to 5 mm in diameter, followed by extirpation of all visible bulbils. Care is required in finding all bulbils while not damaging the developing flowers, which are smaller than the bulbils and much more fragile. After the initial removal of bulbils, a second removal 3 to 7 days later is useful to extirpate bulbils that typically remain.

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

231

Fig. 5.3. Garlic flower development. Early (a and b) and late (c and d) flowering in the same clone where bulbils have been removed (a and c) or retained (b and d). Fertilized ovules swelling late in the season for clones with few (e) and many (f) flowers.

232

P. SIMON AND M. JENDEREK

The most promising flowers for successful seed production usually have pigmented (purple or red) anthers and are male fertile (Etoh 1983a; Konvicka 1984; Etoh 1986; Etoh et al. 1988; Pooler and Simon 1994; Hong and Etoh 1996), although some male fertile plants with yellow anthers do occur and such plants produce seeds (Pooler and Simon 1994; Jenderek 1998). Furthermore, male sterile plants can still be female fertile and set seed. This fact creates an opportunity to produce F1 hybrid garlic seed by isolating a male sterile/female fertile clone with a fully fertile clone during pollination. Pollen stainability and germination varied widely among those garlic clones first used for seed production. For example, Hong and Etoh (1996) and Jenderek and Hannan (2000) observed pollen stainability from 0% up to 98.3%. Pollen germination rate is generally lower than stainability. Pooler and Simon (1994) found from 0% to 10.5% germination, while 0.7% to 49.7% was reported by Hong and Etoh (1996) and Jenderek and Hannan (2000). Although male fertility is associated mainly with purple anthers, stainable pollen in yellow anthers was as high as 68.6% with pollen germination rates of 30.1% in the latter study. To ensure garlic seed production, several authors have used multiple pollination (repeated hand pollination of the same flowers over several days) (Etoh 1983b; Pooler and Simon 1994; Etoh 1997). It is not clear that this practice always improves the seed set. Several pollination vectors have been used to produce garlic seed. Collection of anthers or pollen with a forceps, on a glass slide, or small brush has been used on a small scale, usually with repeated pollen application to the same flowers over several days. Emasculation has been used to eliminate self-pollination, but it is a tedious exercise. Honey bees (Apis mellifera), houseflies (Musca domestica), leaf cutter bees (Megachile rotundata), and bluebottle flies (Protophormia terraenovae) have been used successfully as pollinators in isolation cages. As the style of garlic flowers is small and delicate (Fig. 5.1), seed production in some clones suffers from rough handling and honeybee pollination. Success with bluebottle fly pollination has been very good, and this is the main pollination vector currently used in much garlic true seed production. Diseases, pests, and weather can take a significant toll on garlic pollination and seed formation. In particular, floral pests such as thrips and mites can quickly decimate recently pollinated flowers and lead to crop failure. Inclement weather such as heavy wind and rain can also interfere with pollination and damage plants in the field (e.g., Etoh et al. 1988). A serious threat to true seed production occurs with hot weather (over 30°C) persisting for several days during anthesis when pollen via-

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

233

bility declines rapidly, flowers wither, and flower stalks senesce rapidly. Heat sensitivity varies widely among diverse clones. 2. Seed Harvest, Storage, and Germination. Garlic seed harvest is similar to onion seed harvest. Dried inflorescences are crushed to release seeds, and seeds are separated from chaff. Small seeds are frequently inviable and often occur with great regularity, especially in early generation materials. All seeds are planted in hopes of germination, but as selection for greater viability and seedling vigor proceeds in advanced generations, procedures such as seed separation in a wind column are used to remove chaff and light seed and leave heavier, more vigorous seed. The number of seeds produced by an inflorescence depends on several variables, including genotype and growing conditions. Clonal planting material produced up to 50 seeds/umbel (Etoh et al. 1988; Jenderek and Hannan 2002), whereas maternal plants derived from true seeds produced 656 seeds/umbel (Etoh et al. 1988; Jenderek 1998). Garlic seeds are approximately half the size of onion seeds, resembling them in shape and color (Fig. 5.4). The size of true seeds in garlic depends

Fig. 5.4.

Garlic (above) and onion seeds. Bar = 1 mm.

234

P. SIMON AND M. JENDEREK

on the genotype of parental lines. Early publications documenting the ability of certain garlic clones to set a limited number of seeds inspired a series of experiments evaluating conditions and treatments that stimulate seed germination, including hormone treatments, stratification, scarification, and moist chilling. Hormone treatments proved ineffective (Etoh 1983b; Etoh et al. 1988; Inaba et al. 1995). In vitro seed germination allows for easy evaluation of germinating seeds, but requires high labor input with no apparent improvement in germination rate (Pooler and Simon 1994). Like many Allium species, garlic has some seed dormancy, which is reduced by cold treatments. Germination rates in clones that had been long propagated asexually ranged from 10% to 35% with some seeds still beginning to germinate 12 months after onset of germination conditions. Breeding efforts have increased germination rates to at least 65%, and up to 100%. 3. Growing Seedlings to Plants. Vigorous growth of garlic seedlings is difficult to achieve in early generations of seed production. Seedling growth is often weak and slow, so optimal growing medium, moisture, and freedom from pests and diseases is essential. Like onion and leek, garlic seed germinates epigeally with the bent cotyledon first to emerge. As with plant development from cloves or bulbils, leaves arise from the apical meristem to generate 5 to 15 leaves, depending on genotype, temperature, and photoperiod. The time from germination to 3 true leaves can take as long as 4 months. Pseudostem development is not very prominent in small seedlings. If seedlings grow large enough, flowering can occur in the first year in optimal growth conditions with the best genetic selections. Usually only one scape and inflorescence forms in flowering garlic, and that in the second year, however 2–3 stalks per plant may develop. A broad range of phenotypic variation is evident in bulbs that develop from true seed (Fig. 5.5). Garlic seedling vigor and consequently survival usually increases with progressing seed cycles. Populations of all early generation seed derived plants carry, to various degrees, unfavorable characteristics, such as stunted or aberrant growth, or chlorophyll deficiencies. In fact, plants in these early generations can manifest unfavorable growth even in more mature plants. These can include deformed leaves, stunted roots, and limited bulb production (although the latter trait could be a photoperiod response). The appearance of these traits is more frequent in progenies derived from self-pollination (Jenderek 2002), but surviving seedlings and subsequent generations are usually more vigorous in growth, and often they surpass the vigor of plants grown from clonal material.

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

Fig. 5.5.

235

Garlic bulbs from true seed.

VI. PROGRESS IN GARLIC BREEDING AND FUTURE PROSPECTS Before the development of techniques to produce garlic seed in relatively large quantities, garlic breeding was not a realistic possibility. In fact, there is no evidence indicating that sexual reproduction and selection were ever utilized by garlic growers throughout history so that, while garlic has been one of the longest cultivated horticultural crops, breeding the crop has just begun. Clonal selection has been successful in altering some traits in garlic such as clove number and earliness (Burba 1997) and routine treatments to reduce or eliminate viruses clearly improve production (van Dijk 1994; Verbeek et al. 1995; Salomon 2002), but without sexual reproduction, desired traits found in different clones cannot be combined. With the possibility of seed production, garlic breeding can commence. Information on the current status of true seed derived plants is scarce, as most of the large-scale seed production has been performed by private industry. The United States Patent and Trade Office lists three seed derived cultivars to be propagated by cloves. The area in cultivation utilizing those cultivars is not reported.

236

P. SIMON AND M. JENDEREK

A. Selection for Improved Floral Characteristics and Fertility Although the asexually propagated bolting garlic clones used worldwide typically produce inflorescences with predominant bulbils and completely or nearly sterile flowers, selection for improved floral characteristics and fertility has generally been successful. Reduced numbers of small, easy-to-remove bulbils is typically very rapid, so that after two or three generations of seed propagation, bulbil removal is not necessary in some breeding stocks, although they usually are still present. It may, in fact, be possible to breed for bulbil-free strains in some genetic backgrounds. Selection progress for improved male fertility is also rapid, with high levels of pollen stainability common in the first generation progeny of sexual reproduction. It is presumed that the cleansing effect of meiosis to generate balanced gametes immediately results in meiotic stability and female fertility, although published data are lacking. There is cytological evidence that demonstrates that translocations and inversions occur in many Eastern garlic clones, so one can speculate that some fertile seed progeny of these clones may harbor homozygous translocations or inversions that have no effect upon selfing or intercrossing with similar strains. However, upon crossing with strains lacking these aberrations, sterile translocation or inversion heterozygotes would result. No observations of this type have been published to date. However, several molecular markers have been identified to assist in selection for male fertility so that fertility may be predicted long before a plant flowers (Etoh 1985; Hong et al. 1997, 2000b). B. Selection for Improved Seed Size and Vigor Selection for improved seed vigor and size can be dramatic after two or three cycles. Sizable variation in seed size has been noted in early generation seed progeny, and breeding efforts to date have succeeded in selecting for seed size ranging from 339 to 496 seeds/gram (Jenderek 1998). After two to three generations of selection for improved seed germination, germination rates of 65% to 93% have been routinely observed (Inaba et al. 1995; Jenderek 1998), a dramatic improvement over the low rates of 10% to 35% observed in first generation seed. Like any plant breeding effort, selection is not successful in all genetic stocks. Selection efforts to date have proved that the yield of true seeds could be improved significantly. The possible number of seeds produced by one inflorescence can be up to 1200 seeds in umbels with 200 flowers, based on the assumption that one flower has a potential of developing six seeds.

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

237

C. Garlic Breeding Goals A very wide range of phenotypic variation has been observed in true garlic seed progenies, including variation in all of the characters discussed above in describing germplasm variation. Perhaps most interesting is the observation of non-bolting seedlings, since both parents were obviously bolting plants. These are unusual and no reports have described these segregants beyond first- and second-year field observations. In fact, general descriptions of genetic segregation for any trait in garlic are lacking. Most of the variation described has been for some of the typical bulb and flower traits, including outer and inner bulb scale color, clove number and color, time of bulb maturity, time of flowering, stalk length, and flower petal color, pedicel color, number, and ovary wall color. Given the range of flowering habits observed among diverse garlic clones, a better understanding of flowering genetics will be of particular interest. Of most interest to garlic growers are yield and bulb storage parameters, but no reports to date have included data detailing these attributes. A search for genetic resistance to viruses and other diseases ranks high among garlic breeding goals. Another very critical trait, likely associated closely with yield, is seedling vigor. This was a crucial variable contributing to the initial successes in obtaining garlic seed, and it will also be critical for success in developing seed-propagated garlic. Since clonally propagated garlic typically develops from a clove several grams in size, and a seedling has a weight equivalent to only a small fraction of a clove, early seedling vigor will be essential for crop propagation from seed. Perhaps other technologies such as transplanting of seedlings, rather than direct seeding, will also be used to make seed-propagated garlic a viable economic possibility, although this would add back another production cost and hence reduce the economic benefits of a seed-propagated crop. Field trails to evaluate the possibilities for direct seeding are underway and they will apply intense selection pressure to test the feasibility of seed-propagated garlic. D. Garlic Breeding Methods Initial efforts to produce garlic seed included only a very few fertile clones that were interpollinated and then progeny were either interpollinated again or self-pollinated. As garlic breeding proceeds, there are two distinct directions that can be taken: development of new clones for asexual propagation, or development of seed-propagated garlic cultivars. The development of new garlic clones for asexual propagation will follow the same process of evaluation and utilization as is currently being

238

P. SIMON AND M. JENDEREK

used to evaluate new asexually propagated clones that have been acquired from abroad and are being tested for performance. In this system, individual seedlings will be selected based upon superior performance to be used as new clonally propagated cultivars using the same production techniques as have been used until now. The process of adequately testing a new seedling will take several years, since the increase of a single plant to produce an adequate supply of bulbs is necessary to perform replicated field trials, storage trials, and evaluation of added-value characteristics such as phytonutrient content or soluble solids. New cultivars developed this way must fit all existing parameters to meet grower and processor needs, but also exhibit enough superior traits or unique new combinations of traits to warrant considering replacing existing cultivars with them. New clonally propagated cultivars have likely been gradually replacing older ones in Central Asia throughout history, although this process has never been documented. An important question, which has not yet been answered, is whether new clones from seedlings can out-compete existing clonally propagated materials. Preliminary observations seem to indicate this will occur, but the superior traits offered by new clones selected from seedlings, and the number of existing cultivars that will be replaced is yet unknown. Likely, the large-scale production and processing industry will utilize new clonally propagated cultivars, but smaller-scale growers can benefit to the same extent in improving their operations. The new complication that garlic clones derived from seedlings bring to growers is a potentially huge influx of new clones to be tested. Increase of materials and proper testing over several years for large numbers of seedlings is a monumental task, so new clones from seedlings will likely require striking improvements over existing cultivars to warrant such an effort. The development of seed-propagated garlic cultivars is only in its very initial stages and no clear indication of its likelihood of success is possible for several years. The possibility that seed-propagated garlic will have low or no virus contamination is a strong incentive for true seed propagation. The combination of tissue culture to remove viruses, plus “seed” garlic field production costs exceed $2500/ha. If the costs of $300/ha for hybrid onion seed are an indication of how expensive garlic seed may someday be, then the economic incentive for true garlic seed is apparent. Inventory management, storage, and transportation costs also favor garlic seed over clonal propagation. Garlic breeding methods to be used include the usual recurrent phenotypic selection from (at least initially) a broad genetic base of materi-

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

239

als. As self-pollination has been demonstrated to be possible and to result in reasonably vigorous and more uniform families, it seems likely that cultivar development strategies for seed-propagated garlic will include an emphasis on inbred development for hybrid cultivars, like onions. Since male sterile/female fertile clones have been well documented, these clones are an obvious choice for seed parents. As for other outcrossing crop species, hybrid cultivars not only can take advantage of hybrid vigor (which has been widely observed in garlic but not reported), but they also encourage growers to return to seed producers for seed in subsequent production years. With the large investment involved in developing seed-propagated garlic, there is little incentive for large-scale seed producers to develop open-pollinated cultivars that could be seed propagated by the grower, although public sector programs may develop open-pollinated seed-propagated garlic cultivars.

VII. CONCLUSIONS Garlic is a widely recognized and appreciated crop with a long history of asexual propagation. Several inherent aspects of garlic growth and development combined with artifacts of its long asexual reproduction have resulted in a crop in which many clones do not flower, those flowering are nearly or completely sterile, bulbils usually suppress flower maturation, and first-generation seedlings are weak with a high incidence of abnormalities limiting normal growth and development. In spite of these facts, observations and experiments of the last 50 years, and especially the efforts of T. Etoh in the last 20 years, made it apparent that the production of true garlic seed is possible. Access to a diverse range of germplasm, particularly that from near its center of origin, combined with careful application of procedures to enhance seed production and growth, such as bulbil removal and careful seedling husbandry, has set the stage for true garlic seed production of the crop. Thus, the advantages that sexual reproduction brings in generating a balanced genome and combining traits from two unrelated parents could be captured. Utilizing these materials, methods, and meiosis, a small level of success in garlic seed production was realized. Taking advantage of the benefits of these breakthrough efforts, millions of garlic seeds have been generated in the last decade, and garlic breeding is underway. The potential for combining traits of diverse materials to develop new genotypes is only in its infancy, but much genetic variation is apparent and field testing of a seed-produced garlic crop is underway.

240

P. SIMON AND M. JENDEREK

LITERATURE CITED Al-Zahim, M., H. J. Newbury, and B. V. Ford-Lloyd. 1997. Classification of genetic variation in garlic (Allium sativum L.) revealed by RAPD. HortScience 32:1102–1104. Aoba, T. 1966. Studies on bulb formation of garlic plants (in Japanese). J. Japan. Soc. Hort. Sci. 35:284. Battaglia, E. 1963. Mutazione cromosomica e cariotipo fondamentale in Allium sativum L. Caryologia 16:1–46. Bozzini, A. 1991. Discovery of an Italian fertile tetraploid line of garlic. Econ. Bot. 45:436–438. Bradley, K. F., M. A. Rieger, and G. G. Collins. 1996. Classification of Australian garlic cultivars by DNA fingerprinting. Austral. J. Expt. Agr. 36:613–618. Burba, J. L. 1997. Obtencion de nuevas culivares de ajo. 50 Temas Sobre Prod. de Ajo 2:49–53. Cheng, S. S. 1982. Sexual process in garlic (Allium sativum L. cv. ‘Chonan’). Proc. Trop. Region Am. Soc. Hort. Sci. 25:69–72. Darlington, C. D. 1939. The evolution of genetic systems. Univ. Press, Cambridge. Engeland, R. L. 1991. Growing great garlic. Filaree Productions, Okanogan, WA. Etoh, T. 1979. Variation of chromosome pairings in various clones of garlic, Allium sativum L. Mem. Fac. Agr. Kagoshima Univ. 15:63–72. Etoh, T. 1980. An attempt to obtain binucleate pollen of garlic, Allium sativum L. Mem. Fac. Agr. Kagoshima Univ. 16:65–73. Etoh, T. 1982. Development and degeneration of the tapetum in garlic, Allium sativum L. Mem. Fac. Agr. Kagoshima Univ. 18:75–84. Etoh, T. 1983a. Accomplishment of microsporogenesis in a garlic clone. Mem. Fac. Agr. Kagoshima Univ. 19:55–63. Etoh, T. 1983b. Germination of seeds obtained from a clone of garlic, Allium sativum L. Proc. Japan Acad., ser. B. 59:83–87. Etoh, T. 1985. Studies on the sterility on garlic, Allium sativum L. Mem. Fac. Agr. Kagoshima Univ. 21:77–132. Etoh, T. 1986. Fertility of the garlic clones collected in Soviet Central Asia. J. Japan. Soc. Hort. Sci. 55:312–319. Etoh, T. 1997. True seeds in garlic. Acta Hort. 433:247–255. Etoh, T., J. Johjima, and N. Matsuzoe. 1992. Fertile garlic clones collected in Caucasia. p. 49–54. In: P. Hanelt, K. Hammer, and H. Knupffer (eds.), The genus Allium: Taxonomic problems and genetic resources. JPK, Gatersleben, Germany. Etoh, T., E-R-J. Keller, and A. Senula. 2001. Fertile garlic clones in the Gatersleben collection. Mem. Fac. Agr. Kagoshima Univ. 37:29–35. Etoh, T., Y. Noma, Y. Nishitarumizu, and T. Wakamoto. 1988. Seed productivity and germinability of various garlic clones collected in Soviet Central Asia. Mem. Fac. Agr. Kagoshima Univ. 24:129–139. Etoh, T., and H. Ogura. 1977. A morphological observation on the formation of abnormal flowers in garlic (Allium sativum L.). Mem. Fac. Agr. Kagoshima Univ. 13:77–88. Etoh, T., and H. Ogura. 1978. Multivalent chromosomes in garlic, Allium sativum L. Mem. Fac. Agr. Kagoshima Univ. 14:53–59. Etoh, T., and H. Ogura. 1981. Peroxidase isozymes in the leaves of various clones of garlic, Allium sativum L. Mem. Fac. Agr. Kagoshima Univ. 17:71–77. Etoh, T., and P. W. Simon. 2002. Diversity, fertility, and seed production of garlic. p. 101–117. In: H. Rabinowitch and L. Currah (eds.), Allium crop science: Recent advances. CABI Publ., New York.

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

241

Etoh, T., H. Watanabe, and S. Iwai. 2001. RAPD variation of garlic clones in the center of origin and the westernmost area of distribution. Mem. Fac. Agr. Kagoshima Univ. 37:21–27. FAO Yearbook Prod. Stats. 1961–2001, Rome. Fritsch, R. M., and N. Friesen. 2002. Evolution, domestication, and taxonomy. p. 5–30. In: H. Rabinowitch and L. Currah (eds.), Allium crop science: Recent advances. CABI Publ., New York. Garcia Lampasona, S., L. Martinez, and J. L. Burba. 2002. Genetic diversity among selected garlic clones (Allium sativum L.) using AFLP (Amplified Fragment Length Polymorphism). Euphytica accepted for publication. Gohil, R. N., and A. K. Koul. 1971. Desynapsis in some diploid and polydiploid species of Allium. Can. J. Genet. Cytol. 13:723–728. Gori, O. 1983. Ultrastructural changes in the wall of the sporogenous cells in Allium sativum, clone Piemonte during microsporogenesis. Ann. Bot. 51:139–143. Gori, O., and S. Ferri. 1982. Ultrastructural study of the microspore development in Allium sativum clone ‘Piemonte’. J. Ultrastruct. Res. 79:341–349. Gvaladze, G. E. 1961. The embryology of the genus Allium L. (in Russian). Bul. Acad. Sci. Georgian SSR 26:193–200. Gvaladze, G. E. 1965. Vivipary and the capability of generative reproduction in Allium sativum L. Bul. Acad. Sci. Georgian SSR 40:412–427. Hahn, G. 1996. History, folk medicine, and legendary uses of garlic. p. 1–37. In: H. P. Koch and L. D. Lawson (eds.), Garlic: the science and therapeutic application of Allium sativum L. and related species, 2nd ed. Williams and Wilkins, Baltimore, MD. Hanelt, P. 1990. Taxonomy, evolution, and history. p. 1–26. In: H. D. Rabinowitch and J. L. Brewster (eds.), Onions and allied crops, I. botany, physiology, and genetics. CRC Press, Boca Raton, FL. Helm, J. 1956. Die zu Wurz- und Speisezwecken kultivierten Arten der Gattung Allium L. Kulturpflanze 4:130–180. Hong, C-J., and T. Etoh. 1996. Fertile clones of garlic (Allium sativum L.) abundant around the Tien Shan mountains. Breed. Sci. 46:349–353. Hong, C-J., T. Etoh, and S. Iwai. 2000c. An attempt of crossbreeding in garlic. Mem. Fac. Agr. Kagoshima Univ. 36:17–28. Hong, C-J., T. Etoh, B. Landry, and N. Matsuzoe. 1997. RAPD markers related to pollen fertility in garlic (Allium sativum L.). Breed. Sci. 47:359–362. Hong, C-J., H. Watanabe, T. Etoh, and S. Iwai. 2000a. Morphological and karyological comparison of garlic clones between the center of origin and westernmost area of distribution. Mem. Fac. Agr. Kagoshima Univ. 36:1–10. Hong, C-J., H. Watanabe, T. Etoh, and S. Iwai. 2000b. A search of pollen fertile clones in Iberian garlic by RAPD markers. Mem. Fac. Agr. Kagoshima Univ. 36:11–16. Hwang, J. M. 1993. Genetic divergence and classification of garlic cultivars by multivariate analysis (in Korean). J. Kor. Soc. Hort. Sci. 34:257–264. Inaba, A., T. Ujiie, and T. Etoh. 1995. Seed productivity and germinability of garlic. Breed. Sci. 45 (Suppl. 2):310. Ipek, I., A. Ipek, and P. W. Simon. 2003. Comparison of AFLPs, RAPDs, and isozymes for diversity assessment of garlic and detection of putative duplicates in germplasm collections. J. Am. Soc. Hort. Sci. (in press). Jenderek, M. M. 1998. Generative reproduction of garlic (Allium sativum) (in Polish). Zeszyty Naukowe Akademii Rolniczej im. H. Kollataja w Krakowie 57:141–145. Jenderek, M. M. 2002. Development of S1 families in garlic. XXVIth Int. Hort. Congr., Toronto, Canada, 433 (Abstr.)

242

P. SIMON AND M. JENDEREK

Jenderek, M. M., and R. M. Hannan. 2000. Seed producing ability of garlic (Allium sativum L.) clones from two public U.S. collections. Proc. Third Int. Symp. on Edible Alliaceae. Athens, GA, p. 73–75. Jenderek, M. M., and R. M. Hannan. 2002. Fertility, reproductive characteristics and true seed production of the USDA garlic germplasm collection. HortScience (in press). Jones, H. A., and L. K. Mann. 1963. Onions and their allies. Leonard Hill Books, London, UK. Kamenetsky, R., and D. H. Rabinowitch. 2001. Floral development in bolting garlic. Sexual Plant Reprod. 13:235–241. Kamenetsky, R., and D. H. Rabinowitch. 2002. Florigenesis. p. 31–57. In: H. Rabinowitch and L. Currah (eds.), Allium crop science: Recent advances. CABI Publ., New York. Katarzhin, M. S., and I. M. Katarzhin. 1978. Experiments on the sexual reproduction of garlic (in Russian). Byulleten’ Vsesoyuznogo Ordena Lenina I Ordena Druzhby Narodov Instituta Rastenievodstva Imeni N. I. Vavilova 80:74–76. Katarzhin, M. S., and I. M. Katarzhin. 1982. On generative reproduction of garlic (in Russian). Trudy po Prikladnoi Botanike, Genetike I Selektsii 72:135–136. Katayama, Y. 1936. Chromosome studies in some Alliums. J. College Agr. Imp. Univ. Tokyo 13:431–441. Kazakova, A. A. 1971. Most common onion species, their origin and intraspecific classification (in Russian). Trudy po Prikladnoi Botanike, Genetike I Selektsii 45:19–41. Kazakova, A. A. 1978. Allium (in Russian). In: D. D. Brezhnev (ed.), Flora of cultivated plants. Vol. 10. Kolos, Leningrad. Komissarov, V. A. 1964. Evolution of the cultivated garlic, A. sativum L. (in Russian). Izvestiya Timirjazevskoi Sel’ slokhozyaistvennoi Akademii 4:70–73. Komissarov, V. A. 1965. Classification of garlic, A. sativum L. (in Russian). Doklady Sel’ skohozyaistvennoi Akademii Timirjazev 108:351–357. Kononkov, P. F. 1953. The question of obtaining garlic seed (in Russian). Sad I Ogorod 8:38–40 Konvicka, O. 1972. Cytotaxonomische studien von vier sterilen Arten der Gattung Allium. Biol. Plant. 14:62–70. Konvicka, O. 1973. Die Ursachen der Sterilitat von Allium sativum L. Biol. Plant. 15:144–149. Konvicka, O. 1984. Generative Reproduktion von Knoblauch (Allium sativum). Allium Newsl. 1:28–37. Konvicka, O., and A. Levan. 1972. Chromosome studies in Allium sativum. Hereditas 72:12–148. Konvicka, O., F. Nienhaus, and G. Fischbeck. 1978. Untersuchungen uber die Ursachen der Pollensterilitat bei Allium sativum L. Z. Pflanzenzucht 80:265–276. Kothari, I. L., and J. J. Shah. 1974a. Structure and organization of shoot apex of Allium sativum L. Israel J. Bot. 23:216–222. Kothari, I. L., and J. J. Shah. 1974b. Histogenesis of seed stalk and inflorescence in garlic. Phytomorphology 24:42–48. Koul, A. K., and R. N. Gohil. 1970. Causes averting sexual reproduction in Allium sativum Linn. Cytologia 35:197–202. Koul, A. K., R. N. Gohil, and A. Langer. 1979. Prospects of breeding improved garlic in the light of its genetic and breeding systems. Euphytica 28:457–464. Krivenko, A. A. 1938. A cytological study of garlic (Allium sativum L.) (in Russian). Biol. Zhurnal 7:47–68. Kuznetsov, A. V. 1954. Cultivated garlic (in Russian). Selkhzgiz, Moscow. p. 1–119.

5. FLOWERING, SEED PRODUCTION, AND GENESIS OF GARLIC BREEDING

243

Lallemand, J., C. M. Messiaen, F. Briand, and T. Etoh. 1997. Delimitation of varietal groups on garlic (Allium sativum L.) by morphological, physiological, and biochemical characters. Acta Hort. 433:123–132. Maaß, H. I., and M. Klaas. 1995. Infraspecific differentiation of garlic (Allium sativum L.) by isozyme and RAPD markers. Theor. Appl. Genet. 91:89–97. Mann, L. K. 1952. Anatomy of the garlic bulb and factors affecting bulb development. Hilgardia 21:195–251. Mann, L. K., and P. A. Minges. 1958. Growth and bulbing of garlic (Allium sativum L.) in response to storage temperature of planting stocks, day length, and planting date. Hilgardia 27:385–419. Mathew, B. 1996. A review of Allium Section Allium. Royal Botanic Gardens Kew, Richmond, UK. McCollum, G. D. 1976. Onion and allies. p. 186–190. In: N. W. Simmonds (ed.), Evolution of crop plants. Longman, London, UK. Mensinkai, S. W. 1939. Cytogenetic studies in the genus Allium. J. Genet. 39:1–45. Messiaen, C. M., J. Cohat, J. P. Leroux, M. Pichon, and A. Beyries. 1993. Les Allium alimentaires reproduits par voie végétative. INRA, Paris. Novak, F. J. 1972. Tapetal development in the anthers of Allium sativum L. and Allium longicuspis Regel. Experientia 28:363–364. Novak, F. J., and P. Havranek. 1975. Attempts to overcome the sterility of common garlic (Allium sativum). Biol. Plant. 17:376–379. Pooler, M. R. 1991. Sexual reproduction in garlic (Allium sativum L.). Ph.D. thesis. Univ. of Wisconsin–Madison. Pooler, M. R., and P. W. Simon. 1993a. Characterization and classification of isozyme and morphological variation in a diverse collection of garlic clones. Euphytica 68:121–130. Pooler, M. R., and P. W. Simon. 1993b. Garlic flowering in response to clone, photoperiod, growth temperature, and cold storage. HortScience 28:1085–1086. Pooler, M. R., and P. W. Simon. 1994. True seed production in garlic. Sexual Plant Reprod. 7:282–286. Qu-Ying, H., H. Takagi, N. Ogasawara, and T. Etoh. 1994. Development of types of inflorescences of garlic and some Allium species (in Japanese). J. Japan. Soc. Hort. Sci. 63:121–130. Rahim, M. A., and R. Fordham. 1988. Effect of storage temperature on the initiation and development of garlic cloves (Allium sativum L.). Scientia Hort. 37:25–38. Salomon, R. 2002. Virus diseases in garlic and the propagation of virus-free plants. p. 311–327. In: H. Rabinowitch and L. Currah (eds.), Allium crop science: Recent advances. CABI Publ., New York. Senula, A., and R. J. Keller. 2000. Morphological characterisation of a garlic core collection and establishment of a virus-free in vitro genebank. Allium Improv. Newsl. 10:3–5. Sharma, A. K. 1956. A new concept of a means of speciation in plants. Caryologia 9:93–130. Shimada, T., and T. Shozaki. 1954. Studies on the breeding of garlic (Allium sativum L.). I. On the differentiation and growth of flowers, bulblets and bulbs. Agr. Bul. Saga Univ. 2:1–33. Siqueira, W. J., H. M. P. Filho, R. S. Lisboa, and J. B. Fornasier. 1985. Morphological and electrophoretic characterization of garlic clones. Bragantia 44:357–374. Takagi, H. 1990. Garlic Allium sativum L. p. 109–146. In: J. L. Brewster and H. D. Rabinowitch (eds.), Onions and allied crops, v. III. biochemistry, food science, and minor crops. CRC Press, Boca Raton, FL.

244

P. SIMON AND M. JENDEREK

Takenaka, Y. 1931. Further reports of the cytological investigations on the sterile plants (in Japanese). J. Chosen Natur. Histor. Soc. 12:25–41. Takenaka, Y. 1953. Ring formation in the meiosis of Allium scorodoprasum var. viviparum. Ann. Rep. Nat. Inst. Genet. 4:41–43. Tsuneyoshi, T., A. V. Nosov, Y. Kajimura, S. Sumi, and T. Etoh. 1992. RFLP analysis of the mtDNA in garlic cultivars. Japan. J. Breed. 42 (Suppl. 2):164–165. van Dijk, P. 1994. Virus diseases of Allium species and prospects for their control. Acta Hort. 358:299–306. Vanin, A. O. 1947. Why does garlic bolt (in Russian)? Sad i Ogorod 10:36–37. Verbeek, M., P. van Dijk, and P. M. A. van Well. 1995. Efficiency of eradication of four viruses from garlic (Allium sativum) by meristem-tip culture. Eur. J. Plant Path. 101:231–239. Verma, S. C., and R. K. Mittal. 1976. Chromosome variation in the common garlic, Allium sativum L. Cytologia 43:383–396. Vvedensky, A. I. 1935. Allium L. In: V. L. Komarov (ed.), Flora SSSR, Volume 4. Translated into English by H. K. Airy-Shaw as The genus Allium in the USSR. Herbertia 11:65–218 (1944). Weber, E. 1929. Entwicklungsgeschichtliche Untersuchungen uber die Gattung Allium. Bot. Arch. 25:1–44.

6 Cultivar Development of Ornamental Foliage Plants* Richard J. Henny and Jianjun Chen University of Florida, IFAS Mid-Florida Research and Education Center Apopka, Florida 32703

I. INTRODUCTION II. ORIGIN OF NEW CULTIVARS A. Plant Acquisition and Introduction 1. Plant Acquisition 2. Evaluation and Utilization of New Introductions B. Mutations from Vegetative Propagation 1. Sport Selection 2. Somaclonal Variant Selection C. Hybridization D. Transgene Technology E. Plant Patents III. BREEDING TECHNIQUES A. Control of Flowering B. Pollination Methods 1. Genera with Unisexual Flowers 2. Genera with Bisexual Flowers 3. Ferns C. Seed Handling D. Testing and Releasing New Cultivars

*The authors thank Dr. Adelheid R. Kuehnle at the University of Hawaii for providing published and unpublished information used in this review. They also thank Terri Mellich, Edmund Thralls, and Kelly Everitt for critically reading and editing the manuscript. This research was supported by the Florida Agricultural Experiment Station and approved for publication as Journal Series No. R-09116.

Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 245

246

R. HENNY AND J. CHEN

IV. BREEDING OBJECTIVES A. Phenotypic Traits 1. Foliar Variegation and Colors 2. Leaf Shape and Size; Petiole Colors 3. Plant Form 4. Flowers 5. Fragrance 6. Growth Rate B. Stress-related Traits 1. Adaptation to Interior Environments 2. Disease and Insect Resistance 3. Temperature Tolerance V. FOLIAGE EXAMPLES A. Aroids (Araceae) 1. Aglaonema 2. Alocasia 3. Anthurium 4. Dieffenbachia 5. Epipremnum 6. Philodendron 7. Spathiphyllum 8. Syngonium B. Bromeliads (Bromeliaceae) C. Calathea (Marantaceae) D. Dracaena (Dracaenaceae) E. Ferns (Polypodiaceae) F. Ficus (Moraceae) G. Hedera (Araliaceae) H. Palms (Arecaceae) VI. FUTURE PROSPECTS A. Germplasm Collection and Conservation B. Breeding C. Somaclonal Variant Selection D. Transgene Technology LITERATURE CITED

I. INTRODUCTION Defined literally, foliage plants would include all plants grown for their attractive leaves rather than flowers or fruits. In general horticultural terms, however, foliage plants are those with attractive foliage and/or flowers that are produced in containers in shaded greenhouses and used primarily as living specimens for interior decoration or interiorscaping (Chen et al. 2002). In common terminology, foliage plants are referred to as houseplants. Foliage plants in the tropics, however, may also grow under shade as landscape plants. Currently, more than 500 species with various forms, colors, textures, and styles are grown as foliage plants. The majority of them are indige-

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

247

nous to either the tropics or subtropics. The tropics have warm and stable temperatures, an abundant water supply, 12-hour daylength, and a continuous growing season. Tropical foliage plants grow under the tree canopy on the shaded forest floor or grow as lianas (climbing vines) or even live in trees as epiphytes. Characteristics of such foliage plants include their tolerance of low light, sensitivity to chilling temperatures, day neutral photoperiod response, and lack of dormancy. In the subtropics, summer is hot and humid but winter may have frost. Foliage plants originating from this climate are more adaptable to heat, drought, and chilling temperatures and may also show winter dormancy. Desert plants such as succulents and cacti have unique foliage, styles, and shapes that have been evolved for coping with heat and drought stresses. Few foliage plants are native to the temperate climate. English ivy (Hedera helix) is probably the most significant one whose juvenile form has been widely used as either potted or hanging-basket foliage plants. The use of plants indoors began at least 3,500 years ago when the Sumerians and ancient Egyptians started growing small trees in containers. The ancient Chinese expanded the variety of plants used for indoor decoration (Chase 1997). Wealthy merchants of Florence, Genoa, and Venice introduced plants from the East into Europe in the fifteenth century. Plant collectors in Holland and Belgium also imported plants from Asia Minor and the East Indies starting at the time of the Crusades. A desire for exotic plants developed among the aristocracy of France and England by the middle of the sixteenth century. Many wealthy persons of Europe constructed orangeries and conservatories in the seventeenth century. By the following century, an estimated 5,000 species of exotic plants had been brought into Europe (Smith and Scarborough 1981). In the second half of the nineteenth century, foliage plants were gaining popularity in Europe. The grand drawing rooms of Victorian houses had their fill of palms and ferns (Lowe 1861). In the late 1890s, foliage plants from conservatories, botanical gardens, and private estates were generally brought into commercial production, and these plants were sold for use in middle- and upper-class households. At the same period, shiploads of foliage plants from Europe were sold to greenhouse growers in the Northeast United States for continued growth or subsequent resale. In less than two decades, the production of foliage plants moved to California and Florida because of favorable climatic conditions. The Californian industry began with Kentia palm (Howea forsteriana) and pothos (Epipremnum aureum) in the 1920s, and later heartleaf vine (Philodendron scandens oxycardium) and Norfolk Island pine (Araucaria bidwilli) in the 1940s (Smith and Scarborough 1981). Florida growers have cultivated Boston fern (Nephrolepis exaltata) since 1912. Heartleaf vine was introduced to Florida in 1928 and

248

R. HENNY AND J. CHEN

Chinese evergreen (Aglaonema modestum) and rubber plant (Ficus elastica) in the 1930s (Smith and Scarborough 1981). Florida is now the leading state in the production of foliage plants, accounting for more than 55% of the national wholesale value since the 1960s. The steady increase in foliage plant production has been attributed in part to the back-to-earth, back-to-nature movement of the 1970s and continues unabated (Manaker 1997). The wholesale value of foliage plants in the United States rose from $29 million in 1969 to $585 million in 2001 (USDA 2002). Foliage plants enhance the interior environment and fulfill a psychological need by bringing beauty and comfort to our surroundings (Manaker 1997). Furthermore, foliage plants have been shown to be capable of purifying indoor air and have been demonstrated to remove 87% of pollutants such as formaldehyde and benzene from sealed chambers within 24 hours (Wolverton 1989). The beautification of interior environments and purification of indoor air have become key elements in promoting the use of foliage plants. Another important factor in the steady growth of the foliage plant industry is the continued introduction of new plants and development of new cultivars that dramatically expand options for foliage plant usage in interiorscaping (Chen et al. 2002). The aim of this chapter is to review several aspects relating to the development of foliage cultivars, including plant acquisition and introduction, selection of mutations, and interspecific hybridization. Four popular groups (aroids, bromeliads, ferns, and palms) are emphasized along with four other popular genera. Because much research is carried out by private companies, documentation in the scientific literature is uneven and, therefore, anecdotal information is included.

II. ORIGIN OF NEW CULTIVARS There are three main avenues for new foliage plant cultivars to enter the commercial trade: (1) plant acquisition and introduction, (2) selection of natural and induced mutations from established cultivars, and (3) hybridization and progeny selection. A. Plant Acquisition and Introduction Plant acquisition and introduction played important roles in the initial development of the foliage plant industry in the United States and will continue to be important in introducing new species and improving existing cultivars by providing germplasm for breeding. Until recently, neither the U.S. National Plant Germplasm System (NPGS) nor the Inter-

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

249

national Board for Plant Genetic Resources (IBPGR) had been involved in conservation of foliage plant germplasm. The Ornamental Plant Germplasm Center (OPGC), established in 1999 at the Ohio State University, is considering conserving only five aroid genera (Aglaonema, Anthurium, Dieffenbachia, Philodendron, and Spathiphyllum) at this time. Most foliage plant resources have been collected and maintained by private plant collectors or public institutions such as botanical gardens or conservatories. 1. Plant Acquisition. There are two avenues of obtaining new foliage germplasm: direct collection from the wild (usually done in conjunction with knowledgeable botanical garden personnel or avid private collectors) or acquisition of established material from botanic gardens or private collectors. Dr. Thomas B. Croat, one of the world’s leading experts on Araceae, spends months each year in the tropics as part of his work at the Missouri Botanical Garden in St. Louis. He has collected more than 10,000 living specimens and maintains the world’s largest and most comprehensive collection of living aroid plants in the Garden’s greenhouses. Croat’s work resulted in the revision of Anthurium, Dieffenbachia, and Syngonium (Croat 1982, 1983, 1986, 1997). Dr. Frank B. Brown of Valkaria Tropical Garden, Valkaria, Florida, made more than 50 explorations in the jungles of southeast Asia and brought valuable materials of Aglaonema (Brown 2001) to Florida. Calathea owes much of its popularity to collections from Central America (Kennedy 1973). Not all acquisitions involve jungle exploration. For example, Ficus elastica ‘Decora’ was introduced to Florida from a plant collected during a visit to Holland in 1954 (Griffith 1998a). Collected materials are often exchanged among members within individual plant societies such as the American Ivy Society, Bromeliad Society, and International Aroid Society, all based in Florida. Plant materials are also generously shared between private collectors and public institutions such as botanical gardens and universities. For example, foliage plant breeding programs at the University of Florida and University of Hawaii have received great benefit from valuable germplasm resources provided by both private collectors and botanical gardens, particularly the Missouri Botanical Gardens. However, maintaining tender tropical plants in greenhouses is expensive, and it is conceivable that changes in funding or individual research interests could result in a loss of plant diversity in these invaluable collections. 2. Evaluation and Utilization of New Introductions. Most newly collected foliage plants need to be systematically evaluated. This process may be an individual or joint effort between collectors, growers, and/or

250

R. HENNY AND J. CHEN

researchers. Evaluation of a newly collected plant prior to release includes taxonomic identification, methods of propagation, cultivation, and assessment of ornamental value. For example, an exotic plant with wide ovate-lanceolate leaves and orange-colored petioles was collected from Thailand and named Palisota by the collector. A later study by taxonomists determined it to be Chlorophytum orchidantheroids. Although this plant bears seeds, a local tissue culture firm successfully propagated it through tissue culture and provided a large number of liners for evaluation at the University of Florida. The plant was evaluated based on form, color, and style, as well as performance under interior conditions and was shown to be well adapted to a low light intensity of 8 µmol m–2 s–1 (Chen et al. unpublished). This unique plant was later named C. orchidantheroids ‘Fire Flash’ and is now produced as a new foliage plant. B. Mutations from Vegetative Propagation 1. Sport Selection. Mutant clones or sports have been widely used in foliage plant production as a source of new cultivars. Since most foliage plants are propagated vegetatively, spontaneous mutations may accumulate throughout consecutive generations, and offshoots or cuttings generated from the mutated cells may develop into mutant clones. In the literature, spontaneous somatic mutations are often called bud mutations, bud sports, or sports (van Harten 1998). These mutations may be either nuclear or cytoplasmic in origin and may affect an entire bud or other plant organ or only part of an organ, as in the case of chimeras or mosaics (Neilson-Jones 1969; Marcotrigiano 1997). The causes of somatic mutation could be the result of transposon activities, changes in chromosome number, or loss of genes (Pratt 1983; van Harten 1998). Somatic mutations are as common as or more common than germ-cell mutations at rates of 10–3 to 10–5 per locus per generation (Harrison and Fincham 1964; Simmonds 1979). Success in sport selection depends on the availability of highly variable stock plants and the quantity of propagules produced from the stock plants. For example, heartleaf philodendron (Philodendron scandens oxycardium) used to be the most commonly grown foliage plant in Florida from the 1950s to 1960s; millions of cuttings were produced each year. However, no single mutant was selected from the cuttings because this plant is genetically stable. On the other hand, English ivy (Hedera helix), another cutting propagated foliage plant, has more than 200 cultivars released; all of them were selected from sports (Rose 1996). Another factor that affects the success of a selected sport is its relative

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

251

stability in commercial production. Unstable sports are discarded if they fail to maintain anticipated phenotypes over time. 2. Somaclonal Variant Selection. In addition to sports commonly observed in those traditional vegetatively propagated foliage plants, another form of somatic variation or genetic instability occurs in tissue culture and has been termed somaclonal variation (Larkin and Scowcroft 1981). Compared to spontaneous somatic mutations, somaclonal variations usually occur at much higher frequencies (Buiatti and Gimelli 1993). The increased frequencies have been attributed to (a) pre-existing genetic variation in explant tissues, (b) genetic variation induced by mutagenic action of chemical compounds in the culture media, and (c) variation as a response of the plant genome to stress, including DNA methylation, gene amplication, and activities of transposons (Novak 1991). Since many foliage plants are propagated by tissue culture, selection of variants for desired phenotypes has become an important method of new cultivar development. Syngonium provides an excellent example of somaclonal variant selection as a cultivar development tool. The Syngonium pedigree in Fig. 6.1 shows how 22 cultivars, all somaclonal variants, were selected from large populations of tissue cultured material grown in commercial greenhouses. All 22 cultivars can be traced back to the original ‘White Butterfly’ clone. Each variant selected was micropropagated and each remained stable enough to become a named cultivar. Subsequent mutations resulted in additional cultivars. Several Cream Gold Emerald White Butterfly

Pink Allusion

Banana Cream Gold tetra Butterfly

Bob

Bold

(selfed) Berry

Regina Red

Maria Mango Julia Mini Bing Cherry

Red tetra Neon Pink Splash Peaches Red Butterfly

Crème de Menthe Hillary Exotic Sylvia

Avocado Red Plum Red Banana

Pink Tetra Pink Butterfly

Fig. 6.1. Pedigree of somaclonal mutant Syngonium cultivars from tissue culture that have become commercialized. All cultivars can be traced back to ‘White Butterfly’.

252

R. HENNY AND J. CHEN

of the cultivars have pink or reddish coloration in the foliage, which was not evident in ‘White Butterfly’. The first mutant, named ‘Pink Allusion’, was self-pollinated to generate a population that yielded a selection named ‘Regina Red’. When placed into tissue culture, ‘Regina Red’ produced five somaclonal variants of which ‘Pink Splash’ yielded ‘Pink Tetra’ and ‘Pink Butterfly’ (Fig. 6.1). This series of mutations also established that the pink trait evident in the original ‘Pink Allusion’ was genetically stable and could be transmitted through seed. Anthurium somaclonal variants in the Mid-Florida Research and Education Center (MREC) at the University of Florida include ‘Orange Hot’ (a purple-colored form has also been found and is under evaluation) from A. ‘Red Hot’. ‘Diamond Bay’ and ‘Emerald Bay’, two somaclonal variants of Aglaonema, were introduced in 2002 by MREC (Henny et al. 2003a). Many Dieffenbachia cultivars were also released from a selection of somaclonal variants from tissue-cultured populations. For example, Dieffenbachia ‘Tiki’ is a sport derived from D. ‘Memoria Corsii’, and ‘Snow Flake’, a new cultivar in the market, was derived from ‘Tiki’. Selected somaclonal variants, however, if unstable, could be detrimental to a cultivar because nonuniformity in crops becomes a costly liability to the clonal propagator. This was the case for Dieffenbachia hybrids ‘Starry Nights’ (Henny et al. 1989) and ‘Star White’ (Henny et al. 1992b), which were too unstable in tissue culture to be successful commercially. C. Hybridization Foliage plants are predominantly cross-pollinating species. Parents used in foliage plant hybridization are not usually derived from inbred, single-seed descent, or pedigree selection, because inbreeding depression limits development of inbred lines in most foliage plant genera. Traditional breeding through hybridization has focused on maximizing heterozygosity. By intercrossing distinct clones, both of which have desirable characters, populations are created that may be utilized directly for selection of new clones. If the parent clones are heterozygous, each seedling is a potential new cultivar that can be fixed by vegetative propagation. Interspecific hybridization is the most common practice in producing hybrid cultivars in foliage plant breeding. Interspecific hybridization offers opportunities for obtaining gene recombinations not possible with intraspecific hybridization and expands the range of genetic variability beyond that of a single species. Additionally, interspecific hybridization may create hybrids with unique ornamental characteristics by making

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

253

genetic combinations that could not be achieved through intraspecific hybridization. Interspecific hybrids in genera of aroids and bromeliads are discussed in Section V. With the advance of tissue culture, haploid plants can be produced through culture of excised anthers/pollen or ovaries/ovules. By doubling chromosome number, homozygous diploid plants can be procured in a single generation. The fertile homozygous plants can be used for producing inbred lines required to utilize hybrid vigor. Eeckhaut et al. (2001) initiated Spathiphyllum haploid production using in vitro ovary culture. Two doubled haploid genotypes, verified with flow cytometry and AFLP-patterns, were obtained from the cultivar ‘Stefanie’ (Eeckhaut et al. 2001). Such research is critical to develop true breeding parental lines that can be used to exploit heterosis of Spathiphyllum. D. Transgene Technology Transgene technology has been proven to be a powerful method of altering crop characteristics (Hansen and Wright 1999) and should be particularly useful in foliage plant improvement. Foliage plants are not edible and are valued by their esthetic appearance. Transgenic foliage plants would not cause genetic contamination of other crops because most are vegetatively propagated. However, application of transgene technology in foliage plants is quite limited. Anthurium is probably the only foliage plant being successfully transformed thus far (see details in Section VI). E. Plant Patents A total of 416 United States Plant Patents have been issued to 18 genera/groups of foliage plants since 1976 (Table 6.1). Among them, 66 patents were issued from 1976 through 1989 and 243 patents were issued in the 1990s. Current foliage plant patent activity appears to be increasing because 107 patents have been issued between January 2000 to August 2002. The bromeliad group, encompassing five genera (Aechmea, Cryptanthus, Guzmania, Neoregelia, and Viresea), was the most active, with 107 patents issued. Anthurium followed with 98, including several for cutflower cultivars. Spathiphyllum, Aglaonema, Dieffenbachia, Ficus, and Philodendron followed with 55, 35, 31, 25, and 16 patents, respectively. All 416 patents were reviewed to determine the method by which the cultivars were generated. Breeding analysis could be classified into four groups: (1) hybridization, (2) selection of naturally occurring sports,

254

R. HENNY AND J. CHEN

Table 6.1. A summary of all U.S. Plant Patents issued to 18 genera/groups of ornamental foliage plants from 1976 through August 10, 2002. Genus/Group Aglaonema Alocasia Anthurium Bromeliadz Calathea Chlorophytum Cordyline Dieffenbachia Dracaena Epipremnum Ferny Ficus Hedera Palmx Philodendron Shefflera Spathiphyllum Syngonium Totals

1976–1979

1980–1989

1990–1999

2000–2002

Total

0 0 2 0 0 0 0 1 1 0 3 2 0 0 1 1 0 0

2 0 3 3 0 0 0 14 3 0 2 3 1 0 11 2 7 4

29 0 59 68 7 0 2 13 1 0 4 17 0 0 1 3 36 3

4 0 34 36 1 0 0 3 7 0 1 3 0 0 3 2 12 1

35 0 98 107 8 0 2 31 12 0 10 25 1 0 16 8 55 8

11

55

243

107

416

z

Bromeliad included 5 genera (Aechmea, Cryptanthus, Guzmania, Neoregelia, and Viresea). Fern included 7 genera (Adiantum, Asplenium, Cyrtomium, Davallia, Nephrolepis, Platycerium, and Pteris). x Palm included 5 genera (Chamaedorea, Chrysalidocarpus, Howea, Phoenix, and Rhapis). y

(3) selection of somaclonal variants from tissue culture, or (4) selection of mutants induced from irradiated material. Hybridization was the most prevalent activity, with 289 of 416 (69%) patents resulting from crossing and selection (Table 6.2). Anthurium had the highest number of patented hybrids, with 92 of 98 (94%) derived from traditional breeding; bromeliads had 82, followed by Spathiphyllum (47), Aglaonema (32), Dieffenbachia (20), and Philodendron (14). Even with the obvious overall importance of hybridization, only seven genera/groups produced patented hybrids, while eleven genera/groups indicated no hybridization activity. There were 112 (27%) naturally occurring foliage plant sports patented: Ficus and bromeliads each showed 23 patents originating from sports, followed by Dracaena with 12, Dieffenbachia with 11, and Hedera with one. Alocasia, Chlorophytum, Epipremnum, and palms did not have patented sports. Tissue culture resulted in 12 patented foliage plants, including 4 Calathea, 2

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

255

Table 6.2. A listing of the sources of all U.S. Plant Patents issued to 18 genera/groups of ornamental foliage plants from 1976 through August 10, 2002. Sources include selections from hybrid populations, naturally occurring sports, variants from tissue culture produced populations, and irradiation induced mutation.

Genus/Group Aglaonema Alocasia Anthurium Bromeliadz Calathea Chlorophytum Cordyline Dieffenbachia Dracaena Epipremnum Ferny Ficus Hedera Palmx Philodendron Shefflera Spathiphyllum Syngonium Totals

Total issued

Hybrids

Sports

Tissue culture variants

Irradiation

35 0 98 107 8 0 2 31 12 0 10 25 1 0 16 8 55 8 416

32 0 92 82 0 0 0 20 0 0 0 0 0 0 14 0 47 2 289

2 0 4 23 4 0 2 11 12 0 9 23 1 0 1 8 8 4 112

1 0 2 2 4 0 0 0 0 0 1 0 0 0 1 0 0 1 12

0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 1 3

z

Bromeliad included 5 genera (Aechmea, Cryptanthus, Guzmania, Neoregelia, and Viresea). Fern included 7 genera (Adiantum, Asplenium, Cyrtomium, Davallia, Nephrolepis, Platycerium, and Pteris). x Palm included 5 genera (Chamaedorea, Chrysalidocarpus, Howea, Phoenix, and Rhapis). y

Anthurium, 2 Bromeliads, and one each of Aglaonema, fern, Philodendron, and Syngonium. Finally, 3 patents resulted from irradiationinduced mutants (2 Ficus and 1 Syngonium). Aroids (207 of 289 = 72%) and bromeliads (82 of 289 = 28%) accounted for 100% of foliage plant patents that involved hybridization.

III. BREEDING TECHNIQUES A. Control of Flowering Careful planning is required to ensure a sufficient supply of flowers for foliage plant breeding, especially in families as diverse as members of the Araceae and Bromeliaceae. Genera such as Aglaonema and Dieffenbachia,

256

R. HENNY AND J. CHEN

both aroids, naturally produce only 3–4 inflorescences per stem per year, and different species within each genus may not flower simultaneously (Henny 2000). This potential barrier to breeding has been overcome by the use of gibberellic acid (GA3) sprays to stimulate flowering of aroid genera, including Aglaonema, Caladium, Calathea, Dieffenbachia, Spathiphyllum, Syngonium (Henny 1995), and Philodendron (Chen et al. 2003). GA3 has also been used to stimulate flowering of Cordyline (Fisher 1980). Treatment generally consists of a single foliar spray of 250 to 1,000 ppm GA3. GA3 treatment also increases the number of flowers per plant, and, in most cases, induces different species of the same genus to flower simultaneously (Henny 1995). Bromeliads can be induced to flower by treating them with compounds that emit ethylene such as ethephon (chloroethane phosphoric acid). Crops can be sprayed at different times of the year to ensure salable plants in spike or in color at desired seasons or to produce flowers for breeding. Ethylene compounds can be applied at the rate of approximately 2,500 ppm. Treated plants should flower about two months after treatment (Griffith 1998b). B. Pollination Methods 1. Genera with Unisexual Flowers. Aglaonema, Dieffenbachia, Epipremnum, Philodendron, and other aroids possess unisexual flowers that are held on a common structure termed the spadix. Pistillate flowers on the same spadix mature simultaneously, as do staminate flowers. However, inflorescences exhibit protogyny in that pistillate flowers (located on the lower half of the spadix) are receptive before staminate flowers produce pollen (Henny 2000). This dichogamous nature discourages self-pollination and promotes outcrossing. It is necessary to obtain pollen from a separate inflorescence and to manually transfer it to the inflorescence selected as the seed parent. This can be done with a small, soft brush by brushing the pollen into a container. Alternatively, the entire inflorescence may be removed and turned so the spathe is on the bottom and catches any pollen that becomes dislodged from the spadix. Pollination can be performed via the same brush used to collect the pollen. First make the brush sticky by gently brushing it against the stigmatic surfaces, then dip it into the pollen supply, and lightly brush the pollen on the stigmatic surfaces of receptive flowers (Henny 2000). Receptivity of female flowers coincides with unfurling of the spathe. In some Philodendron species, the entire inflorescence becomes warm to the touch as an indicator of receptivity (Chen et al. 2003; McColley and Miller 1965). Receptivity of Aglaonema and Dieffenbachia flowers

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

257

lasts at least 24 hours, as evidenced by pollen germination studies (Henny 1988). Spathes from these two genera normally begin to unfurl at night and pollination can occur anytime the following day. Seed has been obtained from flowers of both genera pollinated one day after spathe unfurling, but the number of seeds is small. Female flower surfaces that have become discolored and mushy are no longer receptive. Following pollination, Dieffenbachia flowers require 100% relative humidity for pollen to germinate (Henny 1980). This can be done by wrapping the entire spadix with moistened paper toweling and enclosing it in a plastic bag. The wrap is removed the next day so that it does not interfere with pollen production. Pollen germination in Aglaonema is greater when provided high humidity (Henny 1985) but is not as sensitive as Dieffenbachia. There are no reports that Philodendron require high humidity for seed set. Palms (Chamaedorea and Rhapis) have relatively large flowers that are easy to manipulate, but the majority produces unisexual flowers and only a few species have bisexual flowers (Wilfret and Sheehan 1981). Species with unisexual flowers are generally monoecious, but a few are dioecious. The sequence of floral opening and inflorescence type must be observed to determine when pollen is shed and when stigmatic surfaces are receptive. Details of palm pollination were documented by Henderson (1986). 2. Genera with Bisexual Flowers. Aroid genera with bisexual flowers include Anthurium and Spathiphyllum; unfurling of the spathe reveals many uniform flowers located along the entire spadix. All flowers on a Spathiphyllum spadix mature simultaneously. In Anthurium, new flowers become receptive each day, beginning at the spadix base and advancing gradually toward the top over a two-week period. Pistillate flower receptivity is indicated by a glistening shine of stigmatic surfaces and stickiness to the touch. It is sometimes accompanied by small drops of exudate. Flowers may remain receptive for more than a day, so timing of pollination is not as critical (Henny 2000). The stigmatic surfaces of both Anthurium and Spathiphyllum become dry and brown before pollen is dehisced; therefore, emasculation is not required. Pollen begins to appear along the spadix, usually at the bottom first and proceeding toward the top. Pollen is available for several days on Anthurium flowers because of the uneven maturation of individual pistillate flowers. A Spathiphyllum inflorescence also will produce pollen over a 2 to 3 day period (Henny 2000). Anthurium pollen is not dispersed by wind, but it can be transferred to the receptive flowers with the fingertips. Spathiphyllum pollen is lighter and tends to be

258

R. HENNY AND J. CHEN

dispersed by air currents, so a brush should be used to collect it in a container before attempting pollination. To achieve maximum seed production, an Anthurium spadix needs to be pollinated more than once, whereas it is possible to fertilize an entire Spathiphyllum spadix with one pollination. No special environmental manipulation is needed to ensure seed set for Anthurium or Spathiphyllum. If pollen is in short supply, it can be stored in a refrigerator. Philodendron and Spathiphyllum pollen may be stored for several days or weeks in this manner (Henny 2000). Aglaonema and Dieffenbachia pollen is short-lived and germination ability declines within 1 to 2 days of storage (Henny 2000). It is best to use fresh pollen from those genera if possible. Pollinated Dieffenbachia develop mature fruits within 10–12 weeks (Henny 2000). Anthurium fruit will require up to 6 months to ripen, while Philodendron fruit vary from 2 to 6 months depending on species. Aglaonema fruits mature in 4–6 months, although some hybrids have taken 1 year to develop ripe fruit. In Dieffenbachia and Aglaonema, the seed coat turns bright red when the seed is mature. Anthurium and Spathiphyllum spadices begin to change color and soften as seeds mature. Bromeliad pollination is not difficult, but it does require close attention to the ripening of the stamens and to the pistil during the short period it may receive pollen. Normally, bromeliad flowers last but a few hours. Exceptions to this rule may be found in many of the Vriesea; if they blossom in cool weather the flower may remain receptive a second day. In some species, the stigma extends out beyond the stamens and such a flower is easy to pollinate, but in many species the stamens may exceed the pistil or may be of even length, which requires emasculation to avoid self-pollination. Pollen can be procured on a small camel hair brush that is used to deposit it on the selected stigma. Dracaena produce flowers that are cream-colored, approximately 1–2 cm across and possess a single stigma and six stamens. Flowers are held on racemes that may be 30–50 cm long, open in early morning, and must be pollinated before they shrivel in early afternoon. A single raceme may hold 200 individual flowers. The style and cucullate (hood-shaped) staminode distinguish Calathea (Marantaceae) from other families. The style is held under tension by the hood-shaped staminode, which has a trigger. The pollen is deposited in a shallow depression on the back of the style just behind the stigma. It is the upward growth of the style that forces the pollen grains from the anther, and onto the stylar depression. When the polli-

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

259

nator inserts its head into the flower in search of the nectar, it depresses the appendage, or “trigger,” on the cucullate staminode, thus releasing the style, which springs forward bringing the stigma in contact with the pollen (from a previously visited flower) on the pollinator’s body and in the same motion depositing its own pollen in the same spot. Manual pollination requires collecting pollen on the tip of a small needle and transferring it to the stigmatic tip of another flower, followed by placing pressure on back of the style, which triggers the style to spring forward, thus preventing visitation by an unwanted pollinator. Although the juvenile forms of Hedera helix used as foliage plants do not flower, mature ivy has gray-green flower buds that appear on the tips of the twigs from mid-summer onward. The flowers open into umbels in September with yellow-green flowers that have five sepals, petals, and stamens. The flowers produce nectar for assorted insects such as wasps, honey bees, and moths that help with pollination. By the end of the year, many clusters of small, hard green berries have formed and, during January and February, they slowly swell and turn black (Rose 1996). 3. Ferns. Polypodiaceae (fern) is a difficult family to breed. Ferns sold as foliage plants or seen in landscapes are in the sporophyte generation of the plant’s life. They reproduce by spores. When mature spores germinate, they produce a prothallium, a very small, flat, green, mosslike structure, which is the gametophyte generation. Sexual reproduction takes place during this generation. Fertilization occurs, and the next sporophyte generation develops when the prothallium is still small. During the gametophyte stage, archegonium (eggs) and antheridium (pollen sacs) are formed on the prothallium. The antheridium produces mobile antherozoids, that swim to the archegonium and fertilize the eggs when the prothallium is covered with a film of water. Techniques must be developed to introduce desired antherozoids of one species or variety onto the prothallium of another species to produce hybrids. These must be introduced at the proper time to allow fertilization of egg cells before antherozoids of the female parent are released from the prothallium. This task is very exacting and difficult so few people have undertaken hybridizing ferns. C. Seed Handling To achieve good germination, aroid seed should be separated from the spadix. Cleaning seeds speeds germination and lessens the chance of disease starting in the decaying fruit. For genera like Dieffenbachia and Aglaonema that have large seeds, the red berry-like fruit is harvested, and

260

R. HENNY AND J. CHEN

the fleshy seed covering is removed. Genera with large numbers of small seeds, such as Spathiphyllum, are more difficult. It is easiest to harvest the entire spadix when mature (indicated by a change in color from green to yellow and a softening of the tissue) and placing it in a plastic bag with a little water. The spadix tissue will decay in a few days, allowing the seeds to be removed by gently washing them on a screen small enough to catch the seeds but letting the rotted spadix tissue fall through. Once cleaned, seeds should be planted before becoming dry. Good germination is achieved if the seeds are sown on the top of moist medium and covered with plastic or some other material to prevent drying. Soil temperature should be kept at a minimum of 21°C. Aroid seed have no dormancy requirements and begin to grow as soon as sown. They can be removed from the germination chambers and repotted once the first true leaves are produced. Most aroid seedlings require at least 1–2 years before they are large enough to be evaluated for their ornamental value. D. Testing and Releasing New Cultivars At the University of Florida, if a desirable foliage plant is identified from a hybrid progeny, it is first propagated vegetatively for further testing. Plants of the hybrid are given to cooperating Florida tissue culture laboratories for establishment. Subsequently, the hybrid is distributed to trial growers around the state who evaluate the plant under commercial growing conditions. In addition, plants are given back to the University for evaluation of growth rate and tolerance to shipping and interior conditions. If the new clone performs well for the trial growers and in University trials, it is named, patented, and released. Industry participation in the propagation and testing process of foliage plant hybrids is critical in the development of new cultivars at the University of Florida.

IV. BREEDING OBJECTIVES Breeding objectives with foliage plants include the improvement of traits related to ornamental value and stress resistance. A. Phenotypic Traits Because the value of foliage plants lies in the esthetic qualities, the improvement of ornamental traits, such as plant form, leaf shape, texture, plant height, shape, thickness, and color, as well as growth rate, has always been important to any breeding program of foliage plants.

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

261

1. Foliar Variegation and Colors. Variegated plants comprise about one-third of the ornamental plants grown commercially (Betrock 1996). Due to their magnificent multicolored leaves, variegated foliage plants, such as Aglaonema, Calathea, Chlorophytum, Cordyline, Dieffenbachia, Dracaena, Epipremnum, and Syngonium, are widely used in interior plantscapes. Consumer preferences for plant characteristics placed leaf variegation as the second most important consideration in the purchase decision (Behe and Nelson 1999). According to Kirk and Tilney-Bassett (1978), variegation can be categorized as either cell lineage or noncell lineage types. Cell lineage variegation occurs in genetic mosaics (individuals with cells of different genotypes), while in noncell lineage variegation, all cells have the same genotype but the genes responsible for the synthesis or destruction of pigments are expressed only in some of the cells (Marcotrigiano 1997). Noncell lineage variegation is expressed in Aglaonema and Dieffenbachia and in both genera the variegation patterns are inherited in simple Mendelian fashion (Henny 1983; Henny 1982; Henny 1986a). The common cause of noncell lineage variegation is the result of differential gene expression (Marcotrigiano 1997). Inheritance of foliar variegation has been studied in Caladium, Aglaonema, and Dieffenbachia. A single dominant gene controls the netted venation pattern of Caladium, with the recessive genotypes having no pattern (Wilfret 1986). The red main vein in Caladium leaves is dominant to green, and white is dominant to both green and red. Red vein was also found to be epistatic to netted venation, in that the homozygous genotype for red veins produces a solid red leaf with a green margin (Wilfret 1986). Red and white leaf spots in Caladium are governed by codominant alleles (Wilfret 1986; Zettler and Abo El-Nil 1979). The presence of foliar variegation in Dieffenbachia and Aglaonema is dominant to non-variegation. A single dominant allele (Pv) determines the presence of a variegation pattern typical for Dieffenbachia maculata ‘Perfection’ (Henny 1982). The same allele controls a similar pattern in D. maculata ‘Hoffmannii’ with slight differences due to modifying genes. A mutation of the Pv allele to Pv1 produced the variegation pattern present in D. maculata ‘Camille’ (Henny 1986a). The Pv1 allele masks expression of the Pv allele in plants containing both alleles. Six different foliar variegation patterns of Aglaonema were governed by multiple alleles at a single locus (Henny 1983a, 1986b). Each distinct pattern was controlled by a separate dominant allele. Alleles were codominant, allowing expression of two variegation patterns in the same plant. Several other variegation patterns in Aglaonema currently being studied appear to be inherited in the same manner.

262

R. HENNY AND J. CHEN

Studies in the inheritance of the white foliar midrib in Dieffenbachia show it is controlled by a single dominant gene (Henny 1983b) linked to the gene controlling foliar variegation. The dominant alleles for each trait were carried on homologous chromosomes. 2. Leaf Shape and Size; Petiole Colors. Some characteristic leaf shapes and sizes are particularly attractive and striking. Changes in leaf size and shape can create new and exciting appearances in hybrids and in aroids; such traits are under multigenic control. Petiole color in Aglaonema, which includes green, pink, white, and russet, is inherited independently of foliar variegation and appears to be due to the interaction of at least two genes. 3. Plant Form. Plant overall form has always been an important trait in foliage plants. There are six principal groups. Upright. Plants such as Dracaena fragrans ‘Massangeana’ are defined as upright, with their dramatic tall and narrow form. Such plants have no hint of a spreading or trailing stem to soften the outline, even after several years of growth. They are suited to be placed in corners of rooms. Compact or Clumping. This form is attributed to two common elements present in foliage plants, including basal shoot formation and dwarf growth habit. The tendency for plants to develop basal shoots, or suckers, is under multigenic control in Anthurium and Dieffenbachia (R. J. Henny unpublished). Highly suckering plants of both genera tend to transmit the trait to hybrids in varying degrees. Production of numerous basal shoots is a highly desirable trait in Aglaonema, Spathiphyllum, and Dieffenbachia. A cultivar with more basal shoots will need fewer cuttings per container to produce a full appearance. Additionally, basal shoots can be removed and used as propagules. Trailing. Stems that cascade down or are prostrate are typical of such plants as Hedera helix and Ficus pumila. Stems of the spider plant (Chlorophytum comosum) strike upward initially then bend gracefully as they develop, giving the effect of a fountain. As plants start to mature, they produce propagules on the ends of stems that cascade down to make the plant a true trailer. Climbing. A climber generally has limber stems that will trail if they are not supplied with support. They may have twining stems, clinging ten-

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

263

drils, or aerial roots to support them. Such plants include Epipremnum aureum, Monstera deliciosa, and Philodendron scandens oxycardium. Standards. These are treelike plants that have a main stem and a branching head. Ficus benjamina is a popular example. Architectural. The term “architectural” applies to forms that are unusual and dramatic. The shape and outline of such plants is bold and eyecatching. They are always large plants, tending to be tall rather than spreading, such as Chamaerops humilis, the European fan palm, which has big, boldly cut, fan-shaped leaves and sturdy, hairy roots. 4. Flowers. Foliage plants that are also commercially valued for their flowers include Anthurium, Aphelandra, Bromeliads, and Spathiphyllum. In Anthurium, two genes, M and O, are responsible for production of the five major spathe colors of Anthurium andraeanum (Kamemoto and Kuehnle 1996). The colors and their respective genotypes are: red (MMOO, MMOo, or MmOO); pink (MmOo); orange (mmOo, mmOO, or mmoo); coral (mmoo); and white (mmoo or Mmoo). Breeding objectives concerning flowering foliage plants are to increase flower number and longevity and to expand the range of flower colors. In addition, the ability to continue flowering under interior conditions is important in Anthurium and Spathiphyllum. A six-month evaluation of five Anthurium cultivars under a light intensity of 16 µmol m–2 s–1 (Chen et al. 1999) showed that monthly new leaf growth ranged from 1.2 to 5.4 and new flower appearance from 1.4 to 4.7 among cultivars. ‘Red Hot’ showed the best flowering and growth performance with a weekly average flower count of 4.7 and 5.4 new leaves. The leaves were dark green and shiny, and the flowers were good quality. These large differences among cultivars in interior performance indicated there are good possibilities for selecting Anthurium cultivars that will continue to grow and flower under low interior light levels. 5. Fragrance. The consumer often desires fragrance in flowers. First generation progeny analyses from 22 crosses between non-fragrant and fragrant parents indicated that multiple genes govern scent characteristics in Anthurium (Kuanprasert and Kuehnle 1999). Spathiphyllum flowers are also naturally very fragrant for about a two-week period once the spathe opens to expose the spadix. However, no specific breeding work has been done with fragrance as a goal for Spathiphyllum. 6. Growth Rate. Foliage cultivars vary in growth rate. Anthurium ‘Krypton’ produced a dry weight of 38 g, whereas A. ‘Tropic Fire’ had a dry

264

R. HENNY AND J. CHEN

weight of 24 g when both began with the same fresh weights and were grown under identical conditions for 8 months (Chen et al. unpublished). Dieffenbachia ‘Tropical Star’ became marketable one month earlier than D. ‘Snowflake’ and D. ‘Exotica Perfection’ when grown under an ebb-and-flow fertigation system in which fertilizer and water supplies were controlled (Chen et al. unpublished). Many foliage plants grow slowly, so improving growth rate (i.e., reducing the time from transplanting to finishing) will reduce production time and increase profits. B. Stress-related Traits Foliage plants often experience physical, chemical, and biotic stresses in both production and interiorscaping. These stresses include drought, high or low temperatures, inappropriate radiation, low or excessive nutrient levels, agrochemicals, disease, and pests. The important stresstolerance traits in foliage plant production and utilization are listed below. 1. Adaptation to Interior Environments. The ability of foliage plants to adapt to interior environments and maintain their esthetic appearance is one of the most important traits (Chen et al. 2001b). Indoor selection criteria designed for evaluating a plant’s interior performance in our program include a light intensity of 8 or 16 µmol m–2 s–1 for 12 hr a day, a temperature from 20 to 24°C, relative humidity of 40 to 50%, and a CO2 concentration of 600 µL L–1. Interior performance is evaluated using several traits, including leaf yellowing, leaf drop, loss or reduction in foliar variegation, elongation of internodes (i.e. stretching), changes in overall plant configuration or form, change in leaf or flower color, flower longevity, loss of flowering, or development of physiological disorders, diseases, or pests. Good interior plants should be able to maintain their esthetic appearance for at least six months after installation in the interior environment. Light intensity is the most crucial factor that limits plant performance in interior environments. Plants that can be maintained under 8 µmol m–2 s–1 can be used in low light areas such as offices, hotel hallways, or corners of conference rooms. Those maintained under 16 µmol m–2 s–1 are suitable for higher light locations such as airport waiting areas, malls, or bright living rooms. General information with regard to the interior performance of common foliage plant genera is available (Manaker 1997), but the genetic mechanisms underlying their adaptation to the indoor environment are

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

265

largely unknown. Plants tolerant to the lowest light levels include Aglaonema, Calathea, Dracaena, Epipremnum, Nephrolepis, and Syngonium. However, other genera contain species or cultivars that express different tolerances to interior low light conditions. For example, newly released Dieffenbachia ‘Snowflake’ exhibited much better interior performance than the more common ‘Perfection Compacta’ (Chen et al. unpublished data). Cultivars with better interior performance could be potential parents for breeding future hybrids. 2. Disease and Insect Resistance. Foliage plant production requires a warm and humid environment. These are ideal conditions for rapid increase and spread of bacterial, fungal, and viral diseases. Disease problems are more common in production than in interiors where conditions are cooler and drier. Most fungal diseases are controlled with fungicides, but chemical controls are usually ineffective against bacteria or viruses (Chase 1997). Plants infected with bacteria or viruses must be destroyed and only clean materials should be used as a source of propagules (Chase 1981). In an attempt to reduce chemical use in nursery crop production, breeding for disease resistance has become one of the major objectives in foliage plant improvement. With the availability of diverse plant species or cultivars, a wide range of resistance has been identified in many foliage plant genera. Resistant accessions can be used as donors for improving disease resistance. Current programs at the University of Florida include breeding for resistance in Anthurium to Xanthomonas bacterial rot (Norman et al. 1999a), Spathiphyllum to Cylindrocladium fungal root and petiole rot (Norman et al. 1999b), and Syngonium to Myrothecium fungal leaf spot (Norman and Henny 1999). Breeding efforts at the University of Hawaii include the improvement of Anthurium resistance to Xanthomonas campestris pv. dieffenbachia and the nematode Radophlus similis (Kuehnle et al. 2001; Wang et al. 1998). Insect and mite problems can develop rapidly in large populations of foliage plants. The major insect pests for foliage plants include aphids, caterpillars, mealybugs, scales, and thrips (Hamlen et al. 1981; Baker 1994). Adequate chemical controls exist for most insect and mite problems and, if crops are routinely monitored, good control is possible (Short et al. 1999). Use of pest-free propagules and maintenance of sanitary conditions in production areas are essential for reduction of potential pest problems (Hamlen et al. 1981). The great challenge to insect control, however, is foliage plants in interior environments such as public conservatories, homes, hotels, office buildings, restaurants, shopping malls, hospitals, and schools where few chemicals have been

266

R. HENNY AND J. CHEN

cleared for interior plant use. The best long-term solution is the development of cultivars that resist infestation in interior environments. Chen et al. (2001d) evaluated resistance of different Anthurium cultivars to infestation of either banded greenhouse thrip (Hercinothrips femoralis) or two-spotted spider mite (Tetranychus urticae) in interior conditions. Cultivars strongly resistant to one pest were not resistant to the other. However, some cultivars showed moderate resistance to both pests, suggesting that genetic improvement of resistance to the banded greenhouse thrip and the two-spotted spider mite is possible. 3. Temperature Tolerance. Because of their tropical or subtropical origin, foliage plants are sensitive to chilling temperatures. Chilling in foliage plants is defined as a temperature that is cold enough to cause injury but not cold enough to freeze the plant tissue, usually ranging from just above 0° to 15°C (Chen et al. 2001e). Chilling injury can be visible, ranging from water-soaked patches or necrotic lesions on leaves and also invisible, mainly in reducing plant growth rate. For example, chilling injury on Spathiphyllum appears at 7°C with injured leaves becoming necrotic and dry. There is no visible injury immediately following exposure to 10°C, but plant growth index can be reduced by up to 50%, depending on cultivar as a delayed expression (Qu et al. 2000). Aglaonema injury occurs at 13°C, characterized by dark and greasyappearing patches on the surface of leaves (Chen et al. 2001a). Tissue collapse in older leaves is a typical symptom in Dieffenbachia (Conover and Poole 1974). Chilling injury is a significant cause of loss of foliage plants not only in production, but also in transportation and interiorscaping. Chilling injury accounts for nearly 50% of all transportation damage claims (Conover 1980). After evaluating 22 species in simulated shipping experiments, Poole and Conover (1993) concluded that the best shipping temperatures are in the range of 15 to 18°C. Chilling injury to interior plants often occurs when interior temperatures fail to maintain an appropriate level either in winter or summer (Manaker 1997). Genetic variation in chilling resistance exists in foliage plants. Chen et al. (2001a) evaluated 10 Aglaonema cultivars and found that ‘Silver Queen’, one of the most popular cultivars in the foliage plant industry, was extremely sensitive to chilling. ‘Silver Queen’ had 30% of leaves injured 10 days after chilling plants to 13°C. ‘Maria’, a cultivar well known for its chilling resistance, was not the most resistant one tested. Ten days after chilling at 2°C, 32% of Maria’s leaves were injured, but there was no discernable injury on ‘Emerald Star’, ‘Stars’, or ‘Jewel of India’, three recently released hybrids. Use of resistant cultivars may

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

267

greatly reduce chilling injury instances during production and transportation and also conserve energy used during the greenhouse production phase.

V. FOLIAGE EXAMPLES The following discussion will focus on cultivar development of important genera. These genera collectively account for at least 60% of the total U.S. wholesale value of ornamental foliage plants (USDA 1998). Cultivars may originate from plant collection, traditional breeding, and selection of mutants from traditional propagation or from somaclonal variants generated from tissue culture. A. Aroids (Araceae) 1. Aglaonema. The genus Aglaonema, commonly referred to as Chinese evergreen, is comprised of 21 species native to southeast Asia, from northeastern India, across southern China, into Indonesia and New Guinea (Mayo et al. 1997). Most Aglaonema species are open-pollinated; a few may exhibit apomixis, such as Aglaonema costatum ‘Foxii’. Propagation of Aglaonema is by seeds or vegetative tissue, mainly by tip cuttings or division. The basic number of chromosomes could be x = 6, with subsequent polyploidy in many cases (Jones 1957). Both Aglaonema haenkii and A. simplex have a 2n chromosome number of 60, and A. commutatum has 2n = 120 (Jones 1957). However, 2n chromosome number of 40 was also reported in A. pictum Kunth and A. oblongifolium (Kunth) Schott. (Marchant 1971). Speculations on the difference in chromosome numbers among Aglaonema species include chromosome doubling and even dysploidy (Tischler 1954). Aglaonema is one of the most widely used plants in interiorscape due to its ability to tolerate low light and low humidity and its resistance to diseases and pests. The number of commercial cultivars increased from 10 in 1975 to 36 at the end of the 1990s (Table 6.3). Commercial growers add and discard cultivars on a regular basis. For example, when cultivars listed in the FNGA (Florida Nurserymen and Growers Association) Locator 1998–1999 and 1999–2000 are compared, five cultivars listed in 1998–99 were no longer listed, and four new cultivars were added in 1999–2000. Almost all Aglaonema cultivars are hybrids developed through traditional breeding. Current breeding activities are mainly focused on generating novel foliar variegation patterns, petiole colors, increased branching, and chilling resistance (Henny 2000).

268

R. HENNY AND J. CHEN

Table 6.3. Changes in cultivar numbers of major foliage plant genera or groups in commercial production of Florida.z Foliage plant Aglaonema Alocasia Anthurium Calathea Dieffenbachia Dracaena Epipremnum Fern Ficus Hedera Palm Philodendron Spathiphyllum Syngonium

1975

1988–89

1998–99

10 0 0 3 7 12 3 10 14 1 7 16 4 7

17 2 12 14 29 21 4 21 32 12 19 24 24 14

36 14 36 21 23 33 4 54 46 40 22 18 51 17

z

FNGA (Florida Nurserymen and Growers Association) Locator 1975, 1988–1989, and 1998–1999, Orlando, FL.

Aglaonema hybrids are almost exclusively selected from interspecific hybridization. Species commonly used in interspecific hybridization include A. nitidum, A. commutatum, A. costatum, and A. rotundum. The most common hybrids grown in Florida from the 1960s to the 1980s were Aglaonema ‘Fransher’ (A. treublii × A. marantifolium tricolor), A. ‘Parrot Jungle’, and A. ‘Silver King’ (A. curtisii × A. commutatum ‘Treubii’), developed by Nat Deleon of Miami, Florida, and A. ‘Silver Queen’ (A. commutatum ‘Treubii’ × A. nitidum ‘Curtisii’), developed by Bob McColley of Bamboo Nursery in Orlando, Florida. Currently popular hybrids come from breeding programs at the University of Florida’s MREC in Apopka and also from southeast Asia, for example, India, Indonesia, and the Philippines. MREC hybrids include Aglaonema ‘Silver Bay’ (Henny et al. 1992a) that has a medium green edge overlaid with a gray-green center. ‘Golden Bay’ (Henny and Chen 2001) is a white-stemmed cultivar and has very bright cream and green color variegation. ‘Emerald Bay’ has a white and green speckled stem, while ‘Diamond Bay’ displays a clear central gray stripe against a dark green leaf blade (Henny et al. 2002a). Sunshine Foliage World, Zolfo Springs, Florida, introduced 30 new cultivars developed by breeders in Thailand. These cultivars, including ‘Jubilee Petite’, ‘Peacock’, ‘White Rain’, ‘White Lance’, ‘Brilliant’, ‘Illumination’, ‘Black Lance’, ‘Green Lady’, ‘Patricia’, and ‘Stars’, have different sizes, shapes, and

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

269

variegation patterns of leaves and white, green, or pink petioles. ‘Emerald Star’ and ‘Jewel of India’ are two cultivars developed by breeders in India. 2. Alocasia. This genus of 70 species is native to southeast Asia from India to the Philippines and New Guinea (Bailey and Bailey 1976). Common species used as foliage plants include A. cucullata, A. cuprea, A. indica, A. longiloba, A. lowii, A. macrorhiza, A. portei, A. sanderiana, A. thibautiana, A. veitchii, A. watsoniana, and A. zebrina. Alocasia cucullata and A. macrorhiza are giant taro, cultivated also for their edible rhizomes. Chromosome number of most species are 2n = 28, but A. odora is 2n = 56, and A. lowii is 2n = 70, suggesting that the basic number of chromosomes is x = 7 (Marchant 1971). Alocasia has unisexual flowers. Hybrid alocasias have been known for about 100 years and were first produced by European horticulturists (Reark 1951). Hybrids that were popular in the 1950s were A. × Amazonia (A. lowii Var. Grandis × A. sanderiana), A. × Cantrieri (A. cupera × A. sanderiana), A. × Sedenii (A. lowii × A. sanderiana), and A. × Mortefontainensis (A. lowii × A. sanderiana). Seed production is one method of propagation, but the common method is through division of offsets. Even though alocasias are valued for their variegated colorful leaves, they were not included in a recent book Tropical Foliage Plants: A Grower’s Guide (Griffith 1998b). Recent renewed interest in alocasias is largely due to the release of more than 40 new cultivars in the last four years. These cultivars are somaclonal variants and have different colors, shapes, and sizes of leaves embellished with unique variegation, providing valuable additions for interiorscaping. The most notable cultivars include ‘Bako Park’, ‘Black Stem’, ‘Black Velvet’, ‘Corozon’, ‘Elaine’, ‘Fantacy’, ‘Frydek’, ‘Grandis’, ‘Nobilis’, ‘Polly’, ‘Purple Prince’, ‘Sarian’, ‘White Knight’, and ‘Wentii’. Because of the effectiveness of tissue culture in generating desirable mutants, cultivar development in Alocasia now relies primarily on the selection of somaclonal variants. 3. Anthurium. This is the largest genus in the Araceae, consisting of about 1000 species (Croat 1992). The distribution of this genus ranges from northern Mexico and the Greater Antilles to southern Brazil, northern Argentina, and Paraguay (Croat 1983, 1986). Anthuriums are valued for their exotic shape, colorful spathe and spadix with great longevity, and also the attractive foliage of some species. Species from this genus have been used commercially as cut flowers, potted flowering foliage, or potted foliage plants. Anthurium has bisexual flowers and, in many species, emit odors while blooming. Therefore, Anthurium species are

270

R. HENNY AND J. CHEN

largely cross-pollinated by insects. Propagation can be performed through either seed, division, or tissue culture. Commercial cultivars used for cut-flower production are mainly from species of A. andraeanum, whereas potted flowering cultivars were initially derived from A. scherzerianum (Kamemoto and Kuehnle 1996). Species that are produced for their attractive foliage include A. warocqueanum, A. crystallinum, and A. magnificum. Some species, such as A. armeniense, A. fragrantissimum, and A. lindenianum, emit a pleasant fragrance (Kuanprasert and Kuehnle 1999), and these species may be useful for breeding for floral fragrance. Because of its richness in germplasm resources and economic importance as cut flowers, potted flowering foliage, or potted foliage plants, Anthurium is probably the most widely studied genus in relation to cultivar development. Areas of investigation include plant collection and identification (Croat 1983, 1986), isozymes or DNA markers for relatedness determination (Kobayashi et al. 1987; Ranamukhaarachchi et al. 2001), plant introductions (Dressler 1980; Kamemoto and Kuehnle 1996), liquid raft culture conservation of germplasm (Kuehnle, personal communication), traditional breeding (Kamemoto and Kuehnle 1996; Henny 2000), micropropagation (Matsumoto and Kuehnle 1997), somatic embryogenesis (Kuehnle et al. 1992), protoplasm fusion (Kuehnle 1997), and genetic transformation (Kuehnle et al. 2001). In most species studied, the chromosome number is 2n = 30 (Sheffer and Kamemoto 1976a; Sheffer and Croat 1983). Pollination between A. andraeanum and A. scherzerianum has never been successful due to their distant relationship since A. andraeanum belongs to the section Calomystrium, but A. scherzerianum is from the section Porphyrochitonium (Kamemoto and Sheffer 1978; Sheffer and Kamemoto 1976b). The discovery of A. amnicola Dressler, a species belonging to the section Porphyrochitonium, was made in Cocle del Norte, Panama, in 1972 by Dressler (1980) at the Smithsonian Tropical Institute in Balboa, Panama, and its introduction to cultivation made a great difference in potted Anthurium cultivar development. The plant is not only compact and attractive but also can be easily crossed with species from the section Calomystrium, including A. andraeanum. Another species, A. antioquiense, also belonging to section Porphyrochitonium, was found on the Pacific coast of Colombia. In addition to its desired phenotype and crossability like A. amnicola, this species is also tolerant of the bacterial blight caused by Xanthomonas campestris pv. dieffenbachiae. Currently, most breeding programs for potted Anthurium cultivar development use A. antioquiense and A. amnicola as parents because both species are dwarf and highly floriferous (Kamemoto 1981; Kamemoto

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

271

and Sheffer 1978). These species cross with species from section Calomystrium such as A. andraeanum and yield interspecific hybrids with a wide range of spathe colors. One example is Anthurium ‘Red Hot’ (Henny 1999), which was developed from an initial interspecific cross of A. andraeanum with A. amnicola. ‘Red Hot’ branches freely and exhibits a compact growth habit with each pot averaging five red flowers. It is now one of the most popular cultivars in the foliage plant trade. The reader is referred to Breeding Anthuriums in Hawaii by Kamemoto and Kuehnle (1996), an excellent treatise on breeding for cut flower and potted Anthurium use. In addition to traditional breeding, tissue culture propagation of interspecific hybrids has led to the selection of a large number of somaclonal variants and subsequent release of many new Anthurium cultivars. Oglesby Plants International, Inc., Altha, Florida, successfully introduced the Lady series starting with ‘Lady Jane’ in 1985, followed by ‘Lady Anne’, ‘Lady Beth’, and ‘Lady Carmen’. All the Lady series are dwarf and compact cultivars with deep green lanceolate foliage contrasting to either white, pink, or red spathes; the Lady series rapidly gained market share as potted flowering foliage plants. Potted Anthurium cultivar development is also extremely active in Europe, especially in the Netherlands, but information on breeding and cultivar development is not readily available. Anthurium is the only foliage plant genus thus far that has been genetically transformed. Kuehnle and Sugii (1991) presented the first evidence of tumor formation and nonpaline production in Anthurium andraeanum Hort. when co-cultivated with Agrobacterium tumefaciens strains A281 and C58 in an induction medium containing acetosyringone. Subsequently, using nontumorigenic Agrobacterium strain LBA4404, containing the vir-helper plasmid (pAL4404) in strain Ach5 chromosomal background, Kuehnle et al. (2001) developed an effective method for transformation of different Anthurium cultivars. Testing transgenic plants expressing the att gene showed greater decreases in symptoms and in the number of inoculated Xanthomonas campestris pv. dieffenbachiae strain D150 compared to the control plants (Kuehnle et al. 2001). The successful transformation of Anthurium offers additional avenues for improving cultivars of not only this genus but also other important aroid genera discussed in this review, and possibly foliage plant genera from other families. 4. Dieffenbachia. The genus Dieffenbachia, commonly known as dumb cane, is composed of about 30 species native to moist, lowland tropical forests of Central and South America (Bailey and Bailey 1976). Flowers

272

R. HENNY AND J. CHEN

are unisexual with dichogamy in nature; thus, Dieffenbachia is predominately cross-pollinated. The chromosome number of most Dieffenbachia species is 2n = 34 (Jones 1957). A significant amount of Dieffenbachia propagation has shifted from tip cuttings to tissue culture, partly in an attempt to develop stock plants free of systemic bacteria and viruses. Through hybridization and selection of sports and somaclonal variants, almost 100 cultivars have been introduced over the years, of which only about 20 cultivars are grown commercially (Table 6.3). Interspecific hybridization is the primary means of generating new cultivars. Inheritance of foliage variegation has been shown to be dominant to non-variegation and a single dominant allele (Pv) in interaction with modifying genes determines variegation pattern of D. maculata (Henny 1982, 1986a). Basal shoot formation is controlled by multiple genes. Breeding programs have been focused on the improvement of variegation patterns concomitant with increased basal shoot formation. Nine hybrids have been released at the University of Florida including ‘Triumph’, ‘Victory’, ‘Tropic Star’, ‘Starry Nights’, ‘Star White’, ‘Star Bright’, ‘Sparkles’, ‘Tropic Honey’, and ‘Sterling’ (Henny et al. 2003b). These hybrids have different variegation patterns, large leaves with short petioles, and, in most cases, they produce basal shoots freely. E. J. Frazer in Brisbane, Australia, bred several hybrid cultivars, including ‘Tropic Breeze’, ‘Tropic Rain’, ‘Tropic Dawn’, and ‘Tropic Forest’, that have been introduced into the foliage plant industry by Twyford International Inc., Apopka, Florida. Selection of sports is another avenue of cultivar development in Dieffenbachia. At least 11 patented cultivars were selected from spontaneous mutations (Table 6.3). Most of them were derived from D. amoena and D. maculata. For example, ‘Tropic Snow’ is a sport of D. amoena, and ‘Tropic Sun’ and ‘Maroba’ are sports of ‘Tropic Snow’. Cultivar ‘Marianne’ is a mutant of D. maculata ‘Perfection Compacta’; ‘Camille’ is a sport of ‘Marianne’. As tissue culture becomes a routine method of Dieffenbachia propagation, more somaclonal variants have been generated. Selection of a desired somaclonal variant has proved to be a more effective way of cultivar development. For example, ‘Rebecca’s Jewel’ was selected from somaclonal variants rising from tissue culture of D. maculata ‘Camille’. 5. Epipremnum. Epipremnum (pothos) is indigenous to southeast Asia and the Solomon Islands (Huxley, 1994). Epipremnum has about 10 species (Bailey and Bailey 1976), but only E. aureum is widely grown. In fact, E. aureum is one of the most popular houseplants worldwide. Pothos has unisexual flowers and is propagated primarily by single or

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

273

double eye cuttings. Tissue culture has not been used commercially, and there is no information in the literature regarding pothos breeding. Three cultivars (‘Golden Pothos’, ‘Marble Queen’, and ‘Jade’) have been dominant in the market for decades. The only new cultivar recently released is ‘Neon’ with uniform yellowish-green foliage. Qu et al. (2002) successfully regenerated pothos from petiole and leaf explants. There is potential for developing somaclonal variants by screening explants generated from tissue culture. 6. Philodendron. The genus Philodendron contains 700 or more species, making it the second largest genus in the Araceae (Croat 1997). Philodendrons are native to tropical America and comprise a conspicuous component of the native flora because of their abundance, climbing habit, and large, showy leaves. Philodendron formerly dominated all other genera of tropical ornamental foliage plant production, accounting for 50% in 1950 and 36% in 1967 of the national wholesale value of foliage plants in the United States (Smith and Strain 1976; McConnell et al. 1989). Documented interspecific hybridization within the genus Philodendron dates to 1887 in Florence, Italy (Wilfret and Sheehan 1981) with the production of P. corsinianum (P. lucidum × P. cariaceum). The first U.S. hybrid, P. mandaianum (P. hastatum × P. erubescens) was developed by Manda in 1936. Beginning in 1951, most Philodendron breeding was conducted by Bob McColley of Apopka, Florida. He classified Philodendron into three groups based upon the sexual compatibility of the species (McColley and Miller 1965). The first is the self-heading group, growing upright on their own, which is represented by P. wendlandii and the hybrid ‘Black Cardinal’. The second group is the erect-arborescent type, such as P. selloum, which appear self-heading when young, but assume more woody and treelike as they mature. While plants within each group cross freely, no successful crosses have been made between the two groups. This is in part because the chromosome number in the first group is 2n = 34 but 2n = 36 in the second group (Jones 1957). The third group is the vining or scandent type, such as P. scandens oxycardium (heartleaf philodendron); it is commonly grown as hanging baskets. The chromosome number of P. scandens oxycardium is 2n = 32 (Jones 1957). Philodendron scandens has not been successfully self-pollinated or crossed with any plants from the first two groups, indicating that it may be sterile. McColley made about 30 interspecific hybrids using the self-heading species. Some of his hybrids, such as ‘Autumn’, ‘Black Cardinal’, ‘Imperial Green’, ‘Imperial Red’, ‘Moonlight’, ‘Prince of Orange’, ‘Red Empress’, and ‘Red Emerald’, are still popular.

274

R. HENNY AND J. CHEN

7. Spathiphyllum. Spathiphyllum, commonly called peace lily, is one of the most popular foliage plants in the trade due to its dark green foliage, long-lasting showy white flowers, and ease of growing. Peace lily originates in the damp, tropical forest habitats of the Americas, the Philippines, and Indonesia (Mayo et al. 1997). Its survival in the tropical forest understory has enabled it to adapt to interior low light conditions. Spathiphyllum has bisexual flowers and is naturally pollinated by bees (Williams and Dressler 1976). The chromosome number of Spathiphyllum is 2n = 30 (Jones 1957). Plants are propagated either through seeds, division, or tissue culture. Spathiphyllum is a typical example of how a valuable foliage plant can be quickly accepted by the foliage plant industry. In the early 1970s, there were only two cultivars available: ‘Clevelanii’ and ‘Manua Loa’. Now, the number of Spathiphyllum cultivars exceeds 50 in Florida alone (Table 6.3). The surge of new cultivars is largely attributed to traditional breeding, creation of interspecific hybrids, and selection of somaclonal variants from tissue culture. Cultivars may be divided into three classes based on plant size (Griffith 1998b): large, medium, and small. ‘Sensation’ is the largest cultivar in production at about 1.5 m tall. It was the result of a cross of ‘Mauna Loa Supreme’ and ‘Fantastica’. The largest cultivars are usually grown in containers with a diameter of 24 cm or larger. The medium-size cultivars are generally produced in containers from 15- to 25-cm in diameter. Among them, ‘Tasson’ (derived from a cross of ‘Manua Loa’ and ‘Wallisii’) was popular for at least 10 years. The small-size cultivars (i.e., ‘Petite’) are usually grown in containers 15 cm in diameter or smaller. 8. Syngonium. Syngonium, whose name refers to the cohesion of the plant ovaries in Greek, has been a popular houseplant genus for many years. Syngonium, commonly known as arrowhead vine or nephthytis, is native to the region from Mexico to Panama and consists of about 33 species (Croat 1982), but only one species (S. podophyllum) is cultivated. Chromosome counts are 2n = 24, 26, 28, or 30 (Marchant 1970; Pfitzer 1957; Sharma 1970). Flowers are unisexual, and inflorescences are protogynous, becoming receptive 1–2 days before the staminate flowers shed pollen (Croat 1982). As mentioned earlier, cultivars are largely selected from somaclonal variants rising from tissue culture. Current breeding research at the University of Florida is aimed at developing Syngonium hybrids resistant to fungal and bacterial diseases (Norman et al. 2002). Screening work is being conducted using 15 Syngonium species and 20 accessions. Syngonium flowers can be induced by gibberellic acid (GA3) sprays (Henny et al. 1999), and its use is necessary to obtain simultaneous flowering of different species.

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

275

B. Bromeliads (Bromeliaceae) Bromeliaceae have more than 50 genera containing more than 2,700 species native to tropical North and South America (Benzeng 2000). Their exotic appearance, graceful symmetry, and potential for yearround flowers make bromeliads profitable as ornamentals. Plants can remain in bloom for weeks or months under interior conditions. The rise in importance of Bromeliads has been recent, since they were not listed in 1988 as a major group of foliage plants in the Census of Horticulture Specialties of the U.S. Department of Agriculture’s National Agricultural Statistics Service (McConnell et al. 1989). However, they were listed in 1998 and now account for 5.1% of the total national foliage plant wholesale value (USDA 1999). Most species of Bromeliaceae are predominately outcrossing (McWilliams 1974). Protogyny occurs in species of Tillandsia, protandry in Vriesea, self-incompatibility in Ananas, and even dioecy in some species of Catopsis (Benzeng 2000). Some species cultivated by ants (e.g., Aechmea mertensii and A. tillandsioides) regularly set self-seeds (Madison 1979). A few species of Guzmanala, such as G. graminifolia, are also selfed under natural conditions. Bromeliads that are produced as foliage plants come primarily from the genera Aechmea, Cryptanthus, Guzmania, Neoregelia, Nidularium, Tillandsia, and Vriesea. Species of Aechmea, Guzmania, Neoregelia, Nidularium, Tillandsia, and Vriesea are largely epiphytic, while species from Cryptanthus are mostly terrestrial. The chromosome number of the previously listed seven genera is predominately 2n = 25, with exception of Cryptanthus, which is 2n = 34 (Marchant 1967). Bromeliads may be propagated from seeds, offsets (lateral shoots or suckers), or tissue culture. Breeding of bromeliads started in Europe in the late 1880s. Eduard Morren did his first cross between Vriesea psittacina and V. carinata in 1879, which resulted in V. ‘Morrenian’ (Samyn and Thomas 1997). Hybridization and progeny selection were popular in Europe, especially in Belgium, France, Germany, and the Netherlands. Breeding of bromeliads in the United States among breeders, collectors, and hobbyists started in the early 1990s. Henry Nehring (1853–1929) described bromeliads grown in Florida. Julian Nally and Nat Deleon were among the earlier breeders who developed many hybrids in the United States. Most hybrids are selected from progenies of interspecific crosses. Intergeneric crosses between Cryptanthus beuckerii and Billbergia nutans have been successful, but the flowers have no ornamental value. Thousands of hybrids have been developed in the last 20 years. The Bromeliad Society International produced its first checklist of hybrids entitled International Checklist of Bromeliad Hybrids. The Bromeliad

276

R. HENNY AND J. CHEN

Cultivar Registry (BCR), compiled by Don A. Beadle (1998), lists all registered hybrids. C. Calathea (Marantaceae) Calathea is the largest genus in the family Marantaceae and is composed of 100 species native to tropical America in moist to swampy forest habitats (Bailey and Bailey 1976). Most calatheas are grown for their brilliant patterns of leaf color, texture, and elegance. Only C. crocata is produced for both its foliage and erect orange-red flowers. The showy portion of the flower usually consists of sterilized, often petaloid staminodes, and pollination is performed by bees (Kennedy 1973). Most species are crosspollinated, with a few that are truly cleistogamous such as C. panamensis (Kennedy 1973). Reported chromosome numbers include 2n = 18 for C. albertii, 2n = 16 for C. cylinerica, and 2n = 22, 24, 26, and 28 for C. nigricans, C. picturata, C. leucostachys, and C. musaica, respectively (Bisson et al. 1968; Mahanty 1970). Calathea propagation is traditionally from division, but cultivars of C. picturata, C. vandenheckei, and C. argentea can be propagated from seeds. Some cultivars are now propagated through tissue culture. In 1975, only three species (C. insignis, C. makoyana, and C. roseo-picta) were commonly grown (Table 6.3). Now, more than 12 species are in cultivation. The collection and introduction of new species not only expanded the number of cultivars in the foliage plant trade but also greatly broadened the genetic resources of Calathea for breeding. Marantaceae expert Dr. Helen Kennedy discovered 11 new species of Calathea in Panama and Costa Rica (Kennedy 1973). Interspecific hybridization by plant hobbyists and collectors have contributed significantly to the cultivar development. Calathea ‘Royale’ is a hybrid selected from a cross of C. roseo-picta and C. veitchiana. The leaves have a bluish-green upper surface and a yellow to silver band fans around the margin, and a light green pattern spreads along the midrib. The lower leaf surface is rich burgundy. Other popular interspecific hybrids include C. ‘Odora’, C. ‘Corona’, and C. ‘Medallion’. At least three cultivars (‘Cora’, ‘Silvia’, and ‘Angela’) were developed by selection of sports from C. rose picta by Magdalena J. M. van Rijn of the Netherlands. Calathea now is largely propagated through tissue culture. This has led to a number of new cultivars originating as somaclonal variants that were selected by tissue culture companies (Twyford Plant Laboratories, Inc., and Agri-Starts, Inc.) located in Apopka, Florida. These new cultivars include ‘Dottie’, ‘Saturn’, ‘Eclipse’, ‘Rosy’, ‘Artic Blush’, ‘Helen’, ‘Corona’, ‘Cynthia’, ‘Loeneri’, ‘Maria’, ‘Picta Royale’, ‘Rosy Roseo Picta’, ‘Tigrinum’, and ‘Wilson’s Princep’.

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

277

D. Dracaena (Dracaenaceae) Dracaena encompasses 60 species, all of which, except D. americana, are indigenous to tropical Africa and Asia (Hutchinson 1986). Hutchinson (1959) assigned Dracaena to Agavaveae, but Takhtajan (1980) placed it in a separate family, Dracaenaceae. Recent phylogenetic analyses using internal transcribed spacer (ITS) rDNA sequences of 40 taxa in Agavaceae justified this placement in Dracaenaceae (Bogler and Simpson 1996). Most species of Dracaena develop inflorescences consisting of loose umbels or clusters of greenish-white or cream-colored flowers, sometimes delightfully scented. The panicle is terminal and bracteate with two or more flowers per bract. Flowers are bisexual and small. Crosspollination occurs naturally in the wild. The chromosome number of D. draco is 2n = 40, while the chromosome numbers for other Dracaena species are unknown (Borgen 1969). Plants are propagated by either tip or cane cuttings. Seven Dracaena species are grown mainly as foliage plants: D. cincta, D. deremensis, D. fragrans, D. marginata, D. reflex, D. sanderiana, and D. surculosa (godseffiana). These species are particularly favored by interiorscapers because of their diverse shapes, colors, forms, and growth habits and their ability to survive under low-light conditions for extended periods. There is no known organized breeding program devoted to Dracaena, and the increase of cultivars is predominately due to the selection of sports by commercial growers. Dracaena fragrans appears to sport frequently since six cultivars (‘Kanzi’, ‘Jelle’, ‘Lemon Surprise’, ‘Golden Coast’, ‘White Jewel’, and ‘Janet Craig Gomezii’) have been patented. Variability in the variegation patterns has also been exploited in D. deremensis. The cultivar ‘Warneckii’ has lanceolate leaves with milky thin bright green margins. ‘Bausei’ differs only in having a narrower center with the white bands closer together. Leaves of ‘Roehrs Gold’ have a broad yellow center, bordered by white lines and edged with green, and ‘Janet Craig’ is a green sport of this species. Other characteristics have been selected, such as the pendant corrugated leaves of ‘Lognii’. One interspecific hybrid, Dracaena masseffiana (D. massangeana × D. godseffiana), was developed in the 1800s. This hybrid is sterile, has almost no ornamental value, and is not grown commercially. We have treated this plant with colchicine in vitro but have not yet produced any fertile explants (Henny, unpublished research). One ex-plant was found with a very high degree of foliar variegation and is being evaluated for possible release (Henny, unpublished). Foliar variegation in sports of D. fragrans ‘Massangeana’ is chimeral and not seed transmitted. All offspring from self-pollination of this cultivar lack foliar variegation.

278

R. HENNY AND J. CHEN

E. Ferns (Polypodiaceae) Fern refers to a group of primitive plants belonging to the Division Pteridophyta that have specialized vascular systems but are seedless and flowerless. It has been estimated that 10,000 to 12,000 species and 230 to 250 genera of fern exist throughout temperate and tropical zones (Huxley 1994). The genera Adiantum, Asplenium, Cyrtomium, Davallia, Nephrolepis, Platycerium, and Pteris are cultivated as foliage plants. Gametophytic chromosome numbers (n) of Adiantum 30, 60, 90, 120, and 180 (Singh and Roy 1969; Verma and Goloknath 1967), Asplenium 36, 72, and 108 (Lovis 1968; Morzenti 1967), Cyrtomium 41, 82, and 123 (Mitui 1968), Davallia 40 (Dujardin and Tilquin 1971), Nephrolepis 41 (Roy et al. 1971), Pteris 29, 58, and 87 (Kurita 1967; Mitui 1968) were reported among respective species. Ferns are propagated by division, spores, or tissue culture. Hybridization of different fern taxa may occur naturally. For example, Asplenium trichomanes subsp. quadrivalens is likely derived from taxa identical or close to Asplenium trichomanes subsp. trichomanes and A. trichomanes subsp. inexpectans (Vogel 1995). Some early breeding activities in ferns included a staghorn fern hybrid (Platycerium ‘Cass’ hybrid) developed by combining spores of P. grande, P. alcicorne, P. stemaria, and P. hillii. One intergeneric and several interspecific hybrids involving Asplenium exist; these are A. sollerense (A. majoricum × A. petrarchae), A. orelli (A. majoricum × A. trichomanes subsp. quadrivalens), A. litardierei (A. petrarchae subsp. petrarchae × A. trichomanes subsp. inexpectans), A. lessinese (A. fissum × A. viride), and Asplenoceterach barrancense (Asplenium majoricum × Ceterach orricinarum) (Wilfret and Sheehan 1981). Boston fern (Nephrolepis exaltata var. Bostoniensis) was the first foliage plant to be commercially propagated in vitro (Hartman and Zettler 1986). Since then, fern cultivar development has largely relied on selection of somaclonal variants from tissue culture. For example, more than 30 Boston fern cultivars are produced in Florida, most of which originated as somaclonal mutants from tissue culture. Some cultivars have also originated as sports. Collectively, fern cultivars used as foliage plants increased from 10 in 1975 to 54 in 1998–1999 in Florida (Table 6.3). Ceratopteris richardii, or C-Fern, has been used extensively as a model system for studying gametophyte development (Banks 1999). Methods used for selection of mutants in the C-Fern could be suitable for developing other new fern cultivars. F. Ficus (Moraceae) The genus Ficus, the Latin name of fig, encompasses more than 800 species. Figs are woody trees, shrubs, or vines native to Asia, Africa, Aus-

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

279

tralia, and Central and South America (Huxley 1994). Ficus species are either monoecious or dioecious with a chromosome number of 2n = 26 (Condit 1969). Each fig species is exclusively pollinated by a unique agaonid wasp (Agaonidae) species, whose offspring feed only on the tissue of the characteristic fig fruits. Recent mtDNA studies suggest that the fig-agaonid interaction might play an important role in fig species diversification (Cook and Lopez-Vaamonde 2001; Nason et al. 1998). Cultivated species are either produced for edible fruit or possess ornamental value. Fig species grown as foliage plants initially include F. benjamina, F. elastica, F. lyrata, and F. retusa, but several new fig species have been introduced in recent years. Ficus binnedijkii ‘Alii’, which means “king” in Hawaiian, was introduced to the U.S. continent in the late 1980s. Ficus microcarpa originates from southeast Asia and is a giant tree in its native habitat and invasive in some introduced regions, but is cultivated as a miniature bonsai-like plant. Ficus pumila, a vine-type fig native to southeast Asia grown in hanging baskets, was introduced in the early 1980s. Ficus salicifolia, the willow leaf fig, originates from Indonesia and was introduced in the late 1990s. Documented breeding activities on figs have only been reported on F. carica, a fruit tree fig (Storey 1975). There are no known breeding programs devoted to improvement of Ficus for use as foliage plants. New fig cultivars mainly come from the selection of sports from mutations or somaclonal variants. For example, F. benjamina ‘Monique’ and ‘Wiandi’ were sports of ‘Exotica’ and ‘Natasha’, respectively, and were selected by Huub van Diemer in Holland. Ficus benjamina ‘Indigo’ and ‘Midnight’ were sports of F. benjamina ‘Exotica’, selected by Jan van Geest in Holland. These new cultivars were introduced to the United States by Miami Agri-Starts, Inc., Homestead, Florida. Additional new Ficus cultivars include F. benjamina ‘Midnight Princess’ and ‘Too Little’; F. binnendijkii ‘Alii’, ‘Amstel King’; and F. elastica ‘Cabernet’, ‘Sylvie’, and ‘Melany’. These new cultivars are not only esthetically appealing but also perform much better in interior low-light environments than F. benjamina ‘Common’ (Chen et al. 2001c). In 1975, two F. benjamina, four F. elastica, one F. lyrata, and one F. retusa cultivars were in the Florida foliage plant trade. Currently, cultivars of F. benjamina, F. elastica, F. lyrata, and F. retusa number fourteen, five, four, and two, respectively. G. Hedera (Araliaceae) The genus Hedera, commonly called ivy, is native to Europe, northern Africa, and western Asia (Huxley 1994). As stressed previously, English ivy (H. helix) is probably the most significant species native to temperate climate (Rose 1996) and used extensively as a foliage plant by

280

R. HENNY AND J. CHEN

commercial growers worldwide. Algerian ivy (H. canariensis) is a distant second. Two noticeable characteristics of ivy are its evergreen leaves and the fact that it develops both juvenile and adult foliage. The juvenile leaves, those borne by the plant in its creeping or early climbing stages, are usually three to five lobed in H. helix. The adult leaves are produced on stiff, nonclimbing and rootless stems and are usually elliptic-lanceolate and not lobed. Flowers are produced on globose umbels, sometimes solitary but usually in compound panicles, each umbel carrying 10–15 flowers. Pollination is accomplished by flies, wasps, and bees (Rose 1996). Hedera helix, H. azorica, H. maroccana, H. nepalensis, and H. rhombea are diploid species with the chromosome number of 2n = 48; H. canariensis var. algeriensis and H. hibernica are tetraploid, 2n = 96 (Rose 1996). Hedera helix in the foliage plant industry is grown primarily for its juvenile leaves, which are variegated or have different shapes. According to the American Ivy Society (Naples, Florida), ivy leaf shapes can be classified into nine categories in reference to the Pierot System (Pierot 1974): variegated (V), bird’s foot (BF), fan (F), curlies (C), heart-shapes (H), miniature (M), ivy-ivies (I), adult (A), and oddities (O) (www.ivy.org). Plants with different leaf shape and variegation patterns, if stable in propagation, could potentially become new cultivars. The American Ivy Society is the International Registration Authority for new ivy cultivar registration. English ivy is propagated by cutting; tissue culture has not been used commercially. Rose (1996) listed more than 200 cultivars of H. helix varying in leaf sizes, shapes, colors, and variegation patterns; all were selected from sports. More than 40 cultivars of English ivy are grown in Florida. H. Palms (Arecaceae) Palms are woody monocotyledons, consisting of 200 genera and about 2,600 species (Jones 1995) and mainly distributed throughout subtropical and tropical regions of the world (Huxley 1994). Palms produce unisexual or bisexual flowers on one plant or on separate plants. Most palms flower regularly each year, and the transference of pollen from the stamens to the stigma is via insect vectors (beetles or small bees). Palms can be propagated sexually from seed or asexually by division or tissue culture. Palms are excellent decoration plants for both indoor and outdoor environments. The genera that have been used for interior plantscaping include Chamaedorea, Chrysalidocarpus, Howea, Phoenix, and Rhapis. Chamaedorea, commonly known as parlor or bamboo palm, is native to Central and South America. At least eight species (C. cataractarum, C.

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

281

elegans, C. erumpens, C. geonomiformis, C. metallica, C. microspadix, C. seifrizii, and C. tepejilote) are currently used as foliage plants for interiorscaping. Chrysalidocarpus lutescens, also called areca palm, is native to Madagascar and is the most widely grown palm for foliage use. Howea is from Lord Howe Island, off the eastern coast of Australia. Howea forsterana is the species in the foliage plant trade. Phoenix was the ancient Greek name for the date palm. It is native to Africa and Asia, with two species, P. roebelenii and P. reclinata, used mostly as interior foliage plants. Rhapis (Lady palm) has been cultivated since the 1600s. Lady palms are distributed naturally mainly in China, Laos, Vietnam, and Thailand. Rhapis excelsa is the principal species grown in Florida, whereas R. humilis is more commonly grown in cooler climates. All the aforementioned palm species bear seeds at maturity except for Rhapis humilis, which rarely produces seed. Palm flowers are relatively large and easy to handle for breeding purposes. The number of palm cultivars produced in Florida remains relatively constant at around 20. Extensive research has been conducted on the oil palm (Elaeis guineensis), and significant progress has been made on cultivar development (Soh et al. 2003). These include the use of molecular markers to determine genetic diversity (Barcelos et al. 1998; Mayes et al. 2000), development of inbred lines, test of combining ability, production of interspecific hybrids, introduction of palm germplasm for improving disease resistance and other agronomic traits, tissue culture, embryogenesis, microspore, and anther culture to generate haploid or double haploid plants (Breure and Verdooren 1995; Chavez and Sterling 1991; Richardson 1995). These methods suggest potential avenues for research in breeding new palm cultivars for use as foliage plants.

VI. FUTURE PROSPECTS The wholesale value of foliage plants in the United States rose from $13 million in 1949 to $585 million in 2001 (Smith 1980; USDA 2002). As discussed above, cultivar development has played an important role in the steady growth of the foliage plant industry. New cultivar releases and new plant introduction provide products for new uses, thus increasing production. The future of the industry is bright because interiorscape with foliage plants has become an integral part of contemporary design in daily life. With increasing desire by consumers for novelty and new uses of living specimens in interior decoration, more new foliage plants and new cultivars should be developed and released. To meet the upcoming challenges, the following issues may need to be considered:

282

R. HENNY AND J. CHEN

A. Germplasm Collection and Conservation Foliage plants are important members of tropical and subtropical ecosystems; their collections should follow the provisions of the Convention on Biological Diversity at Rio de Janeiro (Anon. 1992). Germplasm of foliage plants are primarily maintained by plant collectors and botanical gardens. It would be very beneficial to breeders and researchers if the Ornamental Plant Germplasm Center (OPGC) would become actively involved in foliage plant conservation and provide researchers and breeders with convenient access. In addition, the term “foliage plant” refers to a very diverse group of plant genera, many of which have not been the subjects of detailed botanical research. Developing uniform standards for foliage plant evaluation regarding taxonomy, cytology, fertility, and crossability, as well as standardized procedures in plant introduction, are needed. B. Breeding Interspecific hybridization has been and will still be an important method of new cultivar development. However, other breeding methods in foliage plants should be considered. Some possibilities would include utilization of hybrid vigor obtained from crossing homozygous lines produced through in vitro culture of anther or ovaries, and population improvement by recurrent selection. Induced mutation using either physical or chemical agents for foliage plant cultivar development should be exploited, since many other ornamental plant cultivars are derived from this method (Langton 1987). Additionally, there are almost no breeding activities in many important foliage plant genera such as Dracaena, Ficus, Hedera, and various ornamental palms. C. Somaclonal Variant Selection The evidence that a large number of foliage plant cultivars are selected from somaclonal variants is probably the best example of how this new source of genetic variation is used in cultivar development. Research is needed to document the causes and frequencies of somaclonal variant occurrence using different explants and to develop procedures for generating the variants in different species. With more foliage plants becoming tissue-cultured, more cultivars will be selected from somaclonal variants. It has been reported that variant occurrence frequency can be accelerated when explants or calli are treated by mutagenic agents

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

283

(Maliga et al. 1981; van Harten 1998). It should be possible to increase variant occurrence frequency by introducing mutagenic compounds into culture media. D. Transgene Technology As stressed earlier, transgene technology could be particularly promising in foliage plant cultivar development since most foliage plants can be regenerated in tissue culture. It is possible that genes proven to be economically significant in improving agronomic crop traits after being transformed can be directly used for cultivar improvement in foliage plants. McCown (1997) listed several genes for improving plant forms. The expression of iaaM and iaaH for overproduction of indole-3-acetic acid (IAA) could lead to the increase of apical dominance with short internodes and small narrow leaves. On the other hand, transformation of iaaL gene could reduce the level of endogenous free IAA; as a result, more basal shoots could be formed, giving a dwarf and compact appearance. Isopentenyl transferase (ipt) gene transformation has been shown to reduce apical dominance, increase basal shoots, and more importantly, increase chlorophyll content and delay leaf senescence. Flower shapes and colors are important traits to Anthurium, Aphelandra, bromeliads, and Spathiphyllum. With the increased understanding of flower development and biochemistry of pigment biosynthesis, transgene technology may become an important tool to manipulate shapes and colors of these foliage plants’ flowers. Transgene technology should be equally useful for engineering foliage plants in resistance to diseases and insects in both production and interiorscape.

LITERATURE CITED Anon. 1992. Handbook on convention on biological diversity at Rio de Janeiro. Secretariat of the Convention. Bailey, L. H., and E. Z. Bailey. 1976. Hortus III. Macmillan, New York. Baker, J. R. 1994. Insect and related pests of flowers and foliage plants. North Carolina Cooperation Extension Service. North Carolina State Univ., Raleigh. Banks, J. A. 1999. Gametophyte development in ferns. Annu. Rev. Plant Physiol. Plant Mol. Biol. 50:163–168. Barcelos, E., G. Second, F. Kahn, P. Amblard, P. Lebrun, and M. Seguin. 1998. Molecular markers applied to the analysis of genetic diversity and to the biogeography of Elaeis. p. 191-201. In: Proc. of the Seminar on Evolution, Variation, and Classification of Palms. Memoirs of New York Botanical Garden. Beadle, D. A. 1998. The Bromeliad Cultivar Registry. The Bromeliad Society Intl., Venice, FL.

284

R. HENNY AND J. CHEN

Behe, B., and R. Nelson. 1999. Consumer preferences for geranium flower color, leaf variegation, and price. HortScience 34:740–742. Benzeng, D. H. 2000. Bromeliaceae: Profile of an adaptive radiation. Cambridge Univ. Press, Cambridge, UK. Betrock, I. 1996. Plantfinder. Betrock Information System, Hollywood, FL. Bisson, S., S. Guillemet, and J. L. Hamel. 1968. Contribution a l’etude caryo-taxnomique des Scitaminees. Mem. Mus. National Hist. Naturelle Ser. B. 18:59–133. Bogler, D. J., and B. B. Simpson. 1996. Phylogeny of Agaraceae based on ITS rDNA sequence variation. Am. J. Bot. 83:1225–1235. Borgen, L. 1969. Chromosome number of vascular plants from the Canary Islands, with special reference to the occurrence of polyploidy. Nytt Magasin Botanik 16:18–121. Bouman, H., and G. De Klerk. 1997. Somaclonal variation. p. 165–183. In R. L. Geneve, J. E. Preece, and S. A. Merkle (eds.), Biotechnology of ornamental plants. CAB Intl., Wallingford, UK. Breure, C. J., and L. R. Verdooren. 1995. Guidelines for testing and selecting parent palms in oil palm. Practical aspects and statistical methods. ASD Oil Palm Papers 9:1–68. Brown, F. B. 2001. The amazing Aglaonema houseplant to the world. Valkaria Tropical Gardens, Valkaria, FL. Buiattia, M., and F. Gimelli. 1993. Somaclonal variation in ornamentals. p. 5–24. In: T. Schiva and A. Mercuri (eds.), Proc. XVII Symp. of the European Assoc. Plant Breeding—Creating variation in ornamentals. Sanremo, Italy. Chase, A. R. 1981. Common diseases of foliage plants. Florida Foliage 7:39–56. Chase, A. R. 1997. Foliage plant diseases: Diagnosis and control. APS Press. St. Paul, MN. Chavez, C. E., and F. Sterling. 1991. Variation in the total of unsaturated fatty acids in oils extracted from different oil palm germplasms. ASD Tech. Bul. (Costa Rica) 3:5–8. Chen, J., R. J. Henny, R. D. Caldwell, and C. A. Robinson. 2001a. Aglaonema cultivar differences in resistance to chilling temperatures. J. Environ. Hort. 19:198–202. Chen, J., R. J. Henny, and D. B. McConnell. 2002. Development of new foliage plant cultivars. p. 446–452. In J. Janick and A. Whipkey (eds.), Trends in new crops and new uses. Timber Press, Inc. Portland, OR. Chen, J., R. J. Henny, D. B. McConnell, and R. D. Caldwell. 2003. Gibberellic acid affects growth and flowering of Philodendron ‘Black Cardinal’. Plant Growth Regulation (in press). Chen, J., R. J. Henny, D. B. McConnell, and T. A. Nell. 2001b. Cultivar differences in interior performances of acclimatized foliage plants. Acta Hort. 543:135–140. Chen, J., R. J. Henny, C. A. Robinson, T. Mellich, and R. D. Caldwell. 1999. Potted Anthurium: An interior-flowering foliage plant. Proc. Fla. State Hort. Soc. 112:280–281. Chen, J., T. A. Nell, R. J. Henny, C. A. Robinson, and R. D. Caldwell. 2001c. Light levels in influencing production and subsequent interior performances of Ficus cultivars. HortScience 36:600 (Abstr.). Chen, J., L. S. Osborne, R. J. Henny, R. D. Caldwell, and C. A. Robinson. 2001d. Evaluating Anthurium cultivar resistance to mites and thrips. Greenhouse Product News 11(2):16–17, 94. Chen, J., L. Qu, R. J. Henny, C. A. Robinson, and R. D. Caldwell. 2001e. Chilling injury in tropical foliage plants: I. Spathiphyllum. Florida Cooperative Extension Service, IFAS, Univ. of Florida Edis Pub. ENH841. Condit, I. J. 1969. Ficus: The exotic species. Univ. of California Division of Agricultural Sciences. Conover, C. A. 1980. The shipping environment. Foliage Dig. 3(2):4–5.

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

285

Conover, C. A., and R. T. Poole. 1974. Foliage collapse of Dieffenbachia picta ‘Perfection’ during propagation. SNA Nursery Res. J. 1(1):1–6. Cook, J. M., and C. Lopez-Vaamonde. 2001. Fig biology: Turing over new leaves. Trends in Ecology and Evolution 16:11–13. Croat, T. B. 1982. A revision of Syngonium (Araceae). Ann. Missouri Bot. Gard. 68: 565–651. Croat, T. B. 1983. A reversion of the genus Anthurium (Araceae) of Mexico and Central America. Part I. Mexico and Middle America. Ann. Missouri Bot. Gard. 70:211–420. Croat, T. B. 1986. A reversion of the genus Anthurium (Araceae) of Mexico and Central America. Part II. Panama, Nonogr. Syst. Bot. Mo. Bot. Gard. 14:1–204. Croat, T. B. 1992. Species diversity of Araceae in Colombia: preliminary survey. Ann. Mo. Bot. Gard. 79:17–28. Croat, T. B. 1997. A revision of Philodendron subgenus Philodendron (Araceae) of Central America. Missouri Botanical Garden Press, St. Louis, MO. DeRoles, S. C., M. R. Boase, and I. Koncazk. 1997. Transformation protocols for ornamental plants. p. 87–119. In: R. L. Geneve, J. E. Preece, and S. A. Merkle (eds.), Biotechnology of ornamental plants. CAB Intl., Wallingford, UK. Dressler, R. L. 1980. A new name for the dwarf purple Anthurium. Aroideana 3:55. Dujardin, M., and J. P. Tilquin. 1971. In IOPB chromosome number reports XXXIII. Taxon 20:609–614. Eeckhaut, T., S. Werbrouck, J. Dendauw, E. V. Bockstaele, and P. DeBergh. 2001. Induction of homozygous Spathiphyllum wallisii genotypes through gynogenesis. Plant Cell, Tissue Organ Culture 67:181–189. Evans, D. A. 1989. Somaclonal variation—genetic basis and breeding applications. Trends Gen. 5:46–50. Fehr, W. R. 1987. Principles of cultivar development. Vol. 1. Theory and technique. Macmillan, New York. Fisher, J. B. 1980. Gibberelin-induced flowering in Cordyline (Agavaceae). J. Expt. Bot. 31:731–735. FNGA (Florida Nurserymen and Growers Association) Locator 1975, 1988–1989, 1998–1999, 1999–2000, 2000–2001. Orlando, FL. Griffith, L. P. 1998a. Roots of the foliage plant industry. Grower Talks (July):74–80. Griffith, L. P. 1998b. Tropical foliage plants: A grower’s guide. Ball Pub., Batavia, IL. Hamlen, R. A., D. W. Dickson, D. E. Short, and D. E. Stokes. 1981. Insects, mites, nematodes, and other pests. p. 428–479. In. J. N. Joiner (ed.), Foliage plant production. PrenticeHall, Englewood Cliffs, NJ. Hansen, G., and M. S. Wright. 1999. Recent advances in the transformation of plants. Trends Plant Sci. 4:226–231. Harrison, B. J., and J. R. S. Fincham. 1964. Instability at the PAL locus in Antirrhinum majus. 1. Effects on frequencies of somatic and germinal mutation. Heredity 19:237–258. Hartman, R. D., and F. W. Zettler. 1986. Tissue culture as a plant production system for foliage plants. p. 293–299. In: R. H. Zimmerman, R. J. Griesbach, F. A. Hammerschlag, and R. H. Lawson (eds.), Tissue culture as a plant production system for horticultural crops. Martinus Nijhoff Pub., Dordrecht, The Netherlands. Henderson, A. 1986. A review of pollination studies in the Palmae. Bot. Rev. 52:221–259. Henny, R. J. 1980. Relative humidity affects in vivo pollen germination and seed production in Dieffenbachia maculata ‘Perfection’. J. Am. Soc. Hort. Sci. 105:546–548. Henny, R. J. 1982. Inheritance of foliar variegation in two Dieffenbachia cultivars. J. Hered. 73:384.

286

R. HENNY AND J. CHEN

Henny, R. J. 1983a. Inheritance of foliar variegation in three Aglaonema species. J. Hered. 74:475–476. Henny, R. J. 1983b. Inheritance of the white foliar midrib in Dieffenbachia and its linkage with the gene for foliar variegation. J. Hered. 74:483–484. Henny, R. J. 1985. In vivo pollen germination of Aglaonema affected by relative humidity. HortScience 20:142–143. Henny, R. J. 1986a. Inheritance of foliar variegation in Dieffenbachia maculata ‘Camille’. J. Hered. 77:285–286. Henny, R. J. 1986b. Single locus, multiallelic inheritance of foliar variegation in Aglaonema. J. Hered. 77:214–215. Henny, R. J. 1988. Pollen germination in Aglaonema flowers of different ages. HortScience 23:218. Henny, R. J. 1995. Stimulating flowering of ornamental aroid genera with gibberellic acid: A review. Proc. Fla. State Hort. Soc. 108:23–24. Henny, R. J. 1999. ‘Red Hot’ Anthurium. HortScience 34:153–154. Henny, R. J. 2000. Breeding ornamental aroids. p. 121-132. In: D. J. Callaway and M. B. Callaway (eds.), Breeding ornamental plants. Timber Press, Portland, OR. Henny, R. J., and J. Chen. 2001. ‘Golden Bay’ Aglaonema. HortScience 36:1142–1143. Henny, R. J., J. Chen, and D. J. Norman. 2003a. ‘Emerald Bay’ and ‘Diamond Bay’ Aglaonema. HortScience (in press). Henny, R. J., D. J. Norman, and J. Chen. 2003b. Progress in ornamental aroid breeding research. VIII Intl Aroid Conference. Missouri Botanical Garden, St. Louis, MO. (in press). Henny, R. J., D. J. Norman, and M. E. Kane. 1999. Gibberellic acid-induced flowering of Syngonium podophyllum Schott. ‘White Butterfly’. HortScience 34:676–677. Henny, R. J., R. T. Poole, and C. A. Conover. 1989. ‘Starry Nights’ Dieffenbachia. HortScience 24:522–524. Henny, R. J., R. T. Poole, and C. A. Conover. 1992a. ‘Silver Bay’ Aglaonema. HortScience 27:1238. Henny, R. J., R. T. Poole, and C. A. Conover. 1992b. ‘Star White’ Dieffenbachia. HortScience 27:82–83. Horsch, R. B., R. T. Fraley, S. C. Rogers, P. R. Sanders, A. Lloyd, and N. Hoffman. 1984. Inheritance of functional foreign genes in plants. Science 223:496–498. Hutchinson, J. 1959. The families of flowering plants. Vol. 2. Monocotyledonous. Clarendon Press, Oxford, UK. Hutchinson, J. 1986. Dracaenas in west Africa. Clarendon Press, Oxford, UK. Huxley, A. 1994. The new Royal Horticultural Society dictionary of gardening. Macmillan, London. James, C. 1997. Global status of transgenic crops in 1997. ISAAA Briefs No. 5, ISAAA, Ithaca, NY. Jones, D. L. 1995. Palms throughout the world. Smithsonian Institution Press, Washington, DC. Jones, G. E. 1957. Chromosome numbers and phylogenetic relationship in the Araceae. Ph.D. Diss., Univ. Virginia, Charlottesville. Kamemoto, H. S. 1981. Anthurium breeding in Hawaii. Ariodeana 4:77–86. Kamemoto, H. S., and A. R. Kuehnle. 1996. Breeding Anthuriums in Hawaii. Univ. Hawaii Press, Honolulu. Kamemoto, H. S., and R. D. Sheffer. 1978. A new species hybrid, Anthurium scherzerianum × Anthurium wendlingerii. HortScience 13:177–179. Kennedy, H. 1973. Notes on Central American Marantaceae I. New species and records from Panama and Costa Rica. Ann. Missouri Bot. Gard. 60:413–426.

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

287

Kennedy, H. 1974. Pollination and systematics of the “closed-flowered” species of Calathea (Marantaceae). Ph.D. Thesis, Univ. of California, Davis. Kirk, J. T. O. and R. A. E. Tilney-Bassett. 1978. The plastids—their chemistry, structure, growth and inheritance. Elsevier/North Holland Biomedical Press, Amsterdam. Kobayashi, R. S., J. L. Brewbaker, and H. Kamemoto. 1987. Identification of Anthurium andraeanum cultivars by gel electrophoresis. J. Am. Soc. Hort. Sci. 112:164–167. Kuanprasert, N., and A. R. Kuehnle. 1999. Fragrance quality, emission, and inheritance in Anthurium species and hybrids. Aroideana 22:48–62. Kuehnle, A. R. 1997. Progress in protoplast isolation and culture from axenic tissue of hybrid Anthurium. Aroideana 20:29–36. Kuehnle, A. R., and N. C. Sugii. 1991. Introduction of tumors in Anthurium andraeanum by Agrobacterium tumefaciens. HortScience 26:1325–1328. Kuehnle, A. R., F. C. Chen, and N. C. Sugii. 1992. Somatic embryogenesis and plant regeneration in Anthurium andraeanum hybrids. Plant Cell Rep. 11:438–442. Kuehnle, A. R., F. C. Chen, and N. C. Sugii. 2001. Transgenic Anthurium. p. 4–15. In: Y. P. S. Bajaj (ed.), Biotechnology in agriculture and forestry. Vol. 48. Transgenic crop III. Springer-Verlag, Berlin-Heidelberg. Kurita, S. 1967. Chromosome numbers of some Japanese ferns. Annu. Rep. Foreign Students. Col. Chiba Univ. 2:57–61. Langton, F. A. 1987. Breeding for improved ornamental plants. p. 159–180. In: A. J. Abbott and R. K. Atkin (eds.), Improving vegetatively propagated crops. Academic Press, New York. Larkin, P. J., and W. R. Scowcroft. 1981. Somaclonal variation—a novel source of variation from cell culture for plant improvement. Theor. Appl. Genet. 60:197–214. Lovis, J. D. 1968. Artificial reconstruction of a species of fern Asplenium adulterinum. Nature 217:1163–1165. Lowe, E. J. 1861. Beautiful leaved plants. Goombridge and Sons, London. 144 p. Madison, M. 1979. Additional observation on ant-gardens in Amazonas. Selbyana 5:107–115. Mahanty, H. K. 1970. A cytological study of the Zingiberales with special reference to their taxonomy. Cytologia 35:13–49. Maliga, P., V. A. Sidorov, A. Cseplo, and L. Menczel. 1981. Induced mutations: A tool in plant research. p. 339–352. Intl. Atomic Energy Agency, Vienna. Maluszynski, M., L. Van Zanten, A. Ashri, H. Brunner, B. Ahloowalia, F. P. Zapata, and E. Weck. 1995. Mutation techniques in plant breeding. p. 489–504. In: Induced mutations and molecular techniques for crop improvement. Proc. FAO/IAEA Symp., Vienna. Manaker, G. H. 1997. Interior plantscapes: Installation, maintenance, and management. 3rd ed. Prentice-Hall, Upper Saddle River, NJ. Marchant, C. J. 1967. Chromosome evolution in the Bromeliaceae. Kew Bul. 21:161–168. Marchant, C. J. 1970. Chromosome variation in Araceae I: Pothoeae to Stylochitoneae. Kew Bul. 24:315–322. Marchant, C. J. 1971. Chromosome variation in the Araceae II. Richardieae to Colocasieae. Kew Bul. 25:27–56. Marcotrigiano, M. 1997. Chimeras and variegation: Patterns of deceit. HortScience 32:773–784. Matsumoto, T. K., and A. R. Kuenle. 1997. Micropropagation of Anthurium. p. 14–29. In: Y. P. S. Bajaj (ed.), Biotechnology in agriculture and forestry. Vol. 40. High Tech and Micropropagation VI. Springer-Verlag, Berlin. Mayes, S., P. L. Jack, and R. H. V. Corley. 2000. The use of molecular markers to investigate the genetic structure of an oil palm breeding programme. Heredity 85:288–293.

288

R. HENNY AND J. CHEN

Mayo, S. J., J. Bogner, and P. C. Boyce. The genera of Araceae. Royal Botanic Gardnes, Kew. 370 p. McColley, R. H., and H. N. Miller. 1965. Philodendron improvement through hybridization. Florida State Hort. Soc. 78:409–415. McConnell, D. B., R. W. Henley, and C. B. Kelly. 1989. Commercial foliage plants: Twenty years of changes. Proc. Fla. State Hort. Soc. 102:297–303. McCown, B. H. 1997. Biotechnology approaches to modifying plant form. p. 199–212. In: R. L. Geneve, J. E. Preece, and S. A. Merkle (eds.), Biotechnology of ornamental plants. CAB Intl. Wallingford, UK. McWilliams, E. I. 1974. Evolutionary ecology. p. 33–55. In: L. B. Smith and R. J. Downs (eds.), Pitcairnioideae (Bromeliaceae). Flora Neotropica Monograph 14, Part 1. Hafner Press, New York. Mitui, K. 1968. Chromosomes and speciation in ferns. Sci. Rep. Tokyo Kyoiku Daigaku Sect. B 13:285–333. Morzenti, V. M. 1967. Asplenium plenum: A fern which suggests an unusual method of species formation. Am. J. Bot. 54:1061–1068. Nason, J. D., E. A. Herre, and J. L. Hamrick. 1998. The breeding structure of a tropical keystone plant resource. Nature 391:685–687. Neal, M. C. 1965. Gardens of Hawaii. Bishop Museum Press, Honolulu. Neilson-Jones, W. 1969. Plant chimeras. 2nd ed. Methuen, London. Norman, D. J., and R. J. Henny. 1999. Conquering diseases of tropical ornamentals. p. 71–83. In: K. M. Heinz (ed.), Proc. Fifteenth Conference on Insect and Disease Management on Ornamentals. Society of American Florists. Norman, D. J., R. J. Henny, and J. M. F. Yuen. 1999a. Disease resistance of fourteen pot Anthurium cultivars to Xanthomonas campestris pv. dieffenbachia. HortScience 34:721–722. Norman, D. J., R. J. Henny, and J. M. F. Yuen. 1999b. Resistance in twenty-two cultivars of Spathiphyllum to cylindrocladium root rot. HortTechnology 9:68–69. Norman, D. J., R. J. Henny, and J. M. F. Yuen. 2002. Screening for resistance to Myrothecium leaf spot among Syngonium species and cultivars. HortScience (in press). Novak, F. J. 1991. In vitro mutation system for crop improvement. p. 327–342. In: Plant Mutation Breeding for Crop. Proc. FAO/IAEA Symposium, Vienna. Perry, F. 1979. Beautiful leaved plants. Scolar Press, London, UK. Peschke, V. M., and R. L. Phillips. 1992. Genetic implications of somaclonal variation in plants. Adv. Genet. 30:41–75. Pfitzer, P. 1957. Chromosomenzahlen von Araceen. Chromosoma 8:436–446. Pierot, S. 1974. The ivy and care of ivy and ivy topiary. Macmillan, New York. Poole, R. T., and C. A. Conover. 1993. Effects of low temperature storage on quality of twenty-one foliage plants. Univ. Fla., IFAS, CFREC-Apopka Res. Rep. RH-93-6. Pratt, C. 1983. Somatic selection and chimeras. p. 172–173. In: J. N. Moore and J. Janick (eds.), Methods in fruit breeding. Purdue Univ. Press, West Lafayette, IN. Qu, L., J. Chen, R. J. Henny, R. D. Caldwell, and C. A. Robinson. 2000. Response of Spathiphyllum cultivars to chilling temperatures. Proc. Fla. State Hort. Soc. 113: 165–169. Qu, L., J. Chen, R. J. Henny, Y. Huang, R. D. Caldwell, and C. A. Robinson. 2002. Thidiazuron promotes adventures shoot regeneration from pothos (Epipremnum aureum) leaf and petiole explants. In Vitro Cell. Dev. Biol. Plant 38:268–271. Ranamukhaarachchi, D. G., R. J. Henny, C. L. Guy, and Q. B. Li. 2001. DNA fingerprinting to identify nine Anthurium pot plant cultivars and examine their genetic relationship. HortScience 36:758–760.

6. CULTIVAR DEVELOPMENT OF ORNAMENTAL FOLIAGE PLANTS

289

Reark, J. B. 1951. Cultivation of the genus Alocasia in Florida. Proc. Fla. State Hort. Soc. 66:326–331. Richardson, D. L. 1995. The history of oil palm breeding in the United Fruit Company. ASD Oil Palm Papers, 11:1–23. Rose, P. 1996. The gardener’s guide to growing ivies. A David and Charles Book, Devon, UK. Roy, R. P., B. M. B. Sinha, and A. R. Sakya. 1971. Cytology of some ferns of Kathmandu valley. Brit. Fer Gaz. 10:193–199. Samyn, G., and F. Thomas. 1997. Thirty years of bromeliad breeding at the research station for ornamental plant growing at Melle, Belgium. J. Bromeliad Soc. 47:120–124. Sharma, A. K. 1970. Annual Report, 1967–1968. Res. Bul. Univ. Calcutta (Cytogenetics Lab) 2:1–50. Sheffer, R. D., and T. B. Croat. 1983. Chromosome numbers in the genus Anthurium (Araceae) II. Am. J. Bot. 70:857–871. Sheffer, R. D., and H. Kamemoto. 1976a. Chromosome numbers in the genus Anthurium. Am. J. Bot. 63:74–81. Sheffer, R. D., and H. Kamemoto. 1976b. Cross compatibility in the genus Anthurium. J. Am. Soc. Hort. Sci. 101:709–713. Short, D. E., J. F. Price, R. W. Henley, and L. S. Osborne. 1999. Commercial interiorscape insect management suggestions. Florida Cooperative Extension Service, IFAS, Univ. Fla. Edis Pub. SS-ENY-914. Simmonds, N. W. 1979. Principles of crop improvement. Longman, London. Singh, J. B., and S. K. Roy. 1969. Cytotaxonomy of Adiantum in India. Nucleus 12:173–177. Smith, C. N. 1980. Evolution of the Florida foliage plant industry. Proc. Fla. State Hort. Soc. 93:208–210. Smith, C. N., and E. F. Scarborough. 1981. Status and development of foliage plant industries. p. 1–39. In: J. Joiner (ed.), Foliage plant production. Prentice-Hall, Englewood Cliffs, NJ. Smith, C. N., and J. R. Strain. 1976. Market outlets and product mix for Florida foliage plants. Proc. Fla. State Hort. Soc. 89:274–278. Soh, A. L., G. Wong, T. Y. Hor, C. C. Tan, and P. S. Chew. 2003. Oil palm genetic improvement. Plant Breed. Rev. 22:165–219. Stewart, F. C. 1978. Ontogeny of the primary body on chimeral forms of higher plants. p. 133–160. In: S. Subtlety and I. M. Sussex (eds.), The clonal basis of development. Academic Press, New York. Storey, W. B. 1975. Figs. p. 568–589. In: J. Janick and J. N. Moore (eds.), Advances in fruit breeding. Purdue Univ. Press, West Lafayette, IN. Takhtajan, A. 1980. Outline of the classification of flowering plants. Bot. Rev. 46:223–359. Tischler, G. 1954. Das problem der basis-chromosomenzahlen bei den angiospermengattungen und-familien. Cytologia 19:1–10. USDA (United States Department of Agriculture). 1998. Floriculture crops 1997 Summary. Washington, DC. USDA (United States Department of Agriculture). 1999. Floriculture crops 1998 Summary. Washington, DC. USDA (United States Department of Agriculture). 2002. Floriculture crops 2000 Summary. Washington, DC. Van Harten, A. M. 1998. Mutation breeding: Theory and practical application. Cambridge Univ. Press, Cambridge, UK. Veilleux, R. E., and A. A. T. Johnson. 1998. Somaclonal variation: Molecular analysis, transformation interaction, and utilization. Plant Breed. Rev. 16:229–268.

290

R. HENNY AND J. CHEN

Verma, S. C., and M. Goloknath. 1967. In IOPB chromosome number reports X. Taxon 16:146–157. Vogel, J. C. 1995. Multiple origins in polyploids in European Asplenium (Pteridophyta). Ph.D. Diss. Cambridge Univ., Cambridge, U.K. Wang, K. H., A. R. Kuehnle, and B. S. Sipes. 1998. In vitro tolerance and resistance to burrowing nematode, Radopholus similis, in Anthurium species. Euphytica 103:23–28. Wilfret, G. J. 1986. Inheritance of leaf shape and color patterns in Caladium (Araceae). HortScience 21:750. Wilfret, G. J., and T. J. Sheenhan. 1981. Development of new foliage plant cultivars. p. 126–136. In: J. N. Joiner (ed.), Foliage plant production. Prentice-Hall, Englewood Cliffs, NJ. Williams, C. A., and R. Dressler. 1976. Euglossine pollination of Spathiphyllum (Araceae). Selbyana 1:349–356. Wolverton, B. C. 1989. Interior landscape plants for indoor air pollution abatement. Interiorscape 8(6):37–62. Zettler, F. W., and M. M. Abo El-Nil. 1979. Mode of inheritance of foliage color in Caladium. J. Hered. 70:433–435

7 Preservation of Genetic Resources in the National Plant Germplasm Clonal Collections* Gayle M. Volk and Christina Walters National Center for Genetic Resources Preservation United States Department of Agriculture, Agricultural Research Service 1111 S. Mason Street, Fort Collins, Colorado 80521

I. INTRODUCTION II. CLONES AS GENETIC RESOURCES III. MAINTENANCE OF GENETIC DIVERSITY IN CLONAL COLLECTIONS A. Wild Germplasm B. Evaluation of Diversity 1. Phenotypic Diversity 2. Ecophysiological Diversity 3. Genetic Diversity C. Core Collections IV. CLONAL COLLECTIONS IN THE NPGS A. Active Collections B. Base Collections 1. Seed Storage 2. Pollen Storage 3. Vegetative Propagule Storage

*We appreciate the assistance of the following curators from NPGS active sites and clonal repositories who provided the data presented in Tables 7.1 and 7.2: John Bamberg, Phil Forsline, Ricardo Goenaga, L. J. Grauke, Barbara Hellier, Kim Hummer, Bob Jarret, Robert Krueger, Warren Lamboy, Alan Meerow, Brad Morris, Gary Pederson, Roy Pittman, Joseph Postman, Ray Schnell, Chuck Simon, Susan Stieve, David Tay, Mark Widrlechner, and Francis Zee. We thank the following NCGRP employees who assisted with information retrieval, data collection, and reviewing: Jennifer Crane, Julie Fleming, Lisa Hill, Julie Laufmann, Andrea Lawrence, Christopher Richards, Leigh Towill, and John Waddell. Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 291

292

G. VOLK AND C. WALTERS

V. CRYOPRESERVATION PRINCIPLES A. Freezing 1. Damage 2. Protection B. Desiccation 1. Damage 2. Protection VI. CRYOPRESERVATION: VARIABLES TO CONSIDER A. Explant B. Degree of Tolerance 1. Physiologically Achieved Tolerance 2. Externally Imposed Tolerance C. Dehydration D. Cooling Rates E. Recovery VII. APPLICATION OF CRYOPRESERVATION TECHNOLOGIES TO VEGETATIVE MATERIALS A. Germplasm Classification B. Cryopreservation Methods 1. Dormant Buds 2. Axenic Shoot Tips 3. Somatic Embryos VIII. CONCLUSIONS LITERATURE CITED

I. INTRODUCTION The availability of novel plant germplasm for breeding programs has been critical to the success of American agriculture. Breeding programs have used germplasm that is publicly available to develop high-yielding cultivars of rice, wheat, barley, sorghum, cotton, soybean, bean, pea, peanut, sugarbeet, and sunflower that exhibit remarkable levels of resistance to pests and disease (Shands and Weisner 1991, 1992; Sharma et al. 1999; Holbrook et al. 2000). Most geneticists keep plant materials on site for breeding purposes; however, their lifelong collections are often unappreciated and discarded at the time of retirement. The U.S. government developed a plant germplasm repository system in 1946 to protect and distribute genetic resources valuable to American agriculture. Germplasm accessions are acquired through donations (plant exchange) and trips (plant exploration) to find new materials from both national and international sources. The USDA-Agricultural Research Service-National Plant Germplasm System (USDA-ARS-NPGS) maintains approximately 450,000 accessions of economically important plant genera at about 25

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

293

ex situ locations (Shands et al. 1989; White et al. 1989; Eberhart and Bockelman 1994). A small part of the collection consists of voucher seeds or plants for patents and crop germplasm registrations. Consulting with Crop Germplasm Committees (CGC) within the United States, curators and scientists within the NPGS collaborate with germplasm experts from other country-specific germplasm systems and germplasm consortiums, as well as the seven International Germplasm Centers overseen by the Food and Agriculture Organization of the United Nations. NPGS curators play an important role by ensuring that plant germplasm collections remain accessible, viable, and healthy. Curators evaluate growth, reproductive habits and disease susceptibility, and other traits of interest. They provide collection information to the public and germplasm to bona fide scientists. The inherent value of a collection is based on many factors. Collections usually contain named cultivars that perform exceptionally under cultivated conditions and are the culmination of university or federal breeding programs. These accessions are invaluable for researchers screening superior germplasm for high yields under alternative growing conditions. Since these cultivars are highly selected, they generally represent a narrow gene pool base. A broader base of genetic diversity is obtained by collecting landraces or non-domesticated plants from the wild. These accessions may provide genes that are not available in elite lines and are valuable for those genes rather than the specific genotype. A single gene or series of genes may provide disease resistance or improved agricultural characteristics to cultivated lines. Desirable genes can be introgressed into elite cultivars to improve plant and yield characteristics through either traditional breeding or genetic engineering programs (Shands and Weisner 1991, 1992; Tanksley and McCouch 1997). The value of landraces and wild accessions has increased as native habitats have been destroyed and new genomic tools have enabled detection of valuable genes that are masked by undesirable phenotypes. The discovery of new genes and the potential to move genes across plant families has expanded the opportunities of breeding programs and novel uses for plants. Strategies to preserve the genetic composition of an accession are dependent upon the accession type (elite line versus wild), the reproductive biology of the species (self-fertile versus obligate outcrosser), and the heterogeneity of the accession. Many elite, self-compatible, inbred accessions (such as the grains) are produced from genetically identical seeds. Often, seeds resulting from controlled fertilization events within

294

G. VOLK AND C. WALTERS

these accessions are uniform and representative of the parent plants. The genetic integrity of elite, obligate outcrossing accessions (such as alfalfa, watermelon) is maintained by controlling the pollination during seed increases by enclosing plants and pollinators within cages. Wild accessions tend to represent more genetic diversity among individuals within seed populations than is found among seeds of accessions from seed-propagated elite germplasm. In accessions that represent wild germplasm, it is critical to store the diversity represented by the population (as defined by a specific region, progeny, etc.) as opposed to storing specific genetic combinations of individuals (clones or inbred lines). Maintaining the genetic composition of these heterogeneous wild accessions involves preserving gene frequencies (preventing genetic drift) so that rare alleles are not lost from the population (Schoen et al. 1998). Controlled pollinations among individuals of an accession maintain allelic frequencies during seed increases (Crossa 1989). About 10% of the NPGS collection is composed of plants that are propagated vegetatively, rather than through seeds. The methods of preservation and regeneration of these clonal collections differ from those of seeds. This review presents data on the size and nature of vegetatively propagated materials at the NPGS active sites. Data were compiled from curators and also from publicly available databases. Genetic diversity is captured by preserving either elite genotypes or by maintaining allelic frequencies within wild accessions. Long-term preservation of vegetative propagules is most effectively achieved through cryopreservation. Thus, the principles, techniques, and obstacles involved in plant cryopreservation are described.

II. CLONES AS GENETIC RESOURCES In the NPGS, most collections are maintained as seed; however, more than 28,000 accessions are vegetatively maintained and are propagated by cuttings, grafted buds, bulbs, runners, corms, or rhizomes. Most of the temperate and tropical fruit and nut crops as well as some ornamentals in clonal collections are maintained as plants in the field, greenhouse, or in tissue culture. In some species such as garlic and pineapple, few or no seeds are formed, so materials must be propagated vegetatively from bulbs or cuttings.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

295

Specific allelic combinations (i.e., genotypes) are preserved in accessions that are propagated clonally. In most cases, accessions represent elite genetic combinations (such as those found in a ‘Delicious’ apple) that cannot be identically regenerated by seed. Mature plantings of clonal collections enable phenotypic evaluation trials to identify unique germplasm characteristics. Some key descriptors include (1) disease resistance, (2) insect and mite resistance, (3) soil pest resistance and tolerance to soil environments, (4) plant cold hardiness and tolerance to other environmental stresses, (5) plant size, shape, and productivity, and (6) crop quality and nutritional factors (Westwood 1989). Desirable lines are used as parents in breeding programs. When materials are maintained as mature field plants, male and female gametes are readily available to breeders. If germplasm is only available as seeds, it could take many years for a seedling to reach maturity and be useful for breeding purposes (Towill 1988a). Some populations of plants endemic to the United States are rare or endangered. For example, Zizania texana is an aquatic grass that grows along a four-mile stretch of the San Marcos River in Texas (Walters et al. 2003). Since few plants and/or seed exist, a series of clones is grown ex situ to maintain the existing genetic diversity of the endangered species. Clones can be vegetatively propagated until enough individuals or seeds are available for re-establishment in the wild (Schoen and Brown 2001). The genera included in the NPGS clonal collections are shown in Table 7.1. For each genus, the location of the collection and the number of species (as well as the total number of species known) are listed. Curators have provided data regarding the number of unique accessions per genus and the number of those that could be considered “wild” or landrace. Currently, wild accessions are maintained as seeds or vegetative explants (discussed later) and the number of these accessions that are maintained as seed or as clones is given, as well as the environment in which clones are kept (field, greenhouse, screenhouse, or tissue culture). The extent of duplicate, backed-up accessions of each collection is indicated. Finally, plant type (temperate vs. tropical) and seed type (orthodox vs. recalcitrant) are characteristics that help determine the approach taken for cryopreservation. There are many additional ornamental and floricultural genera that are maintained in the NPGS that could be considered clonal collections. These additional collections are listed in Table 7.2.

Table 7.1.

Genus

Clonal crop collections in the NPGS, 2002.

Common name

Actinidia

Kiwifruit

Allium, wild Ananas Annona

Garlic, etc. Pineapple Sweetsop Annona Peanut Breadfruit Pawpaw Starfruit

Arachis Artocarpus Asimina Averrhoa Bambusa, etc. Canarium Carica Carya Citrus, etc. Corylus Cydonia Cynodon Digitaria Dimocarpus Diospyros Ficus Fortunella Fragaria Garcinia Humulus Ipomoea Juglans Litchi Macadamia Malus Mangifera Manilkara Mentha Morus Musa Nephelium Olea Passiflora Persea Phoenix Phyllostachys, etc. Pistacia Pouteria Prunus Prunus cerasus Psidium Punica Pycnanthemum Pyrus Ribes Rubus Saccharum Solanum Theobroma Tripsacum Vaccinium Vitis Zoysia A B C D E F G

Bamboo, tropical Pili Nut Papaya Pecan, hickory Citrus Hazelnut Quince Bermudagrass Crabgrass Longan Persimmon Fig Kumquat Strawberry Garcinia Hops Sweet potato Butternuts Walnut Lychee Macadamia Apple Mango Sapodilla Mint Mulberry Banana, plantain Rambutan Olive Passionfruit Avocado Date Bamboo Pistachio Mamey sapote Prunus Sour cherry Guava Pomegranate Mountain mint Pear Current, gooseberry Blackberry, raspberry Sugarcane Potato Cacao Tripsacum Blue, cranberry Cold-hardy grape Grape Zoysia

C

Total acc. D

Wild acc. E

Elite acc. F

Nonseed acc. G

Field H

17 17 94 6 9 9 56 7 5 2 2 9

57 57 213 8 43 43 70 18 9 2 2 34

136 41 919 168 13 71 9712 40 61 41 25 97

100 20 0 20 0 0 9412 6 10 0 0 0

36 21 219 148 13 71 300 34 51 41 25 97

126 41 219 168 13 71 300 40 54 41 25 97

131 41 219 0 13 71 0 40 61 41 25 97

7 7 17

7 7 17

18 11 28

23 113 934

5 26 534

18 87 400

23 113 934

23 113 934

Riverside Miami Corvallis Corvallis Griffin

81 18 9 1 8

83 83 9 1 8

138 138 12 1 9

743 73 518 51 1190

20 0 0 10 20

723 73 518 41 1170

713 73 510 45 1190

743 73 518 51 0

Griffin Hilo Miami Davis Davis Riverside Corvallis Mayaguez Corvallis Griffin

35 1 1 3 3 6 13 9 2 78

35 1 1 9 49 6 13 9 2 78

67 1 1 59 178 6 22 48 3 115

510 22 27 37 129 11 1635 9 185 1110

7 1 0 11 1 3 1000 0 100 423

503 21 27 26 128 8 635 9 85 683

510 22 27 37 129 11 585 9 150 687

0 22 27 37 129 11 100 9 0 0

Corvallis Davis Hilo Miami Hilo Geneva Miami Mayaguez Corvallis Davis Mayaguez

2 16 1 1 4 35 11 3 18 4 2

16 16 1 1 4 35 11 3 18 5 2

22 22 1 1 9 58 21 8 31 9 19

67 420 78 24 33 3925 159 8 469 21 112

50 192 0 0 2 1622 5 0 25 0 0

17 228 78 24 31 2303 154 8 444 21 112

67 420 78 24 33 2421 159 8 418 21 112

67 420 78 24 33 2421 159 8 0 21 112

Hilo Davis Hilo Miami Riverside Griffin/ Byron Davis Mayaguez Davis Geneva

4 1 5 2 2 23

4 1 5 2 7 23

21 9 90 13 13 31

60 133 63 233 69 97

5 3 21 0 1 0

55 130 42 233 68 97

60 133 63 233 69 97

60 133 0 233 69 97

8 4 64 1

8 4 78 1

10 28 167 1

217 26 1174 81

68 0 278 0

149 26 896 81

217 26 1174 81

217 26 1174 81

Hilo Davis Corvallis

4 1 17

6 1 17

27 2 17

59 143 84

2 1 10

57 142 74

59 143 25

59 143 0

Corvallis Corvallis

26 89

26 90

34 117

1753 881

1000 500

753 381

1477 513

1596 770

Corvallis

174

177

361

1635

1000

635

585

772

Miami Sturgeon Bay Mayaguez Miami Corvallis

16 126

16 126

28 384

1665 5440

1200 4730

465 826

1665 710

1665 0

1 13 62

1 13 62

8 16 86

187 272 1111

0 272 1000

187 0 111

187 272 509

187 272 611

Active site

A

B

Corvallis Davis Pullman Hilo Mayaguez Miami Griffin Hilo Corvallis Miami Hilo Mayaguez

15 6 93 6 4 9 56 4 5 2 1 8

Hilo Hilo Brownwood

Geneva

18

33

50

1172

347

825

1112

1172

Davis Griffin

33 4

33 4

50 5

2731 45

155 7

2576 38

2731 45

2731 0

Number of species represented in this genus at the active site. Number of species represented in this genus in the NPGS (including smaller collections at other locations). Number of species in this genus taxonomically. Total number of unique accessions within this clonal collection, according to the curator. Number of total unique accessions that are considered “wild” by the curator. Number of total unique accessions that are considered “elite” by the curator. Number of accessions that are maintained vegetatively at the active site.

Seed longterm backup L

Plant type M

Seed Type N

Curator

0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%

0% 0% 24% 0% 0% 0% 88% 0% 0% 0% 0% 0%

temp temp temp trop trop trop trop trop temp trop trop trop

orthodox orthodox none, orth recal recal recal none recal recal recal recal none

Hummer, K. Simon, C. Hellier, B. Zee, F. Goenaga, R. Schnell, R. Pittman, R. Zee, F. Hummer, K. Schnell, R. Zee, F. Goenaga, R.

0% 0% 0%

0% 0% 0%

0% 0% 0%

trop trop temp

recal recal recal

Zee, F. Zee, F. Grauke, L.

6% 0% 90% 10% 0%

0% 0% 8% 0% 0%

0% 0% 11% 0% 0%

0% 0% 0 0% 0%

trop trop temp temp temp

recal recal recal orthodox orthodox

Krueger, R. Schnell, R. Hummer, K. Postman, J. Morris, B.

0 0 0 0 0 0 1050 0 35 423

0% 0% 0% 0% 0% 0% 50% 0% 50% 100%

0% 0% 0% 0% 0% 0% 13% 0% 57% 3%

0% 0% 0% 0% 0% 0% 0% 0% 5% 0%

0% 0% 0% 0% 0% 0% 24% 0% 0% 42%

temp trop trop temp temp trop temp trop temp temp

orthodox recal recal recal recal recal orthodox recal orthodox none

Morris, B. Zee, F. Schnell, R. Simon, C. Simon, C. Krueger, R. Hummer, K. Goenaga, R. Hummer, K. Jarret, R.

0 0 0 0 0 0 0 0 196 0 in prog

0 0 0 0 0 1504 0 0 51 0 0

50% 0% 23% 0% 0% 0% 30% 25% 25% 0% 0%

0% 0% 0% 0% 0% 0% 0% 0% 43% 0% 0%

0% 0% 0% 0% 0% 80% 0% 0% 4% 0% 0%

0% 0% 0% 0% 0% 30% 0% 0% 0% 0% 0%

temp temp trop trop trop temp trop trop temp temp trop

recal recal recal recal recal orthodox recal orthodox orthodox orthodox recal

Hummer, K. Simon, C. Zee, F. Schnell, R. Zee, F. Forsline, P. Schnell, R. Goenaga, R. Hummer, K. Simon, C. Goenaga, R.

0 0 63 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0

0% 0% 0% 8% 100% 51%

0% 0% 0% 0% 0% 0%

0% 0% 0% 0% 0% 0%

0% 0% 0% 0% 0% 0%

trop temp trop trop trop temp

recal orthodox recal recal orthodox none

Zee, F. Simon, C. Zee, F. Schnell, R. Krueger, R. Pederson, G.

0 0 0 0

0 0 0 0

0 0 0 0

0% 0% 6% 8%

0% 0% 0% 0%

0% 0% 0% 62%

0% 0% 0% 0%

temp trop temp temp

orthodox recal orthodox orthodox

Simon, C. Goenaga, R. Simon, C. Forsline, P.

0 0 32

0 0 9

0 0 59

0% 0% 0%

0% 0% 10%

0% 0% 0%

0% 0% 0%

trop temp temp

recal orthodox orthodox

Zee, F. Simon, C. Hummer, K.

0 296

234 41

276 368

75% 75%

15% 4%

3% 1%

0% 0%

temp temp

orthodox orthodox

Postman, J. Hummer, K.

0

199

1050

30%

25%

2%

24%

temp

orthodox

Hummer, K.

0 0

0 826

0 4730

0% 94%

0% 0%

0% 0%

0% 52%

trop temp

orthodox none

Schnell, R. Bamberg, J.

0 0 300

2 0 150

0 0 602

0% 0% 50%

0% 0% 25%

0% 0% 0%

0% 0% 10%

trop trop temp

recal orthodox orthodox

Goenaga, R. Schnell, R. Hummer, K.

0

0

60

17%

0%

0%

0%

temp

orthodox

Forsline, P.

in prog 45

0 0

0 0

7% 0%

0% 0%

0% 0%

0% 51%

temp temp

orthodox orthodox

Simon, C. Morris, B.

Green/ screen house H

Tissue culture H

0 0 0 168 0 0 300 0 0 0 0 0

H I J K L M N

Seed H

U.S. backup I

Shortterm backup J

0 0 0 168 0 0 0 32 0 0 0 0

10 0 700 0 0 0 9412 0 7 0 0 0

50% 0% 42% 0% 0% 0% 0% 100% 90% 0% 0% 51%

0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%

0 0 0

0 0 0

0 0 0

0% 0% 14%

315 0 86 0 1190

0 0 43 0 0

30 0 8 6 0

510 0 0 0 0 0 772 0 68 0

0 0 0 0 0 0 195 0 88 683

0 0 0 0 0 0 0 0 459 0 0

Longterm backup K

The form in which accessions are maintained at the active site. Percentage of collection that is maintained at a second site in the United States (usually non-federally). Percentage of vegetative collection that is maintained in vitro at the NCGRP. Percentage of vegetative collection that is maintained at the NCGRP in long-term storage (usually cryogenically). Percentage of the seed portion of clonal collection that is maintained at the NCGRP. Plant type (temperate or tropical). Seed type (orthodox or recalcitrant).

298

Table 7.2.

Ornamental germplasm collections in the NPGS.

Genus

Family

Common name

Abies Acer Aglaonema Alstroemeria Amelanchier Anthoxanthum Anthurium Arbutus Bactris

Pinaceae Aceraceae Araceae Amaryllidaceae Rosaceae Gramineae Araceae Ericaceae Palmaceae

Fir Maple Evergreen Alstroemeria Shadbush Grass Anthurium Madrone Palm

Begonia Berberis Betula Brugmansia Buddleja Callicarpa Carpinus Castanea Catalpa Ceanothus Celtis Cercis Chrysanthemum Clematis Cornus Corylopsis

Begoniaceae Berberidaceae Betulaceae Solonaceae Loganiaceae Verbenaceae Betulaceae Fagaceae Bignoniaceae Rhamnaceae Ulmaceae Leguminosae Compositae Ranunculaceae Cornaceae Hamamelidaceae

Begonia Barberry Birch Brugmansia Butterfly-bush Beauty-Berry Hornbeam Chestnut Catalpa Buckbrush Hackberry Redbud Chrysanthemum Clematis Dogwood Winter Hazel

Active site Natl. Arboretum Natl. Arboretum Columbus, Miami Columbus Corvallis Griffin Columbus, Miami Corvallis Miami Hilo Columbus Natl. Arboretum Natl. Arboretum Miami Natl. Arboretum Natl. Arboretum Natl. Arboretum Corvallis Arboretum Corvallis Natl. Arboretum Natl. Arboretum Ames Natl. Arboretum Natl. Arboretum Natl. Arboretum

Number of accessionsz 32 105 — — 57 5 — 9 16 14 — 15 6 10 5 21 40 14 5 35 24 34 5 19 33 5

299

Cotinus Cotoneaster Crataegus

Anacardiaceae Rosaceae Rosaceae

Cotinus Cotoneaster Hawthorn

Cunninghamia Cuphea Cupressus Dendranthema Deutzia Dieffenbachia Duchesnea Elaeagnus Eremechloa Euonymus Euphorbia Exochorda Gaultheria Gaylussacia Gleditsia Halesia Helianthus Hemerocallis

Toxodiaceae Lythraceae Cupressaceae Compositae Saxifragaceae Araceae Rosaceae Elaeagnaceae Gramineae Celastraceae Euphorbiaceae Rosaceae Ericaceae Ericaceae Leguminosae Styracaceae Compositae Lilaceae

China-fir Cuphea Cypress Chrysanthemum Deutzia Dieffenbachia Mock-strawberry Oleaster Grass Spindle tree Poinsettia Pearl Bush Salal Black Huckleberry Honey locust Silverbell/Snowdroptree Sunflower Daylily

Hydrangea Ilex Iris Juniperus Laburnum Lagerstroemia Larix Ligustrum Lilium

Saxifragaceae Aquifoliaceae Iridaceae Cupressaceae Leguminosae Lythraceae Pinaceae Oleaceae Liliaceae

Hydrangea Holly Iris Juniper Laburnum Crape myrtle Larch Privet Lily

Natl. Arboretum Natl. Arboretum Natl. Arboretum Corvallis Natl. Arboretum Ames Natl. Arboretum Columbus Natl. Arboretum Columbus, Miami Corvallis Natl. Arboretum Griffin Natl. Arboretum Columbus Natl. Arboretum Corvallis Corvallis Natl. Arboretum Natl. Arboretum Ames Natl. Arboretum Columbus Natl. Arboretum Natl. Arboretum Columbus Natl. Arboretum Natl. Arboretum Miami Natl. Arboretum Natl. Arboretum Columbus

6 42 48 8 5 14 12 — 24 — 7 9 12 15 — 5 27 14 6 88 24 7 — 21 39 — 46 5 10 6 8 — (continued )

300

Table 7.2.

(continued )

Genus

Family

Common name

Lindera Lonicera

Lauraceae Caprifoliaceae

Spice bush Honeysuckle

Magnolia Malpighia

Magnoliaceae Malpighiaceae

Magnolia Magnolia

Mespilus Nyssa Ostrya Oxydendrum Paspalum Pelargonium Penniesetum Peraphyllum Phalaenopsis Philadelphus Philodendron Picea Pinus Platycladus Plumeria Prunus Pyracantha Quercus Rhamnus Rhododendron Rhus

Malpighiaceae Nyssaceae Betulaceae Ericaceae Gramineae Geraniaceae Gramineae Rosaceae Orchidaceae Saxifragaceae Araceae Pinaceae Pinaceae Cupressaceae Apocynaceae Rosaceae Rosaceae Fagaceae Rhamnaceae Ericaceae Anacardiaceae

Medlar Tupelo Hop-Hornbeam Sourwood, Sorrel tree Grass Geranium Grass Wild crabapple Phalaenopsis Mock-orange Philodendron Spruce Pine Arborvitae Plumeria Prunus Firethorn Oak Buckthorn Rhododendron Sumac

Active site Natl. Arboretum Natl. Arboretum Corvallis Ames Natl. Arboretum Miami Hilo Corvallis Natl. Arboretum Natl. Arboretum Natl. Arboretum Griffin Columbus Griffin Corvallis Columbus Natl. Arboretum Columbus Natl. Arboretum Natl. Arboretum Natl. Arboretum Miami Natl. Arboretum Natl. Arboretum Natl. Arboretum Natl. Arboretum Natl. Arboretum Ames

Number of accessionsz 22 22 28 6 11 15 7 22 10 14 5 14 — 23 6 — 18 — — 65 7 120 54 5 35 13 56 10

Rosa Salix Sambucus Sapium Sophora Sorbus Spathiphyllum Stenotaphrum Styrax Syringa Tilia Tsuga Ulmus Urochloa Vetiver Viburnum

Rosaceae Saliaceae Caprifoliaceae Euphorbiaceae Leguminosae Rosaceae Araceae Gramineae Styracaceae Oleaceae Tiliaceae Pinaceae Ulmaceae Gramineae Gramineae Caprifoliaceae

Grass Snowbell Lilac Linden, Basswood Hemlock Elm Grass Grass Viburnum

Weigela Zanthoxylum Zelkova

Caprifoliaceae Araceae Ulmaceae

Weigela Calla Elm-like

z

Rose Willow Elderberry Tallow Sophora Mountain Ash

Data is unavailable for collections that are newly proposed.

Natl. Arboretum Ames Corvallis Natl. Arboretum Natl. Arboretum Corvallis Columbus, Miami Griffin Natl. Arboretum Natl. Arboretum Natl. Arboretum Natl. Arboretum Ames Griffin Griffin Ames Natl. Arboretum Natl. Arboretum Natl. Arboretum Natl. Arboretum

118 57 127 5 6 223 — 22 15 30 25 28 23 5 14 8 87 23 9 8

301

302

G. VOLK AND C. WALTERS

III. MAINTENANCE OF GENETIC DIVERSITY IN CLONAL COLLECTIONS Strategies to collect and maintain genetic diversity within clonal collections of the NPGS are based on what is considered valuable: specific genetic combinations in elite germplasm or genes conferring desirable traits in wild accessions. Using the geographical, lineage, and historical data available in accession records, the collection of new accessions can be targeted toward novel germplasm types. Strategies to collect wild germplasm must also consider the native range of the species and the number and form of individuals available, as well as material transfer agreements and phytosanitary agreements (if outside the United States). A. Wild Germplasm Collections of wild relatives provide a larger genepool for genetic improvement than collections comprised of only elite accessions. For example, wild apple species demonstrated resistance to apple scab (Venturia inaequalis), cedar apple rust (Gymnosporangium juniperivirginianae Schw.), and fire blight (Erwinia amylovora, Burill) (Hokanson et al. 1997; Luby et al. 2001). To ensure that rare desirable genes are captured from these genepools, the genetic diversity of wild populations must be adequately represented. NPGS is a consortium of ex situ reserves where plants are grown and maintained outside of their native habitats. The ex situ strategy is a necessity since most of the crops important for American agriculture are not endemic to the United States. However, there are several genera in the NPGS clonal collections that have species native to the United States (e.g., Allium, Asimina, Carya, Fragaria, Humulus, Ribes, Rubus, Solanum, Vaccinium, and Vitis) that provide an opportunity to preserve genetic diversity using an integrated conservation strategy that includes both ex situ and in situ reserves (White et al. 1989; Pavek and Garvey 1999). The establishment of in situ reserves ensures long-term access to the wild populations and is particularly useful in cases where high levels of genetic diversity exist in wild populations. They are also a cost-effective method for preserving many wild genotypes. In some species (e.g., Allium), native populations have higher fecundity than plants maintained ex situ at the active site. Maxted and Hawkes (1997) emphasize the importance of considering potential economic uses, conservation status, genetic and ecogeographic distinction, biological and cultural importance, and costs when in situ reserves are designed. Currently, propagules are maintained at active sites, and when regeneration is necessary, materials are collected from the in situ sites. In the NPGS, in situ

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

303

pilot projects are underway for Allium, Carya, and Solanum. When genetic diversity studies are complete, populations will be selected for in situ reserves (Pavek and Garvey 1999). When materials are collected internationally, they are frequently brought to the United States in the form of seeds. There are fewer quarantine restrictions for new germplasm accessions that are seed accessions because diseases are often not transmitted in the seed. Once in the United States, curators plant seeds and evaluate traits of seedlings and, after several years, mature plants bear fruit. There are few guidelines for determining the number of seeds or seedlings necessary to capture and represent the diversity in the original wild population or the overall diversity of the species (Brown et al. 1997; Schoen and Brown 2001). On collection trips, curators or collectors may acquire seeds from many locales within a region and often only a few of these can be included in field evaluations. When possible, remaining seed is stored for later use and/or distributed to other sites for evaluation. Wild accessions in clonal collections were brought to the NPGS as either cuttings or seeds from plants growing in situ. Depending upon the natural distribution range and the number of individuals available, wild accessions can be maintained in the collection as clones or as a population of seeds or seedlings. When seeds are used as the propagule, ideally multiple seeds are planted to represent an accession (seedlings) so the phenotypic and genetic diversity can be evaluated. However, space limitations prevent ideal practices and usually between one and ten seedlings are planted to represent a wild accession since curators are often limited by field space or resources. Often, these numbers are inadequate to fully represent the diversity present in the wild materials and sampling biases occur. Most curators strive to acquire new germplasm to broaden diversity represented in the NPGS collections. However, as collections grow and resources remain constant, preservation priorities must be determined. To reduce maintenance costs for some collections, such as pear, banana, macadamia, guava, longan, lychee, canarium, rambutan, and starfruit, collections are consolidated and often only a single individual represents each accession in field plantings. Forsline and Way (1993) discuss the rationale for the apple collection consolidations that have occurred. It was determined that at least 2100 of the 2600 clonal accessions of Malus were classified as Malus × domestica. These accessions represented a relatively narrow genetic base. At least 325 of those accessions were selected for deaccessioning since many represented sports of standard cultivars or demonstrated little phenotypic uniqueness. Distribution records for these accessions were kept so deaccessioned cultivars could possibly be obtained from non-federal sources in the future. Future field reductions

304

G. VOLK AND C. WALTERS

may occur with the successful implementation of the cryogenic storage of apple budwood (Forsline et al. 1998). Accessions that are rarely requested may be maintained cryogenically at the Geneva, New York, site and regenerated by grafting if required for research purposes. By reducing the maintenance requirements for these apple cultivars, more space is available to study apple germplasm collected from the wild. Wild relatives of crop species are often sources of allelic richness; however, due to the extreme levels of heterozygosity within populations, they are more difficult to maintain. It is a challenge to adequately represent the wild relatives of an agriculturally important species within a collection with the limited resources available to NPGS sites (Schoen and Brown 1995). Fourteen germplasm collections were surveyed to determine the extent of cultivated and wild accessions represented in the NPGS (Greene and Morris 2001). Cultivated accessions comprised most of the selected collections. This suggests that historically collections have been assembled for crop improvement rather than to capture genetic diversity. However, germplasm representing wild species in Malus, Allium, and Solanum was requested just as frequently as accessions from cultivated species, indicating that there are users of this diverse germplasm. Many of the related wild species were listed as endangered. Although Malus was the only genus that is maintained primarily clonally that was included in the study performed by S. L. Greene and J. B. Morris, data available from GRIN (Germplasm Resources Information Network; www.ars-grin.gov/npgs/) comparing cultivated (elite) and wild species of clonal crops indicate that many of the clonal collections have a relatively low representation of diverse species (Table 7.1). Currently, crop germplasm committee members (usually represented by federal, state, and private researchers) work with curators to identify gaps and identify duplication in the NPGS (USDA, ARS, National Genetic Resources Program 2002). To broaden the species representation within each genus, Greene and Morris (2001) proposed that curators receive technical input from CGC members as well as botanists and conservationists. This input would also broaden the spectrum of users with a voice in germplasm collection decisions. With greater species representation in germplasm collections, the NPGS would promote its mission of preserving genetic resources for future generations. B. Evaluation of Diversity Collections should provide users with access to desirable plants with an array of diverse traits. Existing elite lines are already valued for these traits and provide a genetic background for future improvements. While these collections are valuable for phenotypic differences, there may be little

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

305

genetic variability among elite lines (Hokanson et al. 1997). Traditionally, genetic diversity has been determined by measuring phenotypic variability. Recent advances in genetic analyses and the availability of regional ecophysiological data have enabled curators to use other techniques to infer the genetic similarity among accessions (Schoen and Brown 1993). 1. Phenotypic Diversity. Many germplasm collections have been surveyed for phenotypic characters that demonstrate the diversity present within a species. Descriptors of morphological or other identifying traits are publicly available through GRIN. If a breeder identifies an accession of interest from a well-documented germplasm collection, additional accessions that have similar regions of origin can be targeted for screening to determine the level of diversity present for a given character. While phenotypic data are invaluable, they may not always represent the genetic potential of wild accessions in a collection (Tanksley and McCouch 1997; Merila and Crnonkrak 2001). In some species (such as common bean), similar phenotypes may be evolutionarily unrelated (Marita et al. 2000). 2. Ecophysiological Diversity. The environment influences the extent of geographic differentiation in populations (Greene et al. 1999a). Knowledge of the primary and secondary centers of origin as well as early domestication events allow plant exploration teams to collect materials that are likely to be well adapted to a given environment. Since detailed environmental data are available for most regions of the world, the geographical information system (GIS) can provide data such as precipitation, temperature, elevation, soil type, and hardiness zone for the region that encompasses the natural habitat of a particular species (Pollak and Corbett 1993). When accurate passport data for wild accessions are available, gaps in collections can be targeted during exploration trips by overlaying the geographical attributes of the countries of origin and the native sources of existing accessions. Niche environments could be targeted on specific collection trips to acquire germplasm that may have novel genes of interest. Curators must decide which accessions should be included in the NPGS collections. Greene et al. (1999b) suggest that high priority be given to accessions that have passport data that reflect population adaptation, desirable ecotypes that represent novel ecogeographical classes, and potential for population uniqueness that results in geographic distances or barriers that may prevent gene flow. Accessions that do not meet these criteria and are collected from disturbed areas should have lower priority since there is no way to tell the level of adaptation (Greene et al. 1999b). In sweet potato, where geographic origin and morphological descriptor data have been correlated, the collections designed using an ecogeographic

306

G. VOLK AND C. WALTERS

approach had more accessions with desirable phenotypic characters such as disease resistance, salt tolerance, storage root dry matter content, and vegetative period (Huaman et al. 1999). In some cases, for example potato, geographic distances among accessions in a population do not reflect genetic differences (del Rio et al. 2001). 3. Genetic Diversity. The Genomics Age has brought a plethora of new tools for the evaluation, conservation, and utilization of genetic resources. These tools are currently being applied to various clonal collections to identify what is unique or redundant, to infer phylogenetic relationships among species or populations, and to evaluate the genetic structure of populations or accessions (genetic variability within and among populations or accessions). These applications can be used to describe the genetic diversity captured by collections. Various marker systems can be applied depending on cost and prior knowledge of the genome architecture within the species. Studies of population structure frequently use neutral markers (no gene product) such as amplified fragment-length polymorphism (AFLP) analyses and single sequence repeats (SSRs) that are polymorphic, and considered “hyper-variable” (Bretting and Widrlechner 1995; Kresovich et al. 1997; Antolin 1998; Schaal and Olsen 2000; Charlesworth et al. 2001). The depth of genetic representation can be identified using population genetics methods and these data can be used to help target new accessions for acquisition or to identify accessions that are very closely related (Lamboy et al. 1996; Chavarriaga-Aguirre et al. 1999). Fingerprinting techniques, whereby individuals are identified by specific alleles using a suite of polymorphic markers, must be used with caution to determine duplications within the collections, since only a finite number of polymorphisms are scored. Nonetheless, closely related accessions can be identified through fingerprinting, giving curators and researchers a better understanding of the diversity within collections so that they can be managed efficiently without negative impacts on overall genetic diversity (Hokanson et al. 1998; Lamboy and Alpha 1998). C. Core Collections Core collections have been designated for many large germplasm collections so that diversity represented in the collection can be captured with a subsample of accessions. By including approximately 10% of the accessions, core collections are more manageable than entire collections for researchers who wish to evaluate a collection for the presence of a particular trait of interest. Core collections are often determined based on random sampling of accessions that represent geographical

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

307

regions or phenotypical attributes of a collection (Brown 1989; Holbrook et al. 1993; Eberhart and Bockelman 1994; Jaradat 1995; Tohme et al. 1995; Clark et al. 1997; Ortiz et al. 1998; van Hintum 1999; Skinner et al. 1999; Huaman et al. 1999, 2000; Balakrishnan et al. 2000; Upadhyaya et al. 2001). Increasingly, molecular methods have been used to ensure that the genetic diversity of a collection is represented in core collections of germplasm (Lerceteau et al. 1997; ChavarriagaAguirre et al. 1999; Dean et al. 1999; Hu et al. 2000; Marita et al. 2000). Ideally, a core collection encompasses the spectrum of genetic diversity within a collection rather than just extremes (Brown 1995). The relative percentage of wild and elite accessions within a core collection varies considerably among crops. Generally, wild germplasm contains a much higher degree of diversity than that found in selected elite cultivars (Brown et al. 1997) and so one might design core collections around wild accessions or landraces to get the broadest sampling of genetic diversity. However, the composition of core collections for clonals may differ from seed collections since individual clones are valued for their phenotypic differences as well as genetic distinctions. The 15 core collections (Table 7.3) established for clonal genera listed in Table 7.1 were selected primarily on phenotypic and geographical attributes Table 7.3.

NPGS clonal core collections (data from GRIN).

Genus Arachis Corylus Cydonia Fragaria Humulus Juglans (Corvallis) Malus Mentha Pycnanthemum Pyrus Ribes Rubus Solanum Vaccinium Vitis (Geneva)

Total accessions

Accessions in core collection

Total accessions considered wild or landrace

9712 518 51 1635 185 67 3925 469 84 1753 881 1635 5440 1111 1172

798 165 46 517 82 14 205 49 30 232 322 527 3315 369 127

n/a 6 0 54 1 0 7 12 11 28 114 232 n/a 152 47

*% according to the curator.

Core accessions considered wild or landrace (%)

Total accessions considered wild or landrace (%) [*]

4 0 10 1 0 3 24 37 12 35 44

1 [0] 0 [20] 3 [61] 0 [54] 0 [74] 0 [41] 3 [53] 13 [12] 2 [57] 13 [57] 14 [61]

41 37

14 [9] 4 [30]

308

G. VOLK AND C. WALTERS

of the accessions. According to GRIN, the number of wild or landrace accessions represented in those core collections is less than 50% of the core collection, even though the number of wild accessions present in these collections (according to the curator) is often greater than 50% of the total unique collection accessions. Core collections are considered dynamic. Therefore, as more genetic data becomes available and our concepts of diversity change, core collections may be redefined to represent the broadest base of genetic diversity. It is likely that core clonal collections will be comprised of both seed and clonal accessions. Core clonal accessions maintained as plants are valuable for evaluations and breeding, while seeds in long-term storage preserve the allelic diversity of accessions. If valuable characters are identified in field materials, additional seeds from these collections could be evaluated to find additional variation. Original seeds, rather than populations resulting from these individuals, should be evaluated to reduce costs (Brown 1995). IV. CLONAL COLLECTIONS IN THE NPGS A. Active Collections The four Plant Introduction (PI) stations (Pullman, Washington; Geneva, New York; Ames, Iowa; Griffin, Georgia), the Clonal Repositories (Geneva, New York; Corvallis, Oregon; Davis, California; Riverside, California; Hilo, Hawaii; Miami, Florida; Mayaguez, Puerto Rico; and Brownwood, Texas), and several other sites (such as the National Arboretum, Washington, DC; Sturgeon Bay, Wisconsin; Palmer, Alaska) are considered the “active” collections for germplasm in the United States. These sites are charged with maintaining, evaluating, and distributing germplasm to research scientists. A base collection of germplasm is maintained at the National Center for Genetic Resources Preservation (NCGRP, formerly National Seed Storage Laboratory, NSSL) in Fort Collins, Colorado (Roos et al. 1996; Reed 2001a). This collection consolidates the entire NPGS collection in low temperature, quiescent storage conditions for long-term safety and economy. Since clonal collections are usually kept as living accessions either in field, screenhouse, greenhouse, or in tissue culture conditions, they are some of the most expensive (in terms of labor and space) collections within the NPGS. For example, it is estimated that it costs $50–75 per year to maintain an apple tree in the field and around $100 per year to maintain each of the small fruit accessions in Corvallis, Oregon (Hokanson et al. 1997; Hummer and Reed 1999). While most of the clonal

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

309

germplasm collections are maintained at the Clonal Repositories, there are some clonal portions of seed-based collections that are curated elsewhere, such as Solanum (potato), Arachis (rhizomatous peanut), Ipomoea (sweet potato), and Allium (garlic). Of the more than 100 clonal genera maintained in the NPGS, 65 of the more agriculturally common crop collections are listed in Table 7.1. Most of the clonal collections are maintained at only one location since it is very laborious to tend to orchards or screenhouses containing more than a thousand unique accessions. Plants in these collections are frequently grown in duplicate in the field or in the screenhouse or greenhouse to ensure that no phenotypic differences exist and to guard against potential losses. Elite clonal accessions are often grown as duplicates, while wild accessions within clonal collections are grown as families of siblings at the repository. Since the collection is at a single location, traits such as growth habits, fruiting characteristics, or other phenotypes of interest to breeders can be compared among accessions growing in a similar environment. There are drawbacks to having an entire collection reside at a single location. Catastrophes such as disease or natural disasters can cause serious damage to outdoor collections. For example, the Geneva, New York, Malus repository was once severely infected with fire blight (Erwinia amylovora, Burill). Since back-up materials were available, the orchard is now being re-established. In Davis, California, Pierce’s disease (Xylella fastidiosa) has the potential of devastating the Vitis collection because a very efficient vector, the glassy-winged sharpshooter (Homalodisca coagulata), has been identified in the adjacent county. Furthermore, many of the repositories could be damaged by environmental disasters such as flooding, earthquakes, hurricanes, severe cold spells, or tropical storms. Collections are at risk if they are not preserved at a second location. As shown in Table 7.1, there are several genera (Actinidia, Averrhoa, Juglans, Prunus, and Vitis) that are maintained at more than one site. In most cases, the collection was split because some species of a genus were more suited to the growth environment provided at an alternative site. The split testifies to the broad geographic distribution of a genus and its representation in the NPGS collection. As shown in Table 7.1, sites maintain their materials in different forms. Some sites, such as Davis, California, Riverside, California, and Brownwood, Texas, keep accessions solely in the field, greenhouse, or screenhouse. Others, such as the clonal potato collection in Sturgeon Bay, Wisconsin, and the sweet potato collection in Griffin, Georgia, are exclusively in tissue culture. Miami, Florida, Mayaguez, Puerto Rico,

310

G. VOLK AND C. WALTERS

and Hilo, Hawaii, have portions of collections in both field and tissue culture. Davis, California, and Riverside, California, collections are kept in either greenhouse, screenhouse, or field conditions. Corvallis, Oregon, and Geneva, New York, sites maintain their collections as plants in either field, screenhouse, greenhouse, or tissue culture conditions, but also have some accessions within clonal collections stored as seeds. B. Base Collections Base collections differ from active collections in that they house complete collections but do not actively grow or evaluate individual accessions. Base collections are an ideal way to back up active sites since they represent the diversity present within the “active” collections. Accessions stored in the base collection are safe from disasters that may compromise the collection at the active site. NCGRP serves as the base collection for the NPGS and stores germplasm under conditions that prolong life spans and reduce costs for growth and regeneration. Clonal accessions can be stored in the base collection as vegetative propagules (in liquid nitrogen or as slow-growth tissue culture), pollen, or seeds. As described previously, the appropriate propagule (vegetative or sexually derived) depends on whether the genes or the genotypes are the conservation targets. Procedures to place vegetative tissues into the appropriate storage conditions are labor-intensive and prone to variable success depending on accession. Different strategies can be used to preserve genetic diversity or temporarily back up collections while better technologies are developed. Some clonal collections in the NPGS have a significant number of species or wild accessions (Table 7.1). The genetic diversity of clonal accessions with multiple wild accessions may best be represented by storing seeds or pollen. In most cases, pollen and seeds are much easier to store than vegetative propagules because of their inherent ability to survive the desiccation and freezing stresses implicit to cryopreservation. In some cases, there are numerous seeds remaining after exploration trips, and these seeds can easily be included in base collections. If wild accessions can be regenerated from seed without loss (or inadvertent introgression) of genetic diversity, then consideration should be given to storing these accessions as seeds, rather than as clones. Juglans and Carya are two clonal collections that have reliable protocols for pollen preservation. Protocols are available for seed preservation of wild accessions for the following clonal collections: Actinidia, Asimina, Carica, Citrus, Diospyros, Frageria, Humulus, Malus, Mentha, Passiflora, Pistacia, Prunus, Pyrus, Ribes, Rubus, Tripsacum, Vaccinium, and Vitis.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

311

1. Seed Storage. Technologies for storing seeds of most crops are well established. Conventional storage conditions at the NCGRP are designed for crop seeds that survive drying to low water contents (5+/–2%), usually by equilibration at 25% RH, and can then be placed in freezers at –18°C (Hong and Ellis 1996). At the NCGRP, storage is relatively inexpensive (about $0.04 per year per accession) and, dependent upon variety, seeds with initial high quality are expected to survive for several decades, if not centuries (Walters et al. 1998). The NCGRP also stores some crop seeds above liquid nitrogen (about –150°C) in cryovats (Stanwood 1985) (Fig. 7.1). The longer shelf-life achieved by the lower storage temperature reduces the cost of regeneration and the risk of genetic shifts (Eberhart et al. 1991; Stanwood and Sowa 1995). The NCGRP initiated its cryogenic program in 1984 (Stanwood et al. 1986). Results-todate demonstrate longer viability of cryogenically stored seeds (Stanwood and Sowa 1995; Wheeler et al. unpubl.), though the theoretical longevity is still conjectural. Not all seeds survive routine storage protocols used at the NCGRP. Seeds from many clonal species exhibit recalcitrant characteristics (Table 7.1). Recalcitrant seeds can be stored for six months to two years at temperatures between –5° and 25°C (dependent upon species) at 98% relative humidity (Vertucci et al. 1991). These seeds do not survive drying (Engelmann 1997) and so they cannot survive subfreezing temperatures without ultra-rapid cooling protocols (Wesley-Smith et al. 2001a) or application of exogenous cryoprotectants (Kioko et al. 1998; Walters et al. 2003). Cryopreservation of recalcitrant seeds is facilitated by the removal and dehydration of embryonic axes (Pence 1995; Beardmore

Fig. 7.1. Cryotanks at the NCGRP for preservation of seed, pollen, and vegetative propagules.

312

G. VOLK AND C. WALTERS

and Vong 1998; Vertucci et al. 1991; Wesley-Smith et al. 1992, 2001a,b). Thus, even though some species in the clonal collections produce recalcitrant seeds, there is potential for cryogenic storage of wild accessions such as Arachis, Asimina, Carica, Carya, Citrus, Diospyros, Juglans, Passiflora, and Saccharum as seeds (Table 7.1). Though still in the experimental stage, the NCGRP has initiated a program to place some recalcitrant seeds into cryogenic storage. In many clonal genera, particularly Prunus and Vitis, all the seeds originally collected (often very few) were planted in field collections. In other cases, numerous seeds collected from the wild were planted and grown for evaluation purposes, but only a few trees were selected for inclusion in the clonal collection. Thus, the few trees in the field represent the diversity available for that accession and each tree is particularly valuable since only a single or a few individuals may represent an entire species. Many of these individuals are now mature and reproductive. Preserving these species accessions using F1 generation seed is possible but would require controlled crosses since many of the accessions are interfertile and are planted closely together in the field. Seeds from open pollinations would be an easier source of germplasm that would preserve maternally inherited genes; however, valuable recessive alleles may easily be masked by pollen from another accession. To limit cross-contamination among accessions, wild accessions with only one or few individuals are usually vegetatively propagated. Future research efforts will determine how many seeds are needed to represent the genetic diversity of a wild accession in biparental matings within a small population of trees or vines. Detailed theoretical analyses have been modeled to address how many pollinations and offspring are necessary to capture most of the genetic diversity of a wild accession, maintain gene frequencies, and prevent loss of rare alleles for some crops. Unfortunately, the number of parents traditionally predicted in a model is usually much higher than what is available within clonal plantings (Crossa et al. 1993; Brown et al. 1997; Crossa and Vencovsky 1999). 2. Pollen Storage. If seed is unavailable or difficult to store, pollen may provide an alternative form of germplasm to preserve genetic diversity (Namkoong 1981; Schoenike and Bey 1981). Often, pollen is easily collected, dried, and stored (Towill 1985; Towill and Walters 2000). For many crops, distribution of pollen is desirable for breeders (Hoekstra 1995). Rather than waiting the multiple years for a grafted bud or seedling to flower, pollen can be used immediately in a breeding program for crops such as grape, pear, Prunus, and apple (Visser 1955; Parfitt and Almehdi 1983; Parfitt and Almehdi 1984; Ganeshan and Alexander 1990; Moriguchi 1995). Pollen may also serve as a source of

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

313

genetic diversity for perennials in in situ reserves. Although the seed physiologies of some temperate genera, such as Juglans, Carya, and Corylus, are not fully characterized, these genera have pollen that can be easily collected and stored (Haunold and Stanwood 1985; Luza and Polito 1988; Connor and Towill 1993; de Boucaud and Brison 1995; Towill and Walters 2000; Sparks and Yates 2002). Some tropical crops, such as avocado and papaya, have established pollen preservation methods (Sedgley 1981; Ganeshan 1985; Ganeshan and Alexander 1986). Actinidia, Carica, Dimocarpus, Diospyros, Passiflora, and Pistacia are additional NPGS clonal genera that could be preserved in part as pollen when pollen cryopreservation techniques are established. Since gametes only contain a partial set of genetic information and since they are technically not germinal cells that can be used to produce an individual, pollen collection and storage should only be considered as a complementary strategy to preserve genetic diversity. While the protocols for pollen dehydration and cryopreservation are relatively straightforward for many temperate species, procedures are not in place for many species of tropical origin (Towill 1985; Hanna and Towill 1995). In addition, pollen preservation may not be the method of choice for species that make minimal amounts of pollen. 3. Vegetative Propagule Storage. Many plants can be propagated vegetatively as dormant buds, axenic shoot tips, or meristematic materials. However, routine protocols for the storage of these materials are not available. Further development is needed before methods are routinely applied to all accessions in clonal collections. There are several strategies that can be used with variable success depending on the species, its adaptation to low temperatures, and its amenability to microculture environments. A simple approach to preserving some forms of clonal germplasm is to exploit the dormancy or low temperature tolerance of the resting stage (overwintering organs such as corms, bulbs, dormant buds) of a plant by placing non-axenic materials at reduced temperatures (usually just slightly below 0°C). Reduced temperature storage of non-axenic materials may decrease the expense involved in planting some vegetatively propagated field crops that must be regrown annually (e.g., garlic can be stored at –3°C for over 1 year). Storage temperatures (and water content) of bulbs and buds must be optimized to be low enough to maintain dormancy but high enough to prevent desiccation from extracellular ice growth and survival for one to five years is achievable. Cold storage can be used for woody perennial buds (Moriguchi 1995), stem or root structures (Volk et al. 2003), and recalcitrant seeds (Tompsett and Pritchard 1998).

314

G. VOLK AND C. WALTERS

Another method commonly used to provide short-term storage for backup of clonal material is axenic microculture (Reed 1992; Acheampong 1996; Jarret and Towill 1996). Duplicate tissue cultured plants can be kept at a second location when sterilization and media formulations for explant introduction, proliferation, and rooting are established. The time between transfers to fresh media can be increased to up to several years when cultures of temperate crops are kept under conditions that suppress growth (often by decreasing temperatures) (Withers 1980; Reed 1992; Reed 1999; Wu et al. 2001a). The above techniques provide only a partial solution to the problems of storing vegetative materials in base collections because they still require substantial labor when it is time to regenerate or subculture. The frequency with which materials need attention (every few months or years) makes the cost of the short-term solution prohibitive if vegetative material from every clonal accession represented in the base collection were backed up in this manner. To reduce maintenance costs, the base collection at NCGRP uses cryogenic storage to ensure that vegetative materials are safely stowed. As expected for seeds and other life forms that are cryogenically stored, longevities of vegetative materials stored in liquid nitrogen are expected to be very long (Towill 1988a).

V. CRYOPRESERVATION PRINCIPLES Cryogenic storage promises to virtually stop deterioration that results from the degrading reactions that occur at higher temperatures. Without proper treatment, exposure to extremely low temperatures is usually lethal to most cells. Germplasm maintenance requires living tissues so that germplasm supplies are replenishable and gene expression, not just the genes, can be studied. Hence, cryogenic storage of germplasm mandates that plants can be revived following exposure to, storage in, and rewarming from a cryogen such as liquid nitrogen. A. Freezing Actively metabolizing plant tissues generally have high water concentrations, which serves as a solvent for biochemical reactions and provides turgor for growth and rigidity. Lethal, intracellular ice crystals form when these untreated, fully hydrated plant tissues are exposed to low temperatures (Steponkus 1985a). The field of cryobiology largely focuses on ways to prevent intracellular ice nucleation and growth in cells.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

315

1. Damage. The formation of intracellular ice is dependent upon the freezing temperature of the aqueous solution within the cell, effectiveness of nucleators, and cooling rate. The presence of solutes within cells lowers the freezing temperature of the protoplasm. Colligative effects of compatible solutes, such as sugars, proteins, or amino acids, can only reduce the freezing temperature of water by a few degrees. Stronger surface interactions with water can reduce the freezing temperature further (Gekko and Satake 1981). During cooling, samples usually supercool with nucleation occurring first in extracellular spaces at between –2° and –10°C, depending on the cooling rate and the presence of nucleators. The formation of extracellular ice establishes a water potential gradient that promotes water diffusion to extracellular spaces (Levitt 1980). With slow cooling, water continues to migrate from the cytoplasm to the extracellular spaces between cells. The intracellular water concentration and thereby the intracellular freezing temperature is lowered (Levitt 1980). Very slow cooling or prolonged storage at temperatures between –2° and –20°C can allow too much water to migrate to extracellular spaces, thereby imposing desiccation stresses on the cytoplasm. If temperature decreases faster, the cytoplasm water may not diffuse across the membrane quickly enough and intracellular ice may form, usually causing cell death. The mechanisms by which intracellular ice kills cells remains conjectural. Large crystals may pierce membranes or cause demixing and dissolution of cellular constituents (Levitt 1980; Steponkus 1985b, 1990). Some intracellular ice is allowable (Vertucci 1989b; Vertucci et al. 1991; Wesley-Smith et al. 1992), but ice crystals must be sufficiently small to preclude damage (Bald 1987; Wesley-Smith 2001). Insufficient nucleators, antifreeze proteins, or rapid cooling can suppress extracellular ice formation and lead to supercooling of both intra and extracellular solutions. Supercooling may be an excellent protective mechanism for mild or short-term excursions to sub-freezing temperatures (George et al. 1982; Rajashekar et al. 1982; Hoshino et al. 1999). However, supercooled cells are particularly prone to freezing damage during cryogenic treatments. Nucleation of extracellular spaces at temperatures of about –10°C or lower in non-desiccated cells usually results in lethal, spontaneous intracellular freezing (Steponkus 1990; Uemura et al. 1995). 2. Protection. Freezing stresses can be alleviated when intracellular ice formation is avoided. Ice formation can be reduced by the presence of anti-freeze proteins, but this usually confers protection over a limited temperature range (Antikainen and Griffith 1997). Ice formation is also limited when cryoprotectants are present, though the mode of action of

316

G. VOLK AND C. WALTERS

these molecules is poorly understood (Fahy 1984, 1986; Fahy et al. 1987; Hey and MacFarlane 1996). Ice is least likely to form when the water concentration within cells is low. Cellular dehydration can be accomplished by drying at ambient conditions or by slow cooling in the presence of extracellular ice. Dehydration protects cells from intracellular ice formation by decreasing the freezing temperature and increasing the intracellular viscosity (Wesley-Smith et al. 2001a; Walters et al. 2003). Increased viscosity slows the rate of molecular rearrangements required for crystallization. Molecules stay in a semi-random conformation that eventually solidifies into a molecular glass. These glass, or vitrified, solutions are said to be “kinetically-stable” because molecular rotation and diffusion are effectively prevented (Williams 1994). However, subzero aqueous glasses are “thermodynamically-unstable” because of the tendency, albeit slow, to move toward the equilibrium state, which is ice. Because motion of water and other macromolecules is restricted, cells can remain viable in a stable glassy state (Fahy et al. 1984; Franks 1985; Williams 1994). Vitrification temperature for pure water is approximately –140°C and this temperature has become the standard storage temperature for hydrated tissues. Glass formation is promoted when a supercooled liquid is cooled faster than the rate of ice crystal formation. Pure water must be cooled at about 20,000°C⋅sec–1 to –140°C to form a glass (Franks 1985; Bald 1987). Dehydrated cells of seeds and pollen are so viscous that glasses readily form, even at relatively high temperatures (Wesley-Smith et al. 1992; Williams 1994; Buitink et al. 1998). While cells of orthodox embryos survive with water contents of 5 to 10% moisture when exposed to liquid nitrogen in a manner that is nearly independent of cooling rate (Vertucci 1989b), cooling rate becomes increasingly important as cells become more hydrated. Cells containing more than 40% moisture require cooling rates to –140°C of between 100 and 500°C⋅sec–1 (Wesley-Smith et al. 1992, 2001a). Vegetative propagules with higher water contents are difficult to cool fast enough to prevent ice formation. In these systems, exogenous protectants are added and glass formation is a careful balance between the addition of solutes, the level of dehydration, and the cooling rate. The effects of various solutes on solution viscosity and/or suppression of ice formation are poorly understood. The method used to vitrify clonal propagules is dependent upon the extent to which they can be desiccated. Dormant budwood from some temperate fruit crops, such as apple and sour cherry, has been preserved in liquid nitrogen when buds have a 25–30% moisture content (Table 7.4). In the dormant budwood systems, this moisture content is attained

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

317

Table 7.4. Changes in the moisture contents of explants (fresh weight basis) during processing for storage.

Explant Pollen Orthodox embryos Recalcitrant embryos Dormant buds

Genus

Form

Carya, Malus Phaseolus, Zea

Air dried Field dried

Aesculus, Citrus

Shoot tips

Malus, Prunus cerasus Mentha, Ipomoea

Shoot tips

Mentha

Starting moisture (%)

Preferred preservation moisture (%)

5–10 5–10

5 5

Fresh

40–75

Varies

Field materials

45–55

25–30

70

20

70–90

*

0.75M Sucrose alginate beads Freshly dissected

*Difficult to detect since tissue samples are small and immersed in cryoprotectants.

by drying the buds at –3°C prior to cooling slowly at 1°C per hour. In smaller axenic shoot tip systems, more rapid cooling rates can be used (0.5 to 2°C⋅min–1), which still enables some dehydration to occur prior to plunging into liquid nitrogen. In these systems, the threshold concentration for vitrification is achieved by the initial slow cooling before materials are exposed to liquid nitrogen (Sakai 1993). Alternatively, in highly hydrated or chilling sensitive explants, vitrification is promoted by the addition of concentrated cryoprotectant solutions (solutions that do not freeze when cooled to very low temperatures) and rapid cooling (>100°C⋅sec–1) (Fahy et al. 1987). Small propagules (low thermal masses) and rapid plunges into supercooled liquid nitrogen are frequently needed to achieve the rapid cooling rates required for explant survival (Vertucci 1989a; Wesley-Smith et al. 2001a). There is no time for intracellular ice crystals to nucleate and propagate within cells, or if ice crystals form they do not grow to sizes that cause damage. B. Desiccation Cells are protected from direct freezing damage by removing water through dehydration. However, dehydration can be damaging in itself since water plays a structural role, and provides turgor and hydrophilic and hydrophobic associations within cells. Cellular ultrastructure and membrane function can be compromised as cells are dehydrated. In

318

G. VOLK AND C. WALTERS

some cases, the increased cellular solute concentration may lead to a toxic accumulation of solutes (Finkle et al. 1985). The water content and cooling rate combinations that prevent freezing and dehydration damages are currently determined empirically (Wesley-Smith et al. 1992). Future studies will combine knowledge of the biophysical properties of water within cells with physiological understanding of critical water content for desiccation damage, hydraulic conductivity of water, and thermal loads that affect cooling rate. 1. Damage. Desiccation damages appear to be a combination of metabolic stresses (reduced cellular expansion, division, respiration, and biochemical syntheses), mechanical stresses (loss of turgidity, cellular contraction), and phase changes in membrane lipids (Vertucci and Farrant 1995; Walters et al. 2002). Maintaining the integrity of the plasma membrane is critical for cell survival (Steponkus 1990; Bryant et al. 2001; Hoekstra et al. 2001). Cellular contraction during dehydration can result in permanent loss of plasma membrane surface area if irreversible endocytotic vesiculation occurs (Steponkus 1990; Uemura and Steponkus 1999). Freeze-induced osmotic contraction often leads to lysis and death when cells are rehydrated. For example, non-cold-acclimated rye protoplasts shrink during freezing and burst as they regain volume during warming. However, cold-acclimated protoplasts return to original volumes with intact membranes when they are warmed after freezing events (Steponkus 1985a; Uemura and Steponkus 1999). Other lethal damage occurs when membrane systems fuse as membranes are forced into very close proximity when cells are dried (Webb and Steponkus 1993; Bryant et al. 2001). 2. Protection. In some plants, mild chilling or drought treatments induce metabolic changes that allow plants to tolerate cold and desiccation. Many of these responses are similar to those acquired during the development of desiccation tolerant seeds and include late embryogenesis abundant (LEA) proteins, chaperones, water channel proteins, proteinases, protein kinases, phospholipase C, transcription factors, detoxification enzymes, and osmoprotectant synthases (Cohen and Bray 1992; Crowe et al. 1998; Shinozaki and Yamaguchi-Shinozaki 1998; Pammenter and Berjak 1999; Scott 2000; Ditzer et al. 2001; Oliver et al. 2001; Hilbricht et al. 2002). Upon acclimation to stress, cells have higher phospholipid:sterol ratios and a high level of di-unsaturated fatty acids in the plasma membrane (Uemura et al. 1995; Walters et al. 2003). Thus, cold-acclimated rye protoplasts form exocytotic extrusions of the plasma membrane

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

319

instead of endoplasmic vesicles. Since these extrusions are continuous with the plasma membrane, the surface area of the plasma membrane is recovered as protoplasts are thawed, thus preventing expansioninduced-lysis (Uemura and Steponkus 1999). In addition, the accumulation of peroxidases within acclimated plant cells detoxifies free radicals that are produced during dehydration (Leprince et al. 1996; Walters et al. 2003). Successful cryopreservation requires a balance that minimizes both freezing and desiccation stresses. In the case of orthodox seed development, this balance is achieved during the developmental process of seed maturation when seeds acquire extreme tolerance to desiccation. In vegetative tissues such as some woody dormant buds, it is partially acquired through cold or drought acclimation processes. Other tissue types, such as actively growing shoot tips, can be exposed to specific solutions and cryoprotective treatments to exogenously induce the cellular changes needed to survive the desiccation and freezing processes involved in cryopreservation. The desiccation and freezing tolerances of NPGS clonal germplasm accessions are highly diverse, ranging from cold tolerant species growing in Geneva, New York, to cold sensitive, tropical species in Hilo, Hawaii, Miami, Florida, or Mayaguez, Puerto Rico. Development of successful cryopreservation methods for such a diverse collection requires a careful consideration of the innate mechanisms of protection, the physiology of the propagules, and an optimization of exogenous cryoprotective treatments. VI. CRYOPRESERVATION: VARIABLES TO CONSIDER A. Explant Cryopreservation successes are dependent upon achieving minimal ice formation during the freezing process. The probability of survival is maximized when the selected explant is amenable to the externally imposed stresses of dehydration or cryogenic temperatures. The available form of parent material (field, screenhouse, greenhouse, tissue culture) is a key consideration for cryopreservation. While field materials are often readily available in adequate amounts, they may have high levels of microbial infestation and must be decontaminated during processing. It is also difficult to ensure that field materials are free of viruses. Some field materials lend themselves to dormant bud techniques, providing they are properly acclimated at the active site. Screenhouse and greenhouse materials can be maintained virus-free; however, they are still

320

G. VOLK AND C. WALTERS

prone to endogenous microbial contamination and must undergo a sterilization process either before or after cryopreservation. Since cryopreservation protocols rely on large numbers of explants, the space required to grow enough parent plant material may be quite extensive. Materials maintained in tissue culture are generally axenic. They can be proliferated through hormone manipulations, so explants are usually plentiful. However, the process of decontaminating non-axenic materials and developing initiation and proliferation media may be expensive and time-consuming. In some species, such as mint and strawberry, the presence of endogenous bacteria is a source of contamination in tissue culture (Reed et al. 1995; Tanprasert and Reed 1998). Although there are chemicals that limit the spread of bacteria in cultures, the stresses endured by the plants growing with these media additives may decrease their vigor and thus reduce the survival rate after preservation. The moisture content of propagules before preservation varies from 5 to 10% in dry seeds to 90% moisture in fully hydrated shoot tips (Table 7.4). Each of the propagule types can be successfully dried to a moisture level that depends on degree of acclimation, tissue types, and external treatments. Propagules that are dried through natural developmental processes tend to be more amenable to cryoprotective manipulations. This is probably a result of the accumulation of solutes that replace water and provide additional protection. Explant size is also a key factor in determining the cryopreservability of materials. This has been demonstrated for recalcitrant embryos and presumably similar considerations are needed for vegetative materials that are more hydrated than embryo tissues (Wesley-Smith et al. 2001a). Some of the variability experienced in shoot tip procedures is directly related to the size of the extracted propagule. While larger propagules tend to be more robust and can tolerate the physical damages involved in dissection, cryoprotectant diffusion rates may not be sufficient for survival after exposure to liquid nitrogen. In contrast, very small explants are more likely to survive the cryopreservation process, but are susceptible to the physical damages of extraction and exposure to toxic cryoprotectants. The physiological condition of propagules plays a key role in the health and survival after cryopreservation. Age is influenced by the length of time materials have been maintained in culture, length of prestorage before exposure to cryogenics (in the form of bulbs, seeds, or dormant buds), and the conditions under which the propagules have been stored (temperature, humidity, etc.). Often explants excised from newly introduced, vigorous parent plants can withstand the stresses of dehydration, cooling, and warming better than those that originate from cultures in a less optimal state (B. M. Reed, pers. commun.). However,

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

321

apple shoot tips excised from parent materials stored in tissue culture for extended periods of time before use had higher rates of survival than materials originating from younger cultures (Wu et al. 1999). The higher survival rates may have been caused by the lower water contents present in the older parent cultures. Storing materials at temperatures that reduce, but do not completely inhibit, metabolic processes, prolongs their ability to survive cryoexposure. The length of time that materials can be stored prior to use is dependent upon the source of the material, the storage conditions, and the inherent properties of the germplasm accession. B. Degree of Tolerance 1. Physiologically Achieved Tolerance. Many explants acquire a certain degree of freezing and desiccation tolerance through physiological processes. These physiological changes help explants tolerate liquid nitrogen temperatures during cryopreservation. For example, many seeds undergo a drying process as they mature. This programmed maturation includes a period of dry matter accumulation when storage reserves and LEA proteins are formed. Solutes accumulate within cells and may enhance survival of seeds under dry conditions (Walters et al. 2002). Embryos have few vacuoles, resulting in less shrinkage than occurs in highly vacuolated cells (Farrant and Walters 1998). By optimizing the moisture content and storing at subzero temperatures, mature, orthodox seeds can usually be preserved for decades. Analogously, the process of cold acclimation induces physiological changes in vegetative materials that improves the survival of materials after the dehydration, cooling, and warming procedures. Vegetative materials accumulate sugars and other solutes during the cold acclimation process (Sakai and Yoshida 1968). Vegetative organs can also accumulate LEA proteins, most frequently in response to desiccation stresses. These have been characterized most thoroughly in resurrection plants that are able to survive dehydration from full turgor down to 4% of its original water content (Cohen and Bray 1992; Michel et al. 1994; Scott 2000; Ditzer et al. 2001; Hilbricht et al. 2002). There are some germplasm collections, such as Carya, Pyrus, and Prunus, that are fairly cold hardy, yet they do not receive enough cold at the active site to fully acclimate (Brownwood, Texas; Corvallis, Oregon; and Davis, California, respectively). Efforts are underway to determine if full acclimation can be induced artificially by chilling either budwood or potted plant materials in coolers or growth chambers. If this is possible, dormant budwood methods could possibly be applied to these large collections (L. E. Towill, pers. commun.). As the process of

322

G. VOLK AND C. WALTERS

dormancy in tree crops becomes better understood, alternative methods can be proposed to artificially acclimate materials for application in cryopreservation. The preculture conditions of both vegetative parent and explant materials have a definite impact on the success of cryopreservation efforts. By cold-acclimating some temperate genera, improved survival rates are observed. Chilling treatment at 3° to 5°C for one day to four weeks has been demonstrated to be effective in axenic apple, blackberry, garlic, kiwifruit, and Prunus (Reed 1988, 1993; Niwata 1995; Helliot and de Boucaud 1997; Chang and Reed 1999; Wu et al. 1999; Wu et al. 2001b). 2. Externally Imposed Tolerance. Cellular dehydration and improved freezing tolerance can occur in materials that are physiologically acclimated after exposure to sugars or cryoprotectant solutions. Exposure to high concentrations of osmotica results in higher survival rates after cryopreservation in some species, such as potato (Grospietsch et al. 1999). These treatments may mimic cold and drought acclimation processes. Tolerance to freezing can be improved in banana, date palm, sugarcane, and Ribes by pretreating materials with 0.3 to 0.75 M sucrose (Panis and Swennen 1985; Bagniol and Engelmann 1991; Gonzalez-Arnao et al. 1996; Panis et al. 1996; Dumet et al. 2000). For plants, two of the more commonly used cryoprotectants are PVS2 (Plant Vitrification Solution 2: 30% glycerol, 15% ethylene glycol, 15% DMSO, 14% sucrose) used for vitrification and PGD (10% PEG, 10% glycerol, 10% DMSO) (Sakai et al. 1990, 1991; Chang and Reed 1999) used for two-step cooling. Other cryoprotectant solutions have also been used successfully in vitrification and slow cooling methods of plant protoplasts and shoot tips (Langis and Steponkus 1991; Sakai 1993). Cryoprotectants are osmotically active and dehydrate cells to limit intracellular ice formation (Finkle et al. 1985; Sakai 1993). Sugars contained in these solutions as well as some detergents or amphiphilic substances may penetrate cell membranes and also help to stabilize lipids and proteins during the dehydration, cooling, and warming processes. Better understanding the biophysical characteristics, such as rates of permeation in cells, phase behavior, and rheological properties of the individual components and composite solutions of cryoprotectants will enable improved cryopreservation techniques to be developed for diverse accessions. Prolonged exposure or high concentrations of cryoprotectants such as DMSO, glycerol, and ethylene glycol damages cells, and it is unclear whether this arises from desiccation stress or a direct chemical effect (Fahy 1984). Although cryoprotectants can be shown to influence behavior of isolated enzymes (Jackson and Mantsch 1991) and membrane structure (Yu and Quinn 1998), it is unknown whether these events are

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

323

a basis of toxicity. Glycerol, DMSO, and ethylene glycol are highly effective in allowing cells to survive freezing but can damage membranes (Finkle et al. 1985; Fahy 1986). Optimization of cryoprotectant concentration and timing is critical, since some membrane damage may increase the permeability of cells and permit entry of larger cryoprotective compounds, while excessive exposure to cryoprotectants is lethal (Finkle and Ulrich 1979; Finkle et al. 1985). C. Dehydration Successful protection from freezing damage depends on the rate and extent of cellular dehydration prior to cryoexposure. This rate is dependent upon the size of explant, the type of drying procedure, and the permeability of the system to water loss. The extent of drying is a combination of difference in water potentials inside and outside the cell and the permeability of the cell to water loss. Drying is often accomplished by exposing hydrated tissues to a defined relative humidity (RH). Desired moisture levels can be precisely achieved by equilibrating materials at a known relative humidity and temperature, as has been demonstrated for orthodox seeds (Walters 2002). Alternatively, water content of explants is manipulated by drying for a specified time at a low relative humidity, assuming time courses have been previously determined and are repeatable. The latter is frequently used for excised embryos of recalcitrant seeds (Pammenter et al. 1998; Pence 1995; Wesley-Smith et al. 2001a; but see Liang and Sun 2000) and shoot tips encapsulated into alginate beads (Takagi 2000; Towill and Bajaj 2002). Though rapid drying allows desiccationsensitive cells to survive to lower water contents (Pammenter et al. 1998; Walters et al. 2001), drying is uneven and may exacerbate desiccation or freezing damage if parts of explants are over- or under-dried, respectively (Pammenter et al. 1998, Wesley-Smith et al. 2001b). Encapsulating explants is believed to provide more gradual, but uniform, water adjustments (Takagi 2000). Slow cooling in a two-step cooling protocol allows water to migrate to extracellular spaces. The second step is a plunge into liquid nitrogen. Two-step cooling is frequently used with a diverse array of germplasm and the optimum cooling rates for the first step are species dependent largely because of differences in the water permeability among cells. In the case of dormant woody materials, nodal sections are dried to about 25–30% moisture and cooled at about 0.5 to 1°C⋅hr–1 to –30°C before plunging into liquid nitrogen (Forsline et al. 1998). Shoot tips in a cryoprotectant solution can also be dehydrated by slow cooling (0.5° to 1°C per minute). Ice is usually nucleated at a temperature just below the

324

G. VOLK AND C. WALTERS

freezing point of the cryoprotectant solution. Cryovials containing shoot tips are then cooled at the same rate to –30° to –35°C. Since the solution surrounding explants has already been nucleated and the explants are sufficiently dehydrated, they can survive a plunge into liquid nitrogen within cryovials (Reed 1988, 1993; Chang and Reed 1999). A preculture step in highly osmotic media (often sucrose) also serves to dehydrate shoot tips. The relatively low membrane permeability of some solutes such as sucrose and polyethylene glycol (PEG) within pretreatment solutions causes water to move extracellularly and cells become dehydrated. Although the kinetics of water movement from hydrated cells deep within compacted apical shoot tips are not fully characterized, this method provides improved survival after cryopreservation in crops such as banana, sugarcane, and olive (GonzalezArnao et al. 1996; Panis et al. 1996; Martinez et al. 1999). D. Cooling Rates The success of cryopreservation procedures is dependent upon the rate at which explants are cooled to cryogenic temperatures. The effects of cooling rate can be divided into four categories: (1) very slow cooling (< 0.02°C⋅min–1) or prolonged exposure to temperatures between –5° and –35°C can result in desiccation damage in desiccation-sensitive cells. Membrane and protein damage may result from close appression of molecules or elevated salt concentrations (described previously in the section on desiccation damage). Cells with relatively high water permeability may be more damaged by slower cooling than cells with low water permeability. Slow cooling can cause desiccation damage in cells with relatively high water permeabilities, while it can enhance survival of cells with low water permeabilities (Vertucci 1989a). (2) Higher cooling rates allow cells to dehydrate sufficiently to avoid intracellular ice but not so much that they are damaged by desiccation. This optimum cooling rate ranges from 0.5 to 3°C⋅min–1, depending on the species as well as the genotype, physiological status, explant size, cell size, surface/volume ratio, and hydraulic conductivity of the membrane (Sakai and Yoshida 1967; Mazur 1969). Cells with greater water permeabilities tolerate higher cooling rates. (3) If cooled at rates faster than optimum, water does not migrate sufficiently from the cell and the cell is killed by intracellular ice formation. Cooling rates within this supra-optimal region range from 1°C⋅min–1 to 2500°C⋅min–1, again dependent on the properties of the cell or explant. (4) Finally, cooling rates that bring cells to the glass transition temperature faster than it takes for intracellular ice to nucleate and grow can alleviate both desiccation and freezing dam-

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

325

age. This ultra-rapid cooling is the principle behind freeze-fracture electron microscopy (Bald 1987). Required cooling rates vary according to the intracellular viscosity, which is largely a function of the water content of cells (Buitink et al. 1998) and the presence of protectants (Franks 1985; Fahy et al. 1987). Survival of dried tissues is nearly independent of cooling rate, while hydrated tissues must be cooled faster than 6000°C⋅min–1. Vitrification methods of hydrated materials rely on a very rapid plunge into liquid nitrogen. The type of vessel as well as the size and water content of the explant determines the rate of cooling (Table 7.5) (Bald 1987; Wesley-Smith et al. 2001a). Even fairly hydrated small explants (1 mm3) will vitrify when the cooling process is very rapid (WesleySmith et al. 2001a). With larger, hydrated explants, it becomes increasingly difficult to achieve a cooling rate that is fast enough to prevent ice crystal formation. The type of freezing vessel affects the cooling rate of explants as shown in Table 7.5. Very rapid cooling rates can be achieved by placing explants on aluminum foil or on grids that allow direct exposure to partially solidified liquid nitrogen. The use of polyethylene straws or cryovials reduces the rate of cooling because they are not as conductive and they are larger. While the type of cooling vessel is critical to achieve the very rapid cooling rates needed for survival of larger, hydrated explants (up to 8 mm3), smaller, dehydrated explants (1 mm3) can be successfully cryopreserved using a wider assortment of vessels. Rapid cooling is a critical part of the second step in two-step cooling procedures. It is not difficult to achieve rapid cooling in systems at subzero temperatures where the extracellular solution is already frozen because the thermal conductivity of ice is much less than that of liquid solutions. By studying cooling and warming processes with respect to the biophysical responses within cells, methods can be developed to avoid events that are lethal (Benson et al. 1996; Dumet and Benson 2000; Dumet et al. 2000; Martinez et al. 2000).

E. Recovery Sample warming rates must also be optimized to achieve high success rates after cryopreservation (Sakai 1966, 1985; Bagniol and Engelmann 1992). In many techniques, sealed cryovials are warmed by immersion in a 40–45°C water bath for a few minutes (warming rate is between 1 and 3°C⋅sec–1). If the samples are warmed too slowly, ice crystals may form (devitrification) and/or recrystallize during warming (Meryman

326 Table 7.5.

Cooling rates achieved for different containers, plant materials, and cryogens.

Cooling mediaz

Cooling vessel

Cooling rate (°C⋅sec–1)

Measurement range (°C)

Reference

Foil strip (Aluminum)

LN Supercooled LN

–106 –111

0 to –150 0 to –150

Waddell (pers. commun.) Waddell (pers. commun.)

Continental Plastics Polyethylene straw (1.2 mL, 6.4 cm) 3M Polyolefin straw (1 inch diameter) with dry pea seeds

LN Supercooled LN Vapor phase LN

–43 –43 –1

0 to –150 0 to –150 0 to –150

Waddell (pers. commun.) Waddell (pers. commun.) Vertucci 1989a

Corning cryovial (1.2 mL) with 1 ml liquid Corning cryovial (1.2 mL) with 1ml liquid cooled to –35°C Corning cryovial (1.2 mL) with whole dry pea seeds

LN Supercooled LN LN

–3 –4 –5

0 to –150 0 to –150 0 to –150

Waddell (pers. commun.) Waddell (pers. commun.) Waddell (pers. commun.)

LN

–3

0 to –150

Vertucci 1989a

Individual pea seed

LN

–12

0 to –150

Vertucci 1989a

Inside an axis (dry mass 6 mg) 1.5 g/g H2O 0.8 g/g H2O 0.5 g/g H2O 1.5 g/g H2O 0.8 g/g H2O 0.5 g/g H2O

Supercooled LN, plunged at 1.2 m⋅s–1 Supercooled LN, plunged at 1.2 m⋅s–1 Supercooled LN, plunged at 1.2 m⋅s–1 Isopentane –140°C, plunged at 1.2 m⋅s–1 Isopentane –140°C, plunged at 1.2 m⋅s–1 Isopentane –140°C, plunged at 1.2 m⋅s–1

–94 –170 –283 –52 –61 –110

0 to –110 0 to –110 0 to –110 0 to –110 0 to –110 0 to –110

Wesley-Smith et al. 2001 Wesley-Smith et al. 2001 Wesley-Smith et al. 2001 Wesley-Smith et al. 2001 Wesley-Smith et al. 2001 Wesley-Smith et al. 2001

z

Liquid nitrogen (LN) was supercooled by applying a vacuum.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

327

and Williams 1985; Rasmussen and Luyet 1969; Franks 1985; Vertucci 1989b). Once warmed, vitrified samples are usually diluted into a high sucrose solution prior to plating onto appropriate media for shoot induction. Incubation for thirty minutes in 1.2 M sucrose is believed to slow water entry into cells, alleviating damage from rapid cell expansion. Growth media optimization plays a key role in improving plant regeneration after cryopreservation. For example, an ammonia-free media improved growth for recovering sweet potato explants (Pennycooke and Towill 2001). In garlic explants, both the primary meristem and the axillary shoots will elongate in media containing cytokinins. By increasing the number of regenerating shoots, the chances of regenerating a plant is higher than when only one shoot forms from an explant (G. Volk, unpubl.). Explants recovering from liquid nitrogen exposure have an increased susceptibility to photooxidative damage (Benson 1990). Exposure to light after desiccation and low temperature stresses results in the production of reactive oxygen species and free radicals that can cause permanent damage to enzymes, membranes, and chromosomes (Wolff et al. 1986; Chan 1987; McKersie et al. 1988; Dean et al. 1993; Dizdaroglu 1994; Halliwell and Gutteridge 1999; Leprince et al. 2000). The involvement of ethylene with reactive oxygen species production in programmed cell death-induced cells have led some to speculate that programmed cell death plays a role in plant cells damaged by cryoexposure (Benson 1990). Regeneration of plants from callus has been shown to increase the rates of somoclonal variation in resulting propagules, though a quantitative assessment of the risk is not presented (Scowcroft 1984; Lazanyi 1994). To limit somoclonal variation, it is preferable to use media that do not induce callus formation. Some reports indicate that there is still some variability that occurs after freezing, but this is thought to be a result of variations present in the original explants (Ashmore 1997). Somoclonal variation may be an intrinsic risk of in vitro culture and cryopreserving explants. When quantitative assessments of the risk are known, the number of stored explants can be increased to ensure that true-to-type regenerants can be easily retrieved. Some cryopreservation methods are robust for a given species (Malus), while others, such as potato, pear, and Rubus, yield highly variable survival rates across the many diverse accessions within the collection (Towill 1984, 1988b; Reed 1993). Some of the challenges in obtaining uniform results will be overcome when the chemical and biophysical aspects of preconditioning, dehydration, cooling, and warming are better understood (Towill 1990). Ideally, all of these stages of the

328

G. VOLK AND C. WALTERS

cryopreservation process should be optimized for the highest success rates, but a thorough evaluation of even one stage may yield dramatic improvements in survival. As methods are optimized, clonal germplasm collections will be cryopreserved more efficiently. VII. APPLICATION OF CRYOPRESERVATION TECHNOLOGIES TO VEGETATIVE MATERIALS A. Germplasm Classification Since the physiologies of vegetatively propagated germplasm collections are very diverse, ranging from cold-hardy temperate accessions to cold-sensitive tropical accessions, selection of the appropriate and applicable cryopreservation method is critical. Dormant bud techniques are most uniformly successful for woody collections such as Malus and Prunus cerasus that undergo extensive cold acclimation during the fall and winter seasons (Forsline et al. 1998; Towill and Forsline 1999). Axenic shoot tip systems are successful for most temperate plants as well as some tropical collections (Niino et al. 1992; Niino and Sakai 1992; Sakai 1984; Suzuki et al. 1997). Temperate accessions include genera that can survive chilling to some extent. These include most of the collections maintained at the Corvallis, Oregon, and Davis, California, Repositories, as well as Allium and Solanum (Niwata 1995; Helliot et al. 1997; Niino et al. 1997; Hirai and Sakai 1999; Makowska et al. 1999; Shatnawi et al. 1999). Although there are reports of successful vitrification of shoot tips from tropical species, these collections can often be difficult to cryopreserve (Gonzalez-Arnao et al. 1998; Thinh et al. 1999). Tropical accessions are maintained primarily at Riverside, California, Hilo, Hawaii, Miami, Florida, and Mayaguez, Puerto Rico, Clonal Repositories. Table 7.6 classifies the NPGS clonal collections with elite genotypes according to proposed strategy for preservation using a dormant bud or axenic shoot tip approach (or both) and if they are considered temperate or tropical in nature. B. Cryopreservation Methods 1. Dormant Buds. The dormant budwood cryopreservation method is less labor intensive than tissue culture techniques. Budwood samples are collected directly from the field between the times that cold acclimation is maximized and veralization requirements are satisfied. One method uses scions that are cut into one-inch sections, dehydrated at –5°C until the moisture is approximately 25–30% (fresh weight basis),

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT Table 7.6.

329

Potential propagules for the cryopreservation of NPGS clonal collections.

Temperate dormant bud

Temperate shoot tip

Tropical shoot tip

Asimina Carya Cydonia Juglans Malus Morus Olea Prunus Pyrus Vitis Pistacia Corylus

Actinidia Allium Corylusz Diospyros Diospyros Ficus Fragariaz Humulus Menthaz Olea Pistacia Prunus Punica Pycnanthemum Pyrusz Ribesz Rubusz Solanumz Vacciniumz Vitisz

Ananasz Annona Arachis Artocarpusz Averrhoa Bamboo Canarium Carica Citrus Dimocarpus Fortunella Garcinia Ipomoeaz Litchi Macadamia Mangifera Manilkara Musa Nephelium Passiflora Persea Phoenix Pouteria Psidium Saccharum Theobroma Tripsacum

z

At least a portion of the collection is maintained in tissue culture conditions.

packaged, cooled at 1°C⋅h–1 to between –25° and –35°C, and then plunged into vapor phase above liquid nitrogen (–165° to –180°C). For recovery, sections are warmed slowly, rehydrated in moist peat, and then grafted to appropriate rootstocks (Stushnoff and Seufferheld 1995; Forsline et al. 1998; Towill and Forsline 1999). Since the buds are grafted, new trees are established within a single season and can be transferred to the field. Budwood resulting from the growth of grafted material can be distributed to germplasm customers during the following season. Even within a genus like Malus, there are some accessions that have low survival rates after the cryopreservation process. Cold-tender accessions tend to have a lower recovery than cold-hardy accessions (Forsline et al. 1998). Similar results have been obtained over multiple seasons and it is clear that there is a genetic basis for the success or failure of this

330

G. VOLK AND C. WALTERS

method. Difficult accessions often can be cryopreserved using alternative methods, such as an axenic shoot tip system (Stushnoff and Seufferheld 1995). Efforts to cryopreserve collections require decisions to be made about which method is most likely to be successful for the most accessions and then which alternative methods can be developed for the remaining accessions. A factor limiting the success of the dormant bud method in some genera is the lack of reliable grafting procedures. Sometimes the cryopreservation process is successful based on viability assessments of electrolyte leakage or cambium browning, but the grafting recovery success rate is low. Some of the species that tend to be difficult to graft may be recovered using tissue culture methods. Buds can be sterilized, dissected, and introduced into tissue culture. Once established, they can be rooted and transferred to soil. This method is dependent upon the existence of tissue culture media formulae and elimination of endogenous bacteria and fungi. There is an increased length of time before retrieved materials can be distributed to customers. 2. Axenic Shoot Tips. Optimized two-step cooling, vitrification, or encapsulation-dehydration cryopreservation methods for shoot tips of a specific accession can yield highly successful results. However, finding the optimum method to apply across many accessions within a species is a challenge. Some reports compare the relative success achieved with common cryopreservation methods (Reed 2001b; Wu et al. 2001b). In the case of ‘Golden Delicious’ apple buds, vitrification resulted in an almost two-fold higher survival than the encapsulationdehydration technique (Wu et al. 2001a). Reed (2001b) compared vitrification, two-step cooling, and encapsulation cryopreservation techniques using in vitro Ribes shoot tips and found significant differences between the success of the various procedures at the two locations, but that the vitrification and encapsulation-dehydration methods were more repeatable than the two-step cooling method. The variable successes obtained at the two laboratories support the importance of personnel training, equipment, and well-defined protocols for successful cryopreservation (Reed et al. 2001b). While comparative studies are very useful in determining preferred cryopreservation methods, they can be misleading because the optimization of a single variable for a given protocol may drastically improve results. Two-Step Cooling. Sterile, dissected explants are placed on media containing some cryoprotective components (usually high sugar concentration; see Section VIB) for a few days. They are then exposed to a

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

331

cryoprotectant such as PGD (10% PEG8000, 10% glucose, 10% DMSO) for an hour on ice in cryovials. Samples are cooled slowly (0.5 to 1°C⋅min–1) to –35°C (with a nucleation step at approximately –9°C) and then plunged into liquid nitrogen and stored. Vials are thawed for 1 minute in 45°C water and kept in 23°C water until the ice melts. Cryoprotectants are removed, shoot tips are drained, and plated on recovery medium for regrowth (Reed 1988, 1993; Chang and Reed 1999). This method relies upon relatively non-toxic cryoprotectants; but its success is limited to temperate species (Reed 2001a). Preincubation media and duration, exposure time to cryoprotectants, and cooling rate may need to be optimized among accessions of a species. Vitrification. Shoot tips or lateral buds are excised and pretreated by chilling for 2 days or by exposure to high sucrose solutions or cryoprotectants. Often, explants are treated with a solution of sucrose/glycerol for up to several hours immediately before immersion in the cryoprotectant solution (at 0°C). A common cryoprotectant is PVS2 (30% glycerol, 14% sucrose, 15% DMSO, 15% ethylene glycol) (Sakai et al. 1990, 1991). Tissues can be adequately dehydrated by optimal exposure to PVS2. Explants are placed either on foil strips, into vials, or into freezing straws before they are plunged into liquid or partially solidified liquid nitrogen. For recovery, samples are thawed, diluted in 1.2 M sucrose, then plated onto medium. The size and condition of explant, preincubation solutions types, durations, and temperatures, length of exposure to PVS2, rate of cooling, and warming must all be optimized for this protocol to be successful. The success of vitrification is thought to be dependent upon the biophysical attributes of the PVS2 solution (Sakai et al. 1990, 1991). Exposure of explants to high-osmoticum solutions prior to PVS2 immersion results in a dehydration effect critical for high survival. The combined effect of dehydration and cryoprotection allows glasses to form when small explants are rapidly cooled (Steponkus et al. 1992). Vitrification has been successfully applied to many species, including those of tropical origin such as banana, cassava, pineapple, taro, orchids, and yam (Sakai 2000). Encapsulation-Dehydration. Once explants are dissected from axenic materials, they are encapsulated into calcium-alginate beads. These 5 mm diameter beads partially control the diffusion rates of chemicals into the plant material. Encapsulated explants are gradually exposed to sugar solutions of increasing osmotic potential. Beads are then air-dried to 20–30% moisture prior to plunging into liquid nitrogen. For recovery,

332

G. VOLK AND C. WALTERS

vials containing beads are thawed in warm water; the beads are removed and placed onto media. The encapsulation-dehydration procedure relies on a gradual incorporation of sugars and then a slow drying process to induce the cellular changes necessary for successful cryopreservation. This method has been successful for a number of genera, including strawberry (Hirai et al. 1998), kiwifruit (Bachiri et al. 2001; Wu et al. 2001b), potato (Hirai and Sakai 1999), sugarcane (Gonzalez-Arnao et al. 1999), apple (Zhao et al. 1999), and grape (Plessis et al. 1993; Zhao et al. 2001). Modifications of this protocol involve treating alginate beads with various cryoprotectant solutions prior to dehydration and freezing. 3. Somatic Embryos. There are numerous reports on the cryopreservation of somatic embryos or embryogenic suspensions in the literature (Bajaj 1995). Somatic embryos are embryogenic tissue culture lines often derived from immature cotyledons. Because somatic embryos often exhibit similar levels of tolerance as their zygotic embryo counterparts, they may provide an effective means to cryopreserve clonal materials that have low survival using traditional methods (de Boucoud and Brison 1995; Jekkel et al. 2002). For survival, the root and shoot apical meristems as well as the hypocotyls must survive cryogenic manipulations. Direct regrowth (non callus) is observed primarily with immature embryos (Dereuddre and Kaminski 1992). Larger embryos tend to form callus. While optimized cryopreservation methods have been developed for many genera, they are successful only if methods for successful embryo formation and subsequent plant regeneration are well established. The expense of developing these systems for the diversity within NPGS clonal collections has limited their use.

VIII. CONCLUSIONS Genetic resources in the NPGS are an international treasure available for breeding and research. It is critical that plant materials remain viable and usable. Accessions maintained at the NCGRP duplicate collections at active sites and insure against the loss of genetic resources when unexpected events occur. Each collection should be managed based on the physiological and genetic characteristics of the plants. However, the diverse array of materials in the NPGS collections requires a broad understanding and application of principles to ensure safe preservation in a consolidated backup collection. Seeds may be a more appropriate storage propagule for wild accessions that are currently maintained clonally. Genetic diversity of germplasm

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

333

that is valued for genes rather than specific combinations of genes can be preserved as seeds or pollen if the appropriate number of crosses are made and offspring collected. Other considerations, such as time to maturation, need to be evaluated so that germplasm is readily available to breeders in the appropriate form. The greatest challenge to successfully maintaining vegetative materials in the base collection is finding the appropriate propagule and developing protocols that maintain viability when tissues are subjected to extreme desiccation and low temperature stresses inherent in cryopreservation.

LITERATURE CITED Acheampong, E. 1996. In vitro genebank management of clonally propagated crops under minimal conditions. In: F. Engelmann (ed.), Management of field and in vitro germplasm collections. Proc. of a consultation meeting, 15–20 January 1996, CIAT, Cali, Colombia. Intl. Plant Genetic Res. Inst., Rome. Antikainen M., and M. Griffith. 1997. Antifreeze protein accumulation in freezing tolerant cereals. Physiol. Plantar. 99:423–432. Antolin, M. F. 1998. Genetic assessment of strains, varieties and ecotypes. In: T. Gass, W. Podyma, J. Puchalski, and S. A. Eberhart (compilers), Challenges in rye germplasm conservation. Proc. of an international conference on crop germplasm conservation with special emphasis on rye, and an ECP/GR Workshop, 2–6 July 1996, Warsaw/ Konstancin-Jeziorna, Poland. Intl. Plant Genetic Res. Inst., Rome. Ashmore, S. E. 1997. Status report on the development and application of in vitro techniques for the conservation and use of plant genetic resources. Intl. Plant Genetic Res. Inst., Rome. Bachiri, Y., G. Q. Song, P. Plessis, A. Shoar-Ghaffari, T. Rekab, and C. Morisset. 2001. Routine cryopreservation of kiwifruit (Actinidia spp) germplasm by encapsulationdehydration: Importance of plant growth regulators. Cryo-Letters 22:61–74. Bagniol S., and F. Engelmann. 1991. Effects of pregrowth and freezing conditions on the resistance of meristems of date palm. Cryo-Letters 12:279–286. Bagniol S., and F. Engelmann. 1992. Effect of thawing and recovery conditions on the regrowth of meristems of date palm (Phoenix dactylifera L.) after cryopreservation in liquid nitrogen. Cryo-Letters 13:253–260. Bajaj, Y. P. S. 1995. Cryopreservation of plant cell, tissue, and organ culture for the conservation of germplasm and biodiversity. p. 3–28. In: Y. P. S. Bajaj (ed.), Biotechnology in agriculture and forestry 32. Springer, New York. Balakrishnan, R., N. V. Nair, and T. V. Sreenivasan. 2000. A method for establishing a core collection of Saccharum officinarum L. germplasm based on quantitative-morphological data. Genetic Res. Crop Evol. 47:1–9. Bald, W. B. 1987. Quantitative cryofixation. Adam Hilger, IOP Publishing Ltd. Philadelphia. Beardmore, T., and W. Vong. 1998. Role of the cotyledonary tissue in improving low and ultralow temperature tolerance of butternut (Juglans cinerea) embryonic axes. Can. J. For. Res. 28:903–910.

334

G. VOLK AND C. WALTERS

Benson, E. E. 1990. Free radical damage in stored plant germplasm. Intl. Board for Plant Genetic Res., Rome. Benson, E. E., B. M. Reed, R. M. Brennan, K. A. Clacher, and D. A. Ross. 1996. Use of thermal analysis in the evaluation of cryopreservation protocols of Ribes nigrum L. germplasm. Cryo-Letters 17:347–362. Bretting, P. K., and M. P. Widrlechner. 1995. Genetic markers and plant genetic resource management. Plant Breed. Rev. 13:11–86. Brown, A. H. D. 1989. The case for core collections. p. 136–156. In: A. H. D. Brown, O. H. Frankel, D. R. Marshall, and J. T. Williams (eds.), The use of plant genetic resources. Cambridge Univ. Press, New York. Brown, A. H. D. 1995. The core collection at the crossroads. p. 3–19. In: T. Hodgkin, A. H. D. Brown, T. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Intl. Plant Genetic Res. Inst., Wiley, New York. Brown, A. H. D., C. L. Brubaker, and J. P. Grace. 1997. Regeneration of germplasm samples: Wild versus cultivated plant species. Crop Sci. 37:7–13. Bryant, G., K. L. Koster, and J. Wolfe. 2001. Membrane behaviour in seeds and other systems at low water content: the various effects of solutes. Seed Sci. Res. 11:17–25. Buitink, J., M. M. A. E. Claessens, M. A. Hemminga, and F. A. Hoekstra. 1998. Influence of water content and temperature on molecular mobility and intracellular glasses in seeds and pollen. Plant Physiol. 118:531–541. Chan, H. W.-S. 1987. Autoxidation of unsaturated lipids. Academic Press, London. Chang, Y., and B. M. Reed. 1999. Extended cold acclimation and recovery medium alteration improve regrowth of Rubus shoot tips following cryopreservation. Cryo-Letters 20:371–376. Charlesworth, D., B. Charlesworth, and G. A. T. McVean. 2001. Genome sequences and evolutionary biology, a two-way interaction. Trends Ecol. Evol. 16:235–242. Chavarriaga-Aguirre, P., M. M. Maya, J. Tohme, M. C. Duque, C. Iglesias, M. W. Bonierbale, S. Kresovich, and G. Kochert. 1999. Using microsatellites, isozymes and AFLPs to evaluate genetic diversity and redundancy in the cassava core collection and to assess the usefulness of DNA-based markers to maintain germplasm collections. Mol. Breed. 5:263–273. Clark, R. L., H. L. Shands, P. K. Bretting, and S. A. Eberhart. 1997. Germplasm regeneration: Developments in population genetics and their implications. Crop Sci. 37:1–6. Cohen, A., and E. A. Bray. 1992. Nucleotide sequence of an ABA-induced tomato gene that is expressed in wilted vegetative organs and developing seeds. Plant Mol. Bio. 18:411–413. Connor, K. F., and L. E. Towill. 1993. Pollen-handling protocol and hydration/dehydration characteristics of pollen for application to long-term storage. Euphytica 68:77–84. Crossa, J. 1989. Methodologies for estimating the sample size required for genetic conservation of outbreeding crops. Theor. Appl. Genet. 77:153–161. Crossa, J., and R. Vencovsky. 1999. Sample size and variance effective population size for genetic resources conservation. Plant Genetic Res. Newslett. 119 Supp.:15–25. Crossa, J., C. M. Hernandez, P. Bretting, S. A. Eberhart, and S. Taba. 1993. Statistical genetic considerations for maintaining germ plasm collections. Theor. Appl. Genet. 86:673–678. Crowe, J. H., J. F. Carpenter, and L. M. Crowe. 1998. The role of vitrification in anhydrobiosis. Annu. Rev. Physiol. 60:73–103. De Boucaud, M. T., and M. Brison. 1995. Cryopreservation of germplasm of walnut (Juglans species). Biotechnol. Agr. Forestry 32:129–147. Dean, R. E., J. A. Dahlberg, M. S. Hopkins, S. E. Mitchell, and S. Kresovitch. 1999. Genetic redundancy and diversity among ‘Orange’ accessions in the U.S. National Sorghum Collection as assessed with simple sequence repeats. Crop Sci. 39:1215–1221.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

335

Dean, R. T., S. Gieseg, and M. J. Davies. 1993. Reactive species and their accumulation on radical-damaged proteins. Trends Biochem. Sci. 18:37–41. del Rio, A. H., J. B. Bamberg, Z. Huaman, A. Salas, and S. E. Vega. 2001. Association of ecogeographical variables and RAPD marker variation in wild potato populations of the USA. Crop Sci. 41:870–878. Dereuddre, J., and M. Kaminski. 1992. Applications of thermal analysis in cryopreservation of plant cells and organs. J. Therm. Anal. 38:1965–1978. Ditzer, A., H.-H. Kirch, A. Nair, and D. Bartels. 2001. Molecular characterization of two alanine-rich Lea genes abundantly expressed in the resurrection plant C. plantagineum in response to osmotic stress and ABA. J. Plant Physiol. 158:623–633. Dizdaroglu, M. 1994. Chemical determination of oxidative DNA-damage by gaschromatrography mass-spectrometry. Oxygen Radicals Bio. Syst. 234:3–16. Dumet, D., and E. E. Benson. 2000. The use of physical and biochemical studies to elucidate and reduce cryopreservation-induced damage in hydrated/desiccated plant germplasm. p. 43–56. In: F. Engelmann and H. Takagi (eds.), Cryopreservation of tropical plant germplasm. Current research progress and application. Japan Intl. Research Center for Agricultural Sciences, Tsukuba, Japan/Intl. Plant Genetic Res. Inst., Rome. Dumet, D., W. Block, R. Worland, B. M. Reed, and E. E. Benson. 2000. Profiling cryopreservation protocols for Ribes ciliatum using differential scanning calorimetry. CryoLetters 21:367–378. Eberhart, S. A., and H. E. Bockelman. 1994. Plant germplasm resources. Proceedings 2nd international Triticeae symposium. Logan, Utah. June 20–24, 1994. p. 90–100. Eberhart, S. A., E. E. Roos, and L. E. Towill. 1991. Strategies for long-term management of germplasm collections. p. 135–145. In: D. A. Falk and K. E. Holsinger (eds.), Genetics and conservation of rare plants. Oxford Univ. Press, New York. Engelmann, F. 1997. Importance of desiccation for the cryopreservation of recalcitrant seed and vegetatively propagated species. Plant Genetic Res. Newslett. 112:9–18. Fahy, G. M. 1984. Cryoprotectant toxicity: biochemical or osmotic? Cryo-Letters 5:79–90. Fahy, G. M. 1986. The relevance of cryoprotectant “toxicity” to cryobiology. Cryobiology 23:1–13. Fahy, G. M., D. I. Levy, and S. E. Ali. 1987. Some emerging principles underlying the physical properties, biological actions, and utility of vitrification solutions. Cryobiology 24:196–213. Fahy, G. M., D. R. MacFarlane, C. A. Angell, and H. T. Meryman. 1984. Vitrification as an approach to cryopreservation. Cryobiology 21:407–426. Farrant, J. M., and C. Walters. 1998. Ultrastructural and biophysical changes in developing embryos of Aesculus hippocastanum in relation to the acquisition of tolerance to drying. Physiol. Plantar. 104:513–524. Finkle, B. J., and J. M. Ulrich. 1979. Effects of cryoprotectants in combination on the survival of frozen sugarcane cells. Plant Physiol. 63:598–604. Finkle, B. J., M. E. Zavala, and J. M. Ulrich. 1985. Cryoprotective compounds in the viable freezing of plant tissues. In: K. Kartha (ed.), Cryopreservation of plant cells and organs. CRC Press, Boca Raton, FL. Forsline, P. L., and R. D. Way. 1993. Apple accessions of low priority targeted for removal from the National Plant Germplasm System. Fruit Var. J. 47:204–214. Forsline, P. L., L. E. Towill, J. W. Waddell, C. Stushnoff, W. F. Lamboy, and J. R. McFerson. 1998. Recovery and longevity of cryopreserved dormant apple buds. J. Am. Soc. Hort. Sci. 123:365–370. Franks, F. 1985. Biophysics and biochemistry at low temperatures. Cambridge Univ. Press, New York.

336

G. VOLK AND C. WALTERS

Ganeshan, S. 1985. Storage and longevity of papaya (Carica papaya L. ‘Washington’) pollen: I. Effect of low temperature and humidity. Gartenbauwissenschaft 50:227–230. Ganeshan, S., and M. P. Alexander. 1986. Effect of freeze-drying on pollen germination in vitro in papaya (Carica papaya L. cv. Washington) and tomato (Lycopersicon esculentum Mill. cv. Arka Vikas). Gartenbauwissenschaft 51:17–20. Ganeshan, S., and M. P. Alexander. 1990. Fertilizing ability of cryopreserved grape (Vitis vinifera L.) pollen. Vitis 29:145–150. Gekko, K., and I. Stake. 1981. Differential scanning calorimetry of unfreezable water in water-protein-polyol systems. Agr. Biol. Chem. 45:2209–2217. George, M. F., M. R. Becwar, and M. J. Burke. 1982. Freezing avoidance by deep undercooling of tissue water in winter-hardy plants. Cryobiology 19:628–639. Gonzalez-Arnao, M. T., T. Moreira, and C. Urra. 1996. Importance of pregrowth with sucrose and vitrification for the cryopreservation of sugarcane apices using encapsulationdehydration. Cryo-Letters 17:141–148. Gonzalez-Arnao, M. T., M. M. Ravelo, C. U. Villavicencio, M. M. Montero, and F. Engelmann. 1998. Cryopreservation of pineapple (Ananas comosus) apices. Cryo-Letters 19:375–382. Gonzalez-Arnao, M. T., C. Urra, F. Engelmann, T. Ortiz, and C. de la Fe. 1999. Cryopreservation of encapsulated sugarcane apices: Effect of storage temperature and storage duration. Cryo-Letters 20:347–352. Greene, S. L., and J. B. Morris. 2001. The case for multiple-use plant germplasm collections and a strategy for implementation. Crop Sci. 41:886–892. Greene, S. L., T. C. Hart, and A. Afonin. 1999a. Using geographic information to acquire wild crop germplasm for ex situ collections: I. Map development and field use. Crop Sci. 39:836–842. Greene, S. L., T. C. Hart, and A. Afonin. 1999b. Using geographic information to acquire wild crop germplasm for ex situ collections: II. Post-collection analysis. Crop Sci. 39:843–849. Grospietsch, M., E. Stodulkova, and J. Zamecnik. 1999. Effect of osmotic stress on the dehydration tolerance and cryopreservation of Solanum tuberosum shoot tips. Cryo-Letters 20:339–346. Halliwell, B., and J. M. C. Gutteridge. 1999. Free radicals in biology and medicine. 3rd ed. Oxford Univ. Press, Oxford. Hanna, W. W., and L. E. Towill. 1995. Long-term pollen storage. Plant Breed. Rev. 13:179–207. Haunold, A., and P. C. Stanwood. 1985. Long-term preservation of hop pollen in liquid nitrogen. Crop Sci. 25:194–196. Helliot, B., and M. T. de Boucaud. 1997. Effect of various parameters on the survival of cryopreserved Prunus ferlenain in vitro plantlets shoot tips. Cryo-Letters 18:133–142. Helliot, B., M. Brison, and M. T. de Boucaud. 1997. Cryopreservation of in vitro grown shoot tips of an interspecific Prunus rootstock by slow prefreezing method. Acta Hort. 447:539–541. Hey, J. M., and D. R. MacFarlane. 1996. Crystallization of ice in aqueous solutions of glycerol and dimethyl sulfoxide. Cryobiology 33:205–216. Hilbricht, T., F. Salamini, and D. Bartels. 2002. CpR18, a novel SAP-domain plant transcription factor, binds to a promoter region necessary for ABA mediated expression of the CdeT27-45 gene from the resurrection plant Craterostigma plantagineum Hochst. Plant J. 31:293–303. Hirai, D., and A. Sakai. 1999. Cryopreservation of in vitro-grown meristems of potato (Solanum tuberosum L.) by encapsulation-vitrification. Potato Res. 42:153–160. Hirai, D., K. Shirai, S. Shirai, and A. Sakai. 1998. Cryopreservation of in vitro-grown meristems of strawberry (Fragaria × ananassa Duch.) by encapsulation-vitrification. Euphytica 101:109–115.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

337

Hoekstra, F. A. 1995. Collecting pollen for genetic resources conservation. p. 527–550. In: L. Guarino, V. R. Rao, and R. Reid (eds.), Collecting plant genetic diversity. Technical guidelines. CAB Intl., New York. Hoekstra, F. A., E. A. Golovina, and J. Buitink. 2001. Mechanisms of plant desiccation tolerance. Trends Plant Sci. 6:431–438. Hokanson, S. C., A. K. Szewe-McFadden, W. F. Lamboy, and J. R. McFerson. 1998. Microsatellite (SSR) markers reveal genetic identities, genetic diversity and relationships in a Malus × domestica borkh. core subset collection. Theor. Appl. Genet. 97:671–683. Hokanson, S. C., J. R. McFerson, P. L. Forsline, W. F. Lamboy, J. J. Luby, A. D. Djangaliev, and H. S. Aldwinckle. 1997. Collecting and managing wild Malus germplasm in its center of diversity. HortScience 32:173–176. Holbrook, C. C., W. F. Anderson, and R. N. Pittman. 1993. Selection of a core collection from the U.S. germplasm collection of peanut. Crop Sci. 33:859–861. Holbrook, C. C., M. G. Stephenson, and A. W. Johnson. 2000. Level and geographical distribution of resistance to Meloidogyne arenaria in the U.S. peanut germplasm collection. Crop Sci. 40:1168–1171. Hong, T. D., and R. H. Ellis. 1996. A protocol to determine seed storage behaviour. In: J. M. M. Engels and J. Toll (eds.), IPGRI Tech. Bul. 1. Intl. Plant Genetic Res. Inst., Rome. Hoshino, T., M. Odaira, M. Yoshida, and S. Tsuda. 1999. Physiological and biochemical significance of antifreeze substances in plants. J. Plant Res. 112:255–261. Hu, J., J. Zhu, and H. M. Xu. 2000. Methods of constructing core collections by stepwise clustering with three sampling strategies based on the genotypic values of crops. Theor. Appl. Genet. 101:264–268. Huaman, Z., C. Aguilar, and R. Ortiz. 1999. Selecting a Peruvian sweetpotato core collection on the basis of morphological, eco-geographical, and disease and pest reaction data. Theor. Appl. Genet. 98:840–844. Huaman, Z., R. Ortiz, and R. Gomez. 2000. Selecting a Solanum tuberosum subsp. andigena core collection using morphological, geographical, disease and pest descriptors. Am. J. Potato Res. 77:183–190. Hummer, K. E., and B. M. Reed. 1999. Establishment and operation of a temperate clonal field genebank. p. 29–31. In: F. Engelmann (ed.), Management of field and in vitro germplasm collections. Proc. of a consultation meeting, 15–20 January 1996, CIAT, Cali, Columbia. Intl. Plant Genetic Res. Unit, Rome. Jackson M., and H. H. Mantsch. 1991. Beware of proteins in DMSO. Biochim. Biophy. Acta. 1078:231–235. Jaradat, A. A. 1995. The dynamics of a core collection. p. 179–186. In: T. Hodgkin, A. H. D. Brown, T. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Intl. Plant Genetic Res. Inst., Wiley, New York. Jarret, R. L., and L. E. Towill. 1996. Maintenance of sweet potato germplasm in the S-9 germplasm collection in Griffin, Georgia, USA. In: F. Engelmann (ed.), Management of field and in vitro germplasm collections. Proc. consultation meeting, 15–20 January 1996, CIAT, Cali, Colombia. Intl. Plant Genetic Res. Inst., Rome. Jekkel, Z., J. Kiss, G. Gyulai, E. Kiss, and L. E. Heszky. 2002. Cryopreservation of somatic embryos from Aesculus hippocastanum L. (Horse Chestnut). In: L. E. Towill and Y. P. S. Bajaj (eds.), Cryopreservation of plant germplasm II. Springer-Verlag, Berlin. Kioko, J., P. Berjak, N. W. Pammenter, M. P. Watt, and J. Wesley-Smith. 1998. Desiccation and cryopreservation of embryonic axes of Trichilia dregeana Sond. Cryo-Letters 19:15–26. Kresovich, S., J. R. McFerson, and A. L. Westman. 1997. Using molecular markers in genebanks: identity, duplication, contamination and regeneration. In: W. G. Ayad, T. Hodgkin, A. Jaradat, and V. R. Rao (eds.), Molecular genetic techniques for plant

338

G. VOLK AND C. WALTERS

genetic resources. Rep. of an IPGRI workshop, 9–11 October 1995, Rome. Intl. Plant Genetic Res. Inst., Rome. Lamboy, W. F., and C. G. Alpha. 1998. Using simple sequence repeats (SSRs) for DNA fingerprinting germplasm accessions of grape (Vitis L.) species. J. Am. Soc. Hort. Sci. 123:182–188. Lamboy, W. F., J. Yu, P. L. Forsline, and N. F. Weeden. 1996. Partitioning of isozyme diversity in wild populations of Malus sieversii L. and implications for germplasm collection. J. Am. Soc. Hort. Sci. 121:982–987. Langis, R., and P. L. Steponkus. 1991. Cryopreservation of rye protoplasts by vitrification. Plant Physiol. 92:666–671. Lazanyi, J. 1994. Somaclonal variation in the progeny of CHI-31 maize inbred line derived from tissue culture. In: Y. P. S. Bajaj (ed.), Biotechnology in agriculture and forestry, Vol. 25. Maize. Springer-Verlag, Berlin. Leprince, O., F. J. M. Harren, J. Buitink, M. Alberda, and F. A. Hoekstra. 2000. Metabolic dysfunction and unabated respiration precede the loss of membrane integrity during dehydration of germinating radicles. Plant Physiol. 122:597–608. Leprince, O., G. A. F. Hendry, N. M. Atherton, and C. Walters-Vertucci. 1996. Free radicals and metabolism associated with the acquisition and loss of desiccation tolerance in developing seeds. Biochem. Soc. Trans. 24:451–455. Lerceteau, E., T. Robert, V. Petiard, and D. Crouzillat. 1997. Evaluation of the extent of genetic variability among Theobroma cacao accessions using RAPD and RFLP markers. Theor. Appl. Genet. 95:10–19. Levitt, J. 1980. Responses of plants to environmental stress. Vol. 1. Chilling, freezing and high temperature stresses. Academic Press, New York. Liang, Y. H., and W. Q. Sun. 2000. Desiccation tolerance of recalcitrant Theobroma cacao embryonic axes: the optimal drying rate and its physiological basis. J. Expt. Bot. 51:1911–1919. Luby, J., P. Forsline, H. Aldwinckle, V. Bus, and M. Geibel. 2001. Silk road apples—Collection, evaluation, and utilization of Malus sieversii from central Asia. HortScience 36:225–231. Luza, J. G., and V. S. Polito. 1988. Cryopreservation of English walnut (Juglans regia L.) pollen. Euphytica 37:141–148. Makowska, Z., J. Keller, and F. Engelmann. 1999. Cryopreservation of apices isolated from garlic (Allium sativum L.) bulbils and cloves. Cryo-Letters 20:175–182. Marita, J. M., J. M. Rodriguez, and J. Nienhuis. 2000. Development of an algorithm identifying maximally diverse core collections. Genetic Res. Crop Evol. 47:515–526. Martinez, D., R. Arroyo-Garcia, and M. A. Revilla. 1999. Cryopreservation of in vitro grown shoot-tips of Olea europaea L. var. Arbequina. Cryo-Letters 20:29–36. Martinez, D., M. A. Revilla, A. Espina, and J. R. Garcia. 2000. Differential scanning calorimetry applied to the storage at ultra low temperatures of olive and hop in vitro grown shoot-tips. Thermochimica Acta 349:147–151. Martinez, D., M. A. Revilla, A. Espina, E. Jaimez, and J. R. Garcia. 1998. Survival cryopreservation of hop shoot tips monitored by differential scanning calorimetry. Thermochimica Acta 317:91–94. Maxted, N., and J. G. Hawkes. 1997. Selection of target taxa. p. 43–68. In: N. Maxted, B. V. Ford-Lloyd, and J. G. Hawkes (eds.), Plant genetic conservation: The in situ approach. Chapman and Hall, New York. Mazur, P. 1969. Freezing injury in plants. Ann. Rev. Plant Physiol. 20:419–448. McKersie, B. D., T. Senaratna, M. A. Walker, E. J. Kendall, and R. Hetherington. 1988. Deterioration of membranes during aging in plants: Evidence for free radical mediation.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

339

p. 412–465. In: L. D. Nooden and A. C. Leopold (eds.), Senescence and aging in plants. Academic Press, New York, Merila, J., and P. Crnonkrak. 2001. Comparison of genetic differentiation at marker loci and quantitative traits. J. of Evolut. Biol. 14:892–903. Meryman, H. T., and R. J. Williams. 1985. Basic principles of freezing injury to plant cells: Natural tolerance and approaches to cryopreservation. In: K. K. Kartha (ed.), Cryopreservation of plant cells and organs. CRC Press, Boca Raton, FL. Michel, D., A. Furini, F. Salamini, and D. Bartels. 1994. Structure and regulation of an ABA-responsive and desiccation responsive gene from the resurrection plant Craterostigma plantagineum. Plant Mol. Biol. 24:549–560. Moriguchi, T. 1995. Cryopreservation and minimum growth storage of pear (Pyrus species). Biotechnol. Agri. Forestry 32:114–125. Namkoong, G. 1981. Methods of pollen sampling for gene conservation. p. 74–76. In: E. C. Franklin (ed.), Pollen management handbook, U.S. Dept. Agr. 587, Washington, D.C. Niino, T., and A. Sakai. 1992. Cryopreservation of alginate-coated in vitro-grown shoot tips of apple, pear, and mulberry. Plant Sci. 87:199–206. Niino, T., A. Sakai, H. Yakuwa, and K. Nojiri. 1992. Cryopreservation of in vitro-grown shoot tips of apple and pear by vitrification. Plant Cell, Tissue Organ Culture 28: 264–266. Niino, T., T. Kazuo, M. Suzuki, S. Ohuchi, J. Magoshi, and T. Akihama. 1997. Cryopreservation of in vitro grown shoot tips of cherry and sweet cherry by one-step vitrification. Scientia Hort. 70:155–163. Niwata, E. 1995. Cryopreservation of apical meristems of garlic (Allium sativum L.) and high subsequent plant regeneration. Cryo-Letters 16:102–107. Oliver, A. E., O. Leprince, W. F. Wolkers, D. K. Hincha, A. G. Heyer, and J. H. Crowe. 2001. Non-disaccharide-based mechanisms of protection during drying. Cryobiology 43: 151–167. Ortiz, R., S. Madsen, and D. Vuylsteke. 1998. Classification of African plantain landraces and banana cultivars using a phenotypic distance index of quantitative descriptors. Theor. Appl. Genet. 96:904–911. Pammenter, N. W., and P. Berjak. 1999. A review of recalcitrant seed physiology in relation to desiccation-tolerance mechanisms. Seed Sci. Res. 9:13–27. Pammenter, N. W., V. Greggains, J. I. Kioko, J. Wesley-Smith, P. Berjak, and W. E. FinchSavage. 1998. Effects of differential drying rates on viability retention of recalcitrant seeds of Ekebergia capensis. Seed Sci. Res. 8:463–471. Panis, B., and R. Swennen. 1985. Cryopreservation of germplasm of banana and plantain (Musa species). p. 381–397. In: Y. P. S. Bajaj (ed.), Biotechnol. Agr. Forestry 32. Springer, New York. Panis, B., N. Totte, K. Van Nimmen, L. A. Withers, and R. Swennen. 1996. Cryopreservation of banana (Musa spp.) meristem cultures after preculture on sucrose. Plant Sci. 121:95–106. Parfitt, D. E., and A. A. Almehdi. 1983. Cryogenic storage of grape pollen. Am. J. Enol. Vitic. 34:227–228. Parfitt, D. E., and A. A. Almehdi. 1984. Liquid nitrogen storage of pollen from five cultivated Prunus species. HortScience 19:69–70. Pavek, D., and E. Garvey. 1999. The American wild relatives of crops: In situ conservation guidelines. Report from the in situ subcommittee of the Plant Germplasm Operations Committee, USDA-ARS. Pence, V. C. 1995. Cryopreservation of recalcitrant seeds. p. 29–50. In: Y. P. S. Bajaj (ed.), Biotechnol. Agri. Forestry 32. Springer, New York.

340

G. VOLK AND C. WALTERS

Pennycooke, J. C., and L. E. Towill. 2001. Medium alterations improve regrowth of sweet potato (Ipomoea batatas L.) shoot tips cryopreserved by vitrification and encapsulationdehydration. Cryo-Letters 22:381–389. Plessis, P., C. Leddet, A. Collas, and J. Dereuddre. 1993. Cryopreservation of Vitis vinifera L. cv Chardonnay shoot tips by encapsulation-dehydration: Effects of pretreatment, cooling and postculture conditions. Cryo-Letters 14:309–320. Pollak, L. M., and J. D. Corbett. 1993. Using GIS datasets to classify maize-growing regions in Mexico and Central America. Agron. J. 85:1133–1139. Rajashekar, C., M. N. Westwood, and M. J. Burke. 1982. Deep supercooling and cold hardiness in Genus Pyrus. J. Am. Soc. Hort. Sci. 107:968–972. Rasmussen, D., and B. Luyet. 1969. Complementary study of some non-equilibrium phase transitions in frozen solutions of glycerol, ethylene glycol, glucose and sucrose. Biodynamica 10:319–331. Reed, B. M. 1988. Cold acclimation as a method to improve survival of cryopreserved Rubus meristems. Cryo-Letters 9:166–171. Reed, B. M. 1992. Cold storage of strawberries in vitro: A comparison of three storage systems. Fruit Var. J. 46:98–102. Reed, B. M. 1993. Responses to ABA and cold acclimation are genotype dependent for cryopreserved blackberry and raspberry meristems. Cryobiology 30:179–184. Reed, B. M. 1999. The in vitro genebank of temperate fruit and nut crops at the National Clonal Germplasm Repository-Corvallis. p. 132–135. In: F. Engelmann (ed.), Management of field and in vitro germplasm collections. Proc. Consultation Meeting, 15–20 January 1996, CIAT, Cali, Colombia. Intl. Plant Genetic Res. Unit, Rome. Reed, B. M. 2001a. Implementing cryogenic storage of clonally propagated plants. CryoLetters 22:97–104. Reed, B. M. 2001b. Validation of cryopreservation protocols for plant germplasm conservation: a pilot study using Ribes L. Biodiver. Conserv. 10:939–949. Reed, B. M., P. M. Buckley, and T. N. DeWilde. 1995. Detection and eradication of endophytic bacteria from micropropagated mint plants. In Vitro Cell. Dev. Biol. 31:53–57. Roos, E. E., L. E. Towill, C. T. Walters, S. A. Blackman, and P. C. Stanwood. 1996. Preservation techniques for extending the longevity of plant tissues. p. 158–204. In: T. F. Stuessy and S. H. Sohmer (eds.), Sampling the green world. Columbia Univ. Press, New York. Sakai, A. 1966. Survival of plant tissues at super-low temperatures. IV. Cell survival with rapid cooling and rewarming. Plant Physiol. 41:1050–1054. Sakai, A. 1984. Cryopreservation of apical meristems. Hort. Rev. 6:357–372. Sakai, A. 1985. Cryopreservation of shoot-tips of fruit trees and herbaceous plants. In: K. Kartha (ed.), Cryopreservation of plant cells and organs. CRC Press, Boca Raton, FL. Sakai, A. 1993. Cryogenic strategies for survival of plant cultured cells and meristems cooled to –196°C. In: Cryopreservation of plant genetic resources. Japan Intl. Coop. Agency, Tokyo. Sakai, A. 2000. Development of cryopreservation techniques. p. 1–7. In: F. Engelmann and H. Takagi (eds.), Cryopreservation of tropical plant germplasm. Current research progress and application. Japan Intl. Res. Center Agri. Sci., Tsukuba, Japan/Intl. Plant Genetic Res. Inst., Rome. Sakai, A., S. Kobayashi, and I. Oiyama. 1990. Cryopreservation of nucellar cells of naval orange (Citrus sinensis Osb. Var. brasiliensis Tanaka) by vitrification. Plant Cell Rep. 9:30–33. Sakai, A., S. Kobayashi, and I. Oiyama. 1991. Survival by vitrification of nucellar cells of naval orange (Citrus sinensis var. brasiliensis Tanaka) cooled to –196°C. J. Plant Physiol. 137:465–470.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

341

Sakai, A., and S. Yoshida. 1967. Survival of plant tissue at super-low temperature. VI. Effects of cooling and rewarming rates on survival. Plant Physiol. 42:1695–1701. Sakai, A., and S. Yoshida. 1968. The role of sugar and related compounds in variations of freezing resistance. Cryobiology 5:160–174. Schaal, B. A., and K. M. Olsen. 2000. Gene genealogies and population variance in plants. Proc. Natl. Acad. Sci. (USA) 97:7024–7029. Schoen, D. J., and A. H. D. Brown. 1993. Conservation of genetic richness in wild crop relatives is aided by assessment of genetic markers. Proc. Natl. Acad. Sci. (USA) 90:10623– 10627. Schoen, D. J., and A. H. D. Brown. 1995. Maximising genetic diversity in core collections of wild relatives of crop species. p. 55–76. In: T. Hodgkin, A. H. D. Brown, Th. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Intl. Plant Genetic Res. Inst., Wiley, New York. Schoen, D. J., and A. H. D. Brown. 2001. The conservation of wild plant species in seed banks. BioScience 51:960–966. Schoen, D. J., J. L. David, and T. M. Bataillon. 1998. Deleterious mutation accumulation and the regeneration of genetic resources. Proc. Natl. Acad. Sci. (USA). 95:394–399. Schoenike, R. E., and C. F. Bey. 1981. Conserving genes through pollen storage. p. 72–73. In: E. C. Franklin (ed.), Pollen management handbook. U.S. Dept. of Agriculture 587, Washington, D.C. Scott, P. 2000. Resurrection plants and the secrets of eternal leaf. Ann. Bot. 85:159–166. Scowcroft, W. R. 1984. Genetic variability in tissue culture: Impact on germplasm conservation and utilization. IPGRI Technical Report. IPGRI, Rome, Italy. Sedgley, M. 1981. Storage of avocado pollen. Euphytica 30:595–599. Shands, H. L., and L. E. Weisner (eds.). 1991. Use of plant introductions in cultivar development. Part 1. CSSA Special Publ. 17. Crop Sci. Soc. Am., Inc., Madison, WI. Shands, H. L., and L. E. Weisner (eds.). 1992. Use of plant introductions in cultivar development. Part 2. CSSA Special Publ. 20. Crop Sci. Soc. Am., Inc., Madison, WI. Shands, H. L., P. J. Fitzgerald, and S. A. Eberhart. 1989. Program for plant germplasm preservation in the United States: the U.S. National Plant Germplasm System. In: L. Knutson and A. K. Stoner (eds.), Biotic diversity and germplasm preservation. Global imperatives. Kluwer Academic Publ., Dordrecht, The Netherlands. Sharma, S. B., M. A. Ansari, K. S. Varaprasad, A. K. Singh, and L. J. Reddy. 1999. Resistance to Meloidogyne javanica in wild Arachis species. Genetic Res. Crop Evol. 46:557–568. Shatnawi, M. A., F. Engelmann, A. Frattarelli, and C. Damiano. 1999. Cryopreservation of apices of in vitro plantlets of almond (Prunus dulcis mill.). Cryo-Letters 20:13–20. Shinozaki, K., and K. Yamaguchi-Shinozaki. 1998. Molecular responses to drought stress. In: K. Satoh and N. Murata (eds.), Stress responses of photosynthetic organisms. Elsevier, New York. Skinner, D. Z., G. R. Bauchan, G. Auricht, and S. Hughes. 1999. A method for the efficient management and utilization of large germplasm collections. Crop Sci. 39:1237–1242. Sparks, D., and I. E. Yates. 2002. Pecan pollen stored over a decade retains viability. HortScience 37:176–177. Stanwood, P. C. 1985. Cryopreservation of seed germplasm for genetic conservation. In: K. K. Kartha (ed.), Cryopreservation of plant cells and organs. CRC Press, Boca Raton, FL. Stanwood, P. C., and S. Sowa. 1995. Evaluation of onion (Allium cepa L.) seed after 10 years of storage at 5, –18, and –196°C. Crop Sci. 35:852–856. Stanwood, P. C., E. E. Roos, and L. E. Towill. 1986. Cryopreservation of plant germplasm: A pilot research project. FAO/IBPGR Plant Genetic Res. Newslett. 65:18–19.

342

G. VOLK AND C. WALTERS

Steponkus, P. L. 1985a. Cryobiology of isolated protoplasts: Applications to plant cell cryopreservation. In: K. K. Kartha (ed.), Cryopreservation of plant cells and organs. CRC Press, Boca Raton, FL. Steponkus, P. L. 1985b. Fundamental aspects of cryoinjury as related to cryopreservation of plant cells and organs. Biotechnology in plant science. Academic Press, San Diego, CA. Steponkus, P. 1990. Cold acclimation and freezing injury from a perspective of the plasma membrane. In: F. Katterman (ed.), Environmental injury to plants. Academic Press, San Diego, CA. Steponkus, P., R. Langis, and S. Fujikawa. 1992. Cryopreservation of plant tissues by vitrification. Adv. Low-Temp. Biol. 1:1–61. Stushnoff, C., and M. Seufferheld. 1995. Cryopreservation of apple (Malus species) genetic resources. In: Y. P. S. Bajaj (ed.), Cryopreservation of plant germplasm I. SpringerVerlag, Berlin. Suzuki, M., T. Niino, T. Akihama, and S. Oka. 1997. Shoot formation and plant regeneration of vegetative pear buds cryopreserved at –150 degrees C. Soc. Hort. Sci. 66:29–34. Takagi, H. 2000. Recent developments in cryopreservation of shoot apices of tropical species. In: F. Engelmann and H. Takagi (eds.), Cryopreservation of tropical plant germplasm. Current research progress and application. Japan Intl. Research Center for Agri. Sci., Tsukuba, Japan/Intl. Plant Genetic Res. Inst., Rome. Tanksley, S. D., and S. R. McCouch. 1997. Seed banks and molecular maps: unlocking the genetic potential from the wild. Science 277:1063–1066. Tanprasert, P., and B. M. Reed. 1998. Detection and identification of bacterial contaminants of strawberry runner explants. Plant Cell, Tissue Organ Culture 52:53–55. Thinh, N., H. Takagi, and S. Yashima. 1999. Cryopreservation of in vitro-grown shoot tips of banana (Musa spp) by vitrification method. Cryo-Letters 20:163–174. Tohme, J., P. Jones, S. Beebe, and M. Iwanaga. 1995. The combined use of agroecological and characterisation data to establish the CIAT Phaseolus vulgaris core collection. p. 95–107. In: T. Hodgkin, A. H. D. Brown, Th. J. L. van Hintum, and E. A. V. Morales (eds.), Core collections of plant genetic resources. Intl. Plant Genetic Res. Inst., Wiley, New York. Tompsett, P. B., and H. W. Pritchard. 1998. The effect of chilling and moisture status on the germination, desiccation tolerance and longevity of Aesculus hippocastanum L. seed. Ann. Bot. 82:249–261. Towill, L. E. 1984. Survival at ultra-low temperatures of shoot tips from Solanum tuberosum groups Andigena, Phureja, Stenotomum, Tuberosum, and other tuber-bearing Solanum species. Cryo-Letters 5:319–326. Towill, L. E. 1985. Low temperature and freeze-/vacuum-drying preservation of pollen. p. 171–198. In: K. K. Kartha (ed.), Cryopreservation of plant cells and organs. CRC Press, Boca Raton, FL. Towill, L. E. 1988a. Genetic considerations for germplasm preservation of clonal materials. HortScience 23:91–95. Towill, L. E. 1988b. Survival of shoot tips from mint species after short-term exposure to cryogenic conditions. HortScience 23:839–841. Towill, L. E. 1990. Cryopreservation. p. 41–69. In: J. H. Dodds (ed.), In vitro methods for conservation of plant genetic resources. Chapman and Hall, London. Towill, L. E., and Y. P. S. Bajaj (eds.). 2002. Cryopreservation of plant germplasm II. Springer-Verlag, New York. Towill, L. E., and P. L. Forsline. 1999. Cryopreservation of sour cherry (Prunus cerasus L.) using a dormant vegetative bud method. Cryo-Letters 20:215–222.

7. PRESERVATION OF GENETIC RESOURCES IN THE NATIONAL PLANT

343

Towill, L. E., and C. Walters. 2000. Cryopreservation of pollen. p. 115–129. In: F. Engelmann and H. Takagi (eds.), Cryopreservation of tropical plant germplasm. Current research progress and application. Japan Intl. Research Center for Agri. Sci., Tsukuba, Japan/Intl. Plant Genetic Res. Inst., Rome. Uemura, M., and P. L. Steponkus. 1999. Cold acclimation in plants: Relationship between the lipid composition and the cryostability of the plasma membrane. J. Plant Res. 112:245–254. Uemura, M., R. A. Joseph, and P. L. Steponkus. 1995. Cold acclimation of Arabidopsis thaliana. Plant Physiol. 109:15–30. Upadhyaya, H. D., P. J. Bramel, and S. Singh. 2001. Development of a chickpea core subset using geographic distribution and quantitative traits. Crop Sci. 41:206–210. USDA, ARS, National Genetic Resources Program. 2002. Germplasm Resources Information Network—(GRIN). [Online Database] National Germplasm Resources Laboratory, Beltsville, Maryland. Available: http://www.ars-grin.gov/npgs/aboutcgc.html. Van Hintum, Th. J. L. 1999. The core selector, a system to generate representative selections of germplasm collections. Plant Genetic Res. Newslett. 118:64–67. Vertucci, C. W. 1989a. Effects of cooling rate on seeds exposed to liquid nitrogen temperatures. Plant Physiol. 90:1478–1485. Vertucci, C. W. 1989b. Relationship between thermal transitions and freezing injury in pea and soybean seeds. Plant Physiol. 90:1121–1128. Vertucci, C. W., and J. M. Farrant. 1995. Acquisition and loss of desiccation tolerance. In: In: J. Kigel and G. Galili (eds.), Seed development and germination. Marcel Dekker, New York. Vertucci, C. W., P. Berjak, N. W. Pammenter, and J. Crane. 1991. Cryopreservation of embryonic axes of an homeohydrous (recalcitrant) seed in relation to calorimetric properties of tissue water. Cryo-Letters 12:339–350. Visser, T. 1955. Germination and storage of pollen. Medederlingen van de Landbouwhogeschool te Wageningen/Nederland 55:1–68. Volk, G., K. Cross, and W. Lyons. 2003. Low temperature storage of garlic for spring planting. HortScience (in press). Walters, C. 2002. Guidelines for seed storage. In: E. Guerrant, K. Havens, and M. Maunder (eds.), Strategies for survival. Island Press, Covelo, CA. Walters, C., J. M. Farrant, N. W. Pammenter, and P. Berjak. 2002. Desiccation stress and damage. In: M. Black and H. W. Pritchard (eds.), Desiccation and plant survival. CABI Publ., Wallingford, UK. Walters, C., N. W. Pammenter, P. Berjak, and J. Crane. 2001. Desiccation damage, accelerated aging and respiration in desiccation sensitive and tolerant seeds. Seed Sci. Res. 11:135–148. Walters, C., E. E. Roos, D. H. Touchell, P. C. Stanwood, L. Towill, L. Wiesner, and S. A. Eberhart. 1998. Refrigeration can save seeds economically. Nature 395:758. Walters, C., D. H. Touchell, P. Power, J. Wesley-Smith, and M. F. Antolin. 2002. A cryopreservation protocol for embryos of the endangered species Zizania texana. CryoLetters 23:291–298. Webb, M. S., and P. L. Steponkus. 1993. Freeze-induced membrane ultrastructural alterations in rye (Secale cereale) leaves. Plant Physiol. 101:955–963. Wesley-Smith, J. 2001. Investigations into the responses of axes of recalcitrant seeds to dehydration and cryopreservation. Univ. Natal, Durban, South Africa. Wesley-Smith, J., C. W. Vertucci, P. Berjak, N. W. Pammenter, and J. Crane. 1992. Cryopreservation of desiccation-sensitive axes of Camellia sinensis in relation to dehydration, freezing rate and the thermal properties of tissue water. J. Plant Physiol. 140:596–604.

344

G. VOLK AND C. WALTERS

Wesley-Smith, J., C. Walters, N. W. Pammenter, and P. Berjak. 2001a. Interactions among water content, rapid (nonequilibrium) cooling to –196°C, and survival of embryonic axes of Aesculus hippocastanum L. seeds. Cryobiology 42:196–206. Wesley-Smith, J., N. W. Pammenter, P. Berjak, and C. Walters. 2001b. The effects of two drying rates on the desiccation tolerance of embryonic axes of recalcitrant jackfruit (Artocarpus heterophyllus Lamk.) seeds. Ann. Bot. 88:653–664. Westwood, M. N. 1989. Maintenance and storage: Clonal germplasm. Plant Breed. Rev. 7:111–128. White, G. A., H. L. Shands, and G. R. Lovell. 1989. History and operation of the National Plant Germplasm System. Plant Breed. Rev. 7:5–56. Williams, R. J. 1994. Methods for determination of glass transitions in seeds. Ann. Bot. 74:525–530. Withers, L. A. 1980. Tissue culture storage for genetic conservation, IBPGR Tech. Rep., IPGRI, Rome. Wolff, S. P., A. Garner, and R. T. Dean. 1986. Free radicals, lipids and protein degradation. Trends Biochem. Sci. 11:27–31. Wu, Y., F. Engelmann, Y. Zhao, M. Zhou, and S. Chen. 1999. Cryopreservation of apple shoot tips: Importance of cryopreservation technique and of conditioning of donor plants. Cryo-Letters 20:121–130. Wu, Y., Y. Zhao, F. Engelmann, and M. Zhou. 2001a. Cryopreservation of kiwi shoot tips. Cryo-Letters 22:277–284. Wu, Y., Y. Zhao, F. Engelmann, M. Zhou, D. Zhang, and S. Chen. 2001b. Cryopreservation of apple dormant buds and shoot tips. Cryo-Letters 22:375–380. Yu, Z-W., and P. J. Quinn. 1998. The modulation of membrane structure and stability by dimethylsulfoxide. Molecular Membrane Biol. 15:59–68. Zhao, Y., Y. Wu, F. Engelmann, and M. Zhou. 2001. Cryopreservation of axillary buds of grape (Vitis vinifera) in vitro plantlets. Cryo-Letters 22:321–328. Zhao, Y., Y. Wu, F. Engelmann, M. Zhou, D. Zhang, and S. Chen. 1999. Cryopreservation of apple shoot tips by encapsulation-dehydration: Effect of preculture, dehydration and freezing procedure on shoot regeneration. Cryo-Letters 20:103–109.

Subject Index Volume 23 A Aglaonema breeding, 267–269 Alocasia breeding, 269 Amaranth cytoplasm, 191 Anthurium breeding, 269–271 B Bean: breeding, 21–72 rhizobia interaction, 21–72 Biography, Coyne, D.E. 1–19 Breeding: Aglaonema, 267–269 Alocasia, 269 Anthurium, 269–271 bean, 21–72 bromeliad, 275–276 Calathea, 276 cytoplasmic DNA, 175–210 Dieffenbachia, 271–272 Dracaena, 277 Epepremnum, 272–273 fern, 276 Ficus, 278-279 foliage plant, 245–290 Hedera breeding 279–280 palms, 280–281 Philodendron, 273 rice, 73–174 Syngonium, 274 Bromeliad breeding, 275–276

Cytoplasm breeding, 175– 210 D Dieffenbachia breeding, 271–272 Dracaena breeding, 277 E Epepremnum breeding, 272–273 F Fern breeding, 276 Ficus breeding, 278-279 Floral biology, garlic, 211–244 Foliage plant breeding, 245–290 G Garlic breeding, 211–244 Genes, Rhizobium, 39–47 Genetics: cytoplasm, 175–210 rhizobia, 21–72 Germplasm preservation, 291–344 Grain breeding, rice, 73–174 H Hedera breeding, 279–280 Hybrid and hybridization, rice, 73–174

C

M

Calathea breeding, 276 Coyne, Dermot E. (biography), 1–19

Maize cytoplasm, 189 Molecular markers, rice, 73–174

Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 345

346

SUBJECT INDEX

N

S

National Clonal Germplasm Repository (NCGR), preservation, 291–344

Seed, garlic, 211–244 Somaclonal variation, foliage breeding, 245–290 Syngonium breeding, 274

O Ornamental breeding, foliage plants, 245–290

T Taxonomy, garlic, 215–218

P Palm breeding, 280–281 Philodendrum breeding, 273 Potato cytoplasm, 187–189

V Vegetable breeding: bean, 21–72 garlic, 211–244

R Reproduction: foliage plant, 255–259 garlic, 211–244 Rhizobia, 21–72 Rice: cytoplasm, 189 hybrid breeding, 73–174 molecular markers, 73–174

W Wheat cytoplasm, 189–190

Cumulative Subject Index (Volumes 1–23) 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 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 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 pigmentation, maize aleurone, 8:91–137 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: genetics, 9:333–366 rootstocks, 1:294–394 transformation, 16:101–102 Apricot transformation, 16:102 Arachis, see Peanut in vitro culture, 2:218–224 Artichoke breeding, 12:253–269 Avena sativa, see Oat Azalea, mutation breeding, 6:75–76

B Bacillus thuringensis, 12:19–45 Bacterial diseases: apple rootstocks, 1:362–365 cell selection, 4:163–164 cowpea, 15:238–239 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 transformation, 16:105–106 Barley: anther culture, 15:141–186 breeding methods, 5:95–138

Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 347

348 Barley (cont.) 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 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 Bringhurst, Royce S., 9:1–8 Burton, Glenn W., 3:1–19 Coyne, Dermot E., 23:1–19 Downey, Richard K., 18:1–12 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 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 Birdsfoot trefoil, tissue culture, 2:228–229 Blackberry, 8:249–312 mutation breeding, 6:79 Black walnut, 1:236–266 Blueberry: breeding, 13:1–10 rabbiteye, 5:307–357

CUMULATIVE SUBJECT INDEX Brachiaria, apomixis, 18:36–39, 49–51 Bramble transformation, 16:105 Brassica, see Cole crops Brassicaceae: incompatibility, 15:23–27 molecular mapping, 14:19–23 Brassica: napus, see Canola, Rutabaga rapa, see Canola 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 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 cytoplasmic DNA, 12:175–210 diallel analysis, 9:9–36 Dieffenbachia, 271–272 doubled haploids, 15:141–186 Dracaena, 23:277 durum wheat, 5:11–40 Epepremnum, 23: 272–2mn epistasis, 21:27–92 exotic maize, 14:165–187

CUMULATIVE SUBJECT INDEX fern, 23:276 fescue, 3:313–342 Ficus, 23:276 foliage plant, 23:245–290 forest tree, 8:139–188 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 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 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 palms, 23:280–281 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

349 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 rose, 17:159–189 rutabaga, 8:217–248 sesame, 16:179–228 snap pea, 21:93–138 somatic hybridization, 20:167–225 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 sugar cane, 16:272–273 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 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 C Cactus: breeding, 135–166 domestication, 135–166 Cajanus, in vitro culture, 2:224 Calathea breeding, 23:276 Canola, R.K. Downey, designer, 18:1–12 Carbohydrates, 1:144–148

350 Carbon isotope discrimination, 12:81–113 Carnation, mutation breeding, 6:73–74 Carrot breeding, 19: 157–190 Cassava, 2:73–134 Castanea, see Chestnut Cell selection, 4:139–145, 153–173 Cereal breeding, see Grain breeding Cereal diversity, 21:221–261 Cherry, see Sweet cherry 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, 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 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 Cryopreservation, 7:125–126,148–151,167 buds, 7:168–169 genetic stability, 7:125–126

CUMULATIVE SUBJECT INDEX 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 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 rose, 17:169–171 rye, 13:265–292 Saccharum complex, 16:273–275 sesame, 16:185–189 triticale, 5:41–93; 8:54 wheat, 5:12–14; 10:5–15; 11:225–234 Cytoplasm: breeding, 23: 175–210 cybrids, 3:205–210; 20:206–209 molecular biology of male sterility, 10:23–51 organelles, 2:283–302; 6:361–393 pearl millet, 1:166 petunia, 1:43–45 wheat, 2:308–319 D Dahlia, mutation breeding, 6:75 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 blackberry, 8:291–295

CUMULATIVE SUBJECT INDEX 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 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 in land races, 21:221–261 DNA methylation, 18:87–176 Doubled haploid breeding, 15:141–186 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 soybean breeding, 4:203–243 wheat adaptation, 12:135–146 Durum wheat, 5:11–40 Duvick, Donald N. (biography), 14:1–11 E Elaeis, see Oil palm Embryo culture: in crop improvement, 5:181–236 oil palm, 4:186–187 pasture legume hybrids, 5:249–275 Endosperm: maize, 1:139–161 sweet corn, 1:139–161 Endothia parasitica, 4:355–357 Epepremnum breeding, 23:272–273 Epistasis, 21:27–92. Evolution: coffee, 2:157–193 grapefruit, 13:345–363

351 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 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 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 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 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

352 Fruit, nut, and beverage crop breeding (cont.) cactus, 20:135–166 cherry, 9:367–388 citrus, 8:339–374 coffee, 2:157–193 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 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 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 host-parasite genetics, 5:393–433 lettuce, 1:286–287 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 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

CUMULATIVE SUBJECT INDEX 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 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 host-parasite genetics, 5:415–428 maize mobile elements, 4:81–122 salt resistance, 22:389–425 transformation by particle bombardment, 13:231–260 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

CUMULATIVE SUBJECT INDEX 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 cytoplasm, 23:175–210 DNA methylation, 18:87–176 durum wheat, 5:11–40 forest trees, 8:139–188 fruit crop transformation, 16:87–134 gene action, 15:315–374 herbicide resistance, 11:155–198 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 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 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 potato disease resistance, 19:69–165

353 potato ploidy manipulation, 3:274–277; 16:15–86 quality protein in maize, 9:183–184 quantitative trait loci, 15:85–139 reproductive barriers, 11:11–154 rhizobia, 21–72 rice, hybrid, 17:15–156, 23:73–174 rose, 17:171–172 rutabaga, 8:217–248 salt resistance, 22:389–425 snap pea, 21:110–120 sesame, 16:189–195 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 wheat rust, 13:293–343 white clover, 17:191–223 yield, 3:21–167 Genome: Glycine, 16:289–317 Poaceae, 16:276–281 Genotype x 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

354 Germplasm (cont.) guayule, 6:112–125 isozyme, 6:18–21 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 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 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 oat, 6:167–207 pearl millet, 1:162–182 rice, 17:15–156 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 wild rice, 14:237–265 Grapefruit: breeding, 13:345–363 evolution, 13:345–363 Grape, transformation, 16:103–104 Grass breeding: breeding, 11:251–274 mutation breeding, 6:82 recurrent selection, 9:101–113 transformation, 13:231–260

CUMULATIVE SUBJECT INDEX 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 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 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 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

CUMULATIVE SUBJECT INDEX 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 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 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

355 rutabaga, 8:240–241 sweet potato, 4:336–337 transformation fruit crops, 16:113 wheat, 22:221–297 white clover, 17:209–210 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 transformation, 16:104 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 pasture legumes, 5:237–305 Vigna, 8:19–42 Legume tissue culture, 2:215–264 Lethal equivalents and genetic load, 10:93–127 Lettuce: genes, 1:267–293 breeding, 16:1–14; 20:105–133 Linkage: bean, 1:76–77 isozymes, 6:37–38 lettuce, 1:288–290 maps, molecular markers, 9:37–61 petunia, 1:31–34

356 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 carbohydrates, 1:144–148 cytoplasm, 23:189 doubled haploid breeding, 15:141–186 exotic germplasm utilization, 14:165–187 high oil, 22:3–4 honeycomb breeding, 18:226–227 hybrid breeding, 17:249–251 insect resistance, 6:209–243 male sterility, 10:23–51 mobile elements, 4:81–122 mutations, 5:139–180 origins, 20:15–66 overdominance, 17:225–257 protein, 1:103–138 quality protein, 9:181–216 recurrent selection, 9:115–179; 14:139–163 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 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 soybean, 21:277–291 wheat, 2:303–319 Malus spp, see Apple Malus ×domestica, see Apple Malvaceae, molecular mapping, 14:25–27 Mango transformation, 16:107 Manihot esculenta, see Cassava Medicago, see also Alfalfa

CUMULATIVE SUBJECT INDEX in vitro culture, 2:229–234 Meiosis, petunia, 1:14–16 Metabolism and heterosis, 10:53–90 Microprojectile bombardment, transformation, 13:231–260 Mitochondria genetics, 6:377–380 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 molecular mapping, 14:13–37; 19:31–68 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 quantitative trait loci, 15:85–139 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 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

CUMULATIVE SUBJECT INDEX 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 gametoclonal variation, 5:359–391 homeotic floral, 9:63–99 induced, 2:13–72 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

357 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 Nicotiana, see Tobacco Nodulation, soybean, 11:275–318 O Oat, breeding, 6:167–207 Oil palm: breeding, 4:175–201 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 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

358

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 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 transformation, 16:102 Peanut: breeding, 22:297–356 in vitro culture, 2:218–224 Pear transformation, 16:102 Pearl millet: apomixis, 18:55–56 breeding, 1:162–182 Pecan transformation, 16:103 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 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 Pistil, reproductive function, 4:9–79 Pisum, see Pea Plant introduction, 3:361–434; 7:9–11, 21–25 Plant exploration, 7:9–11, 26–28, 67–94

CUMULATIVE SUBJECT INDEX Plantain breeding, 2:135–155; 14:267–320; 21:1–25 Plastid genetics, 6:364–376, see also Organelle Plum 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 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 reproductive barriers, 11:98–105 sweet potato, 4:371 triticale, 5:11–40 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

CUMULATIVE SUBJECT INDEX 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 genotype x environment interaction, 16:135–178 heritability, 22:9–111 overdominance, 17:225–257 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 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 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 molecular markers, 73–174 photoperiodic response, 3:74, 89–92

359 Rosa, see Rose 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 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 cell, 4:139–145, 153–173 honeycomb design, 13:87–139; 18:177–249 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 recurrent restricted phenotypic, 9:101–113 recurrent selection in maize, 9:115–179; 14:139–163 Sesame breeding, 16:179–228 Sesamum indicum, see Sesame

360 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: 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 Spelt, agronomy, genetics, breeding, 15:187–213 Sprague, George F. (biography), 2:1–11 Sterility, see also Male sterility, 11:30–41

CUMULATIVE SUBJECT INDEX Starch, maize, 1:114–118 Statistics: advanced methods, 22:113–163 history, 17:259–316 Strawberry: biotechnology, 21: 139–180 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: mutation breeding, 6:82–84 and Saccharum complex, 16:269–288 Sweet cherry: pollen-incompatibility and selffertility, 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

CUMULATIVE SUBJECT INDEX Saccharum complex, 16:270–272 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: alfalfa, 10:190–192 cereals, 13:231–260 fruit crops, 16:87–134 mushroom, 8:206 rice, 17:179–180 somaclonal variation, 16:229–268 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 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

361 Unreduced and polyploid gametes, 3:253–288; 16:15–86 Urd bean, 8:32–35 V Vaccinium, see Blueberry Variance estimation, 22:113–163 Vegetable breeding: artichoke, 12:253–269 bean, 1:59–102; 4:245–272 bean (tropics), 10:199–269 beet (table), 22:257–388 carrot 19: 157–190 cassava, 2:73–134 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 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 Vogel, Orville A. (biography), 5:1–10 Vuylsteke, Dirk R. (biography), 21:1–25

362 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

CUMULATIVE SUBJECT INDEX 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 1–23)

Abdalla, O.S., 8:43 Acquaah, G., 9:63 Aldwinckle, H.S., 1:294 Anderson, N.O., 10:93; 11:11 Aronson, A.I., 12:19 Ascher, P.D., 10:93 Ashri, A., 16:179 Baggett, J.R. 21:93 Baltensperger, D.D., 19:227 Basnizki, J., 12:253 Beck, D.L., 17:191 Beebe, S., .23:21–72 Beineke, W.F., 1:236 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 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., 221 Brown, J.W.S., 1:59 Brown, S.K., 9:333,367

Burnham, C.R., 4:347 Burton, G.W., 1:162; 9:101 Burton, J.W., 21:263 Byrne, D., 2:73

Dodds, P.N., 15:19 Donini, P., 21:181 Draper, A.D., 2:195 Dumas, C., 4:9 Duncan, D.R., 4:153

Campbell, K.G., 15:187 Cantrell, R.G., 5:11 Carvalho, A., 2:157 Casas, A.M., 13:235 Cervantes-Martinez, C.T., 22:9 Chen, J., 23: 245 Chew, P.S., 22:165 Choo, T.M., 3:219 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 Condon, A.G., 12:81 Cooper, R.L., 3:289 Cornu, A., 1:11 Costa, W.M., 2:157 Cregan, P., 12:195 Crouch, J.H., 14:267 Crow, J.F., 17:225 Cummins, J.N., 1:294

Echt, C.S., 10:169 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

Dana, S., 8:19 De Jong, H., 9:217 Deroles, S.C., 14:321 Dhillon, B.S., 14:139 Dickmann, D.I., 12:163 Ding, H., 22:221

Gai, J. 21:263 Galiba, G., 12:115 Galletta, G.J., 2:195 Gmitter, F.G., Jr., 8:339; 13:345 Gold, M.A., 12:163

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, R.L.S., 17:191 Frei, U., 23:175 French, D.W., 4:347

Plant Breeding Reviews, Volume 23, Edited by Jules Janick ISBN 0-471-35421-X © 2003 John Wiley & Sons, Inc. 363

364 Goldman, I.L. 19:15; 20:67; 22:357 Gradziel, T.M., 15:43 Gressel, J., 11:155; 18:251 Gresshof, P.M., 11:275 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 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 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 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 Hunt, L.A., 16:135 Hutchinson, J.R., 5:181 Hymowitz, T., 8:1; 16:289 Janick, J., 1:xi; 23:1 Jansky, S., 19:77 Jayaram, Ch., 8:91 Jenderek, M.M., 23:211 Johnson, A.A.T., 16:229; 20:167 Jones, A., 4:313 Jones, J.S., 13:209 Ju, G.C., 10:53

CUMULATIVE CONTRIBUTOR INDEX 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 Knox, R.B., 4:9 Koebner, R.M.D., 21:181 Kollipara, K.P., 16:289 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 Lamborn, C., 21:93 Lamkey, K.R., 15:1 Lavi, U., 12:195 Layne, R.E.C., 10:271 Lebowitz, R.J., 3:343 Lehmann, J.W., 19:227 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 Lukaszewski, A.J., 5:41 Lyrene, P.M., 5:307 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 Maas, J. L., 21: 139 Maheswaran, G., 5:181 Maizonnier, D., 1:11 Marcotrigiano, M., 15:43 Martin, F.W., 4:313 Matsumoto, T.K. 22:389 Medina-Filho, H.P., 2:157

Miller, R., 14:321 Mondragon Jacobo, C. 20:135 Morrison, R.A., 5:359 Mowder, J.D., 7:57 Mroginski, L.A., 2:215 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 Palmer, R.G., 15:275, 21:263 Pandy, S., 14:139 Pardo, J. M., 22:389 Parliman, B.J., 3:361 Paterson, A.H., 14:13 Patterson, F.L., 22:221 Peairs, F.B., 22:221 Pedersen, J.F., 11:251 Perdue, R.E., Jr., 7:67 Peiretti, E.G., 23:175 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 Proudfoot, K.G., 8:217 Rackow, G., 18:1 Raina, S.K., 15:141 Ramage, R.T., 5:95 Ramming, D.W., 11:1 Ratcliffe, R.H., 22:221 Ray, D.T., 6:93 Redei, G.P., 10:1 Reimann-Phillipp, R., 13:265

CUMULATIVE CONTRIBUTOR INDEX 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 Ron Parra, J., 14:165 Roos, E.E., 7:129 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 Schroeck, G., 20:67 Scott, D.H., 2:195 Seabrook, J.E.A., 9:217 Sears, E.R., 11:225 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 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, S.E., 6:361

Snoeck, C., 23:21 Socias i Company, R., 8:313 Sobral, B.W.S., 16:269 Stalker, H.T., 22:297 Soh, A.C., 22:165 Sondahl, M.R., 2:157 Spoor, W., 20: 1 Steadman, J.R., 23:1 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 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 Thompson, A.E., 6:93 Towill, L.E., 7:159, 13:179 Tracy, W.F., 14:189 Tsaftaris, A.S., 18:87 Tsai, C.Y., 1:103

365 Wallace, D.H., 3:21; 13:141 Wan, Y., 11:199 Waters, C., 23:291 Weeden, N.F., 6:11 Wehner, T.C., 6:323 Wenzel, G. 23:175 Westwood, M.N., 7:111 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, G.C., 12:81 Wu, L., 8:189 Wu, R., 19:41 Xin, Y., 17:1,15 Xu, S., 22:113 Xu, Y., 15:85; 23:73

Ullrich, S.E., 2:13 Vanderleyden, J., .23:21 Van Harten, A.M., 6:55 Varughese, G., 8:43 Vasal, S.K., 9:181; 14:139 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

Yamada, M., 19:191 Yan, W., 13:141 Yang, W.-J., 10:53 Yonemori, K., 19:191 Yopp, J.H., 4:203 Yun, D.-J., 14:39 Zeng, Z.-B., 19:41 Zimmerman, M.J.O., 4:245 Zohary, D., 12:253