Plant Breeding Reviews (Volume 25)

PLANT BREEDING REVIEWS Volume 25

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 25 M. Gilbert I. L. Goldman C. H. Michler

PLANT BREEDING REVIEWS Volume 25

edited by

Jules Janick Purdue University

John Wiley & Sons, Inc.

This book is printed on acid-free paper. Copyright © 2005 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: 0471666939 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: Stanley J. Peloquin Potato Geneticist and Cytogeneticist

viii 1

Rodomiro Ortiz, Luigi Frusciante, and Domenico Carputo

2. Politics of Plant Breeding

21

Cary Fowler and Richard L. Lower I. II. III. IV.

Introduction Germplasm, Plant Breeding, and the Fight for Rights The Debate Over Biotechnology Plant Breeders’ Choices Literature Cited

3. Doubled Haploids in Genetics and Plant Breeding

22 25 37 42 52

57

Brian P. Forster and William T. B. Thomas I. II. III. IV. V.

Introduction Doubled Haploid Technology Doubled Haploid Populations in Genetics Doubled Haploids in Breeding Prospects Literature Cited

4. Biochemistry and Genetics of Flower Color

57 58 63 72 79 80

89

R. J. Griesbach I. II. III. IV.

Introduction Flavonoid Chemistry Anthocyanin Biosynthesis Mendelian Inheritance

89 91 94 100 v

vi

CONTENTS

V. Transgene Technology Literature Cited

5. The Influence of Mitochondrial Genetics on Crop Breeding Strategies

105 108

115

Sally A. Mackenzie I. II. III. IV.

Introduction Structure of the Mitochondrial Genome in Plants Cytoplasmic Male Sterility Occurrence and Developmental Implications of Nuclear-Cytoplasmic Incompatibility V. Some Implications of Cytoplasmic Genetics for the Plant Breeder Literature Cited

6. Genetic and Cytoplasmic-Nuclear Male Sterility in Sorghum

115 117 119 125 127 131

139

Belum V. S. Reddy, S. Ramesh, and Rodomiro Ortiz I. II. III. IV. V. VI. VII. VIII. IX.

Introduction Genetic Male Sterility (GMS) Cytoplasmic-Nuclear Male Sterility (CMS) Molecular Characterization of Cytoplasms DNA Polymorphism and Mapping Restorer Genes Factors Influencing CMS Systems Use Diversification of CMS Systems Heterosis and Hybrid Development Conclusion Literature Cited

7. Improving Drought Tolerance in Maize

140 140 146 150 151 152 162 166 167 167

173

T. Barker, H. Campos, M. Cooper, D. Dolan, G. Edmeades, J. Habben, J. Schussler, D. Wright, and C. Zinselmeier I. II. III. IV.

Introduction Physiology of the Response of Maize Under Drought Experimental Methods Applied Breeding Methods

175 178 194 217

CONTENTS

V. Molecular Breeding VI. Conclusions Literature Cited

8. The Origins of Fruits, Fruit Growing, and Fruit Breeding

vii

230 241 242

255

Jules Janick I. Introduction II. The Horticultural Arts III. Origin, Domestication, and Early Culture of Fruit Crops IV. Genetic Changes and Cultural Factors in Domestication Literature Cited

256 260 272 308

Subject Index

323

Cumulative Subject Index

325

Cumulative Contributor Index

342

List of Contributors Barker, T., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131, [email protected] Campos, H., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131 Carputo, Domenico, Departmento of Soil, Plant and Environmental Sciences, University of Naples “Federico II”, Via Università, 100, 80055 Portici, Italy Cooper, M., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131 Dolan, D., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131 Edmeades, G., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131 Forster, Brian P., Scottish Crop Research Institute, Invergowrie, Dundee, DD2 5DA, Scotland, UK, [email protected] Fowler, Cary, Center for International Environment & Development Studies (NORAGRIC), Agricultural University of Norway, P.O. Box 5003, 1432 Aas, Norway, [email protected] Frusciante, Luigi, Departmento of Soil, Plant and Environmental Sciences, University of Naples “Federico II”, Via Università, 100, 80055 Portici, Italy Griesbach, R. J., Floral and Nursery Plants Research Unit, U.S. National Arboretum, U.S. Department of Agriculture, Agricultural Research Service, Beltsville, MD 20705-2350, [email protected] Habben, J., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131 Janick, Jules, Purdue University, Department of Horticulture and Landscape Architecture, 625 Agriculture Mall Drive, West Lafayette, Indiana 479072010, [email protected] Lower, Richard L., Dept. of Horticulture, 1575 Linden Drive, University of Wisconsin, Madison, WI 53706 [email protected] Mackenzie, Sally A., Plant Science Initiative, University of Nebraska, Lincoln, Nebraska 68588-0660, [email protected] Ortiz, Rodomiro, International Institute of Tropical Agriculture (IITA), Eastern and Southern Africa Regional Center (ESARC), Plot 7, Bandali Rise, Bugolobi, Kampala, Uganda, and International Institute of Tropical Agriculture (IITA), Nigeria; IITA c/o Lambourn Ltd. (UK), 26 Dingwall Road, Croydon, CR9 3EE, UK, [email protected] Ramesh, S., Global Theme–Crop Improvement, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India.

viii

LIST OF CONTRIBUTORS

ix

Reddy, Belum V. S., Global Theme–Crop Improvement, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru 502 324, Andhra Pradesh, India Schussler, J., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131 Thomas, William T. B., Scottish Crop Research Institute, Invergowrie, Dundee, DD2 5DA, Scotland, UK Wright, D., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131 Zinselmeier, C., Pioneer Hi-Bred International, Inc., 7300 NW 62nd Ave., Johnston, IA 50131

Stanley J. Peloquin

1 Dedication: Stanley J. Peloquin Potato Geneticist and Cytogeneticist Rodomiro Ortiz* International Institute of Tropical Agriculture (IITA), Nigeria; IITA c/o Lambourn Ltd. (UK), 26 Dingwall Road, Croydon, CR9 3EE, UK Luigi Frusciante and Domenico Carputo Departmento of Soil, Plant and Environmental Sciences, University of Naples “Federico II”, Via Università, 100, 80055 Portici, Italy

Dr. Stanley J. Peloquin is known to plant breeders for his decisive contributions to genetic enhancement of potato (Solanum tuberosum L.) using haploids, 2n gametes, and wild Solanum species; for his pioneering work on potato cultivation through true seed; and as mentor of a new generation of plant breeders worldwide. The genetic enhancement of potato, the fourth most important food crop worldwide, benefited significantly from Peloquin’s work on ploidy manipulations led by the genetic knowledge he and his co-workers, mostly former graduate students, created and systematically transformed into applied breeding methods. His scientific papers, book chapters, classes, seminars, and talks on potato as a model plant for genetic breeding and evolutionary research are a source of inspiration to all researchers in crop improvement.

*The authors thank Dr. Phil Simon (Univ. of Wisconsin, Madison) for assistance with this manuscript. Plant Breeding Reviews, Volume 25. Edited by Jules Janick. ISBN 0471666939 © 2005 John Wiley & Sons, Inc. 1

2

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

BIOGRAPHICAL SKETCH Stan was born in Barron, Wisconsin, in 1921. He went to grade school in three small towns in Wisconsin and to a small high school at Ondossagon near Lake Superior in northern Wisconsin, where he was taught to get out of northern Wisconsin to make a living. There were 20 in his graduating class. After graduating with a degree in chemistry in 1942 from River Falls State College (now the University of Wisconsin—River Falls), he joined the U.S. Navy and served on a destroyer in the South Pacific for 3-1/2 years during the Second World War. Upon his release from the service, he enrolled at Marquette University, Milwaukee, Wisconsin, and obtained a MSc degree in biology in 1948. He then enrolled in the University of Wisconsin and studied genetics under the guidance of Dr. R. A. Brink and Dr. D. C. Cooper and was awarded a PhD degree in genetics in 1952. His thesis was entitled “Abnormal Embryo and Endosperm Development in Zea mays Following the Use of Pollen that has been Exposed to Mustard Gas.” From 1951 to 1956, he taught biology at Marquette University. From 1957, when he joined the faculty at the University of Wisconsin, Madison, until his retirement in 1994, he taught genetics to undergraduates in the Biocore Biology Program, and two graduate level courses: cytogenetics, and, with Dr. Ted Brigham, Chromosome Manipulations in Plants. He married Helga Sorensen and reared three sons: Philip, John, and James. Some time after her passing, he married Virgie Eastburn Fry, who is the mother of two children: David and Diane. Stan and Virgie are the proud grandparents of Brandon and Brienne (from Philip), and Christopher, Julia, and Melissa (from Diane), and two great grandchildren, Gavin and David. In 1984, Dr. Peloquin was elected to the U.S. National Academy of Sciences and was awarded a life membership by the Potato Association of America. In 1983, Dr. Peloquin was named the Campbell Bascom Professor by the University of Wisconsin and received Laurea Honoris Causa from the Università degli Studi di Napoli “Federico II” (Italy) in 2002. These recognitions he has received for his innovative research and scientific leadership are an acknowledgment of his superb skills as a potato geneticist and breeder. Peloquin always points out that potato should be seen not only as an important food (fresh or processed) but also as the raw material for the starch-processing industry, as feed because its vines can be fed to animals, and as a potential resource for medicine because of the compounds in its true seed, and as an ornamental.

1. DEDICATION

3

RESEARCH Peloquin’s achievements are the result of fundamental, ingenious scientific insight, using potato genetics as the focal point for his research. Stan integrates his efforts with a broad range of networking activities, creating a world-famous “school” composed largely of his graduate students and research fellows. This intense collaboration has had a great impact in generating basic knowledge and for achieving practical methods for genetic enhancement. Peloquin’s early efforts were in broadening the basic knowledge about cytogenetics and genetics of potato. Stan’s collaboration with Robert W. Hougas at the beginning of his career was an important first step, which led to more than 40 years of seminal potato research. Peloquin and Hougas, sometimes with early students and other collaborators, published 19 papers together that established a foundation of potato reproductive biology, genetics, and the genetics of reproductive biology that supported the broad and deep fundamental and applied developments that were to follow. Because of these early successes, Stan Peloquin set up germplasm enhancement methods relying on scaling up and scaling down chromosome sets via ploidy manipulations. Such ploidy manipulations are easily achieved in potato to transfer genes from wild Solanum species to the primary crop gene pool—particularly alleles for improving horticultural traits. His students in the classrooms of the University of Wisconsin still remember one of his favorite comments when teaching cytogenetics: “The best plant with which to manipulate individual chromosome numbers is wheat. The best plant with which to manipulate sets of chromosomes is potato.” As Stan pointed out in many of his reviews: “The ability to obtain plants with the gamete chromosome number (haploids) and gametes with the plant chromosome number (2n gametes) is the basis for the ease of manipulating whole sets of chromosomes in potato.” Haploid and 2n Gamete Cytology and Cytogenetics Scaling down the ploidy of the tetraploid potato (2n = 4x = 48 chromosomes) to the diploid level is achieved routinely by producing potato haploids (2n = 2x = 24 chromosomes). Maternal haploids can be easily obtained through parthenogenesis after interspecific hybridization of tetraploid cultivars with pollen of S. phureja Juz. et Buk. Haploid frequency is affected by both the maternal genotype and the pollen source,

4

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

and both seed parent and pollen source influence the success of haploid production. Peloquin and co-workers indicated that the endosperm associated with a haploid embryo was always hexaploid, which clearly demonstrated the union of the two chromosome sets from S. phureja with the polar nuclei, and lack of fertilization of the egg. Hence, the pollen source influences haploid frequency via its effect on the endosperm. Paternal haploids are also obtained via anther culture, but maternal haploids offer more advantages for potato breeding because paternal haploid production requires gene(s) for androgenic competence, which are not always available in all tetraploid potato cultivars. Gametes with the sporophytic chromosome number are referred to as 2n gametes. Some authors called them “numerically unreduced gametes,” but Peloquin (and thereafter the students under his mentorship) avoided this term because normal gametes in any species have the haploid (n) number; i.e., 2n gametes would be 2x in diploids, 4x in tetraploids, and so on. Another reason he cautioned against the use of the term is that while chromosome number is not reduced in the formation of 2n gametes, the so-called “reduction division” typically does occur. The failure of students to appreciate this fact leaves them with an incomplete impression of the larger process. Premeiotic, meiotic, and postmeiotic abnormalities during gamete formation are correlated with the production of 2n gametes and there are at least six distinct possible modes of 2n gamete formation: premeiotic doubling, first division restitution (FDR), chromosome replication during meiotic interphase, second division restitution (SDR), postmeiotic doubling, and apospory. FDR and SDR mechanisms are respectively the most common modes of 2n pollen and 2n eggs formation in potato. In potato, FDR 2n gametes transmit on average 80% of the heterozygosity of the diploid progenitor to the tetraploid hybrid offspring. In contrast, SDR 2n gametes transfer on average less than 40% of the diploid heterozygosity to tetraploid hybrids. Parallel orientation of the spindles in the second meiotic division is the most frequent mechanism of 2n pollen formation in most tuber-bearing Solanum spp. This meiotic abnormality is under the genetic control of the recessive gene ps (parallel spindles), which appears to be ubiquitous among Solanum species, but 2n pollen frequency could be affected by variable expressivity and incomplete penetrance. Omission of the second division after a normal first division appears to be the most common mode of SDR 2n egg formation in potato haploids, and haploidspecies hybrids, which is controlled by a recessive meiotic mutant (os) in diploid potato. Genetic background and environment may affect the

1. DEDICATION

5

expressivity of this gene that also shows incomplete penetrance, and the frequency of modifier genes may enhance 2n egg expressivity. Ploidy Manipulations for Genetic Enhancement The findings of mechanisms that underlie parthenogenesis for producing maternal haploid plants and the formation of gametes with the parental chromosome number (or 2n gametes), established Peloquin’s international prestige and credentials. Ploidy manipulations with haploids, 2n gametes, and wild species still remains today as one of the most impressive and exciting crop germplasm enhancement methods ensuing from cytogenetics research. In Stan’s words “the potato is unsurpassed in the facility with which sets of chromosomes can be manipulated. This allows a germplasm enhancement strategy that involves species, haploids, 2n gametes and endosperm balance number (EBN). The species are the source of genetic diversity, haploids provide a method for ‘capturing’ the diversity, and 2n gametes and EBN are involved in an effective and efficient method of transmitting diversity to cultivars.” There are two main methods for ploidy manipulations in potato: unilateral sexual polyploidization (4x-n gametes × 2x-2n gametes or vice versa) and bilateral sexual polyploidization (ensuing from crosses between 2x-2n gametes producing parents). For these breeding schemes, the diploid progenitors are developed by crossing potato haploids with tuber-bearing diploid species (2n = 2x = 24 chromosomes). Maternal haploids, which are easily extracted through parthenogenesis from most tetraploid cultivars, are crossed with diploid species for breeding at the diploid level. The locally adapted haploid-species hybrids are selected because they possess 2n gametes, acceptable tuber characteristics, and sometimes additional desired attributes, e.g., disease or pest resistance. Most of the hybrids ensuing from sexual polyploidization matings in potato are tetraploids. Triploid hybrids from tetraploid-diploid or diploid-diploid crosses are very rare due to a strong triploid block in potato. The best diploid parents for tetraploid-diploid crosses are those producing FDR 2n gametes because hybrid vigor associated with high yields may be maximized by multi-allelism per locus in potato. Not surprisingly, tetraploid hybrids derived from diploid progenitors producing FDR 2n pollen often outyield their half-sib tetraploid hybrids from intermating tetraploid progenitors. However, maximum heterozygosity does not appear to be universal and depends on the genetic background of the crossing material, because of the importance of adaptation for the optimum expression of hybrid vigor in potato.

6

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

With this knowledge, Stan and his “school”—particularly in Italy, Poland, and the Centro Internacional de la Papa (CIP, Lima, Peru)—were able to develop new potato genotypes that combine high and stable yield, plus disease or pest resistance, which also allow the widening of potato growing in areas of the world that were previously unsuitable for this crop. The potato genotypes ensuing from their work were amenable to both fresh table markets and the chipping industry. Some of these materials became parents of cultivars now grown across the world. One potato cultivar (‘Snowden’) ensuing from conventional breeding work of Peloquin and colleagues at the University of Wisconsin is a leading chipping cultivar in North America. Endosperm Balance Number (EBN) and Hybridization Barriers The endosperm is a distinct trait among the Angiosperms, which results from double fertilization; one male gamete unites with the egg to form the zygote and the other male gamete with the central cell to form the endosperm, thereby making this tissue necessary for normal seed development. As Peloquin pointed out many times, “One of my colleagues defines endosperm as the tissue which along with potato feeds the world.” Endosperm research by him and co-workers at the University of Wisconsin led to the Endosperm Balance Number (EBN) hypothesis to explain endosperm development in interploidy crosses, both intraspecific and interspecific. Under this theory, normal endosperm development only occurs when a balance of 2 EBN from the female parent matches with 1 EBN from the male parent in the resulting endosperm. Any deviations from this 2 maternal:1 paternal EBN ratio leads to faulty endosperm and lack of normal seed. For example, in 4x × 2x crosses and 2x × 4x crosses, endosperm development is regularly abnormal since the female:male ratios are 4:1 and 1:1, respectively. However, if 2n gametes function in the diploid, the ratio is 2:1 and endosperm development is normal. The fact that Mexican tetraploid species are easy to cross with most diploid species from South America—yielding triploid hybrids— suggested that both species sets possess 2 EBN, while they are unable to cross with Andean tetraploid species that are 4 EBN. This endosperm dosage system is typical of species possessing a “triploid block.” However, triploids from crosses between tetraploids and diploids arise occasionally from misfertilization, mitotic abnormalities in the gametophyte, and/or mitotic misdivisions in the endosperm. The above results led to the hypothesis—tested thereafter with success by many researchers worldwide, that EBN and 2n gametes are more

1. DEDICATION

7

important than actual chromosome number for predicting the success of crosses and the ploidy of the resulting progeny. The EBN of a species, which initially was thought to be controlled by a few genes rather than by the whole genome, determines, therefore, the effective ploidy, and natural gene flow may occur among species with distinct chromosome numbers but the same EBN. The EBN can be assigned for most Solanum species on the basis of the crossing behavior of a species with a standard species of known EBN. Further research by some of his former co-workers demonstrated that three genes control EBN in potato, and gave convincing evidence for the participation of the EBN incompatibility system and 2n gametes in the origin and evolution of polyploids in tuber-bearing Solanum species. In this regard, the EBN determined gene pools among potato and wild Solanum species. The primary gene pool consists of old and modern tetraploid cultivars, tetraploid Andean landraces, and tetraploid breeding populations (i.e., 4 EBN polysomic polyploid tetraploid species). Diploid cultivars or breeding populations and diploid tuber-bearing Solanum species (2 EBN) producing 2n gametes and hexaploid (4 EBN) species also belong to this primary gene pool. The secondary gene pool includes disomic tetraploid (2 EBN) and diploid (1 EBN) tuber-bearing Solanum species, which may cross with the crop primary gene pool after isolation barriers (mainly due to EBN) are overcome. Wild diploid nontuber bearing Solanum species (1 EBN) of the series Etuberosa are in the tertiary potato gene pool, which could only cross with the primary crop gene pool through bridge species and embryo rescue. Other researchers expanded the EBN theory to many plant taxa, as a unifying concept to predict endosperm function in intraspecific-interploidy or interspecific crosses. True Potato Seed (TPS) Potato production from true (sexual) seed was not new because this propagation system was used in the Andes by the Incas. Andean farmers still grow potato from true seed for disease elimination, stock rejuvenation, and creation of new cultivars. However, another major global impact of Peloquin’s research was the development of breeding methods for using true seed rather than tubers for growing potato—particularly in warm tropical environments, where potato growers are affected by the high cost of seed-tubers, and lack of clean planting materials because of high pest pressure. TPS lowers production costs, reduces the incidence of pests such as viruses that are not transmitted by true seed, and allows true seed to be the source of planting material even if parental

8

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

plants are diseased. TPS technology enables low-income small landholders in the developing world to grow potato, thereby expanding the geographic range of this crop worldwide, especially in locations where transport and cold storage of seed-tubers are not feasible. The production of potatoes from true seed has increased dramatically in areas of India, Bangladesh, and China where they are grown between two crops of rice. Potato cultivars in modern high-input agricultural systems are homogeneous tetraploid genotypes. These cultivars, which are generally produced by cross-pollination, show a great uniformity due to vegetative propagation by tubers. In this agricultural system, tubers are harvested from a potato plant that grew from a single-sprouted tuber. Hence, TPS cultivars must combine plant characteristics for true seed production, and for tuber production from true seed propagules. Conventional potato breeding for vegetative propagation relies on selecting desired allelic combinations for further clonal multiplication, while breeding for potato production from true seed should be based on phenotypically uniform hybrid offspring with a high frequency of favorable heterogeneous allelic combinations. The genetic improvement for TPS production must, therefore, consider a breeding strategy for a sexually propagated crop that is grown for the harvest of its vegetative part, i.e., the tuber. TPS hybrids from crosses between tetraploid parents are the most popular for potato production from true seed. Although heterogeneous gametes may result from meiosis of heterozygous progenitors, some tetraploid hybrid offspring from such crosses show very uniform plant and tuber phenotypes. Tetraploid clones are selected according to their reproductive and agronomic characteristics. Earliness, desired tuber characteristics such as color, shape, number and size, profuse flowering, fertility, and high berry and seed set are the most important attributes of the selected tetraploid parents. Specific crosses for commercial production of TPS hybrids are recommended after testing specific combining ability between locally selected parents. Peloquin and his co-workers suggested other methods for producing tetraploid TPS hybrids through unilateral sexual or bilateral sexual polyploidization. A high frequency of 2n gametes will be very important for commercial TPS hybrids derived from sexual polyploidization after pollinating by hand. Furthermore, recurrent selection effectively increases 2n gametes frequency in diploid potatoes. Nonetheless, production costs of TPS can be lowered by eliminating pollen collection and hand pollination. Hence, since the 1980s Peloquin and his co-workers investigated new breeding systems for alternative TPS production. The cheapest material for potato production from true seed will be derived

1. DEDICATION

9

from open pollination. Open-pollinated TPS costs are significantly lower than those of TPS hybrids because the labor skills and other additional required investments are smaller for its production. However, TPS hybrids always outyield open-pollinated and selfed offspring. The heterozygosity level in the tetraploid sporophyte affects the phenotypic expression of the diploid gametophytic generation in potato; e.g., open pollinated berry and fruit sets are correlated with pollen stainability. It appears that natural pollinators of potato such as bumblebees prefer male fertile plants, especially when pollen viability is at its highest. This finding may explain why tuber yield does not always decrease after successive open-pollinated generations derived from heterogeneous true seed offspring. A synthetic cultivar propagated by open pollination may be, therefore, achievable for potato production from TPS. Open-pollinated offspring could result from selfing but the percentage of hybrid offspring may be increased in open-pollinated derived seedlings because of the variable rate of outcrossing in cultivated tetraploid potato. Selection of vigorous plants in artificial mixtures of hybrid and self-pollinated TPS generations proved, therefore, to be successful in the identification of hybrids in respective TPS generations. The identification of TPS hybrids may be improved by adding a marker gene in the selection scheme, e.g., interplanting a diploid first division restitution 2n pollen-producing male progenitor with a genetic marker (yellow tuber flesh color) among tetraploid female progenitors permitted the selection of TPS hybrids among derived open-pollinated offspring from true seed harvested from the female progenitors. Such TPS hybrids (ensuing from open pollination) had better seedling vigor, more flowers, higher pollen stainability, larger open-pollinated berry set, and greater tuber yield than their open-pollinated half-sibs. Also, the elimination of weak plants during population thinning in the TPS nursery and subsequent interplant competition in densely sown beds for production of tuberlets (the first generation tubers from seedlings) reduce dramatically the frequency of inbred genotypes. There are some disadvantages for potato production from true seeds using transplanted seedlings, e.g., a reliable water supply is required for the successful establishment of seedling in the production field, and labor costs should be added for transplanting. Likewise, potato tuber yield from tuberlets is always higher than that recorded in TPS seedlings, but specific gravity does not change in both generations. Many small tubers are harvested from TPS seedlings, whereas few normal-size tubers are often obtained from tuber propagules. In addition, TPS seedlings require 2 to 3 extra weeks for tuber maturation as compared to tuber propagules. Hence, tuberlets rather than TPS seedlings

10

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

appear to be the most promising propagules for potato production from true seed. Furthermore, tuberlets sold as “seed tubers” are suitable for developing a seed tuber system. Off-season production of tuberlets in well-controlled nursery beds will be another advantage of this seed system, because it provides propagules for the next planting season or in environments where direct sowing of TPS does not seem feasible due to short water supply, high labor cost for thinning and manual weeding, or short growing season. Lastly, selection in the seedling generation may improve the performance of tuberlets from open-pollinated and hybrid offspring. Selection for tuber color, shape, and skin in the TPS seedling nursery may lead to agronomically uniform tuberlets. Peloquin and co-workers reported higher tuber yields in selected tetraploid hybrid offspring from bilateral sexual polyploidization than in those of commercial cultivars. However, both tetraploid and diploid hybrids ensue from diploid-diploid crosses in potato, because polyploid frequency depends on the percentage of 2n gametes in both parents. The formation of first division restitution 2n eggs through desynapsis offers an option for only tetraploid progeny ensuing from bilateral sexual polyploidization. Through their work on TPS breeding, a new method of producing inexpensive tetraploid hybrid true seed offspring through bilateral sexual polyploidization and natural insect pollination was suggested. It consists of using unrelated, locally adapted diploid haploidspecies hybrids as parents—according to their combining ability, with profuse flowering, attractiveness to bumblebees (natural pollinators), and other desired attributes. The diploid male parent has high male fertility and very high frequency of (or almost only) first division restitution 2n pollen, and a heterozygous monogenic dominant marker tightly linked to the centromere. The female parent combines male and female fertility, self-incompatibility, very high frequency (or almost only) 2n egg production, and lacks a dominant marker (recessive genotype). Both diploid haploid-species parents are grown following an interplanting field design to allow natural pollination and gene flow between them. TPS are harvested only from female parents and are grown in a nursery to eliminate those showing poor vigor and lacking dominant phenotype of the male progenitor. Hence, the tuberlet harvest will include mostly (if not only) tetraploid hybrids for potato production in the field. With this TPS scheme, emasculation, pollen collection, and hand pollination are eliminated, thereby reducing by 50% the costs of producing hybrid tetraploid seed. It would be desirable to select diploid parents that are able to set 10,000 hybrid seeds per plant with this method for producing TPS.

1. DEDICATION

11

THE MENTOR Stan Peloquin was among the most prolific trainers of graduate students in the history of the University of Wisconsin—Madison. A total of 93 graduate students (51 PhD, 42 MSc) and 27 visiting scientists from 34 countries were guided by Stan. Many of them returned to their country of origin or pursued a professional career in international agriculture. Through their work, they help to enhance food production in the developing world. Many of his students are now heads of advanced research organizations or lead international agricultural research institutes. Stan has the ability to infuse a “love” for research in science to young professionals. He is fiercely loyal and devoted to the students’ best interest. His enthusiasm and broad interest stimulate not only his students but also colleagues and peers. His former students and visiting research fellows know Stan as someone who enjoys the company of others, has strong education principles, and has a great desire to convey his knowledge. He derives intense satisfaction from the success of each and every student or visiting scientist who comes to his laboratory at the University of Wisconsin. His students and visiting fellows can never forget that their professional education and careers were developed, molded, and launched under his strong and vibrant personality. Many feel that much of their success was related not only to what they learned from him about science but also about life. We all learned one principle: “hard work always pays off.” Professor Stanley J. Peloquin has devoted a lifetime to merging basic cytogenetic and genetic research to achieve direct applicability in crop improvement. His plant breeding philosophy of “putting genes into a usable form” is a prime example of farsightedness in science. His efforts are an inspiration to us all and we proudly dedicate this volume of Plant Breeding Reviews to him and his achievements.

SELECTED REFERENCES Arndt, G. C., and S. J. Peloquin. 1990. The identification and evaluation of hybrid plants among open pollinated true seed families. Am. Potato J. 67:293–304. Arndt, G. C., J. L. Rueda, H. M. Kidane-Mariam, and S. J. Peloquin. 1990. Pollen fertility in relation to open pollinated true seed production in potatoes. Am. Potato J. 67:499–505. Ascher, P. D., and S. J. Peloquin. 1966. Effect of floral aging on the growth of compatible and incompatible pollen tubes in Lolium longiflorum (Thumb.). Am. J. Bot. 53:99–102. Ascher, P. D., and S. J. Peloquin. 1966. Influence of temperature on incompatible and compatible pollen tube growth in Lolium longiflorum. Can. J. Genet. Cytol. 8:661–664.

12

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

Ascher, P. D., and S. J. Peloquin. 1968. Pollen tube growth and incompatibility following intra- and inter-specific pollinations in Lolium longiflorum. Am. J. Bot. 55:1230–1234. Ascher, P. D., and S. J. Peloquin. 1970. Temperature and the self-incompatibility reaction in Lolium longiflorum Thumb. J. Am. Soc. Hort. Sci. 95:586–588. Buso, J. A., F. A. S. Aragao, F. J. B. Reifschneider, L. S. Boiteux, and S. J. Peloquin. 2002. Assessment under short-day conditions of genetic materials derived from three potato breeding strategies 4x–4x (intra-Tuberosum), 4x–2x (FDR 2n-pollen) and 4x–4x (diplandrous tetraploid). Euphytica 126:437–446. Buso, J. A., F. J. B. Reifschneider, L. S. Boiteux, and S. J. Peloquin. 1999. Effects of 2n pollen formation by first-meiotic division restitution with and without crossover on eight quantitative traits in 4x–2x potato progenies. Theor. Appl. Genet. 98:1311–1319. Buso, J. A., L. S. Boiteux, and S. J. Peloquin. 1999. Comparison of haploid TuberosumSolanum chacoense versus Solanum phureja-haploid Tuberosum hybrids as staminate parents of 4x–2x progenies evaluated under distinct crop management systems. Euphytica 109:191–199. Buso, J. A., L. S. Boiteux, and S. J. Peloquin. 1999. Evaluation of 4x–2x progenies from crosses between potato cultivars and haploid tuberosum—Solanum chacoense hybrids under long day conditions. Ann. Appl. Biol. 135:35–42. Buso, J. A., L. S. Boiteux, and S. J. Peloquin. 1999. Multitrait selection system using populations with a small number of interploid (4x–2x) hybrid seedlings in potato: Degree of high-parent heterosis for yield and frequency of clones combining quantitative agronomic traits. Theor. Appl. Genet. 99:81–91. Buso, J. A., L. S. Boiteux, and S. J. Peloquin. 2000. Heterotic effects for yield and tuber solids and type of gene action for five traits in 4x potato families derived from interploid (4x–2x) crosses. Plant Breed. 119:111–117. Buso, J. A., L. S. Boiteux, G. C. C. Tai, and S. J. Peloquin. 1999. Chromosome regions between centromere and proximal crossovers are the physical sites of major effect loci for yield in potato: genetic analysis employing meiotic mutants. Proc. Natl. Acad. Sci. (USA) 96:1773–1778. Buso, J. A., L. S. Boiteux, G. C. C. Tai, and S. J. Peloquin. 2000. Direct estimation of the effects of meiotic recombination on potato traits via analysis of 4x–2x progenies from synaptic mutants with 2n-pollen formation by FDR without crossing over. Theor. Appl. Genet. 101:139–145. Camadro, E. L., and S. J. Peloquin. 1980. The occurrence and frequency of 2n pollen in three diploid Solanums from northwest Argentina. Theor. Appl. Genet. 56:11–15. Camadro, E. L., and S. J. Peloquin. 1981. Cross-incompatibility between two sympatric polyploid Solanum species. Theor. Appl. Genet. 60:65–70. Camadro, E. L., and S. J. Peloquin. 1982. Selfing rates in two wild polyploid Solanums. Am. Potato J. 59:197–203. Carputo, D., A. Barone, T. Cardi, A. Sebastiano, L. Frusciante, and S. J. Peloquin. 1997. Endosperm balance number manipulation for direct in vivo germplasm introgression to potato from a sexually isolated relative (Solanum commersonii Dun.). Proc. Natl. Acad. Sci. (USA) 94:12013–12017. Carputo, D., L. Frusciante, and S. J. Peloquin. 2003. The role of 2n gametes and endosperm balance number in the origin and evolution of polyploids in the tuber-bearing Solanums. Genetics 163:287–294. Carputo, D., T. Cardi, L. Frusciante, and S. J. Peloquin. 1995. Male fertility and cytology of triploid hybrids between tetraploid Solanum commersonii (2n=4x=48, 2 EBN) and Phureja-Tuberosum haploid hybrids (2n=2x=24, 2 EBN). Euphytica 83:123–129. Chujoy, J. E., and S. J. Peloquin. 1986. Barriers to interspecific hybridization between Solanum chacoense Bitt. and S. commersonii Dun. Am. Potato J. 63:416–417.

1. DEDICATION

13

Cipar, M. S., S. J. Peloquin, and R. W. Hougas. 1964. Inheritance of incompatibility in hybrids between Solanum tuberosum haploids and diploid species. Euphytica 13:163–172. Cipar, M. S., S. J. Peloquin, and R. W. Hougas. 1964. Variability in the expression of selfincompatibility in hybrids tuber-bearing diploid Solanum species. Am. Potato J. 41:163–172. Cipar, M. S., S. J. Peloquin, and R. W. Hougas. 1967. Haploidy and the identification of self-incompatibility alleles in cultivated potato groups. Can. J. Genet. Cytol. 9:511–518. Concilio, L., and S. J. Peloquin. 1987. Evaluation of yield and other agronomic characteristics of true potato seed families and advanced clones from different breeding schemes. p. 262–263. In: G. J. Jellis and D. E. Richardson (eds.), The production of new potato varieties: Technological advances. Cambridge Univ. Press, Cambridge. Concilio, L., and S. J. Peloquin. 1987. Tuber yield of true potato seed families from different breeding schemes. Am. Potato J. 64:81–85. Concilio, L., and S. J. Peloquin. 1991. Evaluation of the 4x × 2x breeding scheme in a potato breeding program adapted to local conditions. J. Genet. Breed. 45:13–18. Darmo, E., and S. J. Peloquin. 1990. Performance and stability of nine 4x clones from 4x–2x crosses and four commercial cultivars. Potato Res. 33:352–360. Darmo, E., and S. J. Peloquin. 1991. Use of 2x Tuberosum haploid-wild species hybrids to improve yield and quality in 4x cultivated potatoes. Euphytica 53:1–9. Davies, C. S., M. J. Ottman, and S. J. Peloquin. 2002. Can germplasm resources be used to increase the ascorbic acid of stored potatoes? Am. J. Potato Res. 79:295–299. de La Puente Ciudad, F., and S. J. Peloquin. 1976. Haploids of group Andigena. Am. Potato J. 51:278. den Nijs, T. P. M., and S. J. Peloquin. 1977. 2n gametes in potato species and their function in sexual polyploidization. Euphytica 26:585–600. den Nijs, T. P. M., and S. J. Peloquin. 1977. Polyploid evolution via 2n gametes. Am. Potato J. 54:377–386. den Nijs, T. P. M., E. F. Leue, and S. J. Peloquin. 1980. Topiary: a mutant character in Solanum infundibuliforme. J. Hered. 71:57–60. den Nijs, T. P. M., E. F. Leue, M. Iwanaga, and S. J. Peloquin. 1978. Detection of the mode of 2n egg formation by pollen stainability distributions. Am. Potato J. 55:387–389. Desborough, S., and S. J. Peloquin. 1966. Disc electrophoresis of tuber proteins from Solanum species and interspecific hybrids. Phytochemistry 5:727–733. Desborough, S., and S. J. Peloquin. 1967. Esterase isozymes from Solanum tubers. Phytochemistry 6:989–994. Desborough, S., and S. J. Peloquin. 1968. Disc-electrophoresis of proteins and enzymes from styles, pollen and pollen tubes of self-incompatible cultivars of Lolium longiflorum. Theor. Appl. Genet. 38:327–331. Desborough, S., and S. J. Peloquin. 1968. Potato variety identification by use of electrophotetic patterns of tuber proteins and enzymes. Am. Potato J. 45:220–229. Desborough, S., and S. J. Peloquin. 1969. Acid gel electrophoresis of tuber proteins from Solanum species. Phytochemistry 8:425–429. Desborough, S., and S. J. Peloquin. 1969. Tuber proteins from haploids, selfs and cultivars of Group Tuberosum separated by acid gel electrophoresis. Theor. Appl. Genet. 39:43–47. Frusciante, L., A. Barone, C. Conicella, and S. J. Peloquin. 1989. Use of different breeding schemes and selection of parental lines for TPS production. p. 178–181. In: K. M. Louwes, H. A. J. M. Toussaint, and L. W. L. Dellaert (eds.), Parental line breeding and selection in potato breeding. Pudoc, Wageningen, Netherlands. Groza, H. I., B. D. Brown, D. Kichefski, S. J. Peloquin, and J. Jiang. 2002. Red Companion: A new red potato variety for fresh market. Am. J. Potato Res. 79:133–137.

14

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

Hanneman, R. E. Jr., and S. J. Peloquin. 1981. Genetic-cytoplasmic male sterility in progeny of 4x–2x crosses in cultivated potatoes. Theor. Appl. Genet. 59:53–56. Hanneman, R. E., and S. J. Peloquin. 1967. Crossability of 24-chromosome potato hybrids with 48-chromosome cultivars. Eur. Potato J. 10:62–73. Hanneman, R. E., and S. J. Peloquin. 1968. Ploidy levels of progeny from diploid-tetraploid crosses in the potato. Am. Potato J. 45:255–261. Hermundstad, S. A., and S. J. Peloquin. 1985. Germplasm enhancement with potato haploids. J. Hered. 76:463–467. Hermundstad, S. A., and S. J. Peloquin. 1986. Tuber yield and tuber traits of haploid-wild species F1 hybrids. Potato Res. 29:289–297. Hermundstad, S. A., and S. J. Peloquin. 1987. Breeding at the 2x level and sexual polyploidization. p. 197–210. In G. J. Jellis and D. E. Richardson (eds.), The production of new potato varieties: Technological advances. Cambridge Univ. Press, Cambridge. Hopper, J. E., and S. J. Peloquin. 1968. X-ray inactivation of the stylar component of the self-incompatibility reaction in Lolium longiflorum. Can. J. Genet. Cytol. 10:941–944. Hopper, J. E., and S. J. Peloquin. 1976. Analysis of stylar self-incompatibility competence by use of heat induced inactivation. Theor. Appl. Genet. 47:291–297. Hopper, J. E., P. D. Ascher, and S. J. Peloquin. 1967. Inactivation of self-incompatibility following temperature pretreatments of styles in Lolium longiflorum. Euphytica 16:215–220. Hougas, R. W., and S. J. Peloquin. 1957. A haploid plant of the potato variety Katahdin. Nature 180:1209–1210. Hougas, R. W., and S. J. Peloquin. 1958. The potential of potato haploids in breeding and genetic research. Am. Potato J. 35:701–707. Hougas, R. W., and S. J. Peloquin. 1960. Crossability of Solanum tuberosum haploids with diploid Solanum species. Eur. Potato J. 3:325–330. Hougas, R. W., and S. J. Peloquin. 1962. Exploitation of Solanum germplasm. p. 21–24. In: D. S. Correll (ed.), The potato and its wild relatives. Texas Research Foundation, Renner, Texas. Hougas, R. W., S. J. Peloquin, and A. C. Gabert. 1964. The effect of the seed-parents and pollinator on haploid frequency in Solanum tuberosum. Crop Sci. 4:593–595. Hougas, R. W., S. J. Peloquin, and R. W. Ross. 1958. Haploids of the common potato. J. Hered. 49:103–106. Iwanaga, M., and S. J. Peloquin. 1979. Synaptic mutant affecting only megasporogenesis in potatoes. J. Hered. 70:385–389. Iwanaga, M., and S. J. Peloquin. 1982. Origin and evolution of cultivated tetraploid potatoes via 2n gametes. Theor. Appl. Genet. 61:161–169. Iwanaga, M., R. Ortiz, M. S. Cipar, and S. J. Peloquin. 1991. A restorer gene for geneticcytoplasmic male sterility in cultivated potatoes. Am. Potato J. 68:19–28. Jansky, S. H., G. L. Yerk, and S. J. Peloquin. 1990. The use of potato haploids to put 2x wild species germplasm in a usable form. Plant Breed. 104:290–294. Johnston, S. A., T. P. N. den Nijs, S. J. Peloquin, and R. E. Hanneman Jr. 1980. The significance of genetic balance to endosperm development in interspecific crosses. Theor. Appl. Genet. 57:5–9. Kichefski, D. F., A. A. Quinn, and S. J. Peloquin. 1976. The effectiveness of selection during early clonal generations in varietal breeding. Am. Potato J. 53:370–371. Kidane-Mariam, H. M., and S. J. Peloquin. 1974. The effect of direction of hybridization (4x × 2x vs. 2x × 4x) on yield on cultivated potatoes. Am. Potato J. 51:330–336. Kidane-Mariam, H. M., and S. J. Peloquin. 1975. Method of diplandroid formation and yield of progeny from reciprocal (4x–2x) crosses. J. Am. Soc. Hort. Sci. 100:602–604.

1. DEDICATION

15

Kidane-Mariam, H. M., and S. J. Peloquin. 1975. Potato clonal variation with respect to the relation between fruit set and tuber yield. E. African Agr. For. J. 40:256–260. Kidane-Mariam, H. M., G. C. Arndt, A. C. Macaso-Khwaja, and S. J. Peloquin. 1985. Comparisons between 4x × 2x hybrid and open-pollinated true potato seed families. Potato Res. 28:35–42. Kidane-Mariam, H. M., G. C. Arndt, A. C. Macaso-Khwaja, and S. J. Peloquin. 1985. Hybrid vs. open-pollinated TPS families. p. 25–37. In: Innovative methods for propagating potatoes. Centro Internacional de la Papa, Lima, Perú. Kotch, G. P., and S. J. Peloquin. 1987. A new source of haploid germplasm for genetic breeding research. Am. Potato J. 64:137–141. Kotch, G. P., R. Ortiz, and S. J. Peloquin. 1992. Genetic analysis by use of potato haploid populations. Genome 35:103–108. Leue, E. F., and S. J. Peloquin. 1980. Selection for 2n gametes and tuberization in Solanum chacoense. Am. Potato. J. 57:189–195. Leue, E. F., and S. J. Peloquin. 1982. The use of topiary gene in adapting Solanum germplasm for potato improvement. Euphytica 31:65–72. Lynch, D. R., D. Kichefski, S. J. Peloquin, C. S. Schaupmeyer, L. la Croix, D. Waterer, W. T. Andrew, J. Holley, N. Crowe, B. McConnell, G. A. Nelson, and B. Rex. 1991. Niska: A maincrop chipping potato cultivar with high specific gravity and good quality after storage. Am. J. Potato Res. 68:143–149. Lynch, D. R., S. J. Peloquin, L. M. Kawchuk, C. A. Schaupmeyer, J. Holley, D. K. Fujimoto, D. Driedger, D. Waterer, J. Wahab, and M. S. Goettel. 2001. AC Glacier Chip: a highyielding chip cultivar for long-term storage. Am. J. Potato Res. 78:327–332. Macaso-Khwaja, A. C., and S. J. Peloquin. 1983. Tuber yield of families from openpollinated and hybrids true potato seed. Am. Potato J. 60:645–651. Masson, M. F., and S. J. Peloquin. 1987. Heterosis for tuber yields and total solids contents in 4x × 2x FDR-CO crosses. p. 213–227. In: G. J. Jellis and D. E. Richardson (eds.), The production of new potato varieties: Technological advances. Cambridge Univ. Press, Cambridge. Mendiburu, A. O., and S. J. Peloquin. 1976. Sexual polyploidization and depolyploidization: Some terminology and definitions. Theor. Appl. Genet. 48:137–143. Mendiburu, A. O., and S. J. Peloquin. 1977. Bilateral sexual polyploidization in potatoes. Euphytica 26:573–583. Mendiburu, A. O., and S. J. Peloquin. 1977. The significance of 2n gametes in potato breeding. Theor. Appl. Genet. 49:53–61. Mendiburu, A. O., and S. J. Peloquin. 1979. Gene centromere mapping by 4x–2x matings in potato. Theor. Appl. Genet. 54:177–180. Mendiburu, A. O., S. J. Peloquin, and D. W. S. Mok. 1974. Potato breeding with haploids and 2n gametes. In: K. J. Kasha (ed.), First Intl. Symp. Haploids in Higher Plants. Univ. of Guelph, Ontario, Canada. p. 249–258. Mok, D. W. S., and S. J. Peloquin. 1975. Breeding value of 2n pollen (diplandroid) in tetraploid × diploid crosses in potato. Theor. Appl. Genet. 46:307–314. Mok, D. W. S., and S. J. Peloquin. 1975. The inheritance of three mechanisms of diplandroid (2n pollen) formation in diploid potatoes. Heredity 35:295–302. Mok, D. W. S., and S. J. Peloquin. 1975. Three mechanisms of 2n pollen formation in diploid potatoes. Can. J. Genet. Cytol. 17:217–225. Mok, D. W. S., H. K. Lee, and S. J. Peloquin. 1974. Identification of potato chromosomes with giemsa. Am. Potato J. 51:337–341. Mok, D. W. S., S. J. Peloquin, and A. O. Mendiburu. 1976. Genetic evidence for mode of 2n pollen formation and S-locus mapping in potatoes. Potato Res. 19:157–164.

16

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

Mok, D. W. S., S. J. Peloquin, and T. R. Tarn. 1975. Cytology of potato triploids producing 2n pollen. Am. Potato J. 52:171–174. Mok, I. C., and S. J. Peloquin. 1982. Sexual polyploidization and protein diversity in potato. Am. Potato J. 59:480. Mortenson, L. R., S. J. Peloquin, and R. W. Hougas. 1964. Germination of Solanum tuberosum pollen on artificial media. Am. Potato J. 41:322–328. Okwuagwu, C. O., and S. J. Peloquin. 1981. A method of transferring the intact parental genotype to the offspring via meiotic mutants. Am. Potato J. 58:512. Ortiz, R., and S. J. Peloquin. 1991. A new method of producing 4x hybrid true potato seed. Euphytica 57:103–107. Ortiz, R., and S. J. Peloquin. 1991. Breeding for 2n egg production in haploid × species 2x potato hybrids. Am. Potato J. 68:691–703. Ortiz, R., and S. J. Peloquin. 1992. Associations between genetic markers with quantitative traits in potato. J. Genet. Breed. 46:395–400. Ortiz, R., and S. J. Peloquin. 1992. Recurrent selection for improvement of 2n gamete production in 2x potatoes. J. Genet. Breed. 46:383–390. Ortiz, R., and S. J. Peloquin. 1993. Mapping the flower pigmentation locus in potato. J. Genet. Breed. 47:171–173. Ortiz, R., and S. J. Peloquin. 1993. Population improvement in the development of 2x parents in potato using exotic germplasm. J. Genet. Breed. 47:81–88. Ortiz, R., and S. J. Peloquin. 1994. Effect of sporophytic heterozygosity on the male gametophyte of the tetraploid potato (Solanum tuberosum). Ann. Bot. 73:61–64. Ortiz, R., and S. J. Peloquin. 1994. Manipulaciones de ploidía en el mejoramiento genético de la papa. Turrialba 43:196–209. Ortiz, R., and S. J. Peloquin. 1994. Use of 24-chromosome potatoes (diploids and dihaploids) for genetical analysis. p. 133–154. In: J. E. Bradshaw and G. R. Mackay (eds.), Potato genetics. CAB International, Wallingford, Oxon, UK. Ortiz, R., D. S. Douches, G. P. Kotch, and S. J. Peloquin. 1993. Use of haploids and isozyme markers for genetic analysis in the polysomic polyploid potato. J. Genet. Breed. 47:283–288. Ortiz, R., M. Iwanaga, and S. J. Peloquin. 1993. Male sterility and 2n pollen in 4x progenies derived from 4x × 2x and 4x × 4x crosses in potato. Potato Res. 36:227–236. Ortiz, R., M. Iwanaga, and S. J. Peloquin. 1994. Breeding potatoes for developing countries using wild tuber bearing Solanum spp. and ploidy manipulations. J. Genet. Breed. 48:89–98. Ortiz, R., M. Iwanaga, and S. J. Peloquin. 1997. Evaluation of FDR diploid and tetraploid parents in potato under two contrasting day length environments. Plant Breed. 116:353–358. Ortiz, R., R. Freyre, S. J. Peloquin, and M. Iwanaga. 1991. Adaptation to day length and yield stability of families from 4x × 2x crosses in potato. Euphytica 56:187–198. Ortiz, R., S. J. Peloquin, R. Freyre, and M. Iwanaga. 1991. Efficiency of potato breeding using FDR 2n gametes for multi trait selection and progeny testing. Theor. Appl. Genet. 82:602–608. Parfitt, D. E., and S. J. Peloquin. 1977. Variation of vine and tuber yield as a function of harvest date and cultivar. Am. Potato J. 54:410–417. Parfitt, D. E., and S. J. Peloquin. 1981. The genetic basis for tuber greening in 24chromosome potatoes. Am. Potato J. 58:299–304. Parfitt, D. E., and S. J. Peloquin. 1982. Yield trials and economic analysis to select cultivars for silage production. Am. Potato J. 59:395–401. Parfitt, D. E., S. J. Peloquin, and N. A. Jorgensen. 1982. The nutritional value of pressed potato vine silage. Am. Potato J. 59:415–423. Peloquin, J. G., S. J. Peloquin, and D. W. S. Mok. 1975. Vertical slab polyacrylamide gel electrophoresis of tuber proteins. Am. Potato J. 52:280–281.

1. DEDICATION

17

Peloquin, S. J. 1979. Breeding methods for achieving phenotypic uniformity. In: Planning Conf. on the Production of Potatoes from True Seed. Centro Internacional de la Papa, Lima, Perú. p. 151–155. Peloquin, S. J. 1981. Chromosomal and cytoplasmic manipulations. p. 1–17. In: K. J. Frey (ed.), Plant breeding II. Iowa State Univ., Ames, IA. Peloquin, S. J. 1982. Meiotic mutants in potato breeding. In: G. P. Redei (ed.), Stadler Genetic Symp. St. Louis, Missouri. p. 99–109. Peloquin, S. J. 1982. New approaches to breeding for the potato of the year 2000. In: W. L. Hooker (ed.), Proc. Intl. Potato Center 10th Annual Congress. Centro Internacional de la Papa, Lima, Perú. Peloquin, S. J. 1983. Genetic engineering with meiotic mutants. p. 311–316. In: D. Mulcahy and E. Ottaviano (eds.), Pollen biology and use in plant breeding. Elsevier Pub. Co, Amsterdam. Peloquin, S. J. 1984. Utilization of exotic germplasm in potato breeding: Germplasm transfer with haploids and 2n gametes. p. 147–158. In: W. L. Brown (ed.), Conservation and utilization of exotic germplasm to improve varieties. Pioneer Hi-Bred Intl. Inc., Des Moines, IA. Peloquin, S. J. 1986. Chromosome engineering with meiotic mutants. p. 47–52. In: D. Mulcahy and E. Ottaviano (eds.), Biotechnology and ecology of pollen. Springer Verlag, New York. Peloquin, S. J., A. C. Gabert, and R. Ortiz. 1996. Nature of pollinator effect in potato (Solanum tuberosum L.) haploid production. Ann. Bot. 77:539–542. Peloquin, S. J., and R. Ortiz. 1992. Techniques for introgressing unadapted germplasm to breeding populations. p. 485–508. In: H. T. Stalker and J. P. Murphy (eds.), Plant breeding in the 1990s. CAB International, Wallingford, Oxon, UK. Peloquin, S. J., and R. W. Hougas. 1958. Fertility in two Solanum tuberosum haploids. Science 128:1340–1341. Peloquin, S. J., and R. W. Hougas. 1959. Decapitation and genetic markers as related to haploidy in Solanum tuberosum. Eur. Potato J. 2:176–183. Peloquin, S. J., and R. W. Hougas. 1960. Genetic variation among haploids of the common potato. Am. Potato J. 37:289–297. Peloquin, S. J., C. O. Okwuagwu, E. F. Leue, S. A. Hermundstad, D. M. Stelly, S. H. Schroeder, and J. E. Chujoy. 1984. Use of meiotic mutants in breeding. p. 65–74. In: Present and Future Strategies for Potato Breeding and Improvement. Centro Internacional de la Papa, Lima, Perú. Peloquin, S. J., G. C. Arndt, and H. M. Kidane-Mariam. 1985. Utilization of ploidy manipulations for true potato seeds. In: Innovative Methods for Propagating Potatoes. Centro Internacional de la Papa, Lima, Perú. p. 17–24. Peloquin, S. J., G. L. Yerk, and J. E. Werner. 1989. Ploidy manipulations in potato. p. 167–178. In: K. W. Adolph (ed.), Chromosomes: Eukaryotic, prokaryotic and viral. CRC Press, Boca Raton, FL. Peloquin, S. J., G. L. Yerk, J. E. Werner, and E. Darmo. 1989. Potato breeding with haploids and 2n gametes. Genome 32:1000–1004. Peloquin, S. J., J. E. Chujoy, and J. E. Werner. 1992. 4x hybrid progeny from 2x–2x crosses in potato. p. 23–29. In: A. Mariani and S. Tavoletti (eds.), Gametes with somatic chromosome number in the evolution and breeding of polyploid polysomic species: Achievements and perspectives. Consiglio Nazionale delle Ricerche, Perugia, Italy. Peloquin, S. J., J. E. Werner, and G. L. Yerk. 1990. The use of potato haploids in genetics and breeding. p. 79–92. In: P. K. Gupta and T. Tsuchiya (eds.), Chromosome engineering in plants. Elsevier Pub. Co., Amsterdam.

18

R. ORTIZ, L. FRUSCIANTE, AND D. CARPUTO

Peloquin, S. J., L. S. Boiteux, and D. Carputo. 1999. Meiotic mutants in potato: Valuable variants. Genetics 153:1493–1499. Peloquin, S. J., R. W. Hougas, and A. C. Gabert. 1966. Haploidy as a new approach to the cytogenetics and breeding of Solanum tuberosum. p. 21–29. In: R. Riley and K. R. Lewis (eds.), Chromosome manipulations and plant genetics. Oliver & Boyd, Edinburgh. Peloquin, S. J., S. H. Jansky, and G. L. Yerk. 1989. Potato cytogenetics and germplasm utilization. Am. Potato J. 66:629–638. Peloquin, S. J., T. P. M. den Nijs, and A. O. Mendiburu. 1978. Sexual polyploidization versus somatic doubling. Proc. 7th EAPR Triennal Conf. p. 201–212. Pérez-Ugalde, O., R. W. Hougas, and S. J. Peloquin. 1964. Fertility of S. phureja-haploids and S. tuberosum F2 hybrids. Am. Potato J. 41:256–262. Quinn, A. A., and S. J. Peloquin. 1973. Use of experimental tetraploids in potato breeding. Am. Potato J. 50:415–420. Quinn, A. A., D. W. S. Mok, and S. J. Peloquin. 1974. Distribution and significance of diplandroids among Solanums. Am. Potato J. 51:16–21. Ross, R. W., S. J. Peloquin, and R. W. Hougas. 1964. Fertility of hybrids from Solanum phureja and haploid S. tuberosum matings. Eur. Potato J. 7:81–89. Schonnard, G., and S. J. Peloquin. 1991. Performance of true potato seed families. I. Effect of level of inbreeding. Potato Res. 34:397–407. Schonnard, G., and S. J. Peloquin. 1991. Performance of true potato seed families. II. Comparison of transplants versus seedlings. Potato Res. 34:409–418. Schroeder, S. H., and S. J. Peloquin. 1983. Seed set in 4x–2x crosses as related to 2n pollen frequency. Am. Potato J. 60:527–536. Serquén, F. C., and S. Peloquin. 1996. Variation for agronomic and processing traits in Solanum tuberosum haploids × wild species hybrids. Euphytica 89:185–191. Simon, P. W., and S. J. Peloquin. 1976. Pollen vigor as a function of mode of 2n gamete formation in potatoes. J. Hered. 67:204–208. Simon, P. W., and S. J. Peloquin. 1977. The incidence of paternal species on the origin of callus in anther culture of Solanum hybrids. Theor. Appl. Genet. 50:53–56. Simon, P. W., and S. J. Peloquin. 1980. Inheritance of electrophoretic variants of tuber proteins in Solanum tuberosum haploids. Biochem. Genet. 18:1055–1061. Souter, B. W., J. C. Daws, and S. J. Peloquin. 1980. 2n pollen formation via parallel spindles in the potato cultivar Sebago. Am. Potato J. 57:449–455. Stelly, D. M., and S. J. Peloquin. 1986. Diploid female gametophytes formation in 24chromosome potatoes. Genetic evidence for the prevalence of second meiotic division restitution mode. Can. J. Genet. Cytol. 28:101–108. Stelly, D. M., and S. J. Peloquin. 1986. Formation of 2n megagametophytes in triploid tuber-bearing Solanums. Am. J. Bot. 73:1351–1363. Stelly, D. M., and S. J. Peloquin. 1986. Screening for 2n female gametophytes, female fertility and 2x × 4x crossability in potatoes (Solanum spp.). Am. Potato J. 62:519–529. Stelly. D. M., S. J. Peloquin, R. G. Palmer, and C. F. Crane. 1984. Mayer’s hemalum-methyl salicytate: A stain-clearing technique for observation within whole ovules. Stain Technol. 59:155–161. Thill, C. A., and S. J. Peloquin. 1995. A breeding method for accelerated development of cold chipping clones in potato. Euphytica 84:73–80. Thill, C. A., and S. J. Peloquin. 1995. Inheritance of potato chip color at the 24 chromosome level. Am. Potato J. 71:629–646. Wangenheim, K. H. von, S. J. Peloquin, and R. W. Hougas. 1960. Embryological investigations in the formation of haploids in the potato (Solanum tuberosum). Z. Vererbungslhre 91:391–399.

1. DEDICATION

19

Watanabe, K. N., M. Orrillo, S. Vega, M. Iwanaga, R. Ortiz, R. Freyre, G. L. Yerk, S. J. Peloquin, and K. Ishike. 1995. Selection of diploid potato clones from diploid (haploid × species) F1 hybrid families for short day conditions. Breed. Sci. 45:341–347. Watanabe, K., and S. J. Peloquin. 1989. Occurrence of 2n pollen and ps gene frequencies in cultivated groups and their related wild species in tuber-bearing Solanums. Theor. Appl. Genet. 78:329–336. Watanabe, K., and S. J. Peloquin. 1991. The occurrence and frequency of 2n pollen in 2x, 4x, and 6x wild, tuber-bearing Solanum species from Mexico and Central and South America. Theor. Appl. Genet. 82:621–626. Watanabe, K., and S. J. Peloquin. 1992. Cytological basis of 2n pollen formation in a wide range of 2x, 4x, and 6x taxa from tuber-bearing Solanum species. Genome 36:8–13. Watanabe, K., S. J. Peloquin, and M. Endo. 1991. Genetic significance of mode of polyploidization: somatic doubling or 2n gametes. Genome 34:28–34. Werner, J. E., and S. J. Peloquin. 1987. Frequency and mechanisms of 2n egg formation in haploid Tuberosum-wild species F1 hybrids. Am. Potato J. 64:641–654. Werner, J. E., and S. J. Peloquin. 1990. Inheritance of two mechanisms of 2n egg formation in 2x potatoes. J. Hered. 81:371–374. Werner, J. E., and S. J. Peloquin. 1991. Occurrence and mechanisms of 2n egg formation in 2x potato. Genome 34:862–870. Werner, J. E., and S. J. Peloquin. 1991. Potato haploid performance in 2x × 4x crosses. Am. Potato J. 68:801–811. Werner, J. E., and S. J. Peloquin. 1991. Significance of allelic diversity and 2n gametes for approaching maximum heterozygosity in 4x potatoes. Euphytica 58:21–29. Werner, J. E., and S. J. Peloquin. 1991. Yield and tuber characteristics of 4x progeny from 2x × 2x crosses. Potato Res. 34:352–357. Yeh, B. P., S. J. Peloquin, and R. W. Hougas. 1964. Meiosis in Solanum tuberosum haploids and haploid-haploid F1 hybrids. Can. J. Genet. Cytol. 6:393–402. Yeh, B. P., S. J. Peloquin, and R. W. Hougas. 1965. Pachytene chromosomes of the potato (Solanum tuberosum Group Andigena). Am. J. Bot. 52:1014–1020. Yerk, G. L., and S. J. Peloquin. 1988. 2n pollen in eleven 2x, 2 EBN wild species and their haploid-wild species hybrids. Potato Res. 31:581–589. Yerk, G. L., and S. J. Peloquin. 1989. Comparison of 2n and non-2n pollen producing haploid × wild species hybrids in potato. J. Hered. 80:450–453. Yerk, G. L., and S. J. Peloquin. 1989. Evaluation of tuber traits of 10, 2x (2EBN) wild species through haploid × wild species hybrids. Am. Potato J. 66:731–739. Yerk, G. L., and S. J. Peloquin. 1990. Performance of haploid × wild species 2x hybrids (involving five newly evaluated species) in 4x × 2x families. Am. Potato J. 67:405–417. Yerk, G. L., and S. J. Peloquin. 1990. Selection of potato haploid parents for use in crosses with 2x (2 Endosperm Balance Number) wild species. Crop Sci. 30:943–946.

2 Politics of Plant Breeding Cary Fowler Center for International Environment & Development Studies (NORAGRIC), Agricultural University of Norway, P.O. Box 5003, 1432 Aas, Norway Richard L. Lower Dept. of Horticulture, 1575 Linden Drive, University of Wisconsin, Madison, WI 53706

I. INTRODUCTION II. GERMPLASM, PLANT BREEDING, AND THE FIGHT FOR RIGHTS A. Historical B. “Rights” in the International Arena C. Expanded Claims of Rights D. The FAO Treaty on Plant Genetic Resources and the Global Plan of Action 1. The Treaty 2. The Global Plan of Action E. The International Treaty on Plant Genetic Resources for Food and Agriculture 1. The Treaty 2. Expected Impact of the Treaty III. THE DEBATE OVER BIOTECHNOLOGY A. Hurdles to Construct, Hurdles to Jump B. Taking a Step Backward IV. PLANT BREEDERS’ CHOICES A. Public vs. Private B. GMOs or Non-GMOs C. Intellectual Property Protection and Rights D. Looking Ahead LITERATURE CITED

Plant Breeding Reviews, Volume 25. Edited by Jules Janick. ISBN 0471666939 © 2005 John Wiley & Sons, Inc. 21

22

C. FOWLER AND R. LOWER

I. INTRODUCTION Over the past century, plant breeding has contributed to and benefited from the growth of the biological sciences that support both the science and art of plant improvement. Major building blocks include contributions from genetics, statistics, and supporting biological curricula. In the past 30 years, the rapid increase in molecular sciences, principally genomics, paired with advanced technologies in information processing and communication have fostered the development and usage of plant breeding tools. The uses of molecular marker techniques are an example of these contributions to plant breeding. The development of germplasm collections and the establishment of gene banks throughout the world have improved the conservation and increased the availability of genetic resources. These events coupled with the development of teams of scientists in related sciences, such as plant pathology, entomology, plant physiology, and biochemistry, have resulted in major advances in plant breeding. Free exchanges of germplasm and information have been major factors in the success and growth of plant breeding. In spite of the timely advances in plant breeding and its supporting/cooperating sciences, the luxury of and access to plant collections, the myriad of molecular tools and techniques, and in spite of the tremendous and increasing need the world has for the products of plant breeding, the practicing plant breeder is likely to encounter complex and often frustrating obstacles and dilemmas in his or her professional career and research. More and more frequently these difficulties are not technical in nature. At their heart, they are political. Politics and plant breeding have an ancient, if unappreciated and uncomfortable, relationship. Liberally interpreted, the genetic vestiges of this history—those associated with the spread of agriculture, and the first plant collecting expeditions 4500 years ago—are still with us today in a world in which virtually every country is highly dependent on crops that originated outside its own region (Plucknett et al. 1987; Palacios 1998). Positing such a lengthy history depends on how one defines “plant breeding” and “politics.” No such definitional problems exist today, however, if we assert that politics is a part of plant breeding, and viceversa. The entire profession of plant breeding and virtually each of its component activities is currently a matter of political concern and/or controversy. Consider: • The acquisition and transfer of plant genetic resources—the “raw material” for plant breeding and the first step in the breeding

2. POLITICS OF PLANT BREEDING

•

•

•

•

23

process—is the subject of two international laws and several international contractual agreements. Moreover, the subject continues to evoke strong emotions and is still under active dispute. The priorities and goals of plant breeders are under constant scrutiny. Not infrequently, the “choices” made by breeders are painted in stark and evocative terms. Are plant breeders producing hybrids or open-pollinated cultivars? Genetically modified organisms (GMOs) or non-genetically modified organisms (non-GMOs)? Cultivars tailored for small, resource-poor farmers and marginal ecological niches, or varieties aimed more at commercial agriculture and high-input systems? Cultivars with or without intellectual property rights? Cultivars released by the public or by the private sector or jointly? To many, each of the questions has a single “correct” answer. A constant undercurrent in deliberations over funding and prioritysetting concerns the division of labor between public and private plant breeding efforts. Recent surveys show declining numbers of public plant breeders and increasing numbers in the private sector (Frey 1996). As public efforts depend on public funding, politics is always present. The political process determines how much funding there will be, to which crops funding will go, and how far “upstream” or “downstream” the research should be, i.e., how far from a commercial product or a consumer. Indeed, the public views may be changing as public funding is declining and products of breeding programs come to be viewed as potential sources of financial support for research programs. The methodology and tools used by the breeders are also under question. Are “participatory” methods employed to produce cultivars with the farmers? Are certain biotechnologies used or eschewed? How is field testing done and where is it done? Guidelines vary greatly from country to country, and even within countries, and apply particularly when biotechnological tools are used and when genetically modified organisms are to be marketed. Marketing of seed is highly regulated and controlled by laws that have not always been welcomed or accepted in the seed trade. Quarantine laws, straightforward and laudable on the surface, are frequently the subject of manipulation for political and trade purposes. Questions of nomenclature—for example, producing and selling something called Basmati rice in the U.S.—can also be controversial (Tripp 1997). Marketing seed via grower contracts or agreements on the usage of the resulting crop can be very restrictive and are common with GMOs.

24

C. FOWLER AND R. LOWER

• Even questions of “rights” are now routinely raised in regard to the practice of plant breeding. For instance, what rights do suppliers of germplasm (or their descendants) have to “benefit-sharing” associated with the final product, a cultivar produced in a breeding program? How is the “Right to Food,” an initiative being promoted by a number of countries in international forums, related to the rights (or obligations) of plant breeders? Note that the “right to food” has been proclaimed in a number of texts, including the Universal Declaration of Human Rights (1948), the International Covenant on Economic, Social and Cultural Rights (1976), and the Rome Declaration on World Food Security issued during the World Food Summit (1996), and is a subject of current discussion and debate in various intergovernmental forums. In addition, 22 countries have enshrined this right in their constitutions. • Ethical concerns are being raised in a political fashion particularly as it regards the use of biotechnologies and the notion of what is “natural,” or not. While many consumers may think of “natural” as being both safe and “better,” many scientists would be uncomfortable in defining such terms with any scientific precision; and many would observe that numerous “naturals” are toxic and some can even be deadly. • We live in changing times. During the past three decades numerous changes in mechanisms for intellectual property rights have taken place. Over the same time frame, dramatic advances in biotechnologies and in related molecular sciences and information technologies have greatly altered and affected the way plant breeding is conducted. Handling these new technologies, when generated in the public sector, is causing considerable stress within public institutions. • Finally, the structure of plant breeding is undergoing a major transformation. As a result of declining political influence, public sector activities are in a state of decline, worldwide. Budgets are down. Fewer students are showing up at universities to study plant breeding, and certain kinds of crops and breeding activities are being abandoned, or “rationalized” as it is more commonly explained. Moreover, the glitz and glamour of some of the biotechnologies continue to attract many who likely would have been plant breeders in earlier decades. In this paper, we examine some, but not all of the political issues facing plant breeding, as there are far too many. We have selected several that we believe are most interesting to the actual practice of plant breed-

2. POLITICS OF PLANT BREEDING

25

ing in the future. We have chosen not to address in a detailed manner at least one subject that meets these criteria—the issue of the role, status, and funding of public sector plant breeding, though we have touched upon such subjects as they relate to the United States. This is a critically important subject and one that has been developed by numerous authors. We believe it is worthy of discussion. But, unlike the other topics we have chosen to address, this one is predominantly country-specific. The reality is different in the U.S. and Norway and different still in other developed and developing countries. Suffice it to say that the public sector has a virtual monopoly in training new plant breeders, in taking on certain kinds of fundamental basic research, and in working with crops, regions, farming systems, markets, and types of farmers (e.g., the resource-poor) that are not and cannot be priorities for the private sector. This is a broad and vitally important “mandate.” At the international level, it is addressed by international agricultural research centers (the Consultative Group on International Agricultural Research System (CGIAR), Asian Vegetable Research and Development Center (AVRDC), and the Tropical Agriculture Research and Higher Learning Center (CATIE)) and at the national level by national agricultural research systems, including universities. In recent years, all of these have suffered dramatic funding cuts that concern us greatly. However, for pragmatic reasons, we have chosen to concentrate in this paper on issues that can be dealt with more generically, those that are more international in scope.

II. GERMPLASM, PLANT BREEDING, AND THE FIGHT FOR RIGHTS A. Historical By the time the first Europeans had set foot in the New World, maize, for example, had already migrated to the high mountain valleys of the Andes as well as to jungle areas of the Amazon, and was being grown under irrigation in the dry coastal lands of western South America (Plucknett et al. 1987). The spread of many food crops was “natural.” We know of no disputes involved. People and plants had been on the move for thousands of years. While maize spread throughout the world, reaching West Africa in the second half of the 16th century, and while grains whose Center of Diversity was the Near East found their way into Europe, Asia, and the Americas (Columbus introduced wheat along with chickpea, melons, onions, grape vines, and sugar cane on his second voyage), the acquisition,

26

C. FOWLER AND R. LOWER

transfer, and spread of other species was anything but routine. Some of the earliest plant collecting expeditions dating back 3500 years were mounted by armies, which is precisely what it would take to collect materials in some countries today! Moreover, during the colonial era, a number of species—principally of industrial, plantation, or medicinal crops—were acquired by questionable means. Accusations of theft are not uncommon in the histories of the time (Wolf and Wolf 1936). Colonial powers were interested in establishing production under their own control for certain high-value crops. In some cases they openly desired monopoly (e.g., for indigo) and attempted to establish it. As early as 1556, for instance, Spain’s Council of the Indies, convened in Madrid, passed a law making it illegal for foreigners to explore for plants in Spain’s New World possessions (Haughton 1979). Those who were subjugated and from whom the planting materials were acquired frequently did not give their consent (Brockway 1979; Crosby 1986). Our purpose is not to delve into ancient history; it is to make a simple point in introducing the subject of contemporary conflicts over plant genetic resources, namely that questions of ownership and rights have been an important feature of agriculture for centuries. Disputes have been the norm, not the exception, historically. The colonial era was characterized by the desire to obtain particular species. Mendelian laws were unknown and there were few efforts to acquire and limited ability to use intra-specific diversity, one exception being France’s interest in broadening the variability of certain economic species (Plucknett et al. 1987). The development of sizeable urban markets in Europe and North America in the 1800s provided the impetus for the rise of commercial agriculture. Varietal differentiation became commercially important. In the U.S., the government distributed millions of packets of seeds of countless cultivars or “landraces” acquired abroad and encouraged experimentation by farmers. The establishment of county agricultural fairs in the mid-1800s gave added encouragement to both experimentation and to the “fixing” of ideal traits. Farm journals and the ever-popular Farmer’s Almanac were replete with advice on how farmers could improve their crops. And a significant portion of the seed used was produced on the farm (USDA 1918; Elder 1921). Today, interestingly, much of the same information is available on hundreds of Internet sites. Cultivars were valued for different traits including their keeping and shipping qualities. By the turn of the century (19th to 20th), for example, there were 60,000 refrigerated railway cars operating in North America, and the Illinois Central was boasting of its 25-car Thunderbolt train, the first “all-strawberry train in the nation” designed to link the farms of southern Illinois with Chicago (Stover 1975; Taylor 1901). Despite the

2. POLITICS OF PLANT BREEDING

27

absence of large, formal “scientific” (Mendel-informed) plant breeding programs, crop cultivars—such as those tailored for shipping—were well established, and certain ones of them were known and appreciated in the marketplace. By 1925 one publication lists 1362 cultivars of strawberry, for example, but an early 20th century U.S. government survey documents the astounding existence of more than 7000 distinct, named cultivars of apples in the U.S. in use in the 1800s (Ragan 1926). In this atmosphere, it should come as little surprise that certain interests—in this case, the nursery trade—began to claim rights and lobby for legal recognition of their ownership of varieties. The first such proposal was made by the Committee on Registration of the Peninsula Horticultural Society in the U.S. in 1891, though prior to this there had been other efforts to achieve control through regulations standardizing nomenclature. Such proposals were made and rejected a number of times before the U.S. Congress finally passed the Plant Patent Act of 1930 (Fowler 2000). This legislation covered asexually-reproducing species excluding, in practice, some tuber crops used for food, such as potatoes. It was only a matter of time before pressure mounted to pass similar legislation for sexually-reproducing (i.e., seed) crops. In 1961, representatives of six European countries gathered to create the Union for the Protection of New Varieties of Plants (UPOV). The UPOV “process” created a consensus among commercial breeding interests about how new varieties should be legally protected through the establishment of intellectual property rights. The UPOV solution took protection out of the realm of patent law by creating its own sui generis system. Plant Breeders Rights—“Plant Variety Protection” in the U.S.—was born, based on the criteria of the new cultivar being distinct, uniform, and stable. It was well recognized that farmers themselves, even in developed countries, were producing new varieties of value. Speaking of the 19th century, Anderson and Brown (1952a,b) recognized the contributions of farmers and observed that “there is detailed evidence of the purposeful blending of diverse cultivars . . . the controlled breeding of new varieties by farmers themselves was more frequent than anyone would believe.” They went on to note the “highly elaborate methods of selection.” Nevertheless, such cultivars fell short of the new legal requirements for recognition. B. “Rights” in the International Arena While this is not the place for a detailed exploration of the immediate roots of contemporary conflicts in the 1980s and 1990s over plant genetic resources (PGR), it seems evident that the content, flavor, and tone of the controversies that arose over intellectual property rights in the U.S. and

28

C. FOWLER AND R. LOWER

Europe and over PGR more generally at the Food and Agriculture Organization (FAO), were inspired by earlier political movements (e.g., civil rights and environmental), as well as by the politics of the Cold War and the political discourses over development and “under-development” (Fowler 1994). It was also rooted in a certain sense of history, as recounted above. Several small but energetic non-government organizations (NGOs) began to agitate first for increased attention to and funding for PGR conservation programs. Principally, these were the International Coalition for Development Action, the National Sharecroppers Fund/Rural Advancement Fund (later RAFI), and Genetic Resources Action International (GRAIN). These groups were also concerned about two other matters: (1) the perceived imbalance created by the recognition of the rights of plant breeders (mainly in developed countries) without a corresponding recognition of the contributions and rights of (mainly developing) countries and farmers who had “created,” conserved, and ultimately donated the myriad landraces used in such breeding programs, and (2) the perceived diminution of the public domain—of what was then described as the “common heritage of mankind”—by the privatization of the PGR through intellectual property rights. These concerns led to proposals (introduced by Mexico) for a binding legal agreement on plant genetic resources at FAO, one that would fix the status of PGR as a resource in the public domain, and would recognize the existence of “Farmers’ Rights” in juxtaposition to Plant Breeders Rights. Years of acrimonious debate followed at FAO, divided strictly along developed/developing (or north/south) country lines. The early FAO battles produced two outcomes, controversial at the time, but less so now in retrospect. First, an intergovernmental Commission on Plant Genetic Resources was created. Second, governments adopted a non-binding International Undertaking on Plant Genetic Resources. The Commission, which met for the first time in 1985, was to serve as a specialized forum for the discussion and resolution of PGRrelated problems. The Undertaking, adopted in 1983, was to serve as a statement of how the international community viewed PGR—it viewed PGR as a “common heritage.” A subsequent resolution (Resolution 5/89) by the FAO Conference defined Farmers’ Rights as “rights arising from the past, present and future contributions of farmers in conserving, improving, and making available plant genetic resources, particularly those in the centres of origin/diversity.” Resolution 5/89 constituted an important milestone. The concept of Farmers’ Rights had now found a place in intergovernmentally approved text. But, it was a Pyrrhic victory for developing countries and NGOs. A UN document observing where rights arise from is not the same as defin-

2. POLITICS OF PLANT BREEDING

29

ing what those rights actually are. Lacking a definition, Farmers’ Rights was a concept and only a concept. There was nothing legally enforceable. It was an empty vessel into which everyone was free to provide content. Initially, however, this was not of great concern to its proponents, as Farmers’ Rights had not been envisaged as a system of legal “rights” but as a concept or political slogan designed to pressure governments into increased support for PGR conservation “in support” or “recognition” of those undefined rights (Fowler 1998). Common heritage came with common responsibilities, responsibilities that were not being met. Resolution 3/91, adopted in 1991, made the link explicit by stating that “Farmers’ Rights will be implemented through international funding on plant genetic resources, which will support plant genetic conservation and utilization programmes, particularly, but not exclusively, in the developing countries.” C. Expanded Claims of Rights Claims of “rights” are typically asserted and contested at the national level. In many countries, “rights” are enshrined in constitutions. Politically speaking, additions/alterations to the list are not considered trivial. Nevertheless, in the relatively restricted field of genetic resources, the contest over rights has often taken place in international forums, in part because those advocating such rights lack “critical mass” politically to take up the fight country-by-country. As the mid-1990s approached, there was no obvious way to continue the Farmers’ Rights debate at FAO. An accommodation, of sorts, had been made in the available negotiating document—the International Undertaking and associated resolutions. FAO was moving on to other, rather more pragmatic and scientific concerns, such as the development of a Global Plan of Action for Plant Genetic Resources. While not devoid of politics, the Plan was not the best or most logical vehicle for a renewed fight over Farmers Rights’ or any other kind of rights. Moreover, countries were busy asserting sovereignty over their own PGR and their interest in devolving rights and power to farmers was not high. NGOs, the leading advocates of Farmers’ Rights in the first place, were still passionate about the subject, but Farmers’ Rights competed with other political campaigns concerning intellectual property rights, biopiracy, the CGIAR, and the World Trade Organization. In this context—and lacking any legal/political forum on which to focus, the concept of Farmers’ Rights began to evolve. One NGO publication from 1995 (GRAIN 1995) lists and seems to support an astonishing array of rights: Farmers’ Rights, Expanded Farmers’ Rights, Community Rights, Community Intellectual Rights, Communal Rights, Peoples’

30

C. FOWLER AND R. LOWER

Rights, Indigenous Peoples’ Rights, Indigenous Community Rights, Local and Indigenous Community Rights, Traditional Resource Rights, Collective Rights, Neighboring Rights, Development Rights, Human Rights, Religious Rights, Heritage Rights, Territorial Rights, Territorial and Tenurial Rights, and, “bundles” of rights! Rather than become more focused and precise—more defined—Farmers’ Rights became more diffuse and fuzzy. Smith (1997) cataloged 16 different claims being associated with the term “Farmers’ Rights” alone, ranging from moral principles of recognition and rights to compensation, subsidies and land, to provision of funding for maintenance of current lifestyles, participatory breeding, and genetic resource conservation. In addition, support for such rights and interests now led some proponents to become highly critical of genebanks, plant breeding and breeders and formal scientific research systems, and to advocate instead for a combination of in-situ and on-farm initiatives coupled with associated “rights.” While the in-situ approach has much to commend itself, we note that its promotion by certain groups has frequently been framed in political terms as being in contrast to and in some ways in opposition to ex situ approaches. It is this political context—and its implications for plant breeding and plant breeders—that we draw attention to here. Otherwise, we do not see it as being within the scope of this particular article to address the merits of the case for or against any “rights.” A more interesting shift was and still is taking place. However imprecisely defined the above “rights” might be when found in various documents, many are taking on the flavor of being a form of intellectual or property rights. Far from emphasizing the “common heritage” history and quality of the PGR, many groups seem now to treat the resource as a commodity, one to be owned, sold, or bartered by its owners. That such a position has evolved in reaction to more dominant intellectual property rights systems seems obvious. Yet, it is also ironic, given the history of the debate and indeed of the actors themselves as opponents of intellectual property rights. GRAIN has gone so far as to describe as “biopiracy” an agreement whereby PGR collections held by the CGIAR Centers are placed essentially in the public domain for the international community under the auspices of FAO (GRAIN 2002). In the early 1990s, they had been among the groups pressuring in favor of this agreement. It is difficult to say whether the dispute could be resolved by national promulgation of rights or intellectual property laws applicable to landraces/farmers’ varieties, or whether the real problem could only be addressed by more fundamental shifts in political and economic power. Without resolution, however, advocacy groups are unlikely to devote time or resources to supporting ex situ conservation efforts or plant breeding efforts (unless the latter are strictly on their terms, i.e., partic-

2. POLITICS OF PLANT BREEDING

31

ipatory with farmers, and in accordance with the “rights” noted previously). Thus, PGR conservation and public sector plant breeding has lost a potentially powerful advocate. And, it may in many cases have picked up an energetic opponent, as the CGIAR has experienced in recent years. At the national level, the effects of restrictive access laws and political pressures have combined to limit access to PGR, especially from insitu sources. This situation may persist as long as the question of rights is unresolved, unless the new International Treaty on Plant Genetic Resources takes hold and is seen to provide an equitable solution to the issues of access and benefit sharing.

D. The FAO Treaty on Plant Genetic Resources and the Global Plan of Action 1. The Treaty. Neither the FAO Undertaking (a non-binding agreement) nor the resolutions on Farmers’ Rights (which failed to include an operational definition of the term) served to dispel the underlying tensions. Heightened awareness of the importance of “biodiversity” coupled with historic sensitivities associated with its appropriation and use (noted above) led to negotiations culminating in the Convention on Biological Diversity (CBD) in 1992. The CBD turned the world of the FAO Undertaking upside down, by reasserting claims of national sovereignty over genetic resources. In the space of a decade, developing countries shifted from advocating “common heritage,” to demanding “national sovereignty.” NGOs faced a dilemma: abandon what they perceived to be the politically progressive notion of common ownership, or abandon their alliance with developing countries. Most chose the former, quietly avoiding any public recognition of the obvious conflict and subsequent political flip-flop. Many countries proceeded to try to “commoditize” PGR, exactly what NGOs had complained about when the private sector claimed ownership over varieties through intellectual property rights. The CBD was based, according to John Barton, Professor of Law at Stanford University, on a fundamental misconception at least in regards to PGR of agricultural species—the belief that there was a commercial market for the resources (Agres 2003). While access to PGR was to be granted by “prior informed consent” and on the basis of “mutually agreed terms,” no one was buying. Negotiators, however, had recognized that the CBD could not appropriately apply to existing collections of PGR and they specifically invited FAO to initiate negotiations on this subject. FAO got on with the task. In the meantime, access/transfers of genetic resources slowed to a trickle. A study of the Andean region found that several countries had rejected all applications for access and

32

C. FOWLER AND R. LOWER

that for the region as a whole “no case can be identified where benefit sharing has actually taken place.” (Correa 2003) FAO’s strategy was two-pronged. It would provide a forum for the development of two related agreements: a binding legal agreement on access and benefit-sharing, and a Global Plan of Action for the conservation and utilization of PGR. It was assumed by many that the “benefitsharing” arrangement would provide the means for implementing the Global Plan of Action, implementation of which would generate benefits. Both documents would be formally adopted at a large intergovernmental conference in Leipzig in 1996. This, however, was not to be. 2. The Global Plan of Action. FAO established a Secretariat to draft a major report on the state of the world’s plant genetic resources for food and agriculture as well as the Global Plan of Action itself, and to prepare for the Leipzig Conference. The State of the World’s Plant Genetic Resources, the most comprehensive assessment undertaken to date, was prepared largely on the basis of information in 154 country reports submitted by the countries themselves, as well as additional information provided by scientists, in particular from the International Plant Genetic Resources Institute (IPGRI) and other centers of the Consultative Group on International Agricultural Research. NGO concerns by this time had come to focus almost exclusively on political matters (e.g., Farmers’ Rights) and though formally invited and facilitated to do so, they provided few inputs to the State of the World report or to the prescriptive measures found in the Global Plan of Action. The State of the World’s Plant Genetic Resources functioned as a “problem statement,” identifying gaps in conservation, use and, to a much lesser degree, in the distribution of benefits from this use. It clearly indicated that additional efforts (and funding) were needed, but it just as clearly revealed that better coordination and priority-setting could, even with current levels of funding, solve many problems and bring substantial benefits. The Global Plan of Action is organized into four major categories: • • • •

In Situ Conservation and Development Ex Situ Conservation Utilization of Plant Genetic Resources Institutions and Capacity Building

Within these general categories are 20 concrete “activities.” For each activity, the Global Plan of Action provides: • An assessment of the situation, drawing on the State of the World report

2. POLITICS OF PLANT BREEDING

• • • • • •

33

Long-term objectives Intermediate objectives Policy/strategy Capacity requirements/recommendations Research/technology recommendations And, coordination/administrative recommendations

Negotiations on a binding legal agreement—what would eventually become the International Treaty—bogged down and no draft was ready by the time the Leipzig Conference opened in June, 1996. Instead, the Conference concentrated on the Global Plan of Action, which it adopted, and 150 countries formally agreed on what needed to be done to conserve and better utilize plant genetic resources. For the thoughtful observer, however, one nagging question remained: By the time the Treaty was adopted would countries still support the Global Plan, and would they see its implementation as providing the type and size of benefits they required in order to resolve the access/benefit-sharing conundrum? Would the Treaty “package” feature the Plan “front and center,” or would it simply give a nod to it and proceed to try to lock in financial benefits as did the Convention on Biological Diversity? A partial answer to these questions was quick in coming. The Secretariat’s estimate of the costs of implementing the Global Plan of Action was prevented from being put on the table. A number of countries were opposed to considering it. It seems that they saw the value of the Plan not in terms of the benefits that implementation would bring, but in terms of how big the budget would be for implementation. And it was too small. Thus they continued to feel that they were giving more away in PGR than they were receiving in dollars for the Plan’s implementation. E. The International Treaty on Plant Genetic Resources for Food and Agriculture 1. The Treaty. The acrimonious 7-year negotiation of the International Treaty came down to one simple question: how much access for how much benefits? It was understood that countries would provide access to genetic resources. But, what kind of access would this involve, how many crops would this arrangement cover, and what would the terms of access be? Would specific acts of access be linked to concrete benefits that would accrue to the provider? In this negotiation, developing countries were at something of a disadvantage, and not simply because of any differences of “power” (decisions are taken by consensus at FAO). They were still operating under the notion that their PGR were a highly coveted commodity with a substantial

34

C. FOWLER AND R. LOWER

monetary value. They were also smarting from the belief that they were being exploited—that modern cultivars protected by intellectual property rights were being marketed to them at high prices when, in fact, these cultivars were based on genetic materials donated to developed countries’ breeders by developing countries and developing country farmers. In retrospect, it is possible to question whether such analyses (irrespective of their merits) and the emotions they generated served the negotiators well in the negotiations. Was, for example, the Treaty a good vehicle for redressing the types of grievances harbored by developing country delegates? Did it lead them to develop the best and most practical remedies? The jury is still out, at least on the latter question. Developing countries also saw themselves as donors, not as recipients, of germplasm. Consequently, they started with a hard bargaining position. FAO studies, however, revealed that countries—including developing countries—are highly interdependent in terms of PGR. And, the State of the World report indicated that much of the diversity had already “escaped” to genebanks outside of the control of individual developing countries. The implications were reasonably clear: one of the major benefits of the Treaty would be to guarantee access. Access itself was a benefit for everyone. Developed countries started with an equally hard bargaining position. In essence they were returning to the view of “common heritage” in the International Undertaking, which they had fought strongly against a decade earlier. They wanted the new Treaty to guarantee access to all crops. Europe countered Africa’s proposal of an agreement on 6 crops with a proposal for nearly 300 crops, including a few that most delegates (including many Europeans!) had never heard of, and some, such as kudzu, that others would just as soon forget. The sheer breadth of the European proposal, however, served to reinforce developing country suspicions that “their” genetic resources were valuable and sought after by commercial interests. In the end, a measure of common sense prevailed on all sides, and the International Treaty was adopted in 2001, the first treaty of the new millennium (FAO 2001). The Treaty is a watershed in international relations regarding plant genetic resources. It is not altogether inaccurate to say that it is the first international agreement addressing what could be described as a 3000-year running dispute. At the heart of the International Treaty is the creation of a Multilateral System of access and benefit-sharing. This System encompasses 35 crops and a number of temperate forage species. The 35 include most major food crops, but exclude soybean, tomato, groundnut, and tropical forages, as well as most industrial crops (rubber and oil palm, for example). A tremendous asset associated with the Treaty is the genetic

2. POLITICS OF PLANT BREEDING

35

resources, mostly of the world’s major food crops, that are held at the centers of the Consultative Group on International Agricultural Research (CGIAR). Historically, these have been considered as an international heritage and have been freely available to everyone, most recently under the terms of a formal agreement between FAO and the centers in which it is agreed that the centers are holding the materials “in trust” for the benefit of the international community. Several other gene banks outside of the CGIAR system also share the CGIAR philosophy regarding the nature of germplasm being held and its availability. These are usually, but not always, national collections that may be shared widely, but that are nevertheless subject to political whims. For the agreed crops, the Treaty provides for “facilitated” access to PGR that is under the management and control of a Contracting Party (a country that ratifies the Treaty). Access is consistent with respect for applicable intellectual property rights, and is at the discretion of the developer during the period of development. The benefit-sharing provision of the Treaty is activated when someone acquires material from the Multilateral System (for example, from a genebank of a country that has ratified the Treaty), incorporates that genetic material into a product that is a PGR (e.g., a cultivar or line as opposed to a breakfast cereal), commercializes it, and then protects it (through intellectual property rights, for instance) in a way that restricts subsequent access and use of the product. In practice this is assumed by many to mean that if someone takes out a utility patent on the product, benefit-sharing will be required, whereas if one takes out Plant Breeders Rights, benefit-sharing will not be required, as this form of IPR explicitly allows the protected cultivar to be used for further research and breeding. The Treaty calls for a “fair and equitable” share of benefits “in line with commercial practice” to be paid into a fund which in turn will be used to support genetic resources related work, primarily in developing countries and countries with economies in transition. The all-important details of what this means precisely have been left to the first meeting of the Treaty’s Governing Body. The Governing Body, at its first meeting, is tasked with agreeing on a Material Transfer Agreement (MTA) to be used when materials within the Treaty’s Multilateral System are transferred. The Governing Body, composed of the countries that have ratified the Treaty, will convene within a year of the date when the Treaty comes into force, which will take place 90 days after the 40th country has ratified it. We expect the Treaty to come into force sometime in 2004. 2. Expected Impact of the Treaty. How the Treaty will affect the “politics of plant breeding” remains to be seen; however, we would argue that the effects will ultimately be determined by three factors:

36

C. FOWLER AND R. LOWER

1. The number of countries that ratify the Treaty and their composition. Canada, Ethiopia, and India have already ratified and EU countries are expected to join shortly. But, will the U.S., Japan, and China ratify? The absence of major players will weaken the Treaty. Not only do the U.S. and Japan have major germplasm collections, they are also among the few countries that have intellectual property rights systems that might trigger the Treaty’s benefit-sharing mechanism. 2. The nature of the benefit-sharing agreement yet-to-be-made. Unless this agreement provides for realistic and concrete benefits, some countries (those who do not see the value of access) may choose not to ratify. Moreover, lack of benefits will undermine the Treaty and eventually lead countries to look for other ways in which to gain benefits from their agro-biodiversity. This could lead to the reemergence of conflict. 3. The scope of the Treaty’s Multilateral System. As noted above, 35 crops are covered, including rice, wheat, maize, potato, and the Brassica family. It is a great start, but for the majority of these crops the genetic resources are widely dispersed and rather freely available already, though it can be quite difficult to access materials, even of these crops, from in-situ conditions. For many, access from ex situ sources has never been problematic, especially for crops held by Centers of the CGIAR. Crops not included within the Multilateral System will be accessed for the most part under terms of the Convention on Biological Diversity, that is, on the basis of prior informed consent and mutually agreed terms. In the past, this has meant very limited access at best. Thus, breeders working with soybean, tomato, onion, tobacco, sugar cane, and so on will find plant collecting to be difficult and access from ex situ sources to be far from assured. Assembling genetic resource collections from scratch in order to initiate a breeding program—for example, on an underutilized or “minor” crop—will be almost impossible. Thus, unless the scope of the Treaty is expanded, the Treaty will serve to circumscribe future breeding efforts to crops whose collections fall under the Treaty—collections that are held by Contracting Parties to the Treaty of crops that are within the Multilateral System. The question then is how much is the Treaty’s scope likely to broaden in the future to encompass more crops. Our reluctant conclusion is not much, at least within the next decade. The Treaty requires complete consensus for changes to the list of crops it covers. One objection is enough to keep a crop out of the Multilateral System. It will take at least a decade before all countries realize what seems obvious to us: that there is less to gain from trying to make carrot genetic resources a marketable commodity, for exam-

2. POLITICS OF PLANT BREEDING

37

ple, than there is in promoting improvement of carrots through the exchange of materials. Despite shortcomings, the Treaty should help defuse and de-politicize tensions over plant genetic resources. Making access and exchange of plant genetic resources more routine will certainly contribute positively to plant improvement efforts. The resolution of long-running disputes and the establishment of international law and norms for access and benefit-sharing also clears the way for an additional and much needed initiative, specifically the creation of the Global Crop Diversity Trust, an endowment fund to support the long-term conservation of PGR, most notably of collections maintained in accordance with the rules of the Treaty. Prior to the adoption of the Treaty, it was hard to argue (or raise money) for an international fund to support “international” collections, when countries were asserting national sovereignty over these materials. Those that assert sovereignty must be prepared to pay the costs of maintenance themselves. Under the Treaty, however, certain crop collections are, in effect, placed in the public domain and it makes sense— as original proponents of the International Undertaking argued—that there is a common and public responsibility vis-à-vis these plant genetic resources. The Trust, created by FAO and the CGIAR, and linked to and made politically feasible by the Treaty, could provide security of funding that no PGR genebank collection has ever had heretofore (Plucknett et al. 1987). To succeed, however, the Trust will need to establish strict criteria that allow it to maintain a focus on supporting the scientific conservation of diversity, not the more politically-enticing conservation of extraneous genebanks. In addition, and in light of “9/11,” the international community cannot ignore the dangers posed to PGR collections as a result of terrorist actions, catastrophes, and so on. This aspect of politics, never a particular concern to plant breeders before, cannot now be ignored. The time may have come to establish a “fail-safe” world PGR collection. In this regard, the offer made by Norway some years ago to host a facility in a secure underground mineshaft in the permafrost of Svalbard near the North Pole could be revisited. The costs would be low; the insurance benefits enormous.

III. THE DEBATE OVER BIOTECHNOLOGY A. Hurdles to Construct, Hurdles to Jump Prior to the 1980s and the development and employment of the new biotechnologies, major innovations in the practice of plant breeding were assimilated with little evident public concern. The notion of

38

C. FOWLER AND R. LOWER

“improvement” as it applied to crop varieties and animal breeds was already well established before Mendel’s laws of heredity were rediscovered in 1900. Darwin had recognized it explicitly in the first chapter of Origin of Species. To be sure, neither Darwin’s theories nor Mendel’s work were immediately accepted in the scientific community. While most quickly agreed that Mendel’s findings explained inheritance in peas, some thought it might apply only to peas or to a limited number of plants. No one, however, questioned the “application” of Mendel in plant breeding, or opposed the breeding strategy advocated by Beal in the 1870s (Fitzgerald 1990). No one, it seems, thought that greater control and precision in the “improvement” (breeding) process was somehow violating the “natural order” or that the products of Mendelian-informed plant breeding were automatically unnatural or unsafe because of the way in which they were produced. More to the point, the same can be said of George Schull’s crossing of inbred lines to produce hybrids at Cold Spring Harbor, Long Island, and Donald Jones’ subsequent work double-crossing inbreds using four inbred lines (Kloppenburg 1988). The age of hybrids in plant breeding was ushered in with silent consent from the consuming public and approval from the farmers. But times were different then. World trade was less developed and regulated and corporate, consumer, and environmental interests in this area were immature at best. The new biotechnologies were introduced into a totally different political and economic context. Political critiques of the goals, techniques, and impacts of plant breeding were well established by the 1980s (Griffin 1972; Palmer 1972a,b; Cleaver 1972; Hightower 1973; George 1977; Lappé et al. 1977; Mooney 1979). Yet, compared to the current discourse on biotechnology, there was something quite different about these early critiques. The issues being raised principally concerned hunger, the economic plight of small farmers, and the rising power of the private sector. The contemporary critique is not so much about the lack of food as it is about the safety of the food being offered, and the potential harm that its production might cause to the environment. In some quarters the debate is also about how GM foods are produced, about the technology and not just its product. An early White Paper published by the U.S. National Academy of Sciences concluded that safety assessments of introducing recombinant DNA-engineered organisms into the environment “should be based on the nature of the organism and the environment into which the organism is introduced, and not on the method by which it was produced.” (NAS 1987) Nevertheless, it is evident that GM products are subjected to a different kind of scrutiny precisely because of the methods used.

2. POLITICS OF PLANT BREEDING

39

Technologies are developed for a purpose, even if all consequences of deployment can never be fully discerned by the scientists involved. As Burns and Ueberhorst (1988) note: “Technologies are instruments of social action with consequences. The introduction of new technologies and the development of technological systems entail more than setting up and using new machines and other physical artifacts. It entails social re-organizing and the making of new rules.”—or as Winner (1987) puts it, “in a fundamental sense . . . determining things is what technology is all about.” But “determining” and “rule-making” are rarely trivial processes when major technologies are involved and when significant social, economic, and legal changes are unfolding. Perhaps this is as it should be. The implication, however, is that those who introduce technologies can never manage how they are perceived, accepted, or regulated, or be in complete control of the full range of consequences the technologies ultimately have. Unlike the earlier “Green Revolution,” biotechnology’s “Gene Revolution” has been taking place mostly in the private sector. The technologies are being developed primarily by the private sector, and many are proprietary, outside the public domain. A sense of the scale and speed of these developments can be gathered from a simple comparison. During the period 1981–1985, a total of 257 patents were issued in the U.S. containing the terms “rice,” “wheat,” or “corn” plus “gene.” During the period 1996–August 2001, 11,475 were issued. In addition, backlogs of patent applications are growing and the time required for processing them has increased to over two years (Rader 2000). The asymmetry of claims reflects the asymmetry of funding. In 1998, the CGIAR spent $25 million on biotechnology research, much to the consternation of biotech critics. But, Monsanto spent $1.26 billion that year (Pardey and Beintema 2001). The mobilization of capital to fund such research combined with the use of the utility patent system in the U.S. not just to protect products, but to lay claim to basic technologies, has raised concerns over the creation of an “anti-commons,” and among other things, over whether the public sector will be able to participate in a meaningful way or will be technologically confined (Heller and Eisenberg 1998; Falcon and Fowler 2002). A variety of “solutions” have been offered ranging from ”no patents on life” to the creation of biotechnology “patent pools.” (Resnik 2003) Castagnera et al. (2002) question whether the public sector is prepared to deal with the challenges of managing intellectual property rights. A survey of public plant breeders in the U.S. revealed that close to half had experienced difficulties in obtaining genetic stocks from private companies, 45% indicated that this had interfered with their research, and almost a quarter indicated that it

40

C. FOWLER AND R. LOWER

had interfered with the training of graduate students (Price 1999). Little wonder then that in a more recent survey of U.S. Land Grant College scientists, the majority see shrinking budgets as an obstacle to their seizing the opportunities provided by biotechnology and view the private sector as the greatest beneficiary of the new technologies and the public sector as being least likely to benefit (Ag Biotechnology Implementation Task Force 2003). B. Taking a Step Backward In the space of a few short years, the amount of land devoted to genetically modified crops has increased dramatically, to approximately 145 million acres worldwide, including 45% of the soybean, 11% of the corn, 20% of the cotton, and 11% of the rapeseed (USDA/USTR 2003). While a great deal of resources have been devoted to the development of these varieties, with obvious implications for the plant breeders of the crops, the trend toward GM crops has not been without controversy. Strong consumer opposition has emerged, particularly in Europe. After approving 18 biotech products, and with 13 applications in the pipeline, the European Union initiated an informal moratorium in June of 1999. This prompted sharp reactions, and eventually a challenge by the U.S. in the World Trade Organization over unfair trade practices. While some are convinced that no evidence exists indicating that such crops pose any danger to human health or the environment, others are not so sure and cite the “precautionary principle” to support a “go slow” approach. Meanwhile, the rhetoric has reached the boiling point. GM opponents have labeled the crops “poison,” described them evocatively as “Frankenfoods” and as “worse than nuclear weapons or radioactive wastes” (Conko and Prakash 2002). GM proponents have retorted that opponents are “Luddites” who are guilty of “murder” for seeking to stop GM food relief to the hungry in Zambia. The verbal food fight is perpetuated and fueled by political leaders, commercial interests, and well-funded advocacy groups on both sides of the divide. While direct causal links are difficult to prove, it seems almost selfevident that the controversy has harmed the reputation of plant breeders and breeding and contributed to reductions and distortions in funding for public sector work. Goodman (2002) notes the worrying downward trend in training of new plant breeding Ph.D.s, which raises the question of whether this too might be linked. It is difficult to discern a silver lining. With polls showing that a large percentage of the American population does not “believe” in Darwin, it is hard to argue that the public debate in the U.S., or in Europe, is based on scientific literacy. As the mother of a

2. POLITICS OF PLANT BREEDING

41

wheat breeder friend of ours remarked recently: “Son, I’ve made a decision: in the future I’m not going to eat any food with genes in it.” The debate, obviously, is not just about science. Some people do not like Picasso no matter how much art experts say that his is great art. Similarly, a sizeable portion of the public sees little compelling reason to welcome genetically modified crops. To them, such foods do not serve their interests or solve any problems they believe they have. The antipathy is so pronounced that there is speculation that “GM concerns could be spilling over to adversely affect demand for non-GM products” such as “food products from animal industries that use GM foodstuffs, and honey . . .” (Foster et al. 2003). While some advocacy groups seem to think that there is something inherently unnatural and bad about the technology itself, the general public’s view seems to be more nuanced. Biotechnology applications in medicine, for instance, do not uniformly or automatically generate instant opposition. In the field of medicine, many applications are viewed as addressing real problems and providing tangible benefits. Moreover, the “risks” are borne by those who have the problem and need a solution. Society at large is not an unwitting guinea pig. Thus, the movement in Europe and elsewhere to require labeling strikes a cord. In a recent commentary, Taylor (2003) points out that “people won’t accept biotechnology under pressure” and argues that the U.S., for example, “should not be on the wrong side of the choice issue.” But, to make an informed choice, the public must become more informed. Therein lies the difficulty as well as the power and danger of the images conjured up by both sides in the debate. Before they give their support, the public must be convinced that plant breeding using the new biotechnologies brings benefits, not just to the bottom line of corporations or even to farmers, but to consumers through better, cheaper, and more nutritious food products, or through positive impacts on the environment. Until that day comes (and because of the intense publicity surrounding these issues), plant breeders and breeding will be inextricably linked with the public’s wary view of biotechnology. The effect will not just be directed to the “image” of the profession, but through funding mechanisms and regulatory measures, it will also have an impact on the scientist’s “freedom to operate,” including the freedom to choose which technology is most appropriate for solving a particular problem. Increased support to public sector research programs would be one step toward generating real and visible benefits, according to Taylor. At the same time, safety issues must be addressed. Assurances from some scientists that there is “no problem” are not sufficient in a world in which sophisticated technologies have frequently been discovered by the public to come with unintended and unforeseen consequences.

42

C. FOWLER AND R. LOWER

Strengthening the capacity of scientific institutions to assess products properly—in particular in developing countries—is a prerequisite for uncoerced support in some countries. USAID has provided a $15 million grant for supporting biosafety work in developing countries through a program led by the International Service for National Agricultural Research, ISNAR (ISNAR 2003). In the U.S., Congressman Dennis Kucinich (Ohio) has recently introduced a series of broad-ranging bills that address a potential framework for genetically modified plants, animals, and bacteria. They can be found at http//Thomas.loc.gov.

IV. PLANT BREEDERS’ CHOICES A. Public vs. Private Perhaps one of the first choices a plant breeder makes is to choose the type of research and employment that he or she is interested in, public or private. Frey (1996) surveyed plant breeding interests in the U.S. and found that there were 2241 scientific years devoted to plant breeding in 1994. Of these, 1499 were in private companies and, more specifically, 545 were involved in field corn breeding (with 94% of those involved in field corn breeding being in the private sector). The remainders were in public venues, such as universities and their State Agricultural Experiment Stations (SAESs) and the United States Department of Agriculture–Agriculture Research Service (USDA–ARS). Heisey et al. (2001) in Public Sector Plant Breeding in a Privatizing World flatly state: “Public sector agricultural research in general, and public plant breeding research in particular, is in trouble in both industrialized and developing nations.” They call attention to and emphasize the dramatic growth in the private sector and increase in support of private plant breeding. Further, they discuss the factors that have brought about these changes and how they affect the plant breeding world. They state that public sector plant breeding in industrialized nations is “entering into increasingly turbulent times” and a cloudy future. Declining budgets are a major problem and responsible for the bleak future. Under-investment in agricultural research is ubiquitous in both developed and developing countries and can only be explained as a consequence of short-sightedness and lack of political power of those who support such research. Significant social returns might be possible if appropriate areas of research, such as those that involve components of clear public good, are chosen for future action areas. Pinstrup-Andersen and Cohen (2000) observe that private and social rates of return on investment in agricultural research

2. POLITICS OF PLANT BREEDING

43

still exceed 20% a year, compared with long-term real interest rates of 3–5% for government borrowing. The demarcation line between public and private plant breeding is probably clearer now than in recent history. The choice is likely to revolve around several options and some of them are discussed in the following synopsis. As the roles of public and private plant breeders have shifted or evolved it is clear that private industry goals are driven and motivated by profit. Private programs have a stronger funding or support base and a more clearly defined research goal than that of public breeding programs. Private interests are directed or formulated to capture market share and make money for the investors. This posture drives companies and their plant breeders to crops that have substantial acreage or seed usages. Frey (1998) estimates that 80% of private plant breeding effort is directed toward cultivar development. The target for the private breeder is usually a cultivar that outperforms the market standard or offers some other advantage such as improved or new pest resistance or nutritional quality. (IPRs also enter into this situation and they will be discussed in another section of this paper.) Support for public breeding, meanwhile, is declining in terms of real money. This trend appears to have begun in the 1990s in the U.S. Heisey et al. (2001) and Frey (1997) reported that funding for public plant breeding research and development had not kept pace with the increasing costs of conducting breeding programs. Because of reduced availability in public support for plant breeding, these scientists have sought out both traditional and non-traditional sources of support. This usually involves grant writing, which takes considerable time and effort away from plant breeding. Public plant breeders have been ascribed numerous roles by various authors (Coffman 1998; Eberhart et al. 1998; Frey 2000; Heisey et al. 2001), but most agree that the prime role for public plant breeders is the education and training of the world’s future plant breeders. All authors suggest that public plant breeders may play important roles in developing and advancing new technologies and in the breeding of minor crops. Further, Frey discussed the need for a formal plan to promote the breeding of minor crops and made suggestions for facilitating or increasing the role for plant breeders, both public and private, in meeting the needs for food, health, fiber and other industrial uses, medicines, and environmental remediation or enhancement. The public sector plays the major role in the collection, conservation, and distribution of genetic resources. Private sector activities in long-term conservation of plant genetic resources for food and agriculture are modest (FAO 1998). Even though there are numerous new technologies available, the science and

44

C. FOWLER AND R. LOWER

art of plant breeding, whether public or private, will be critical to the development of the world’s crop needs for food, fiber, and medicine. Although most observers expect the public plant breeder to continue to have a major role in the future, we have great concern about how to increase the funding necessary to complete the task of educating and training new plant breeders and thus secure the foundation of both public and private breeding. Tracy (2004) argues that chances for increased public support are strengthened if the public is educated about the positive contributions that plant breeding plays in enhancing food security and sustainability. Further, he reviews the positive aspects of the independence of the public breeder and the roles that the public breeder can play in educating new breeders in the development of cultivars because of the long-term nature/continuity of public programs. B. GMOs or Non-GMOs “Choices” for plant breeders and their clientele are subject to new and changing rules, laws, and even viewpoints. A case in point is the following example of how one industry ventured into the biotechnology arena and then backed away due to consumer response to GMOs. The pickling cucumber processing industry in the U.S. had long sought plant resistance to nematodes and pickle worms and to soil-borne pathogens that caused considerable damage to fruit. The industry was also hampered by weed populations as it changed from hand to machine harvesting. Teams of plant breeders and other scientists screened genetic resources from all over the world, but could not find resistance to either the major nematode species or the pickle worm. After little success in solving these seemingly intractable problems via traditional breeding methods, the industry (in the early 1990s) turned to the possibilities of genetic engineering. They made contacts with the players in the biotechnology businesses to gain access to the Bt gene and gene(s) for herbicide resistance, as well as for genetic constructs that might be of assistance to plant breeders in resolving some of the problems noted above. Concomitantly, they helped support public research to develop a successful transformation system for cucumber. For a number of reasons, the biotech companies demurred and showed little interest in this processing industry’s problems. Undaunted, the processing industry continued to support biotech research, mostly on the identification and use of molecular markers for marker-aided selection to enhance the success of plant breeding, in addition to the more routine conventional plant breeding. In the late 1990s, as consumer backlash to genetically modified organisms became more pronounced and more frequent, the cucumber processing industry withdrew its financial support of research dealing

2. POLITICS OF PLANT BREEDING

45

with GMOs. Although the industry continues to be plagued by these same pest problems, it continues to seek solutions to them through pre-GMO forms of pest control. Plant breeders that work hand in hand with the processing industry have shown great flexibility and durability in their efforts to find pest resistance even as their choices have become more limited. The interactions of “conventional” plant breeders and biotech’s genetic engineers will continue to be critical and important to the success of both public and private breeding efforts to develop improved cultivars. The positive synergistic potential of conventional and biotechnological approaches is frequently overlooked as the focus has shifted to the prophesied “miracles” of biotech and the currently more tangible transfer of single genes to make a new cultivar immediately better than the donor recipient. Without the interaction of the plant breeder (as well as researchers from many other disciplines), the possibilities of failure of such a new biotech innovation are heightened, as highlighted in the review of the Flavr Savr tomato (Charles 2001). The roles of biotechnology in agriculture are an enticing and interesting subject. Biotech interacts with plant breeding and nearly all other aspects of growing, handling, and marketing a crop. The biology and breeding of a genetically modified crop is sometimes quite simple when compared to the actions necessary to comply with the safety regulations and legal implications of a genetically modified crop. Fortunately, the literature covering these subjects is common and too voluminous to begin listing citations. Thus, our treatment of the subject is at best cursory. C. Intellectual Property Protection and Rights Fretz and MacKenzie (2002) discussed the roles of both the public scientist and the administrator in regard to intellectual property in agricultural research and development. Under the U.S. system, it is the responsibility of a public institution (and its plant breeders) to look for ways and methods to deploy new discoveries for economic, social, health, and environmental benefits. Traditionally, public institutions have openly shared research findings (public goods). Indeed, this behavior was expected by government and by the tax-payer. However, over the past two or three decades the funding at public institutions for research programs has changed dramatically. Research budgets have faced new and numerous fiscal constraints and, in addition, the cost of doing research has increased. Thus, a public institution is now forced to look into ways of protecting intellectual property and to have as one of its goals the generation of returns/revenues to support its research agenda. The public institution has the moral responsibility to transfer its research products to the public. This can be accomplished in different

46

C. FOWLER AND R. LOWER

ways. The product may be dealt with as a public good and made available to all or it may be protected as intellectual property. If the latter is chosen, the form of protection may be by claiming the intellectual property only to assure that the general public has access to it (so-called “defensive” patenting), or, by protecting the invention with the intent of commercialization. The plant breeder is faced with the dilemma of determining how best to market new developments whether they are in the form of germplasm or technological innovations. Now, at nearly every institution of higher learning in the U.S., there is a technology transfer office that assists with the movement of new products into the market chain. Financial returns resulting from plant breeding inventions are usually distributed according to institutional policy and the formulas for distribution vary greatly across institutions. However, it is a concern to administrative management that the plant breeders are rewarded for their innovations and inventions, but there is also a concern that the reward does not become the principal driving force behind the research program. Knudsen et al. (2000) discuss the intellectual property protection options for plant breeding achievements in the U.S. Similar options (breeders’ rights) are available in many other industrialized nations as well. Details of the various options are also discussed. Most plant breeders, but especially those in the public sector and those that are members of small businesses or are self-employed, will have the opportunity to exercise the protection option that best fits their needs. Jondle (2003) recently addressed members of an American Society for Horticultural Science (ASHS) workshop on the patentability of naturally occurring genes. He listed protection options as follows: • • • •

Plant patents Utility patents Plant Variety Protection Certificates (“Breeders Rights”) UPOV Contract law (varies from state to state) Written contracts Material Transfer Agreements (MTAs) Restrictive use wording on bags, tags, and signed invoices Trade secrets (covered by state laws) • Trade names and trademarks Jondle emphasized the differences between utility patents (UPs) and the Plant Variety Protection Certificates (PVPCs) and explained that UPs have a broader scope than PVPCs and a very narrow research exemption that does not allow for commercialization or farmers exemption. The UP is the strongest form of intellectual property protection for cultivars,

2. POLITICS OF PLANT BREEDING

47

inbreds, and hybrids. He argues that patenting cultivars is likely to stimulate plant breeding and research and that patents provide a legal form for germplasm exchange via licensing. The use of restricted wording on bags, tags, and invoices, along with similar wording in written contracts and MTAs, allows for strong protection of genetic material. Enforcement, however, is another subject. We believe that it is fair (accurate) to say that these restrictions are frequently overlooked by both growers and breeders and suspect that violations of the wording may also occur rather frequently. The field of intellectual property rights is controversial, however, and divergent viewpoints are easy to find. Jaffé (2000) notes the uncertainty that exists concerning the level of the economic impact of utility patenting. And, the U.K.’s Council of the Royal Society, for example, has recently issued a report (2003) that called attention to the “tension” that IPRs can create between the private profit and public good, and noted their potential to “hinder the free exchange of ideas and information on which science thrives.” The debate, which began in earnest with the adoption of the Paris Convention for the Protection of Industrial Property in 1883, continues! Despite on-going disputes over the efficacy and effects of IPRs, it is obvious that plant breeders have a wide-ranging set of choices for protecting their inventions and developments. Each choice has advantages and disadvantages and it behooves plant breeders, and in some cases, their technology transfer offices, to determine which choice best fits each particular invention. There is not a “one size fits all” philosophy for inventions. Lesser and Mutschler (2002) reviewed the types of IPRs available to plant breeders in the U.S. They analyzed the entire PVP data set from 1970 to 2001 and the use of utility patents on some species from 1985 to 2001. Soybeans (19%), maize (12%), bread wheat (7.7%), cotton (6.1%), and lettuce (5.4%) account for half of the applications for PVPs. They also note that patents on maize and soybeans increased rapidly in the late 1980s (maize) and the early 1990s (maize and soybeans) and discuss the reasons for these increases (changes in types of protection available, results of legal cases, and transgenic crops). Types of protection are also influenced by reproductive biology, mechanisms of seed production, and economics of the crop. Prior to the use of PVP and UPs, plant breeders, at times both public and private, shared information and germplasm freely or with as little as a verbal agreement and a handshake. This level of cooperation was also prevalent in regional and national trials at which breeders agreed to the use of their material (trial entries) in other breeding programs. Once entered into some trials, the germplasm was freely available to other contributors. Although courtesy announcements of the use of

48

C. FOWLER AND R. LOWER

breeding material from another program were respected, they were not required. We do not know of any similar arrangements in the 21st century. The level of protection(s) now afforded to plant breeding products and the requirement at some levels of both public and private programs to use PVPCs or UPs or some type of contract minimizes the opportunity for the type of germplasm and information exchange that was available 20–30 years ago. No doubt this lost opportunity is not only one of reduced germplasm sharing and fewer improved products, but also the loss of the synergism that results from bringing more brain power to the development of a common goal—the superior genetic performer. Factor into this type of behavior the need for programs to generate their own support, and the transfer of genetic material between competing programs has slowed to at best a trickle of what it was 30 years ago. D. Looking Ahead In reviewing the literature for this effort, we were pleased at the agreement found in previous publications about the role(s) and goals of the public plant breeder and how he or she should garner support to accomplish these targets (Coffman 1998; Eberhart et al. 1998; Fretz and MacKenzie 2002; Frey 1997, 1998, 2000; Heisey et al. 2001). There is no quarrel with the view that one of the primary goals and roles of the public plant breeder is the training and education of future plant breeders. But there are fewer and fewer opportunities to acquire basic training and as Heisey et al. point out, “to the extent that plant breeding skills are not firm-specific, firms will not invest optimally in training . . .” One of the difficulties here is the downsizing of many public breeding programs because of the escalating costs associated with those programs and the lack of political commitment toward maintaining them amidst competing demands on limited public resources going to agriculture. Field and laboratory programs are both time-consuming and costly. The hands-on training that is critical to the student is lacking in many programs now, and the new plant breeder, although perhaps better educated in the sciences and with a better tool box than that of his or her predecessor, may lack the training necessary to improve and grow the crop of interest through breeding. Duvick (1998) stated that . . . a well-rounded plant breeder today must know many more kinds of science and technology than were required as recently as 10 years ago, not to mention 20 or 30 years earlier. In actuality, no one person can master all of the science and technology that is needed for efficient plant breeding for the late 1990s. At best, a well-rounded plant breeder will know

2. POLITICS OF PLANT BREEDING

49

some of the specialties, and know how to converse and collaborate with specialists in other parts of plant breeding.

The 21st century public plant breeder will be trained in how to develop and utilize new technologies that will enhance the progress of the breeding program and increase the chances of success. Likewise, he or she will likely be involved in germplasm collection, development, preservation and enhancement of the gene pool, breeding methodologies, and variety development. Although the majority of new cultivars for most major and some minor crops will result from efforts in the private sector, the public breeder will have opportunities to develop some of the major and many of the minor crop cultivars if so inclined. Referring to crops as “major” and “minor” is confusing, as the reference could refer to size of production area, nutrition contribution, value of a specific crop, or other descriptors. The private sector is interested in the economics of the crop and its future. On this basis, some so-called major crops today may be called minor in the future and vice-versa. We suspect that there is a shift in emphasis in the private sector toward the high-value market crops and away from other crops that are not as profitable. This has implications for progress in agriculture with the other crops and for the public sector. We also sense that this reality is not clearly understood or acted upon by policy makers, funding organizations, or indeed even by people in the public sector. If such a shift does occur in the private sector, a number of important crops, both major and minor, will by default fall solely into the hands of the public sector, which, unfortunately, is not adequately staffed or funded to meet the challenge. In addition, the public breeding ground will also give chances for seeking answers to seemingly intractable problems, both short-term and long-term, such as technological glitches in the lab and field. As they have in the past, public breeders can and will be expected to continue to make major contributions to genetic resources, including increased pest resistances, improvements in nutritional content and beneficial health components, greater adaptability of cultivars, and environmental preservation and stability. At a recent plant breeding workshop, Ryder (2003) expressed concerns felt by many plant breeders. He challenged the intellectual property guidelines that have been thrust upon the plant breeding community by the legal profession, the courts, and business interests. This suggests another role for plant breeders that starts with understanding those guidelines and their constraints on both public and private plant breeders. Protection of intellectual property to assure returns on investment, although a goal of private industry, is not always in the best interest of

50

C. FOWLER AND R. LOWER

the public. The opportunity to enhance and further science, and the knowledge that undergirds it, is a continual goal for all plant breeders, as well as a justifiable professional and personal reward and a tangible public good that supports future plant breeding. Frey (1997, 1998, 2000) offers several initiatives to keep private and public sectors interactive and cooperative as they solidify and make positive the future for plant breeding. For example, the need for cooperative ventures between sectors to insure that the world does not forget about the roles of minor crops in the supply of food for healthy diets, nutrition, and antioxidant and health-promoting foods. Many of the International Agricultural Research Centers (IARCs) have long advocated for the need to insert minor crops into the monoculture cropping systems that exist in many developing countries of the world. The minor crops offer much in the way of food and diet diversity, crop rotation, soil, and environmental quality and maintenance. (See: http://www.avrdc.org for information about the importance of vegetable crops, including for instance the role of leafy vegetables in the diet.) Yet, relatively little investment is made in these crops. The potential of many minor crops has not been tapped at all, as quite a few have probably never been the subject of a formal, scientific plant breeding program. Even crops of major economic and nutritional importance (e.g., yams, bananas/plantains, minor millets) receive surprisingly little investment and thus may be worked with by no more than a half-dozen plant breeders around the world, if that. Public-private sector interactions ebb and flow with opportunities. However, some universities and their SAESs, i.e., Cornell and Texas A&M University, have fostered solid relationships between their public plant breeders and the private sector and these plant breeders will continue to enhance contributions to the health of the populace and the health and quality of the environment. Perhaps these models deserve consideration by other institutions as they promote increased cooperative efforts between sectors and generate support for public efforts. The trend line for the number of practicing public plant breeders is going down, while the numbers of private breeders are increasing. Since the late 1970s these trends have been accompanied by an increase in mergers between seed companies and buyouts of one by another. In this same timeframe the vertical integration of biotech companies with seed companies that utilize and market biotech intellectual property has been common. We believe that this decrease in total number of seed companies has led to decreased coverage of many crop species by plant breeders, most of which are among the so-called minor crops. Numerous vegetable crops were bred by no fewer than 20 seed companies in the

2. POLITICS OF PLANT BREEDING

51

1970s and today the number for many crops is no more than three. Contributions of these crops to health and nutrition, to increased sustainability, to local niche markets, and to special economic considerations may result in a re-focusing of breeding efforts and lead to increased improvement efforts by plant breeders in the future. The public breeder will be called upon to respond to the needs of growers of peripheral or “orphan” crops, some of which are actually very important when judged by virtually any set of criteria. New genetic technologies might create opportunities for substantial advancement in the improvement of such crops currently experiencing “under-investment” in research and development (Naylor et al. 2004). In short, new opportunities for the public sector will surely emerge, sparked by the presence of new tools and new demands created by changes in production and environmental systems. But, will the public sector be in a position to respond? We are deeply troubled by several related phenomena and their effects on both public and private plant breeding. First, the lack of support for public sector plant breeding and the need for researchers to generate funding via IPRs. Second, the belief held by many that the public sector should not compete with the private sector. Finally, the growing necessity that the public sector must behave in a quasi-commercial manner in order to secure funding for its own survival. Increasingly, the public sector is expected to generate support, but usually only on crops on which there is little or no money to be made. These situations threaten the existence of public breeders and their historic mission. The decline in the number of plant breeders in the public sector is accompanied by a changing education and training pattern at the university level. As noted previously, the plant breeding graduates of this millennium will likely not have the same experiences of growing and knowing their crops of interest that were standard for their major professors trained only a few decades earlier. New high-profile courses at the molecular level have replaced specialized courses in plant growth, structure, and supporting courses in associated areas such as soils, pathology, entomology, and food science. Tracy (2004) points out that today’s new plant breeders are “weaker, in agricultural sciences, quantitative thinking, whole plant biology, and selection theory.” This situation reinforces the critical role—and the challenge—of the public institution in the education and training of new plant breeders. Long-term breeding programs, both field and laboratory, are critical to the training of students. However, the costs of these programs continue to grow and many of them are downsized annually to meet budget restraints. University administrations have not been convinced of the deleterious effects of changes to and constraints on plant breeding programs. Funding allocations remain marginal

52

C. FOWLER AND R. LOWER

when compared to the size of the job. Despite current trends in funding, plant breeders will continue to be the link between the new technologies and the acceptance of a final product for agricultural growers, processors, and consumers. Plant breeders need not fear about their future livelihood as a profession, but they do need to tell their story to their administrators and other supporters in a more convincing manner. We are hopeful that the pendulum swing will reverse its direction and the number of public plant breeders will rebound. We trust that the public demand for agricultural sustainability, food security, and safety will eventually create political pressures that will support public sector efforts and lead to a larger and vibrant cadre of public plant breeders to serve the needs of the greater system. The plant breeder, involved as he or she is in one of the world’s most important endeavors, is never far removed from the political context of the work. At no point since the rediscovery of Mendel’s laws has plant breeding had so many tools and opportunities at its disposal. Yet, never have so many constraints been placed on the art, science, and practice of this ancient and noble profession. Whether, how, and to what end plant breeders will be able to continue to serve society productively will be answered in large part by the accommodation that is being and will be made between plant breeding and politics.

LITERATURE CITED Ag Biotechnology Implementation Task Force. 2003. Survey http://www.cals.ncsu.edu/ agcomm/biotech/ Agres, J. 2003. Biodiversity treaty called disastrous. The Scientist. Sept. 10. www.biomedcentral.com/news/20030910/03/ Anderson, E., and W. Brown. 1952a. The history of common maize varieties of the United States corn belt. Agr. Hist. 26:2–8. Anderson, E., and W. Brown. 1952b. Origin of corn belt maize and its genetic significance. In: J. Gowan (ed.), Heterosis. Iowa State College, Ames. p. 124–148. Brockway, L. 1979. Science and colonial expansion: The role of the British Royal Botanic Gardens. Academic Press, New York. Burns, T., and R. Ueberhorst. 1988. Creative democracy: Systematic conflict resolution and policymaking in a world of high science and technology. Praeger, New York. Castagnera, J., C. Fine, and A. Belfiore. 2002. Protecting intellectual capital in the new century: Are universities prepared? Duke Law Technol. Rev. (June). Charles, D. 2001. Lords of the Harvest. Pereus Publishing, Cambridge, MA 02142. p. 126–148. Cleaver, H. 1972. Contradiction of the green revolution. American Economic Review, Vol. 62, Number 2. p. 177–186. Coffman, W. R. 1998. Future of plant breeding in public institutions. Corn and Sorghum Res. Conf., Am. Seed Trade Assoc., Chicago.

2. POLITICS OF PLANT BREEDING

53

Committee on Registration of the Peninsula Horticultural Society. 1891. Registration of new fruits. The American Garden, Vol. 12. No 6. p. 338–339. Conko, G., and C. Prakash. 2002. The attack on plant biotechnology. p. 179–218. In: R. Bailey (ed.), Global warming and other eco-myths. Prima Publishing/Random House, New York. Correa, C. 2003. The access regime and the implementation of the FAO treaty on plant genetic resources for food and agriculture in the Andean group countries. Unpubl. manuscript for a study sponsored by SAREC (Sweden) and executed by the International Potato Center. Council of the Royal Society. 2003. Keeping science open: The effects of intellectual property policy on the conduct of science. The Royal Society, London. Crosby, A. 1986. Ecological imperialism: The biological expansion of Europe, 900–1900. Cambridge Univ. Press, Cambridge, UK. Duvick, D. N. 1998. Future sources of plant breeders for industry. Production and Res. Conf., Texas Seed Trade Assoc., College Station. Eberhart, S. A., H. L. Shands, W. Collins, and R. L. Lower. 1998. Intellectual property rights III. Global genetic resources: Access and property rights. CSSA Misc. Pub., Madison, WI. Elder, G. 1921. Seed marketing hints for the farmer. Farmers’ Bul. 1232. USDA. Government Printing Office, Washington, DC. Falcon, W., and C. Fowler. 2002. Carving up the commons: Emergence of a new international regime for germplasm development and transfer. Food Policy 27:197–222. Fitzgerald, D. 1990. The business of breeding: Hybrid corn in Illinois, 1890–1940. Cornell Univ. Press, Ithaca, NY. Food and Agriculture Organization of the United Nations. 1998. The state of the world’s plant genetic resources for food and agriculture. FAO, Rome. Food and Agriculture Organization of the United Nations. 2001. International treaty on plant genetic resources for food and agriculture. FAO, Rome. Foster, M., P. Berry, and J. Hogan. 2003. Market access issues for GM products: Implications for Australia. Australian Bureau of Agricultural and Resource Economics Report for the Dept. Agr. Fisheries and Forestry, Canberra, Australia. Fowler, C. 1994. Unnatural selection: Technology, politics, and plant evolution. Gordon and Breach Science Publ., Yverdon, Switzerland. Fowler, C. 1998. Rights and responsibilities: Linking conservation, utilization, and sharing of benefits of plant genetic resources. Intellectual property rights. III. Global genetic resources: Access and property rights. Crop Sci. Soc. Am., Am. Soc. Agron., Madison. Fowler, C. 2000. The Plant Patent Act of 1930: A sociological history of its creation. J. U.S. Patent and Trademark Office Soc. 82:621–644. Fretz, T. A., and D. R. MacKenzie. 2002. The public university, intellectual property and agricultural R&D. In: Max Rothschild and Scott Newman (eds.), Intellectual property rights in animal breeding and genetics. CAB Int., Oxon, UK. Frey, K. J. 1996. National plant breeding study. I. Human and financial resources devoted to plant breeding research and development in the United States in 1994. Iowa Agr. Home Econ. Special Rep. 98, Iowa State Univ., Ames. Frey, K. J. 1997. National plant breeding study. II. National plan for promoting breeding programs for minor crops in the U.S. Iowa Agr. Home Econ. Special Rep. 100, Iowa State Univ., Ames. Frey, K. J. 1998. National plant breeding study. III. National plan for genepool enrichment of U.S. crops. Iowa Agr. Home Econ. Special Rep. 101, Iowa State Univ., Ames. Frey, K. J. 2000. National plant breeding study. IV. Future priorities for plant breeding. Iowa Agr. Home Econ. Special Rep. 102, Iowa State Univ., Ames.

54

C. FOWLER AND R. LOWER

Fuglie, K. O., and D. E. Schimmelpfennig. 2000. Public–private collaboration in agricultural research. Iowa State Univ. Press, Ames. George, S. 1977. How the other half dies: The real reasons for world hunger. Allanheld Osmun, Montclair, NJ. Goodman, M. 2002. New sources of germplasm: Lines, transgenes, and breeders. In: J. Martinez (ed.), Simposio El Fitoramiento ante los Avances Científicos y Tecnológicos (2 Sept.), Saltillo, Mexico. GRAIN. 1995. Towards a biodiversity community rights regime. Genetic Resources Action Int., Barcelona. GRAIN. 2002. Biopiracy by another name? A Critique of the FAO-CGIAR Trusteeship System. Seedling (Oct.) www.grain.org/seedling/seed-02-10-2-en.cfm. Griffin, K. 1972. The green revolution: An economic analysis. United Nations Research Institute for Social Development, Geneva. Haughton, C. 1979. Green immigrants. Harcourt Brace Jovanovich, New York. Heisey, P., Srinivasan, C., and C. Thirtle. 2001. Public sector plant breeding in a privatizing world. Agr. Inform. Bul. 772. Resource Economics Division, Economic Research Service, U.S. Department of Agriculture, Washington, DC. Heller, M., and R. Eisenberg. 1998. Can patents deter innovation? The anticommons in biomedical research. Science 280:698–701. Hightower, J. 1973. Hard tomatoes, hard times. Schenkman Publ., Cambridge. ISNAR (International Service for National Agricultural Research). 2003. Group Awarded $15 Million for Work on Strategies and Policies. Press Release. (June 9) Jaffé, A. 2000. The U.S. Patent System in transition: Policy innovation and the innovation process. Res. Policy 29. Jondle, R. 2003. How to apply for a patent related to a naturally occurring gene. HortScience 38:142 (Abstr.). Kloppenburg, J. 1988. First the seed: The political economy of plant biotechnology, 1492–2000. Cambridge Univ. Press, Cambridge. Knudsen, M. K., R. L. Lower, and R. Jones. 2000. State agricultural stations and intellectual property rights. p. 199–215. In: K. O. Fuglie and D. E. Schimmelpfennig (eds.), Public-private collaboration in agricultural research. Iowa State Univ. Press, Ames. Lappé, F., J. Collins, and C. Fowler. 1977. Food first: Beyond the myth of scarcity. Houghton Mifflin, New York. Lesser, W., and M. Mutschler. 2002. Lessons from the patenting of plants. In: M. Rothschild and S. Newman (eds.), Intellectual property rights in animal breeding and genetics. CAB Int., Oxon, UK. Michaels, P. 2003. Is Europe returning to the dark ages? Cato Institute. (15 February). http://www.cato.org/dailys/02-15-03.html Mooney, P. 1979. Seeds of the earth: Private or public resource? Canadian Council for Int. Co-operation and the Int. Coalition for Development Action, Ottawa. National Academy of Sciences (NAS). 1987. Introduction of recombinant DNA-engineered organisms into the environment: Key issues. National Academy Press, Washington, DC. Naylor, R., W. Falcon, R. Goodman, M. Jahn, T. Sengooba, H. Tefera, and R. Nelson. 2004. Biotechnology in the developing world: A case for increased investments in orphan crops. Food Policy (forthcoming). Palacios, X. F. 1998. Contribution to the estimation of countries’ interdependence in the area of plant genetic resources. Background Study Paper 7, Rev. 1. FAO Commission on Genetic Resources for Food and Agriculture, Rome. Palmer, I. 1972a. Food and the new agricultural technology. United Nations Research Institute for Social Development, Geneva.

2. POLITICS OF PLANT BREEDING

55

Palmer, I. 1972b. Science and agricultural production. United Nations Research Institute for Social Development, Geneva. Pardey, P., and N. Beintema. 2001. Slow magic: Agricultural R & D a century after Mendel. IFPRI Food Policy Statement 36. International Food Policy Research Institute, Washington, DC. Pinstrup-Andersen, P., and M. Cohen. 2000. Biotechnology and the CGIAR. Conf. sustainable agriculture in the next millennium: The impact of modern biotechnology on developing countries. Brussels, 28–31 May. Plucknett, D., N. Smith, J. T. Williams, and M. Anishetty. 1987. Gene banks and the world’s food. Princeton Univ. Press, Princeton, NJ. Price, S. C. 1999. Public and private plant breeding. Nature Biotechnol. 17(10):938. Rader, R. 2000. Trends in biotechnology patenting. Am. Chem. Soc. Natl. Meeting. Washington, DC. Ragan, W. 1926. Nomenclature of the apple: A catalogue of the known varieties referred to in American publications from 1804 to 1904. Bureau of Plant Industry, Bul. 56. Government Printing Office, Washington, DC. Resnik, D. 2003. A biotechnology patent pool: An idea whose time has come? J. Philo. Sci. Law 3(Jan.). http://www.psljournal.com/archives/papers/biotechPatent.cfm Rothschild, M., and S. Newman. 2002. Intellectual property rights in animal breeding and genetics. CAB Int., Oxon, UK. Ryder, E. J. 2003. Nature’s gifts: Are they intellectual property? HortScience 38:742 (Abstr.). Smith, S. 1997. Farmers’ rights and benefit sharing. Background Paper for the Intellectual Property Rights III Symp. Washington (June 4–6). Stover, J. 1975. History of the Illinois Central Railroad. Macmillan Publ., New York. Taylor, M. 2003. Rethinking US leadership in food biotechnology. Nat. Biotechnol. 21(8). Taylor, W. 1901. The influence of refrigeration on the fruit industry. Yearbook of Agriculture, 1900. U.S. Government Printing Office/USDA, Washington, DC. Tracy, W. F. 2004. Proceedings of the Seeds and Breeds Summit, Washington, DC. September 8–9, 2003. Published by the Rural Advancement–USA, Pittsboro, NC. In press. Tripp, R. (ed.). 1997. New seed and old laws: Regulatory reform and the diversification of national seed systems. Intermediate Technology Publications, London. USDA. 1918. Seed reporter. 2(2) (August 10). Bureau of Markets. Government Printing Office, Washington, DC. USDA/USTR (United States Department of Agriculture, United States Trade Representative). 2003. U.S. and cooperating countries file WTO case against EU moratorium on biotech foods and crops. Press Release. Washington, DC (May 13). Winner, L. 1987. Autonomous technology: Technics-out-of-control as a theme in political thought. MIT Press, Cambridge. Wolf, H., and R. Wolf. 1936. Rubber: A story of glory and greed. Covici Friede, New York.

3 Doubled Haploids in Genetics and Plant Breeding* Brian P. Forster and William T.B. Thomas Scottish Crop Research Institute, Invergowrie, Dundee, DD2 5DA, Scotland, UK

I. INTRODUCTION II. DOUBLED HAPLOID TECHNOLOGY III. DOUBLED HAPLOID POPULATIONS IN GENETICS A. Classic Studies B. Early Studies on Doubled Haploid Populations C. Bulked Segregant Analysis D. Genetic Maps E. Mapping of Quantitative Traits F. Genomics IV. DOUBLED HAPLOIDS IN BREEDING A. Early Breeding Attempts B. Elite Crossing C. Comparison of Breeding Methods D. Cultivar Releases E. Backcross Conversion V. PROSPECTS LITERATURE CITED

I. INTRODUCTION Doubled haploids (DHs) are produced from haploids. Haploid cells occur naturally in the gametophytic phases of higher plants, in their ovules and pollen. It is the cells of these tissues that are the main targets for doubled haploidy. By manipulating the environment of the gametic cells it is possible to divert development to produce embryos rather than mature pollen grains or ovules. The microspores of developing pollen grains are *Acknowledgments: The Scottish Crop Research Institute receives grant-in-aid from the Scottish Executive Environment and Rural Affairs Department. The authors are also grateful to members of the EU COST Action 851 “Gametic cells and molecular breeding for crop improvement” for valuable comment. Plant Breeding Reviews, Volume 25. Edited by Jules Janick. ISBN 0471666939 © 2005 John Wiley & Sons, Inc. 57

58

B. FORSTER AND W. THOMAS

of particular interest since plants generally produce pollen to excess. Microspore embryogenesis therefore has the potential to generate several hundred DH plants from a single anther. Induced or spontaneous chromosome doubling of haploid embryogenic cells produces doubled haploid plants. These are completely homozygous, true breeding, and consequently of great importance in genetics and plant breeding. The relevance of doubled haploids (DHs) to plant breeding has increased markedly in recent years owing to the expansion of application to over 200 species and the development of more reliable protocols. Doubled haploids can be exploited not only to produce new cultivars, but also to construct genetic maps, locate genes of agronomic and economic importance, identify markers for trait selection, and to increase plant breeding efficiency. The benefits accrue from the efficiency of DH technology in producing completely homozygous lines, vital resources in plant breeding and genetics. The potential of doubled haploidy has been the subject of various reviews (Choo et al. 1985; Dunwell 1985; Kasha et al. 1989; Pickering and Devaux 1992; Touraev et al. 2001; Maluszynski et al. 2003a). The recent rapid expansion of doubled haploid technology, which now includes over 200 plant species (Maluszynski et al. 2003b) sets the stage for greater DH deployment in a range of species and disciplines. We describe the exploitation of DHs from classic genetic studies to contemporary plant breeding and attempt to gauge current and future impact.

II. DOUBLED HAPLOID TECHNOLOGY Doubled haploids can be produced via in vivo and in vitro systems. Haploid embryos are produced in vivo by parthenogenesis, pseudogamy, or chromosome elimination after wide crossing. The haploid embryo is rescued, cultured, and chromosome-doubling produces doubled haploids. The in vitro methods include gynogenesis (ovary and flower culture) and androgenesis (anther and microspore culture). Methods can be complex and exacting. Critical stages include: growing conditions, pre-treatment of floral parts (before and after collection of material for culture), physiology of donor plants, genotype, media composition, incubation, regeneration, and chromosome doubling (Kasha and Reinbergs 1981; Pickering and Devaux 1992; Jain et al. 1996). Androgenesis, where available, is the preferred method (Sopory and Munshi 1996). In 1922, Blakeslee et al. reported the finding of a haploid plant in Jimson weed, Datura stramonium. It took another 40 years to develop laboratory methods for the routine production of Datura haploids via embryogenesis in cultured anthers (Guha and Maheswari 1964, 1966). This pioneering work led to the testing of many species, but responses to the protocols used have been extremely varied across the range of

3. DOUBLED HAPLOIDS IN GENETICS AND PLANT BREEDING

59

species. Tobacco, rapeseed, and barley are currently among the most responsive species, and although these are considered model species for DH production, response is further dependent on genotype. Legumes remain a particularly difficult group. There has been a progressive trend in reclassification from recalcitrant to responsive to highly responsive. Barley, for example, was considered recalcitrant to androgenetic protocols before the patented incorporation of maltose in anther/microspore medium (Hunter 1987), which paved the way to large-scale commercial application. A typical scheme for barley anther culture is given in Fig. 3.1. Legumes have been a notoriously difficult group, but in 1994 Ormerod and Caligari reported successful anther and microspore culture of lupin. Potato moved from recalcitrant to responsive in anther culture (Rokka et al. 1996) and, more recently, improved microspore culture methods have been described for wheat (Liu et al. 2002). Problems in doubled haploid production vary among species. For most species androgenesis is the preferred route, but for some (e.g., onion), the only practical method is gynogenesis. For large trees there is the practical problem of controlling the donor plant environment and materials for culture (flower buds) are restricted to outdoor collections during spring months. A frequent problem in cereals and other grasses is genotype dependency; many genotypes regenerate high frequencies of albino plants from cultured tissues (e.g., Holme et al. 1999). The production of doubled haploids using wide crossing can circumvent many of these problems. In barley (Hordeum vulgare), haploids can be produced by wide crossing with the related species H. bulbosum, fertilization is effected, but during the early stages of seed development the H. bulbosum chromosomes are eliminated leaving a haploid embryo. The embryos can be rescued and cultured, and plants derived from them artificially doubled (Devaux 2003; Hayes et al. 2003). The bulbosum technique has been very successful in producing DH cultivars (the majority of barley cultivars, 74/115, listed in the web-site http://www.scri.sari.ac.uk/assoc/cost851/Default.htm have been produced by this method). Wide crossing is also common in haploid/doubled haploid production in bread wheat (Inagaki 2003), macaroni wheat (Jauhar 2003), and triticale (We ¸dzony 2003); in these protocols maize is commonly used as a pollinator. Pollination with maize is also successful in oat haploid production (Rines 2003). There are, however, disadvantages in wide crossing: it is inefficient when compared with a responsive anther/microspore method, more crossing is required, specialized growing conditions are needed to grow both the target (female) and inducer (pollinator) species, haploid embryos must be rescued and cultured before they die, and at some stage artificial chromosome doubling (often involving toxic chemicals) is required. Table 3.1 lists species in which doubled haploidy is an active area in European research and breeding. We have tried to order the species

60

B. FORSTER AND W. THOMAS

Doubled haploid production in a barley cross via anther culture

A

2

1 B

3

4

2

1

3

C

4 D 1

1

2 2

3 Fig. 3.1. Doubled haploid production in barley via anther culture. There are four basic components in the production of a barley DH population. A. Crossing: typically the DH population is derived from a cross; this can be done at any generation. Crossing normally consists of emasculation of the female parent (1), selection and preparation of a male parent (dehiscing ear, 2), controlled pollination (3), and protection (4). B. Donor plant production: In order to save time, developing seed from crosses can be extracted from ears (1) and immature embryos excised and cultured (2). The progeny is grown out in controlled environment conditions (3) that maximize the harvest of anthers containing responsive microspores. Spikes are harvested when developing microspores are at the uni-nucleate stage (4). C. Anther culture: anthers (or ears) are pre-treated by either cold or starvation treatments (1) and then transferred to specialized media to induce embryogenesis of the microspores (2). The green shoots that arise are transferred to another medium to promote root growth (3). D. Bulking of DH lines: rooted plantlets are transferred to pots, usually in a glasshouse (1), where they are grown on to maturity. In barley, 60–90% of plants are spontaneously doubled and the production of excess plants precludes the need for chemicalinduced doubling, though this is required in other species. Sufficient seed can be produced in the glasshouse for field trial evaluation in plots in the next generation (2).

Table 3.1.

Current status of doubled haploid production in some target species.

Species Common name (Latin name) Tobacco (Nicotiana tabacum) Rapeseed (Brassica napus) Cauliflower, broccoli, Brussels sprouts, cabbage, etc. (Brassica oleracea) Barley (Hordeum vulgare) Wheat (Triticum aestivum) Maize (Zea mays) Triticale (×Triticosecale) Eggplant (Solanum melongena) Rice (Oryza sativa) Pepper (Capsicum annuum) Cucumber (Cucumis sativus) Onion (Allium cepa) 61

Production capacity (no. plants produced)

Cultivars developed via doubled haploidy

>10,000

Yes

Genotype

>10,000

Yes

>10,000

Yes

Genotype, germination of DH embryos, polyploids produced Genotype, germination of DH embryos, polyploids produced

>10,000

Yes

>10,000

Yes

>10,000

?

Anther culture, microspore culture, wide crossing Anther culture

>10,000

?

>5,000

Yes, F1 hybrids

Anther culture, haploid somaclone Anther culture

>1,000

Yes

>1,000

Yes, F1 hybrids

Genotype, embryo germination

Gynogenesis

>1,000

Yes, F1 hybrids

Genotype, inbreeding depression

Flower culture, gynogenesis

>1,000

No

Genotype, inbreeding depression, low fertility, chromosome doubling

Common methods Microspore culture, anther culture Microspore culture, anther culture Microspore culture, anther culture

Anther culture, microspore culture, wide crossing Anther culture, microspore culture, wide crossing Inducer lines

Major problems

Genotype, albinism, laboratory specificity Genotype, chromosome doubling, albinism Chromosome doubling Genotype, albinism, chromosome doubling Chromosome doubling, genotype Genotype

(continued )

62

Table 3.1.

Continued

Species Common name (Latin name) Turnip rape (Brassica rapa) Mustard (Brassica juncea) Ethiopian mustard (Brassica. carinata) Rye (Secale cereale) Oat (Avena sativa) Ryegrass (Lolium perenne) Potato (Solanum tuberosum) Sugar beet (Beta vulgaris) Asparagus (Asparagus officinalis) Citrus/clementine (Citrus spp.) Flax (Linum usitatissimum) Apple (Malus spp.) Cork Oak (Quercus suber) Aspen (Populus spp.)

Production capacity (no. plants produced)

Cultivars developed via doubled haploidy

Microspore culture, anther culture Microspore culture, anther culture Microspore culture, anther culture Anther culture

>1,000

Yes

Genotype

>1,000

Yes

Genotype

?

No

Genotype, albinism

42°C) at Bhavanisagar (Tamil Nadu, India), Reddy and Stenhouse (1994, 1996) reported that A1 was more stable for maintaining male sterility than others, whereas A3 stability was greater than that of A2 and A4, and that of A2 better than that of A4. The tapetum was intact and pollen was sterile in A2 male-sterile lines in winter (< 10°C), while partial or complete degeneration of tapetum and pollen grains were fertile in summer (> 42°C), indicating the unstable nature of male sterility in the A2 CMS system (Devi and Murthy 1993). At Patancheru, the low temperature-induced female sterility in A1 CMS female lines was similar to the line 296A, which indicated that sterility may be reduced significantly by using their non-parental single cross F1 male-sterile lines (Reddy 1992). In the sorghum breeding program at ICRISAT, the frequency of maintainer lines observed in A1 (Table 6.4)

154

B. REDDY, S. RAMESH, AND R. ORTIZ

Table 6.4. Maintainers and restoration frequency in sorghum A1 cytoplasm in rainy and post-rainy seasons at Patancheru (Andhra Pradesh, India). Frequency Season (2002)

A1 line

Total tested

Maintainers

Restorers

Rainy

ICSA 56 ICSA 84 ICSA 101 CK 60 A Total

75 66 87 49 277

0.71 0.83 0.84 0.55 0.75

0.29 0.17 0.16 0.45 0.25

Post-rainy

ICSA 1 ICSA 9 ICSA 101 ICSA 88005 Total

39 39 200 21 299

0.62 0.87 0.95 0.90 0.89

0.38 0.13 0.06 0.10 0.11

and A2 CMS systems (Table 6.5) was higher in the post-rainy season (< 10°C) than in the rainy season (Reddy et al. 2003). However, Indian researchers have reported higher fertility restoration in the A2 CMS system in the post-rainy season than in the rainy season (U.R. Murthy, pers. comm.). Thus, stability of the expression of male sterility may vary with the temperature as well as type of cytoplasm. Research involving the same CMS lines in both seasons may provide a better understanding of the stability of different CMS systems in different seasons. Table 6.5. Maintainers and restoration frequency in sorghum A2 cytoplasm in rainy and post-rainy seasons at Patancheru (Andhra Pradesh, India). Frequency A2 line

Total tested

Maintainers

Restorers

Rainy 1999

MR 750 ICSA 94003 Total

130 140 270

0.62 0.63 0.62

0.38 0.37 0.38

Post-rainy 1999

MR 750 ICSA 88004 ICSA 94001

19 110 20

0.47 0.45 0.85

0.53 0.55 0.15

Post-rainy 2000

ICSA 38 ICSA 743 ICSA 88001 Total

133 72 34 388

0.97 1.00 0.85 0.79

0.03 0.00 0.15 0.21

Season

6. GENETIC AND CYTOPLASMIC-NUCLEAR MALE STERILITY IN SORGHUM

155

B. Effect of CMS Systems on Economic Traits The observed frequency of segregation for tall and dwarf plants in crosses of two dwarf isocytoplasmic lines carrying A1 cytoplasm and two tall tropical landraces (IS 2317 and IS 35613) confirmed that height was controlled by four recessive non-linked genes (Murthy 1986). However, in crosses between dwarf isocytoplasmic lines of A2 cytoplasm and two landraces, the segregation pattern of dwarf and tall deviated significantly from the four gene theory, indicating an effect of the A2 cytoplasm on plant height. A comparative assessment of five pairs of sorghum iso-nuclear A1 and A2 CMS lines in Mexico revealed that CMS did not have any effect on days to flowering (Table 6.6) (Williams-Alanis and Rodriguez-Herrera 1992). Similarly, considerable variation was observed at the ICRISAT sorghum breeding program between the available male-sterile lines and maintainer lines in the A1 CMS system for flowering. In the early group, a few A-lines tended to be late by a day or two, but in the medium and late maturity groups, A-lines tended to be significantly late in flowering, and there was a tendency of increased delay in flowering in A-lines with the increased maturity period (Table 6.7). B-lines had more open panicles than those from A-lines. Rodriguez-Herrera et al. (1993) also reported a delay in flowering of A- lines (A2) compared to their maintainer (B) counterparts. Spikelet damage and adult emergence of midges was significantly lower on midge-resistant B-lines (PM 7061 and PM 7068) than their corresponding A-lines, and vice versa in the midge-susceptible parental lines (296A and ICSA 42) (Sharma et al. 1994; Sharma 2001). At Patancheru, the maintainer lines (B) flowered early by one or two days and had more open panicles than those of their A-lines. Further, A1 cytoplasm was more susceptible to shoot fly than the maintainer line cytoplasm, while the reverse was true for stem borer resistance (Reddy Table 6.6. Days to anthesis in sorghum iso-nuclear CMS lines A1 and A2 at three planting dates in Mexico. Source: Williams-Alanis and Rodriguez-Herrera (1992). 20 February

8 March

23 March

Lines

A1

A2

A1

A2

A1

A2

LRB-1104A LRB-1102A LRB-1110A LRB-1106A E-15A Means Significance

84 83 84 79 87 83 NS

83 82 85 82 86 83

80 80 79 78 81 79 NS

79 76 82 80 79 79

76 74 78 73 76 75 NS

77 75 76 76 78 76

156

B. REDDY, S. RAMESH, AND R. ORTIZ

Table 6.7. Frequency of sorghum male sterile and maintainer lines differing in days to 50% flowering at Patancheru (Andhra Pradesh, India).

Difference in days (A-B)

Early maturity group (< 67 days)

Medium maturity group (67–74 days)

Late maturity group (>74 days)

–2 –1 0 1 2 3 Total number tested

0.00 0.00 0.74 0.24 0.03 0.00 34

0.02 0.08 0.41 0.40 0.09 0.00 108

0.04 0.08 0.31 0.35 0.19 0.04 26

et al. 2003). This finding has significance in developing shoot fly and stem borer resistant hybrids. Evaluation of five pairs of sorghum isonuclear A1 and A2 CMS lines in four locations of Tamaulipas (Mexico), viz. Rio Bravo (irrigated), El Tapo (drought), El Canelo (drought), and Guelatao (drought), during the fall summer season of 1992 indicated significant differences between A1 and A2 CMS lines for grain yield only in drought conditions (RodriguezHerrera et al. 1993) (Table 6.8). However, no significant differences were Table 6.8. Economic traits as influenced by iso-nuclear A1 and A2 sorghum CMS lines evaluated in four locations in Tamaulipas (Mexico). Source: Rodriguez-Herrera et al. (1993).

Grain yieldz (t ha–1)

Days to floweringz

Plant heightz (cm)

Panicle lengthz (cm)

Panicle exertionz (cm)

Rio Bravo A1 A2

NS NS

82 ab 84 a

NS NS

30 ab 33 a

NS NS

El Tapon A1 A2

2.3 b 2.1 b

75 b 78 a

137 b 135 b

NS NS

15 ab 11 b

El Canelo A1 A2

1.3 b 1.5 a

83 a 83 a

NS NS

27 b 30 a

NS NS

Guelatao A1 A2

0.34 a 0.25 b

82 ab 83 a

NS NS

NS NS

NS NS

Location Isogenic lines

z

Means followed by same letter at each location are not significantly different; NS = non-significant.

6. GENETIC AND CYTOPLASMIC-NUCLEAR MALE STERILITY IN SORGHUM

157

Table 6.9. Grain yield and agronomic characteristics of sorghum isonuclear hybrids (in A1 and A2 CMS backgrounds) evaluated in ten environments in Northern Mexico. Source: Williams-Alanis et al. (1993). Characteristics Grain yield (kg ha–1) Days to flowering Plant height (m) Panicle length (cm) Panicle exertion (cm) z

A1

A2

CV (%)

4195 az 76 b 1.4 a 30 a 17 a

4210 a 77 a 1.4 a 30 a 17 a

23 4 7 15 33

Means in rows followed by same letter are not significantly different

found between A1 and A2 CMS lines for plant height, panicle length, and panicle exertion. In yet another study using 32 isonuclear A1 and A2 CMS line-based hybrids evaluated in 10 environments in Northern Mexico during the fall–winter season of 1990, 1991, and 1993 under irrigated and dry conditions, Williams-Alanis et al. (1993) reported an absence of significant differences between A1 and A2 CMS lines-based hybrids for grain yield, plant height, panicle length, and panicle exertion (Table 6.9). Evaluation of two sets of 36 hybrids obtained by crossing two different sets of six A1 and A2 isonuclear CMS lines with common three dual restorers at Patancheru during the post-rainy season of 2001 and the rainy season of 2002 indicated an absence of significant differences between A1 and A2 CMS systems for mean performance for traits such as days to 50% flowering, plant height and grain yield, lodging resistance and aphid resistance (Tables 6.10 and 6.11). Although hybrids based on Table 6.10. Effect of A1 and A2 CMS systems on mean performance for grain yield and other economic traits in sorghum during post-rainy season of 2001 at Patancheru (Andhra Pradesh, India). Trait

A1

A2

Difference

Significance

Days to 50% flowering Plant height (m) Grain yield (t ha–1) Plant agronomic performancez Lodging resistancey Aphid resistancex Seed set under open pollination Seed set upon selfing

69.20 2.11 7.01 2.41 1.98 3.11 92.04 84.63

69.37 2.10 6.80 2.65 2.04 3.09 91.20 78.80

–0.17 0.01 0.21 –0.24 0.06 0.09 0.84 5.83

NS NS NS * NS NS * *

* and NS indicate significant at P = 0.05 and non-significant 1–5 scale where 1 = good and 5 = poor y 1–5 scale where 1 = less than 10% plants lodged and 5 = more than 80% plants lodged x 1–5 scale where 1 = leaf free from aphids damage and 5 = more than 60% leaf area damaged z

158

B. REDDY, S. RAMESH, AND R. ORTIZ

Table 6.11. Effect of A1 and A2 CMS systems on mean performance for sorghum grain yield and other economic traits in the rainy season of 2002 at Patancheru (Andhra Pradesh, India). Trait

A1

A2

Difference

Significance

Days to 50% flowering Plant height (m) Plant agronomic performancez Grain yield (t ha–1)

68.44 2.58 1.65 5.71

68.59 2.55 1.78 5.72

–0.15 0.05 –0.13 0.01

NS NS * NS

*and NS indicate significant at P = 0.05 and non-significant 1–5 scale where 1 = good and 5 = poor

z

A2 cytoplasm showed superior plant agronomic performance, A1 based hybrids excelled in seed set under open pollination as well as selfing. A comparative evaluation of A1 and A3 cytoplasm-based iso-nuclear sorghum-sudan grass hybrids at the University of Nebraska field laboratory, Ithaca during 1990 and 1991 by Pedersen and Toy (1997) revealed that cytoplasm had no effect on days 50% flowering, plant height, dry matter of forage yield, in vitro dry matter digestibility and protein content (Table 6.12). However, while fertility restoration was equivalent in A1- and A3-based hybrids, it was significantly lower in a few A3-based hybrids. Recently, by evaluating a set of 12 isonuclear hybrids each in A1, A2, and A3 cytoplasmic background at Weslaco and the Texas Agricultural Experimental Station farm located near College Station (Texas) during 1998 and 1999, Moran and Rooney (2003) reported that A1, A2, and A3 cytoplasms had no effects on plant height and had minimal practical effect on days to anthesis (Table 6.13). However, grain yield in A3 cytoplasmic background was significantly reduced as compared with A1 Table 6.12. Effect of A1 and A3 cytoplasms on forage traits in sorghum-sudan grass hybrids evaluated during 1990 and 1991 at Univ. of Nebraska. Source: Pedersen and Toy (1997).

Cytoplasm source A1 A3 Significancez z

Days to 50% flowering

Plant height (m)

Dry matter yield (t ha–1)

In vitro dry matter digestibility yield (t ha–1)

Cut 1

Cut 2

Cut 1

Cut 2

Cut 1

Cut 2

Cut 1

Cut 2

72 72 NS

1.77 1.79 NS

2.16 2.13 NS

3.52 3.50 NS

5.85 5.97 NS

637 635 NS

602 607 NS

12.2 12.3 NS

9.4 9.5 NS

NS indicates non-significant

Crude protein (%)

6. GENETIC AND CYTOPLASMIC-NUCLEAR MALE STERILITY IN SORGHUM

159

Table 6.13. Effect of A1, A2, and A3 cytoplasms on average grain yield, days to anthesis, and plant height in iso-nuclear grain sorghum hybrids evaluated during 1998 and 1999 at Weslaco and the Texas Agricultural Experimental Station farm. Source: Moran and Rooney (2003). Cytoplasm source Trait Average grain yield (t ha–1) Days to anthesis Plant height (m) z

A1

A2

A3

5.01 az 70 a 1.44 a

4.92 a 70 a 1.43 a

4.74 b 71 b 1.42 b

Means followed by same letter are not significantly different

and A2 cytoplasm-based hybrids. Although the specific reason for the reduced yield of A3 hybrids is not known, seed set data indicated that it was not associated with fertility restoration. C. Restorer Gene Frequency The availability of restorers determines the extent of the use of various CMS systems in hybrid seed production. Scheuring and Miller (1978) reported a frequency of 0.62 restorers and 0.23 maintainers on milo (A1) cytoplasm in the world collection of 3,507 sorghum accessions. The work carried out at ICRISAT showed a restoration frequency of 0.9 on A1, 0.5 on A2, 0.1 on A3, and 0.3 on A4, when 48 germplasm lines were test crossed onto A1, A2, A3, and A4 CMS systems (Reddy et al. 2003). Senthil et al. (1998) found that the frequency of restoration was 0.15 on A1, 0.04 on A2, 0.01 on A3, and 0.03 on A4 CMS systems. Both results suggest that the restorer frequency is highest on A1, and lowest on A3 CMS system. Hence, considering the restoration frequency, A1 CMS system provides the widest possible choice in selecting restorers. D. Cytoplasm Effects on Heterosis for Economic Traits Even if all the requirements are met in a CMS system, the existence of economically viable heterosis ultimately determines the use of a CMS system. In sorghum, as indicated earlier, the A1 cytoplasm is more stable than other alternative cytoplasms. The frequency of restorer genes for A1 CMS is higher than with others. The heterosis estimates reported for grain yield using A1 CMS system vary. For example, results from the Indian National Program Testing showed that the standard heterosis with A1 CMS system for grain yield ranged from 18 to 31% in the rainy season, and from 19 to 29% in the post-rainy season in the years 1999

160

B. REDDY, S. RAMESH, AND R. ORTIZ

and 2000. The heterobeltiosis (or highest parent heterosis) estimates for the same period ranged from 15 to 26% in the rainy season and 1.5 to 11% in the post-rainy season (Reddy et al. 2003). Siddiq et al. (1993) reported that heterobeltiosis was 38% for grain yield in the rainy season. Similar studies with alternate CMS systems are limited. Senthil et al. (1998) also reported that the A1 CMS system produced a higher number of heterotic combinations than the A2, A3 or A4 system. Kishan and Borikar (1989a) observed that A2-based hybrids had larger grains and higher yields than A1- and A4-based hybrids. Based on testing of 15 hybrids derived from three isonuclear male-sterile lines and five common restorers, the A4-based hybrids were inferior to others for grain yield in the rainy season. However, another report indicated that A4-based hybrids had higher grain yield and larger grain size than A1 hybrids during the post-rainy season study (Kishan and Borikar 1989b). The favorable and higher frequency of mid-parent heterosis observed in the landrace-based hybrids for several of the post-rainy season traits such as grain and fodder yields in the Advanced Hybrid Trials (AHT) and Landrace-based Advanced Hybrids and Parents Trials (LRAHPT) at Patancheru and Nandyal (India) during the post-rainy season (Table 6.14) indicates the potential for exploitation in the field for post-rainy season adaptation. Evaluation of two sets of 36 hybrids obtained by crossing two different sets of six A1 and A2 isonuclear CMS lines and three dual common restorers at Patancheru during the post-rainy season of 2001 and the rainy season of 2002 indicated an absence of significant differences between A1 and A2 CMS systems for mean heterosis (%) for any of the traits (Tables 6.15 and 6.16).

Table 6.14. Standard heterosis of landrace-based sorghum hybrids over M 35-1 (check) in landrace advanced hybrid trials at two locations during post-rainy season at Patancheru (Andhra Pradesh, India). Hybrids showing standard heterosis (%)

Range of standard heterosis (%)

Location

Year

Grain yield

Fodder yield

Grain yield

Fodder yield

Advanced Hybrid Trial (AHT)

ICRISAT, Patancheru, India

1993

79

45

3.8–49.2

0.4–40

Landrace Advanced Hybrids and Parents Trial (LRAHPT)

Nandyal, India

1995

100

97

23.2–81.3

0.3–118

Trial

6. GENETIC AND CYTOPLASMIC-NUCLEAR MALE STERILITY IN SORGHUM

161

Table 6.15. Effect of A1 and A2 CMS systems on mean heterosis for grain yield and other economic traits in sorghum during post-rainy season of 2001 at Patancheru (Andhra Pradesh, India). Mean heterosis (%) Trait

A1

A2

Difference

Significancew

Days to 50% flowering Plant height (cm) Grain yield (t ha–1) Plant agronomic performancez Lodging resistancey Aphid resistancex Seed set under open pollination Seed set upon selfing

–6.67 47.18 31.78 –1.62 62.57 –6.22 4.05 9.04

–6.45 46.61 27.87 7.88 67.93 –6.34 3.11 1.99

0.22 –0.58 –3.90 9.51 5.37 –0.12 –0.93 –7.05

NS NS NS NS NS NS NS NS

z

1–5 scale where 1 = good and 5 = poor 1–5 scale where 1 = < 10% plants lodged and 5 = > 80% plants lodged x 1–5 scale where 1 = leaf free from aphids damage and 5 = > 60% leaf area damaged w NS indicates non-significant y

It may be advantageous to use the A2 CMS system among the alternate cytoplasms available after considering the restoration frequency, development of high-yielding male-sterile lines, and hybrid performance when using A1 restorers. However, the A2 CMS system is not popular, as the anthers in A2 male-steriles, unlike the A1 male-steriles, mimic the fertile or maintainer lines and lead to complex monitoring of the purity of hybrid seed production. Extensive research is underway at ICRISAT and among Indian researchers for the development of A2 cytoplasmbased hybrids. Based on A2 CMS systems, the hybrid ‘Zinza No. 2’ was released in China for commercial cultivation. This hybrid is now grown Table 6.16. Effect of A1 and A2 CMS systems on mean heterosis for grain yield and other economic traits in sorghum during rainy season of 2002 at Patachenru (Andhra Pradesh, India). Mean heterosis (%) Trait

A1

Days to 50% flowering Plant height (m) Plant agronomic performancez Grain yield (t ha–1)

–3.27 42.46 0.72 43.79

z y

1–5 scale where 1 = good and 5 = poor NS indicates non-significant

A2 –3.071 41.23 10.26 44.49

Difference

Significancey

0.20 –1.23 9.55 0.70

NS NS NS NS

162

B. REDDY, S. RAMESH, AND R. ORTIZ

in an area of 200,000 ha, accounting for one sixth of total sorghum area in China (Liu Qing Shan et al. 2000).

VII. DIVERSIFICATION OF CMS SYSTEMS Most of the sorghum hybrids worldwide depend on the A1 CMS system, which may lead to cytoplasm uniformity among hybrids for this crop (Moran and Rooney 2003; Reddy and Stenhouse 1994). This uniformity may increase the risk of outbreak or quick transmission of cytoplasmically inherited susceptibility to pests or diseases, as reported in maize with Helminthosporium leaf blight, which was linked to T-cytoplasm (Kishan and Borikar 1988) and pearl millet downy mildew disease epidemic (C. T. Hash 1998, pers. commun.). There are no reports of such an occurrence of cytoplasm-linked diseases or pests in sorghum. However, cytoplasmic diversity should be kept in this crop to avoid such hazards that may arise in future. In addition to causing uniformity of cytoplasm in the hybrids, the use of milo (A1) cytoplasm-based hybrids restricts the nuclear diversity (Schertz and Pring 1982). With the introduction of additional cytoplasmicnuclear male sterility systems, new parental combinations should be possible. Hence, new and alternate cytoplasms have been sought with additional nuclear diversity, and are studied in several countries (Rao 1972; Nagur and Menon 1974; Tripathi 1979; Quinby 1981; Schertz and Pring 1982; Rao et al. 1984), based on restoration patterns in testcrosses made with common pollinators. The frequency of maintainer genes in a diverse range of improved populations and breeding lines has a direct bearing on the success of genetic diversification of A-lines. Conversely, the frequency of restorers for several established CMS sources directly influences the use of such diversified male steriles in grain hybrid development. In addition, as already discussed in previous sections, the factors influencing the use of alternate CMS systems that include stability of their male sterility, their effect on economic traits, and extent of heterosis for economic traits, should be analyzed to make a rational judgment about CMS diversification. Often such analyses are not carried out, leaving new useful CMS systems unused or underused. A. Germplasm Base and Traits Although the need to diversify the cytoplasm base of seed parents has been recognized, research at ICRISAT showed that the kafir-based crosses with CK 60B produced higher frequency of B-lines than the

6. GENETIC AND CYTOPLASMIC-NUCLEAR MALE STERILITY IN SORGHUM

163

caudatum-based B-lines with A1 CMS, which essentially derives its male sterility maintainer genes from kafir. This finding, coupled with the reduced frequency of restoration and the difficulties in distinguishing the fertiles from male steriles among alternate CMS systems at field level, forced many programs to depend mostly on the A1 CMS system for the production of hybrids (Moran and Rooney 2003). There are four major national programs (USA, India, China, and Australia) and an international program (ICRISAT) engaged in diversifying the base of the male steriles and hybrids. All the national programs depend mostly on kafir-milo based male-sterile lines. The U.S. program diversified its kafir-milo (durra) male-sterile system by crossing with other kafir or bicolor lines in red- or brown-grain color background up to the mid1980s (Schertz et al. 1997). The U.S.-based program influenced the sorghum breeding programs in China and Australia. Later on, the U.S.based program further diversified its female parent base by crossing the kafir-bicolor B-lines with kauras (durra-caudatum) and other caudatum lines of African origin (Duncan et al. 1991; Schertz et al. 1997). This approach led to the development of a series of white-grain color malesterile lines (e.g., Tx 623A, Tx 624A). The Indian program introduced CK 60A (milo-kafir) from the USA in the early 1960s, and developed 2219A (kafir-shallu), 2077A, 3675A, and 3677A among other kafir types in the late 1960s. Later on, male steriles were diversified by further selection and identification of maintainers in the crosses of the kafir Blines with Indian durras, which are usually restorers on the milo-kafir system (Rao 1972). The most significant heterotic female line developed from such a program is 296A. However, this line was reported to be temperature-sensitive and seed production based on this line, therefore, became difficult due to low temperature-induced female sterility. The national programs diversified their male-sterile line base mostly for grain yield or the yield components such as grain number, grain size, etc. (Rao 1972; Duncan et al. 1991). Little effort was directed toward breeding male-sterile lines for resistance to abiotic and biotic stresses during the 1970s and 1980s (J. W. Stenhouse, pers. comm.). During late 1970s, ICRISAT began the development of high-yielding male-sterile lines on the A1 CMS system in white-grain background by extensively using kafir-caudatum-guinea crosses, and selecting whitegrain male-sterile lines for grain yield. This program resulted in the development of a series of male steriles such as ICSA 1 to ICSA 110 and ICSA 88001 to ICSA 94013. In the late 1980s, ICRISAT made a major shift in its emphasis on the improvement of A1 CMS male-sterile lines for resistance to various abiotic and biotic yield constraints. Considering the high correlation between per se performance of lines and hybrids (Rao 1972), ICRISAT followed simultaneous selection, test-crossing,

164

B. REDDY, S. RAMESH, AND R. ORTIZ Table 6.17. Trait-specific resistant male sterile lines developed by ICRISAT sorghum breeding program at Patancheru (Andhra Pradesh, India). Trait

Lines

Downy mildew Anthracnose Leaf blight Rust Grain mold Shoot fly Stem borer

ICSA 201 to ICSA 259 ICSA 260 to ICSA 295 ICSA 296 to ICSA 328 ICSA 329 to ICSA 350 ICSA 351 to ICSA 408 ICSA 409 to ICSA 463 ICSA 464 to ICSA 474 ICSA 475 to ICSA 487 in post-rainy season ICSA 488 to ICSA 545 ICSA 546 to ICSA 565 in rainy season ICSA 566 to ICSA 599 ICSA 600 to ICSA 614 ICSA 615 to ICSA 637 ICSA 638 to ICSA 670 ICSA 671 to ICSA 674 ICSA 675 to ICSA 68

Midge Head bug Striga Acid soils Early maturity Bold grain Tillering Stay green

and backcrossing methods to convert the maintainer lines selected for resistance to individual stresses and grain yield, under specific screenings in trait-specific breeding populations (Reddy et al. 1992). As a result, several trait-specific resistant male-sterile lines were developed (Table 6.17) Each ICRISAT breeding population involved different source materials, namely guinea for grain mold, durra for shoot fly, stem borer, and Striga, and caudatum-guinea for stay-green lines. B. Other Cytoplasms Breeding programs in the USA also developed a few high-yielding A2 and A3 CMS lines (Duncan et al. 1991). ICRISAT has converted some of the high-yielding A1 CMS restorers into A2 (ICSA 688 to ICSA 738), A3 (ICSA 739 to ICSA 755), and A4 (ICSA 756 to ICSA 767) CMS lines. Most of these lines were developed from kafir-durra-caudatum crosses. C. Information Management and Knowledge Sharing through the ICRISAT Website The development, characteristics, and pedigrees of the above malesterile lines improved at ICRISAT are described fully in the ICRISAT web page, which can be accessed under crop at http://www.icrisat .org/text/research/grep/homepage.htm (Mahalakshmi et al. 2002). Infor-

6. GENETIC AND CYTOPLASMIC-NUCLEAR MALE STERILITY IN SORGHUM

165

mation about the performance of individual lines for traits of interest can be displayed, and further information on parents used in the breeding program can be obtained by clicking on the icon for the parent, which is linked to the germplasm passport data. Seeds of both breeding lines can be obtained by accepting the material transfer agreement on line and placing the request. Other information available in the sorghum web page includes gene bank germplasm and its score collection, crop diseases, pest, parasitic weeds, and plant nutritional disorders. The on-line system from the main crop page allows a query using the common or scientific name for a pest. There is also a two-module on-line learning system on sorghum that is based on sorghum practices ensuing from ICRISAT training manuals on breeding and seed technology in sorghum. D. Effect of Genetic Background Nuclear genetic background also has profound influence on male sterility. The fertility maintenance patterns differ from cross to cross, although the parents involved in such crosses by themselves are maintainers. Work at ICRISAT showed that the frequency of non-maintainers in B × B crosses ranged from 0 to 25%. For example, all the progenies were male-sterility maintainers in a cross of ICSB 554 × ICSB 79. On the other hand, in a cross like ICSB 583 × ICSB 64, nearly 20% of the progenies failed to maintain male sterility. The extent of the influence of genetic background thus determines the effectiveness of diversification of malesterile lines. At ICRISAT, it is also observed that the maintainer gene frequency among the progenies of B × B crosses depends on the tester (male-sterile line) used, indicating the role of the “residual” cytoplasm factors accumulated in the converted male-sterile lines. It is useful to use, as much as possible, the original CMS source. For example, to diversify A1 CMS lines, it may be desirable to convert to female the original male-sterile line CK 60A. E. Seed Parents’ Purity Maintaining the purity in hybrid parents is important because impurity may arise due to mechanical admixture of the parents among themselves or with other lines. Impurity can also arise due to occurrence of pollen shedders in A-line, which depends on the CMS systems and the seed production season. The pollen shedding revertants may be due to mutation occurring either in cytoplasm or nuclear genes. The cytoplasm fertile revertants can be pulled out at flowering before they contribute to seed setting to maintain the purity of seed. If a nuclear gene mutation is the cause of reversion to male fertility, A-line seed purity can be

166

B. REDDY, S. RAMESH, AND R. ORTIZ

maintained by making plant-to-plant crosses between A- and B-line progenies for two successive generations.

VIII. HETEROSIS AND HYBRID DEVELOPMENT Restorer and male-sterile lines should be as divergent as possible to gain maximum advantage in heterosis (Allard 1960). There is a considerable body of evidence to suggest that per se performance of the parents is highly correlated with hybrid performance (Rao and Rana 1982). Besides, other characters such as flowering behavior, pollen quantity (in R-lines), seed setting (in A-lines), and relative plant height of A- and Rlines also influence the hybrid seed production. A-lines with transpiration efficiency, OA > staygreen), depending on the types of environments to which progenies were exposed over the breeding program cycles. C. Stress-sensitive Periods in the Life of the Crop In older maize germplasm not subject to a long history of improvement, yield is reduced two to three times more when water deficits coincide with flowering compared with other growth stages. A drought-sensitive period from a few days before silking to about 25 days after has been identified, and water deficits during this time can sharply reduce grain number per plant (Claasen and Shaw 1970; Shaw 1977; Herrero and Johnson 1981; Grant et al. 1989; NeSmith and Ritchie 1992). Sensitivity during this period may be greater in maize than other cereals due to its monoecious habit and its relatively synchronous development of many florets on a single ear. Factors affecting kernel set under drought are reviewed in Section III D. In modern hybrids, considerable improvement has occurred in yield under stress at flowering. In a recent study 18 ERA hybrids released between 1953 and 2001 (approximately 3 hybrids per decade) were subjected to a period of approximately 50 days without irrigation in a rain-free environment. Genetic gain per year during this period was greatest under unstressed conditions (211 kg ha–1 yr–1) and declined by 41% under flowering stress and by 77% under late grain fill stress (Table 7.1). Data show that gains in tolerance of grain yield to flowering stress have outpaced those for stress tolerance during the latter half of grain filling. This was accompanied by systematic reductions in the anthesis-silking interval (ASI) with relatively little change in staygreen or in weight per kernel under grain filling and terminal stress (Fig. 7.3). Thus, susceptibility to stress at flowering has kept pace with increases

182

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

Table 7.1. Yield reduction and genetic gain observed in 18 hybrids that were released between 1953 and 2001, when exposed to a stress imposed at different growth stages. Data are from two trials conducted in a dryland environment, 2001–2003. Period of major drought stress

Variable Stress started (°C days) Stress relieved (°C days) 50% anthesis (°C days) Mean yield (t ha–1) Percent of control Gain (kg ha–1 yr–1) 2 R

Control

663 15.91 100.0 211 0.93***

Early grain fill

Mid grain fill

Late grain fill

Terminal

257

354

456

583

769

772

861

956

1103

1283

681

676

670

662

660

Flowering

6.91 43.4 124 0.47**

4.95

4.73

6.26

10.98

31.1

29.7

39.3

69.0

92 0.42**

91 0.60***

48 0.71***

77 0.83***

in yield potential, while performance under late-season stress has not improved as rapidly as yield potential. Elite hybrids show considerable variation in drought susceptibility at flowering and in ASI, but the best of these show little yield loss when drought coincides with flowering (Bruce et al. 2002). D. Tolerance 1. Seedling Establishment. Adequate stands are critical to high yields. Plant establishment may be severely reduced by drought when a dry spell is encountered after initial planting rains. Research in tropical maize directed toward improving seedling establishment under drought stress showed only modest increases in survival under water deficit (Bänziger et al. 1997b). The authors concluded that selection for improved survival and biomass production under post-emergence drought stress is difficult since environmental variation is high in field screens, and because natural selection may have exploited much of the naturally occurring genetic variation for these traits. Drought during establishment is a rare occurrence in temperate production zones, and selection for seedling tolerance therefore is not a priority in Corn Belt germplasm.

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

183

85 65

Flowering stress Slope = NS (R2 = 0.01)

45

50 25 0 –25

25 1950

7

5

1970

1980

1990

Well-watered Slope = 0.120 yr–1 (R2 = 0.62***)

0.45

Flowering stress Slope = 0.035 yr–1 (R2 = 0.51***)

3

1 1950

Filling, terminal Slope = 0.022 yr–1 (R2 = 0.10 NS) 1960

1970

1980

1990

No flowering stress Slope = –0.87°Cd yr–1 (R2 = 0.73***)

–50 1950

2000

2000

Weight per kernel (g)

Stay green visual score

9

1960

Flowering stress Slope = –1.50°Cd yr–1 (R2 = 0.66***)

75 ASI (°Cd)

Yield as percent of well-watered

100 Filling and terminal stress: Slope = –0.53% yr–1 (R2 = 0.65***)

1960

1970

1980

1990

2000

Well-watered/flowering Slope = 0.0024g yr–1 (R2 = 0.64***)

0.40 0.35 Filling/terminal stress Slope = 0.0004 g yr–1 (R2 = 0.17 NS)

0.30 0.25 0.20 0.15 1950

1960

1970

1980

1990

2000

Year of release of hybrid

Fig. 7.3. Grain yield of hybrids exposed to grain filling/terminal and flowering drought stress, expressed as a percent of the well-watered control, and anthesis-silking interval (ASI), visual stay green score, and weight per kernel vs. year of release. The 18 hybrids, a subset of the ERA hybrid set released from 1953 to present, were evaluated in Chile in 2003.

2. Vegetative Tolerance. Maize is relatively tolerant of stresses that occur after establishment but prior to flowering. At this stage cell expansion is very sensitive to plant turgor (Hsiao et al. 1970) and the primary stress effects are reductions in leaf area and plant height since leaves are expanding and stalks are elongating. Internode elongation patterns are a sensitive stress-timing diagnostic since any reduction in internode length is usually permanent. Poor husk cover is another byproduct of drought effects on cell expansion. Prophyll elongation occurs up to a week earlier than ear elongation. If a water deficit is relieved following the peak of husk growth, the ear may grow beyond the husk tip, creating an easy entry point for diseases and insects. This can negatively impact grain yield data quality under water stress. Drought prior to flowering reduces plant biomass because intercepted radiation is less and because radiation use efficiency (RUE) declines with stomatal closure and reduced gas exchange. A 20% reduction in RUE

184

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

and a 7% reduction of intercepted radiation were reported in tropical maize subject to a severe flowering stress (Bolaños and Edmeades 1993a), although the effects on radiation interception became cumulatively greater during grain filling. Earl and Davis (2003) reported a 5–20% reduction in intercepted radiation but a 21–61% reduction in RUE under severe stress imposed in two successive seasons. There is also a concomitant reduction in leaf chlorophyll (Chaves et al. 2003). Yellowing of young leaves and a 15–25% reduction in chlorophyll levels are often observed in vegetative drought-stressed plants (G. Edmeades unpubl. data, 2002). Pre-flowering drought tolerance reflects a capacity to generate and maintain a large flux of current assimilate to the developing ear. Leaf rolling is commonly observed during the vegetative period and contributes to decreased light interception. Its occurrence varies greatly among genotypes and has usually been selected against in breeding programs that target improved drought tolerance (Bolaños and Edmeades 1996). Rolling typically begins mid-day a few days after first observing reduced plant height in water stressed plots. It reflects plant water status, leaf morphology and the presence of bulliform cells. However, the role of leaf rolling is being reconsidered because it is observed more frequently in newer hybrids than in older ERA hybrids (Fig. 7.4). Apparently, elite hybrids can reduce radiation interception and water use by leaf rolling to conserve water yet generate sufficient assimilate flux to the ear for adequate kernel set. Proclivity to roll is undoubtedly aided by the more upright leaf habit of hybrids released since 1970 (M. Cooper, unpubl. data, 2003). Prior to flowering, maize roots increase extensively in length, number and dry weight, and their capacity to capture water normally reaches a maximum near flowering (Mengel and Barber 1974). Excess or too little soil resistance to root growth appear to send signals, perhaps abscisic acid (ABA), that reduce top growth and directly affect yield (Passioura 2002). Root length density (cm length cm–3 soil) decreases exponentially with depth in the soil in most maize cultivars (J. Bolaños, unpubl. data, 1989; Gregory 1994). Intense rooting near the soil surface is of limited value to the plant under prolonged water stress, and is less necessary for nutrient capture in modern production systems because fertilizer is placed close to the plant. Recurrent selection for eight cycles in one tropical maize population exposed to severe drought stress at flowering resulted in a 35% root biomass reduction in the upper 0.5 m of soil (Bolaños et al. 1993) even though grain yield under flowering stress increased by 40% (Bolaños and Edmeades 1993a). Richards (1991) also reported improved stress tolerance in wheat when roots were reduced

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

Flower, early fill and midfill stages Slope = –0.035 yr–1 (R2 = 0.79***)

40 Canopy temperture (°C)

Leaf rolling score

8

7

6

5

4 1950

1960

1970

1980

1990

2000

185

Flower, early fill and midfill stages Slope = –0.027°C yr–1 (R2 = 0.48**)

35

30

Well-watered control Slope = –0.010°C yr–1 (R2 = 0.17 NS)

25 1950

1960

1970

1980

1990

2000

Year of hybrid release

Fig. 7.4. Visual leaf rolling score (1 = severely rolled; 9 = completely unrolled) and canopy temperature observed on three occasions shortly after solar noon in 18 ERA hybrids grown under three water stress treatments imposed at different development stages prior to or shortly after flowering, Chile 2003. Canopy temperatures were observed one day prior to relieving the stress.

in the top 30 cm of soil. Limited data suggest that selection for high stable yield in temperate maize has led to deeper rooting and conserved water use, possibly aided by increased leaf rolling under stress. Increased capacity to access soil water is supported by a significant trend in lower leaf temperature in more recent hybrids of the ERA set grown under stress (Fig. 7.4). In drying soils, even root distribution with depth without an increase in root biomass would be desirable. There is little evidence that a further increase in rooting intensity over 1 cm cm–3 will improve water uptake (Ludlow and Muchow 1990), although maximum N and P uptake may call for root length densities of 2 to 5 cm cm–3. Increasing fine root numbers with depth seems fully justified (Boyer 1996) provided water is present at those depths. This is supported by recent results from simulation in wheat (King et al. 2003). Since measuring roots is difficult in a breeding program, several researchers have used electrical capacitance measurements for rapid root dimensions assessment. Although electrical capacitance and total root volume are positively correlated (van Beem et al. 1998), the measurement does not assess root distribution (CIMMYT, unpubl. data, 1998). 3. Reproductive Processes Partitioning. Breeding impact on maize reproductive processes has been significant. Duvick (1992, 1997), Castleberry et al. (1984), and Russell (1985) have documented improved yield potential over time in both public and private lines. Improved hybrids have also exhibited altered

186

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

partitioning. A clear example of this change can be seen in ERA hybrid data (Duvick 1997; Duvick et al. 2004). For example, there has been a shift in partitioning away from tassel development (as measured by decreased tassel size) and toward increased ear growth (as measured by increased ear number per plant and decreased ASI). Other examples include: (1) drought tolerant CIMMYT populations partition more assimilate to ears than drought susceptible populations (Edmeades et al. 1993; Lafitte and Edmeades 1995); (2) barrenness associated with high-density planting is due to a shift in partitioning away from the ear and toward the stalk (Edmeades et al. 1999); and (3) improved yield under drought stress can be achieved by partitioning stem-infused sucrose to reproductive structures (Zinselmeier et al. 1995). Therefore, tassel, root, and stalk tissue may be perceived as competing structures vying for assimilate with the immature ear near anthesis. Anthesis-to-silking Interval. Grain yield per plant is closely associated with kernel number per plant, especially when the crop is subjected to stress around anthesis. Bolaños and Edmeades (1996) observed genetic correlations between grain yield and kernels per plant of 0.86 to 0.88 under severe flowering drought stress. Otegui et al. (1995) reported similar results. Utilizing managed drought stress locations, Pioneer Hi-Bred International, Inc. (PHI) elite commercial and pre-commercial maize hybrids have been evaluated for flowering stress tolerance. Results have shown that although shortened relative to older genetics, ASI in newer hybrids can be lengthened in susceptible genotypes under stress resulting in significant yield losses. Given the ease of measuring ASI and the association between a short ASI and increased grain yield under stress (Edmeades et al. 2000b), this trait is routinely measured in precommercial hybrid evaluations. Pre-commercial hybrids that are screened in managed drought stress testing sites and advanced to commercial status may be less likely to have a long ASI when subjected to flowering drought stress in grower fields than other commercial hybrids that were not evaluated under managed drought stress conditions. Ear Growth Rate (EGR). Kernel number per plant is positively correlated to plant and ear growth rates around anthesis (Tollenaar et al. 1992). Kernel set increases and ASI is shortened as plant and ear growth are increased near flowering. Kernel set is increased as source is increased during the critical flowering period (Schoper et al. 1982; Zinselmeier et al. 2002) and when source is decreased during this time, kernel set decreases (Early et al. 1967; Borrás et al. 2004). Genotypic variation exists for EGR at a given source level, and gains in EGR and kernel set may be made. Ribaut et al. (CIMMYT, unpubl.

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

187

data) have shown that ASI and EGR are highly correlated with grain yield in a tropical recombinant inbred line (RIL) population subjected to flowering drought stress. At PHI, selections were made for rapid and slow EGR at flowering under irrigated nursery conditions in a B73*MO17 syn 4 RIL population and then evaluated for flowering drought tolerance under managed drought stress conditions (Zinselmeier, unpubl. data, 2002). Yield levels were reduced approximately 50% by the flowering drought stress treatment (Fig. 7.5). Mean yield of the rapid EGR selections under drought stress was greater than that of the slow EGR selections (p = 0.05). Grain yield was highly correlated with barrenness (data not shown), suggesting that selection for EGR improved drought tolerance by enhancing partitioning to the ear. Although more difficult to quantify than ASI, EGR is a trait associated with stress tolerance that may be useful in some screening programs or in attempts to understand the biology of stress tolerance. Lizaso et al. (2003) recently used the coincidence of pollen and emerged silks to successfully model kernel set at low pollen concentrations typical of stressed fields. In an open-pollinated population, asynchrony in pollen supply and silk emergence may be relatively unlikely since there is a wide range of flowering dates in the field, but asynchrony may be an issue in a large field of a uniform single cross hybrid. Maintaining the partitioning of assimilate to the ear ensures adequate ear growth and allows synchronous floral development of the staminate and pistillate flowers (short ASI) and appears to be important for adequate kernel set in maize. Given the importance of ear growth, gene

Full Irrigation

Stress 9.4

12.6 LSD0.05 = 12

a

LSD0.05 = 26.6

a

Yield (t/ha)

9.4

c 6.3

c

Yield (t/ha)

b 6.3 b c

3.1

c

3.1

0

0 Check

Fast Parent Material Type

Slow

Check

Fast Parent Material Type

Slow

Fig. 7.5. Relationship between ear growth rate (EGR) and yield from B73*Mol7 recombinant inbred lines. Tails (n = 10) were selected under irrigated nursery conditions over 3 years for ear growth rate at anthesis, classified as a Fast or Slow EGR entry and evaluated for yield under fully irrigated and flowering drought stress conditions, Chile, 2001. Performance of parents and elite inbred checks are included for comparison.

188

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

expression profiling of immature maize ears subjected to drought stress at flowering has been conducted to determine pathways and genes associated with reduced ear growth. Initial trials were conducted in greenhouse/growth chamber conditions where the rooting volume was restricted (20-L volume pot). Water deficits sufficient to completely inhibit photosynthesis were achieved within five days and expression profiling of immature ears showed large changes in gene expression (Sun et al. 1999; Zinselmeier et al. 2002; Yu and Setter 2003). Classic “stress responsive” genes were identified as differentially expressed under drought stress, including the ABA pathway (rab17/dehydrin3 and rab 15), cell cycle (cyclin D, Cks1, MCM5, and CDK) and sugar metabolism (glu1 and hexose-transporter). Initially, these studies generated enthusiasm for the combination of expression profiling, physiology, and natural genetic variation for drought tolerance. However, when similar studies were conducted in the field where root volume was essentially unrestricted and the kinetics of stress was different (four to five weeks to completely inhibit photosynthesis instead of five days), a different story emerged. Habben (2001) showed that many fewer genes showed differential expression when stress was applied gradually (Table 7.2). Using rab17 as an example, classically drought responsive genes showed no differential expression when stress was imposed slowly under field conditions (Table 7.3). Genes showing differential expression under short-term stress were usually not a subset of the genes showing differential expression under long-term stress in the field. These results led us to conclude that field environments are unique from greenhouse/growth chamber environments and will require a re-examination of the metabolism that constitutes the physiology. Essential to this reexamination will be the use of new tools such as expression profiling and metabolic profiling to measure stress effects on gene expression at the cellular or organ level within diverse genotypes at multiple develTable 7.2. Effect of stress kinetics on differential gene expression of immature maize ears at anthesis. Greenhouse experiments were conducted with buckets and the transient water stress was sufficient to inhibit photosynthesis within 5 days after withholding water (i.e., Rapid stress kinetics). Field experiments required >5 weeks to deplete soil moisture sufficiently to inhibit photosynthesis (Slow stress kinetics). Stress category

Stress kinetics

% genes changed

Shading Drought (bucket) Density Drought (field)

Rapid Rapid Slow Slow

18 27 1 5 weeks to deplete soil moisture sufficiently to inhibit photosynthesis (Slow stress kinetics). Stress category

Stress kinetics

Organ

Fold increase

Drought (bucket) Drought (field) Drought (bucket) Drought (field) Drought (bucket) Drought (field)

Rapid Slow Rapid Slow Rapid Slow

Seed-pedicel Seed-pedicel Pedicel Pedicel Leaf Leaf

11 NSz 71 NS 134 NS

NS = not significant at p ≤ 0.1

z

opmental stages. Genetic transformation will also be critical in confirming the effect of these genes and pathways. Silk Growth. Maize silks are sensitive to changes in plant water status (Tatum and Kehr 1951; Herrero and Johnson 1981) and their growth pattern under stress is important in determining kernel set. Mechanistically, silk growth patterns may affect kernel set directly if silk exertion is not synchronous with pollen shed. Does this trait impact current elite genetics under flowering drought stress? Examining silk exertion relative to pollen shed of maize hybrids grown under flowering drought stress suggests that this trait shows significant variation in elite genetics (J. Schussler, unpubl. data). When considering a genetic solution to the problem of kernel set, one needs to ask whether the addition of kernels under water limiting conditions is wise. If a transient stress at flowering is expected, we believe that additional kernel set will improve yield given the plants’ ability to fill these additional kernels after the reduction of stress. Schussler et al. (2002) demonstrated that this concept is applicable to some germplasm grown under high plant-density stress. However, if a terminal stress is expected, this strategy will likely not improve yield. Silk growth patterns may also alter kernel set indirectly. Carcova and Otegui (2001) demonstrated that basal kernels may suppress kernel tip development in some genetic backgrounds and the suppression may be sufficient to prevent these tip kernels from being harvestable. Consequently, a breeding or transgenic strategy to enhance yield stability may be to speed silk exertion of tip silks relative to basal silk exertion, leading to more synchronous pollination of basal and tip spikelets.

190

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

Kernel Set. Water deficit at flowering can increase the frequency of abnormal embryo sac development and kernel abortion (Moss and Downey 1971; Damptney and Aspinall 1976; Westgate and Boyer 1986). Abortive mechanisms may occur as early as 2–3 days after pollination (Westgate and Boyer 1986). The lag phase of kernel development (usually defined as 0–14 days after pollination) is considered key in determining successful kernel set. Critical physiological, genetic, and cytological events occur during this period and are not reversible. Although the extent of natural variation for some of these key processes is not known, genetic solutions may be achievable. For example, manipulation of genes that control the cell cycle may govern the extent to which endosperm cells endoreduplicate (Setter and Flannigan 2001; Yu and Setter 2003). 4. Osmotic Adjustment. Capacity to accumulate osmotically active solutes (e.g., organic acids, sugars, ions) in the cell generates a gradient in water potential from cell to soil that increases cell turgor in the presence of available water. In turn, this maintains meristematic activity and leaf area expansion while delaying foliar senescence. OA has been associated with stable hybrid maize grain yields across moderately dry environments in Argentina (Lemcoff et al. 1998), but appears to have less impact under severe stress (Bolaños and Edmeades 1991). Maize OA is modest (0.4 Mpa) relative to rice, wheat, and sorghum (> 1 Mpa) (Ludlow and Muchow 1990; Nguyen et al. 1997). This mechanism may serve as a stabilizer of growth only under moderate and transient stresses. Serraj and Sinclair (2002) suggest that the principal value of OA lies in maintaining root growth in dry soils. It is likely that selection for stability under multi-environment testing has already exploited much of the useful genetic variation for OA in temperate maize. Further increases in OA under drought in roots will likely be from transgenes linked to stress inducible, root-preferred promoters. 5. Growth Regulators as Signals. Phytohormone concentrations vary with stress level and can have profound growth effects. ABA concentration, often considered a growth regulator fostering survival rather than productivity, increases significantly under stress, leading to root growth maintenance, stomatal closure, leaf shedding, dormancy and possibly tip kernel abortion in maize (Ober and Setter 1992; Jones and Setter 2000; Setter et al. 2001; Ober and Sharp 2003). Improved yield under stress has also been associated with reduced ABA leaf concentrations in tropical maize (Mugo et al. 2000). Recent research suggests that ABA may function to limit ethylene production since ethylene increases under

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

191

stress, inhibiting root growth and building up in rolled leaves (Sharp 2002). Heat stress reduces cytokinin concentrations in developing kernels (Jones and Setter 2000), and water stress likely has similar effects. Enhanced cytokinin biosynthesis by over-expression of transgenes driven by stress-inducible promoters in the roots could provide a positive signal to stimulate both root and shoot growth, provided the translocation of cytokinin to the developing grain did not induce undesirable side effects such as precocious germination on the ear. 6. Visual and Functional Staygreen. Drought accelerates leaf senescence, especially in the lower leaves. These leaves have a relatively minor role in assimilation in a normal canopy but serve as a source of remobilized assimilates and nutrients for younger leaves and developing grain. Breeding programs usually report delayed senescence as visual staygreen scores. Greenness per se does not guarantee continued C assimilation, and the distinction between visual staygreen and functional staygreen (i.e., high carbon exchange ratio, CER) is critically important (Thomas and Howarth 2000). Staygreen, however, has been associated with improved performance under drought (Borrell et al. 2000, 2001) and N stress (Bänziger and Lafitte 1997; Bänziger et al. 2002). There has been significant improvement in staygreen scores with selection in temperate germplasm evaluated under well-watered conditions (Fig. 7.3). Gain, however, was much less under grain filling and terminal stresses, a difference that is reflected also in small changes in weight per kernel under these stresses. Heterosis delays leaf death under drought in maize (CIMMYT, unpubl. data). Since N cannot be taken up readily from dry soil, a large grain sink may also accelerate leaf senescence. For a leaf area index of 4, up to 30% of the N in foliar biomass would be required if yields under drought stress increased by 0.8 t ha–1 and leaves were the only source of N for grain growth (Chapman and Edmeades 1999). Attention must be paid to slowing the decline in functional staygreen (best measured as CER) with age under the highly photo-oxidizing conditions that prevail in drought-stressed canopies exposed to high radiation levels. Modern hybrids grown under high density or drought had a reduced rate of decline in staygreen when compared with a hybrid released 30 years previously (Tollenaar 1992; Nissanka et al. 1997; Valentinuz 2002). Exploration of antioxidant enzymes such as superoxide dismutase (SOD) to control the concentration of active oxygen species (O2 and H2O2 mainly) and the metabolism of antioxidants ascorbate and glutathione remains a research priority. Over-expression of

192

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

SOD has been reported to increase chilling tolerance in maize and drought tolerance in wheat (Noctor and Foyer 1998). Furthermore, there is evidence that specific metabolites such as glycinebetaine serve as protectants for enzymes and membranes against deleterious desiccation and photoxidation effects (Makela et al. 1996). Late embryogenesis abundant proteins increase sharply in tissues during desiccation or chilling, are upregulated by ABA, and are thought to stabilize plasma membrane structure (Campbell and Close 1997). However, formation of these compounds may represent an emergency response by the plant under severe drought conditions that rarely occur in a temperate production environment. While the small increase in CER provided by protectants in plants exposed to flowering stress may prevent barrenness in some cases, it is likely to be of limited value under severe terminal drought stress. 7. Remobilization of Stem Reserves. This is generally considered to be of greater importance in other cereals, such as wheat, than in maize, although in both species the process helps maintain a reasonably constant rate of grain filling under stress. Blum (1998) identified genetic variability for remobilization in wheat by applying chemical defoliants midway through grain filling (Blum 1998). A larger flux of remobilized assimilate might be expected from taller vs. shorter genotypes, but taller maize varieties have not proven more stable in yield under drought than their shorter counterparts (Fischer et al. 1983; Edmeades and Lafitte 1993). Stem reserves are much less effective at maintaining kernel set at flowering than in stabilizing filling rate (Schussler and Westgate 1991a,b, 1994, 1995; Blum 1997). Finally, capacity to remobilize stem reserves is likely to induce stem lodging in temperate maize, and any effort to screen for this trait must be done concurrently with evaluation of lodging susceptibility. 8. General Stress Tolerance. The strong correlation between kernel numbers and grain yield under any stress that affects photosynthesis per plant during flowering suggests that improvement for drought tolerance may increase tolerance to a number of other stresses. Drought tolerance has been positively correlated with high plant density-stress tolerance (Dow et al. 1984). There have been concomitant increases in plant densitystress and drought tolerance in temperate hybrids released over the past 50 years (Edmeades et al. 2003). Grain yield improvement and ASI reduction under drought through selection were accompanied by yield increases and reduced ASI under low N in four populations (Bänziger et al. 1999) and under high plant density (Mugo et al. 2003). As the stress

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

193

level intensifies, general stress tolerance may become less important and tolerance mechanisms specific to that stress may become more important. Until a severe stress intensity is reached, however, selection for rapid ear growth rate using ASI and barrenness as external indicators should confer broad tolerance to an array of abiotic stresses that have their primary effect through reduced plant assimilation during flowering and grain fill (Edmeades et al. 2000b). We contend that carefully controlled water deficits generated in a rain-free environment provide the most effective means to display genetic variation for these traits. E. Relative Importance of Secondary Traits in Selection for Drought Tolerance Traits described in this section emphasize underlying mechanisms that confer stable grain yield, allowing us to dissect this complex trait and identify the key control points for further physiological and genetic analysis. Secondary traits may also enhance direct selection, allowing the breeder to develop an ideotype for specific environments that differ in drought stress timing and intensity (Rasmusson 1991; Chapman et al. 2003). There is a well-developed body of theory surrounding the use of secondary traits in breeding (Falconer and MacKay 1996). Genetic gain in the primary trait (GYSTR) is greater from indirect selection for a secondary trait such as ASI, than from direct selection when [hGY < |rG × hST|] where hGY and hST are the square roots of heritability for the primary and secondary traits, and rG is the genetic correlation between them (Bänziger and Lafitte 1997). It is rare to satisfy this condition under drought except where stress reduces GYSTR by 80% or more and its heritability falls sharply (Bolaños and Edmeades 1996). There are a few cases, however, where gain from selection for yield alone has been compared with gains from selection for [yield + secondary traits]. In an evaluation of a series of progeny trials conducted within a recurrent selection scheme for tolerance to low N, Bänziger and Lafitte (1997) reported that gain for GYSTR was increased by 14% when secondary traits were included during selection. Secondary traits have maximum utility in selection when they are (1) genetically associated with grain yield under stress, (2) highly heritable, (3) cheap, fast to measure and nondestructive, (4) stable over the measurement period, (5) observed at or before flowering so that undesirable parents are not crossed, (6) not associated with yield loss under unstressed conditions, and (7) represent an actual measurement rather than a subjective score (Edmeades et al. 2000a). Based on current knowledge (see Bolaños and Edmeades 1996) only GYSTR, ASI, barrenness, kernels per ear and staygreen are candidates for

194

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

inclusion in such a selection index. While other secondary traits discussed may have potential in a selection index, most are not currently considered because they do not meet, or have not been evaluated for, the above criteria.

III. EXPERIMENTAL METHODS A. Target Environments and Environmental Characterization Maize breeding programs are designed to change the genotype and phenotype for multiple traits. However, the genetic basis for most important traits and their contributions to grain yield within the context of a TPE is not yet well understood. Lacking detailed knowledge of the gene-tophenotype relationships, it is difficult to predict many of the new trait phenotypes resulting from genotypes created over multiple breeding cycles. Consequently, maize breeding programs, regardless of strategy, rely heavily on direct evaluation of the new genotypes for their trait phenotypes within multi-environment trials (METs) that are expected to represent target production conditions. To predict and achieve high realized genetic gain for yield and yield stability, METs should be conducted with low experimental error and with sampling of environments and management conditions that predict the expected performance of the genotypes within key target environments, or more broadly in the TPE (Comstock 1977; DeLacy et al. 1996). With regard to experimental error, we are concerned with micro-environmental variation and other sources of withinexperiment measurement effects that impact experimental precision. Environment sampling controls the effects of genotype-by-environment (G*E) interactions that impact the accuracy of the differences observed in the METs relative to true values in the target environments and TPE. Conducting METs to achieve high levels of precision and accuracy is particularly challenging for drought-affected environments where experimental errors are often high and heterogeneous, and G*E interactions are common and poorly understood. Experimental error and G*E together contribute to low heritability and low realized genetic gain for yield and other performance-related traits discussed in Section II. Many authors have discussed issues regarding trial precision and accuracy when testing genotypes that have already been created (e.g., DeLacy et al. 1996; Smith et al. 2002a,b). Comstock (1977, 1996) articulated these points relative to the creation of new genotypes by the breeding process. Predicting performance of current genotypes and the capacity to create new genotypes are both important considerations in

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

195

the design and implementation of a successful breeding program (see Section IV C and D). For commercial hybrid maize, we create and test both new inbreds with superior breeding value and new hybrids with superior performance for grain yield and drought tolerance component traits (Section IV A). Here we consider some general aspects of G*E interaction and experimental error and implications for the design and analysis of METs to predict the performance of maize genotypes in a drought-affected TPE. This is done within the context of a framework for evaluating and improving conventional maize breeding strategy (Section IV) and considering the opportunities to use molecular technologies to enhance the breeding strategy (Section V). For important quantitative traits, the phenotype is many levels of organization removed from that of allelic variation at the genome level. However, defining a breeding target requires joint consideration of the genetic changes created by breeding strategies and the effects of these changes on phenotypic performance. Technology only recently advanced to the level required to examine the detailed genetic architecture of traits and breeding strategy impact extending from the genome through to the phenotypic variation of quantitative traits. In practice hybrids are grown across a range of environmental conditions within a geographically defined target domain. At the phenotypic level, this is represented as a linear statistical model: pijk = m + ej + gi + (ge)ij + eijk

[1]

where pijk is the trait phenotypic value for measurement k taken on genotype i in environment j, m is the grand mean over all observations, e j is the marginal effect of environment j, gi is the marginal effect of genotype i, (ge)ij is the genotype-by-environment interaction effect for genotype i in environment j, and eijk is the residual effect for observation k on genotype i in environment j, often referred to collectively as experimental error. Note that in this equation and the following discussion, random effects are indicated with an underline. Nyquist (1991) and Holland et al. (2002) discussed alternative forms of this linear model. From equation [1] we see that predicting the outcomes of selection among genotypes and differences among a given set of genotypes within a target environment or the TPE, i.e., the differences in gi + (ge)ij effects, is complicated by the presence of G*E interactions and experimental errors. To consider the impact of these effects on breeding, we can extend the statistical model from the genotypic to the genic (gene action) level. For example, in relation to equation [1] we can represent this extension of the linear model as:

196

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

(

pijk = µ + e j + α q + ηq

)

i

(

+ (αe)qj + (ηe)qj

)

ij

+ ε ijk

[2]

where we have partitioned genotypic effects gi into an additive (a q) and non-additive (hq) components for the alleles of gene (or quantitative trait locus, QTL) q possessed by genotype i. Similarly, we have partitioned the G*E interaction effect (ge)ij into an additive-by-environment effect (a e)qj and a non-additive-by-environment effect (he)qj for the alleles of gene q possessed by genotype i in environment j. Explicit examples of this extension of G*E interaction analysis from the genotypic to the QTL level are discussed by van Eeuwijk et al. (2002). From the above discussion, we recognize that predicting differences in genotypic performance or selection response requires an understanding of the genetic architecture of the traits, the multi-genic composition of the genotypes tested, and the types of environments and their frequencies of occurrence within the TPE (Cooper and Hammer 1996; Schlichting and Pigliucci 1998; Cooper 1999; Cooper and Podlich 2002; van Eeuwijk et al. 2002). Selection pressure directs desired changes in allele frequency in the germplasm pool and the creation of specific gene combinations. From a population improvement perspective, the objective is to increase the frequency of “favorable” alleles across the segregating loci that contribute to the genetic variation for the target traits (Chapman et al. 2003). At the theoretical level these changes in allele frequency may be considered as trajectories of genetic change within the genetic space available to the breeding program. This provides a quantifiable framework for modeling how effectively breeding programs achieve their objectives (Cooper et al. 2002). A major research challenge for breeding programs today is to combine molecular and phenotypic information to understand and efficiently achieve preferred genetic trajectories. The TPE may be international, national, local, or in some cases a series of non-contiguous regions with a common environmental constraint, e.g., drought, disease incidence, or acid soils. Comstock (1977) introduced the concept of a TPE to describe the range of environmental conditions encountered over time within the target region of a breeding program. This concept can be extended to quantify the TPE in terms of the major types of environments and their frequencies of occurrence within the defined target region (Cooper and Podlich 1999; Chapman et al. 2000a,b,c; Löffler et al. 2003). Environment classes may recognize the role of abiotic and biotic stresses in generating the important genotypeby-environment interactions that complicate selection (Chapman et al. 2002, 2003). Combining a characterization of the TPE with an understanding of the trait genetic architecture provides a basis for evaluating the merits of alternative selection (Cooper and Podlich 1999; Podlich et al. 1999) and breeding strategies (Podlich and Cooper 1998, 1999).

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

197

Selection pressure is applied assuming that there is a positive association between phenotypic performance and the value of the alleles segregating for the trait. As the complexity of the target genotypeenvironment system increases, e.g., in the presence of single or multiple abiotic stresses, the relationship between the phenotype and genotype becomes more complex. At the genic level, G*E interactions can change the value of alleles across different environmental conditions. Interdependencies between traits contributing to adaptive responses under stress conditions can result in epistatic interactions among genes such that the relative value of alternative alleles at one locus is conditioned by the allelic conformation of other loci (Holland 2001; Chapman et al. 2002). Therefore, the breeder’s ability to direct changes in allele frequency is highly conditional upon how phenotypic determinations are made. Values of selected genotypes within a large and diverse TPE are determined by a MET sample of the TPE. These MET samples can vary and their representativeness of the TPE can significantly influence genetic trajectories achieved by a breeding program (Cooper and Podlich 1999). Much work has been done on the influence of test environment on selection response. However, most breeding programs do not quantify the relationship between the conditions imposed in the MET test environments and those that make up the TPE. Breeding programs generally utilize environments that are readily available at a limited set of research stations. In the worse case scenario, research station environments may be atypical of on-farm conditions within the TPE and result in selection and genetic trajectories that do not optimize performance in the TPE (Cooper and Podlich 1999). One strategy to avoid these limitations is to utilize a large number of “randomly sampled” environments representative of the TPE. While this strategy will work (e.g., Duvick et al. 2004), it does not enable a knowledge-based breeding strategy by improving our understanding of the relationships between allelic variation and phenotypic performance in the TPE. An alternative strategy is to establish a set of well-characterized managed environments to represent the key types of environments within the TPE (Fischer et al. 1983, 1989; Cooper et al. 1995; Fukai et al. 1999). Given that drought-resistant genotypes must perform well in both stressed and non-stressed environments, and that the TPE will encompass non-stressed environments, we increase the likelihood that the trials cover such a range of environments by strategically manipulating available water throughout the growing season (see Section III B). As a prelude to the discussion of G*E interactions, experimental errors, experimental design and statistical analysis in maize breeding experiments, we give an overview, without derivation, of the key components

198

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

of a prediction framework for studying selection response in a TPE. See details in DeLacy et al. (1996). A basic prediction equation to estimate expected selection response is: 2

∆GM = iM hM sp(M)

[3]

where ∆GM is the predicted selection response based on the conditions sampled in the MET, iM is the standardized selection differential applied 2 on the phenotypes measured in the MET, hM is the heritability in the MET, and sp(M) is the standard deviation of the phenotype in the MET. Here we have used a subscript M to indicate that the j = 1,…, J environments sampled in the MET are taken to be a random sample of environments representing the distribution of environmental types that constitute the TPE. We may consider this as a basic equation in the sense that it is often used to predict expected improvements in broad adapta2 tion within the context of the TPE. Here the heritability component (hM ) is related to the model given in equation [1]. The line-mean heritability can be estimated in several ways (Nyquist 1991; Holland et al. 2002) by using appropriate estimates of variance components based on the specified linear model and its alternative forms. For example, for a given 2 set of genotypes created by the breeding program, line-mean hM(1) is defined as:

σ g2

2 hM (1) =

σ 2g

σ 2ge σ 2ε + + J JK

[4]

2

where hM (1) is a measure of degree of genotypic discrimination (or repeatability of the differences among the reference set of genotypes), s g2, s 2ge and s 2e are the genotypic, genotype-by-environment interaction and residual variance components, respectively, defined in relation to their respective effects included in equation [1], J is the number of environments sampled in the MET, and K is the number of replicates per envi2 ronment. Narrow-sense heritability (hM (2)) can be defined at the gene-effect level as:

σ α2

h2M (2) =

σ 2g

σ 2ge σ 2ε + + J JK

[5]

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

199

where s a2 is the additive component of genetic variance defined in relation to the additive effect in equation [2] and is a component of the total genotypic variance. Narrow sense heritability predicts the expected transmission of average gene effects across generations in a defined reference population. The purpose for introducing equations [3], [4], and [5] is to emphasize the critical influence that both G*E interaction and experimental error sources of variation have on the breeder’s capacity to predict differences among genotypes and therefore the selection response based on the results of J environments sampled in METs. 2 Throughout the following we consider the hM as a general case, recognizing key issues that impact predicted selection response at both the genic and genotypic levels, but without giving the explicit derivations for both cases represented by equations [1]–[4]. Following Falconer (1952) and later developments given in Falconer and MacKay (1996), an alternative form of the basic selection response equation has been given where response is the result of indirect selection. We are interested in the indirect response in the TPE based on selection in the MET, which is written as DGT|M such that: DGT|M = iMhThMrg(TM)sp(T)

[6]

where, iM is the standardized selection differential applied to the phenotypic data in the MET, hT and hM are the square root of heritability in the TPE and MET, respectively, rg(TM) is the genetic correlation between trait performance measurements in the TPE and MET, and sp(T) is the standard deviation of trait phenotypic measurements in the TPE. This representation of indirect selection response demonstrates its relationship to direct selection response and indicates its value within the quantitative framework for examining the effects of G*E interactions and experimental errors on predicting performance and selection response in target environments. A specific application of indirect selection occurs when selecting in managed environments of the MET to achieve a response to selection within target production environments of the TPE. Equation [6] allows explicit consideration of G*E interaction effects on the genetic correlation of trait performance measurements between pairs of environments or following classification of environment-types between pairs of sets of environment types (e.g., flowering water deficit, grain-filling water deficit, well-watered conditions) by estimation and inspection of rg(TM) (Cooper and DeLacy 1994; van Eeuwijk et al. 2001; Smith et al. 2002a). Equations [3] and [6] are similar since the genetic variance component from a multi-environment analysis of variance

200

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

represents an average of all of the genetic covariances between all possible pairs of environments that constitute the MET (Cooper and DeLacy 1994; Cooper et al. 1996). Therefore, with the appropriate model assumptions, equation [6] constitutes an expanded form of equation [3] whenever there is specific interest in the impact of G*E interactions on genetic correlations between the environments, and the conditions they represent, sampled in a MET. The impact of interactions on genotypic performance can be examined by a range of graphical tools developed within the broader family of pattern analysis methods. Using the framework discussed above, we now turn our attention to a more detailed consideration of G*E interactions and experimental errors in maize drought experiments. B. Experimental Design and Analysis Management to achieve a uniform plant water deficit across the experimental area is critical to optimize the design of drought METs. Analysis methods cannot compensate for poorly managed trials. However, there are several useful design and analysis techniques given the types of variation and non-uniformity that commonly occur even in the bestmanaged trials. 1. Extraneous Variation. Variation associated with experimental management and measurement techniques can be random or systematic; for example, systematic variation occurs when experiment operations are aligned with the rows or columns of a field trial. This is often referred to as extraneous variation (Gilmour et al. 1997). To illustrate this, Fig. 7.6A shows results of a trial where yield trends caused by the water management regime are apparent. Irrigation was supplied via drip tape for each column and charged from a pressurized supply manifold located at the bottom of the graphic. Due to frictional loss of water pressure in drip tapes along the columns, water delivery was slightly higher closer to the manifold. In water deficit treatments where plants were forced to use previously applied water during growth, variation in pre-stress irrigation supply resulted in differential plant water deficits down the length of the columns leading to consistently lower yields at the field end furthest from the irrigation supply (top of the graphic). There were also differences between the columns: the left side of Fig. 7.6A indicates a lower yielding section of the field in comparison to the right side. This gradient was likely due to differences in water supply among drip tapes or to gradients in soil water holding capacity. Such row and column effects are not as noticeable in the experiment shown in Fig. 7.6B.

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

A

B

50

49

Rep 1

41

Rep 2

5–7 32

16–17

35 15–16 33 31

3–5

17–18

39 37

35

29

43

10–12

8–10

Yield

Rep 2

38

45

Row

Row

41

Yield

Rep 1

47

47 44

201

14–15

29 26

0–2

27

23

13–14

25 1

3

5

7

9

Column

2

4

6

8

Column Irrigation Source

Irrigation Source

Fig. 7.6. Level plots of the raw yield data (tons per hectare) for two experiments, where darker colors indicate higher yielding plots. (A) An example where yield trends are introduced due to the water management regime; (B) An example where no trend is present.

Randomized complete block (RCB) designs help control differences between replicates by randomizing genotypes assigned to plots within each complete block. Other experimental designs better cope with extraneous variation (see Basford et al. 1996). A common alternative design for plant breeding is the resolvable alpha design (Patterson and Williams 1976). These are formed by dividing each complete replicate into sets of smaller blocks to adjust for heterogeneity between the smaller blocks. The designs can be created by the commercially available software, Alpha+ (Williams and Talbot 1993). This software also can be used to create row-column designs, which utilize two sets of smaller blocks in each complete replicate: one set that is aligned with the columns of the experiment and one that is aligned with the rows (Nguyen and Williams 1993; John and Williams 1995; Basford et al. 1996). Row-column designs

202

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

are often appropriate for plant breeding programs since many activities such as planting, harvesting, and applications of fertilizer, herbicide, and pesticide are aligned with rows or columns of the experiment and can cause extraneous variation aligned with these factors. 2. Spatial Variation. In addition to extraneous variation, experiments may exhibit spatial variation with global or local trends where phenotypic measurements are related to position in the field (Gleeson and Cullis 1987; Cullis and Gleeson 1991; Cressie 1993; Gilmour et al. 1998; Smith et al. 2002a). For a spatially correlated process, the correlation between phenotypic values is higher for adjacent plots than for plots far apart. Smith et al. (2002a) describe how global spatial trends can be accounted for by including appropriate model terms and appropriate covariance structures for the residual error to accommodate a local (or natural) stationary trend (Gilmour et al. 1997). A common model is the autoregressive covariance structure for residuals in the rows and columns (AR1*AR1) of an experiment (or with AR1 on only rows or columns), which expresses the covariance between plot residuals as a function of the number of rows and columns separating the plots (see Smith et al. 2002a for details). Variograms indicate the correlation of residuals for plots that are separated by an arbitrary number of rows and columns. The variogram for an independent process is constant for plots separated by any number of rows and columns above zero; such an ideal variogram was shown in Smith et al. (2002a). Analyses of the two experiments in Fig. 7.6 are shown in Table 7.4 as examples to illustrate concepts utilizing data from our work with drought treatments. This is not intended to be an exhaustive treatment; the references provided in the preceding paragraphs are appropriate to review for more detail. Table 7.4 displays results from alternative models: (i) RCB analysis, (ii) RCB analysis with random row and column effects, and (iii) RCB analysis with random row and column effects and the AR1*AR1 covariance structure for the residuals. Note that the models for data in the second experiment (Fig. 7.6B) do not include an effect for replication since this was not found to be a significant source of variation in preliminary investigations. For each model, the deviance and variance component values are shown in addition to the autoregressive coefficients. The deviance can be used to compare models that have the same fixed effects structure and nested random effects. For example, with a change in two degrees of freedom, a difference in deviance of 5.99 (95th percentile of Chi-Square distribution with 2 degrees of freedom) between models (i) and (ii) or between models (ii) and (iii) would be considered a

Table 7.4. Several analyses of the data in Fig. 7.6A and 7.6B are shown. Results from (i) an RCB analysis, (ii) an RCB analysis with random row and column effects, and (iii) an RCB analysis with random row and column effects and the AR1 * AR1 covariance structure for the residuals. The models shown in (iv-a) and (iv-b) are the final analyses of those considered. Ind = independent errors; AR1 * AR1 = autoregressive covariance among row and column residuals.

Data Fig. 7.6A

Fig. 7.6B

Model

Error

(i) Rep + Entry

ind

(ii) Rep + Row + Col + Entry

ind

(iii) Rep + Row + Col + Entry

AR1 * AR1

(iv-a) Rep + Col + Entry

AR1 * AR1

(i) Entry

ind

(ii) Entry + Row + Col

ind

(iii) Entry + Row + Col

AR1 * AR1

(iv-b) Entry + Col

ind

Deviance (df)

Vg (S.E.)

Ve (S.E.)

1839.05 (220) 1762.28 (218) 1755.11 (216) 1755.20 (217)

318.0 (115.0) 307.3 (90.5) 267.5 (82.5) 267.6 (82.3)

1065.0 (117.0) 565.9 (67.7) 687.8 (104.5) 718.2 (89.7)

1023.92 (165) 1013.48 (163) 1012.73 (161) 1013.51 (163)

168.56 (42.81) 173.60 (43.23) 170.39 (42.55) 173.92 (43.29)

99.10 (12.49) 83.69 (11.78) 81.50 (11.50) 84.53 (10.92)

R: Vrow (S.E.) C: Vcol (S.E.)

Autoregressive coefficients

NA

NA

R: 123.9 (61.1) C: 407.3 (229.8) R: 23.7 (61.1) C: 387.0 (227.2)

NA

C: 383.7 (226.2)

R: 0.21 (0.08) C: 0.24 (0.11) R: 0.22 (0.08) C: 0.27 (0.08)

NA

NA

R: 0.85 (4.85) C: 16.73 (12.35) R: 3.49 (5.41) C: 16.29 (11.90)

NA

C: 16.71 (12.36)

R: –0.05 (0.11) C: –0.11 (0.13) NA

203

204

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

significant improvement. For the data in Fig. 7.6A, the deviance indicates that model (iii) would be chosen among the three considered. This is not surprising given the column variance component and the two autoregressive parameters, which are both more than twice their standard errors. However, the variance component for the row factor in model (iii) is smaller than its standard error and it should be omitted from the statistical model. Results for this model are shown as model (iva) for the data from Fig. 7.6A. For the data used in Fig. 7.6B, the deviances indicate that model (ii) would be the best choice among the first three considered, but again the row factor should be omitted from the model, i.e., as in model (iv-b). While model selection can be based on quantitative measures (Table 7.4), the variogram is useful to evaluate and select alternative models. Figs. 7.7A and 7.7B show variograms corresponding to the first model (i) and final models (iv-a, iv-b) for each experiment. In Fig. 7.7A (i), differences among the columns of the experiment are reflected by the pat-

A

B Model (i)

Model (i) 120

1600 1400 1200 1000 800 600 400 200 0

100 80 60 40 6 5

20

15 2 5

1 00

4 14

3

10

Row

5

0

4

n lum Co

12

3 10

Row

8

6

2 4

2

1 00

n lum Co

Model (iv-b)

Model (iv-a) 120

1600 1400 1200 1000 800 600 400 200 0

100 80 60 40 6 5 4

15

3

10

Row

2 5

1 00

n lum Co

20

5

0

4 14

12

3 10

Row

8

6

2 4

2

1 00

n lum Co

Fig. 7.7. Variograms corresponding to the first model (i) and final models (iv-a, iv-b) for the yield data displayed graphically in Figures 7.6A and 7.6B.

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

205

tern in the plot factor. This trend is removed in the variogram (Fig. 7.7A (iv-a)) corresponding to the model (iv-a) and is much closer to the “ideal” variogram that is flat at all row and/or column lags above zero. Fig. 7.7B (i) and (iv-b) display variograms for the first and final models applied to the data in Figure 7.6B. In this case the variogram in Fig. 7.7B (iv-b) corresponds to the best choice from among the four models, although the difference between the two variograms is not nearly as striking as for Fig. 7.7A. Fig. 7.8 contains scatter plots showing the impact that improved experimental analysis has on the best linear unbiased predictors (BLUPs) of genotype yield for these data. Figs. 7.8A and 7.8B are graphs of the BLUPs from the final models in Table 7.4 (models (iv-a) and (iv-b)) versus the BLUPs based on the initial models (model (i)) from data shown in Fig. 7.6A and 7.6B, respectively. From Fig. 7.8, it is evident that there has been a greater impact on the adjusted means under model (iv-a) using the data from Fig. 7.6A. These examples are intended to show an application to drought research where experimental design and analysis helped reduce experimental error. Given that extraneous variation and spatial trends are common in drought experiments and drought-affected METs, maize breeders will be assisted by combining the use of incomplete block experimental designs (e.g., alpha and row-column designs) and appropriate spatial analysis methodology to adjust the raw data for sources of extraneous variation and global and local trends. Ultimately this will A

B 17 BLUPs from model (iv-b)

BLUPs from model (iv-a)

6.0 5.5 5.0 4.5 4.0 3.5

16

15

14

3.0 3.0

3.5 4.0 4.5 5.0 5.5 BLUPs from model (i)

6.0

14 15 16 BLUPs from model (i)

17

Fig. 7.8. (A) Scatter plot of the yield BLUPs from the final model in Table 7.4 (model (iv-a)) versus the BLUPs based on the initial model (model (i)) using the data shown graphically in Fig. 7.6A. The line y = x, where the two sets of BLUPS would be identical is shown for reference. (B) Scatter plot for the data shown in Fig. 7.6B. It is evident that there has been a greater impact on the adjusted means under model (iv-a) using the data from Fig. 7.6A.

206

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

reduce the influence of experimental error and increase trait heritability estimates (equations [4] and [5]). Increased heritabilities have consequences for the predicted selection response (see Section III A). For example, using equation [5] to compute heritability for data in Fig. 7.6A on a single environment and replicate basis, we estimate 1.26/(1.26 + 4.23) = 0.23 under model (i) and 0.27 under model (iv-a). Bänziger and Lafitte (1997) reported similar analysis effects when conducting selections under low N. Improved experimental analysis increases both heritability and predicted selection response (equation [3]). In the previous examples experiments were designed to achieve maximum gains from the row-column and spatial analyses. However, it is possible to include factors that account for systematic extraneous variation in experiments that were not designed as such through a procedure analogous to the concept of post-blocking. Qiao et al. (2000) used empirical results from a long-term wheat study and Smith et al. (2002b) used theoretical arguments and simulation to show that analysis adjustments to the raw data are expected to improve the correlation between the estimates obtained from analysis of METs and the true values in the target environments and thus the TPE. These findings support the argument that experimenters will generally better understand genotypic effects and G*E interactions using data that has been adjusted for natural and extraneous variation. C. Describing, Understanding, and Managing G*E in Drought Trials Once an optimal analysis is identified for each environment, this information can be incorporated into a multi-environment analysis using appropriate mixed models. In such analyses there are a number of options for modeling the variance-covariance structure between environments (van Eeuwijk et al. 2001). Compound symmetry, also called uniform correlation, is the default variance-covariance structure in many statistical analysis packages. This structure implies that the genetic variance is the same for each environment and that the genetic correlation is the same for each pair of environments. These assumptions are generally not true and rarely appropriate especially when environments are substantially different, such as combinations of well-watered and drought-stressed environments. One alternative to the compound symmetry structure is the unstructured variance-covariance matrix. This allows a different genetic variance for each environment and a different genetic correlation for each pair of environments. However, the variance-covariance matrix has a large number of parameters that can cause difficulty for model fitting. van Eeuwijk et al. (2001) describe a class of mixed multiplicative mod-

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

207

els or factor-analytic models that allow genetic variances and covariances to be dependent on some (unobserved) quantitative descriptors of the environments. This class of models is flexible, and contains fewer parameters than the unstructured variance-covariance matrix. To illustrate multi-environment analysis techniques, we use data from a trial containing 54 hybrids evaluated in two replicates of six different environments. Duvick et al. (2004) have discussed these ERA hybrids within the context of long-term improvements for yield from a commercial breeding program. The objective of our experiment was to study the ERA hybrids under specific drought and density-stress conditions. Instead of relying on a set of randomly sampled environments to capture the diverse and hard-to-predict drought environments, water and density treatments were imposed to help assure that our MET contained the key environments of interest. As key environment types and their frequencies of occurrence in the TPE are better understood, managed stress experiments like this one may reduce some trial-and-error or augment METs that are based on “random” samples of environments from the TPE. In the design of any targeted MET regime, consideration should be given to the genetic correlation between trait performance of genotypes in the managed environments and the production environments in the TPE as emphasized in equation [6]. The six environments in this experiment were combinations of two different density treatments (High Density = 91,000 plants/ha, Low Density = 46,000 plants/ ha) and three different irrigation treatments (well watered, flowering drought stress, and grain filling drought stress). This experiment was designed as a row-column design in each treatment, but initial analyses indicated that there were no significant design effects or spatial correlation (Fig. 7.9). Therefore, all subsequent analyses use a RCB analysis. We have found this to be the exception rather than the rule and recommend the use of improved experimental designs for most drought studies. For example, a row-column design was needed in the second year of this experiment (data not shown). The form of the final model used for the analysis of these data is: yijkl = m + Dj + Wk + (WD)jk + (r/WD)jkl + gi + (gD)ij + (gW)ik + (gWD)ijk + eijkl where yijkl represents the phenotypic data (e.g., yield) for the l th replicate of the i th genotype in density (D) treatment j and water (W) treatment k. Random effects are underlined. Note that we classified hybrids as random effects (Smith et al. 2002a). In practice, consideration should be given to appropriate classification of fixed vs. random effects given the structure and objectives of an experiment. The choice of random hybrid effects is somewhat conservative since the resulting BLUPs of

208

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

High

49 47

High

Low

Rep 1

45 Yield

43

Rep 2

41

14–17

Row

39 37

11–14 Rep 1

35 33

8–10

31 Rep 2

29

4–7

27

Low

1–4

25

Low

High

Well-Watered

Filling Stress

Flowering Stress

23 55

59

63

67

71

75

79

83

87

91

95

Column Fig. 7.9. Level plot of raw yield data (tons per hectare) for experiment with 54 hybrids in six different environments (combination of high density = 91,000 plants/ha and low density = 46,000 plants/ha with flowering stress, filling stress, and well-watered treatments), with darker shades of gray indicating higher-yielding plots.

hybrid means are more “shrunk” back towards an overall mean than if hybrids were fixed effects. Preliminary investigations showed that a uniform correlation structure between environments, which is the default in many software applications, was appropriate for these data. This is due to strong genotypic effects that exist in this set of hybrids spanning the years 1930–2002 and would not be expected for data sets where genotypes were more alike. Fig. 7.10 shows scatter plots and high correlations between the BLUPs for each pair of environments included in this managed stress study. The large genetic variance attributed to improvements in grain yield over time also contributes to higher heritability for yield (0.89). Sophisticated graphical tools have been developed to visualize the genotype-environment matrices of trait data generated from METs, e.g., the biplot (Kroonenberg 1995; DeLacy et al. 1996; Yan and Kang 2003). The biplot in Fig. 7.11A

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

8

11 14

0.90

0.91

0.93

0.87

0.82

High Fill

0.93

0.86

0.92

0.80

High WW

0.94

0.94

0.91

Low Flower

0.91

0.89

Low Fill

0.93

12 16

15

6 8

11

4 6 8

8

4 6 8

2

High Flower

4 6 8

6

4 6 8

209

8 11

Low WW 2 4 6 8

8

12 16

6

8 10

Fig. 7.10. Scatter plots and correlations between the yield BLUPs for each pair of environments defined and shown in Fig. 7.9. High = High density; Low = Low density; Flower = Flowering drought stress; Fill = Grain filling drought stress; WW = Well-watered control.

contains all 54 ERA hybrids and the primary feature is an increase in average yield in all six environments for later release dates. Fig. 7.11B contains only those hybrids with a release date of 1990 or later. There appears to be one environment that has a particularly low genetic correlation with the other five environments. In this experiment there are several hybrids that perform differently in the high density, grain filling stressed environment than in the other five environments. The reasons for this reduced genetic correlation are a potential subject for further investigation This discussion and example has emphasized appropriate designs and analyses for a MET and graphical interpretation of subsequent results. With this particular data set, we can evaluate how representative the drought MET is of the TPE since these hybrids have been tested extensively in a number of experiments from the TPE (Duvick et al.

210

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

A All Data (1930 – 2002) 33P66_1997

3

3378_1983 3335_1995

Dimension 2 (3.8%)

2 339HYB_1942 336HYB_1940

LD/W Watered

317OLD_1937 351OLD_1934

1

3377_1982 3376_1965 3206_1962 3541_1975 34N44_2002 307HYB_1936 3306_1963 347HYB_19503301A_1974 34B23_1999 3334_1969 33R77_2001 LD/Filling ReidYD_1930 3279_1992 340HYB_1941 301B_1952 329HYB_1954 328HYB_1959 354A_1958 LD/Flowering 330HYB_1939 344HYB_1945 352HYB_1946 KrugYD_1930 322HYB_1936

0

HD/W Watered HD/Flowering

3390_1967 3379_1988 354_1953 3388_1970 3489_1994 3571_1968 3382_1976 3529_1975 3366_1972 3475_1984 3417_1990 3618_1961 350B_1948 3517_1971 34G13_2000 34M95_2001 33P67_1999 34H31_2002

–1

HD/Filling

3394_1991 3362_1989 34G81_1997 33G26_1998 3563_1991

–2 –2

–1

0 1 2 Dimension 1 (91.6%)

3

B 1990-2002 Data 3 HD/Filling

Dimension 2 (17.2%)

2

33G26_1998

34M95_2001

1

33R77_2001

HD/Flowering

3394_1991 34H31_200234B23_1999 3563_1991 3279_1992

HD/W Watered

33P67_1999 34G13_2000 3489_1994 3417_1990

0

3335_1995

34G81_1997 34N44_2002

LD/Filling LD/W Watered

–1 LD/Flowering

–2 33P66_1997

–3 –3

–2

–1 0 1 Dimension 1 (58.4%)

2

3

Fig. 7.11. Biplots of the yield BLUPs for the hybrids from Figure 7.9. (A) Contains all 54 hybrids and it can be seen that the primary feature is an increase in average yield, in all six environments, for later release dates. (B) Contains only those hybrids with a release date of 1990 or later.

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

211

14

Grain Yield Johnston YldOpt (t/ha)

Flowering_YldOpt Filling_YldOpt Well-Watered_YldOpt

12

10

8

6

4 1:1 2 2

4

6

8

10

12

14

16

18

Grain Yield Woodland YldOpt (t/ha)

Fig. 7.12. Scatter plot of the yield BLUPs of each hybrid under optimal density conditions (YldOpt) for each water regime versus the BLUPs obtained in Duvick et al. 2004, which are representative of the TPE. Flowering = Flowering drought stress; Filling = Grain filling drought stress; Well-watered = low stress control. Data from Johnston, Iowa, and Woodland, California.

2004). Fig. 7.12 contains a scatter plot of BLUPs under optimal density conditions for each water regime versus BLUPs obtained under optimal density, i.e., the density under which the hybrids achieved their highest yield, in Duvick et al. (2004), which are representative of the TPE. This comparison suggests that grain yield of the hybrids under managed drought regimes is positively associated with grain yield in target environments of the U.S. Corn Belt. D. Hybrid Characterization: Measures of Drought Tolerance 1. Need for Managed Stress Approach. One challenge in dissecting the basis for genotypic differences in drought tolerance is the identification and management of uniform and reproducible water deficit environments. Historically, the use of many yield test locations within a breeding program was assumed to incorporate a percentage of these environments. In practice, this has proven challenging due to unpredictable rainfall

212

BARKER, CAMPOS, COOPER, DOLAN, EDMEADES, HABBEN, SCHUSSLER, WRIGHT, & ZINSELMEIER

patterns in the U.S. Corn Belt and a natural bias to conduct yield trials in areas with high yield potential. An alternative approach to extensive sampling is to develop managed stress locations with carefully designed reproducible levels of timing, duration and intensity of water deficits. This is best accomplished by choosing locations with historically low rainfall during the maize growing season. Ideally, these locations will also possess deep, well-drained soils with uniform water holding capacity in the maize root zone. These factors are critical since small differences in uniformity lead to large differences in plant water deficits under drought stress resulting in unmanageable levels of variability for yield and secondary traits of interest. A dependable high-quality water source is also necessary to control prestress irrigation and “load” soils to specific levels of water holding capacity and to provide adequate irrigation for irrigated control plots. Extreme precision in irrigation application is required, necessitating the use of drip tape or overhead irrigation systems. Knowledge of evapotranspiration rates and hybrid phenology must be combined in order to “stage” plants for specific drought stress levels at specific developmental stages (see Section II C). This can only be accomplished on a routine basis in relatively rare environments where sunshine, temperature, and wind are fairly constant during a growing season and across growing seasons. Preliminary studies to confirm the value of potential sites, such as the example depicted in Figure 7.9, should be conducted for 1–3 years prior to undertaking managed stress studies. 2. Execution of Managed Stress Approach. Drought tolerance can be realized through a variety of putative mechanisms occurring at various plant growth and development stages (see Section II D). Prioritization of drought tolerance trait targets and selection strategies must be dictated by (1) the relative predominance of drought timing and severity in the TPE, and (2) the probability of genetic gain due to utilization of natural variation or candidate transgenes. Flowering Drought Stress. Due to the determinate flowering habit of maize, water deficits occurring during anthesis can be severe (see Section II C and D 3). Ear development and associated silk exertion can be significantly delayed while tassel development and pollen dehiscence are far less affected. Increased ASI under flowering drought stress results in a reduction in the number of pollinated ovaries since some silks do not emerge at all or are delayed beyond the window of pollen availability. Selection for reduced ASI in combination with grain yield has been effective for increasing drought tolerance in maize (see Section II C). In order to characterize hybrid differences for this trait, water must be

7. IMPROVING DROUGHT TOLERANCE IN MAIZE

213

withheld such that the plants utilize most of the stored soil water prior to flowering. An appropriately timed drought stress treatment will result in plants that are shorter, exhibit leaf rolling, lower leaf senescence, and increased ASI and barrenness compared to fully irrigated plants. This results in greater plant-to-plant variability, an indication that some individuals within the population cannot compete with their neighbors. In contrast, hybrids tolerant to flowering drought stress typically maintain more uniform plant-to-plant growth with little barrenness. Other within-ear traits also affect reproductive efficiency. These include the rate of silk exertion, length of silk receptivity under plant water deficits, and amount of post-pollination kernel abortion. These processes establish yield potential during the critical anthesis period. Under managed flowering drought stress conditions, kernel number per ear is highly correlated with grain yield per plant (Fig. 7.13). Data are from individual plants of two different hybrids exposed to different levels of flowering stress. Data at the low end of the scale are from severely stressed plants, data in the middle are from plants receiving a less severe drought stress at flowering, and data at the highest point of the scale are from plants in the fully irrigated control environment. Kernel number per ear increased as stress was minimized.

250

r2 = 0.98, P