WORLD BANK DISCUSSION PAPER N O. 367 WorK in progress W DP! 36 for public discussion Easing Barriers to Mlovement of Plant Varieties for Agricultural lD)evelopment Ji (/j/t/ . / Recent World Bank Discussion Papers No. 298 Uganda's AIDS Crisis: Its Implications for Development. Jill Armstrong No. 299 Towards a Payments System Lawfor Developing and Transition Economies. Raj Bhala No. 300 Africa Can Compete! Export Opportiunities and Challengesfor Garments and Home Produlcts in the European Market. Tyler Biggs, Margaret Miller, Caroline Otto, and Gerald Tyler No. 301 Review and Outlookfor the World Oil Market. Shane S. Streifel No. 302 The Broad Sector Approach to Investment Lending: Sector Investment Programs. Peter Harrold and Associates No. 303 Institutional Adjuistment and Adjusting to Institutions. Robert Klitgaard No. 304 Putting Institutional Economics to Work: From Participation to Governance. Robert Picciotto No. 305 Pakistan's Public Agricultutral Enterprises: Inefficiencies, Market Distortions, and Proposalsfor Reform. Rashid Faruqee, Ridwan Ali, and Yusuf Choudhry No. 306 Grameen Bank: Performance and Stability. Shahidur R. Khandker, Baqui Khalily, and Zahed Khan No. 307 The Urugutay Rouind and the Developing Economies. Edited by Will Martin and L. Alan Winters No. 308 Bank Governance Contracts: Establishing Goals and Accountability in Bank Restruicturing. Richard P. Roulier No. 309 Public and Private Secondary Edutcation in Developing Countries: A Comparative Study. Emmanuel Jimenez and Marlaine E. Lockheed with contributions by Donald Cox, Eduardo Luna, Vicente Paqueo, M. L. de Vera, and Nongnuch Wattanawaha No. 310 Practical Lessons for Africafrom East Asia in Industrial and Trade Policies. Peter Harrold, Malathi Jayawickrama, and Deepak Bhattasali No. 311 The Impact of the Uruigutay Rouind on Africa. Peter Harrold No. 312 Procuirement and Disbuirsement Manualfor Projects with Community Participation. Gita Gopal No. 313 Harnessing Informationfor Development: A Proposalfor a World Bank Grouip Strategy. Eduardo Talero and Philip Gaudette No. 314 Colombia's Pension Reform: Fiscal and Macroeconomic Effects. Klaus Schmidt-Hebbel No. 315 Land Quality Indicators. Christian Pieri, Julian Dumanski, Ann Hamblin, and Anthony Young No. 316 Sustainability of a Government Targeted Credit Program: Evidencefrom Bangladesh. Shahidur R. Khandker, Zahed Khan, and Baqui Khalily No. 317 Selected Social Safety Net Programs in the Philippines: Targeting, Cost-Effectiveness, and Optionsfor Reform. Kalanidhi Subbarao, Akhter U. Ahmed, and Tesfaye Teklu No. 318 Private Sector Development During Transition: The Visegrad Countries. Michael S. Borish and Michel Noel No. 319 Edutcation Achievements and School Efficiency in Rural Bangladesh. Shahidur R. Khandker No. 320 Household and Intrahouisehold Impacts of the Grameen Bank and Similar Targeted Credit Programs in Bangladesh. Mark M. Pitt and Shahidur R. Khandker No. 321 Clearance and Settlement Systemsfor Securities: Critical Design Choices in Emerging Market Economies. Jeff Stehm No. 322 Selecting Development Projectsfor the World Bank. Jean Baneth No. 323 Evaluating Public Spending: A Framework for Putblic Expenditutre Reviews. Sanjay Pradhan No. 324 The Bangladesh Rural Advancement Committee's Credit Programs: Performance and Sustainability. Shahidur R. Khand- ker and Baqui Khalily No. 325 Institutional and Entrepreneutrial Leadership in the Brazilian Science and Technology Sector:Setting a New Agenda . Edited by Lauritz Holm-Nielsen, Michael Crawford, and Alcyone Saliba No. 326 The East Asian Miracle and Information Technology: Strategic Management of Technological Learning. Nagy Hanna, Sandor Boyson, and Shakuntala Gunaratne No. 327 Agricutltural Reform in Riussia: A Viewfrom the Farm Level. Karen Brooks, Elmira Krylatykh, Zvi Lerman, Aleksandr Petrikov, and Vasilii Uzun No. 328 Insuring Sovereign Debt Against Default. David F. Babbel No. 329 Managing Transboundary Stocks of Small Pelagic Fish: Problems and Options. Max Aguiero and Exequiel Gonzalez No. 330 China: Issutes and Options in Greenhoiuse Gas Emissions Control. Edited by Todd M. Johnson, Junfeng Li, Zhongxiao Jiang, and Robert P. Taylor No. 331 Case Studies in War-to-Peace Transition: The Demobilization and Reintegration of Ex-Combatants in Ethiopia, Namibia, and Uganda. Nat J. Colletta, Markus Kostner, Ingo Wiederhofer, with the assistance of Emilio Mondo, Taimi Sitari, and Tadesse A. Woldu No. 333 Participation in Practice: The Experience of the World Bank and Other Stakeholders. Edited by Jennifer Rietbergen- McCracken (Continuted on the inside back cover) WORLD BANK DISCUSSION PAPER NO. 367 Easing Barriers to Movement of Plant Varieties for Agricultural Development Edited by David Gisselquist Jitendra Srivastava The World Bank Washington, D.C. Copyright © 1997 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing July 1997 Discussion Papers present results of country analysis or research that are circulated to encourage discussion and comment within the development community. To present these results with the least possible delay, the typescript of this paper has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. ISSN: 0259-210X David Gisselquist is a consultant in the World Bank's International Economics Department. Jitendra Srivastava is principal agriculturalist in the Agriculture and Forestry Systems Division in the World Bank's Agriculture and Natural Resources Department. Library of Congress Cataloging-in-Publication Data Easing barriers to movements of plant varieties for agricultural development / edited by David Gisselquist, Jitendra Srivastava. p. cm. - (World Bank discussion paper ; no. 367) Papers from an international workshop organized by the World Bank, June 1995. Includes bibliographical references. ISBN 0-8213-3991-5 1. Seeds-Government policy-Developing countries-Congresses. 2. Seed technology-Developing countries-Congresses. 3. Seed industry and trade-Developing countries-Congresses. 4. Seed distribution-Developing countries-Congresses. I. Gisselquist, David. II. Srivastava, Jitendra, 1940- . III. World Bank. IV. Series: World Bank discussion papers ; 367. SB114.D44E37 1997 338.1'8-dc2l 97-23700 CIP Contents List of tables, figures, and boxes v List of abbreviations vi Foreword vii Acknowledgments viii Abstract ix 1. Workshop Overview and Recommendations I David Gisselquist 2. Regulations and Recommendations for Reform 19 Comparing EU and US Seed Regulatory Systems 21 Jim Elgin Recommendations to Increase the Flow of Seed Technology to Farmers in Developing Countries 23 Mark Condon Recommendations for National Private Seed Industry Development 27 Federico Poey Three Main Barriers: Weak Protection for Intellectual Property, Unreasonable Phytosanitary Rules, and Compulsory Variety Registration 29 Jon Geadelmann World Phytosanitary System: Problems and Solutions 33 Dena McGee Seed Regulatory Frameworks and the Availability of Crop Varieties 39 Robert Tripp Streamlining for a Demand-Driven Seed Sector: A Framework for Policy Reform 53 Jitendra 1? Srivastava iii CONTENTS 3. Supply of New Varieties 59 Prospects and Constraints for Hybrid Maize in Developing Countries 61 Roberto WAnsaldo and Ray Riley An International Strategy for Cotton Breeding, Variety Introduction, and Seed Supply 71 Donald A. Pallin Assessment of the International Transfer of Wheat Varieties 73 Mywish Maredia, Richard Ward, and Derek Byerlee Genetic Resources, International Organizations, and Rice Varietal Improvement 87 -E. Evenson and D. Goltin 4. Country Experiences with Reform 103 Seed Industry Regulations and Seed Industry Development in India 105 Pramod K Agrawal Impact of Turkey's 1980s Seed Regulatory Reforms 113 David Gisselquist and Carl Pray Reform, Regulations, and Recent Developments in the Seed System in Peru 133 Joseph Cortes 1990 Seed Policy Reform in Bangladesh 139 David Gisselquist Annex 1: Workshop Program 145 Annex 2: List of Participants 147 Annex 3: List of Other Papers Presented and Distributed at the Workshop 149 iv List of Tables, Figures, and Boxes Tables 1.1 Regulatory barriers and recommended reforms 3 1.2 Government options to limit farmer risks 16 2.1 Comparing EU and US seed regulations 21 2.2 Proposed database on seed-borne diseases 35 2.3 Seed regulatory deficiencies affecting the availability of crop varieties 49 3.1 African and Asian maize area by grain color and ecology, ca 1990-91 66 3.2 Asia hybrid maize area, 1992 66 3.3 Hybrid adoption to meet Asian maize consumption in 2003 68 3.4 Classification of spring wheat megaenvironments (MEs) used by CIMMYT wheat program 75 3.5 Regression results of potential spillovers at the megaenvironment level using ISWYN data, 1980s 78 3.6 Estimated spillover matrix for wheat improvement research at the global megaenvironment level 80 3.7 Varieties released in developing countries, classified by type of technology transfer embodied, 1966-90 81 3.8 Percent of wheat varieties released in developing countries by type of technology transfer and region, 1965-90 82 3.9 Classification of NARS by the extent of released varieties based on direct transfers, 1966-90 82 3.10 Numbers of varieties included in the data set, by country of release and by time period of release 91 3.11 International genetic resources flows by time period 92 3.12 Matrix of varietal borrowing 93 3.13 Matrix of parental borrowing 95 3.14 Routes of varietal release-descriptive statistics 98 4.1 Seed distribution of field crops of improved cultivars in India 107 4.2 Seed requirement and availability for field crops, 1992-93 108 4.3 Market analysis of hybrid vegetable seeds in India, 1993-94 110 4.4 Varieties introduced by year and crop, 1982-1994 116 4.5 Commercial Seed Production, 1980-1993 118 4.6 Seed production plans by type of firm, 1994 119 4.7 Seed exports and imports, 1980-1993 121 4.8 Maize yield response function, 1961-1991 123 4.9 Calculating net benefit from hybrid maize, 1990-92 123 4.10 Vegetable production and export, 1979-1992 126 4.11 Number of private seed enterprises in Peru, 1988-1995 134 4.12 Characteristics of seed enterprises by department, 1995 135 4.13 Source of cultivars in use in Peru 136 4.14 Volume and value of US seed exports to Peru, 1991/92-1993/94 137 V LIST OF TABLES, FIGURES. AND BOXES Figures 3.1 Data driven decisions 63 3.2 Multi-location testing to evaluate performance 63 3.3 African maize ecology map 65 4.1 Hybrid maize Seed Production, Imports, and Exports, 1980 to 1992 120 4.2 Hybrid Sunflower Seed Production, Imports, and Exports, 1980 to 1992 120 4.3 Maize Yield: Actual and Projected, 1961 to 1991 124 4.4 Maize Yields in the Mediterranean Region, 1961 to 1991 124 Boxes 2.1 Most common regulatory barriers to seed movement 54 2.2 Global trends in the seed industry 55 2.3 Needs of seed producers 56 Abbreviations ADB Asian Development Bank CIAT Centro Internacional de Agricultura Tropical CIMMYT Centro Internacional de Mejoramiento de Maiz y Trigo IARC international agricultural research centers NARS national agricultural research system OPV open pollinated variety PVP plant variety protection SOE state owned enterprises TEBD Seed Industry Association t/ha tons per hectare TIGEM Agricultural State Enterprises TZDK Agricultural Supply Organization WB World Bank vi Foreword The papers in this document come out of research agencies in the largest developing an international workshop on seed policies held countries. Even when government research is at the World Bank in June 1995. Papers and effective and well-funded, improving farmer discussions at the workshop-with prominent access to private varieties can boost farm participation by representatives from private productivity and incomes. In particular, success seed companies and developing country govern- with many minor crops, including high value ments-have strong implications for Bank horticultural crops such as fruit trees, vegeta- advice on seeds and agricultural technology bles, and flowers, is severely constrained, if not transfer. Recommendations from the workshop impossible, without the active participation of ask donors and governments to work together private seed companies to deliver the latest to revise regulatory and policy obstacles to world-class technology. private technology transfer in agriculture, Deregulation will be particularly important induding especially obstacles to variety intro- for smaller and poorer developing countries, ductions. such as in Africa and Central America. In these Good policies would allow private compa- countries most technology must come from nies of all sizes to assist with technology transfer outside, but markets are too small to persuade and seed distribution to promote higher yields companies to invest the time and money neces- and crop diversification. Although government sary to overcome regulatory barriers. research agencies in many countries have had a Similarly, for countries in East and Central major impact on farm-level technology, particu- Europe, reducing controls on variety transfer larly for major crops, maintaining an adequate can facilitate integration of their technically flow of new technology stretches the resources advanced but isolated seed industries into the of even the most successful government international seed industry. Masood Ahmed Director International Economics Department vii Acknowledgments From planning the workshop through Forno, Alan Winters, and Yahaya Doka. preparation of this volume of collected papers, Vandana Pradhan provided crucial administra- we are indebted to many friends and associates tive support during preparation of the work- for advice and support. John Nash, Carl Pray, shop. We also thank the many people who con- Jitendra Srivastava, and David Gisselquist guid- tributed to the workshop as chairpersons, pre- ed initial plans for the workshop. Three divi- senters, discussants, and attendees, as listed in sions in the World Bank - Agriculture and annexes 1 to 3. Special thanks go to Jennifer Forestry Systems, International Trade, and Ngaine, Sarah Lipscomb, and Jeff Hayden, who Training - cooperated in planning and fund- prepared tables, diagrams, and text for publica- ing the workshop, with leadership from Doug tion. viii Abstract Private seed trade can be an important Foreign varieties are available for interna- channel to deliver new agricultural technology tional borrowing when governments allow. -especially new varieties - to farmers in Even with pervasive government limits on developing countries. However, many devel- private variety introductions, a large share of oping countries enforce seed regulations and improved varieties released in developing coun- other policies that obstruct private companies tries comes directly from foreign crosses, and from operating and delivering new technology. most of the rest depends at least in part on The World Bank organized an international foreign parents. Papers at the workshop demon- workshop in 1995 to review seed policies and strate that roughly half of the wheat and a to develop recommendations to ease barriers to quarter of the rice varieties which government variety introductions into developing countries. research agencies have approved and released in This volume presents recommendations and developing countries have come direcdy from selected papers from the workshop. foreign crosses by IRRI, CIMMYT, and other Papers and discussions at the workshop governments. Private companies similarly locate identified reforms to speed the flow of private breeding in selected countries, distributing lines seed technology into developing countries. Key to other countries for testing and introduction recommendations advise governments to: (a) (depending on market opportunities). Papers at work more with the private seed industry; (b) the workshop describe private company strate- allow private companies to introduce new vari- gies for maize and cotton. eties without prior government approval (vol- Although many developing governments untary variety registration); (c) make seed certi- severely limit private seed company activities, fication optional rather than compulsory; (d) policies are not uniform. From the early 1980s, give private companies access to public varieties an increasing number of developing countries and breeding lines; (e) establish legal processes have been allowing private seed companies to for companies to register ownership of varieties; introduce new varieties and otherwise con- and (f) focus phytosanitary controls on realistic tribute to agricultural development. Papers at pest and disease threats, and remove other non- the workshop described seed policies, reforms, tariff barriers on seed imports. and impacts in India, Turkey, Peru, and Bangladesh. ix Workshop Overview and Recommendations David Gisselquist The purpose of this volume is to discuss regulations governing the trade of plant varieties and to recommend steps governments could take to facilitate the movement of new plant varieties to farmers in developing countries. Recommendations are based on papers presented and discussed at the workshop, Easing Barriers to Movement of Plant Varieties for Agricultural Development. The workshop, sponsored by the World Bank, brought together represen- tatives from private seed companies, governments, and donor organizations, along with consultant and university agricultural experts.1 Participants from private companies played a large part in workshop deliberations, challenging others to recognize difficulties in existing seed policies. Sections 2 through 4 of this volume present selected papers from the workshop dealing with regulations and recommended reforms, international sources for new seed technology, and country experiences with reform. Seed policy reforms proposed in this volume are partic- ularly important for Africa, a continent of relatively small countries, each of which has limited funds for public research and also small seed markets that, taken alone, are David Gisselquist is a consultant with the World Bank's International Economics Department. 2 WORKSHOP OVERVIEW AND RECOMMENDATIONS of limited interest to private companies. Compulsory variety registration Recommendations from the workshop, as out- Governments of many developing countries lined in this document, challenge African strictly regulate introduction of new varieties, governments to open borders to new varieties, banning sale of seeds until some government giving farmers immediate access to regional agency has tested performance of the variety technology, and also creating larger regional (often for two to six or more years) and a gov- markets that will attract private domestic and emient committee reporting to the minister of foreign companies to enter and compete. agriculture has approved the variety. Performance tests address the question of Policy obstacles and recommendations for whether or not the candidate new variety has reform value for farmers in cultivation and use. Workshop participants described policy Performance or VCU tests often compare the obstacles to movement of plant varieties - and candidate variety in side-by-side field trials with especially movement through private seed com- standard or check varieties. panies - and recommended regulatory reforms Governments that control the introduction (see papers in Section 2). This sub-section sum- of new varieties maintain lists of allowed vari- marizes the obstacles and recommendations and eties, adding to these lists as they approve new organizes them into eight general topics (see varieties. Many governments use variety also table 1.1). controls to block (or delay) entry of private varieties, approving only varieties coming out of Lack of government contacts with private seed public sector research programs. Often, lists of industry approved varieties are short and filled with rela- The private international seed industry tively old varieties. In Ghana, for example, the offers research, market access, policy advice, National Seed Committee as of 1985 allowed and training through multiple channels, includ- only five improved varieties for maize, none for ing private seed companies and also national cassava, and two for sorghum; in Malawi, the and international seed industry associations. 1991 seed list allows eleven varieties for maize The industry works with multinational organi- (only one of which is from a private breeding zations such as the OECD Seed Schemes and program), four for soybean, and one each for International Seed Trade Federation to manage sorghum and pearl millet. Typically, many listed seed trade and other aspects of the private seed varieties are of interest to farmers, and on the industry. The private seed industries in develop- other hand popular or potentially popular ing countries, however, are often small, isolated improved varieties have been left off the list (for from the international industry, and absent example, pajam rice in Bangladesh or magoye from domestic seed policy deliberations. soybeans in Malawi). For unlisted varieties, farmer access to seed is limited to smuggling or Recommendation 1: To foster stronger pri- to informal farmer-to-farmer trade. vate seed industries in developing countries, Even when governments manage systems of governments are encouraged to establish compulsory variety registration so as to allow formal contacts with their own private private companies to enter the market, costs to domestic seed industries, to bring domestic register new varieties (fees, time, etc) limit the and international seed companies into seed number of varieties that companies introduce. policy discussions, and to join supporting When this happens, farmers suffer. Companies international seed organizations (see papers are not able to introduce their best varieties for by Condon and Poey in Section 2). each agro-ecological zone (for example, Ethiopian highland, lowland, and rift valley Table 1.1 Regulatory barriers and recommended reforms Obstacle Recommended government reforms Lack of input from private To improve seed policies, sector into govemment * establish formal contacts with natonal seed industry associations policies * bring nabonal and intemational seed companies into seed policy discussions * join supporting official multinatonal organizations dealing with seeds, such as the OECD Seed Schemes and the Intemational Seed Trade Federation Compulsory variety To encourage research and variety introductions by private companies, NGOs, and registrabon autonomous public organizations (for example, universities): * make variety registration voluntary; in other words, allow companies to sell seed of varieties that govemment has not tested and approved If there is too much oppositon to end compulsory variety registraton, then: * limit compulsory variety registrabon to a few major crops, leaving other crops with voluntary registration; and/or * make registraton automatc for varieties that are already registered in a list of other countries Compulsory seed To favor quality seed and competitive markets: certfication and other * offer seed certificabon, but make it voluntary quality controls * enforce truth-in-labelling; mandate what companies must put on labels, then prosecute if contents do not match labels Lack of private access To encourage private research and to speed delivery of public research results to to public germplasm and farmers: breeders' seeds, * allow anyone (companies, NGOs, etc.) to get IARC lines and other genetic including materials from materials direct from lARCs; and instruct public research agencies to pass them on IARCs at cost * establish policies for public research agencies to sell germplasm and seeds from own research at cost or for a profit Inadequate protection for To encourage variety transfer and research: intellectual property rights in * allow companies to sell hybrid seeds without giving the govemment samples of seed technology valuable parent lines (that is, allow companies to maintain physical control of parent lines) * adopt legislation allowing companies to register ownership of vareties Unreasonable phytosanitary To facilitate safe intemational seed trade: rules and other non-tariff * work with multinational organizatons and the private seed industry to barriers on seed imports and standardize phytosanitary seed tests and to arrange transparent procedures to exports resolve disputes * do away with all other non-tariff barriers on seed imports and exports Over-centralized To favor expansion and development of the informal seed sector administraton of regulatons * deregulate as above, particulariy doing away with compulsory variety registration and compulsory seed certificaton * as far as possible, decentralize administration of remaining regulations to sub- natonal govemments or local offices of central agencies Fragmented regional To encourage expanded flows of new public and private varietes into regional markets: markets: small countries * do away with compulsory variety registration with excessively natonalistic * remove non-tariff barriers on seed imports (except those that address seed systems phytosanitary concems) * cooperate to study regional seed-bome pests and diseases 4 WORKSHOP OVERVIEW AND RECOMMENDATIONS zones) or market niche (for examnple, for food eties without registering them in any way and or feed, early or late maturing, etc). Instead, also without government tests and approvals. companies choose varieties that are good in Registration is available as an option; a compa- some markets and satisfactory in others. ny can register a new variety by submitting its Similarly, compulsory registration makes it own data on the variety to a journal. Seed com- more expensive for companies to replace older panies in EU countries must submit seeds of varieties with newer and better ones. Finally, candidate new varieties to one or another EU farmers do not always want the best variety. government for two years of tests. Tests look at Depending on many factors, such as seed cost performance (VCU tests) and also consider or availability of fertilizers, farmers may prefer whether or not seeds represent a variety (DUS varieties that would not pass official perfor- tests, to see if plants are distinct from other vari- mance tests. eties, uniform, and have stable characteristics Most - but not all - of the participants at over several generations). the workshop criticized performance tests and The EU seed regulatory system is superfi- compulsory variety registration as costly and cially similar to compulsory variety registration unnecessary and recommended alternate proven as practiced in many developing countries. arrangements for managing the introduction of However, the EU system has an important mit- new varieties. The following paragraphs present igating feature: Companies can choose to intro- the overall sense of the workshop in favor of duce a new variety through any EU govern- doing away with variety controls; minority ment. Once approved by any one government, arguments in favor of variety controls are con- the variety goes into an EU Common Catalogue, sidered at the end of this Section. and seeds can be sold throughout the EU. For Seed company representatives taking part in example, seeds of a variety approved in Italy can the workshop described extensive company be sold in Spain or Germany without further tests of new varieties before bringing seeds to tests. Multiple doors to a common list give market (see paper by Ansaldo and Riley in companies some leverage among governments; Section 3). Companies judge seed quality on if one government is difficult, maybe another the basis of data from hundreds or thousands of will be more reasonable. Also, governments are plots, including demonstrations and market tri- competing to keep a share of the EU's private als to determine performance in farmers' fields seed industry. Finally, the EU Common and farmer interest. Official trials, however, are Catalogues provide access to a large multi-coun- often too limited in number and too poor in try market, so that companies have strong quality to support good decisions (see papers by incentives to overcome regulatory barriers. Gaedelmann in Section 2 and Ansaldo and Thus, although EU variety controls delay Riley in Section 3). In addition, official trials introduction of new varieties for several years, divert public and private resources from other their impact is not at all as stultifying as single- activities, such as research and technology dis- country variety controls in small and poor semination. developing countries such as Malawi, Angola, Workshop participants described alternate Madagascar, or Ghana. Even supposing that seed regulatory practices in various countries, tests and decisions are reasonable and unbiased providing a backdrop against which to consider - which has often not been the case - regis- compulsory variety controls in many develop- tration procedures block entry for all varieties ing countries. Jim Elgin, of the United States for which registration costs exceed expected Department of Agriculture (USDA), described profits from seed sale. In small and poor coun- and contrasted US and EU (European Union) tries, companies may be willing to deal with the seed regulations (see Elgin in Section 2). In the expense and bother of VCU tests and registra- United States, companies sell seeds of new vari- tion procedures to introduce new hybrids for EASING BARRIERS TO MOVEMENT OF PIANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 5 major crops such as maize, but not for self- or * go to multi-country variety lists; any open-pollinated varieties (OPVs) of minor government acting on its own could do crops, with their relatively much smaller poten- this by announcing that it will automati- tial seed revenues. cally accept varieties registered in a speci- Other papers at the workshop made it clear fied list of other countries. that voluntary variety registration is practiced by a significant portion of the developing During the 1990s, governments of many world; the pattern is not limited to the United developing countries have been reviewing and States. In 1990, Bangladesh did away with revising seed regulations. Whether or not to compulsory variety registration for all but five make variety registration compulsory has been crops (see paper by Gisselquist in Section 4). one of the issues on the table. For example, the Agrawal (see Section 4) reports that for many Southern African Development Community years the Indian government has not required (SADC) and the Commonwealth Secretariat variety registration for seed sale.2 In countries organized a 1994 workshop on SADC seed pol- with this pattern, seed laws often recognize a icy in Harare, Zimbabwe. Acting on the basis of class of "quality declared" commercial seed, for an expert report that did not discuss the option which governments ask that companies label of voluntary variety registration (Dixit and seeds accurately, but do not require prior gov- Swarup 1993), the workshop endorsed compul- ernment tests of varieties or seed quality. sory variety registration and national variety lists for all SADC countries (Commonwealth Recommendation 2: Governments are Secretariat and SADC 1994, p 11). This unfor- encouraged to make variety registration tunate position was somewhat mitigated by voluntary, allowing companies to sell seeds another workshop recommendation that "laws of new varieties on the basis of company should provide reciprocal recognition of vari- decisions alone (see the first four papers in eties amongst countries since this would facili- Section 2). Official performance tests tate regional seed trade" (Commonwealth could continue on a strictly voluntary Secretariat and SADC 1994, p 2). As discussed basis, and governments could maintain above, the workshop's recommendation for lists of recommended varieties as a compulsory variety registration with a multi- source of information for farmers and for country SADC variety list is a second-best solu- the seed industry, though companies tion. Taken together, all eleven SADC countries would be allowed to sell seed of varieties (Angola, Botswana, Lesotho, Malawi, that are neither tested nor listed. Mozambique, Namibia, South Africa, Swaziland, Tanzania, Zambia, and Zimbabwe) As second best solutions, governments have a population of only 140 million. could: During the 1990s, advances in bio-engineer- ing led to development of transgenic varieties * limit compulsory variety registration to (plants modified to include genes from other one or two crops that are important for species) for commercially significant crops, food security or the economy, leaving including maize and cotton. Bio-engineered companies free to introduce varieties for varieties offer valuable advantages, such as all other crops without government improved resistance to pests through incorpora- interference (see paper by Gisselquist in tion of genes that instruct plants to produce Section 4 describing seed regulatory biopesticides. But experts raise questions about reforms in Bangladesh); public health and environmental impacts of bio-engineered crops - for example, wide- or spread planting of cotton with the gene to pro- 6 WORKSHOP OVERVIEW AND RECOMMENDATIONS duce Bt toxins could induce resistance to Bt (a Compulsory certification boosts seed pro- useful biopesticide) in pest populations. Also, duction costs. Also, the expense and paperwork transgenic genes could escape into wild popula- involved in arranging for government officials tions when crops are planted in their Vavilovian to visit seed plots and to test seed lots are barri- centers. ers to entry into the formal seed sector. The workshop did not develop any specific At the workshop, endorsements for volun- recommendations for regulating introduction tary certification came from donor staff, a of transgenic varieties. Since transgenic varieties USDA seed expert (see Elgin in Section 2), and are so new, there is no consensus on public private sector participants (see papers by health and environmental risks. Currently, Condon and Poey in Section 2). Venkatesan OECD countries regulate transgenic varieties makes a strong case for voluntary certification more stringently than varieties from conven- (Venkatesan, p 35): tional breeding. However, regulations in OECD countries are in a state of flux, and can A seed law which provides for "optional" be expected to evolve rapidly over the next certification would encourage seed pro- decade. Governments of developing countries duction in the informal sector and can generally be encouraged to follow the lead would make the certification agency of OECD countries to design regulations and "earn" its honor, rather than having it to approve specific varieties or genes. given to it by a government fiat; if farm- ers develop confidence in the agency, Caompulsory certification and seed quality tests they might be prepared to pay a higher Paralleling differences in their approach to price for "certified" seed. variety registration, governments follow differ- ent regulatory strategies to promote seed quali- Poey (see Section 2) advised that govern- ty. Elgin describes and compares US and EU ments "audit' or confirm that companies main- systems (see Elgin in Section 2). The United tain adequate quality control," but cautioned States addresses seed quality through truth-in- that "certification should by no means substi- labelling, allowing companies to set their own tute for each company's own internal quality quality standards and to carry out their own control measures." tests, asking only that companies label seeds with accurate information on variety, germina- Recommendation 3: Governments are tion, purity, inert matter, etc. In the United encouraged to make official seed certifica- States, official certification (by state not federal tion available but voluntary. Also, a agencies) is voluntary. The EU, however, truth-in-labelling approach allows gov- demands compulsory certification: companies ernments to promote seed quality and to must arrange for government officials to visit all control fraud with minimal interference seed production plots to check that seeds are of in competitive private markets (see end the correct variety and to test all seeds for quali- of this section.) ty (germination rate, etc). In many developing countries, including Unrealistic phytosanitary rules and other non- India and Bangladesh (see papers by Agrawal tariff trade barriers and Gisselquist in Section 4) certification is vol- Seed imports are important for introducing untary, as in the United States. But many other new varieties and for keeping seed costs down. developing countries follow the EU pattern, Often, imports help a company to introduce making certification compulsory for all com- new varieties to a market; once the size of a mercial seed. market is determined, locally produced seed is EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 7 normally cheaper and replaces imported seed. iff barriers on seed trade (that is, barriers Even so, companies can have difficulty manag- that have nothing to do with phytosanitary ing annual fluctuations in seed production and concerns). demand, leading to unmet demand or large inventories that are expensive to carry. Even without any policy interference to pro- Occasional seed exports and imports help com- tect domestic seed industries, seed import costs panies hold down costs and hence seed prices. heavily favor in-country production in most Papers and discussions at the workshop crit- circumstances. For example, even with liberal icized unrealistic phytosanitary rules as barriers seed import regimes, imports into Thailand and to seed trade (see papers by Condon, Chile supplied only about 1 to 3 percent of Gaedelmann, and McGee in Section 2). To planted seed in the mid-1980s (FAO 1987, p address the risk that imported seeds may intro- 307; Pray 1990, p 195). Nevertheless, in some duce new pests and diseases, virtually all gov- circumstances a case can be made for modest ernments set detailed crop-by-crop rules about tariffs to protect domestic seed production for how imported seeds are to be treated, from major crops. For example, with devaluations, which regions or countries seed imports are not inflation, and price interventions in fertilizer allowed, etc. While there is no debate about the and maize markets, relative maize seed prices in principle of phytosanitary controls, experts Malawi, Zambia, and Zimbabwe in the mid- argue that many specific existing rules do not 1990s diverged widely and fluctuated rapidly. address realistic threats, but are rather an excuse In such circumstances, unrestricted and to block seed imports. Also, current rules often untaxed seed imports could devastate seed pro- overlook real threats. duction in one or several countries, undermin- ing stable domestic supply of seed for major Recommendation 4: To facilitate safe seed food crops. Currently, Malawi and Zambia trade, governments are encouraged. (a) to limit seed imports with non-tariff barriers; in focus phytosanitary systems on realistic pest general, tariffs would be preferable. and disease threats; and (b) to work with international organizations and private Weak intellectualproperty rights seed trade associations to establish transpar- Participants at the workshop argued that ent procedures to appeal and arbitrate dis- weak protection of intellectual property rights agreements about seed rules and decisions. in many developing countries inhibits private sector introduction of new varieties. Depending Many governments protect domestic seed on the crop and country, seed companies have a industries with systems of import permits that number of options to protect intellectual prop- require importers to get prior approval from erty rights in seed technology (see Gaedelmann ministries of agriculture for each order. Tariffs in Section 2). on seeds, however, are characteristically low and For crops such as maize and sunflower, com- are seldom a significant trade barrieL The more panies can package breeding advances into actions taken to protect domestic seed indus- hybrid seed, which is produced from two or tries, the less competitive the seed industries more parent lines and which will not produce tend to become, and this effect is greater for equivalent seed in the next generation. With smaller countries. With protected seed indus- hybrids, companies have biological protection tries, farmers get lower seed value for cost. (through physical control of parent lines), which can be reinforced with legal protection in Recommendation 5: To favor competitive a number of ways (for example, as trade secrets, domestic seed industries, governments are through contracts with seed growers, or by reg- encouraged to do away with other non-tar- istering variety ownership through PVP or 8 WWORKSHOP OVERVIEW AND RECOMMENDATIONS patent law), depending on what laws are avail- As part of the latest GATT accords, govern- able in a country. ments agree to "provide for the protection of For other crops such as potatoes and cotton, plant varieties either by patents or by an effec- for which breeding advances are packaged into tive sui generis system or any combination non-hybrid seed (which anyone can use to pro- thereof" (Trade Related Intellectual Property duce more seed of the same variety), companies Rights [TRIPS] Agreement, part 2, section 5, have no alternative to legal protection through article 27, para 3[b]). With respect to plant PVP or patent laws. If these are not available, a varieties, recommendation 7 asks governments company is not able to block other companies to satisfy their obligations under TRIPS by from multiplying seed for sale. joining UPOV. Clauses in the TRIPS agree- ment give developing countries five to ten years Two recommendations come out of work- to comply. shop discussions on intellectual property rights: Lack ofprivate access to public lines Recommendation 6: Governments are In years past, government research agencies encouraged to stop requiring companies to routinely channeled breeders' seed to seed submit samples of hybrid parent lines parastatals for seed multiplication and sale. before allowing seed sale (for example, as With a trend away from seed parastatals, gov- conditions for variety registration, seed cer- ernment research scientists look for new .tification, etc). These demands under- arrangements to multiply seed of new public mine biological protection for hybrids. varieties. For example, the Ministry of Companies fear that governments will Agriculture in Malawi organizes growers to take make valuable parent lines available to and multiply seed for sale back to the Ministry, breeders in government agencies or other which then offers it to farmers. companies. In recent years, these Government supervision of seed multiplica- demands have blocked private compa- tion, as in Malawi, can be expensive and unnec- nies from introducing hybrids into many essary. Instead, governments can demonstrate countries, including China and Zambia. new varieties and at the same time offer breed- er's seed for sale, leaving to farmers and private Recommendation 7: Governments are seed companies the task to multiply and sell. If encouraged to adopt PVP laws consistent a variety will not go under such conditions, with the 1978 or 1991 Convention of then arguably it is not interesting to farmers UPOV (Union for the Protection of New and not worth further promotion. Plant Varieties). UPOV is an internation- Public research agencies also have al treaty organization, for which mem- germplasm and lines from research that are not bership requires passage of an acceptable suitable for release as varieties. To support pri- PVP law. PVP laws consistent with vate research and to encourage widest use of either Convention allow a company to these materials, governments can make standard register ownership of a variety and then policies to sell them on request and at cost. to block other companies from selling However, some public materials may have a seeds of that variety. The 1991 high market value. In such cases, exceptions can Convention more strictly limits farmer's be made to allow agencies to sell or license privilege (the amount of seed a farmer them for whatever the market will bear, particu- can produce for own use or sale) and larly if materials are expected to go into interna- close breeding (see Gaedelmann in tional trade. Section 2). EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 9 Germplasm and lines from international the economy in general and for seeds in partic- agricultural research centers (IARCs), such as ular. Governments can break down this dualism the International Rice Research Institute by making it easier for informal seed producers (IRRI), are another resource that can support to establish links with government agencies and private as well as public research. Donors spon- to move into the formal sector. sor IARC research for the benefit of farmers Informal seed producers are often skilled and consumers in developing countries. IARCs and respected medium- or large-scale farmers. freely distribute their germplasm and lines If regulations and government agencies allow, (many of which are suitable for immediate many of these informal producers will expand release as varieties without any further breed- operations and join the formal private seed ing) to government research agencies. However, industry. Some regulatory obstacles are com- governments of many developing countries mon to all sectors (for example, requirement to claim monopolies on [ARC materials, denying register as a company, which entails paying access to resident private companies. With VAT or other taxes). In the seeds sector, any monopoly power, public research agencies have number of regulations, such as compulsory seed delayed and even blocked introduction of new certification, can serve to reinforce dualism. varieties from IARC lines; for example, Turkish Some of these regulatory obstacles can be sim- scientists blocked introduction of CIMMYT ply removed (for example, seed certification can wheat in the 1960s. be made voluntary), while others cannot. This leaves governments with a problem: how to Recommendation 8: Government research lower barriers so that a significant share of agencies are encouraged to establish stan- informal seed producers can grow into formal dard arrangements to sell germplasm and seed companies? breeder's and foundation seeds to private Decentralization seems to be part of the companies. Standardizing commercial solution: Peru's recent strategy to work with pri- arrangements provides a framework for vate seed companies through committees at public-private cooperation, supporting department level (a level of sub-national gov- in-country private research and expedi- ernment equivalent to provinces or states) facil- tiously channeling advances from public itated the emergence of more than 150 small sector research into farmers' hands. seed companies in the early 1990s (see paper by Cortes in Section 4). In India and Turkey (see Recommendation 9: Governments are papers by Agrawal and by Gissdlquist and Pray, encouraged to allow resident private com- respectively, in Section 4), sub-national govern- panies free access to IARC germplasm and ments have responsibility for important aspects lines. India, for example, allows resident of seed regulation as well; in both countries, companies direct access to ICRISAT private seed industries include large numbers of materials on the same terms as govern- local companies. ment research agencies. Recommendation 10: To enabk infermal Over-centralized administration ofseed regulations seed producers to graduate into theformal In many developing countries, the formal seed sector, governments are encouraged- (a) commercial sector, public and private together, to cut seed regulations; and (b) as far as provides less than 10 percent of planted seed, possible to decentralize administration of while the informal sector, including farmer remaining regulations to states or other saved seed and farmer-to-farmer sales, accounts sub-national governments that are more for most planted seed. The large size of the accessible to smal seedproducers. informal seed sector is a symptom of dualism in 10 WORKSHOP OVERVIEW AND RECOMMENDATIONS Small countries, fragmented markets and fruits. In 1993 alone, the OECD's List of Private seed companies balance expected Cultivars Eligible for Certification added more profits against costs when considering whether than 400 new varieties for maize and more than or not to enter a market or to introduce a new 100 each for sugar beet, wheat, and sunflower variety. For smaller and poorer countries, (this list may not be complete, since US law potential profits are relatively small, so that allows companies to sell seeds of varieties which even very modest regulatory barriers may com- are not listed). pletely block private companies from entering Corresponding to this steady flow of new or introducing new varieties, particularly self- varieties, older ones fall out of favor. In indus- and open-pollinated varieties of minor crops. trial countries, the expected market life for a For example, private seed companies in India, new variety is less than ten years. Steady with a population of 900 million, have relative- advances in breeding achieve yearly increases in ly strong incentives to introduce new varieties yields and other improvements so that even because the market is so large. In contrast, seed very successful varieties soon lose their market markets in Malawi and Zambia, with popula- appeal as even better ones come available. tions of roughly 10 million each, are only about Thinking of the supply of new varieties as a one percent as large as in India; hence, incen- flow rather than single varieties has implications tives for private companies to enter and to for agricultural development strategy. For a introduce new varieties are relatively low. developing country, the challenge is not to Throughout sub-Saharan Africa, only a small introduce particular new technologies-such as minority of countries has more than 15 million a handful of improved varieties for a few major people, and only one exceeds 100 million. crops-but rather to improve farmer access to One strategy for relatively small countries in the flow of new seed technology from laborato- Africa (and also Central America) to improve ries and research plots throughout the world. farmer access to new varieties is to create Certainly, only a portion of the world's flow regional markets for seed technology, whose size of new varieties is relevant for any particular would attract private companies. Similarly, cre- developing country, depending on environment ating regional markets would give farmers and other considerations. However, with the unimpeded access to varieties from public expansion of public and private sector breeding breeding in multiple regional countries. programs in low latitudes over the last several decades, the share of world seed technology Recommendation 11: African governments suitable for developing countries has increased are encouraged to create regional markets over time. For wheat and rice, direct interna- for seed technology by: (a) doing away with tional variety transfers have accounted for a compulsory variety registration and conse- large share of improved varieties introduced quent single-country lists of allowed vari- into developing countries (see papers by eties; (b) moving towards regional PVP Maredia, Ward and Byerlee and by Evenson registration; (c) cooperating to study region- and Gollin in Section 3). For hybrid maize and al seedborne pests and diseases; and (d) cotton, private companies have varieties suitable doing away with non-tariff barriers (except for many ecological zones and market prefer- reasonable phytosanitary controls) on ences, such as white maize for sub-tropical regional seed trade. regions of Africa (see papers by Ansaldo and Riley and by Pallin in Section 3). More than 40 International flows of new seed technology percent of Pioneer Hi-Bred's maize research Farmers in industrial countries have access locations are in developing countries, at low lat- to a steady and large flow of new varieties for all itudes that facilitate breeding for other develop- major and minor crops, including vegetables, ing countries. Depending on market size, com- EASING BARRIERS TO MOVEMENT OF PIANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 11 panies will also breed for specific country or enforced [World Bank 1997b, vol 2, annex 2, regional markets if the company's existing vari- p4]). Second, agriculture is a state subject, so eties are not satisfactory. that autonomous state research organizations The potential contribution of foreign vari- are able to make their own decisions about eties depends in part on a country's size. Small breeding and variety release. Third, seed certifi- countries share ecological zones with neighbors, cation has been voluntary, which lowers barriers have small seed markets, and small government to entry and seed production costs for private research budgets. If farmers in small countries companies. And finally, India is a large country, such as Malawi are to gain access to an equiva- with a population greater than all of Africa or lent flow of new varieties as available to farmers Latin America. in larger and more populous countries such as Even before regulatory reforms from the Brazil, Mexico, the United States, or India, then end of the 1980s, Indian farmers had access to international variety transfer will necessarily seeds of several thousand varieties for major supply most of those varieties. and minor crops from thousands of breeders in federal and state agricultural research organiza- Country experiences with refonn tions and universities, and there was a large Over the past fifty years, developing coun- and competitive domestic private seed industry tries have practiced a range of seed regulatory as well. Compared to current systems in most regimes, with some such as Thailand more open African countries, the Indian seed system even all along and others such as Ghana consistently before reform was far superior. Even though closed. Also, in some countries, such as Chile most breeding was in the public sector, it was and Mexico, regulatory reforms have moved competitive rather than centrally controlled. countries from closed to open policies. This Varieties could follow market demand from range of experience provides information about one state to another. Private companies could the impact of regulatory systems and other con- introduce new varieties-from other states or ditions on seed industries and agricultural countries or from own research-without hav- development. ing to seek government approval, and they Papers presented and distributed at the could sell seeds across state borders. In con- workshop described regulatory reforms and trast, in much of Africa, governments control their impact in India, Turkey, Peru, and variety introductions and discourage cross-bor- Bangladesh. der trade, limiting farmers to public sector varieties bred and released in single countries, India most of which are smaller than an average India for many years pursued self-sufficiency Indian state. and socialism, discouraging seed imports and India's long experience with voluntary vari- forcignswhile promoting public ey registration and seed certification provides fetorseign c rompanies and tradc. In many solid evidence that even in poor countries sector seed production and trade. In many other developing countries, similar policies left farmers do not suffer when governments allow farmers with a short and often poor selection of farmers and markets-rather than government varieties. In India, however, several mitigating regulators-to assess variety performance and factors limited the damage to farmers. First, seed quality. A 1997 World Bank report pre- variety registration has been voluntary, so that pared with FAO assistance states that for India no single committee or agency has had (World Bank 1997a, p 16): authority to decide which varieties can be intro- . . voluntary certification of notified duced (except that a Cotton Control Act varieties (ie, those released officially empowers government to decide which cotton through the variety release system) and varieties can be grown, but that Act is not 12 WORKSHOP OVERVIEW AND RECOMMENDATIONS truthful labeling of hybrids and other Officials in Turkey's Ministry of Agriculture seed produced by the private sector (ie, pushed through seed regulatory reforms in the voluntary variety registration and seed early 1980s. In designing seed reforms, govern- certification) provides adequate protec- ment officials sought advice from private com- tion to consumers without unduly panies as well as the International Finance inhibiting producers. This system Corporation (which sub-contracted advice to should be retained. the Industry Council for Development). Turkey removed seed price controls, encouraged for- India's primary seed and economic reforms eign seed companies to enter, and modified but from the late 1980s included: improved private did not abandon compulsory variety registra- company access to varieties from public breed- tion (that is, the government allowed compa- ing (through sale of breeders' seed); reduced nies to do their own variety trials, cut testing to entry barriers for foreign vegetable varieties two years, approved most varieties submitted (through seed imports); and reduced entry bar- for registration, and expanded use of "produc- riers for foreign seed companies. With these tion permits" to allow companies to sell seeds new policies, private companies have improved of unregistered varieties). the menu of varieties available to Indian farm- With reforms, the number of improved vari- ers, introducing foreign varieties as well as new eties available to farmers increased dramatically varieties from expanded in-country research. - for example, from 1982 to 1987 the number From 1988 to 1995, the cumulative number of of improved varieties for sunflower increased varieties released by private companies from in- from three to thirty, and for soybeans from two country research more than doubled to about to more than forty. From the early 1980s, pri- 300. vate companies have introduced most new vari- eties, most of which have been direct transfers Turkey from foreign countries. Turkey shares latitudes Prior to 1980, Turkey similarly favored self- with many industrial countries, and has been sufficiency and socialism in the seed industry. able to source new varieties from the United Government parastatals dominated seed pro- States and European countries as well as from duction and trade. However, the situation was Mexico and other countries at lower latitudes. more closed than in India; government research From the early 1980s, the private seed was centrally controlled, and variety registration industry increased from a very few firms focus- and seed certification were compulsory. With ing on vegetables to approximately eighty a all these restraints, farmers had access to few decade later. Most major multinationals have a interesting varieties for many crops, and private presence through joint ventures and licensing, commercial seed production was extremely lim- although there are also a few subsidiaries. As of ited. the mid-1990s, private companies dominate Entering the 1980s, widespread smuggling commercial seed sales for maize, sunflower, of vegetable seeds and government inability to potatoes, soybeans, and vegetables, while paras- popularize hybrid maize demonstrated the fail- tatals dominate seeds for wheat, barley, and cot- ure of existing policies to give farmers access to ton. Sugarbeet seed is dominated by a joint advanced technology. Also, millions of Turks venture. had travelled to Western Europe and had seen It is difficult to measure the impact of how things work in industrial market Turkeys 1980s regulatory reforms from aggre- economies. Although general macroeconomic gate data on crop agriculture, since many other reforms from 1980 had liberalized trade and factors such as changes in macroeconomic poli- investment, special regulations continued to cies get in the way. However, it is possible to see block private seed production and trade. impact of reforms on specific crops. Regressions EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 13 in the paper by Gisselquist and Pray in Section varieties released by public sector research insti- 4 show a roughly $100 million increase in tutes; for vegetables, variety lists protected annual net farm income from maize due to seed monopolies and oligopolies for private compa- reforms. Other estimates show smaller increases nies having sole agent contracts with foreign in income from vegetable and seed exports due seed companies. to seed reforms. Toward the end of the 1980s, renewed inter- The paper also considers possible costs or est in crop diversification-along with general losses to the domestic seed industry and to interest in privatization and liberalization- farmers due to seed reforms. Despite entry of stimulated the Ministry of Agriculture to multinationals, domestic seed companies have re-examine seed policies to see what could be increased seed production and sales. Increases done to speed entry of new varieties through in seed exports suggest that Turkeys seed indus- private seed companies. This review led to a try has been able to compete in international Ministry of Agriculture seed policy paper that markets as well. Finally, evidence suggests that recommended seed trade liberalization to give farmer losses from inappropriate varieties have farmers in Bangladesh access to the best vari- been lower with post-reform private seed trade eties available in world markets. Acting on than with pre-reform public seed trade, and internal recommendations, the Ministry of seed quality has not been a problem. Agriculture used its authority under existing seed legislation to issue orders ending controls Peru on varieties allowed for import and trade. Toward the end of the 1980s, Peru took The Ministry exempted five crops from several steps to shift seed production from the these reforms, maintaining compulsory variety public sector to a competitive private sector: registration for rice, wheat, jute, sugarcane, and specifically, it eliminated two seed production potatoes. Some experts defended variety con- parastatals and created public-private commit- trols for these major crops with the argument tees in multiple departments (a division of the that a poor or disease-susceptible variety could country equivalent to state or province) to assist have wide impact on production. This logic is development of private seed enterprises. not entirely convincing, because poor varieties Reforms at the end of the 1980s facilitated would be unlikely to gain popularity. The gov- the emergence of more than 150 new seed pro- ernment has mechanisms other than compulso- ducers through the mid-1990s. Regional seed ry variety registration to limit area planted to committees arrange (voluntary) seed certifica- varieties with severe disease susceptibility (for tion and in some cases have seed drying and example, extension advice or even short lists of processing equipment available for seed produc- varieties not allowed). Probably the real reason ers to use. These regional seed committees have for continuing variety controls for only five been accessible to even very small seed produc- crops was that these were of most interest to ers, lowering barriers to entry into the formal public sector scientists, and reformers felt the sector. A large majority of new seed companies need to defuse opposition from that quarter. have less than ten hectares in seeds (see table Since the early 1990s, private companies 4.12). and several non-governmental organizations (NGOs) have been testing and introducing new Bangladesh varieties for sunflower, maize, other field crops, Through the 1980s, Bangladesh enforced and vegetables. Companies have also lobbied compulsory variety registration, maintaining for relaxing variety constraints on potato lists of allowed varieties for all crops, induding Imports. vegetables and tree crops. For major and minor In several respects, 1990 seed reforms in field crops, lists consisted almost entirely of Bangladesh can serve as a model for reforms in 14 WORKSHOP OVERVIEW AND RECOMMENDATIONS other developing countries. First, the Ministry Willfarmers losefirom poor varieties and of Agriculture introduced reforms through seeds? Some experts argue that farmers in internal Ministry orders, with authority from developing countries will lose if govern- existing seed legislation. Standard seed legisla- ments relax compulsory variety registra- tion in many countries gives ministries of agri- tion and compulsory seed certification; culture authority to control seed imports, to although markets can regulate seed qual- notify or otherwise list crops for which compul- ity in developed countries, conditions are sory variety registration is necessary, and to different in developing countries, so that exercise judgement and authority in other areas more stringent regulatory measures are of seed regulation. Introducing change through appropriate. ministerial orders is easier and quicker than try- ing to design and draft new seed legislation. * Will technology transfer damage in-coun- Amendments or even new seed laws may be try research? Many government scientists called for in some cases (for example, to intro- worry that government financial support duce plant variety protection), but these take for public research will fall if private seed time. companies are allowed to introduce new Second, reforms emphasized seed trade lib- varieties for major crops. eralization rather than privatization, removing barriers to private seed import and variety Are appropriate varieties available? introduction, but leaving contentious debates After WWII, the insight that breeding for about parastatal seed production to be decided low latitudes and other specific conditions was over time. With the emphasis on liberalization, necessary to generate varieties suitable for devel- farmers gain improved access to seeds and vari- oping countries led to establishment of the eties through private sector channels without International Rice Research Institute (IRRI) losing access to seeds through public channels. and other international agricultural research Virtually the only opposition to 1990 seed centers (IARCs) as well as donor support for reforms in Bangladesh came from public sector national agricultural research systems (NARS). scientists. Seed trade liberalization allows pri- Despite the substantial number of public as vate companies to enter and expand; over time well as private breeding programs for low lati- expansion of the private seed industry can be tudes that have been established over the past expected to change the character of the debate forty years, some experts continue to argue that about parastatal seed production. Further, lead- climate, diseases, tastes, and other factors inhib- ing with seed trade liberalization rather than it transfer of varieties among developing coun- privatization does not threaten the stability of tries so much that regulatory reform would (public sector) seed supply for major crops. have little or no impact. Papers in Sections 3 and 4 confront this argument, demonstrating Reservations about regulatory reform that varieties are available and do move when During the workshop, participants discussed regulations allow. reservations or questions as potential obstades Several workshop participants argued that to reform: whether or not varieties move, a case for regula- tory reform can be built on the premise that * Are appropriate varieties available? Some reform does no harm. However, that approach experts argue that seed regulatory reform may not be sufficient to change policy; unless is not worth worrying about since vari- one can make a case that private introductions eties must be locally bred for climates, of new foreign varieties (or varieties from pri- tastes, and other specific conditions in vate research in-country) will follow regulatory each country. reform, inertia may be easier. Donors might not EASING BARRIERS TO MOVEMENT OF PIANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 15 be interested to push reform if experts thought not be trusted to offer good varieties and seed, it would have no impact. that they might try to sell bad varieties to For a developing country, the potential recoup breeding costs, and that they may sell impact of foreign varieties coming in through bad seed (low germination rate, etc). This was a direct transfer (with in-country screening, but minority position at the workshop, with most without further crosses) depends on many fac- participants favoring markets and alternate reg- tors, including size of country, quality of in- ulatory strategies to ensure good varieties and country breeding programs, quality of foreign seeds. research in comparable latitudes, and other cir- Arguments to deregulate seed markets do cumstances. Measures of agro-ecological dis- not turn on whether or not companies are to be tance (see Maredia, Ward and Byerlee in trusted. Markets work because farmers will stop Section 3) compare average performance of buying seed from companies offering poor vari- local vs foreign varieties; however, it is not the eties and poor quality. Also, even without com- average that is transferred, but rather the indi- pulsory variety registration and seed certifica- vidual. tion, governments have other options to rein- A farmer wants the best variety for a partic- force market rejection of poor varieties and ular field and cropping pattern. The chance poor quality seed (see table 1.2). that public sector in-country research supplies The hypothesis that farmers will lose if gov- that variety decreases for smaller countries, ernments do not directly control variety and minor crops, and less effective public research seed quality can be tested against experience in systems. Turning the question around, what is developing countries such as India, Bangladesh, the chance that a new variety will exhaust its Thailand, and many others with voluntary vari- market potential within one country? That will ety registration and seed certification. Such a similarly depend on the size of the country, the test has not been done. Instead, advocates for share of cultivated area planted to the crop, and government controls cite selected anecdotes to the extent and quality of research in countries show that companies sell bad seed. Examined with comparable environments. closely, most of these anecdotes can be reinter- Further studies on variety movement might preted to show that public agencies buy bad look at share of new varieties borrowed vs bred seed. For example, the government of in-country for major and minor field crops and Bangladesh imported low quality maize seed vegetables, based on planted area, size of coun- from a Thai company around 1990. That gov- try, etc. Data on source of varieties (domestic vs ernments often buy bad seed (from private foreign) for developing countries that regulate companies as well as from parastatals) demon- variety movement could be compared with sim- strates that they cannot be relied on to make ilar data on source of varieties for countries in good decisions about variety and quality, and if the EU and states in the United States or India, anything supports arguments against compulso- where seed laws allow varieties and seeds to ry variety registration and certification. move freely across country or state boundaries. Furthermore, risks are not all on one side. In Vegetable seed companies which distribute evaluating what might happen with seed dereg- international seed catalogues might be willing ulation, risks of loss from poor varieties and to share data on international markets for indi- poor quality seeds can be balanced against risks vidual varieties. of foregone gains from suppressed new varieties or continued reliance on relatively poor quality Willfarmers lose from poor varieties and seed? farmer-saved and parastatal seed if seed reforms Advocates for compulsory variety registra- are not implemented. tion and compulsory seed certification for Finally, arguments that farmers might lose developing countries argue that companies can- from poor varieties and poor quality seed can 16 WORKSHOP OVERVIEW AND REcoMMENDATioNs Table 1.2 Government options to limit farmer risks (Consistent with voluntary variety registration and seed certification) Risk Govemment options to promote quality seed Poor variety * promote free entry for companies and varieties, to ensure that farmers have choice within a competitive market * advise farmers about variety performance, generally recommending better varieties, but possibly discouraging some (for example, because of disease susceptibility) * continue government breeding Mis-labelled variety * follow a truth-in-labelling strategy: require companies to name the or mixture of variety (if it is a registered variety) on the label; if a package is mis- varieties labelled, the govemment has a legal basis on which to prosecute Poor germination, * follow a truth-in-labelling strategy: require companies to state presence of weed minimum germination, maximum percent of weed seeds, etc., on the seeds, other analytic label; if contents do not match the label, the govemment can measures of low prosecute; and/or quality * set minimum germination rates and other standards for various crops (this is more restrictive than truth-in-labelling) Seed imports * enforce phytosanitary rules to block seeds with pests and diseases introduce pests and (note: this allows imports of unregistered varieties and uncertified diseases seed) be considered against the percent of planted industrial as well as agricultural technology. seed coming from the formal private sector. In Probably the best answers to date come from the poorest countries where markets supposedly research on the impact of India's industrial work least efficiently (and where opponents of technology policy on technology transfer. seed regulatory reform argue that farmers face Before late-1980s reforms, India discouraged the greatest risk) the percent of planted seed commercial technology imports in industry, coming through private formal channels seldom hoping to stimulate in-country research. Other exceeds low single digits. In short: losses with developing countries have followed similar poli- deregulation are undocumented and not bal- cies. anced against losses due to over-regulation; gov- In a paper distributed at the workshop, ernments have alternate instruments; and risks Brian Fikkert presents evidence that India's pro- of farmer loss from seed deregulation are grossly tectionist policies worked poorly: controls on overblown. the purchase of foreign technology led to small increases in in-country research but had rela- Will technology transfer damage in-country tively large negative impacts on company prof- research? its. In other words, buying foreign technology This is a general question that applies to is more efficient than trying to duplicate it. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 17 However, the debate over whether or not to is possible. Some of the most important policy buy foreign industrial technology has been less obstacles to variety movement - such as com- about the benefits of foreign technology than pulsory variety registration for specified crops about the terms on which it is introduced. - take the form of government orders from Those opposing commercial transfer (that is, ministries of agriculture, and these can be technology purchase) have most often wanted removed expeditiously through new orders to get the same technology through non-com- from the same ministries. However, ministers of mercial transfer (for example, pirating and edu- agriculture may be reluctant to act on technical cation). matters without support from agricultural In agriculture in developing countries, experts. opposition to private technology transfer has Thus, the challenge for reform involves: been more general and more persistent than in raising awareness of current seed policies that industry. In many countries, public sector damage agricultural development and offering scientists have opposed all private variety intro- designs for reforms to put matters right. This ductions, without making distinctions among publication aims to do both. commercial transfers (varieties acquired Introducing regulatory reform through gov- through payment of royalties), non-commercial ernment orders leaves existing seed laws in transfers (for example, foreign public varieties place. Whether or not these laws are ideal, leav- acquired free of cost), and introduction of new ing the task of changing seed laws aside until a varieties from private in-country research. competitive private seed industry has emerged Government scientists may be motivated in and is able to take part in policy debates can part by perceived self-interest, fearing that often be good strategy. Without the persistent budget cuts for public sector research will come presence of an active and interested private seed with seed regulatory reforms, as happened in industry association, it is doubtful that well- Turkey, Mexico, and Chile. Seed reforms also meaning experts will be able to design and push challenge the status of government scientists as through reasonable, open, and workable seed benefactors and protectors of peasant farmers. laws and amendments. Writh seed regulatory reforms, governments spend less on variety performance tests and compulsory certification programs; hence, there Notes is no inherent reason for reforms to come with 1. Three divisions of the World Bank - International budget cuts for public research. With private Trade, Agriculture and Forestry, and Training - provid- ed financial and administrative support. companies introducing new varieties, returns to public research could increase or decrease; pre- 2. Although the government has allowed the sale of sumably, returns will fall in some areas (for domestically produced seeds without variety registration, example, maize breeding) and rise in others (for other regulations have blocked seed imports and discrimi- nated against foreign seed companies. For example, major- example, research on soil fertility). Competition ity foreign-owned seed companies are required to certify and communication between public and private all seed, even seed producea in India; this entails prior researchers could improve management of pub- variety registration. Reportedly this rule is not consistently enforced. From the late 1980s, India has eased barriers to lic research institutions. Nevertheless, public introduction of new varieties through imported seed. sector agricultural scientists are right to worry about budget cuts because governments may References cut research budgets despite evidence of high Commonwealth Secretariat and SADC Food Security returns and effective management. Technical and Administrative Unit. 1994. Harmonization of Seed Laws in the SADC Region, Conclusion report of a regional workshop to discuss a study on harmonization of seed laws, Harare, Zimbabwe, 4-6 Seed policy is an area in which rapid reform October 1994. London: Commonwealth Secretariat. 18 Dixit, R K, and Vishnu Swarup. 1993. Harmonization of Venkatesan, V. 1994. Seed Systems in Sub-Saharan Africa: Seed Laws in the SADC Region. London: Issues and Options, World Bank Discussion Paper Commonwealth Secretariat. 266. Washington DC: World Bank. Food and Agriculture Organization (FAO). 1987. FAO World Bank (WB). 1997a. Implementation Completion Seed Review 1984-85. Rome: FAO. Report: India: Third National Seeds Project (Credit 1952-IN). Washington, DC: World Bank Pray, Carl E. 1990. "The Potential Impact of Liberalizing India's Seed Laws," in Food Policy (June 1990), pp . 1997b. India: Cotton and Textile Industries- 193-198. Maximizing the Potential for Growth in a More Competitive Environment. (document #16437-IN). Washington, DC: World Bank. Regulations and Recommendations for Reform In many developing countries, seed regulations discourage seed companies from entering markets and doing what they can to introduce new varieties and to deliver seeds to farmers. Some of the same regulations also discourage efforts of NGOs and autonomous public sector organizations, such as universities, to introduce new vari- eties to farmers. A number of speakers at the workshop described regu- latory barriers and/or proposed reforms. In this Section, the first contribution, by Jim Elgin, describes and com- pares EU and US seed regulations. The next three papers analyze regulations and offer recommendations for reform from various private sector points of view, including: the American Seed Trade Association, a US foundation seed company with international sales, and private seed compa- nies in developing countries. The fifth paper, by Denis McGee, describes shortcomings and proposes solutions for phytosanitary regulations in developing countries. Robert Tripp provides a general overview of seed regu- lations and reforms with special attention to public sector breeding and variety release. To improve seed technology generation and transfer, Jitendra Srivastava proposes policy and regulatory reforms that address "the specific needs of all private players." Comparing EU and US Seed Regulatory Systems Jim Elgin Governments of the EU (European Union) US system is more open, favoring competition, and United States have different seed regulatory lower seed costs, and ease of entry for new seed systems, as shown in the following table. The companies and varieties. Table 2.1 Comparing EU and US seed regulations EU (European Union) Seed Regulatory System US Seed Regulatory System 1. Varety registration is mandatory. Companies are 1. Variety registration is voluntary. Seeds of all not allowed to sell seeds of a variety until the varieties can be sold, whether or not they have variety has been registered. been registered. Voluntary registration is available through public description of the variety in a joumal. 2. Governments or their agents test new varieties to 2. Companies do their own tests. Each seed decide whether or not to register them; that is, merchant decides what to market based on his or whether or not to allow seeds of the variety to be her own criteria; the merchant's integrity is at marketed. stake. For voluntary registration, variety descriptions are from breeder's own data, without any official review. Subsequent court cases challenging ownership of a variety could lead to review of published data. 3. EU Common Catalogues lists all varieties 3. There is no list of registered or approved approved for sale. Catalogues list varieties varieties, since seeds of all varieties are allowed registered in any one EU country. Seeds of listed for sale. varieties are allowed for sale in all EU countries. 4. Registration establishes legal rights. Every 4. Voluntary registration or patents establish legal registered variety is owned by a company or other rights. However, not all varieties are registered or party that pays fees for registering and patented. Some seeds are sold of varieties for maintaining registration. which there is no owner. 5. Seeds must be certified in order to enter the 5. Seed certification or official quality tests are not market. required for seeds to enter the market. However, all seeds must be labelled with information on species, variety (if known), and seed analysis information (germination, purity, weed seeds, inert matter, etc.) Jim Elgin works with the Agricultural Research Service, United States Department ofAgriculture and is the U.S. National Program Leaderfor OECD Seed Certification. I I I Recommendations to Increase the Flow of New Seed Technology to Farmers in Developing Countries Mark Condon The seed industry encourages the following Association of Official Seed Certification reforms to increase rates of introduction and Agencies (AOSCA, a US agency). Joining adoption of new seed technology by farmers in these groups would encourage seed test- the developing world: ing and certification programs that are acceptable to major seed trading coun- 1. Donors and developing country governments tries, which would facilitate seed exports. should promote integration of emerging or developing seed industries into the internation- (ii) International Seed Trade Federation. alpiivate seed industry atfollows: Membership entails the adoption of trade rules and arbitration procedures (a) Developing countries are asked to recog- already accepted by fifty four countries, nize that each country is one of many which would facilitate seed trade with potential markets for investment and trade, those countries. Membership would also and that companies choose where to work provide exposure to seed breeding, pro- and to invest. Therefore, from a competitive ducing, and trading organizations in perspective, developing countries need to other member countries. evaluate their regulatory systems, investment requirements, comparative advantages in (iii) Unionfor the Protection of New Plant seed breeding and production, and other Varieties (UPOV) and International factors that encourage or discourage atten- Association of Plant Breeders for the tion from the world seed community, Protection of Plant Varieties (ASSINSEL). Membership in UPOV entails adoption (b) Developing country governments are of plant variety protection legislation encouraged to seek membership in the fol- consistent with UPOV's 1978 or 1991 lowing international organizations. convention. Membership in ASSINSEL Membership will often entail restructuring provides a framework for technical and of private seed industries and trade along legal consultations to improve intellec- patterns consistent with the international tual property rights for seed technology. private seed industry. (iv) International Plant Protection (i) OECD Seed Schemes, International Convention (IPPC) and regional organi- Seed Testing Association (ISTA), and zations such as North American Plant Protection Organization (NAPPO), European Plant Protection Organization Mark Condon is Vice President, International (EPPO), etc. Membership in IPPC Marketing, American Seed Trade Association would foster rational and transparent (ASTA). 24 RECOMMENDATIONS TO INCREASE THE FLOW OF NEW SEED TECHNOLOGY phytosanitary standards and procedures, rather than variey registration and compulsory which would facilitate international seed seed certification systems. Such an approach trade. Membership in an existing or new would streamline new variety introductions regional phytosanitary organization and would facilitate private seed sector would further facilitate international interactions with farmers to develop prod- seed trade, especially regional trade. ucts that are appropriate for targeted farm- ing systems, while maintaining accountabili- (v) Food and Agriculture Organization ty to host country governments. Companies (FAO). Industrial and developing gov- that repeatedly distribute products that do ernments work together through FAO to not conform to government/international determine policies on issues related to standards on labelling, seed quality, etc, preservation and distribution of would be subject to registration revocation. germplasm. National seed trade associations could be consulted in the verification of a company (c) Donors are encouraged to work more or company's namne. closely with private seed trade associations as follows: (a) If a company registration system is not feasible for a particular country, then the (i) Private seed trade associations in next preferable scheme would be voluntary industrial countries (for example, ASTA variety registration schemes. Information and similar organizations in other indus- from variety performance/registration trials trial countries) are ready to advise and would be available for advisory purposes assist in the establishment of national only. Governments could maintain a list of seed trade associations in each devel- recommended varieties. However, the per- oping country-, to train government and formance trials to determine varieties eligi- company staff on technical and regula- ble to be placed on the recommended list tory matters; and to advise on national must be conducted by an unbiased third seed legislation and policies. party and not by government or public (ii) International Seed Health Initiative, institutions. a joint effort of the American, French, (b) Regardless of the scheme that is adopted Dutch, Japanese, and Israeli Seed Trade (for example, company or variety registra- Associations, promotes scientifically- tion), registration fees should be kept at a based phytosanitary regulations that minimum and registration procedures enhance seed quality and facilitate trade. should be simplified to encourage small and (d) Because of inordinate overhead costs in medium-sized companies to enter the seed sector. Participation and involvement across establishing joint ventures with existingabrdspcumosaltoagecp- public institutions, foreign private compa- niera broad spectrum of small to large compa- nie ofe ree to wor dietymihnes creates a desirable level of competiton nis ftnprefer to work directly with and an industry structure that is able to smaller private seed companies. Accordingly, defend free market policies. donors and developing country governments should emphasize trade and investment lib6r- (c) If compulsory variety registration contin- alization to facilitate entry of private foreign ues in force, seeds and crops produced for companies into emerging seed markets. export should not be subject to these variety restrictions. 2. Developing country governments should direct efoirt to establish company registration systems (d) Regional registration lists and programs EASING BARRiERS TO MOVEMENT OF PLANT VARIETIES FoR AGRICULTURAL DEVELOPMENT 25 (as in the EU) should be encouraged to the 4. Governments of developing countries should maximum extent possible. strengthen intellectual property rights in seed technology, as follows: (e) Governments should make private seed companies responsible for accurate labelling (a) Under no circumstances require compa- and dissemination of information about nies to deposit samples of in-bred lines with their products through simple but clear public agencies as a precondition for seed national laws. sales. These requirements are sometimes enforced in conjunction with compulsory 3. For variety development to meet specific coun- seed registration and certification systems. try needs, donors and developing country gov- ernments should: (b) Adopt either the 1978 or 1991 UPOV Convention, which entails passage of (a) Promote cooperation among public and national legislation establishing plant variety private research organizations - national protection meeting international standards. agricultural research systems, domestic pri- vate seed industry, international private seed 5. Governments of developing countries should industry, and international agricultural adopt liberal policies for seed import and research centers - to prioritize national export, asfollows research objectives and to identify compara- (a) Formulate and communicate transparent tive advantage in research, import and export regulations, especially (b) Consult with national seed trade associa- with regard to phytosanitary requirements; tions in the identification of interested pri- and vate companies. (b) As far as possible, abolish all import and (c) Devise arrangements for private compa- export permit systems. nies to work with NARS and IARCs so as to maximize synergy from cooperative research 6. Donors are encouraged to devise viable politi- activities. cal risk insurance schemes specifically for seed trade and investment. Recommendations for National Private Seed Industry Development Federico Poey The development of a strong private seed 3. Government should motivate and assist pri- system is crucial to improve the delivery of vate R&D by providing ready access to good quality seed to farmers. The level of agri- breeding materials from IARCs and NARS. cultural development and existing seed industry NARS should work hand in hand with local components are unique to each developing seed companies willing to do their own country, so that interventions to improve seed research to develop another source for systems must be tailor-made. The following hybrids and varieties in addition to multina- issues merit special attention: tionals and public sectors, thus contributing to more healthy competition. 1. It is important that the existing or potential private seed sector be properly represented 4. Public sector institutions can offer seed pro- and given a leadership role during early dis- cessing and storage services at cost to emerg- cussions to plan laws, policies, and strategies ing private seed companies, which would to strengthen private seed systems. market seed under their own brand names. 2. A mechanism to strengthen dialogue with 5. Credit and financial assistance should be the private sector is through formation of a available for fixed and working capital for national seedsmen association with a bal- emerging seed enterprises, supporting anced public-private sector participation. investments in infrastructure and in produc- Existing or potential seed producers, distrib- tion and distribution networks. Design of utors, importers, exporters, entrepreneurs, financial services should be adapted to needs and investors should be included in the pri- of the emerging seed industry. vate sector participation. Government should recognize the association as repre- 6. Multinational seed companies should have senting private interests, and encourage ready access to participate in national seed association discussions with government system development through convenient agencies involved in research, production, interactions with local firms such as joint quality control, marketing, regulation, etc. ventures, licensing, etc. 7. Variety registration and seed certification should be properly designed or adjusted to Federico Poey is president ofAGRIDEC help the private seed industry, not to hinder (Agricultural Development Consultants, Inc.), it former senior staff in the Seed Unit of CIAT The principle task of official certification is (Centro Internacional de Agricultura Tropical), to "audit" or confirm that companies main- and frequent consultant on national seed polic , a t .tam adequate qualty control durmg seed industry, and training issues. 28 RECOMMENDATIONS FOR NATIONAL PRIVATE SEED INDUSTRY DEVELOPMENT production, processing, and storage. cial seed and can support marketing efforts Government certification should by no of private seed companies. means substitute for each company's own internal quality control measures. 8. A "cottage seed sub-system " should be pro- moted for small farmers as a mechanism to However, official control should not include introduce new varieties and to improve qual- definition of what varieties are allowed to ity of seed for own-use or local sale. Care enter the market. This decision should be should be taken to avoid unfair competition taken by individual firms according to their with established commercial seed companies. own and available variety performance The sub-system is effective in handling seeds information and marketing objectives. of crops that may not be profitable for com- mercial seed companies, such as beans and In some countries, governments have tied vegetatively reproduced crops, and also in production loans or crop insurance to the supply see rote areas lacking ade- use of certified seeds. These arrangements suap rking infrstructure. encourage farmers to buy certified commer- q m Three Main Barriers: Weak Protection for Intelkctual Property, Unreasonable Phytosanitary Rules, and Compulsory Variety Registration Jon Geadelmann Two types of protection for intellectual ing countries have laws establishing pro- property in seed technology are: biological and cedures for companies to register owner- legal. ship of varieties. Preserving intellectual property * Patents. The US allows utility patents Biological protection is achieved with for normally bred varieties; the EU does nlOt. Both US and EU allow patents for hybrids; a company is able to maintain owner- . . ship of hybrids through physical control of products of non-traditional plant breed- inbred parent lines. ing research, such as for genes from bio- Companies can seek legal protection technology research. through various mechanisms: For non-hybrids, PVP laws are the primary - Trade secrets. In countries with trade mechanism to protect intellectual property. secrets laws, companies are able to seek Seed companies are concerned to clarify two redress from anyone who steals their issues in PVP laws: farmer's privilege and close trade secrets. Companies have to demon- breeding. strate that what has been taken was kept as a secret, ie, that the company made an Farmers privilege effort to prevent its dissemination. PVP laws vary from country-to-country in allowing farmers to save seed for own use and * Contracts and licenses. A company may to sell seed to neighbors. The new US law and a transfer seeds of a variety (for example, recent Supreme Court decision give US farmers an inbred line for hybrid production) the right to plant saved seed but not to sell through a contract that restricts what the saved seed to neighbors; the old US law allowed recipient of the seeds can do with them farmers to sell seed. The EU goes even further, (eg, use only to produce hybrid seeds, requiring farmers to pay royalty to the owner of not seeds of the inbred line). the variety on seed saved and planted. When farmers are allowed to sell seed to * Plant variety protection (PVP). neighbors, the breeder of a variety loses return Industrial countries and many develop- on research investment. This weakens incentive for private research, so that farmers lose in the long run. Jon Geadelmann is a plant breeder at Holden's Foundation Seeds, Inc. 30 THREE MAIN BARRIERS Close breeding sometimes refuses maize seed shipments after With PVP protection, a company that owns finding fusarium kernel rot, although the fun- a variety may lose if another company is able to gus is present in Japan. take that variety, breed to develop a very similar But the lack of good phytosanitary barriers but new variety (for example, with different silk inflicts serious losses. For example, the United color, but otherwise the same), and register States has lost from import of Dutch elm dis- ownership of that new variety. Traditional crite- ease and European corn borer from Europe. ria for a new variety require that seeds presented Currently, phytosanitary rules pay insufficient for testing produce plants that are distinct from attention to maize streak virus, which is limited other varieties, uniform, and stable over time to Africa but could spread. Another pest which (DUS criteria). is limited but could spread is corn lethal necro- To protect variety owners from loss due to sis, which is found in parts of Midwest United close breeding, the 1991 UPOV Convention States. introduced the concept of Essentially Derived Variety (EDV), enjoining member governments Compulsory variety registration to adopt PVP laws that do not allow anyone to Many governments allow seed companies to register ownership of a new variety that is essen- sell seeds of only those varieties that have passed tially derived from a protected (owned) variety. some sort of government-controlled evaluation The measurement of EDV is not yet exactly of the performance of the variety. Variety per- defined; test cases over time may establish formance tests with compulsory variety registra- threshold values for EDV. tion are known as VCU (value in cultivation and use) trials, official trials, etc. Other intellectualproperty issues Systems requiring variety registration as a The 1992 Convention on Biological condition for seed sale are not the same as PVP Diversity gave attention to claims of some gov- Variety registration based on variety performance ernments for payments from private companies tests do not establish ownership, but rather per- for germplasm taken out of their territory that mission to sell a variety. PVP, on the other hand, is used in breeding commercial varieties. can be voluntary and totally unrelated to perfor- Discussions on this issue have continued in mance; for PVP, seeds are tested against DUS FAO. Private companies fear loss of access to and EDV criteria as already discussed. germplasm or charges that would interfere with In the basic system for compulsory variety development and marketing of new varieties. registration, a government agency performs small-plot tests at relatively few locations, com- Phytosanitary regulations on seed import paring new varieties against standard or check We need phytosanitary rules based on scien- varieties. Only those new varieties that exceed tific risk assessment. Current rules may be standards by some threshold value are permit- unreasonable and nonscientific, creating unnec- ted to be sold to the farmers. Traits commonly considered include: yield, maturity, resistance to essary barriers to seed movement on the one hand, and ignoring some real phytosanitary lodging, etc. threats on the other. Proponents of variety registration argue that Unreasonable phytosanitary rules limiting government tests protect farmers from poor varieties and also provide useful information to d td ma timents. or , the farmers to help them choose among varieties by protectionist sentments. For example, te lowed for sale. governments of New Zealand and United States Howed fo r sale. limit maize seed imports based on concern However, there are disadvantages and weak- lii maiz sediprsbsdo. ocr nesses in systems with compulsory variety regis- about corn smuts, although evidence suggests tessins that these are present in both countries. Japan EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 31 * small plot official trials often produce a Sununary suggestions very limited database compared to pri- 1. Pass laws establishing PVP consistent with vate company small plot research tests the 1991 UPOV Convention, that is, that and larger-scale tests; limit a farmers right to sell seed to neigh- bors and that limit close breeding. * small plots do not fully represent the real world, that is, the farmer's field; 2. Develop and maintain reasonable and fair * seed source effects are often confounded phytosanitary regulations and inspections. with genetic performance; 3. Eliminate official trials, or at least eliminate * compulsory variety registration may performance thresholds (as a condition for eliminate lower-performing but much seeds or a variety to be marketed), and use less expensive varieties from the market; only as a source of information for advice to and farmers on choice of variety. * compulsory variety registration diverts government and seed company resources from more productive uses. World Phytosanitary System: Problems and Solutions Denis McGee Seeds can be efficient means of moving mined either by a field inspection of the grow- pathogens between geographical regions. There ing crop or by a laboratory assay of the seeds are many documented introductions of harvested from the field. pathogens from one country to another by The current world phytosanitary system has seeds that resulted in economic loss (Neergaard significant problems. This paper will discuss the 1977). There is no question that regulations are reasons for and consequences of these prob- necessary to protect against the international lems, and suggest possible solutions. spread of plant pathogens. International move- ment of seeds is regulated by a standard phy- Problems with the world phytosanitary tosanitary certificate, as defined by the system International Plant Protection Convention of Problems with the world phytosanitary sys- 1951, which indicates that seeds are substantial- tem fill into five categories, as discussed below. ly free of injurious pests or diseases. It also has a section to deal with specific diseases or pests for Poor understanding of potential economic losses which the importing country may require tests. from introducingpathogens In addition, each country has its own specific This can be attributed to lack of access to requirements regarding type of seeds and information on the pathogen and to insuffi- pathogens that may be admitted as follows: cient research into the epidemiology of seed- • It may totally embargo particular seeds. borne diseases. A consequence of this problem is the proliferation of unjustified phytosanitary * It may require that some seeds be test- regulations that impair the movement of seeds ed after import. worldwide. Resources that could be applied to protecting against the spread of economically * It may ask that some seeds be treated, important seed-borne pathogens are wasted on for example, with a fungicide. pathogens of minimal importance. An example is the requirement that all soybean seed lots Information for the phytosanitary certificate exported from the United States to the is generated and reported by the country European Union (EU) have to be tested for exporting the seed. Seed health may be deter- Pseudomonas syringae pv glycinea, the cause of soybean bacterial blight. This disease is wide- spread in the United States and has no detectable economic impact. It also has been Denis McGee is with the Seed Science Center, present in Europe for many years. Since 1988, Iowa State University. US seed companies have spent approximately 34 WORLD PH-yrOSANITARY SYSTEM: PROBLEMS AND SOLUTIONS $1,500,000 on testing for this pathogen on soy- this occur with considerable frequency in the bean seeds to meet phytosanitary requirements. seed industry. A second consequence of this problem is Lack of knowledge of relationships between that there are seed health tests in common use tolerances in seed assays to risks of transmission of that have never been subjected to international the pathogen to the planted crop standardization. This also is a result of lack of access to infor- mation on the pathogen and of insufficient Different seed treatment regulations among research into the epidemiology of seed-borne countries diseases. One consequence of this problem is This results from variations in regulations that it leads to the use of non-scientifically- and standards for efficacy and environmental based tolerances by default. Pseudomonas risk of pesticides between countries. The main syringae pv glycinea must be detected in each of consequence of this problem is confusion five subsamples of 1 kg of soybean seeds to be regarding the legality of use by the exporting denied a phytosanitary certificate. This toler- country of particular treatments and their rates ance was determined by negotiation between of application. US and EU authorities and has no epidemio- logical foundation as a means of protecting Trade barriers against transmission of the pathogen by seeds. There is strong evidence that use of phy- A second consequence of this problem is that tosanitary rules as trade barriers is becoming sample sizes may be inadequate or imprac- more prevalent with the development of free tical. A grow-out test commonly used to detect trade areas such as NAFTA and the EU, and by Erwinia stewartii in corn seeds uses 400 seeds completion of the recent GATT agreement. (McGee 1988). In the fifteen years the test has Because these agreements remove tariffs as bar- been used in the Iowa State University Seed riers to trade, other instruments such as phy- Testing Laboratory, the pathogen has never been tosanitary certification can be used to achieve detected in a corn seed lot. Recent research indi- the same purpose. cates that 10,000 seeds would have to be tested A consequence of use of phytosanitary rules to give any realistic chance of detection of as trade barriers is widespread cynicism about Erwinia stewartii (Block 1995). the whole phytosanitary system that leads to ineffectual testing of seeds in the exporting Lack ofstandardization in testingprotocols countries and to mistrust during negotiations This problem results from the fact that, between countries on phytosanitary issues. apart from sixty-five working sheets on seed- borne diseases published by the Plant Disease Solutions to phytosanitary problems Committee of the International Seed Testing Five actions are suggested to improve the Association (PDC-ISTA), there has been no world phytosanitary system. systematic effort to develop seed health test methods that are accepted internationally. One Better information sourcesfor seed-bornepathogen consequence of this is that different tests may A good starting point to obtain information be used by exporting and importing countries. on seedborne diseases is An Annotated List of In 1994, Erwinia stewartii was detected in Seed-borne Diseases (Richardson 1990). This Europe by a serological procedure in corn seeds publication lists all seed-bore microorganisms that had been issued a phytosanitary certificate that have been recorded in the scientific litera- based on the grow-out test in the United States. ture as being associated with seeds of all crops. A shipment of 20 metric tons of corn seed had Unfortunately the publication can be misused to be returned to the US. Experiences such as EASING BARRIERS TO MOVEMENT OF PLANT VARIEnES FOR AGiCULTURAL DEVELOPMENT 35 because it does not provide information neces- Provide ready access to scientific infor- sary to justify phytosanitary regulations such as mation needed to make rational and jus- economic risks from introducing the pathogen tifiable decisions on plant quarantine or whether the pathogen can be transmitted by regulations. seeds. Seed-borne microorganisms may be placed on phytosanitary regulations merely Standardxzation ofseed ealth test methods because they are listed in this publication.' The working sheets on seed-borne diseases There is an extensive world literature on seed- published by the PDC-ISTA are the product of borne diseases that contains answers to many of working groups assigned to develop protocols these questions. It remains largely untapped, for each pathogen. This is a thorough process however, due to a lack of systematic organiza- that is a good model for standardization of tion of the information and difficulties in tests. Only sixty-five sheets have been devel- accessing hard copy reference sources, particu- oped over forty years, however. The PDC must larly in developing countries. Recent advances become more productive in the supply of in the electronic media present opportunities to working sheets and in promoting them as disseminate data efficiently and economically methods that should be accepted international- throughout the world. An international data ly if it is to meet current demands for standard- base addressing key aspects of seed-borne ized tests. pathogens could be of great value (table 2.2). Since 1994, European and US seed indus- tries have formed an active international Such a database would provide the follow- committee comprising scientists from the ing: private and public sectors, who are now • An information source of unprecedented working on standardization of methods for csand power that will twenty-five economically important seed-borne compree ns en ess pathogens of vegetables (Wesselling 1996; proviem a u eresoure foreffeci Maddox 1996). It is expected that methods for m,anagement of seed-borne diseases. fedco ed iladdltr - ~~~~~~~~~~~field crop seeds will added later. * Enhancement of the capacity of users to identify seed-borne pathogens, assess Access to worldwid seed tratment regulations their economic impact, and devise con- This is a difficult problem to resolve, trol strategies. because regulations are changing constandy. Making current information available in data- Table 2.2 Proposed database on seed-bome diseases Scope: * All major crops Structure: * Potential economic impact of planting infected seeds * Worldwide distribution * Incidence of seed-bome infection * Effect on seed quality * Transmissibility of the pathogen by seeds * Control by seed treatment * Seed health assays Output: * Electronic and book format 36 WORLD PHYrOSANITARY SYsTFM: PROBLEMS AND SOLUTIONS bases that deal with regulations would greatly criteria of maize chlorotic motde in that the alleviate the problem. Regulatory databases pathogen is limited in distribution and is of such as the EXCERPT system of potential economic importance where found USDA-APHIS are already in existence or are (McGee 1988). being developed. Previous reviews of seed pathology have stressed the importance of establishing inocu- Improved communication and cooperation lum thresholds (tolerances) for seed-borne between phytosanitary authorities in different pathogens (Neergaard 1977; Kuan 1988.). countries Only a handful of the citations examined for One action that would greatly improve the this review addressed this topic, however. environment for communication between Research on inoculum thresholds is both com- countries on phytosanitary issues would be for plex and expensive. However, it is so funda- each country to make a realistic appraisal of the mental to realistic and effective management of diseases present in their country. Authorities seed transmission of plant pathogens that little should not rely solely on published records of improvement in the worldwide seed health sys- pathogens, but accept the findings of "experts" tem will be possible unless priorities in seed who work in these countries and are knowl- pathology research are changed to meet this edgeable about plant disease situations. demand. Revision ofpriorities in seedpathology research Condusion A recent review of the literature on seed The overall goal of a world phytosanitary pathology over the period 1982-94 (McGee system is to protect against the spread of eco- 1995), indicated that 23.5 percent of approxi- nomically important pathogens without posing mately 2,000 citations simply catalogued the unnecessary barriers to worldwide movement of presence of microorganisms on seeds. These seeds. The present world phytosanitary system types of publications are purely descriptive and is not doing this effectively or efficiently. do not address the potential for crop damnage by Resources are being squandered on unnecessary planting diseased seeds or the management of regulations, and many potentially important seed-borne disease. Indiscriminate cataloguing seedborne pathogens are being ignored. of seed-borne microorganisms on seeds Implementation of the suggested improvements obscures seed-borne pathogens that might be of could lead to a system that is more effective in genuine economic importance. A case in point preventing spread of pathogens and less expen- is maize chlorotic mottle which was first report- sive to seed companies and governments ed in 1973 (McGee 1988). No serious effort throughout the world. was made to determine that the pathogen was seed-borne until it caused an epidemic in win- ter nurseries in Hawaii. Because of the concerns References of importing infected seeds into the US, seed Block, C. 1995. 'Biology of Seed Transmission of Erwinia transmission of the pathogen was investigated stewartii in Maize", unpublished PhD dissertation, and shown to occur (Jensen et al 1991). IowaStateUniversiq. Cataloguing of seed-borne pathogens is neces- International Seed Testing Association (ISTA). 'Working sary, particularly for viruses and bacteria which Sheets on Seedborne Diseases." Zurich: ISTA. traditionally have been neglected due to a lack Jensen, S G, D S Wysong, E M Ball, and P M Higley. of adequate seed health assays. Priority, howev- 1991. "Seed transmission of maize chlorotic motde er, should be given to pathogens that meet the virus," Plant Disease 75:497-498. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 37 Kuan, T L. 1988. 'Inoculum thresholds of seedborne McGee, D C. 1988. Maize Diseases: A Refirenee Sourcefor pathogens: overview," Phytopathology 78:867-868. Seed Technologists. St Paul: APS Press. Maddox, D. 1996. "Regulatory Needs for Standardized Neergaard, P. 1977. Seed Pathology, vols I and II. New Seed Health Tests," in D C McGee, ed, Plant York: John Wiley & Sons. Pathogens and the Worldwide Movement of Seeds. St Paul: APS Press. Richardson, M J. 1990. An Annotated List of Seed-borne Diseases, 4th ed. Zurich: ISTA. McGee, D C. 1995. "An epidemiological approach to dis- ease management through seed technology," Annual Wesselling. 1996. "Govemment/Vegetable Seed Industry Review ofPhytopathology 33: 445-466. cooperation in the Netherlands," in D C McGee, ed, Plant Pathogens and the Worldwide Movement of Seeds. St Paul: APS Press. Seed Regulatory Frameworks and the Availability of Crop Varieties Robert Tripp Strategies and visions of agricultural devel- bution. They may be divided into two broad opment vary widely among governments, categories: those regulations that determine the donors, NGOs, and community groups, but type of products (varieties and seeds) that are one goal they are likely to share is the provision available to farmers; and those that are con- of better crop varieties. Despite significant cerned with controlling the quality of the prod- differences regarding the most appropriate ucts once they reach the market. This paper direction for agricultural development, there is addresses the first of these two categories, which little disagreement that farmers deserve access includes: the procedures and practices that to a wider range of crop varieties than they guide the conduct of plant breeding; the rules currently enjoy. One of the impediments to governing the official release of new varieties; achieving this goal is the regulatory frameworks and restrictions on the types of varieties that that govern the conduct of plant breeding, seed may be used or offered for sale. The implica- production, and seed sale. tions of these regulations for seed imports and for the operation of private seed enterprises are Introduction also discussed. This paper outlines the relationship between An examination of seed systems must start seed regulatory frameworks and farmers' access with plant breeding. Although the procedures to seed of crop varieties. The first section intro- of plant breeding are usually not considered a duces the subject of seed regulations and part of formal seed regulations, they are strong- reviews how factors related to demand for seed ly influenced by variety release requirements and varieties affect public and private sector and perceptions of marketing regulations, and roles in national seed systems. The second sec- hence deserve attention here. Public sector tion of the paper discusses how current seed plant breeders are guided by national and insti- regulatory frameworks limit the availability of tutional policies in setting priorities for varietal crop varieties. The final section describes several characteristics and identifying target farming strategies to address these problems. populations. They are also influenced by the requirements of variety release, and will natural- Seed regulatoryframeworks ly conduct their breeding programs to maxi- Seed regulatory frameworks are the rules mize possibilities for approval. Private sector and regulations, and the norms, guidelines and plant breeders concentrate their efforts on crops standard practices, associated with crop variety and varieties where there is the best chance of development, seed production, and seed distri- commercial success, but are often limited by restrictions on access to markets. Variety release requirements, which are usually in the hands of Robert Tripp is on the staff of the Overseas public entities, may also affect private sector Development Institute in London. plant breeding initiatives. 40 SEED REGULATORY FRAMEwoRKs AND THE AVAILABIuTY OF CROP VARIETIES Variety release procedures are often man- entry to seed production, particularly for cer- aged by committees, which can have consider- tain crops. able power. As described by Douglas (1980, p 57), they can: Seeds and varieties establish guidelines for considering vari- Before proceeding with a discussion of the effects of seed regulatory frameworks on the reties fromrpbli adeprmive pathr availability of crop varieties, it will be useful to barietied pre togbera cmms eeinded whether define a few terms. Seed regulatory frameworks varieties are to be recommended, consid- cre, "suitable," or listed as "unsuitable"; are meant to facilitate the operation of national ered"sutabl," r lstedas unsutabe"; seed systems. Seed systems can be divided into establish criteria for accepting varieties as foma (public or prvte) an d info eligible for seed certification; and assume responsibility for an allocation policy for farm or community level) sectors. In addition, seed of new varieties, any discussion of seed systems needs to clearly distinguish between demand for varieties and demand for seed quality. Seed is clearly the rolety e committees ulay akey vehice for any new crop variety, but in many worolekinsthe. conduct of seed regulatoryframe- cases farmers are able to maintain a variety once wvorks. Variety release decisions helthey have acquired an initial amount of seed. ran fvariety ease deson hel dermIne the This is particularly true for self-pollinated crops range of varieties available to farmers. In some sc src rwet nteeisacs cases official release is only required for public demand efor willtbeIloselyrelateatoeth sectr vrieies whle rivte ectr vrieies demand for seed will be closely related to the sector vreelymaries,hIe pither seto privareties development and introduction of new varieties. may e feelymarete. Inothr cses,priate If variety turnover is infrequent, demand on the sector varieties must be submitted to the same release procedures. In addition, seed certifica- formal seed sector will be correspondingly low. tion, which verifies that seed offered for sale In other cases, seed may be purchased every sea- conforms to the chaaceson, because hybrid varieties are used, or because of difficulties in seed storage or process- established in the release procedure, may be ing that make it more likely that farmers will compulsory (in which case only seed of official- look to the formal seed sector for better quality ly released varieties is available) or it may be seed. voluntary. Thus seed demand is a function of farmers' The availability of crop varieties is also interest in new varieties and farmers' require- affected by import regulations. Imported van- ments for an external source of seed of current eties often have to go through the same variety varieties. Distinctions between these two poten- release procedures as domestically produced tial sources of demand are important for under- varieties. In addition, some countries impose standing the overall demand for seed and iden- strict limitations on the importation of seed of tifying the most appropriate orientation for the particular crops, often related to policies that seed sector. protect national seed industries. Finally, seed regulatory frameworks may Public andprivate roles in seed systems impose restrictions on the ability of the private seed sector to compete with public crop vani- The implications of these characteristics of eties. The regulations regarding variety release seed demand for the relative responsibilities of or seed certification may be managed in such a the public and private sector in national seed way that the private seed sector is effectively systems are discussed in several recent studies barred from participation. In other cases, there (Pray and Ramaswami 1991; Jaffee and may be explicit limitations to private sector Srivastava 1994). In general, it is widely recog- nized that public sector seed production and EASING BARRIERS TO MOVEMENT OF PLANT VARIEnEs FOR AGRICULTURAL DEVELOPMENT 41 distribution has often been inefficient, and that are in fact the result of pressure by interest there is much scope for the expansion of the groups, and Bernstein's (1955) analysis of the private seed sector in developing countries. life cycles of several US regulatory agencies Nevertheless, for crops or areas where seed shows how each comes to be captured by the demand is low or uncertain there is little to interests it was established to regulate. attract conventional private initiatives. These Regulations and regulatory agencies can also conditions obtain for less commercial farming assume a bureaucratic life of their own that may areas and for self-pollinated crops, and they work at cross purposes to the original regulato- represent a larger proportion of total seed use ry goals (Mitnick 1980). than is sometimes acknowledged. There is thus The following discussion outlines problems a need for some type of public sector participa- with current seed regulatory frameworks that tion in seed supply, perhaps directly, or more contribute to restrictions on the range of crop probably by supporting or contracting private varieties available to farmers. The problems and community level seed production and dis- indude inadequate assessment of demand for tribution. varieties, lack of transparency in the regulations, In contrast to seed production and distribu- problems with the costs or logistics of imple- tion, plant breeding and variety development menting regulations, and lack of opportunity are areas that will continue to require a direct for wider institutional participation in the regu- role for the public sector in many countries. It latory process. is often difficult for private enterprises to meet the costs of basic plant breeding research. Even Inappropriate breedingpriorities in well developed seed systems, public sector The results of the Green Revolution involvement in plant breeding continues to be brought a number of critiques of plant breeding important (Knudson 1990). Thus our examina- concerning the assignment of priorities between tion of the effect of seed regulations on variety more and less favored environments (Lipton availability needs to pay particular attention to and Longhurst 1989). Although there has been public plant breeding, and to acknowledge that considerable effort in the past two decades to increasing privatization of the seed sector, how- address the needs of farmers in what Chambers ever attractive, will not by itself resolve many of (1991) calls 'complex, diverse and risk-prone" the problems presented by current seed regula- environments, the generally lower utilization of tory frameworks. modern varieties in these farming systems indi- cates that the challenge remains. On the one Problems with current seed regulatory hand, it is undoubtedly more difficult to devel- firameworks op appropriate varieties for marginal environ- Seed regulations are usually established and ments, and some analyses show that the propor- defended on the grounds of consumer protec- tion of resources invested in breeding for these tion. Farmers, and the general public, need to environments is reasonable (Byerlee and Morris be protected from the consequences of wide- 1993). On the other hand, farmers who are spread use of inappropriate varieties, and seed more isolated and command fewer resources are regulations help to address market failures in less likely than better organized, more commer- the processing and management of informa- cial farmers to have their voices heard by public tion. Although this view of seed regulations is research institutions. Resource allocation in accurate in many cases, there are also instances public sector agricultural research is far from where the performance of seed regulatory transparent (Busch and Lacy 1983), and much frameworks is most usefully analyzed using can be done to improve responsiveness to farm- other interpretations of regulatory regimes. ers' concerns. Stigler (1971) proposed that many regulations 42 SEED REGULATORY FRAMEWORKS AND THE AVAILABILITY OF CROP VARIEnES A more immediate indication of inefficien- dure advises the appropriate authorities of the cies in public sector plant breeding are the variety's availability for seed production. Variety instances where farmers select or improve upon release or approval often requires several years discarded breeding lines or materials that are of performance testing, as well as recording the being tested to develop their own varieties. morphological characteristics that distinguish There are a number of examples where such the variety from others. farmer initiative has led to the development of Variety release is usually the result of a com- widely used varieties (Green 1987; Maurya mittee decision. Due (1990) describes the 1989; Salazar 1992). This is evidence that plant process of releasing a bean variety in Tanzania, breeding procedures are often inadequately where the Grain Legume Coordinating organized, or bureaucratically constrained, and Committee must send a recommendation to illustrates the lack of congruence between offi- the Variety Release Committee, which in turn cial release standards and farmer priorities. forwards instructions to the Seed Production Committee. Because funding for such commit- Delays in the variety testingprocess tees is usually insufficient, and they may com- Even when breeders are able to accurately prise representatives from geographically distant assess farmer priorities, the process of variety institutions, meetings are often cancelled or testing can be very protracted. In Kenya (a typi- postponed. The rules and operating procedures cal example), a promising variety must be of variety release committees may also be induded in three years of national performance unclear, and their legal status is often not trials before being considered for release (Ruigu defined; five out of seven southern African 1988, p 138). Because several cycles of multi- countries recently surveyed had variety release plication from breeder seed are required to pro- committees with no legal backing duce sufficient seed for sale or distribution, it is (Commonwealth Secretariat, 1994). more than five years between the time that a This lack of transparency in variety release variety is identified and when it is available to committees allows considerable scope for per- farmers. There are certainly dangers of rushing sonal or professional politics. Ferguson (1994) a variety to release without adequate assess- points out that personal or prestige factors ment, but in many cases requirements for a often play a role in committee decisions, and fixed number of performance trials represent Douglas (1980, p 57) cautions against the prac- bureaucratic whims rather than technical opti- tice of appointing committee members on a mums. In countries where private sector vari- political basis. These committees meet infre- eties must pass through the same testing process quently, and because the deliberations do not before official release, they may face the added represent exceptionally important concerns for disadvantage of having to compete on a playing many of the members, there are opportunities field designed by their rivals in the public sector for the pursuit of particular personal interests, research institution. and less possibility for negotiation and compro- mise. Senior plant breeders may have consider- Variety releaseprocedures able power in determining which junior staff In many countries, a variety cannot be sold are rewarded, or the viewpoints of particular unless it has been officially released. This is par- disciplines, such as plant pathology, may be in or out of favor on a committee. Varieties from ticularly true of public sector varieties, but at times the private sector must submit its varieties the private sector, international research organi- to the same procedures. In some cases there are zations, or NGOs may not be given a fir hear- two distinct steps; a variety is first approved for ing. Cromwell and Wiggins (1993, p 57), for instance, describe the delays in the release of a release, and then a second (separate, although soyabean variety in Bangladsh that had been normally straightforward) notification proce- EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 43 promoted by an NGO. became established before the initiation of for- Even the formal standards used by a variety mal release prodcedures. Similarly, varieties that release committee may be problematic. Joshi have been released in the past may be removed (1995) reports the case of an early maturing from a national register if demand drops or the rice variety in Nepal that, although lower yield- variety is found to be susceptible to pests or dis- ing than other varieties, fit nicely into a rota- eases. tion practiced by some farmers. The variety Kelly and Bowring (1990) and Bould could not be submitted to the release commit- (1992) describe the evolution of the variety reg- tee because of its lower yield, despite evidence istration system in the UK Prior to 1964, there of demand by farmers. Standards for variety were no official systems for registering varieties release may depend on other factors as well; or certifying seed, although voluntary schemes Ulrich et al (1987) analyze the losses to were in practice. The Plant Varieties and Seeds Canadian wheat farmers caused by restrictions Act of 1964 established an index of variety on the release of a variety judged inappropriate names and introduced performance trials, as for Canada's export strategies. well as allowing variety protection under plant Although variety release requirements may breeders rights. The index provided descriptions be defined for major crops, they are often not of each variety which could be used in seed established for crops such as forage species quality control. A variety's inclusion in the (Ferguson 1994), and the uncertainty associat- index did not, however, depend on any prede- ed with this lack of definition often serves as a termined results in performance trials, and the disincentive to plant breeders for developing index was used instead to provide recommenda- new varieties. tions to farmers. Entry to the European The importance of adequate variety testing Community in 1972 means that UK seed law should not be underestimated, however. An now has to conform to EU directives. No vari- accurate assessment of a variety's susceptibility ety can be marketed unless it appears in the EU to disease may require several cycles of observa- Common Catalogue, which is composed of tion, for instance. But the threat of major epi- national lists of member states. A variety on a demics is also increased by the cumbersome national list in an EU country must not only be variety release and distribution procedures cur- distinct, uniform and stable (DUS) but also rently in place in many national programs, leav- show value for cultivation and use (VCU). ing large areas planted to a few varieties (Heisey In India, all public sector varieties must pass 1990). The problem is compounded when through release procedures and be notified for authorities ban older, susceptible varieties that seed production, although the actual registra- farmers are accustomed to without offering ade- tion process is based on a few varietal character- quate replacements. It can be argued that vari- istics, rather than the comprehensive DUS test- ety release can be made more agile without ing practiced in the EU (Turner 1994, p 11). In jeopardizing the legitimate concerns of plant the United States, varietal registration is not protection. required, and determination of performance is left to the market, although several voluntary National Variety Registers variety testing schemes are in operation (Kelly A distinction can be made between variety 1989, p 42). A voluntary scheme operates in release and the function of a national variety New Zealand as well, but before seed can be register that defines the varieties currently eligi- certified the variety must be accepted on a ble for seed production. In some cases a variety national list that requires proof of uniformity may be included on a national register without and stability (Hampton and Scott 1990). having been submitted to testing for variety Variety registers may function in different release; this is especially true for varieties that ways. Gisselquist (1994) distinguishes among: 44 SEED REGULATORY FRAMEWORKS AND THE AvAILABILITY OF CROP VARIETES those countries that require no prior approval related to the International Union for the for variety sale; those where seed sale is restrict- Protection of New Varieties of Plants (UPOV) ed to varieties on national lists; and those that convention that came into force in 1968. The utilize a common regional list, as practiced in recently concluded GATT negotiations require the EU and under consideration by members of member countries to provide intellectual prop- the Southern African Development erty protection for plant varieties, and many Community (SADC) (Commonwealth countries are moving toward establishing a vari- Secretariat 1994). etal protection system. The issue has been the To the extent that crop varieties must pass subject of considerable debate, particularly with through a release procedure and be registered, it respect to possible impacts on biodiversity is legitimate to debate the type and level of (Crucible Group 1994). There is concern that standards to be used. With increasing technical increased protection may threaten farmers' cus- sophistication, the standards and testing proce- tomary practices of saving and exchanging seed, dures can become quite costly. In general, the reorient plant breeding priorities towards the trend is to pass the costs of testing to the breed- more affluent farmers, or even take control of ers and seed companies. local varieties away from farmers. The type of There is considerable debate over the extent plant variety protection to be instituted in to which exceptionally precise standards for response to GATT is still being discussed, how- variety registration or approval are always in ever, and it will be several years until the results farmers' interests. Berg et al (1991), for are clear. In the meantime, however, there is instance, question the emphasis on varietal uni- much that can be done with respect to reform- formity for farmers in environments where the ing conventional seed regulations in order to stability and adaptability of local landraces are achieve wider availability of new crop varieties. due in part to their intrinsic variability. It is worth listening to the views of an eminent Seed certification plant breeder who is concerned about two Seed certification takes place after varieties issues (Simmonds 1979, p 225): become available, but its conduct has consider- able bearing on variety availability and hence it first, the possible stultifying effect of toodervsomdicsonhe.Sdcabee- rigi an aplcto of DU crtei deserves some discussion here. Seed can be cer- (which are leial ratherio n DSritera tified only if the variety has been officially regis- (ich artent); l reron tha agricutura tered. The term "seed certification" is subject to in ie;de dtseveral definitions. In its strict sense (and the of VCU principles (which are certainly one used here) it is a measure of genetic purity, unnecessary and expensive and can be assessing the degree to which the seed conforms regarded as potentially dangerous to the characteristics of the registered variety. farming)... The VCU criterion seems to But certification also often includes measures of have come to stay, however; if interpret- seed physical quality, such as germination ed broadly it may work quite well. but capacity or freedom from weed seed. Grobman the risks and disadvantages of narrow (1992, p 141) is concerned that the two mean- interpretation are real; unrestricted lists, ings "have been hopelessly confounded and even if they were longer, would probably unfortunately integrated into a single process" have served farming better, in many countries, and urges that the second usage be referred to as seed quality control. He Simond's emaks remad intheconex emphasizes that all seed offered for sale needs to of a discussion on plant breeders' rights, where emsubec to see of qual nto verypreisc istncton aongvaritie is be subject to some type of quality control, veryupreci . s edistincioamn varietoieisl preferably managed by seed enterprises them- selves, with the government as the final arbiter. EASING BARRIERS TO MOVEMENT OF PLNT VARIETIES FOR AGRICULTURAL DEVELOPMENT 45 Seed certification, on the other hand, is open to unreasonably high, and seed enterprises may more debate. lose valuable stocks because of the rigid judge- Seed sold in the EU must be officially certi- ments of a seed inspector (Chopra 1986). In fied, while in the United States certification is addition, official seed certification is subject to voluntary (although widely practiced for certain significant abuse; it presents opportunities both crops) and is carried out by private associations. for bribes to pass unacceptable material and for In India, seed of all public sector varieties is additional payments to inspect legitimate seed. supposed to be certified, whether the seed is On the other hand, where certification is produced by public or private enterprises. optional, lack of certification may be equated Private sector varieties (mostly hybrids) are not by farmers to lower quality and may deny the certified because they have not been through seed access to government extension or farm official release procedures. In India, seed that is credit support programs. not certified may be sold as "truthfully labelled seed." A number of countries allow two classes The private seed sector and seed imports of seed to be sold, one which meets certification A range of laws and regulations may be standards and another which has not been sub- used by governments to limit the development jected to official inspection (Bombin-Bombin of a domestic private seed sector or to control 1980, p 7). Garay et al (1988) describes how a the import of foreign seed (Pray 1990; Pray and reorientation of seed regulation procedures in Ramaswami 1991, p 29). In addition, seed reg- Bolivia allowed several certification options; as ulations themselves, presumably established for farmers and seed producers gained more experi- consumer protection regardless of the source of ence, there was a gradual shift to higher stan- seed, can be manipulated or interpreted to sup- dards. port government objectives of discouraging Seed certification may not always be neces- competition from the domestic or foreign pri- sary, and may impose financial or logistic bur- vate sector. dens on national seed systems. Private seed In some countries the private seed sector enterprises will tend to impose their own con- may face considerable obstacles. Until recently, trols on the seed production process to ensure all public sector crop varieties in Zimbabwe that the seed they sell is of adequate genetic have been produced by one company, the Seed purity. The costs of official seed inspection are Coop (Friis Hanssen, 1992). In some countries, normally charged to the seed producer (and variety release is managed by the public sector passed on to the farmer). Agrawal and Tunwar agricultural research institution, which finds (1990) estimate that seed certification costs in itself controlling entry of varieties from its com- India are generally equivalent to 2 to 4 percent petitors in the private sector. There is currently of the sale price of the seed. Certification in one considerable debate in India regarding variety Indian state may not be accepted in another, registration and certification. Private seed com- however, which complicates the movement of panies prefer not to submit their varieties (par- seed across state boundaries. ticularly hybrid parents) to the government for Kelly (1989, p 111) points out that the high registration, partly out of fear that their materi- costs of EU seed certification are ultimately al will be appropriated by other breeding pro- borne by the farming community. In countries grams. Private companies in India are under- where all seed production and inspection is standably concerned about proposals for com- done by the public sector, lack of funds may pulsory seed certification (Turner 1994, p 15), restrict the coverage of the certification service and argue that the bureaucratic requirements of and, where certification is obligatory, limit the such a scheme would surely disrupt seed supply. supply of seed. There is often inconsistency in the manage- Standards used for certification may be ment of seed and varietal imports. In Nepal, the 46 SEED REGULATORY FRAMEWORKS AND THE AVAILABIuTY OF CROP VARIETIES government is discussing the possibility of ond deficiency is that seed regulations are not exempting Indian public sector grain crop vari- well established, widely understood, or consis- eties from passing through the Nepali variety tently interpreted. This causes uncertainties testing and release procedures that are currently regarding possible options to improve the avail- required of them. This would seem a justifiable ability of new varieties or seed. Finally, the procedure, given that these varieties have already management of regulations that affect variety been tested in similar environments. The private availability is often in the hands of small groups sector, however, can import vegetable seed from or committees who may not represent, or have India without any testing requirements, and access to, the interests of important sectors of inappropriate varieties have at times entered the the farming population. market (Rajbhandary 1994). An obvious solution would seem to be the Plant quarantine is a key element of seed introduction of more appropriate regulations, regulations, and is legitimately used to protect a but our emphasis here is rather on strengthen- country from the introduction of pests, diseases ing institutions. Institutional change must or weeds. It also can be used as a pretext for accompany regulatory reform. Indeed it can be limiting seed imports, however. Douglas (1980, argued that regulations are products of institu- p 131) warns against the overzealous use of tions, and that the weaknesses of current regula- plant quarantine laws where there is often little tory frameworks reflect deficiencies in institu- justification. tions. This interpretation has several implica- Finally, it is worth noting that modifications tions for the design of development programs. in seed regulations to facilitate seed trade or to First, strengthening national seed systems will meet domestic industrial agricultural standards be achieved not merely by changing seed regu- may have negative consequences for lations but by changing the institutions that resource-poor farmers. If the strict certification manage those regulations. Second, a focus on standards necessary to promote a country's par- institutional strengthening implies country-spe- ticipation in seed export markets were applied to cific strategies; aid and advice must be tailored all national seed production they would serve as to particular circumstances. Third, there must a disincentive to meeting the seed requirements be a commitment to the long-term develop- of many less commercial farmers. In recognition ment of national institutions, rather than a of this fact, recent discussion of seed law modifi- belief in quick policy fixes. The rest of this sec- cation to facilitate seed trade in southern Africa tion discusses three broad areas of institutional recommended the adoption of separate stan- change related to improving the performance of dards for small-scale seed production programs seed regulatory frameworks. The first area is (Commonwealth Secretariat 1994, p 18). concerned with possible changes in the man- agement and orientation of public sector agri- Addressing deficiencies in seed regulatory cultural research. The second area addresses the frameworks importance of opening variety development There are several common deficiencies evi- and seed production to a wider range of players. dent in seed regulatory frameworks (Tripp The third area explores the possibilities of 1995). Foremost among these is that the seed building regulatory capacity. laws and plant breeding protocols of many developing countries are not consistent with the Improving the public variety delivery system resources or purposes of national institutions. There is an undeniable need to make public They have been modelled on European or plant breeding systems more responsive to the North American examples, without reference to needs of representative farmers. Efforts in adap- the social, economic and technological circum- tive on-farm research (Byerlee et al 1987) or stances of the country (Grobman 1992). A sec- participatory research (Ashby et al 1987) can EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 47 help reorient breeding priorities. Sperling et al that will not immediately benefit from these (1993) describe a bean research program in changes. Private seed production requires rela- Rwanda where farmers visited the research sta- tively large and stable markets. Several examples tion and helped make selections. Maurya et al of private sector success with small farmers (1988) describe how lines from a rice breeding involve the transfer of varieties already devel- program in India were tested on farmers' fields, oped for commercial farmers in similar environ- using farmer management and evaluation. Pain ments; this is the case of the widespread use of (1986) describes the success of location-specific US sorghum hybrids in Mexico, as well as the breeding strategies for rice in Sri Lanka. successful entry of private maize hybrids origi- Innovations in variety development should nally developed for South Africa to drier be accompanied by more flexibility in variety regions of Zimbabwe (Rusike 1995, p 124). release. Zambia has instituted a release policy The fragmented and uncertain markets that that allows varieties that have been tested and characterize much seed demand will have to be found popular with farmers to be multiplied addressed by other means. Sperling et al (1994) and distributed as non-certified seed while they discuss the distribution of seed of bean varieties undergo the testing scheme that is required for developed by the national research program official release (DANAGRO 1988, cited in through merchants in local markets. Grisley Cromwell et al 1992, p 60). (1993) describes the successful distribution of Perhaps most important, public agricultural an initial supply of public sector bean seed in research needs to better define its clients. Zambia, and points to this as a "non-market Current donor policy towards greater private solution" to the problem of variety dissemina- sector participation in seed markets is certainly tion. Several Latin American countries have had appropriate, but an unfortunate consequence is experience with the promotion of artisanal seed that public sector research is sometimes encour- production (Lepiz et al 1994; Janssen et al aged to focus on its (legitimate) collaboration 1992). The recent proliferation of smaller maize with the private sector, at the expense of seed producers (Lopez-Pereira and Filippello addressing its equity responsibilities towards 1995) is also encouraging in this regard. farming populations not likely to be served by In their discussion of developing forage seed private enterprise. The private seed sector can production capacity, Ferguson and Sauma be an important conduit for public varieties, (1993) stress the importance of including seed but the monetary (or at times merely production as one component of broader devel- policy-correct) attraction of earning royalties opment projects. Cromwell and Wiggins from privately marketed varieties can distract (1993) survey a large number of NGO projects public research from its traditional clients promoting local seed production capacity and (Lopez-Pereira and Filippello 1995, p 43). draw conclusions for making such activities There is a difference between making public more economically viable. sector research more streamlined and efficient, In short, farmers' problems of inadequate on the one hand, and encouraging it to be a access to appropriate crop varieties will be ame- pale imitation of private enterprise, on the liorated, but not resolved, by more private sector other. participation. Reform of seed regulatory frame- works should be designed to encourage the Expanding the range of seedproduction options search for innovative seed provision options that There is no doubt that recent changes in are not limited to conventional seed enterprises. policy towards private seed enterprises will result in the wider availability of crop varieties Buildng regulatory capacty for many farmers. But as was pointed out in the Pray and Ramaswami (1991) describe four introduction, there are many crops and farmers stages in the development of national seed 48 SEED REGULATORY FRAMEwoRKs AND THE AVAILABILITY OF CROP VARIETIES industries, beginning with farmer selection and many small firms are exploring the feasibility of supply, progressing through public sector providing seed to sectors of the farming popula- responsibility and then increasing private sector tion currently served only by informal seed sys- participation, and culminating in a mature seed tems. A market's regulatory capacity also industry. This type of evolution will progress at depends on opportunities for consumers to different rates in different countries, and should exchange information and seek alternatives. But be paralleled by a progression through various for many resource-poor farmers, isolation and regulatory formats. Douglas (1980, p 122) cau- poverty constrain their capacity to exert suffi- tions that the design of appropriate seed policy cient influence on the market. depends on a number of factors specific to the Thus some type of public oversight of seed stage of development of national seed sectors, regulations will probably be required in many and emphasizes that, countries. FAO (1993) has developed general guidelines for a "Quality Declared Seed" system seed legislation is not just to protect that offers the possibility of establishing a seed farmers but to promote fair competition regulatory system consistent with national among seed enterprises and other seed resources and markets. It includes the establish- sellers as well. Excessive legislation stifles ment of a list of eligible varieties for seed pro- the industry. duction, with potentially flexible entry require- ments, and a monitoring system for seed quali- As private initiative assumes more respon- ty that shares responsibility with seed producers sibility for seed production and distribution, and merchants. and at times for variety development as well, it Recent efforts by Sasakawa Global 2000 in is reasonable to consider how the market can Ghana to stimulate the development of a pri- also assume more responsibility for regulatory vate seed industry after the collapse of the functions. But market capacity for regulating parastatal Ghana Seed Company place equal the type of product, and its quality, depends emphasis on strengthening the capacity of the very much on location-specific factors. The US government Seed Inspection Unit and develop- seed regulatory system, which emphasizes mar- ing small scale seed growers (Bockari-Kugbei ket control, is often held out as a model, but its 1994). Private sector involvement in regulation success is related to the specific conditions should be encouraged, however. Several coun- under which it operates. Commenting on vol- tries in southern Africa have already begun to untary certification, for instance, Simmonds recognize private certification agencies, for (1979, p 222) links its success to "a discriminat- instance (Commonwealth Secretariat 1994, p ing population of farmer-customers and effi- 13). Garay et al (1988) describe how several cient and competitive breeders." Kloppenburg institutions were invited to participate in seed (1988, p 107) shows that the US experience certification services in Bolivia. was not without its problems, describing the National seed acts often provide general pol- proliferation of companies in the late 1940s icy and an institutional framework for national selling hybrid maize seed, often with multiple seed systems, but vest authority for specific reg- names for one variety. ulations in seed committees or seed boards. The capacity of seed markets for self-regula- Bombin-Bombin (1980) emphasizes that one tion is related to the ability of seed enterprises advantage of such a system is that subsidiary (private firms or producer coops) to establish legislation can be managed by the most appro- reputations (Klein and Leffler 1981), and this priate authority and can be amended or in turn is related to the development of relative- changed more flexibly than the basic seed act ly stable, secure demand. But in many cases, itself. These boards or committees deserve spe- seed demand is still far from well defined, and cial attention when considering regulatory Table 2.3 Seed regulatory deficiencies affecting the availability of crop varieties Deficiency Possible Action 1. Inappropriate priorities for * More emphasis on location-specific plant breeding public sector plant breeding * More use of data on agroecological zones for targeting breeding * More adaptive research, farmer-participatory research * Clearer policy regarding public sector plant breeding mandate * Clearer division of labor, and collaboration, with private sector and community groups 2. Delays in varety testing * Allow more farmer access to public vareties being tested; more pre-release seed multiplication * Confine extended testing or observation to priority plant protection issues * Utilize data from other countries when assessing imported varieties 3. Lack of consistency in variety * Standardize and make more transparent the procedures of variety release procedures release committees * Establish adaquate frequency of committee meetings * Address possible bias in release of non-public varieties 4. Inappropriate standards for * Emphasize performance in local farming systems; allow wider variety release or registration range of criteria * Utilize varietal distinctions [distinct, uniform, and _table (DUS)] consistent with national needs and capacities * Establish a plant breeders' rights system consistent with farmer seed saving and exchange * Utilize only very broad interpretation of value in cultivation and use (VCU) 5. Regulations biased against * Encourage seed imports, consistent with legitimate plant non-public varieties quarantine considerations * Encourage private seed sector participation * Encourage local-level seed production initiatives * Do not require public sector variety release procedures for non- public varieties * In situations where the commercial seed sector is unable to address demand, consider public sector contracting or tendering of private firms or community groups 6. Seed certification requirements * Reconsider certification standards; make them more consistent restrict seed supply with national needs and resources * Make seed certification voluntary * Encourage the licensing of independent certification authorities * Allow the sale of truthfully labelled seed 7. Seed regulation for export * Distinguish seed regulations needed for export and those markets incompatible with local appropriate for small-scale or local seed enterprises seed requirements 8. Insufficient attention to national * Encourage and support national seed boards or seed committees; seed sector development limit their ex-officio membership; emphasize their policy responsibilities 50 SEED REGULATORY FRAMEWORKS AND THE AvAiLABiLrry OF CROP VARIETIES reform (Commonwealth Secretariat 1994; Bernstein, M. 1955. Reglating Business bv Independent Bockari-Kugbei 1994, p 270). Garay et al Commission. Princeton: Princeton University Press. (1988) describe the formation of regional seed Bockari-Kugbei, S. 1994. 'The Role of Small-Scale boards in Bolivia, and as these became effective, Enterprises in African Seed Industries," unpublished the establishment of a national coordinating PhD thesis, Department of Agricultural Economics board was a logical outcome. National seed and Management. Reading: University of Reading. boards or committees must include adequate Bombin-Bombin, L M. 1980. "Seed Legislation," legisla- representation from the private seed sector tive study no 16. Rome: FAO. (Chopra 1986, p 83); place more emphasis on Bould, A. 1992. "Models for Plant Variety Registration committed and experienced membership, and Grant of Plant Breeders Rights in the European rather than ex-officio representation (Poey Community and the Developed World," in ICAR, 1991); and operate with a clearly defined man- Proceedings of the Group Discussion on Management of Change in All India Coordinated Crop Improvement date. Projects. New Delhi: Indian Council of Agricultural Research. Conclusions Busch, L and W Lacy. 1983. Science, Agriculture and the Regulatory reform must begin not on paper Politics of Research. Boulder, Colorado: Westview but with the institutions that are active partici- Press. pants in the seed sector. National seed boards or Byerlee, D, M R Akhtar, and P R Hobbs. 1987. committees need to assume more responsibility "Reconciling Conflicts in Sequential Cropping for policies that stimulate a dynamic national Patterns Through Plant Breeding: The Example of seed sector. They should assure that national Cotton and Wheat in Pakistan's Punjab." Agricultural , , . . , ~~~~~~~~~Systems 24: 291-304. variety release and registration procedures are flexible and transparent, and consistent with Byerlee, D. and M. Morris. 1993. "Have We the needs of the nation's farmers. Less emphasis Underinvested in Research for Marginal should be placed on the specification of seed Environments? The Example of Wheat Breeding in Developing Countries," FoodPolicy 18: 381-393. regulations, and more attention devoted to establishing an institutional environment in Chambers, R. 1991. "Farmer First: A Practical Paradigm which seed enterprises, farmers and consumers for the Third Agriculture," in Altieri, M and S Hecht, eds, Agroecology and Small Farm Development. Boston: are able to negotiate their own regulatory CRC Press. frameworks in the context of an evolving national seed system. Chopra, K R. 1986. "Problems in Seed Production, Certification and Quality Control" in Proceedings of the First National Seed Seminar, Seed Association of India, 27-28 December 1986, New Delhi. References Commonwealth Secretariat. 1994. Harmonization of Seed Agrawal, P K and N S Tunwar. 1990. "Seed Certification Laws in the SADC Region. London: Commonwealth and Quality Control in India." Plant Varieties and Secretariat. Seedes3(3): 165-171. Ashby, J A, C A Quiros and Y M Rivera. 1987. 'Farmer Cromwell, E and S Wiggins, with S Wentzel. 1993. Participation in On-Farm Varietal Trials," Sowing Beyond The State. London: Overseas Agricultural Administration (Research and Extension) Development Institute. Network Discussion Paper No 22. London: Overseas Development Institute. Cromwell, E, E Friis-Hansen, and M Turner. 1992. The Seed Sector in Developing Countries: A Framework for Berg, T, A Bjoernstad, C Fowler, and T Kroeppa .1991. Performance Analysis. London: Overseas Development "Technology Options and the Gene Struggle," NOR- Institute. AGRIC Occasional Papers Series C: Development and Environment No 8. As: Agricultural University of Crucible Group. 1994. Peopk. Plants and Patents: The Norway. Impact of Intellectual Property on Biodiversitv, Conservation, Trade and Rural Sociey. Ottawa: IDRC. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 51 Douglas, J. 1980. Successful Seed Programs. A Planning and Jaffee, S and J Srivastava, J. 1994. "The Roles of the Management Guide. Boulder, Colorado: Westview Private and Public Sectors in Enhancing the Press. Performance of Seed Systems," The World Bank Research Observer 9: 97-117. Due, J. 1990. "Tanzania Seed Release, Multiplication and Distribution Systems," Seed Science and Technology Janssen, W, C A Luna, and M C Duque. 1992. 18: 187-194. "Small-farmer Behaviour Towards Bean Seed: Evidence From Colombia," Journal of Applied Seed Ferguson, J. 1994. 'El Proceso de Liberacion de Nuevos Production 10: 43-51. Cultivares de Forrajeras: Experiencias y Perspectivas," in J E Ferguson, ed, SemiUa de Especies Forrajeras Joshi, K. 1995. 'Participatory and Other Alternative Tropicales, Conceptos, Casos y Enfoque de la Approaches to Seed Production and Distribution in Investigacion y la Produccion. Memorias de la Octava Nepal," draft paper. Reunion del Comite Asesor de la REPT, Noviembre de 1992, pp 167-187. Kelly, A F. 1989. Seed Planning and Policyfer Agricultural Production. London: Belhaven Press. Ferguson, J and G Sauma. 1993. 'Towards More Forage Seeds for Small Farmers in Latin America," in Kelly,A F andJ D C Bowring. 1990. 'The Development Proceedings of the XVII International Grassland of Seed Certification in England and Wales," Plant Congress, 13-16 February 1993, Palmerston North, Varieties and Seeds 3(3): 139-150. New Zealand. Klein, B and K B Leffler. 1981. "The Role of Market FAO. 1993. "Quality Declared Seed," FAO Plant Forces in Assuring Contractual Performance," Journdl Production and Protection Paper 117. Rome: FAO. ofPoliticalEconomy 89(4): 615-641. Friis-Hanssen, E. 1992. Seeds for African Peasants: A Case Kloppenburg, Jr, J. 1988. First The Seed. Cambridge: Study from Zimbabwe. Copenhagen: Centre for Cambridge University Press. Development Research. Knudson, M. 1990. 'The Role of the Public Sector in Garay, A, P Pattie, J Landivar, and J Rosales. 1988. Applied Breeding R & D," Food Policy 15: 209-217. "Setting a Seed Industry in Motion: A Non-Conventional Successful Approach in a Lepiz, R, J Ashby, and J Roa. 1994. "Artisanal Bean Seed Developing Country," Working Document No 57. Production Experiences in Latin America," paper pre- Cali, Colombia: CIAT. sented at Workshop on Integrated Seed Systems, October 1994, Kampala, Uganda. Gisselquist, D. 1994. "Import Barriers for Agricultural Inputs," draft paper, International Trade Division. Lipton, M and R Longhurst. 1989. New Seeds and Poor Washington DC: World Bank. People. London: Unwin Hyman. Green, T. 1987. Farmer-to-Farmer Seed Exchange in the Lopez-Pereira, M and M Filippello. 1995. "Emerging Eastern Hills of Nepal. The Case of 'Pokhreli Masino' Roles of the Public and Private Sectors of Maize Rice. Dhankuta, Nepal: Pakhribas Agricultural Centre. Industries in the Developing World," CIMMYT Economics Working Paper 95-01. Mexico DF: CIM- Grisley, W. 1993.. "Seed for Bean Production in MYT. Sub-Saharan Africa: Issues, Problems, and Possible Solutions," Agricultural Systems 43: 19-33. Maurya, D M. 1989. "The Innovative Approach of Indian Farmers," pp 9-13 in Chambers, R, A Pacey, and L A Grobman, A. 1992. "Fostering a Fledgling Seed Thrupp, eds, Farmer First: Farmer Innovation and Industry," in J R Anderson and C de Haan, eds, Agricultural Research. London: Intermediate Public and Private Roles in Agricultural Development, Technology Publications. Proceedings of the Twelfth Agricultural Sector Symposium. Washington, DC: World Bank. Maurya, D M, A Bottrall, and J Farrington. 1988. "Improved Livelihoods, Genetic Diversity and Farmer Hampton, J G and D J Scott. 1990. "New Zealand Seed Participation: A Strategy for Rice Breeding in Rainfed Certification," Plant Varieties and Seeds 3(3): Areas of India." ExperimentalAgriculture 24: 311-320. 173-180. Mitnick, B M. 1980. The Political Economv of Regulation: Heisey, P, ed. 1990. Accekrating the Transfer of Wheat Creating, Designing, and Removing Regulatory Forms. Breeding Gains to Farmers: A Study of the Dynamic of New York: Columbia University Press. Varietal Replacement in Pakistan. Mexico DF: CIM- MYT. 52 SEED REGULATORY FRAMEWORKS AND THE AVAILABIUTY OF CROP VARIEnES Pain, A. 1986. 'Agricultural Research in Sri Lanka: An Simmonds, N. 1979. Principles of Crop Improvement. Historical Account." Modern Asian Studies 20(4): Harlow, UK Longman. 755-778. Sperling, L M E Loevinsohn, and B Ntambovura. 1993. Poey, F. 1991. 'Philippines: The Development of the 'Rethinking the Farmer's Role in Plant Breeding: Rice and Corn Seed Industry" Agricultural Policy Local Bean Experts and On-Station Selection in Analysis Project, Phase 11, Report No 312. Boston: Rwanda," ExperimentalAgriculture 29: 509-519. Abt Associates. Sperling, L, U Scheidegger, and R Buruchara. 1994. Pray, C. 1990. "The Potential Impact of Liberalizing 'Designing Seed Systems with Small Farmers: India's Seed Laws.' Food Policy 15: 193-198. Principles Derived from Bean Research in the Great Lakes Region of Africa," paper presented at Workshop Pray, C, and B Ramaswami. 1991. A Framework for Seed on Integrated Seed Systems, October 1994, Kampala, Policy Analysis in Developing Countries. Washington, Uzanda. DC: International Food Policy Research Institute. Stigler, G. 1971. 'The Theory of Economic Regulation," Rajbhandary, K 1994. "The Regulatory Frarneworks for Bell Journal of Economics and Management Science 2: Seed Production, Purity and Sale in Nepal," draft 3-21. paper. Tripp, R. 1995. "Seed Regulatory Frameworks and Ruigu, G M. 1988. "The Kenya Seed Industry Evolution: Resource-Poor Farmers: A Literature Review," Current Status and Future Prospects," paper prepared Agricultural Administmtion (Research and Extension) for the seminar on "Technology Development and Network Paper No 51. London: Overseas Changing Seed Supply Systems," 22 June 1988, Development Institute. Tilburg, The Netherlands. Turner, M R. 1994. "Trends in India's Seed Sector." Rusike, J. 1995. "An Institutional Analysis of the Maize paper presented at Asian Seed 94, 27-29 September Seed Industry in Southern Africa," unpublished PhD 1994, Chiang Mai, Thailand. thesis, Department of Agricultural Economics. East Lansing: Michigan State University. Ulrich, A, W H Furtan, and A Schmitz. 1987. "The Cost of a Licensing System Regulation: An Example from Salazar, R. 1992. "Community Plant Genetic Resources Canadian Prairie Agriculture," Journal of Political Management: Experiences in Southeast Asia," in D Economy 95(1): 160-178. Cooper, R Vellve, and H Hobbelink, eds (1992). Growing Diversity: Genetic Resources and Local Food Security. London: Intermediate Technology Publications. Streamliningfor a Demand-Driven Seed Sector: A Framework for Policy Refiorm Jitendra P Srivastava Improved seed - agriculture's oldest tech- up barriers to the entry or acquisition of foreign nology and most important input - has the seed products. This dosed-end stance, associat- power to ensure food supplies, create rural sur- ed with public-sector bias, further exacerbates pluses, and ultimately transform economies. defects and problems in the research-and-devel- The production gains from high-yielding wheat opnlrent base. The results include regulatory and rice since the mid-1960s. in developing barriers that: impede the entry into the seed countries and the increased efficiency of agri- market of foreign seed researchers, traders, and culture (from a diminishing base in terms of private operators; and prevent, limit, or delay number of farmers in industrialized economies) access to superior seed technology by a coun- provide convincing evidence. For the past thir- try's farmers (see Box 2.1). ty-five to forty years, the public sector has played a dominant role in the seed drama, with Changing the status quo international research centers transferring In many developing countries, there is a much-needed, higher-yielding seed for signifi- pressing need for a framework of policies, relat- cant food crops to developing countries ed institutions and infrastructure that establish- through national networks. This recipe yielded es a favorable climate for the entry and utiliza- highly visible results. Provided local testing for tion of advances in seed technology to increase crop adaptability was conducted, and seed pro- the yield and efficiency of crop production. duction plants were in place, it was thought Developing this framework will entail paring that the process would continue indefinitely. down and refining public sector participation, But nature has built-in limitations. Some while strengthening and expanding the roles of crops are not widely grown, nor widely adapt- all private sector participants from subsistence able. While international research has begun farmers and small farmer seed producers to tackling problems of lesser crops, in more domestic seed enterprises and multinational remote and specific environments, gains in new seed companies (see Box 2.3). varieties are subsiding. Also, public sector entrenchment can feature inflexibility, ineffi- Trends, needs, and challenges ciency, and in some cases, irrelevancy with top- Current approaches to seed research, policies down decisions focusing research on seeds that affecting seed flow, and supporting institutions farmers~ ~ ~ ~ ~ ~ ~afecn doe notw want Aupotm thesamttme,nasen farmers do not want. At the samne time, nascent and infrastructure are evolving as global trends national research systems and seed producers impinge on the sector (see Box 2.2). The appar- often become protective of their turf, putting ent new direction points toward demand-driven seed production, more private initiative, more competition - usually from the private sector, Jitendra P Srivastava is PrincipalAgriculturalist, but in command-driven economies that persist, Agriculture and Natural Resources Department, among parastatals - and market pricing in lieu The World Bank, Washington, D. C 54 STREAMLINING FOR A DEMAND-DRIVEN SEED SEcrOR: A FRAMEWORK FOR POLICY REFORM Box 2.1 Most common regulatory barriers to seed movement Impediments to entry of foreign seed traders and researchers and prvate operators in seed markets: * Unreasonable phytosanitary or quarantine restnctions and procedures * Compulsory variety registration * Seed certification and control * No private access to public-sector seed stocks * Inadequate protection of intellectual property rights in seed * Over-centralized administration of regulations Deficiencies stifling farmers' access to superior seed technology: * Inapproprate prorities for public sector plant breeding * Delays in variety testing * Lack of consistency in variety release process * Inapproprate standards for variety release or registration * Regulations biased against non-public varieties and structured for export markets * Seed certification requirements 3 Insufficient attention to national seed sector development Sources. The Agricultural Technology and Services Division (AGRTN) (September 1995): Reform of Seed Regulations - Easing Barriers to Better Seed Varieties, World Bank Agriculture Technology Notes, Number 12, Washington, D.C. Tripp, Robert (1995): Seed Regulatory Frameworks and the Availability of Crop Varieties, Overseas Development Institute, London. of subsidies. Needs and challenges emerge in ous developing countries that churn out too response to these global trends. much seed of varieties farmers don't value Promoting a robust seed sector will entail speak to this critical point. addressing the specific needs of all private sector players (see Box 2.3). For example, subsistence * Access. If desirable seed is available on the farmers primarily require information and research shelf, can farmers get hold of it, in strong linkages to small farmer seed producers, sufficient supply and at an affordable price? while multinational companies require different Questions of access mainly imply control by institutional and policy responses to accommo- government through policy and regulatory date their need to relocate profits and establish mechanisms, which can be conducive or joint ventures with domestic seed enterprises. stifling to seed flow. In specific countries, At the same time, there are challenges common otherwise useful and beneficial technologies to all interested in improving seed research and can be withheld or bypass farmers, leaving production. These include: them less productive and less competitive. * Quality. Is seed technology any good? Does Information. The free flow of information is it do what farmers need and want? Without critical to an improved seed sector. What demonstrable value-added, as defined by kind of information is available and how it farmers themselves, new seed is not neces- is communicated impact the level and quali- sarily better and farmers won't use it. White- ty of participation by all seed producers. elephant seed production plants in numer- Information on improved harvesting prac- EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 55 Box 2.2 Global trends in the seed industry * Globalization of agricultural research * Investments in biotechnology R&D * Shift in crop improvement and R&D from public to prvate sector * Liberalization of trade * Emergence of successful private seed companies * Entry of multinational seed trading companies * Greater attention to informal sector * Debate regarding impact of legislation and trade agreements on intellectual property rights and bio- diversity tices, variety performance trials, external about two-thirds of varieties from the markets, seedborne pathogens, as examples, International Rice Research Institute have a must reach a diverse audience through borrowed parent, that is, an out-of-country diverse media. genetic claim. Unless private firms also have access to such foundation seed, research efforts Enabling services. The best seed available and are stymied. the best-intentioned farmer are not enough. At the same time, public sector institutions The farmer may need credit to buy such should be divested from several activities - seed, advice from extension services on best production, distribution, and marketing -for practices for its use, and adaquate infrastruc- most types of seed over time. The critical path ture, such as roads, communication facili- for public action regarding seeds entails trade ties, and markets, to move his/her products. liberalization of seed for research purposes and Such functions are generally, at least sale, a useful and workable legal environment initially, provided by the public sector, and a for seed flow, and appropriate and transparent critical mass of institutions that provide "rules of the game" for private sector involve- them need to be functioning in order to get ment. The public sector may refocus its seed and agriculture moving. resources and personnel on research on less commercialized crops or adaptation of off-the- Critical issues shelf germplasm for domestic conditions, quali- Having identified common challenges, pub- ty and hazard control, training/promotion and lic, private, national and international interests credit. Nevertheless, in a given economy, the alike must contend with several critical issues private sector may share in these functions. that will define a country's path toward seed If the motive of seed regulations-is to protect sector reform. national agriculture from biological hazards and How can good varieties that outperform seed consumers from unreliable seed, what are ade- in farmers'fields be more efficiently developed? quate and effective measures to ensure public con- Open sourcing of seed is the ideal solution.This fidence in seed? Numerous regulations - many will dissolve the tension between foreign tech- well-intentioned, but several capricious and nology and internal innovativeness. covert - are used in the name of consumer International and domestic researchers, whether protection to keep competitors out of national public or private need each other's talent and seed markets. For example, phytosanitary stan- germplasm. Their research tracks need to con- dards to keep legitimate pests out are employed verge more often than they currently do. Seeds by virtually every country. Yet experts concur, have increasingly complex parentage. For rice, many false issues are pursued, while real and 56 STREAMLINING FOR A DEMAND-DRIVEN SEED SEcrOR: A FRAMEWORK FOR POLICY REFORM Box 2.3 Needs of seed producers Category of Seed Producer Assistance Needs Subsistence farmers who save Information on: own seed * most suitable variety * improved techniques for on-farm seed selection * improved harvest, cleaning and drying methods using local materials * access to small farmer/seed producers seed and trained extension staff Small farmer/seed producer * access to most suitable varieties for locality * training in seed production of specific crops * access to credit * access to seed testing laboratory * assistance in conducting 'on-farm' variety performance trials * close linkage to domestic seed enterprise and/or public research institution Domestic seed enterprise * access to improved varieties/hybrids * easy flow of improved germplasm * hard currency to purchase inputs * credit in local currency * market pricing and demand for improved seeds * access to information on extemal markets * government policies to promote seed export * plant variety protection * government policy incentives to undertake R&D Multinational seed enterprise * minimal restriction on intemational movement of germplasm/seed * government policies and legal framework for joint ventures with domestic companies * ability to relocate plants * plant variety protection * policies encouraging participation of national seed companies Source: Srivastava, Jitendra P. and Steven Jaffee. 1993. Best Practicesfor Moving Seed Technology: New Approaches to Doing Business, World Bank Technical Paper 213, Washington, D.C. serious problems are ignored, in the name of gauging the economic dimensions of such phytosanitary control. Possible solutions threats as priorities in seed pathology research include: upgrading knowledge of biological are revised. threats; using a global database on seedborne Another possibility is self-regulation pathogens and seed treatment regulations; stan- wherein the seed industry devises appropriate dardization of seed health testing methods; and and enforceable local standards that are moni- EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 57 tored and upgraded by seed producers' associa- improved seed? Only about one-third of seed tions, among others. Cross-linkages to national used worldwide is, sold commercially, and of and international associations can further that a paltry 10 percent is sold in developing serve to reinforce standards, professionalize countries. Another third is produced by public conduct, and upgrade research and products sector institutions, and the remaining third is throughout the industry, as it is commercial- saved by farmers for their own use next season. ized. Seed systems do not have to be highly The great bulk (estimated at 80 percent) of developed to benefit from such activity; self- crops cultivated in developing countries come regulation is at the heart of grass-roots from seed that escapes the monetized economy endeavors, in addition to being the hallmark of - the informal sector. With liberalization and the most reputable global seed producers. privatization at the global level, the next big How can the value-added that research yields challenge is to move the great majority of devel- be legally protected and economically captured by oping country farmers, not just those in those who invest in it? Factors surrounding how preferred areas, crops, and irrigated conditions, seeds are replicated (whether easily in farmers' into the commercial mainstream. fields or only in the laboratory) and whether Small-farmer seed producers are emerging as investment in research on crops can be captured transitional go-betweens, connecting previously for economic profit by the researcher are subsistence farming to diversified, commercial receiving increasing attention. Public agriculture (see Box 2.3). Possible mechanisms researchers generally don't want to give away for tapping into this vast reservoir of underuti- seed technology that private firms can make a lization covers a broad range, including creation profit on, nor would private firms want to of cottage-seed subsystems and formation of invest huge sums in seed technology, only to national seed farmers' associations. As the have it pilfered by other agents, thus losing cornerstone of strong, private seed systems they their investment. could contribute: their input for establishing A handy solution is offered by plants them- research priorities; their fields as a source of selves. Seed of self-pollinated crops such as new varieties and hybrids; and their energy as wheat and rice and of open-pollinated crops seed producers, distributors, importers, such as maize and sorghum can be readily sown exporters, entrepreneurs and investors. In short, for seed by farmers, while seeds of hybrids have they are the missing link that will put private to be bought new each planting season because and public research on a common track, seeds lose vigor in the next generation. redesign policy regulation, seek common stan- Accordingly, public sector research focuses on dards and approaches to phytosanitary control, self-pollinated food crops and private firms on and build reputations for reliability and innova- hybrids where research profits are capturable. tion in the marketplace. These divisions are not hard and fast. In large market economies, farmers often buy self-polli- Final comment nated seed from commercial seed producers for The needs, challenges and issues inherent to assured quality and convenience. In large devel- the process of seed sector reform will be suc- oping economies, without shared business cessfully addressed through collaboration and ethics and enforceable legal controls, the participation. Seed producers, from subsistence prospect of having seed technology pirated may farmers to multinational companies, along with be too daunting for private firms to introduce governments and donor agencies have critical self-pollinated seed for sale. roles to play in framing policies that promote a How will global trends affect small-holder robust demand-driven seed sector. farmers in less developed countries who most need Supply of New Varieties Depending on factors such as crop and country size, foreign public and private research produces a flow of new varieties that farmers appreciate - or would, if govern- ments allowed them to be introduced. When foreign-bred lines are not immediately suitable, crosses with local lines can add resistances and other features to create new vari- eties that farmers appreciate - or would, if governments allowed them to be introduced. Four papers in this Section describe private supply of new varieties for maize and cotton and public supply of new varieties for wheat and rice. Both public sector IARCs and private companies locate international breeding programs in a limited number of countries, then distribute new lines for testing in multiple countries. These breeding strategies rely on international movement of varieties to optimize returns to research investment. Other things equal, the faster a new variety spreads throughout its potential (multi-country) market, the greater the return. With seed deregulation as recommended in the previous Section, private companies would be able to locate breeding to serve markets defined by ecological conditions and other technical criteria rather than political boundaries. Also, public varieties from IARCs and some of the stronger national systems could be expected to disseminate faster and further, boosting returns to public research. Prospects and Constraintsfor Hybrid Maize in Developing Countries Roberto W Ansaldo and Ray Riley Pioneer Hi-Bred has been developing, maize markets in Asia. The final part describes promoting, marketing and distributing hybrid constraints and solutions for spread of hybrid seeds since the 1920s in the United States, and maize in developing countries. on a worldwide basis since the late 1960s. Over the years, we have seen many countries go Product development, testing, and through the process of improved seed tech- deployment for developing country markets nology development and adoption. These From a technical and organizational per- comments are based on our experience and spective, delivering improved varieties is a commitment to hybrid seeds over the past multi-faceted process involving: product devel- seventy years. opment, testing, seed production, and market- Hybrid maize seed can be seen as one com- ing. In addition to the purely technical and ponent in a larger system: in many countries organizational challenges to deliver products to maize is an important feed component (energy markets, government policies often impact source) for expanding livestock and poultry Pioneer's ability to serve markets. Policies that industries. We seek to commercialize hybrid impede our ability to effectively serve markets maize seed and other technology so as to move include: unjustified phytosanitary obstacles, a majority of farmer-producers into mainstream unnecessary variety registration requirements, commercial agriculture. In the Philippines, lack of intellectual property protection, and farmers have recently adopted silage technology, lack of regional cooperation, particularly for expanding the market for hybrid maize seed small size of national markets. and lowering livestock production costs. Introduction of silage technology - along with Product development hybrid maize - could have similar impacts in Pioneer has an international network of other developing countries. forty-five maize research locations, more than The first part of this paper describes 40 percent of which are in developing countries Pioneer's research and commercial strategies to (three in Africa, five in Asian developing coun- develop and introduce maize hybrids, with tries, and twelve in Latin America), with an particular attention to developing country annual research budget that has averaged about markets. Parts two and three discuss hybrid $100 million from the late 1980s. In research, the company collaborates with universities and other institutions to augment our in-house Roberto Ansaldo is Director of Government efforts. Affairsfor Developing Countriesfor Pioneer Hi- Product development entails meeting Bred International Ray Riley is Corn Director, market needs, which requires products that North Central Corn Belt, for Pioneer Hi-Bred have "added value" according to farmers' InternationaL 62 PROSPECTS AND CONSTRAINTS FOR HYBRID MAIZE IN DEVELOPING COUNTRIES criteria. When considering the feasibility of As this account shows, Pioneer typically entering and effectively serving new markets, carries out hundreds of on station trials and the company looks at market factors such as: another 1,000 on-farm trials before bringing a new product into full commercial production. * maturity and ecological conditions, Initially, the purpose of these trials is to deter- mine performance of the variety in representa- * product attributes valued in the specific tienvrmnswthhexptdragof markt (or eampe, isec anddisase tive environments with the expected range of, mrkesistance, (ra ealei and dulismasge- stress and disease. A second objective of these resistance, grai color and quality, manage- trials, which is to measure and to develop the market, comes to the fore as testing moves onto * expected yield level, with attention to farms in years R4 and R5. farmer access to fertilizer and other support- The volume and quality of information ing inputs. coming out of Pioneer's own product tests far exceeds that from official government trials. In Based on assessment of what farmers value countries with mandatory variety registration- in existing or new markets, the company is able where governments do not allow sale of seeds to focus information and genetic resources from until they have tested and approved the variety its global research network to develop products -superior varieties can fail due to poor infor- for specific markets. Breeders in each of mation from government tests. For example, if a Pioneer's forty-five research locations develop government bases its decisions on ten replica- new products from Pioneer lines, public materi- tions only, there is a 5 percent chance that the als, and licensed materials (that is, lines for observed yield from those ten plots will be at which Pioneer pays royalties to other organiza- least 7 percent higher or lower than the expected tions). This process is facilitated by close inter- or 'true" yield of the variety. (figure 3.2 shows action and sharing of information and the number of locations required to have 95 germplasm among Pioneer researchers. percent confidence that observed yield is within a given percent of expected yield.) Hence, with Product testing compulsory variety registration, not only do governments second-guess company decisions Once a new product has been developed (in about whether or not a variety wil sell, but they any one of forty-five research stations), it goes frequently do so with inadequate information, in the second year (R2 in figure 3.1) to wide that is far less than companies have on hand area research testing (for example, on five when they make decisions about whether or not research stations x ten plots each). From there, the winners go in the next year, R3, to extended to promena n vety andutom h wide-area testing (for example, eight to ten sta- r tions x ten plots each). In the next year, R4, promising hybrids go into several hundred farm Pd tion trials in "side-by-side" demonstrations with When entering a new market, Pioneer not other varieties in farmers' field, typically in only faces the challenge to develop and test new four-row sub-plots planted to each variety; in products, but also evaluates the sustainability of this first year of on-farm demonstrations, the business. Can seeds be sold at a profit that Pioneer provides free seed. In the next year, R5, will allow for a sustained business operation Is tests expand to 1,000 or more farm plots; farm- there a market for maize that allows the farmer ers buy seed for many of these demonstrations, to generate enough cash to pay for his invest- while Pioneer contributes technical advice. Also ments and to recognize his rewards?: If condi- in R5, Pioneer approves promising products for tions are favorable, Pioneer works out business limited commercial release. arrangements for reliable seed supply. Figure 3.1 Data driven decisions 6000 Example Product Testing and Volume Cycle 5000 * / " 4000 --%of Peak| 3000 -Testing 2000 . 1000 0 R2 R3 R4 R511RC1 RC2 RC3 RC4 RC5 RC6 RC7 RC8I Testing Commercialization P'ion=cr i I. l. Product Life Cycle Stage Figure 3.2 Multi-location testing to evaluate performance 50…-- B U 40… A 30 C 20 - - - - - - R E10… 1 2 5 10 15 20 25 30 40 50 NUMBER OF LOCATIONS ** *Performance difference required for (20:1) confidence of a difference **Number of locations comparing performance Pion= o-WI hE, i. (LSD = .05 @ 150 BU/AC) 64 PROSPECTS AND CONSTRAINTS FOR HYBRID MAIZE IN DEVELOPING COUNTRIES Sales and marketing l loss of physical control of parents (that is, Pioneer expands on-farm tests or demon- government demands that companies deliv- strations into the thousands during the first sev- er samples of hybrid parent lines to certifica- eral years of commercialization (years RC1 and tion or registration agencies, etc); RC2 in Figure 3.1). For these tests or demon- strations, frmers buy seeds, while Pioneer pro- n* ocal sourcng reqirements for seeds or par- vides technical advice. These tests provide farm- ent seeds, whch obstructs company ability . s * , * * * r ~~~~to introduce new hybrids or tO move excess ers with information about variety perfor- seed from country-to-country. This boosts mance, building market demand. In markets with competitive private seed seed costs to farmers; and industries, the expected market life of a maize inadequate protection for intellectual prop- hybrid is about eight years from first commer- erty rights, which discourages companies cial sale. The main cause of this turnover is from bringing in their best materials. genetic advance in breeding. Genetic advances support annual yield gains of 0.5 to 2 percent, The size of the market is a major considera- depending on the market. Gains are toward the tion for Pioneer and other seed companies. high end of that range in markets such as Pioneer weighs costs to enter and serve a mar- Thailand, where research is in early stages of ket - for example, research and development, development and where favorable markets exist overcoming policy obstacles - against market for the increased grain that is produced. In potential. Markets can and sometimes do more established markets, such as the United extend across political boundaries; this is partic- States, gains are lower. With steady genetic ularly true in Europe, where all countries accept advance, even a very good hybrid can be over- varieties approved by any one government. taken and pushed off the market in ten years. Larger markets encourage more research and allow more efficient seed production and sup- Policy-based barriers to serving new markets ply; Pioneer is able to deliver better products at While previous paragraphs describe the lower cost. But when small countries allow technical and organizational challenges involved political boundaries to divide markets by not in developing new hybrids and getting seeds to allowing varieties or seeds from other countries farmers, the world is not so simple. to enter, each of the resulting markets may be Government policies often get in the way, too small to repay Pioneer for research and reg- blocking Pioneer's ability to serve markets. In ulatory expenses required to deliver seeds to many developing countries, policy-based obsta- farmers. This is a problem in much of Africa, des have been so severe that Pioneer - and where political boundaries divide large potential many other private seed companies - simply markets into many small national markets, and leave them alone. In some countries, obstacles where regulatory barriers add unnecessary costs are less severe, so that Pioneer is able to operate, to entering these small national markets. but without full efficiency. Obstacles include: The map in figure 3.3 shows CIMMYT maize mega-environments for sub-Saharan * unrealistic phytosanitary regulations block- Africa. CIMMYT developed these mega-envi- ing seed import, such as quarantine require- ronments to guide research design; each mega- ments for diseases that are endemic to the environment describes an area within which seed importing country; varieties developed with particular characteris- tics can be distributed. From a marketing point * excessive registration requirements, whereby of view, mega-environments suggest size of governments block private company efforts market for new varieties. Clearly, mega-envi- to deliver seeds of new varieties to markets; ronments overlap political boundaries. Grain EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 65 Figure 3.3 African maize ecology map Afrca (19!/9 Ecology rg c y 2sLowland tropics to c oea e hSubtropical e e h Highland tropics i t a Maturity shoul o t BLate 7 o vI ntermediate n |g] Early J_ Source: CIMMYT (Cantro Internacional de Mejoramiento de Maiz y Trigo) maize mega-environments for subSaharan Africa. (1989190) color is another important attribute. Table 3.1 The 1980s saw the introduction of hybrid shows hectares planted to maize according to maize to many of the other Asian nations as grain color and ecology in Asia and Africa. average corn yields began to flatten, and private sector seed companies began to conduct opera- Maize hybrids in Asia: historical perspective tions in many countries. While the seed indus- Maize hybrids were first introduccd into trics and markets in many Asian countries are Asia on a large scale in China in the 1960s. still in the early stages of development, the Many of the hybrids that were successfiil were 1990s should offer exceptional opportunities temperate-zone varieties, and were not useful for increased hybrid use as maize consumption for sub-tropical and tropical areas of other in the area increases, and the need for higher Asian countries. By the early 1970s, the majori- yielding maize increases. Table 3.2 shows the ty of the breeding programs in the other parts datc of introduction for hybrid maize for the of Asia focused chiefly on the developmcnt Of major maize producing countries of Asia as well high-yielding open pollinated varieties (OPVs), as estimates of hybrid maize area for 1992. led by the International Maize and Wheat Improvement Center (CIMMYT). 66 PRosPEcrS AND CONSTRAINTS FOR HYBRID MAIZE IN DEVELOPING COUNTRIES Table 3.1 African and Asian maize area by grain color and ecology, ca 1990-91 (1,000 hectares) Grain color and ecology Africa/Middle East Asia (except China) China white sub-tropical 8,000 840 620 white highland tropical 20 - - white lowland tropical 6,300 2,900 white transitional 1,500 - 1,200 white temperate 2,100 - 2,900 yellow lowland tropical 680 8,800 yellow sub-tropical 580 1,100 2,000 yellow transitional - 860 1,900 yellow temperate 3,300 - 11,600 Table 3.2 Asia Hybrid Maize Area, 1992 hybrid area OPV area Country hybrids introduced (1,000 ha) (1,OO0 ha) Percent hybrid China early 1960s 15,300 5,700 73 India mid-1970s 1,200 4,800 20 Philippines 1981 230 3,100 7 Indonesia 1985 60 2,900 2 Thailand 1980 320 950 25 other Asia - 1,300 1,700 42 Asia total - 18,400 19,100 49 China India China is far and away the largest maize pro- Hybrid use in India is prevalent only in ducer in Asia. Almost three-quarters of its some of the southern and eastern parts of the maize area is currendy planted to hybrid vari- country. Hybrid use is confined mainly at this eties. Yellow dent hybrids from the United time to the provinces of Andhra, Pradesh, States were successfully introduced in the 1960s Karnataka, and Bihar. Constraints on increased in the northern, temperate maize growing areas. hybrid use in India have induded the small size These hybrids continue to be prevalent to this and low incomes of farms, and lack of hybrids day. China's seed industry is currently dominat- suitable for some maize growing areas. ed by the government. The basic mechanisms Investment by private seed companies has been of a functioning seed market are in place, but limited by a challenging business environment. inadequate seed processing and quality assur- ance procedures result in very poor quality Philippines seeds often being used by firmers. The genetic Hybrid maize in the Philippines grew base for maize is also narrow, due to the govern- steadily to about 15 percent of total area in ment monopoly on the seed industry and lack about 1987. Economic distress among farmers of intellectual property protection. in the late 1980s caused hybrid use to decrease somewhat since that time. Volatile maize prices, EASING BARRIERs TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 67 a lack of infrastructure in some areas, and prob- grown to meet increased demand, maize con- lems with access to credit for farmers have sumption has increased dramatically along with slowed hybrid adoption. Maize areas in Luzon other feedstufis. Feed use of maize has grown in and in some key areas of Mindanao use the importance relative to food use. In 1973, food highest proportion of hybrids. uses of maize accounted for 66 percent of total consumption and feed uses accounted for 34 Indonesia percent. In 1993, these numbers reversed, as 36 Hybrid maize use in Indonesia has increased percent is used for food purposes and 64 per- in some major maize growing areas such as cent for feed. West Java, Lampung, Sumatra, and Northern At current rates of growth, consumption of Sumatra. Maize growing areas on other outly- maize in Asia will increase from a 1993 level of ing islands face greater difficulties in maize 153 million metric tons to almost 200 million transportation and marketing, and have limited metric tons by 2003. This represents an hybrid seed maize use. Maize also fits into a rel- increase of 30 percent in the next decade. If atively complex multi-cropping system in many income growth is sustained at levels of the past crop areas, where maize is not the crop of pri- few years, consumption of meat and dairy mary importance. products could increase an even greater amount, putting further pressure on maize pro- Thai/and ducers in Asia to increase productivity. Thai maize farmers use an increasing pro- An important question is how countries in portion of hybrid maize. The central and north Asia will supply maize to meet increased central regions of the country currently use demand. Will maize be imported from interna- hybrids on almost percentofmaizearea. tional markets, or is there potential for hybrids on almost 50 percent of maize area. icesdpouto rmwti h ein Farmers in Thailand have historically purchased sncreased production from within the regiona seeds of improved OPVs on an annual basis. As consumption of maize has increased, Asias However, hybrid varieties from the private sec- production has not kept pace. Net imports have tor are now beginning to dominate the seed been rising gradually in the last decade. Imports maize market. Most maize is sold by farmers are likely continue to grow over the next into a relatively efficient commercial marketing decade. This will be particularly true if yield system, where maize moves into export chan- levels in many Asian countries remain at cur- nels or an ever growing poultry sector. rent low levels. Projections of Asian maize consumption and Hybrid technology to meet conumption needs implications for hybrid seed utilization Rising maize consumption figures show the presence of economic motivation to boost Consumption maize production among the major maize pro- Maize has historically ranked third in ducing countries in Asia. Increases in maize importance among grains behind rice and planted area along with steady yield gains wheat in Asia. However, its importance has fueled production growth in Asia from grown significantly in the last two decades as 1960-90. Maize area in Asia grew 52 percent diets have become increasingly westernized. from 1960-90, but further expansion is difficult Rapid income growth in most countries in Asia due to land constraints and population growth. has resulted in dietary upgrade favoring higher In many countries, maize is already being plant- meat and dairy consumption. ed on highly marginal land. Increased maize For example consumption of both pork and production, if it is to come, will have to come poultry meat have doubled since 1983. To the from higher yields, increasing at a faster rate extent that domestic livestock industries have than in the past. 68 PROSPECTS AND CONSTRAINTS FOR HYBRID MAIZE IN DEVELOPING COUNTRIES More widespread and extensive use of tries to remain self-sufficient. China and hybrid technology stands as a very effective Thailand would have to convert virtually all means of achieving growth in yields on current maize areas to hybrid use, and significant gains maize area in Asia. Pioneer estimates that would have to be made in India, the hybrids achieve yields 50 to 300 percent higher Philippines, and Indonesia for these countries than open pollinated varieties in Asia, utilizing to remain self-sufficient. little more than a different seed at planting time. Constraints and solutions for extending Yield levels for maize producing countries in hybrid maize in developing countries Asia are significantly lower than the world aver- What needs to happen to increase use of age (1986-91 average) of 3.6 tons/hectare hybrid maize seed in developing countries (t/ha). China is the only country with yields where it is not currently prevalent, but where over this level (4.0 t/ha), with Thailand (2.5 there is economic incentive for higher yields? tlha), Indonesia (1.79 t/ha), India (1.4 t/ha), What are the main constraints to increased and the Philippines (1.2 t/ha) falling below the hybrid utilization, and what changes are needed average. Table 3.3 shows what maize consump- in economic and policy environments to allow tion would reach in ten years at current rates of seed markets to develop and farmers to use growth for the largest maize producing coun- hybrid seeds? tries in Asia. Of these countries, China is cur- First, in many developing countries, suspi- rently a net exporter, and the rest are about cion about hybrid seeds continues both out of self-sufficient. ignorance and misinformation. These suspi- Table 3.3 also shows the amount of total cions are often expressed as follows: that hybrid maize area that has to be planted to hybrids for seeds are expensive; that hybrids are big con- countries to be self-sufficient in maize in 2003. sumers of agricultural inputs such-as chemicals Even under generous assumptions of yield and fertilizers; that hybrid seeds are tools of growth with total maize area at 1992 levels, developed countries to control the agricultural over eight million hectares of maize has to be economies of developing countries; etc. converted to hybrids for the illustrated coun- Table 3.3 Hybrid adoption to meet Asian maize consumption in 2003 hybrid area Projected 1992 hybrid needed for self- 2003 hybrid consumption in 1992 hybrid area as a % of sufficiency in 2003 area as a % of 2003 (1,000 ha) area (1,000 ha) of total maize area (1,000 ha) 1992 maize area China 117,000 15,300 73 21,000 100 India 11,800 1,200 20 1,900 32 Philippines 6,300 230 7 1,200 35 Indonesia 6,500 60 2 960 32 Thailand 6,300 320 25 1,300 100 Assumptions: (1) Authors' estimates of the difference between hybrid and OPV yields were used to determine how much production resulted from areas planted to hybrids and OPVs. (2) Hybrid yields are assumed to increase to 2003 at a 2 per- cent annual rate from current levels, and OPV yields at a I percent annual rate; these figures are consistent with long-term yield trends for countries where hybrids or OPVs predominant, respectively. (3) Using figures for current maize total area, the column on the right shows the percent of area that would have to planted to hybrids for countries to be self-sufficient in maize with projected consumption in 2003. EASING BARRIERS TO MOVEMENT OF PIANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 69 The situation calls for a hybrid seed educa- countries have to be prioritized. Maize areas tion program to correct the misinformation of with the best agronomics, the best access to previous years promoted by *ell-meaning but markets, and the most dynamic part of the farm misguided individuals. Governments and population have to be targeted for conversion donors can and should play leading roles in this from OPV to hybrid use. Technology adoption effort. An education program can avoid refer- typically occurs in stages over a period of time, ences to specific brands of hybrid seeds, focus- so attempts to achieve widespread increases in ing on informing farmers about issues such as: hybrid use in all maize growing areas of a coun- try will likely be futile. Rather, a step approach, * the economics of hybrid maize production, targeting those areas most likely to use hybrids, with attention to cost per ton of grain pro- should be used. duced rather than cost per hectare; Fourth, many governments will have to .. th. positive impact ofhybridmaizeonf improve environments to foster a competitive profits; and seed market and to achieve wider use of hybrids. In many cases, investments in infra- * suitability for varying levels of technology structure, post-harvest storage facilities, and (low, medium, or high input) in developing extension services are called for to improve farm countries. productivity and profit. Improved access to credit would in many cases dramatically Attitudes are already changing. Until just a increase farmer use of hybrid seed. Often a couple of years ago, for example, the Philippines more attractive business and legal environment Department of Agriculture promoted OPVs would motivate private seed companies to rather than hybrids in government sponsored increase investment, and ultimately form a raize production programs. Today, however, the viable and competitive seed market. Philippine government pushes hybrids, and even Fifth, age-old obstacles to agricultural devel- CIMMYT promotes hybrid seed technology fior opment (for example, small farm sizes, back- developing countries, ward farmers, etc) should no longer be used as Second, government agricultural and pricing excuses for lack of emphasis on seed market policies can wreak havoc on farmers' decisions development. Seed markets are well established to use hybrid seed. Farmers use good seeds (and in many developing countries despite extreme other inputs, such as fertilizers and chemicals) variations in farm size and sophistication, when they see chances to make money. Like income levels, and the status of agriculture in good investment bankers, farmers exercise good the economy Research is making available many "risk management." Governments complicate viable hybrids for developing countries, and eco- farmers' investment decisions by playing nomic incentives to increase production are evi- around with the question of whether or not to dent. Impediments to hybrid technology adop- import maize. Once maize imports enter the tion can and have been eliminated in many picture, domestic prices drop, and farmers who developing countries despite past reservations invested for high yields may be left holding five about the likelihood of technology transfer. metric tons per hectare with domestic prices half of what they were at time of planting and Closing statement before maize imports. Some governments "tilt the field" in favor of Formulating and adopting consistent agri- local maize breeders and against international cultural policies will lead farmers toward greater companies. Governments do this by requiring productivity and ultimately lower consumer lengthy variety registration procedures or sim- prices. ply by providing no protection for intellectual Third, key maize production areas within property. 70 PROSPECTS AND CONSTRAINTS FOR HYBRID MAIZE IN DEVELOPING COUNTRIES However, maize plants wave no flags! The order to give them the plant characteristics they plant's germplasm is so varied and mixed that it need to be successful. Countries can encourage crosses over many political boundaries. foreign companies to bring in their best materi- Ultimately the most important issue is to get als by establishing systems to protect plant the best possible genetics into farmers' hands in breeder's rights. An International Strategyfor Cotton Breeding, Variety Introduction, and Seed Supply Donald A. Pallin Delta and Pine Land Company (D&PL) is cottonseed technology is pushing D&PL the largest private cottonseed breeding, produc- toward joint ventures, where D&PL has more tion and marketing company in the world. influence over day-to-day operations and can D&PL has commercial cottonseed products on oversee technologies that are becoming more the market in fourteen countries. In the United important; that is, delinting, conditioning, and States, varieties owned or controlled by D&PL treating. Also, the new transgenic varieties will are planted on 70 percent of cotton area. The require more experienced management. company is a leader in bringing new biotech- On first entering into a new country mar- nology cottonseed products to the market. ket, D&PL screens large numbers of cultivars Cotton is *a unique crop in the developing to evaluate competitiveness of its own world. Cotton is a raw material for textile germplasm against existing varieties. If needed, industries within developing countries and an and depending on market opportunities, important earner of foreign exchange. Because D&PL develops new varieties through conven- of cotton's commercial importance, perfor- tional breeding protocols to overcome any mance of the crop from year to year often has shortcomings in its products that are identified social implications. in this screening process. In developing countries where cotton is a Once varieties are identified or developed, major crop, governments characteristically have the company invests in extension-type market- strong interest in and influence over cotton ing to demonstrate advantages of its new prod- production. Whole bureaucracies exist to sup- ucts. port and control production. Self-protective When entering a new market, D&PL evalu- behavior by these bureaucracies sometimes ates local farming and ginning practices in inhibits introduction of new varieties or other terms of their ability to produce high-quality technologies. planting seed. Where seed markets and produc- tion possibilities are satisfactory, D&PL finds a D&PL approach to foreign markets local partner, forms a joint venture, and con- The company's initial international efforts structs delinting and conditioning facilities. were based on license agreements with local The company introduces strict quality assur- seedsmen to produce and market D&PL vari- ance controls and procedures in order to assure eties. However, increasing complexity in consistent production of high-quality seed. D&PL aims for country-by-country self-suffi- ciency in cottonseed production and processing Donald A. Pallin is Vice President and General to keep costs down. Once present in a country, Manager, International Division, Delta and Pine the company introduces new technologies as Land Company. market development warrants. 72 AN INTERNATIONAL STRATEGY FOR COTrON BREEDING, VARIETY INTRODUCTION, AND SEED SUPPLY Condusions and recommendations laws, which are particularly important for cot- Developing nations can participate in and ton, a self-pollinated crop. benefit from new technologies if they will create Corruption, bureaucratic stone-walling, and a level playing field on which all parties who other impediments to market access must be have something to offer can compete on the avoided at all cost or new technologies will pass basis of product performance. This level playing countries by. Government leaders are encour- field comes with free markets. Testing and regis- aged to recognize the potential advantages that tration requirements can be obstacles to free modern technology can bring to their people, markets. Market forces are very efficient in eval- to encourage collaborative efforts by technology uation and accepting or rejecting new technolo- companies and local organizations, and to gies. maintain a friendly and welcome environment D&PL urges governments to enact and for organizations with a proven track record in enforce plant variety protection and patent developing useful technologies. Assessment of the International Transfer of Wheat Varieties Mywish Maredia, Richard Ward, and Derek Byerlee In this paper we present some empirical tem? We then present estimates of actual trans- indicators of potential and actual transfers of fers of wheat varieties from a global perspective wheat varietal technology across agroclimatic and discuss the implications of the findings for environments and countries in order to gain the design of national and international wheat insight into some of the factors determining research systems. We open with a brief discus- spillovers of varietal technology. The analyses sion of the concept of a spillover matrix and presented in this paper also highlight the articu- relate it to the basic model used in this study to lation and degree of complementarity of estimate potential spillover coefficients. national and international crop improvement research efforts as reflected in potential and Spillover matrix: a conceptual frAmework realized international transfers of wheat vari- A critical feature of much agricultural tech- eties. nology is its environmental specificity. Transfer We first use an econometric approach to of agricultural technology, particularly biologi- estimate the coefficients of a global potential cally-based technology such as improved vari- spillover matrix for wheat improvement eties, is limited by the spatial and temporal vari- research. We demonstrate the value of interna- ation of environmental factors such as climate, rional yield trial data in estimating potential biotic and abiotic stresses, photoperiod, and soil spillover effects of an international wheat type. For instance, no one wheat variety will improvement research system to address the fol- excel everywhere and at all times. The concept lowing issue: To what degree is wheat varietal of a spillover matrix, C, makes the notion of technology environment specific?l In other environmental heterogeneity more tractable. words, is there a yield advantage for varieties The matrix C is usually an m x m matrix (where developed for a specific target environment m is the number of agro-ecological environ- compared to varieties developed for other envi- ments) with spillover indexes or coefficients, cij. ronments and by the international research sys- These technology spillover effects, ci, measure the performance of a technology developed for Mywish Maredia is with the Department Of environment i, in environment 'j', in relation to Agricultural Economics, Michigan State the technology developed for environment j. University. Richard Ward is in the Department of Potential sources of technology spill-ins are Crops and Soil Sciences, Michigan State not only research programs in other environ- University. Derek Bycrlee is in the Agriculture and ments, but also international and regional Natural Resources Department of the World Bank. research programs that develop crop varieties not specific to a particular environment in a Due to space constraints the orikinalpaper has country. In such cases, technologies emerging been editedfor this publication. Elipses mark from international programs are considered as places where material has been dekted 74 ASSESSMENT OF THE INTERNATIONAL TRANSFER OF WHEAT VARIEnES an additional source of spill-ins but there is no difficult to delineate geographical areas that corresponding target environment, implying respond differentially to new technologies (that that the spillover matrix need not be square. is, G x E interactions). However, it should not As noted by Pardey and Wood (1994), two be so highly disaggregated that the dimensions major issues need to be addressed in construct- of C become impractically large, leading to esti- ing such a matrix. The first relates to the esti- mation problems. The environmental classifica- mation of spillover coefficients and the second tion system should, therefore be crop specific is the environmental classification system and based on those delineating biological char- employed. These two issues are discussed below acteristics (that is, classification criteria) that within the context of varietal technology. allow for the expression of genotype by envi- ronment interactions. Spillover coefficients Most studies in the past, including In the case of varietal improvement research, Englander (1991) and Davis et al (1987) have the elements ci1 can be defined as the potential either used the Papadakis (1966) or the FAO decline (or increment) in the yields of varieties climatic classification to characterize a location. developed for environment i when evaluated in In this study, however, we use the global environment j (Y- ), relative to the yields of megaenvironment classification system devel- varieties developed for environment j (Y-.): oped by the International Maize and Wheat Improvement Center (CIMMYT)(table 3.4). cii = yijyY. (1) CIMMYT defines megaenvironments specifi- cally for wheat improvement research, in terms Because of environment specificity, it is of areas of similar biotic and abiotic stresses, expected that cij < cjj - that is yields are less in cropping system requirements, and consumer environments for which the varieties are not preferences for types of wheat (Rajaram et al specifically designed. Because of the differential forthcoming). The CIMMYT megaenviron- response of genotypes to environmental varia- ment classification system is used in this study tion (G x E, or genotype by environment inter- instead of Papadakis or FAO, because the latter actions), the matrix C may not be symmetric. is inadequate for a specific commodity like That is to say, cij is not equal to 9i. G x E inter- wheat. Unlike the general Papadakis system, the actions result when test environments are het- CIMMYT system is based explicitly on selec- erogenous for properties which evoke different tive environmental properties such as the mois- responses in the genetic lines evaluated. Those ture and temperature regimes in the winter sea- environmental properties are referred to as son when wheat is grown. It delineates irrigated "selective" to distinguish them from other envi- from non-irrigated areas within an agro-ecolog- ronmental properties (Antonovics et al 1988). ical environment, a distinction that is especially Most studies in the past have used subjective important for wheat. estimates of Yij in order to estimate potential spillover coefficients, c1j. In this study, we use Econometric procedure and data sources econometric procedures to quantify these esti- CIMMYT's International Spring Wheat mates.... Yield Nursery (ISWYN) trial data for the years 1979-80 to 1987-88 were used to estimate a Environmental clasification system global potential spillover effects matrix for The environmental classification system wheat improvement research.2 This data set determines not only the dimensions of the includes more than 24,000 yield observations, spillover matrix but also the biological basis for of which about 23,000 were used after exclud- estimating and interpreting the spillover coeffi- ing all observations pertaining to triticale and cients. Unless disaggregated, it will be extremely durum wheats. Also, local checks were excluded Table 3A Classification of spring wheat m.gaenvironments (MEs) used by CIMMYT wheat program Latitude Moisture Temperature Breeding Representative ME degrees regime a re;me b Sown objectives c locations/regions MEI d < 40 Low rainfall; Temperate Autumn Resistance to Yaqui valley, Mexico; irrigated lodging, SR, and Indus valley, Pakistan; LR Gangetic valley, India; Nile valley, Egypt ME2 <40 High rainfall Temperate Autumn As ME1 + Mediterranean basin; resistance to Southem cone; Andean YR, Septoria highlands; East African spp., Fusarium Highlands spp. and sprouting ME3 < 40 High rainfall Temperate Autumn As ME2 + acid Brazil, Andean; soil tolerance Highlands, Central Africa; Himalayas ME4A c 40 Low rainfall; Temperate Autumn Resistance to Aleppo, Syria; Settat, winter drought, Morroco dominant Septoria spp. and YR ME4B < 40 Low rainfall; Temperate Autumn Resistance to Marcos Juarez, Argentina summer drought, dominant Septoria spp., Fusarium spp., LR and SR ME4C < 40 Mostly residual Hot Autumn Resistance to Indore, India moisture drought ME5A < 40 High rainfall; Hot Autumn Resistance to Poza Rica, Mexico; irrigated; heat, Joydebpur, Bangladesh; humid Helminthosparium Encamation, Paraguay spp., Fusarium spp., and sprouting ME5B < 40 Irrigated; low Hot Autumn Resistance to Gezira, Sudan; Kano, humidity heat and SR Nigeria ME6 > 40 Moderate Temperate Spring Resistance to YR, Harbin, China rainfall; LR, Fusarium summer W., dominant Helminthosparium spp., and sprouting a. RainfalH refers to just before and during the aop cycle. High - > 500mm, low = < 500mm. b. Hot * mean temperature of the coolest month > 17.5oc; cold = < 5.Ooc. c. Factors additional to yield and industrial quality. SR - stem rust, LR = leaf rust, YR = yellow (stripe) rust. d. Further subdivided into: (1) optimum growing conditions, (2) presence of Karnal bunt, (3) late planted, and (4) prob- lems of salinity. Sewrm Byerlee and Moya (1993). 76 ASSESSMENT OF THE INTERNATIONAL TRANSFER OF WHEAT VARIETES because many were either not reported by the are tested) perform relative to varieties devel- cooperators, not identifiable because of lack of oped in other megaenvironments (irrespective information on cross and selection history, or of their country of origin).6 Also, we are inter- were duplicated as one of the nonlocal check ested in the issue of transferability of wheat entries.3 There were 209 unique wheat varieties varieties developed by the international wheat in the 364 entries over the period of eight years. improvement research system spearheaded by The number of different locations in 81 coun- CIMMYT. This international research system tries totaled 195. The trial locations were classi- consists of the collaborative research and testing fied according to CIMMYT's megaenvironr- efforts by CIMMYT and the NARSs around nents discussed above (table 3.4). The wheat the world. Its aim is development of high yield- varieties were classified by their institutional ing, widely adapted wheat varieties that can be origin as either: used by NARSs either as seed products or breeding parents in their wheat improvement * NARS (National Agricultural Research programs. System) varieties (that is, crossed, selected The regression model used to estimate the and tested by national programs), or spillover matrix is shown in Equation 2 (see below), where, * "CIMMYT varieties"4 (developed through the international CIMMYT-NARS research j is the test megaenvironment in which collaborative system; that is, crossed and ini- the yield data point is observed. The tial selections done by CIMMYT but with equations were estimated separately testing conducted by national programs). for the following seven megaenviron- ments - MEI, ME2, ME3, ME4A, The NARS varieties were further classified ME4B, MESA and ME6 described in by their environmental origin based on the table 3.4.7 dominant megaenvironment in the country or region of development and information on the Y) is the observed yield (kg/ha) of the environmental niche (rainfed, irrigated, etc) for hgt g'th entry at the h'th trial location in which the variety was released. CIMMYT vari- environment j and in e'th trial year. eties were classified into two classes: (1) those released in Mexico (CIM 1); and (2) those DLOCh is a vector of dummy variables equal released in countries other than Mexico or not to one if the data point belongs to reeased by any national program (CIM2).5 location h, zero otherwise. The question addressed in estimating a global spillover matrix is: how do varieties DYEAR, is a vector of dummy variables equal developed in a given megaenvironment (that is, to one if the data point belongs to the megaenvironment in which all the varieties trial year t, zero otherwise. H T m Yhgt =a+XbhDLOCb +,c,dYEAR, +v VINT+X,w1DORIG. +r MR+Ehgt h=I t=I i=I forj= 1, 2,...,n (2) EASING BARIUERS TO MOVEMENT OF PLANT VARIETES FOR AGRICULTURAL DEVELOPMENT 77 VINT is a variable to reflect the age or vin- Since the model is estimated separately for tage of a variety approximated by the each megaenvironment, the coefficients for trial year in which the g'th variety DORIG represent the performance of varieties first appeared. of different environmental origins in a given megaenvironment relative to the "home vari- DORIG; is a vector of dummy variables equal eties." The varietal group originating from the to one if the g'th variety belongs to test megaenvironment were considered as the the origin group i (that is, developed benchmark variable (that is, dummy variable for megaenvironment i), zero other- DORIGJ were dropped from the equation for wise. There are nine such dummy each megaenvironment). Therefore, the coeffi- variables; seven correspond to NARS cients of DORIG; are the differential yields varieties classified by their megaenvi- defined as (wJi = Yi- Y--). These coefficients can ronment origin (DOMEI, DOME2, be used to estimate Yij/Yjj to give the elements DOME3, DOME4A, DOME4B, of the potential spillover matrix, cij, based on DOMESA, DOME6); and two cor- the constant Y j (approximated by the arith- respond to CIMMYT varieties metic mean) for each megaenvironment. released in Mexico (DCIM1) and elsewhere or not at all (DCIM2). Empirical results and the estimation of global potential spillover matrix MR is the inverse Mill's ratio (described Model parameters in Equation 2 were esti- further below). mated using the ordinary least squares method. The statistical results of the regression analyses E IS the crror term. are summarized in Table 3.5. The results indi- cate that the inclusion of dummy location vari- Thus, the performance of a variety is ables had a significant positive effect on the R2 assumed to be a function of environmental vari- of all the seven regression models. Except for ables (DLOC, DYEAR) and technology vai- ME1 and ME2, the dummy variables for trial ables (VINT, DORIG). The variables VINT.. .2 ryears also significantly increased the R2 of the and DORIG represent characteristics of a vani- estimated models. etal technology. Since we are using panel data, The coefficient of VINT variable measures the location and year dummies (DLOC and the gain in average yield/ha/year of new var- DYEAR) are included to factor out the site and cties n ia given megaenvironment. Note that time effect on the observed yields.eteinagvnmanirmntNoeht Time yieffectrio dath obsrved yiaraterids d byvar the coefficient is an average for all the varieties The yield trial data are characterized by van- and is not specific to a particular origin group. etal attrition due to the replacement of older Except in ME2 (high rainfall, temperate), ME3 varieties by better yielding varieties in successive (high rainfall, acid soils), and ME4B (low rain- years of the trials. Since the probability of vari- fall,ewinter drough), yed improv are etal attrition is correlated with experimental not significantlydife fmprom zero. Tes response (that is, yield), the traditional statisti- no . non-significant coefficients of VINT variable in cal techniques for panel data estimation will many environments induding MEI (irrigated, providc* biased and inconsistent estimators temperate) confirm the difficulty that wheat (Hsiao 1986). The variable MR (inverse Mill's breeders have faced in maintaining significant ratio) is induded in the equation to correct for growth rate in yield potential since 1980s (Bell this selection bias of non-randomly missing et al 1994). As indicated by coefficients of the varieties in the yield trials conducted over a MR variable (inverse Mill's ratio), there is a pos- number of years (Hsiao 1986). itive and highly significant (in most of the Table 3.5 Regression results of potential spillovers at the megaenvironment level using ISWYN data, 1980s ME1 ME2 ME3 ME4A ME4B ME5A ME6 Independent variables Irigated High rainfall Acid soils Winter drought Early drought High temperatures High latitude 1. Constant a 4880*** 3390*** 336** 2041 *** 1942*** 2221 *** 3394*** 2. Dummies for year R2 change b 0.02 0.02 0.23 0.17 0.17 0.05 0.08 F change c 35*** 32*** 1 84*** 1 44*** 46*** 1 5*** 124*** 3. Dummies for location R2 change b 0.56 0.44 0.27 0.40 0.21 0.29 0.52 F changec 166*** 131*** 287*** 159*** 59*** 11 3*** 154*** 4. VINT a 4.27 31.2 10.9* 2.5 28.1** -2.2 4.7 5. Mill's ratio, MR a 155*** 135*** 111 ** 93 141 ** 97** 87.7** 6. Oigin, DORIG a,d DOMEl: Irrigated 0 -1 89** -406*** -374** -346** 34 -223*** DOME2: High rainfall -232*** 0 -509*** -307* -275* -177 -175** DOME3: Acid soils -507*** -141 0 -568*** -282* -31 1 DOME4A: Winter drought -66 -226* -565*** 0 -483** -154 -259** DOME4B: Early drought -486*** -101 -290** -334* 0 -161 -56 DOMESA: High temperature -593*** -525*** -219 -672*** -328 0 -334** DOME6: High latitude -588*** -395*** -414*** -507*** -270* -264** 0 DCIM1 527*** 490*** -14 20 191 23 -91 DCIM2 227*** 230*** -138 -105 16 7 -131** Number of observations 4641 4248 719 1824 850 935 2913 R2 0.61 0.53 0.78 0.65 0.40 0.53 0.68 *P < 0.05 **P < 0.01 ***P < 0.001 a. Number given is the estimated coefficient (kg/ha). b. Number given is the change in R2 when a given set of dummy variables is entered in the equation that includes all the other variables. c. Number given is the change in F-ratio when a given set of dummy variables is entered in the equation that indudes all the other variables. d. Origin groups DOMEI to DOME6 represent cultivars developed by national programs for respective megaenvironments. DClM1 indicates CIMMYT cultivars released in Mexico and DCIM2 indicates CIMMYT cultivars released in countries other than Mexico or not released anywhere. EASING BARRIERS TO MOVEMENT OF PLANT VARIEnES FOR AGRICULTURAL DEVELOPMENT 79 megaenvironments) relationship between test megaenvironments indicates the potential observed yields and the probability of retention of CIMMYT varieties to spill-over to these test in the trials. megaenvironments.8 The coefficients of origin variables (wi) esti- Akin to previous studies, the spillover coeffi- mate the yield advantage (or disadvantage) of cients are presented in table 3.6 in terms of per- varieties originating in different environments centage coefficients based on the average yields relative to the test environment (kg/ha). The of the benchmark variable (that is, c1. = Y.1/Y11). zeros on the diagonal indicate that the coeffi- Off-diagonal values less than onre indicate that cient of variety group of the same environmen- directly introduced wheat varieties from other tal origin as the test environment is defined as megaenvironment yield less than those devel- the "benchmark" and all the other coefficients oped by local breeding programs in the test in that column represent deviations from that megaenvironment. Similarly, values greater than value. one (as in the case of CIMMYT varieties) indi- The negative values of NARS technology in cate that directly introduced wheat varieties all the megaenvironments confirm the hypothe- from these sources yield more than those devel- sis that varieties developed in a test megaenvi- oped by local breeding programs in the test ronment perform better than varieties devel- megaenvironment. oped in other megaenvironments. For example, The significant yield advantages expressed the second number in the first column shows by varieties developed and evaluated in ME1, that NARSs varieties of ME2 (high rainfall) ori- ME2, ME3 and ME6 (implying less direct gin yield 232 kg/ha less on average in ME1 spill-ins of NARS varieties from other megaen- (irrigated) than the NARSs varieties developed vironments) can be explained by the fact that for MEI (after adjusting for other variables). these megaenvironments are comprised of The strength of this relationship is shown by countries with a strong experience in wheat the fact that nearly all the off-diagonal elements research - for instance, India and Pakistan in are negative and usually statistically significant. ME1 (irrigated), Turkey and Spain in ME2 Significant negative values in a given column (high rainfall), Brazil in ME3 (acid soils) and result from either: (a) both genetic differences the developed countries of Europe and North among varieties and a difference in the selective America in ME6 (high latitude). On the other environment at the test versus origin environ- hand, the "environmental distance" plays a role ments, or (b) only a difference in the genetic in explaining the significant yield advantage properties of the varieties tested. The latter cir- enjoyed by domestic varieties in ME4A and cumstances could reflect different levels of ME4B (drought environments). To a certain breeding success and would result in symmetri- extent this also holds true for ME3 (acid soils) cal relationship such that wji = - wi. The abun- and ME6 (high latitude). For example, grow- dance of negative values both above and below ing conditions in ME3, except for the acid soil, the diagonal show that CIMMYT's megaenvi- are very similar to those in ME2 in terms of ronment system reflects true differences in water supply and temperatures (that is, envi- selective environmental properties. ronmental distance is less). Thus, ME3 vari- The last two rows show that CIMMYT vari- eties perform relatively well in ME2. However, eties perform well in most megaenvironments, in ME3 the soil toxicity adds to the distance especially in MEI (irrigated) and ME2 (high between the two environments and constrains rainfall). For example, CIMMYT varieties the transferability of technology from ME2. released in Mexico (DCIM1) enjoy a yield This is evident from the highly significant yield advantage of 527 kg/ha in MEI (irrigated) disadvantage of ME2 varieties (19 percent) compared to NARS varieties of ME1 origin. when planted in ME3 compared to the small The positive yield advantage of CIMI in many and less significant yield disadvantage of ME3 Table 3.6 Estimated spillover matrix for wheat improvement research at the global megaenvironment level Megaenvironments where varietie' are tested a 1 2 3 4A 4B 5A 6 Origin of vafiety Irrigated High rainfall Acid soils Winter drought Early drought High temperature High latitude ME1 Irrigated 1.00 0.95 0.84 0.90 0.88 1.02 0.94 ME2 High rainfall 0.95 1.00 0.81 0.92 0.90 0.89 0.96 ME3 Acid soils 0.89 0.96 1.00 0.85 0.90 0.98 1.00 ME4A Winter drought 0.99 0.94 0.78 1.00 0.83 0.91 0.93 ME4B Early drought 0.90 0.97 0.89 0.91 1.00 0.90 0.99 ME5A High temperature 0.88 0.86 0.92 0.82 0.89 1.00 0.92 ME6 High latitude 0.88 0.89 0.84 0.87 0.91 0.84 1.00 CIM1 CIMMYT/Mexico 1.11 1.13 0.99 1.01 1.07 1.01 0.98 CIM2 CIMMYT/Other 1.05 1.06 0.95 0.97 1.01 1.00 0.97 a. Yildd expressed rdative to the yidd of cultivars originating in that megaenvironment (= 1.00). EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 81 varieties (4 percent) planted in ME2. The grams in thirty-eight developing countries asymmetry in the environmental distance (Byerlee and Moya 1993). The inventory of all between two environments explains the asym- the wheat varieties released in developing coun- metry in the spillover matrix (that is, C i is tries during the 1966-90 period is classified unequal to C 1). according to the origin of cross as either CIM- If we examine the performance of CIM- MYT variety or NARS variety. This informa- MYT varieties (CIM 1 and CIM2) across tion is used here to assess the direct and indirect megaenvironments, the prominent result of the international transfer of wheat v'arieties. Direct regression analyses is the wide adaptability and transfers are those varieties that are developed transferability of CIMMYT varieties to differ- by research program A and direcdy released by ent environments. The environmental specifici- research program B after local testing; an exam- ty and associated selective environmental het- ple is Pak 81, a CIMMYT bred variety released erogeneity evident in the comparison of NARS in Pakistan. By contrast, a variety used as a par- varieties is minimized when CIMMYT varieties ent material in another country's breeding pro- are compared across different megaenviron- gram is a case of indirect transfer. ments. This points to the success of the interna- Several important changes have occurred in tional research system in overcoming G x E the global system for producing varietal tech- interactions and developing widely adapted nologies since 1966, which reflect the increas- varieties, at least in the irrigated and high rain- ing mobility of wheat varieties across interna- fall environments, which account for about 70 tional boundaries. The proportion of releases percent of the spring wheat production in that are direct CIMMYT transfers has increased developing countries. by 26 percent, and indirect transfers increased by 11 percent. In 1986-90, 85 percent of Evidence of actual international transfer of NARS releases were based on CIMMYT wheat varieties germplasm (that is, direct plus indirect trans- The major source of data for the empirical fers), compared to just 48 percent in 1966-70 analysis of actual spillovers is a 1990 survey (Byerlee and Traxler 1995). conducted by CIMMYT of wheat research pro- By 1990, more than 650 (about 50 percent of total) wheat varieties released in developing Table 3.7 Varieties released in developing countries, classified by the type of technology transfer embodied, 1966-90 Technology transfer group Number of varieties Percentage of varieties 1. Direct Transfer (a) from CIMMYT 575 44.3 (b) from other NARS 79 6.1 2. Indirect Transfer a (a) from CIMMYT 235 22.9 (b) from other NARS 38 2.9 3. Non-transferred varieties b 389 23.9 Total 1,317 100 a. Varieties based on NARS crosses using parent materials from other sources. b. Varieties developed by NARS without direct or indirect transfers. Source: Byerlee and Moya (1993). 82 ASSESSMENT OF THE INTERNATiONAL TRANSFER OF WHEAT VMIETnES Table 3.8 Percent of wheat varieties released in developing countries by type of technology transter and region, 1965-90 Region West Asia & Sub-Saharan North Affica South Asia Latin Ameulca 1. Direct Transfer (a) from CIMMYT 40.1 51.8 41.0 45.8 (b) from other NARS 20.4 26.3 4.8 2.1 2. Indirect Transfer 20.4 8.8 31.0 22.5 from CIMMYT a 3. Non-transferred varieties b 19.1 13.2 23.2 29.6 Total 100 100 100 100 a. Varieties based on NARS crosses using CIMMYT parents. b. Varieties developed by NARS without direct or indirect transfers. Source: CIMMYT Wheat Impact Survey, 1990. Table 3.9 Classification of NARSs by the extent of released varieties based on direct transfers, 1966-90 Less than 25 percent 26 to 50 percent 51 to 75 percent 76 to 99 percent 100 percent Kenya Argentina Ecuador a Guatemala Algeria Peru Brazil Egypt Ubya Bangladesh a Chile Ethiopia a Mexico Bolivia Colombia Iran Morocco Burundi India Jordan Saudi Arabia Lebanon China Pakistan Sudan a Myanmar a (South) Paraguay Yeman a Nepal a Zimbabwe Syria Nigeria Tunisia Tanzania a Turkey Zambia Uruguay a a. Countries with a significant release based on direct transfers from other NARS. Source: Adapted from Byerlee and Moya (1993). countries were directly transferred from CIM- West Asia, North Africa, and Latin America, MYT or other NARSs (table 3.7). Another 25 where a greater number of varieties are based on percent of the wheat varieties were based on direct transfers. indirect transfers from CIMMYT or other The importance of direct CIMMYT trans- NARSs. The use of CIMMYT germplasm as fers is also evident from table 3.9, which indi- direct transfers is consistently high across cates that most of the smaller NARS have regions (table 3.8). The tendency toward using depended on direct transfers for more than half CIMMYT materials as indirect transfer is of their varieties; and a third of these have greater in Asia, where most advanced wheat depended 100 percent on direct spillovers. breeding programs are located, compared with Important exceptions to this rule are Mexico EASING BARRIERS TO MOVEMENT OF PLANT VARIEnES FOR AGRICULTURAL DEVELOPMENT 83 (where CIMMYT is located) and Pakistan matrix are less than those of the diagonal ele- (where the wheat-growing environments are ments. very similar to those in Mexico). There are also The results for wheat presented in this paper, significant country-to-country direct spillovers, do not sustain this "location specificity" argu- as some of the countries with small wheat pro- ment (at least in terms of yields) when the inter- grams, such as Sudan, Bangladesh, and Nepal, national research system is considered as a import large numbers of varieties from neigh- source of research spill-ins. Wheat varieties orig- boring countries. Most larger NARSs, on the inating from collaborative CIMMYT-NARS other hand, have demonstrated a significant international research system have proven to be ability to develop varieties from their own highly transferable within megaenvironments crosses. Access to CIMMYT germplasm is, and across different countries around the world. however, crucial to the wheat improvement The yield advantage of varieties developed by programs in these large countries; even the the international research system, was as high as Indian and Chinese spring wheat programs 13 and 11 percent in the high rainfall and irri- make use of CIMMYT germplasm in more gated environments, respectively. In other than 50 percent of their releases through a megaenvironments (such as low rainfall, acid CIMMYT parent used to produce an adaptive soils, high temperatures, etc), the yields of CIM- variety. MYT varieties, although higher than imported The increased coordination between CIM- NARS varieties, were not significantly different MYT and NARS in the global system of testing from yields of locally developed varieties.... and sharing of germplasm has been an impor- Thus, the overarching result of the global . . tant component of the progress that has been . . analyses is that varieties developed by the achieved. Two factors have been crucial for the international research system perform better large direct and indirect transfer of wheat vari- than or at par with the NARS cultivars in most eties that has occurred: (a) germplasm sharing of the major spring wheat environments indi- through a network of international wheat nurs- cating the success of the international research eries coordinated by CIMMYT scientists since system in developing widely adapted wheat 1964; and (b) ongoing training and scientific varieties. This success in combining high yield exchanges between CIMMYT and NARS potential and wide adaptation can be attributed wheat breeders. In particular, it appears that the to: (a) large number of crosses (12,000 per year) nursery system significantly reduces the transac- made by CIMMYT breeders in Mexico; (b) the tions costs of transferring varieties developed use of "shuttle breeding" that allows CIMMYT anywhere in the world. scientists to alternate selection cycles in differ- ent environments with high yield potential that Conclusions and implications differ in altitude, latitude, photoperiod, tem- Many important results pertaining to the perature, rainfall, soil-type and disease spec- issue of technology transferability emerge from trum, and (c) the wide testing of advanced lines the estimation of the potential spillover matrix in collaboration with NARSs throughout the and actual spillovers at the global and country world (Romagosa and Fox 1993). The compar- levels. Research evaluation models have often ative advantage of this international research used spillover matrices to account for the bene- system lies in its ability to conduct such a large fits of research conducted by other research breeding operation. programs in similar and different environ- However, it should be noted that wheat vari- ments. These estimates have been based solely eties are probably more "environmentally on subjective guesses and on the assumption of robust" than varieties of many other crops in location specificity, which implies that the val- terms of international transferability. because the ues of the off-diagonal elements in the spillover differences among production environments 84 ASSESSMENT OF THE INTERNATIONAL TRANSFER OF WHEAT VARIETIES and local quality preferences are not as marked national wheat research programs. These coun- as in other crops such as rice, maize or beans. tries need not devote resources for each and Evidence on the origin of varieties released in every environmental niche in the country. They developing countries support these results of the can utilize their resources more efficiently by potential spillover matrix. The analysis of actual following a mixed strategy of direct importation spillovers indicate that more than 50 percent of of technology in some environments and local total wheat varieties released in developing development of technologies in other environ- countries in the 1980s were directly transferred ments which are unique to the country. CIMMYT varieties. Also, 10 out of 36 countries There are however, a few caveats to be noted surveyed by CIMMYT, report that 100 percent about this study. First, given the fact that of all wheat varieties released in 1965-90 in ISWYN trials are conducted by CIMMYT for these countries were based on direct transfers the purpose of disseminating its germplasm, from the international research system. Even a there is an overwhelmingly large representation large wheat producing country like Pakistan of CIMMYT varieties (about 50 percent) in the depends heavily on direct transfers from this data analyzed in this paper. international research system. For example, 80 Second, the results are based on the megaen- percent of all varieties released in the Punjab for vironment classification system that may over- the normal planting date in the period 1980-90 look important within megaenvironment varia- were CIMMYT-origin. tions such as late planting in intensively These empirical findings, based on the ori- cropped irrigated areas ... . The transferability gin of the released varieties in developing coun- of CIMMYT varieties may differ within a tries and the coefficients of the spillover matrix megaenvironment depending on the cropping estimated in this paper, suggest that both the system of a region and other country-specific potential and actual spillovers of wheat breed- factors. ing research are larger than have been reported Third, this analysis ignores other important to date. They also suggest the comparative factors like grain color, quality and stability advantage of the international research system which may be important in determining the in producing widely adapted wheat varieties. local acceptability of varietal technology. If the The possibilities of direct transfer of varieties technology available from other sources is high developed by other programs that might be uti- yielding in the local environment but not com- lized effectively in a given area should therefore patible with the socio-economic environment, be taken into consideration in determining the then national programs can justify a local appropriate type of wheat research in a given breeding program on the basis of other traits. environment. Countries where wheat is not an But breeders agree that in field crop like wheat, important crop or where national agricultural yield is the most important trait used in making research systems are not highly developed can decisions about releasing wheat technology to consider the option of direct transfer of varieties farmers. developed by CIMMYT or other national This paper has provided empirical estimates wheat breeding programs as an alternative. This of potential spillover coefficients, which have is especially so for countries where wheat is hitherto been based on subjective guesses. In grown under irrigated or high rainfall condi- the age of shrinking budgets for agricultural tions. These countries can benefit substantially research, national programs will have to seek from only a testing program without incurring advantage of research spill-ins from not only large costs in adaptive breeding (crossing and other NARSs in similar or other environments, selection) research. There are also implications but also from the regional and international for countries with large wheat growing areas or research systems. This paper has demonstrated diverse environments and which have strong the usefulness of national and international EASING BARRIERS TO MOVEMENT OF PLANT VARIETIEs FOR AGRICULTURAL DEvELoPMENT 85 yield trial data and the data on the origin of 8. A note of caution is needed on the comparability of the varieties in providing estimates of potential and coefficients across columns. The values of the coefficients reported in Table 3.5 are relative to the benchmark origin actual spill-ins, from other research programs group (represented by zeros), and are therefore compara- and international research system. Such infor- ble across rows (technologies) but not across columns mation can be used to make strategic decisions (environments). Thus we can say that in ME2, MEI tech- nology yields 189 kgtha less than ME2 technology, but it about the design of crop breeding programs is erroneous to say that MEI technology yields 189 kg/ha both at national and international levels that less in ME2 than in MEI. would lead to a more efficient global system of agricultural research. References Notes Antonovics, J, N C Ellstrand, and R N Brandon. 1988. "Genetic variation and environmental variation: 1. This notion is usually referred to in the literature as Expectations and experiments," in: L D Gotdieb and 'location specific." S K Jain, eds, Plant Evolutionary Biology. London: 2. With the exception of ISWYN year 1982-83, which Chapman and Hall. was not induded because data were incomplete. Bell, M A, R A Fisdher, D Byedree, and K Sayre. 1994. 'Genetic and agronomic contribution to yield gains: 3. Since local checks are likely to be the best cultivars A case study for wheat," draft paper. Mexico: CIM- grown by the farmers in a given location, their exdcusion MYT. from the analysis may bias the results downward. However, local checks are not synonymous with locally Brennan, J P. 1989. 'Spillover effects of international developed varieties. In fact, about 70 percent of the local agricultural research: CIMMYT-based semi-dwarf checks that were reported and identified were CIMMYT wheats in Australia," Agricultural Economics, 3: bred cultivars released by national programs. 323-332. 4. CIMMYT is an international research center with the Byerlec, D, M R Akhtar, and P R Hobbs. 1987. mandate to provide improved germplasm that can be used "Reconciling conflicts in sequential double cropping by a national program either as parent materials in their patterns through plant breeding: The example of cot- breeding program or released after local screening and ton and wheat in Pakistan's Punjab," Agricultural testing. "CIMMYT variety" as used in this paper is a short Systems 24: 291-304. for "advanced breeding line developed by CIMMYT in collaboration with NARSs" and should not be equated Byerlee, D and P F Moya. 1993. "Impacts of international with the notion that these varieties are released by CIM- wheat breeding research in the developing world, MYT in any given country. 1966-1990," Mexico: CIMMYT. 5. CIMMYT's headquarters is located in Mexico. Byerlee, D and Traxler, G. 1995. "National and However, cultivars developed by CIMMYT have to International Wheat Improvement Research in the undergo the same procedure for release in Mexico as they Post-Green Revolution Period: Evolution and would in any other country. Impacts," American Journal of Agricultural Econonomics, 77:268-278. 6. Since the potential transfer of varietal technology is constrained by differences among environments, the Davis, J S .1994. "Spiliover effects of agricultural research; objective is to analyze technology transfer across megaen- Importance for research policy and incorporation in vironments and not across political boundaries (ie, coun- research evaluation models," in: J R Anderson, ed, tries) as done by Englander (1991). Relating the potential Agricultural Technology: Policy Issues for the transferability of a technology to environmental zones is International Community. Wallingford: CAB important because it allows us to determine the yield International, change as a function of variables which are based on G x E knowledge. Moreover, estimates of potential technology Davis, J S, P A Oram, and J G Ryan. 1987. 'Assessment transferability based on political boundaries are often diffi- of agricultural research priorities: An international cult to interpret (since it is very unlikely that a country or perspective." Canberra ACIAR and IFPRI. politically defined region will have a homogenous crop growing environment). Edwards, G W and J W Freebairn. 1984. "The gains from research in tradable commodities," American Journa 7. Because of insufficient number of observations the ofAgrimulrsralEcononomics, 66: 4149. equations were not estimated for two spring wheat megaenvironments defined by CIMMYT (ME4C and ME5B in Table 3.4). 86 ASSESSMENT OF THE INTERNATIONAL TRANSFER OF WHEAT VARIETIES Englander, A S. 1991. "International technology transfer Papadakis, J. 1966. "Agricultural Climates of the World." and agricultural productivity," in: R E Evenson and C Published by author, Buenos Aires. E Pray, eds, Research and Productivity in Asian Agriculture. Ithaca: Cornell University Press. Pardey and Wood. 1994. "Targeting research by agricul- tural environments," in J R Anderson, ed, Agricultural Evenson, R E. 1989. "Spillover benefits of agricultural Technology: Policy issues for the International research: Evidenee from US experience," American Community. Wallingford: CAB International. JournalofAgriculturalEconomics. 71: 447-452. Rajaram, S M, M van Ginkel, and R A Fischer (forthcom-. 1994. "Analyzing the transfer of agricultural ing). 'CIMMYT's wheat breeding mega-environ- technology," in J R Anderson, ed, Agricultural ments," in Proceedings of the International Wheat Technology: Policy Issues for the International Genetics Symposium, 20-25 July 1993, Beijing. Community. Wallingford: CAB International. Romagosa, I and P N Fox. 1993. "Genotype x environ- Evenson, R E and Y Kislev. 1975. Agricultural Researeb ment interaction and adaptation," in M D Hayward, and Productivity. New Haven: Yale University Press. N 0 Dosemark and I Romagosa, eds, Plant Breeding: Principes and Prospects. London: Chapman and Hall. Hsiao, C. 1986. Analysis of Panel Data, Econometric Society Monographs No. 11. Cambridge: Cambridge White, F C and J Havlicek Jr. 1981. 'Interregional University Press. spillover of agricultural research results and intergov- ernmental finance: Some preliminary results," in Maredia, M K 1993. "The Economics of the internation- Evaluation of Agricultural Research, proceedings of a al transfer of wheat varieties," PhD dissertation. East symposium, 12-13 May 1980, Agricultural Lansing: Michigan State University. Experiment Station, University of Minnesota, pp 60-70. Genetic Resources, International Organizations, and Rice Varietal Improvement R. E. Evenson and D. Gollin Rice varietal improvement is an important preserve germplasm and to provide it freely to component of productivity gains in rice pro- the international scientific community- duction... Until the development of the mod- including national germplasm collections. In ern agricultural experiment stations in the mid- addition, IRRI maintains and coordinates a sys- dle of the 19th century, farmers themselves cre- tem of international nurseries, the International ated improved rice cultivars in the form of dis- Network for the Genetic Evaluation of Rice tinct subspecies types or "landraces" of rice. (INGER), through which advanced genetic Modern plant breeders in national agricultural materials are exchanged and evaluated. research programs have been systematically In this paper we . .. conduct a genealogical developing improved rice cultivars in many analysis of released rice varieties from national countries for the past century. Since 1960, the rice breeding programs and IRRI since 1965, International Rice Research Institute (IRRI) when the first modern high-yielding rice vari- located in the Philippines has been providing eties were released, and we trace the "routes" by international leadership and support to rice which rice germplasm is incorporated into genetic improvement. improved varieties... IRRI has a number of programs to facilitate The first part of the paper presents a brief rice genetic improvement. Its own plant breed- discussion of rice breeding methods and the ing program (IRPB) produces improved culti- role of genetic resources. The second discusses vars, both in the form of "varieties" that are databases and reports basic tabulations of ready for use in farmers' fields and in the form genetic resource characteristics of released rice of "advanced lines" suited for use as parent varieties. The third section reports an anialysis material in national plant breeding programs. of "routes" taken by varieties from origin to IRRI maintains an International Rice release. . . The final section summarizes Germplasm Collection (IRGC) designed to economic implications. Robert Evenson teaches in the Economics Rice breeding and genetic resources Department at Yale University. Doug Gollin is in Rice is a self-pollinated crop. Because of the Economics Department at Williams College. this, genetically segregated lines remain geneti- cally stable from generation to generation. The original of this paper is scheduledfor publica- Genetic changes occur mostly through deliber- tion in 'Economic Development and Cultural ate "crossing" - or hybridizing of parental cul- Change." This abridged version is published with tivars - and pureline selection of the resultant thepermission ofEDCC. Elipses mark places offspring (De Datta 1981). This technique of where material has been deleted due to space con- varietal improvement, known as "pedigree straints. Since one table and several sub-sections breeding" is not widely practiced with other have been cut, those which remain have been cereal grains. renumbered. 88 GENETIC RESOURCES, INTERNATIONAL ORGANIZATIONS, AND RICE VARIETAL IMPROVEMENT Varietal improvement efforts in rice, as in all time researchers in tropical Asia were busy seek- other domesticated crops, date back to the ori- ing more productive varieties of rice from the gins of domestication. The very processes of other two main classes or races: the so-called harvesting and planting tend to select for such indica andjavanica rices. traits as non-shattering, uniform maturation, Among the most important goals of rice high germination, and adaptation to local breeding historically has been the development growing conditions. Harvesting for consump- of fertilizer-responsive varieties that are widely tion will also tend to imply selection for high adaptable to regions of different day length and palatability and nutritional value. Farmer selec- sunlight intensity. More recently, breeders have tion over many centuries has produced todays also sought to incorporate multiple disease and "landraces" of rice. pest resistance as well as tolerance of certain Cultivated rice falls into two species, Oryza environmental stresses, such as soil salinity, iron sativa and Oryza glaberrima. The former is the toxicity, drought or flooding.2 common Asian cultigen, while the latter This process has inherently involved the accounts for a small fraction of African rice pro- exchange of germplasm across countries and duction. In addition to these two cultigens, the agroclimatic zones. As early as the 19th century, genus Oryza includes about twenty wild scientists were transporting rice varieties across species, although some scholarly disagreement national borders in an effort to bring in new remains on the exact number (Chang 1985). It and advantageous genetic materials. This effort has been estimated that about 140,000 cultivars attained new levels in the post-World War II of rice exist in the world today; more conserva- period with the initiation of a program by the tive estimates put the figure around 90,000. United Nations Food and Agriculture These include wild species, landraces, and mod- Organization (FAO) to cross indica rice with ern varieties. japonicas. Because the japonica rices tended to Organized breeding efforts probably date yield better than the indicas, this was seen as a back to before 1000 AD in China, with the possible means of increasing yield of rice in development of the Champa varieties of early South Asia and hence as a means of averting ripening, drought-resistant rices. Modern hunger (Dalrymple 1986; Barker, Herdt, and efforts, however, can be traced to the late 19th Rose 1985, pp 54-55). century in several parts of Asia. In temperate In 1950, a japonica-indica crossing program East Asia, the first significant advances were was started at India's Central Rice Research made by Japanese farmers and scientists in Institute in Cuttack. The FAO program devel- developing relatively short-statured and fertiliz- oped several varieties that proved enormously er-responsive varieties called the rono varieties. successful including Mahsuri, which remains These varieties, which included the popular one of the most widely grown varieties in the Shinriki, belonged to the japonica class of rices, world. Perhaps more important, the FAO pro- and were widely cultivated in Japan as early as gram served as a model for a more intensive the 1890s (Chang 1985, pp 441-443). project initiated in 1960 by the Ford and During the early part of the 20th century, Rockefeller Foundations along with other Japanese scientists sought to create similar vari- donor agencies. The new research center was eties that were adapted to Taiwan, which the the International Rice Research Institute Japanese then occupied. The main thrust of this (IRRI), located in Los Banos, the Philippines. research was to create japonica rices that were Many authors have recounted the develop- adapted to the more tropical conditions of ment of the earliest high-yielding rice varieties at Taiwan; the result was a group of varieties IRRI. Hargrove (1981) has pointed out that the referred to as ponlai rices (Dalrymple 1986; first scientists at IRRI had a deliberate blueprint Barker, Herdt, and Rose 1985). At the same to create a semi-dwarf, fertilizer-responsive rice EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 89 variety that would grow well under a wide range Other germplasm collections are maintained of conditions. The eighth cross made at IRRI by various national programs (as in India). turned out to be a near-immediate success; IR8, Most local stations also have short-term storage as it was designated, was a cross between the facilities or keep stocks of some advanced lines. Indonesian indica, Peta, and the Chinese Breeding activities at the local, regional, and semi-dwarf variety, Dee-Geo-Woo-Gen national levels recombine genetic materials. We (DGWG). An earlier Taiwanese cross of can thus see a complex and involved inter- DGWG with the tall variety Tsai-Yuan-Chung change of germplasm throughout the interna- had led to a similar plant, known as Taichung tional system of rice research establishments. Native 1 (TN1). The two new varieties quickly achieved enormous popularity among farmers Rice varieties and characteristics and scientists around Asia, and a large-scale The role of IRRI is to improve rice produc- internationalization of breeding efforts resulted tivity in the rice-producing countries of the (Barker, Herdt, and Rose 1985, pp 56-57). world. IRRI's main program emphasis has been Plant breeders typically rely on field or labo- on the development of new rice varieties that ratory collections of genetic resources for "raw have desirable characteristics. materials." Such collections are described as ex The breeding sequence begins with the situ, as opposed to in situ collections which are development of a strategy; this suggests a preserved in their natural habitat. About 85,000 potentially desirable cross involving some com- landraces and wild species are currently stored in bination of landraces, wild species, advanced a long-term ex situ facility at IRRI, under condi- lines, or released cultivars. The progeny from tions of low temperature and humidity. (This this cross are then evaluated over several collection is one of the primary responsibilities self-pollinated generations. Selection of plants of the IRGC, mentioned above.) A subset of for uniformity is undertaken. . . Promising these cultivars are catalogued according to agro- selections (cultivars or advanced lines) are field nomic and genetic characteristics. tested. Release as a variety is subject to careful The germplasm bank serves two related evaluation. Only a small proportion of crosses functions. One is as a repository for genetic are ultimately released as varieties (less than one materials. By keeping litde-used or uncommon percent for most programs). Most released vari- varieties of rice in a long-term storage facility, eties are planted by farmers although there is the bank serves as a protective facility to pre- considerable variation in the degree of success vent the loss of potentially valuable gene in the field. sources. From 1965 to 1974, IRRI's breeding pro- The second function of the germplasm bank gram released several important varieties direct- is as a "lending library" of genetic resources for ly as IRRI varieties (IR8 to IR36). National use by plant breeders. Breeders request materi- programs were responsible for most varietal als from the bank - either by name or by some release prior to 1974 and for all releases after set of characteristics - and incorporate the 1974. Some IRRI crosses have been released as materials into their programs of evaluation and national varieties (sometimes with IR names), breeding. Unimproved materials from the col- often after selection in the releasing countries. lection are used by breeders at IRRI and are Similarly, crosses originating in a national pro- also sent out freely to scientists around the gram may ultimately be released as varieties in globe. In addition, genetic materials are also another national program. incorporated in the genealogies of improved lines that are sent out from IRRI through the The varietal database INGER.* For this study we have compiled a database for 1,709 modern rice varieties released since 90 GENETIC RESOURCES, INTERNATIONAL ORGANIZAnONS, AND RICE VARIETAL IMPROVEMENT the early 1 960s.3 For each of these released vari- (23 percent) were the result of a cross made eties a complete genealogy was compiled. This outside the releasing country. IRRI was the included the date and origin of the cross on source for 294 (17 percent) of these varieties. which the variety was based, as well as the data Other national programs were the source for 96 and origin of all parents, grandparents and releases. (Table 3.12 provides country detail for other ancestors. Thus ancestry was traced back varieties.) to original landraces or wild species. In addition After IRRI, India was the next largest we were able to determine whether the cross or exporter of varieties, with 28 Indian varieties any ancestors appeared in the INGER nurseries released elsewhere. India was also a large and whether they were selected from these importer of varieties; 70 of its 643 varieties nurseries for crossing.4 originated elsewhere, with 53 from IRRI. Sri Of the 1,709 modern varieties and elite Lankan varieties were released 11 times in other (advanced) lines, 33 were released prior to 1965 countries. Twelve Thai varieties were released in (and thus prior to the release of any IRRI mate- Myanmar (Burma). Myanmar was one of the rials).5 Table 3.10 gives the frequency of release largest importers of rice varieties; 43 of its 76 by country and by time period. For some vari- releases were imported varieties. including vari- eties, release dates were not available; in such eties from Bangladesh, China, India, Indonesia, cases, approximate dates were estimated based IRRI, the Philippines, Sri Lanka, Thailand, and on available information. Vietnam. The data set includes materials from numer- In addition to IRRI's direct role as a source ous countries. but it is relatively more complete of exported varieties, IRRI has served as a con- for rice-producing countries of South and duit through which elite lines have moved from Southeast Asia than for those from other country to country. Even before the establish- regions. India, in particular, is represented in ment of INGER in 1975, IRRI scientists the data set at a level that appears to be dispro- helped test and disseminate elite lines of rice portionately large, with 643 varieties. Although around the world. This function was formalized India's breeding programs have a long and pro- with the inauguration of INGER. Through ductive history, the data set probably reflects a INGER's activities, elite lines and released vari- bias towards India based on the extensive and eties from national research programs have been available data.6 Other countries represented by made available for international testing and large numbers of varieties in the data set evaluation. Participating countries have gained include Korea. with 106 varieties, and the com- access to promising varieties, and in some cases, bined countries of Latin America (239 varieties) they have been able to import them directly and Africa (101 varieties). Japanese varieties from the INGER nurseries. were not induded in this analysis. INGER itself keeps a complete and accurate The data indicate that numbers of released set of data on varietal importing that has varieties rose steadily during the 1970s but have occurred through its programs. INGER has stabilized over the past fifteen years. In some documented more than 300 instances of vari- countries and regions, however, such as Latin eties imported after appearing in INGER trials America, varietal release totals have climbed (Sehsu, personal communication, 1992). Our markedly in the most recent period. study lacks complete data on varietal releases in participant countries, especially in Africa and Internationalflows ofgenetic resources Latin America. Nonetheless, this study was able Table 3.11 reports measures of international to identify nearly 200 instances in which vari- flows of genetic resources associated with the eties could have been imported through released varieties and the parents of the released INGER7 In particular, INGER has played a varieties. Of the 1,709 released varieties, 390 significant role in disseminating IRRI lines. For Table 3.10 Numbers of varieties Included in the data set, by country of release and time period of release Country/Region Pre-1965 1966-70 1971-75 1976-80 1981-85 1986-91 Total Africa 3 7 8 17 26 40 101 Bangladesh 1 7 8 11 4 3 34 Burma 0 4 6 21 37 8 76 China 0 1 8 30 31 12 82 India 10 67 136 139 125 166 643 Indonesia 1 2 5 21 10 9 48 Korea 0 5 11 35 40 15 106 Latin America 7 9 48 32 43 100 239 Nepal 0 0 1 10 4 2 17 Oceania 0 1 4 1 0 0 6 Pakistan 0 4 2 3 3 0 12 Philippines 3 4 13 23 8 2 53 Sr Lanka 3 14 4 8 21 3 53 Taiwan 0 3 0 3 0 0 6 Thailand 1 2 4 8 5 3 23 USA 2 5 18 17 3 6 51 Vietnam 0 16 6 16 16 5 59 Other S.E. Asia 2 1 8 7 6 5 29 Other 0 7 15 15 15 19 71 Total 33 159 305 417 397 400 1709 Table 3.11 International genetic resources flows by time period Country/Region Pre-1965 1966-70 1971-75 1976-80 1981-85 1986-91 Total I. Released varieties, percent based on: IRRI cross 3 25 19 22 18 12 17 Other NAR cross 16 7 6 6 6 5 6 Own NAR cross 81 68 75 72 76 83 77 II. Parents of released varieties, percent based on: IRRI cross 0 24 29 33 23 19 24 Other NAR cross 27 25 21 15 18 20 18 Own NAR cross 73 51 50 52 59 61 58 Table 3.12 Matrix of varietal borrowing (numbers in parentheses represent borrowings through INGER) Latin SE. Sr Oter Amerka Oceania Banladesh Afica Dunma USA China India Indonesia IRRI Korea Asia Nepal Pakistan Philippines Lanka Taiwan Thailand Vietnam Total Other 9 7 0 0 0 1 0 6 16 3 18 6 5 0 0 0 0 1 0 0 71 (2) (2) (9) (13) LatinAAmerica 3 185 0 0 1 0 5 0 2 0 39 1 0 0 0 2 1 0 0 0 239 (1) (2) (1) (2) (15) (1) (22) Oceania 0 0 1 0 0 0 0 0 0 0 5 0 0 0 0 0 0 0 0 0 6 (0) Bangladesh 0 0 0 17 0 0 0 1 4 0 11 0 1 0 0 0 0 0 0 0 34 (1) (3) (4) Africa 1 1 0 0 69 0 0 0 1 0 26 0 0 0 0 1 2 0 0 0 101 (1) (1) (1) (12) (2) (17) Burma 2 0 0 1 0 33 0 1 1 1 18 0 2 0 0 1 2 0 12 2 76 (1) (1) (1) (5) (1) (2) (6) (2) (19) USA 0 1 0 0 0 0 48 0 0 0 2 0 0 0 0 0 0 0 0 0 51 (2) (2) China 0 0 0 0 0 0 0 66 1 0 13 1 0 0 0 0 1 0 0 0 82 (8) (1) (9) IndIa 5 0 0 0 1 0 0 0 573 0 53 0 1 0 1 1 4 2 1 1 643 (33) (3) (1) (37) Indonesia 0 0 0 0 0 0 0 0 0 29 18 0 0 0 0 1 0 0 0 0 48 (13) (13) SE Asia 0 0 0 0 0 0 0 0 1 0 7 0 21 0 0 0 0 0 0 0 29 (2) (2) Korea 0 0 0 0 0 0 0 0 0 0 1 105 0 0 0 0 0 0 0 0 106 (0) Nepal 0 0 0 0 0 0 0 0 0 0 5 0 1 8 0 0 1 0 0 0 17 (2) (1) (3) Pakisan 0 0 0 0 0 0 0 0 0 0 7 0 0 0 5 0 0 0 0 0 12 (0) Philippines 0 0 0 0 0 0 0 0 0 0 25 0 0 0 0 26 0 0 0 1 53 (15) (15) Sri Lanka 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 51 0 0 0 53 1 (0) Taiwan 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0' 0 0 6 0 0 6 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~(0) Thailand 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 23 0 23 (0) VIetnam 0 0 0 0 0 0 0 0 0 0 44 0 0 0 0 0 0 0 0 15 59 (27) (27) Tota 20 194 1 18 71 34 53 73 601 34 294 113 31 8 6 32 62 9 36 19 1,709 (1) (2) (0) (0) (3) (4) (3) (146) (0) (1) (0) (0) (0) (10) (0) (7) (2) (183) 94 GENETIC RESOURCES, INTERNATIONAL ORGANIZATIONS, AND RICE VARIETAL IMPROVEMENT varieties developed at IRRI and released by al flow of germplasm is extraordinary. No coun- national programs, INGER was the apparent try in the data set has failed to take advantage conduit in half of the cases, all of them in the of unimproved or improved germplasm from period, 1976-9 1. other countries. Since 1976, INGER has also become the primary channel through which nationally Genealogicalcomplexity -developed varieties have been transferred from One outcome of the increased flows of vari- one country to another. Since 1976, 37 nation- eties across national borders has been the al program varieties have been imported increasing complexity of the genealogies of rice through INGER During the same period, the varieties. Whereas a released variety in the number of national program varieties imported 1960s might have only three landraces in its through other avenues has diminished from 13 genealogy, more recent releases have 25 or more in 1976-80 to 6 in 1986-91. INGER has landrace ancestors. In a sense, this increasing played an important role in facilitating the pedigree complexity is a measure of breeding transfer of varieties across geographic zones; for inputs in the development of new varieties. instance, both of two Sri Lankan varieties Pedigree complexity can most easily be mea- released in Africa came through INGER, and sured as the number of landraces, pureline both of two Indian varieties released in Latin selections, and mutants that are ultimate ances- America camne through INGER. tors of a released variety. For released varieties Perhaps more remarkable than the direct containing more than one such ancestor, at international flows of varieties has been the least one cross must have been made by breed- international flows of parents of the varieties. ers corresponding to every two ultimate ances- Nearly three-quarters of the varieties in the data tors. Thus, varieties with large number of ances- set (1,263) have at least one imported parent. tors - and hence complex pedigrees - are the Including imported varieties, 810 releases (47 product of intensive breeding efforts... percent) have at least one parent from IRRI, Moreover, the share of ancestors delivered and 619 (36 percent) have at least one parent through IRRI lines has increased over time. For from another national program. Excluding varieties released in 1965-74, IRRI provided 54 imported varieties, more than 500 varieties have percent of ancestors, by number. For varieties at least one parent from IRRI. Excluding both released in 1981-90, IRRI delivered 72 percent imported varieties and those with IRRI parents, of ancestors, by number. Thus, the usefulness of more than 350 released varieties have at least IRRI's breeding lines has not diminished fol- one parent from another national program. lowing the early release of semi-dwarf rices. This indicates that parental importing is taking Instead, IRRI continues to offer national breed- place across national programs on a large scale. ing programs highly useful packages of genetic (Table 3.13 reports country details for material. These packages form the nucleus of parents.)8 contemporary breeding programs in most The extent of international exchange- countries. both of varieties and of parents - implies that National rice improvement programs have a large majority of the varieties in the data set depended to differing extents on IRRI lines as were developed using breeding lines from out- sources of genetic materials. Some countries side the country of release. In fact, only 145 have borrowed many of their released varieties varieties out of 1,709 (8.5 percent) were devel- or parent lines from IRRI, while others have oped entirely from own-country parents, grand- used IRRI materials in conjunction with local parents and other ancestors. Most of these were varieties or other internationally available simple varieties with fewer than four landraces breeding lines. For example, Vietnam and in their pedigree. The extent of this internation- Pakistan have based their modern varieties Table 3.13 Matrix of parental borrowing (numbers in parentheses represent borrowings through INGER) Latin S.E. Sn Other Amenca Oceania Bangladesh Aftica Burma USA China India Indonesia IRRI Korea Asia Nepal Pakistan Philippines Lanka Taiwan Thailand Vietnam 7otal Other 28 2 0 0 0 1 1 0 42 4 48 1 7 0 0 4 0 2 2 0 142 (2) (23) (13) (3) (1) (42) Latin America 116 73 0 0 4 0 22 0 62 3 161 4 0 0 0 8 6 5 10 4 478 (26) (20) (3) (3) (58) (3) (68) (3) (2) (3) (3) (192) Oceania 0 0 0 0 2 0 2 0 5 0 5 0 0 0 0 0 0 1 2 0 12 (0) Bangladesh 4 0 0 1 0 0 2 1 18 2 34 0 2 0 0 2 0 1 0 1 68 (1) (1) (t 1) (9) (1) (1) (24) Afrca 33 3 0 0 34 0 0 2 57 6 42 0 6 0 0 3 3 5 5 3 202 (20) (2) (13) (2) (51) (4) (24) (3) (3) (1) (3) (126) Burma 26 0 0 0 0 1 0 1 68 5 38 0 1 0 0 7 0 2 0 3 152 (15) (68) (1) (29) (5) (3) (121) USA 28 2 0 0 0 0 47 0 15 0 6 0 0 0 0 0 1 1 2 0 102 (1) (15) (4) (2) (22) China 42 0 0 0 0 0 0 20 27 1 72 0 0 0 0 0 1 1 0 0 164 (13) (7) (27) (2) (52) (101) India 150 1 0 3 4 0 2 10 637 14 351 0 23 0 0 15 10 57 7 0 1,284 (32) (2) (4) (1) (3) (101) (7) (174) (13) (6) (4) (14) (4) (365) Indonesia I1 0 0 0 0 0 0 1 22 22 35 0 1 0 0 2 0 0 2 0 96 (3) (12) (1) (14) (30) SE Asia 0 0 0 0 0 0 0 1 9 3 17 0 9 0 0 3 0 2 2 0 46 (5) (7) (1) (2) (15) Korea 0 0 0 0 0 0 0 0 3 0 49 74 0 0 0 0 0 0 0 0 126 (3) (15) (8) (26) Nepal 9 0 0 1 1 0 0 0 7 1 10 0 2 0 0 1 2 0 0 0 34 (4) (4) (2) (10) Pakistan 1 0 0 0 0 0 0 2 6 0 12 0 0 0 0 1 0 0 2 0 24 (3) (4) (7) Philippines 20 0 0 0 0 0 4 0 11 5 45 0 2 0 0 18 0 0 0 0 105 (2) (1) (13) (2) (18) Sri Lanka 24 0 0 0 0 0 1 1 4 3 22 0 0 0 0 0 50 1 0 0 106 (3) (1) (1) (3) (1) (4) (5) (18) Taiwan 2 0 0 0 0 0 0 0 3 0 5 0 0 0 0 0 0 2 0 0 12 (2) (2) Thalland 12 0 0 0 0 0 0 0 12 2 7 0 0 0 0 1 0 1 11 0 46 (5) - (1) (5) (1) (22) Vietnam I1 0 0 0 0 0 0 3 34 3 63 0 1 0 0 1 0 0 0 2 118 (1) (22) (23) (46) Total 517 81 0 5 45 2 81 42 1,042 74 1,017 79 54 0 0 66 73 81 45 13 3,317 (121) (22) (3) (20) (6) (13) (421) (20) (462) (8) (17) (18) (17) (19) (10) (10) (1,187) 96 GENErIC RESOURCES, INTERNATIONAL ORGANIZATIONS, AND RICE VARIETAL IMPROVEMENT almost completely on IRRI lines, but Sri Lanka varieties, is represented only by the Gsv gene used a large pool of other breeding lines as for resistance to grassy stunt virus, biotype 1 sources of germplasm. (Chang and Li 1991). The Indian varieties Ptb 21 and Ptb 18, each of which is an ancestor to Landracepools more than 100 varieties, are both known to Several concepts of germplasm "pools" are have resistance to brown planthoppers, tungro, relevant for our purposes. We can identify all and green leafhoppers (Swaminathan 1984, the landrace ancestors of a released variety. cited in Plucknett et al 1987). Alternatively, we can identify all the released By and large, national program breeders varieties that are descendants of a particular lan- have sought to incorporate the useful genes drace. To use the first approach, the 1,709 vari- from these varieties (and others like them) into eties in the data set trace to a total of 11,592 local varieties, which are well adapted to envi- ancestors. Of these, 8,208 have come through ronmental stresses and ecologies, and which sat- IRRI lines and 3,384 have been acquired inde- isfy local tastes and preferences. In some coun- pendently of IRRI lines.9 tries, breeders have worked directly with these Following the second approach, however, we landraces and pureline selections. More often, find 838 distinct landraces that are ancestors of however, national programs have taken advan- the 1,709 released varieties. Many of these tage of work performed by IRRI breeders in ancestors occur repeatedly in the genealogies of "packaging" these useful traits in improved released varieties. Dee Geo Woo Gen, for exam- lines. This continues to be true even in recent pie, is an ancestor for 1,105 varieties in the data years, despite expanded international access to set; Cina and Latisail each appear in more than the original unimproved varieties. 1,000 varieties. These three landrace selections Evidence for the role of IRRI in packaging are the parents and grandparents of IR-8, and advantageous germplasm can be found in the thus have appeared jointly as ancestors to nearly frequency with which IRRI has been the source 700 varieties in the data set. of the most commonly used ancestors. The 17 In addition to DGWG, Cina, and Latisail, most commonly used ancestors have appeared 14 other varieties have appeared as ancestors of independent of IRRI lines in fewer than 15 more than 100 varieties. These 17 varieties percent of their accumulated occurrences. For together account for 8,439 of the 11,592 accu- example, Oryza nivara has appeared in the mulated ancestors. The frequency with which genealogies of released varieties exclusively via these varieties appear as ultimate ancestors of IRRI lines. Similarly, the Thai variety Gam Pai releases indicates the enormous value of the 15 appears as an ancestor for 139 varieties, but genes they contain. only two varieties (both Thai) have obtained Although a relatively small number of lan- Gam Pai 15 independent of IRRI. A variety draces accounts for such a large amount of the from Andhra Pradesh, India, occurs in the genetic material embodied in the released vari- genealogies of 379 released varieties, but only eties, it would be incorrect to assume that this five releases have made direct use of the figure implies genetic uniformity. In many ancestor. cases, these ancestors appear far back in the Thus, IRRI has played a major role in iden- genealogies of released varieties, so that they tifying sources of advantageous genes and in may be contributing only small amounts of packaging these genes in forms that can be genetic material, or even single genes. readily used by national programs. IRRI con- For example, DGWG probably contributes tinues to function in the same role today, work- little more to most varieties than the ing extensively with unimproved materials. semi-dwarfing gene sd 1. Similarly, Oryza Curiously, however, relatively few additional nivara, which appears as an ancestor of 244 materials have entered the ancestor pool EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 97 through IRRI's efforts since the mid-1970s. The data on landrace pools and on the For each of the 838 ancestors of our released importance of certain individual ancestors has- varieties, our study has identified all of its implications for the management of global released progeny. and has noted the earliest germplasm collections. Some have suggested release date for these progeny... that germplasm collection need not be exhaus- The data indicate that only 80 ancestors tive, since a relatively smaUl sample of rice vari- appeared for the first time through IRRI mate- eties will contain, statistically, almost all the rials. An additional 47 appeared in released alleles in the rice genome. It is clear that "core" national varieties before 1965, and were thus collections offer a useful approach to the man- part of the international varietal pool before agement and handling of germplasm collections IRRI was established. Since 1970, however, 687 (Hodgkin 1991). However, most experts appear ancestors have been added to the ancestral pool to believe that comprehensive collection serves - but only 41 have been added through IRRI an important purpose (Chang 1989). Our lines. research supports this principle. A landrace or This suggests that national programs have wild species containing a single useful gene, continued to make use of IRRI lines that pack- such as 0. nivara, can provide economically age material already used in released varieties. useful traits to hundreds of released varieties. In some cases, these IRRI lines may be re-pack- Although it is true that a relatively small aging IRRI's own old materials; in other cases, sample of the total population of rice landraces IRRI is re-packaging lines used in one country would contain virtually all the genes in the in a form suitable for another country. species, it is equally true that genes for specific However, it must be noted that IRRI's focus on traits are often found only in minuscule semi-dwarf plant types has limited its success in numbers of landraces. This is true because culti- providing a broader landrace base for modern vated rices are self-pollinated, with relatively rice varieties.1I low genetic variability within most localities. Many of the ancestors introduced by However, rice is grown across enough environ- national programs are landraces that are locally ments that significant genetic differentiation adapted to certain micro-environments. Of the can be found across landrace populations of 0. 606 landraces introduced by national programs sativa (Oka 1991 a). Upland rices are even more since 1970, 414 have appeared in the pedigrees polymorphic genetically than are lowland rices, of just one released variety. In effect, then, and wild species are far more polymorphic than national programs have been using IRRI pack- either upland or lowland varieties of Oryza ages of germplasm to upgrade the quality of sativa (Oka 1991a) . local and traditional varieties. Empirical data suggest that wild species, The total pool of ancestors incorporated in isolated landraces, and other such "fringe" released varieties appears to have grown by 25 materials are important sources of resistance to to .50 landrace and pureline ancestors in almost diseases and pests of cultivated rice. For every year since 1970. This suggests that as the example, Oka notes that Chang (1989) available stock of improved varieties expands, reported finding resistance to white-backed farmers have more choices, and can choose planthoppers in only 0.8 percent of the 48,544 locally adapted high-yielding varieties. The varieties of Oryza sativa screened at IRRI, improvements of the original high-yielding whereas 45.4 percent of the 681 varieties of varieties are no longer limited to favorable Oryza glaberrima displayed resistance, as well agro-ecological zones in major rice producing as 46.2 percent of the 437 wild Oryza species countries; instead, an increasing number of tested. Oka concludes that in rice, "wild rela- released varieties is available for widely varying tives and primitive landraces are important production environments, objectives in germplasm collecting and conser- Table 3.14 Routes of varietal release - descriptive statistics Average number Average number of of landraces landraces independent of IRRI Number Percent Percent of Average number of landraces of vareties of varieties Total area area Pre 1976 Post 1976 Pre 1976 Post 1976 with rare trait index a >5.0 IRRIAnger 146 8.5 5177 13.3 n.a 13.2 n.a 0.0 12.55 IRRI/no inger 148 8.7 3959 10.2 5.4 12.4 0.0 0.0 7.66 Otherringer 37 2.2 411 1.1 n.a 4.2 n.a 2.1 3.35 Other/no inger 59 3.5 2954 7.6 4.4 5.2 2.5 1.6 4.14 IRRI parentringer 214 12.5 6570 16.9 n.a 10.4 n.a. 1.2 9.55 IRRI parent/no inger 313 18.3 5589 14.4 5.6 9.5 1.7 1.4 6.53 Other parentringer 208 12.2 4283 11.0 n.a 2.9 n.a 2.5 1.52 Other parent/no inger 151 8.8 3228 8.3 3.4 4.8 3.4 3.8 2.68 IRRI GparenVinger 14 0.8 670 1.7 0.0 7.2 0.0 3.0 6.00 IRRI Gparent/no inger 94 5.5 1436 3.7 7.4 10.7 4.6 3.6 8.93 Other GparentVinger 0 0.0 0 0.0 0.0 0.0 0.0 0.0 0.00 Other Gparent/no inger 180 10.5 1482 3.8 4.4 4.1 4.3 3.8 2.04 Pure national 145 8.5 3121 8.0 3.2 2.6 2.7 2.2 1.10 a. The rare trait index for a landrace is calculated as the ratio of the number of varieties in the data set for which the landrace appears somewhere in their geneologies to the number of varieties for which the landrace is a parent. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 99 vation because they are expected to be ancestor materials are used for very specific treasure-houses of useful genes." traits. They may have been identified through Vaughan (1991) reports that more than special searches ofthe germplasm collection, or 40,000 accessions in the IRGC have been they may be widely recognized for possessing screened for resistance to blast, with compara- particular characteristics. These varieties are ble numbers for bacterial blight, green leafhop- desirable only for one attribute; they may pos- per, and brown planthopper. Vaughan notes sess other undesirable attributes. These lan- that for blast and green leafhopper, resistance is draces or wild species may occur in the pedi- evenly distributed across the germplasm collec- grees of numerous varieties, but they are used tion. For brown planthopper and bacterial only infrequently as parents. 0. nivara, for blight, however, resistance is concentrated in example, has been used solely as a source of materials from given geographic regions and resistance to grassy stunt. Its other characteris- strata of the collection. Some sources of resis- tics make it an undesirable parent. Thus, 0. tance have come from unlikely landraces; for nivara has appeared seven times as a parent in example, Vaughan cites the discovery of cold our data set, but it appears in the genealogies of tolerance genes in an Indonesian variety 244 released varieties. The resulting ratio, (Silewah) from 1,200 m elevation in Sumatra. 244:7, provides an index of its 'rare trait" char- Similarly, a variety from the dry zone of Sri acteristics. . Lanka has been used as a source of genes for flood tolerance. Routes (pathways) from origin to release Saxena (1991) notes that researchers first In order to analyze more formally the identified resistance to brown planthopper in impacts of IRGC, IRPB, and INGER, it is use- the 1960s, with IR26 in 1973 the first ful to trace the routes by which varieties were semi-dwarf variety to show resistance. The sin- released. . . These routes are defined to be gle gene responsible for resistance, Bph-l, sup- mutually exclusive categories, so that each vari- pressed the original biotype of brown planthop- ety in the data set falls into exactly one of the pers. Evolution of new planthopper biotypes following categories: forced researchers to seek a succession of new resistance genes. however. To date, seven resis- Borrowed varieties tance genes have been identified, and resistance 1. IRRI line, borrowed through INGER has been transferred from 0. officinalis to 0. (IRRI/INGER) sativa. 2. IRRI line, borrowed independent of Similarly, Ikeda (1991) reports that resis- INGER (IRRIINO INGER) tance to six strains of bacterial blight has been 3.NGe fro N inG pr) 3. Variety from another national program, bor- found in varying levels in rice varieties from rowed through INGER (OTHER Asian countries. The frequency of resistance to different races of bacterial blight was different NATL/INGER) acrosscountries; for exa , v s wh t 4. Variety from another national program, bor- across countries; for example, varieties with the roe ineedn of INE (TE Xa-3 resistance gene were found in almost all NATL/NO INGER) countries, but with a frequency of 17.2 percent in Indonesia and only 0.3 percent in India. Nationally developed varieties, borrowedparents All of these examples suggest that large Ationast ope varetIes , borrowed germplasm collections are important economi- 5. At least one parent from IRRI, borrowed cally as sources of desirable genes. Evidence to through INGER (IRRI PARENT/INGER) support the importance of large germplasm col- 6. At least one parent from IRRI, borrowed lections can be found in the genealogy data independent of INGER (IRRI analysis of released varieties. Some landrace and PARENT/NO INGER) 100 GENEric RESOURCES, INTERNATIONAL ORGANIZATIONS, AND RICE VARIETAL IMPROVEMENT 7. No IRRI parents, but at least one parent The importance of INGER can be seen by borrowed from another national pro- looking at the time trends on borrowing gram via INGER (OTHER NATL PAR- through INGER Since 1981, more than half of ENT/INGER) released varieties (440 out of 797) have either 8. No IRRI parents, but a least one parent bor- been borrowed through INGER or were bred rowed from another national program inde- from parents borrowed through INGER" II pendent of INGER (OTHER NATL PAR- Table 3.14 reports numbers of landraces and ENT/NO INGER) proportions of rare traits by route and by region. This table shows that IRRI material has Nationally developed varieties andparents, been the conveyor of high landrace content and borrowed grandparents high rare trait content. Data on area planted are 9. At least one grandparent from IRRI, bor- also reported... rowed through INGER (IRRI/GPAR- ENT/INGER) Economic implications 10. At least one grandparent from IRRI, bor- ... Evenson and David (1993) report esti- rowed independent of INGER (IRRI/ mates of modern variety impacts for India, GPARENT/NO INGER) Pakistan, Bangladesh. Philippines, Thailand, 11. No IRRI grandparent, but at least one Indonesia and Brazil. These range from a rela- grandparent borrowed from another tively high value for India to lower values -for national program via INGER (OTHER other countries. The approximate value of mod- GPARENT/INGER) ern varieties in 1990 (in US dollars) in Indica 12. No IRRI grandparents, but at least one rice regions was 3.5 billion dollars. If we con- grandparent borrowed from another sider this to be the cumulated contribution of national program independent of INGER the first 1,400 modern varieties, we obtain an (OTHER GPARENTINO INGER) average value of a released variety of 2.5 million dollars per year and this annual value continues Nationally developed varieties, parents, into perpetuity because we are considering vari- grandparents etal improvements to be additive. . . 13.All parents and grandparents from country Notes of release (PURE NATIONAL) Financial support was provided by the International Rice Research Institute (IRRI). Our thanks to IRRI scientists, In practice, virtually no varieties fell into administrators, and staff for their cooperation, and in par- categories 9 or 11, since INGER has not been ticular to Tom Hargrove and Victoria Cabanilla for use of in existence longenouhtoproidetheir data. Thanks also to M. Add Judd for her research in existence long enough to provide many assistance. We also gratefully acknowledge the comments grandparent materials. Moreover, many vari- of an anonymous (and energetic) reviewer. eties with borrowed grandparents also have bor- rowed parents oae orrwed van- 1. The term pedigree breeding is used by De Datta (1981). With wheat and sorghum, for instance, methods eties. .. of bulk breeding are generally preferred. For our purposes, Since 1970, only 7.8 percent of new vari- these methods make it more difficult to trace flows of spe- eties have been of "pure" national development. cific genetic material. The most significant channels of release have 2. Recent technological advances require us to broaden been the use of IRRI parents. Before 1975, our understanding of germplasm, however. Through tis- IRRI parents were obviously not channeled sue culture of various kinds, a number of types of plant tissue can be used to regenerate plants. In addition, some through INGER, but in recent years, the largest species have traditionally been regenerated from stems, single pathway for developing new varieties has shoots, tubers and cuttings, rather than for seed. been to use IRRI parents taken from INGER Germplasm in its broadest sense is thus any genetic mater- ial that can be used to reproduce adult plants. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 101 3. The study drew heavily on a number of data sets avail- not only in the total set of varieties, but even in a single able through IRRI. The first of these was a list of elite variety. lines and released varieties from more than forty countries. This data set, collected by V L Cabanilla and T R 10. In fact we regard IRRI's poor performance in con- Hargrove for the International Rice Geneology Database tributing to the landrace pool to be a disappointing part of provides information on the parentage and release dates of a program that otherwise has made great contributions. most indica rice varieties since 1968. An accompanying IRRI has introduced a new plant type in recent years and data set, containing more than 6,500 entries, contains this may ultimately allow a broader base for landrace con- breeding records that make it possible to trace complete or tributions by IRRI. It would appear, in retrospect, that partial genealogies for all the elite lines and released vari- IRRI was slow in moving to other plant types and less eties in the first data set. This data set is also based on concerned than it should have been with broadening the work by Cabanilla and Hargrove, although much expan- landrace pool. sion and modification was carried out for this study. These alterations transformed the two data sets into a I1. Note that grandparent varieties will have had a shorter united, self-contained, self-referencing data set. period to have been influenced by INGER because of the time lag between appearance in INGER and their ultimate 4. It was also possible to combine the varietal data with appearance as a grandparent. INGER, however, may have additional data sets from INGER. Two INGER data sets, had a large impact on these flows, even if they were NO from IRRI, were used. The firsr was a list of entries in INGER flows because it stimulated more international INGER since its inception; the second was a list of the searching for genetic resources. Similarly the IRRI lan- nurseries in which these entries were tested. By matching drace pool may also stimulate these flows by inducing the names of varieties to the list of INGER entries, it was national program efforts to complement IRRI materials. possible to infer the indusion of varieties and ancestors in (See Gollin and Evenson 1991.) INGER. 5. These 32 varieties were generally regarded to be early "modern" varieties. References Barker, Randolph and Robert W Herdt with Beth Rose. 6. Most released varieties have been planted on significant 1985. The Rice Economy of Asia. Washington DC: acreage. In India, the Indian Council of Agricultural Resources for the Future. Research reports yields of two leading rice varieties in a number of districts. More than 150 varieties have been Chang, T T. 1985. "Crop History and Genetic noted as important varieties in farmers' trials over the Conservation: Rice - A Case Study," lowa State 1977 to 1989 period. Journal of Research 59(4): 425-456. 7. The criterion used was whether varieties developed in _. 1989. "Domestication and Spread of the one country were released in another country two or more Cultivated Rices," in D R Harris and G C Hillman, years following their appearance in INGER. (Given the eds, Foraging and Farming - The Evolution of Plant omission in our data set of many countries in Africa and Exploitation. London: Unwin Hyman. Latin America, which have imported actively from INGER, the figures appear to be consistent with the data Chang, T T and Cheng-Chang Li. 1991. "Genetics and maintained by INGER.) Since typography and nomencla- Breeding," in B S Luh, ed, Rice: Production, vol 1 ture also make it difficult to match named varieties with (2nd ed). New York: Van Nostrand Reinhold. INGER entries, it is likely that imports through INGER have been undercounted, rather than overcounted, in our Dalrymple, Dana G. 1986. Development and Spread of study. High-Yielding Rice Varieties in Developing Countries. Washington DC: US Agency for International 8. As many as 422 of the varieties based on international- Development, Bureau for Science and Technology. ly exchanged parents may have been developed from materials chosen out of INGER, in the sense that the par- De Datta, Surajit K. 1981. Principes and Practices of Rice ent first appeared in INGER trials four years or more Production. New York: John Wiley & Sons. prior to the release of the variety. About half of the INGER parents were IRRI materials, and half were vari- Evenson, R E and C David. 1993. Adjustment and eties from national programs other than the one of even- Technology: The Case of Rice. Paris: OECD, tual release. Parents chosen out of INGER have steadily Development Center Series. grown to account for larger proportions of borrowing. By 1986-91, as many as 80 percent of parental selection may Evenson, R E and D Gollin. 1991. "Priority Setting for have taken place via INGER. Genetic Improvement Research," unpublished manu- script prepared for workshop on Rice Research 9. Note, however, that the total number of ancestors Prioritization. Los Banos: IRRI includes ancestors that have been replicated many times, 102 GENETIC RESOURCES, INTERNATIONAL ORGANIZATIONS, AND RICE VARIETAL IMPROVEMENT Genetic Evaluation and Utilization (GEU) Program. Hodgkin, T. 1991. "Improving Utilization of Plant 1985. Parentage of IRRI Crosses IRI-IR50,000. Los Genetic Resources Through Core Collections," in Banos: IRRI. Rice Germplasm: Collecting Preservation, Use, proceed- ings of third international workshop. Los Banos: IRRI. . 1990. Parentage of IRRI Crosses IR No 50,001-60,000, unpublished booklet. Los Banos: Oka, W 1. 1991. "Distribution of Genes in Rice IRRI. Populations," in Rice Germplasm: Collecting, Preservation, Use, proceedings of third international Hargrove, Thomas R and Victoria L Cabanilla. 1979. workshop. Los Banos: IRRI. "The Impact of Semidwarf Varieties on Asian Rice-Breeding Programs.' BioScience 29(12): Plucknett, Donald L, Nigel, J H Smith, J T Williams, and 731-735. N Murthi Anishetty. 1987. Gene Banks and the World's Food. Princeton: Princeton University Press. Hargrove, T R, V L Canbanilla, and W R Coffman. 1985. "Changes in Rice Breeding in 10 Asian Saxena, R C. 1991. "Brown Planthopper," in "Evaluating Countries. 1965-84: Diffusion of Genetic Materials, Germplasm for Resistance and Tolerance," in Rice Breeding Objectives, and Cytoplasm," IRRI Research Germnplasm: Collecting, Preservation, Use, proceedings of Paper Series, Number 11. Los Banos: IRRI. third international workshop. Los Banos: IRRI. Hargrove, Thomas R, W R Coffman, and Victoria L Vaughan, D A. 1991. "Gene Distribution in Germplasm Canbanilla. 1980. "Ancestry of Improved Cultivars of Collections," in Rice Germplasm: Collecting, Asian Rice." Crop Science, 20: 721-727. Preservation, Use, proceedings of third international workshop. Los Banos: IRRI. 4 Country Experiences with Reform Some developing countries have all along had more open seed regulatory systems than others, and some have reformed over time. Differences across countries and over time provide windows of opportunity to look at what happens when governments do not control variety intro- duction, require official seed certification, or otherwise interfere in private seed trade. This Section presents four country papers from the workshop. In India, variety registration and seed certification have been voluntary all along. Recent policy changes give -pri- vate companies access to germplasm from government research and ease barriers to entry for imported seed and foreign seed companies. The government of Turkey moved dramatically from blocking to welcoming private seed trade through reforms in the early 1980s. With reforms allowing private compa- nies to introduce new hybrids and varieties for many crops, farmers and consumers have realized large gains in income. In Peru, policy reforms at the end of the 1980s shifted seed production from parastatals to multiple new private companies. Decentralization of responsibility for regula- tion and certification to regional seed committees has brought an influx of small seed producers into the formal seed industry. In 1990, Bangladesh's Ministry of Agriculture, with authority from existing seed legislation, moved from 104 compulsory to voluntary variety registration for introduced new cultivars, including maize all but five crops (rice, wheat, jute, sugarcane, hybrids (field corn and sweet corn), sunflower and potatoes). Following deregulation, private hybrids, and vegetables. seed companies and NGOs have tested and Seed Industry Regulations in Relation to Seed Industry Development in India Pramod K Agrawal India is predominantly an agricultural coun- one of the world's largest programs in variety try. Agricultural policy is primarily aimed at self improvement. More than 5,000 plant breeders -sufficiency in major staple foods of cereals, are working at the agricultural universities and pulses, and oilseeds. Self-sufficiency has been ICAR institutes. The ICAR annual budget for obtained in cereals and oilseeds, but India still 1995-96 is Rs 5.6 billion (US$175 million). depends on imports of pulses. In general, agri- The modest budget (by international stan- cultural development focuses on food crops and dards) with a massive network of research infra- is not export oriented except for tea, coffee, structure has yielded a rich dividend in the spices, and tobacco. sense that India has developed from a food deficit country to an food exporting country. Plant breeding research and the development From the ICAR system more than 2,000 vari- of new varieties eties of different crops have been developed; of course many of them are no longer in cultiva- Public sector tion. In the last thirty years there has been a major reorganization of agricultural research in Private sector India. Research institutes in the central govern- In a survey conducted in 1988 by the author ment have been transferred from the adminis- it was revealed that about ten seed companies trative control of the Ministry of Agriculture to started plant breeding research during the Indian Council of Agricultural Research 1948-1970 and about twenty companies during (ICAR). Individual states have also transferred 1971-1985. Most of these companies had breed- responsibility for agricultural research to the ing programs for agricultural crops (Agrawal agricultural universities. 1991). In recent years, some private companies ICAR has organized "All India Coordinated have been paying increased attention to vegeta- Crop Improvement Projects" (first one in 1957) bles. The number of companies now engaged in to speed up agricultural progress in the country. breeding has increased to forty, out of which ten Today there are thirty-two coordinated crop are working on vegetables. Five to ten percent of improvement projects, fifteen research insti- their revenues are spent on R&D (Arora 1994). tutes, nine national research centres and five Thus plant breeding research in the private sec- project directorates. India has emerged as a tor is comparatively of new origin. leader in the field of varietal development with With the introduction of the new seed poli- cy in 1988 and liberalization of investment pol- icy in 1991 (maximum allowed equity partici- pation for foreign companies and investors Pramod K Agrawal is General Manager of increased from 40 to 51 percent) private seed Proagro-PGS India, Ltd, an Indian Seed companies in India have established contact Company. 106 SEED INDUSTRY REGULATIONS IN RELATION TO SEED INDUSTRY DEVELOPMENT IN INDIA with transnational companies for accessing lished biotechnology laboratories without gov- hybrid parent lines for seed production, mar- ernment support. An exception is Tata Energy keting, and formation of joint ventures. A few Research Centre (TERI), which got funds from transnational companies have opened sub- DBT to establish a tissue culture laboratory for sidiaries in India. rapid multiplication of forestry species. Production and marketing of hybrid agricul- PROAGRO Seed Company, a leading tural seeds (for example, sorghum, pearl millet, research-driven company in India, has formed a maize, sunflower, cotton, castor, paddy) and joint venture with Plant Genetic Systems (PGS) horticultural seeds (for example, tomato, egg of Gent, Belgium. The joint venture, plant, okra, watermelon, etc) have expanded Proagro-PGS India Ltd, has access to the pro- considerably. Private seed companies have prietary technology of PGS, namely: Seed Link developed at least 122 varieties: 55 of vegeta- TM or Barnase (male sterility gene), Barstar bles, 39 of millets, 13 of cotton, nine of (male fertility restorer gene), and Bt genes (for oilseeds, four of fodder, and two of pulses. introducing insect resistance). Proagro-PGS is About 70 percent of these varieties are hybrids creating a 21st century state of the art biotech- (Agrawal 1991). The number of varieties devel- nology laboratory near New Delhi. The joint oped by private sector companies is now esti- venture is expecting to commercialize the mated to be about 300. hybrid mustard (Brassicajuncea) and insect tol- In recent years hybrid paddy varieties have erant tomatoes in India within two to three been developed and marketed by both public years. and private sectors. Private sector hybrid paddy varieties reportedly have performed well in Seed production farmers' fields. Private companies have been more successful in paddy seed production. Public sector Cooperation between public and private seed Essentially the Indian seed industry consists companies and between private seed companies of ICAR institutes, state agricultural universi- and governmental research stations has never ties, and public and private seed companies. been optimal because of conflicting interests in Organized seed supply started with the estab- plant breeding research and seed testing proce- lishment of the National Seeds Corporation in dures. With the changing environment, public 1963. Subsequently thirteen state seed corpora- sector breeding stations should pay more atten- tions and nineteen state seed certification agen- tion to create advanced breeding lines with value cies were established under the National Seeds added (for example, disease resistance) or to Programme with a loan from the World Bank. basic research, results from which could be sold These corporations virtually conduct no to the private sector to support private R&D. research and depend on flows of new cultivars from agricultural universities and ICAR New technologies to support plant breeding Institutes. The potential of biotechnology in agricul- Some agricultural universities and state ture has been recognized both by the public and departments of agriculture are also producing private sectors. Department of Biotechnology and selling seeds, though on a very limited (DBT) and ICAR have played key roles, creat- scale. Their objective is primarily to popularize ing awareness of the potential role for biotech- the varieties and hybrids developed by state nology to support plant breeding and to speed universities and departments of agriculture. up varietal development programs. The DBT has funded the establishment of Departments Private sector of Biotechnology in many universities in India. Private seed companies can be divided into Also, a few private sector companies have estab- the following categories: EAsING BAIUERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 107 1. Companies conducting research, organizing Seed Regulations production, and marketing their proprietary products. The Seeds Act of 1966 The main objective of the Act, which 2. Companies organizing production and mar- became effective in October 1968, is to regu- keting cultivars developed by public sector late the quality of seeds offered for sale. Under organizations. this Act labelling of seeds sold for planting is 3. Companies that have no production base compulsory, while seed certification is volun- but ail i atary. Most of the horticultural and floricultural but are Involved in marketing only. A large seeds sold are labelled, while agricultural seeds number of companies fall in this category. sessl r aeld hearclua ed are certified. For seed certification, prior notifi- cation of the variety is essential, which entails An eroupstImatd 10 troup 20 compnises bong formal release of the variety by the Central to groups 1 and 2. Group 3 is comprised of Sub-Committee on Crop Standards, more than 10,000 distributors and dealers. N otificati on andR s as. Notification and Release of Varieties. The Seed supply process of release is quite elaborate and time- consuming. Over the years the quantity of commercial Most private companies enter their hybrids seed has increased (table 4.1). The annual and varieties in official trials and at the same turnover of seed business through the organized time start marketing the same material as sector is Rs 13 billion (US$ 420 million; table labelled seeds. Companies have their own net- 4.2). The bulk of this business is of agricultural works for testing in different agro-climatic seeds. Just over 3 percent by value is from zones. Thus good planting material reaches hybrid horticultural seeds (table 4.3). farmers sooner than would be the case if com- In volume terms, the Indian seed industry is panies waited for release and notification proce- dominated by wheat and paddy. In value terms, dures. however, almost 50 percent of gross sales are There are nineteen state seed certification hybrids of sunflower, cotton, and pearl millet agencies. There has been lack of trained man- (table 4.2). By rough estimate, public and pri- power for seed certification. Most of the per- vate sectors each account for about 50 percent sonnel in seed certification agencies are drawn of turnover. Most public sector business is in from departments of agriculture, do not possess open pollinated varieties, while the private sec- adequate knowledge of seed certification, and tor emphasizes hybrids. do not appreciate its intricacies. Frequent trans- fers have also hurt the quality of work. Table 4.1 Seed distribution of field crops of Most government officials recognize only Improved cultivars in India certified seed as produced under the Act, Year Quantity (in 1,000 otons) although certification is voluntary. 1980-81 250.1 The Seed Control Order, 1983 1984-85 484.6 In 1983, government passed the Seed 1988-89 502.2 Control Order to regulate the seed sale. This 1992-93 764.7 order brought seeds within the purview of the Sourcesr Agrawal (1991); National Conference on Seeds, Essential Commodity Act of 1955. As per the Agra, February 1993. Act, all seed dealers and storage points are to be licensed. Due to early legal challenge, the Seed Control Order became sub judice and was not Table 4.2 Seed requirement and availability for field crops, 1992-93 Crop Seed rate Total planted seed Commercial seed Commercial seed Seedprice Commercial seed (sown area in millions of ha) (kgha) (1,000 t) supply (1,000 t) as % of planted (rps/t) sales (millions of rps) Wheat (24) 100 2,400 176 7.3 7,000 1,230 Rice (42) 30 1,300 172 13 9,000 1,550 Sorghum (15) 12 170 38.9 (18.9 hy) 22 28,000 hy 529 (20.0 var) 14,000 var 280 Pearl millet (10) 4 42 18.4 44 50,000 920 Maize (6) 20 120 16.1 (6.1 hy) 14 20,000hy 122 (10.0 var) 10,000 var 100 Fodder sorghum 8.0 16,000 0.1 Pulses Chickpea (7.4) 75 560 16.5 3 Pigeon pea (3.6) 20 72 5.30 7 Lentil (1.2) 40 46 1.40 3 Black gram (3.4) 20 68 9.00 13 Green bean (3.4) 20 68 8.20 12 Peas (0.55) 60 2.5 2.50 8 Total 42.9 40,000 1,720 Groundnut (0.44) 150 1,200 82.5 7 3.3 Rape/must'd (5.7) 5 29 9.70 34 40,000 242 Soybean (2.4) 63 150 19.5 13 25,000 390 Sunflower (1.6) 10 16 9.70 59 20,000 1,360 Castor (0.81) 13 10 0.14 28 140,000 12.6 Linseed (1.2) 25 29 147 0.51 90,000 ha 4.4 30,000 Cotton (7.4) 120 150 22.4 (11.2 hy) 15 375,000 hy 4,200 (11.2 var) 25,000 var 280 Total (141) 6,500 12 12,900 EASING BARRIERs TO MOVEMeNT OF PlANr VARIETIES FOR AGRICULTURAL DEVELOPMENT 109 put into force for many years. In 1994, howev- New Industrial Policy, 1991 er, the Supreme Court upheld the legality of the Prior to July 1991, foreign investors were order, so that it became operative from July 1, allowed to have equity participation only up to 1994. 40 percent. From July 1991, the government The order establishes strict controls: dealers decided to allow foreign investors to establish and storage points must display stocks and equity participation up to 51 percent in priority prices daily; government inspectors are free to sectors, including the seed sectoL Non-resident enter and inspect premises where seed is stored Indians are allowed 100 percent participation. and to draw samples; inspectors are empowered Investors must apply for approval from the to stop the sale of seeds if they suspect any mal- Reserve Bank of India, which will review their practice. Therefore, the Seed Control Order is application without referring to other ministries contrary to the current liberalization of the and reply within a very reasonable period, that Indian economy. is, one to four weeks. Breeder' Seed Supply, 1986 Export-Import of Seeds, 1989 For many years, public sector institutions Import and Export of seeds are regulated by supplied breeder's seed to the National Seed the Plant, Fruits and Seeds Order, 1989, made Corporation (a parastatal) for production of under the Destructive Insects and Pests Act, foundation seed, which was sold to the private 1914. Within the Ministry of Agriculture, sector for production of certified seed. With Directorate of Plant Protection, Quarantine these arrangements, spread of improved publi- and Storage regulates import and export of cally-bred varieties into farming communities seeds. For commercial consignments the follow- was slow. Therefore, in a landmark decision, ing steps are required: government decided to sell breeder's seed of publically-bred self-pollinated varieties to pri- General vate seed companies having adequate facilities 1. Registration with Joint Chief Controller and manpower to produce foundation and cer- for Import and Export. tified seed. This decision helped to alleviate the 2 Registration with Reserve Bank of India. shortage of certified seeds and stimulated seed production. Export 1. Registration with Agricultural Processed New Seed Policy, 1988 Food Products Export Development The Indian government severely restricted Authority. seed import and export until 1988. In 2. Export order (from importer) with com- September 1988, the government announced a plete terms and conditions. New Seed Policy with the intention that "farm- 3. Application to seed certification agency ers should have access to the best material (for seed sample to be drawn from lots available," regardless of its origin. Import pro- for laboratory test). cedures were liberalized; most seeds and planti- 4. Application to Directorate of Plant ng materials are now listed as "freely Protection, Quarantine and Storage, importable" items. Also, foreign companies Ministry of Agriculture, for were encouraged to participate in the develop- Phytosanitary Certificate. ment of the seed industry. Now, seeds of many 5. Preparation of Export Documents, foreign companies are available, particularly which includes invoice, packing list, vegetable seeds. shipping bill, and exchange control dec- laration form (GR-Form) issued by Reserve Bank of India. 110 SEED INDUSTRY REGULATIONS IN RELATION TO SEED INDUSTRY DEVELOPMENT IN INDIA Table 4.3 Market analysis of hybrid vegetable seeds in India, 1993-94 Total value Total area Area under Hybnd seed Seed price (Millions of Crop (1,000s of hectares) hybrid (percent) marketed (kgs) (rupees/kg) rupees) bnnjal 447 4.0 2,250 6,000 13.5 cabbage 223 21 15,000 8,000 120.0 cauliflower 421 0.59 1,300 4,000 5.2 chili 551 0.45 800 16,000 12.8 gourd 389 1.2 9,200 1,300 12.0 melon 157 2.1 6,600 2,500 - 16.5 okra 347 0.77 27,000 600 16.2 tomato 454 24 17,000 15,000 255.0 total 79,000 450 Source: Revised from Seed Association of India. 6. Attestation of documents, from PHD * Canalised items (import only through canal- Chamber of Commerce, for certificate of ising agencies designated by the central origin. Government, ie, the State Trading Corporation of India, Ltd, and Hindustan Note: At present there is no seed testing Vegetable Oils corporation, Ltd): seeds of laboratory accredited by ISTA in India groundnuts, safflower, soybean, cotton, sun- to issue ISTA seed analysis certificates for flower, and rapeseed. export. Many seed items are restricted for export, so Imports that exports are permitted only under license. 1. Registration with the National Seeds Important ones are as follows: seeds of castor, Corporation. cotton (except grown under custom produc- 2. Application to the Ministry of tion), Egyption clover, fodder crops, green Agriculture for import permit (from the manure crops, jute, linseed, lucerne, mesta, Directorate of Plant Protection, onion, rubber, all oilseeds, and all pulses. Quarantine and Storage). 3. Purchase contract (with supplier). The Plant Varieties Act, 1993 4. Inspection and sampling of consignment Currently, no law provides for ownership of at the port of entry (Delhi, Bombay, varieties. The Plant Varieties Act (PVP), 1993, Madras, Calcutta, or Amritsar). is under consideration in Parliament. Consignments are released after 30 days, The Act proposes to provides protection to: the period required for grow-out tests. new varieties; extant varieties; and germplasm. Protection is provided for fifteen years from the Lists of items freely importable, canalised, date of registration. To qualify for protection, and restricted are as follows: the variety must be evaluated through an appropriate evaluation system (hopefully Freely importable: seeds of cereals, pulses, through an independent agency having no oilseeds, vegetables, flowers, fruits and interest in plant breeding research), and a pre- plants, tubers and bulbs of flowers, cuttings scribed quantity of seed must be deposited in and saplings. the National Gene Bank. EASING BARRIERs TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 111 Seeds deposited with the National Seed It is difficult to conceive that PVP legisla- Bank will be freely available to researchers for tion in the present form will stimulate plant developing new varieties. Also, if the govern- breeding research in the private sector. Farmers ment determines that domestic or international would have the right to produce and sell pro- markets for the variety are not being met by the tected material without any limit. Also, how current owner, the government may grant can breeders protect their material if seeds of license to any other person to produce and sell protected varieties or lines are to be deposited seed of the variety. with the National Bureau for Plant Genetic Seeds of protected varieties are not to be Resources and Seed Certification Agencies? It sold unless they are certified, which means that will be difficult for a breeder to realize a return the owner of the variety must also provide on investment if government decides to provide parental material to certification agencies. license to any person to multiply and market Farmers retain the right to save, use, seed of a protected variety, as the law provides. exchange, share, and sell seed grown from seed The industry feels that this legislation reflects of protected varieties. the compulsion of article 27, para (3)b, of the Variety owners or seed dealers are to provide Trade Related Intellectual Property Rights compensation to growers if a protected variety (TRIPS) Agreement rather than conviction to does not perform as advertised. provide protection and incentives to plant The National Gene Bank will manage inter- breeders. (The TRIPS segment of the World national germplasm exchange. The Bank will Trade Agreement requires all nations to enact arrange reciprocal exchange with countries or suitable legislation to protect innovations.) agencies that allow free access to germplasm. Finally, regulation to protect transgenic plants is For others, the Bank will negotiate terms and far from clear, as the issue is not mentioned in conditions for germplasm exchange. The Bank the draft Act. will allow unrestricted access to germplasm from international agricultural research centers. References The Act provides for establishing a national Agrawal, P K. 1991. "Seed Industry in India," in Ton authority for plant varieties protection, which Grossman, A Linnemann, and H Wierema, eds, Seed Industry Development in a North-South Perspective. will register varieties, maintain a list of regis- Wangeningen: Pudoc. tered varieties, evaluate materials for granting PVP, and to carry out related functions. Arora, R S. 1994. Personal discussion. The Impact of Turkey's 1980s Seed Regulatory Reform David Gisselquist and Carl Pray Beginning in the early 1980s Turkey pro- Agriculture (MARA)1 authority over seed pro- gressively eased regulations controlling private duction, domestic trade, imports, and exports. sector introduction of new agricultural technol- While this law provides the framework for con- ogy through inputs production and trade. This trol, implementation depends on subsidiary study describes Turkey's 1980s seed regulatory regulations and practices. reforms and estimates the impact of those Variety registration. With respect to variety reforms on variety movement, agricultural pro- registration (that is, listing of crop varieties for ductivity, and incomes. which seed may be sold), Article 4 of Law 308 Turkey lies the same distance from the gives the Minister of Agriculture authority to equator as Spain, northern China, Japan, decide which crops are subject to compulsory Kansas in the United States, and southern registration. MARA has over the years insisted Argentina. The country offers a wide range of on compulsory registration for essentially all climates that supports a varied agriculture-for crops of interest to commercial seed firms. example, along the Mediterranean coast, hot For those crops for which varieties must be dry summers and mild winters support citrus registered, Article 6 gives the Minister authority and other fruit production, while on the to decide whether or not to register each foreign Anatolian Plateau in the center and east of the variety after "adaptation tests in our country country, cold winters and limited annual pre- and after their economic and agricultural values cipitation favor grain and pastures. With its rel- have been determined..." Notably, the decision atively high latitude and wide range of climates, about whether or not to register a variety is not Turkey has been able over the years to adopt based on presence or absence of negative exter- plant varieties, animal breeds, and other agri- nalities, which are not even mentioned; rather, cultural technologies from the United States registration depends on economic performance. and Central and Western Europe as well as Sometime before 1980, MARA began to from Mediterranean and sub-tropical countries. issue "production permits" as a form of second- class registration. Production permits expire Seed reforms after a fixed term (often four years), and some Pre-1980s government controls production permits limit use of the seeds, such as to produce goods for export only. Turkeys seed legislation dates from 1963, Certification. Law 308 authorizes MARA to with Seed Law 308 giving Ministry of certify seed and to set quality standards. MARA exercised this authority by insisting on compul- sory government certification and quality tests David Gisselquist is a consultant with the World for all commercial seed. Bank's International Economics Department and Other Controls. With authority from Law is editor of this volume. Carl Pray teaches in the 308, MARA annually fixed seed prices for most Department ofAgricultural Economics, Rutgers major crops. In addition, Article 11 gives University 114 THE IMPACr OF TuREY's 1 980S SEED REGULATORY REFORM MARA authority to control seed imports and genetic advances in seed technology exports, without limiting controls to phytosani- available elsewhere in the world... The tary issues. Before 1980s reforms, MARA used time span for new variety introduction, this authority to limit private seed trade primar- including hybrids, and their approval by ily to vegetables. the key Variety Registration Committee, is a minimum of three years... Seed com- 1980s reforms welcome the private sector panies operating outside Turkey have Entering the 1980s, MARA limited variety been very reluctant to undergo the approvals for most field crops to those spon- lengthy time period required to obtain sored by government research agencies. Private commercial introduction of hybrids in companies were more active in vegetables, Turkey or to make available the local importing and producing seeds. However, even production of parent lines under the sys- here MARA controls held back new technology. tem of rigid price fixing and the mini- For example, through 1980 MARA had regis- mum margin structure of the industry tered (allowed) only one variety of hybrid (ICD 1982, vol 1, p 35). cucumber and two of hybrid tomato, and trade in these varieties was monopolized by domestic The ICD mission recommended that companies licensed as sole distributors. Most Turkey eliminate price controls on seeds and traded vegetable seed was from local open-polli- allow "free import ... by private organizations nated varieties. A number of sources report sig- for testing and marketing purposes, subject nificant smuggling of vegetable seeds-for only to phytosanitary import regulations" (ICD example, for many years all hybrid watermelon 1982, vol 1, p 75). Notably, allowing free seeds were smuggled from Cyprus. Local firms, import for marketing entails doing away with most prominently May Seed Company, pro- compulsory variety registration. And if seeds of duced seeds for open-pollinated vegetables. unregistered varieties are to be multiplied in- Along with seed smuggling, other evidence country and then sold, compulsory seed certifi- of failed seed policies included government cation would also have to go. inability to popularize hybrid maize and large The government acted quickly on these and expensive government seed agencies serving issues, though reforms fell short of dismantling no more than about 10 percent of planted area. compulsory variety registration and seed certifi- In the early 1980s, frustrated with these and cation. A 1983 ICD paper reports changes 'in other difficulties, officials redesigned policies to the body of regulations pertaining to the seed stimulate rapid expansion of Turkey's private industry," including the following (Grobman seed industry and to invite multinational seed 1983, pp 1-2): companies to participate. Iompanies 18 Turkeyiased t * private companies from October 1982 are In981Tud hel ptrniomt allowed to do their own performance tests Finance Corporation (IFC) to helpfor varieties they would like to introduce private seed production and trade. IFC in turnes t woults to trove asked the Industry Council for Development (companies report results to the govern- (lD2 to .dnirse akesoeaut ment, which continues to decide whether or (ICD)2 to identify seed markets, to evaluate notoapveacvriy) strategies for private participation in the seed industry, and to suggest policy changes (ICD c MARA's Registration Committee is pre- 1982, vol 1, p 2). An ICD team visiting Turkey pared to issue production permits for new in early 1982 reported: varieties "even without testing, provided that evidence is presented that such a variety The government variety testing program has been successfully grown in a similar eco- makes it extremely difficult to introduce logical region"; EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 115 * testing requirements for variety registration Impact on variety movement and seed are lowered from the previous minimum of industry three years to one to three years, depending The chain of causality through which seed on crop and type of seed (for example, one regulatory reform has an impact on agricultural year for hybrids, two years for vegetables); productivity and incomes goes through variety introduction and seed supply. This section * from December 1983, each seed firm is describes the impact of reforms on variety allowed to set its own seed prices, movement and on Turkeys seed industry. The next section looks at the consequent impact of As part of efforts to stimulate private ne vaite.n mrvd edspl npo domestic and foreign investment, MARA and . . ICD organized a National Seed Enterprise ductivity and incomes. Development Workshop in Istanbul in May Variety movement 1984, inviting private Turkish and foreign seed Wity rement companies and soliciting their views. Also, the With seed reforms from the early 1980s, the government relaxed controls on private foreign number of improved varieties allowed for sale investment and made credit available for seed (with either registration or production permits) companies. increased dramatically over the next several Licensing Seed Companies. The government years, as shown in table 4.4. For sunflower, for licenses seed companies in two categories, desig- example, the number of varieties allowed for sale nating stronger ones (having some land and sci- increased from three in 1982 to about thirty in entific staff) as research companies, which 1987. For soy beans, varieties increased over the allows them to do their own variety trials. Other same period from two to more than forty. seed companies must ask research companies or Many of these new varieties have been intro- government farms for help with variety trials. duced and tested at the initiative of private seed Plant Variety Protection. In 1994, the gov- companies. For private seed companies, identi- ernment passed a law establishing a system for fying and introducing new varieties is a com- plant variety protection (PVP), which will allow mon strategy to expand sales. With a good new companies to daim property rights in varieties. variety, a company can expand its market share With implementation of this law, Turkey joins in competition with other companies and can UPOV (International Convention for the also expand the market, coaxing farmers to shift Protection of New Varieties of Plants), a treaty from saved seed or to change crops. organization in the United Nations system that Even for crops that have had strong public sets standards for intellectual property rights in sector involvement in seed trade, 1980s regula- plant varieties. te tory reforms brought significant increases in Subsidies. government from 1985 numbers of varieties available, through public through 1994 has subsidized specific categories as well as private agencies. In wheat, for exam- of seeds (for example, hybrid maize, soybeans, ple, the government allowed sale of seeds of etc) in order to cut seed cost to farmers. The only twenty-one improved varieties in 1982; on government annually announces subsidies per average, the government approved more than kilogram (kg) seed, and then pays that amount five new wheat varieties per year from 1984 to to companies against receipts demonstrating 1994. In 1980, Agricultural State Enterprises sale to farmers. Subsidies have fallen precipi- (TIGEM) produced seeds for fourteen wheat tously- for example, subsidies on hybrid maize and four barley varieties; by 1994, TIGEM's seeds ranged from 12-38 percent during 1987- offerings had increased to forty-two varieties of 1990 fell to about 1 percent during 1992- wheat and nine of barley. 1993, and were eliminated in 1994. In sugar beets, KWS (partner with the gov- ernment in Pan Tohum) responded to competi- Table 4.4 Varieties Introduced by year and crop, 1982-94 New Varieties cf. New Harvested Vaneties vareties area In available a added to 1990 as of 1/6/82- 1/6/84- 1/6/85- 1/6/86- 1/6/87- 1/6/88- 1/8/90- 6/8/91- 21/7/92- OECD list in Crop (hectares) 31/5/82 31/5/84 b 31/5/85 b 31/5/86 b 31/5/87 b 31/5/88 b 31/5/90 b 5/1891 b 20//92 c 2/7/94 c 1993 d wheat 9,400,000 21 1 10 9 2 16 5 16 1 2 164 sunflower 715,000 3 5 8 10 6 7 20 11 7 0 127 conton 641,000 9 2 0 5 1 1 5 1 4 0 18 maize 406 e hybrid 155,000 24 35 24 9 22 22 48 23 1 1 other 360,000 20 1 0 0 4 5 0 0 0 0 sugarbeets 378,000 11 5 7 3 3 1 3 2 0 0 108 tobacco 320,000 31 0 0 6 1 0 2 0 0 0 potatoes 192,000 8 3 6 3 1 0 14 13 8 3 tomatoes 126,000 43 14 10 0 8 11 122 67 3 2 soy beans 74,000 2 27 0 12 4 11 7 9 0 2 69 cucumbers 16,000 1 2 0 2 4 5 57 50 3 0 Note: The major reform that allowed an increase in number of varieties available, as shown in the table, was agreement on the part of key MARA officials from about 1983 to grant production permits (ie, temporary registration) more expeditiously, and to sometimes waive in-country performance tests. a. Cumulative number of varieties available on 31/5/82 indudes varieties either registered or with unrestricted production permits on that date. b. For all other dates, the table shows number of new varieties introduced from the previous date given in the table; number of new varieties is calculated as the sum of new varieties intro- duced with new registration or unrestricted production permits; lists have been scrutinized to avoid counting varieties that were, for example, newly registered but had previously been avail- able with an unrestricted production permit. c. From about mid-1 991, MARA has stopped listing varieties with new unrestricted production permits in Resmi Gazete; hence, figures in the table presumably understate number of new varieties available from mid-1991. d. The last column lists numbers of new varieties added to OECD's List of Cultivars Eligible for Certification in 1993. Since the US does not restrict varieties, US farmers legally have access to all of these new varieties. Farmers in EU countries, which share a Common Catalogue, have access to a large share of these varieties. e. OECD gives no breakdown into hybrid and non-hybrid varieties. Sourers Harvested areas for field crops are from: SIS, 1990 Agricultural Structure and Production, p 4. Total area for cucumbers and tomatoes are from: SIS, General Agricultural Census 1991, p65. Available andnewvarieties forvarious years and periods are from: Resmi Gazete for 31 May 1982, 31 May 1984, 31 May 1985, 31 May 1986, 2July 1987, 31 May 1988, 3 May 1990, 5 August 1991, 20 July 1992, and 20 July 1994. Numbers of cultivars added to OECD's list are from: OECD, List of Cultivars Eligible for Certification 1993 (Paris: OECD, 1993). EASING BARRIERS TO MOVEMENT OF PLANT VARIErIES FOR AGRICULTURAL DEVELOPMENT 117 tion from several fully private competitors by sents twenty-nine companies, most of which selecting additional lines from their breeding are private. The seed industry includes another program that are specially suited for Turkey. fifty or so private companies that have not Even without variety protection, companies joined TEBD. have invested to identify, test, and introduce Along with increase in number of seed com- new varieties for non-hybrid crops, such as panies, private seed sales expanded to dominate wheat and soy beans. During interviews in commercial seed sales for several field crops and 1994, seed company managers reported only to take a share of sales for other crops, as shown one case of a company multiplying seed of a in table 4.5. Most of the increase in private sales variety introduced by another: TIGEM multi- has been for varieties that private companies plied seed of a soy bean variety that May and its have introduced, so that private sales correlate international partner Asgro had introduced. strongly with technology transfer. Seed company staff expect that 1994 passage of For wheat, despite subsidized comnpetition a law establishing PVP will stimulate more from state-owned enterprises, private seed com- effort to identify and introduce non-hybrid panies have introduced new varieties from varieties. CIMMYT and Europe, and more are in the pipeline, including a durum wheat from the Expansion ofprivate seed industry United States. Recognizing that farmers multi- Before the 1980s reforms, the few private ply wheat seed for their own use and sale to seed companies importing and producing seed neighbors, one seed company manager in Turkey focused on vegetables. In addition, described his plan to continuously introduce there were two semi-private seed companies: new varieties, staying ahead of diseases and Pan Tohum, a joint venture between the govern- offering new varieties with attractive features ment's Turkish Sugar Factories Company and (for example, durum wheat with low percent of KWS, a German firm; and BETA, for which the yellow bellies). With farmers planting a million dominant owners are sugar cooperatives, so that tons of wheat seed each year, a company gain- BETA is controlled by Pankobirlic, a govern- ing even a small share of the market could see ment-run apex cooperative. big sales. In 1994, private companies sold In 1980, the government directly managed wheat seed at about double TIGEM seed and or controlled most commercial seed production grain prices. With higher prices, private wheat and trade through cooperatives and state owned seed depends on other factors, such as quality enterprises, including Agricultural Supply and variety, to compete with public seed. Organization (TZDK) and TIGEM. For sugar For selected crops, table 4.6 shows seed pro- beet seeds, Pan Tohum had a monopoly for duction plans by type of firm in 1994. production and import and distributed seeds Numbers do not show actual market share through sugar cooperatives. because some firms do not follow their plan A 1985 World Bank paper reports nine new (for example, TIGEM is unlikely to more than seed companies established from the time seed double wheat seed production from 1993), the reforms began (Freiberg and Grobman 1985, p seed may not all be sold, and some of what is 280). That number increased rapidly as major sold may be exported. These are, however, the US and European-based seed multinationals only numbers available to show market share. along with companies from Israel, Taiwan, Subsidiaries of six foreign firms proposed to Thailand, Japan, Korea, and other countries produce about half of the hybrid maize and established a presence through joint ventures, sunflower seed. Most of the rest of the hybrid licenses, or subsidiaries. In 1986, Turkish seed maize and sunflower seed as well 98 percent of companies founded a Seed Industry the soy bean seed was to be produced by six Association (TEBD); TEBD in 1994 repre- Turkish firms that have joint ventures with for- 118 THE IMPACr OF TURKEY'S 1980s SEED REGULATORY REFORM Table 4.5 Commerciala seed production, 1980-93 (metric tons) Crop 1980 1985 1990 1993 Wheat private sector 0 < 1,000 7,000 10,000 public sector 49,000 188,000 124,000 103,000 Barley private sector 0 0 0 300 public sector 13,000 35,000 16,000 10,000 Hybrid sunflower private sector 0 80 2,600 3,500 public sector 0 10 30 40 OPVsunflower private sector 0 na na na public sector 2,500 na na na total 4,800 1,000 200 Hybrid maize private sector na 1,800 4,500 7,200 public sector na 300 200 200 total 900 Sugar beets private sector na 3,400 na 3,300 public sector na 0 na 3,300 Cotton private sector 0 0 500 300 public sector 35,000 27,000 30,000 31,000 Potatoes private sector na 800 4,200 2,200 public sector na 6,300 900 200 total 2,000 Chickpeas private sector 0 0 0 90 public sector 0 200 200 70 Soybeans private sector 0 800 3,200 3,600 public sector 0 1,100 600 200 Vegetables private sector na 300 600 700 public sector na 30 30 30 total 191 Note: Data have been rounded off to two significant figures for hundreds and above; for hundreds and below, data have been rounded to one significant figure. a. Seed produced by registered seed companies. Source: Various publications of TEBD (Turkish Seed Industry Association). eign firms. Unofficial estimates from the firms markets; for example, the leading maize seed we interviewed suggest that market share for company reports that its market share is being the largest firm in maize is about 30 percent eroded by smaller firms undercutting its price. and in sunflower about 25 to 30 percent. There The large number of companies - local appears to be strong competition in major seed and multinational - involved in the seeds sec- EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 119 Table 4.6 Seed production plans by type of firm, 1994 Type of firrn Maize Sunflower Soybean Wheat Subsidiaries of foreign firms metric tons 6,200 1,500 0 2,500 percent share 50 48 0 1 Joint ventures metric tons 5,500 1,500 3,100 0 percent share 45 49 98 0 Other Turkish private firms metric tons 400 0 0 6,800 percent share 3 0 0 2 Turkish public firms metric tons 310 110 70 275,000 percent share 3 3 2 - Total (metric tons) 12,400 3,100 3,200 284,000 Source: Calculated from MARA, Tohumluk Programi 1994 (Ankara: MARA, 1994). tor in Turkey means that the industry has been locally produced seed was available. During the competitive despite some tendency toward 1960s, imports of Mexican wheat seed played a over-regulation (that is, continuing compulsory similar role, facilitating rapid expansion of area variety registration and seed certification). planted to new varieties. A recent study of maize yields for fifty countries over twenty-five Seed imports and exports years found strong correlation between yield Reforms lowering barriers to varietal intro- increases and volume of seed imports (Pray and duction and to seed imports led to more Echeverria 1992, pp 372-373). imports for selected crops, such as maize and While continuing government efforts to dis- sunflower. Soon, however, local seed production courage seed imports no doubt have some expanded to take care of local demand and then influence, their impact can be over-emphasized. pushed into export markets as well. Figures 4.1 Since the early 1990s, Turkey has been import- and 4.2 show this progression for maize and ing seeds for roughly 1 percent of planted area. sunflower: from 1988 maize seed exports have This can be compared to seed imports into exceeded imports; for sunflowers, seed exports Thailand and Chile, which have more open have exceeded imports from 1990. Once seed policies, for roughly 1 to 3 percent of reforms allowed seed technology to enter, planted area in the mid-1980s (FAO 1987, p Turkey has been able to exploit its comparative 307; Pray 1990, p 195). Seeds are bulky and advantage in seeds based on good climate, sci- costly to move, and even reasonable phytosani- entific skills, and low cost labor. Table 4.7 tary rules add to import costs. Aside from some shows Turkey's exports and imports for all high-value vegetable seeds and seeds that seeds. Chile, another early reformer, shows the require special conditions for production, lower same progression from maize seed imports, to costs with in-country production sooner or local production, and then to exports. later tend to dominate decisions about where to Maize and sunflower seed imports in the produce seeds, so that imports give way to local early 1980s assisted introduction of new vari- production. eties, allowing farmers to adopt them before Figure 4.1 Hybrid maize seed production, imports, and exports, 1980-92 (1,000 metric tons) Thousand MT 10 8 6 4 2 0 1980 1982 1984 1986 1988 1990 1992 Production Imports Exports Source. Data from TEBD. Figure 4.2 Hybrid sunflower seed production, imports, and exports, 1980-92 (1,000 metric tons) 4 3 2 0 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 M Production E Import Export Source: Data from TEBD. EAsING BARRIERS TO MOVEMENT OF PLANT VARIETiES FOR AGRICULTURAL DEVELOPMENT 121 Table 4.7 Seed exports and imports, 1980-93 (in tons, unless otherwise noted) 1980-82 1983-85 1986-88 1989-91 1992-93 Crop average average average average average Comments Wheat & barley export 0 0 600 2,000 700 virtually all in the import 200 17,000 6,000 14,000 2,500 public sector Hybrid maize export 0 0 600 1,200 1,800 for mid-80s imports import 200 1,000 1,500 50 200 fomi-siprt Hybrid sunflower export 0 0 0 300 1,400 prvate sector import 0 500 1,400 300 0 Vegetables export 0 90 90 110 140 pnvate sector import 20 90 140 200 210 Potato import 10 3,500 6,700 7,100 2,100 Soy bean import 1,300 3,400 5,000 2,500 0 Cotton import 100 0 100 500 80 Others export 0 50 40 300 40 import 80 400 200 700 400 Total export volume vegetables 0 90 90 110 140 others 0 50 1,400 3,900 4,100 Total export value a 0 5,000 5,000 8,000 12,000 ($1,000) Total import volume vegetables 20 90 140 200 210 potatoes 10 3,500 6,700 7,100 2,100 others 1,900 22,000 14,000 18,000 3,200 a. Value per ton of seeds exported has been estimated at $50,000 for vegetables, $1,500 for hybrid maize and sunflower, and $250 for wheat and barley, and $500 for all other crops. Source. Volume of exports and imports from various TEBD publications. Values are estimated. Shrinkingpublic sector seed sales As private companies expand, pressure Seed reformers in the 1980s concentrated on mounts for government to limit state-owned building up the private sector rather than tear- enterprise seed production and unfair competi- ing down the public sector. With even reformers tion. During interviews in 1994, managers of afraid that private companies would not be some private seed companies vehemently interested in producing relatively low value seeds objected to state-owned enterprise sales of for open and self-pollinated crops such as wheat, hybrid maize seed. In the 1990s, the govern- barley, and fodder crops, government agencies, ment has been debating plans to privatize induding state-owned enterprises, continued to TZDK, which produces potato and other seeds, produce seed for wheat and other major crops. and also TIGEM, which produces wheat and In addition, the government continues to play a barley seeds. In addition, consistent losses on large role in seed trade through cooperatives. Agricultural Bank loans to cooperatives drain 122 THE IMPACT OF TURKEY'S 1980s SEED REGULATORY REFORM public funds, which pressures government to to maize times annual total fertilizer use (FAO reduce interference in inputs trade. reports that share of fertilizer going to maize However, seed industry sources expect the drops slightly from 1980 to 1988; we use 1980s government through TIGEM to continue to share throughout; see Ange 1994, p 10). For produce wheat and barley seeds in the medium- rainfall, we use national annual rainfall. term future. Also, KWS has just signed a ten- Table 4.8 reports results of this regression. year contract with Pan Tohum, which suggests The coefficient for percent of maize area plant- that the government plans to maintain a large ed to hybrids is positive and significant in all share of the market for sugar beet seeds as well. specifications. The trend variable is positive and Cotton seed production and trade continues to significant and lowers the coefficient for be dominated by state-managed cooperatives. hybrids. Fertilizer is not significant in the regressions except when the trend variable is Reform impact on net incomes: benefits and not included; over the period, national average costs fertilizer use and hence calculated use on maize Since many factors (for example, macroeco- grew slowly. nomic policies, weather, and agricultural pric- Figure 4.3 shows actual yield of maize and ing policies) affect aggregate crop production, it projected yield using estimated coefficients is difficult to estimate the impact of seed reform from specification two and actual values for fer- from aggregate crop data. However, the impact tilizer use, rainfall, and trend, but with zeros for of seed reforms has been uneven across crops, post-reform hybrids. The gap between actual with productivity and incomes from selected and projected yields was greatest in 1990 at crops showing dramatic movement. Hence, almost 2 tons per hectare (t/ha). Impact of although total benefits are difficult to estimate post-reform private hybrids on maize yields in from aggregate crop data, it is possible to esti- Turkey parallels a similar impact of imported mate partial benefits for specific crops. hybrids on maize yields in Greece during the 1970s (see figure 4.4). Virtually all maize in Partial benefits from seed regulatory reform Greece is irrigated, which largely explains why Maize. With a simple yield response func- national average yields for Greece exceed those aion, we estimate the impact of post-reform for Turkey. Figure 4.4 also shows lower maize tion we stimte te imact f pot-reorm yields in Syria, a country which limits variety hybrids on yields. Since Turkey is a small pro- introduction, and higher yields in EUvariet ducer and regularly imports maize, we assume troduction, and free i n EU of that changes in maize productivity and produc- varieties. tion have no impact on price. With some vreis abonhavenout inpuct ost wrice. Wthe cs - During 1990-92, annual gross value of addi- assumptions abouttional production due to post-reform hybrids late the impact of higher yields on farmer net was about ion due over hybrids incomes. was about $130 million ($255/ha over 515,000 incomodel maize yields in tons perhectareas hectares (has)). We did not find cost of produc- We model maize yields in tons per hectare as tinsreshongctdfeecsbtwn a fnctonof hybrids, fertilizer, and rainfall.3 tion surveys showing cost differences between afunction ofhbis etlzr n anal3 local and hybrid maize. We estimate change in Hybrid use is shown as percent of maize area sown to private hybrids (all of which came in net farm income per hectare due to hybrids as after and because of reform; we do not include the value of increased production due to pre-reform [public] hybrids, which covered hybrids less higher cost for hybrid seed and cost only 1 percent of maize area in 1992 [CIM- to harvest and dry increased production (see MYT 1994]). We also include a trend variable Table 4.9). With these costs, we estimate that to capture impact of transport improvements, post-reform hybrids boost annual net farm extension, etc. Annual fertilizer use on maize is incomes in 1990-92 by $ 97 million (see table calculated from 1980 share of fertilizer applied 4.9). Table 4.8 Maize yield response function, 1961-91 In % fertlha In fert/ha Adjusted Specification % hybrid hybrid (kgs) (kgs) Rainfall Trend R2 1. 3.40 - 0.011 - 2.44E-04 - 0.916 (0.474) (0.0019) (7.37E-04) 2. 2.888 - 1.38E-03 - 4.58E-04 5.34E-02 0.924 (0.522) (5.30E-03) (7.14E-04) (2.73E-02) 3. - 0.056 - 0.161 - - 0.868 (0.009) (0.031) Note: Numbers in parentheses are standard deviations for the variables. Sources Percent of maize are planted to private or post-reform hybrids was calculated from TEBD data. Fertilizer use if from FAO data on 1980 percent of fertilizer use on maize (Ange 1994, plO) multiplied by FAO annual data on national fertilizer use. Rainfall is from national annual data. Table 4.9 Calculating net benefit from hybrid maize, 1990-92 Before reform, After reform, Change due item 1980-82 1990-92 Total change to hybrids Average yield volume (metric ton/hectare) 2.18 4.13 1.95 1.43 a value at $178/ton ($/ha) 388 735 347 255 Average fertilizer use on maize volume (kg/ha) 87 120.8 33.8 0 value at 178/ton ($/ha) 15.5 21.5 6.0 0 Seed for representative hectare OPV seed volume at 37 kg/ha (kg/ha) 37 26 (11) (11) OPV seed value at $178/ton ($/ha) 6.6 4.6 (2) (2) hybrid seed volume at 28 kg/ha (kg/ha) 0 8.4 8.4 8.4 hybrid seed value at 2,970/ton ($Iha) 0 25.0 25 25 total seed value for representative hectare ($/ha) 6.6 29.6 23.0 23.0 Harvesting and drying costs, calculated as 1/6th value of production ($/ha) 65 123 58 43 Net gains ($/ha) - - 260 189 Total area planted to maize (hectares) 581,000 515,000 - - Total benefits for 1990-92 maize planted area ($1,000) $134,000,000 $97,000,000 a. Actual yield less projected yield, using estimated coefficients for national average fertilizer and weather. Source: Maize prices, fertilizer prices, and seed rates from: CIMMYT, 1993/94 World Maize Facts and Trend (Mexico: CIMMYT, 1994). Figure 4.3 Maize yields in the Mediterranean region, 1961-91 Mt/ha 12 10- _ 6 _-_ 4 - - 1961 1966 1971 1976 1981 1986 1991 - Spain Greece wF Italy Turkey Syria Source Data from USDA. Figure 4.4 Maize yield: actual and projected, 1961-91 (tonslhectare) 5 4 3 2 1961 1966 1971 1976 1981 1986 1991 ~~~ Actual -----Projected Source~. Actual from SIS, The Summary ofAgricultural Statistics (Ankara: SIS, various years). Projected using coefficients from Table 4.2. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 125 Vegetables for export. Vegetable exports is an With total vegetable exports increasing area where one would expect an impact from much more than $200 million from the early reforms allowing in new varieties as well as new 1980s (see table 4.10), and with evidence that specialty chemicals, particularly plant growth new varieties contributed to exports for particu- regulators, which are important for high value lar vegetables, we estimate that new technology horticulture production. The limited data coming in with 1980s regulatory reforms added which are available confirm an impact, but the at least $100 million to annual vegetable picture is incomplete (see Table 4.10). This sec- exports (about half for fresh and frozen and half tion is based on information from a 1994 FAO for canned vegetables) by the early 1-990s. All of and MARA study, Turkey: Horticulture Subsector this increase does not represent higher incomes, Review: Approaches to Market-Led Development, since exports are pulling resources away from Value of vegetable exports (including tomato other activities. Without trying to work concentrate and other canned vegetables) through a general equilibrium model, we esti- increased from roughly $100 million in 1979- mate that net incomes increase by one-third of 81 to near $400 million in the early 1990s. For the value of additional exports (from a rule of tomato concentrate, exports increased from thumb taken from the Economist). Hence, by nothing in the early 1970s to $89 million in the early 1990s, $33 million is a lower bound 1991. Exports of other canned vegetables dou- estimate of the impact of regulatory reforms on bled to nearly $100 million in two years from annual incomes from vegetable exports; farm- 1989 to 1991. Other vegetable exports increased ers, traders, and processors share the increase in from $60 million in 1981 to $192 million in income. 1991 before falling back somewhat in 1992. Seeds for export. Seed exports depend on Growth was not even across all vegetables, seed imports almost in the same way garment but was rather concentrated in a few. Among exports depend on imports of zippers and canned vegetables, "pickled cucumbers and arti- cloth. Before reforms, when companies could choke hearts packed in glass have become major not freely send seeds and other technology to export items in the last two years" (FAO and Turkey, they did not arrange seed production MARA 1994, annex 4, p 9). Among fresh and in Turkey for international markets. In other frozen vegetables: fresh potato exports went respects - climate, low-cost labor, skills - from $5 million in 1987 to $37 million in Turkey is an attractive country for seed pro- 1991; fresh cucumber exports increased from duction. With 1980s regulatory reforms, $1 million to $13 million over the same period; Turkey's seed exports have grown dramatically, and frozen vegetable exports increased from $7 but from a low base; volume and value are still million to $26 million in the same four years. low. The pattern of export growth suggests that We have data on volume but not value of regulatory reform was a factor. For one thing, seed exports. Table 4.7 reports volume and esti- fresh and frozen vegetable exports grew most mates value of seed exports with some assump- rapidly at the end of the period. Also, we can tions about seed prices. Annual total value of correlate export growth for particular products seed exports grows from zero in the early 1980s with introduction of new varieties. For exam- to an estimated $12 million in 1992-1993. ple, the increase in fresh and pickled cucumber Compared to other agricultural products and exports can be linked to reform, since only one exports, seed exports are a small matter. cucumber variety was allowed for sale in 1982 Assuming, as above, that net income gains are compared to more than 100 in 1992. Similarly, one-third increases in trade, the increase in net the number of potato varieties allowed incomes due to seed exports may be about $ 4 increased very rapidly from 1988, while potato million. exports jumped in 1991. Table 4.10 Vegetable production and export, 1979-1992 Crop and category 1979-81 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Tomato production (1,000 tons) 3,550 5,750 6,000 6,200 fresh export (1,000 tons) 42 78 34 107 fresh export ($US millions) 30 24 13 13 29 Tomato concentrate production (1,000 tons) 155 150 105 164 198 290 250 export (1,000 tons) 124 149 export (see "vegetable exports" below) Potato fresh export (1,000 tons) 44 48 40 22 219 fresh export ($US millions) 5 5 6 5 43 Onions production (1,000 tons) 1,071 1,360 1,550 1,700 fresh export (1,000 tons) 133 164 148 85 202 fresh export ($US millions) 12 15 17 13 33 Cucumbers production (1,000 tons) 503 fresh export (1,000 tons) 3 7 12 23 fresh export ($US millions) 1 4 7 13 Vegetable exports ($US millions) ca 100 a 379 of which: tomato concentrate 74 89 canned vegetables < 50 98 other vegetable exports b 60 c 72 65 81 96 126 96 98 89 109 192 163 of which: fresh vegetables 65 69 65 55 149 frozen vegetables 7 15 15 25 26 dehydrated vegetables 12 12 6 8 13 a. Estimated. b. Data from SIS, reported in GATT, 1994. c. Dgta for 1981 only. Sources: FAO and MARA, Turkey Horticulture Subsector Review: Approaches to Market-led Development (Rome: FAO, 1994), tables 7 and 9, and annex 4, p 9 and table 1; and GATT, Trade Policy Review: Republic of Turkey (Geneva: GATT, 1994), vol 1, pp 59-60. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 127 Total costs ofseed regulatory reform Externalities firom imported pests and diseases. An estimate of the costs of seed regulatory We found no evidence for any increase in losses reform depends on what changes are considered from imported pests and diseases. As of 1994, to be regulatory reforms and what are not. For many experts argue than Turkey's phytosanitary example, someone might think that regulatory rules are in some respects unnecessarily strict reform entails doing away with phytosanitary and suppress seed trade. The manager of a restrictions, and then consider losses from dis- Turkish seed company reported that, notwith- eases introduced through imported seeds to be standing phytosanitary controls, customs offi- one of the costs of reform. To avoid confusions, cials pay more attention to commercial import we darify the definition of regulatory reform. controls than to seed condition. Arguably, gov- Defining regulatory reform. Seed regulatory ernment could improve phytosanitary protec- reform focuses government regulatory activities tion by getting rid of extraneous seed import on efforts to control negative externalities, restrictions (for example, that only seed firms reducing or eliminating other government can import) and focusing on realistic phytosani- interferences in seed production or trade. tary concerns. Hence, seed regulatory reform: Farmer losses due to inappropriate varieties. Evidence suggests that 1980s regulatory reforms * Emphasizes trade and investment liberaliza- reducing the government's role in selecting and tion rather than privatization. Ending a promoting new varieties have reduced farmer public or private monopoly or oligopoly by losses from planting inappropriate varieties. revising regulations to allow free entry and During government crash programs to competition is regulatory reform. On the introduce Mexican wheat in the 1960s and other hand, privatizing a government seed hybrid maize in the 1980s, government agen- parastatal is not regulatory reform, though it cies decided which varieties of seeds to buy and may be worth doing for other reasons. where to distribute them around the country. The government offered incentives in the form * Promotes new channels for introducing new of credit or subsidies to farmers buying govern- varieties, and does not attack existing gov- ment-selected seeds. In both these programs, ernment research. Regulatory reforms ask farmer choice of varieties after several years government scientists to pay more attention showed major differences from initial govern- to their own research, and to spend less time ment allocations, which suggests assignment of and effort interfering with private company some inappropriate varieties and possibly some decisions to introduce new varieties. farmer losses (Frizzell 1968; also personal com- munications from maize seed experts during * Does not address the issue of subsidies. interviews in 1994). Whether or not to subsidize can be an Currently, with private companies having a important question because of impact on larger role in seed trade, markets participate in government spending. However, in most the selection and expansion of new varieties cases, seed subsidies do not seriously inter- over several years; the government no longer fere with private company ability to intro- makes large commitments to new and duce superior varieties. unproven varieties, and we did not see private companies taking big risks promoting unknown From this definition of seed regulatory varieties. This makes sense: private companies reforms, we consider costs of 1980s regulatory want to build public confidence in their brand reforms in Turkey. name; any big failure promoting an inappropri- ate variety would weaken that trust. 128 THE IMPACT OF TURKEY'S 1980s SEED REGULATORY REFORM Before reforms farmers lost by planting infe- Impact on domestic seed industries. In some rior varieties available in Turkey when superior countries, a common argument against regula- (but illegal) varieties were available in world tory reforms is that seed imports will damage markets. In some cases, particularly for high domestic inputs industries. In Turkey, the value vegetables, farmers smuggled seeds to domestic seed industry not only expanded after avoid planting inappropriate (inferior) legal reforms but also boosted export sales. varieties. In this respect also, regulatory reforms Weakening of Public Sector Research. During have reduced planting of inappropriate vari- the 1980s, MARA cut the budget for General eties, and have also reduced smuggling. Directorate of Agricultural Research (GDAR), Farmer losses due to poor quality seed. and new private seed companies hired away Evidence suggests that seed quality has thirty to forty senior GDAR scientists, includ- improved. With regulatory reforms bringing ing leaders of plant breeding programs. rapid expansion of private seed production and Whether or not 1980s changes in GDAR have trade, farmers have shifted a significant share of been a major loss to the country (experts in planted area for some crops - notably maize, Turkey disagree on the issue), changes have sunflower, many vegetables - from own-seed arguably been due more to coincidental budget to private commercial seed. This shift improves cuts than to regulatory reforms. seed quality. Furthermore, reforms have broken Seed cost. Hybrid seeds are much more crop-by-crop public and private monopolies expensive than non-hybrid seeds; furthermore, and oligopolies in commercial seeds; more farmers must replace hybrid seed every year, competitive commercial seed markets favor whereas they can often keep their own non- improvements in quality. Also, Turkey now pro- hybrid seed. However, returns far exceed seed duces seed for export; improving quality to cost. Also, all-or-nothing adoption patterns compete in export markets presumably has a suggest that all farmers, poor as well as rich, positive impact on quality for domestic seed as take advantage of new seed technology. Sugar well. beet area has shifted almost entirely to hybrids, Although TIGEM's seed production has as has sunflower area. Depending on region of fallen by more than 100,000 tons, that is due to the country, maize area has either shifted or not other policy decisions, not regulatory reforms. shifted to hybrids. What happens to TIGEM, and what impact Information on seed costs can be mislead- that has on wheat seed quality is tangential to ing. Switching to hybrids, farmers are able to costs and benefits of regulatory reform. Indeed, reduce seed rate; for example, the seed rate for regulatory reforms have brought in 10,000 tons hybrid sunflower in Turkey is only 3 kg/ha of private wheat seed production, which by compared to 20 kg/ha for OPV seeds; even itself improves wheat seed quality. though hybrid sunflower seeds are 5-10 times The government continues compulsory seed more expensive, the lower seed rate largely certification and official tests. However, one compensates for the price differential. For sugar anecdote suggests that private companies take beet, seed rate has fallen from 30 kg/ha with care of quality even without government multigerm (non-hybrid) seed to 3 kg/ha with involvement. In 1994, employees in one seed new monogerm (hybrid) seed. company stole expensive vegetable seed before packaging and replaced it with cheaper seed. Net impact of regulatory reforms Sale of mixed seed threatened the company's Following the train of evidence from seed reputation. Apprised of its error, the company regulatory reforms through seed trade to crop paid damages to affected farmers. Turkish seed yield and production, we have estimated annual companies maintain quality to protect their benefits from reform for selected crops only - brand names. maize, vegetables for export, and seeds for EAsING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 129 export - at about $130 million. Furthermore, In Turkey, reforms from the 1980s have the costs of regulatory reform, as discussed favored establishment of multiple channels for above, have been non-existent to insignificant agricultural research and technology transfer in in Turkey. Seed quality has most likely the private sector. Before reforms, private indi- improved, and farmers have more information viduals broke the law to test foreign varieties and choice, so there is less risk they will plant with smuggled seeds. Reforms de-criminalized an inappropriate variety. With low to non-exis- these research activities, encouraging private tent costs, the estimate of partial benefits (bene- individuals to scout varieties in other countries, fits for several crops) can be taken as a lower bring seeds back and test them in Turkey, and bound estimate for net benefits. then introduce new varieties to the market. Aggregate net benefits might be much larger After reforms, Turkish seed companies have than the above estimate of partial benefits. We multiplied channels to world technology. For have not tried to estimate benefits for all crops; example, one company in 1994 reported formal for example, we have ignored the impact of connections with at least seven international reforms on vegetable production for local mar- companies and Cornell University as well as $ kets. Also, benefits increase over time as farmers 50,000 annual expenditures to test new vari- and seed companies gain experience with new eties of maize, wheat, and other crops in technology and as seed companies increase Turkey. international contacts and financial capacity. In Turkey not only takes existing technology Turkey, seed regulatory reform has had a posi- but also influences and contributes to interna- tive and significant impact on net income. tional research. Most multinational seed com- panies do their breeding research in one to sev- Conduding observations and eral countries, limiting research in other coun- recommendations tries where they operate to adaptation trials. Both before and after 1980s reforms, most KWS, for example, breeds sugar beets in new technology in Turkey has come from for- Germany but selects lines for Turkey and many eign countries. This observation echoes a simi- other countries. Several Turkish seed companies lar finding by Eduardo Venezian for Chile: "the report sending materials for crossing to the overall most important innovations in Chilean breeding programs of international partners. agriculture post-1960 ... originate mainly After reforms, one multinational has established abroad" (Venezian 1987, pp 107-108). Hence, a breeding program in Turkey to develop sun- reforms did not change the source of new tech- flower varieties resistant to the orobanche para- nology, but did change channels and facilitate site, taking advantage of the presence of the flow. parasite. The varieties to be developed are intended not only for Turkey but also for other Multiplying channelsfor technology transfer countries to which the orobanche parasite is McMahon encourages developing countries expected to spread. to move from the "national institute model" for . . . ...................... Unfinished Business agricultural research toward a system with "many research players" and with "upstream, Although the seed industry has modernized downstream, and horizontal links among insti- and established strong international linkages, tutes, universities, firms, policy-makers, and some current policies continue to limit compe- social groups" (McMahon 1992, p 6). Antholt tition and also limit farmer access to new tech- advocates making room for "institutional plu- nology. ralism" featuring an enhanced role for the pri- Seed import controls block competition. vate sector in generating and spreading technol- The government limits imports to established ogy in agriculture (Antholt 1994, p 16). seed companies, limits the amount that a com- 130 THE IMPACT OF TURKEY'S 1980s SEED REGULATORY REFORM pany can import based on its record of seed these agencies have a role, they are not the only production, and (arguably) enforces some possible institutions that could have an impact. excessive phytosanitary restrictions that do not If governments or donors explicitly want the address realistic pest or disease threat. public sector to monopolize variety introduc- While Turkey from the 1980s has allowed tions, then it makes sense to continue controls multiple new varieties, compulsory variety reg- that block private channels while trying to istration and seed certification have remained in strengthen public research. However, when place. Also, Turkey may be back-sliding toward there is no commitment to a public sector tighter variety controls. Private seed industry monopoly, then it makes litde sense to design sources assert that major companies have no projects or programs to support government trouble getting their varieties approved, either breeding and extension without at the same for production permits or registration. In 1994, time considering relatively inexpensive or even company representatives report that the cosdess seed regulatory reforms to un-block pri- Registration Committee approves about 80 per- vate sector channels. cent of applications for production permits and Taking this approach, programs and projects 95 percent for registration. However, compul- for agricultural research and extension can be sory variety registration may be a more serious re-cast as agricultural technology projects, with barrier to entry for varieties sponsored by newer components to reform seed regulations that and smaller seed companies. block channels for variety transfer through the private sector. A recent World Bank Recommendations Agricultural Technology Project for Mexico Conclusions support the following two rec- illustrates this pattern (World Bank 1992). ommendations for donors and developing countries: Notes I. The Ministry of Agriculture and Rural Affairs (MARA) * Whenever increasing the flow of new vari- dates from a reorganization in 1994, which split the Ministry of Agriculture, Forestry and Rural Affairs. For all eties and other seed technology to farmers is time periods, this paper uses the acronym MARA to desig- an objective of a project or program, donors nate the ministry responsible for agriculture, though the and governments are encouraged to study name and other duties of that ministry have shifted over regulations to assess the extent to which they time. discourage or block private sector introduction 2. ICD is a US-based non-profit organization, supported of new varieties and other seed technology. by large multinationals, with a mandate to promote trade These assessments can be attached to agri- and investment to support development. cultural research projects, trade reform stud- 3. We also looked at impact of irrigation on maize pro- ies, inputs projects, etc. duction. Since maize irrigated area is not available, we used percent of total area under irrigation. 'The coefficient * Depending on the outcome of assessments, for irrigation was not statistically significant. donors and governments are encouraged to promote seed regulatory reforms as opportuni- References ties artse. Agency for International Development (AID). 1988. Development and Spread of Improved Maize Varieties Many programs and projects propose to and Hybrids in Developing Countries. Washington introduce new technology to farmers in devel- DC: AID. oping countries. Often, these programs depend Ange, A. 1994. 'Overview of Agricultural Development, on government research and extension services Fertilizer Use and Fertilizer Prices in Turkey," unpub- to breed and disseminate new varieties. While lished. Rome: Land and Water Development Division, FAO. EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 131 Antholt, Charles H. 1994. Getting Ready for the Twenty- MARA. 1994. "Draft Regulation in Respect of the First Century: Technical Change and Institution Application of the Law to Amend in the Law for Seed Modernization in Agricukure, World Bank Technical Registration, Control and Certification." Ankara: Paper 217, Asia Technical Department series. Directorate of Seed Registration and Certification, Washington DC: World Bank. MARA, May 1994. Aresvik, Oddyar. 1975. The Agricultural Development of MARA. Tohumiuk Programi, various years. Ankara: Turkey (New York: Praeger, 1975). MARA. CIMMYT 1994. 1993/94 Workl Maize Facts and Trends. Organization for Economic Co-operation and Mexico: CIMMYT. Development (OECD., 1993. List of Cultivars Eligible for Cerlification 1993. Paris: OECD. Echeverria, Ruben. 1991. "Impact of Research and Seed Trade on Research Productivity," in Philip Pardey, Pray, Carl E. 1990. "The Potential Impact of Liberalizing Johannes Roseboom, and Jock Anderson (eds), India's Seed Laws," in Food Policy (June 1990), pp Agricultural Research Policy: International Quantitative 193-198. Perspectives, pp 365-395. Cambridge: Cambridge University Press. Pray, Carl E, and Ruben Echeverria. 1988. "Transferring Hybrid Maize Technologr. The Role of the Private Food and Agriculture Organization (FAO). 1987. FAO Sector," in Food Policy (November 1988). Seed Review 1984-85. Rome: FAO. Republic of Turkey. 1963. "The Registration, Control FAO and Ministry of Agricultural and Rural Affairs and Certification of Seeds," law no 308, 1963. (MARA). 1994. Turkey: Horticulture Subsecteor Review: Approaches to Market-Led Development. Rome: State Institute of Statistics (SIS). 1977. 1977 Statistical Investment Centre, FAO/WB Cooperative Yearbook ofTurkey. Ankara: SIS. Programme, FAO. SIS. 1981. The Summary of Agricultural Statistics 1980. Freiberg, S, and A Grobman. 1985. "Development of the Ankara: SIS. Turkish Seed Industry," working paper 2, vol 1, pp 270-322 in: World Bank, Turkey: ASAL: Loan SIS. 1985. 1985 Statistical Yearbook of Turkey. Ankara: Implementation Volume, report no 5576-TU. SIS. Washington DC: World Bank. SIS. 1991. 1991 Statistical Yearbook of Turkey. Ankara: Frizzell, J K. 1968. Introduction of Mexican Wheat in SIS. Turkey, 1967-1968. Ankara: USAID. SIS. 1994a. 1991 General Agricultural Census. Ankara: Gisselquist, David. 1994b. Import Barriersfor Agricultural SIS. Inputs, Occasional Paper, UNDP-WB Trade Expansion Program. Washington DC: World Bank. SIS. 1994b. The Summary of Agricultural Statistics 1992. Ankara: SIS. Grobman, Alexander. 1983. "Toward a New and Dynamic Seed Industry in Turkey: A Review of Key Venezian, Eduardo. 1987. "Chile and the CGIAR- Problems and Policy Considerations," unpublished Centers," CGIAR study paper no 20. Washington paper. New York: ICD. DC: CGIAR Secretariat. Industry Council for Development (ICD). 1982. World Bank. 1985. "Turkey: SAR: Agricultural Sector "Development of the Turkish National Seed Adjustment Loan," Report 5576-TU. World Bank. Industry," report of the ICD mission to Turkey 22 Washington DC. February - 5 March 1982. New York, ICD. ____ -__ 1991. "PCR: Turkey: Agricultural Sector McMahon, Matt. 1992. Getting Beyond the "National Adjustment Loan (Loan 2585-TU)," Report 10025. Institute Model" for Agricultural Research in Latin World Bank. Washington DC. America, Report 20, Regional Studies Program, World Bank. Latin America and the Caribbean . 1992. "SAR: Agricultural Technology Project," Technical Department. Report 9782-ME. World Bank. Washington DC. - I Reforms, Regulations, and Recent Developments in the Seed System in Peru Joseph E. Cortes In the late 1980s, the government of Peru, * Growth of the private seed industry from with the support of the United States Agency seven to sixty-one enterprises by 1994. for International Development, initiated a development project with a seed component in * Grouping of the private sector into associa- an effort to increase private sector production tions of private seed enterprises. and distribution of seeds and to establish a non-gvernmnt qalitycontol sytem.* An increase in seed production from 12,000 non-government quality-control system. ° s @ .' . ~~~~~~~tons in 1990/91 to 17,500 tons in 1991/92; The first two years of the project were dedi- alls in the pro was podc by pri- cated to bringing the public and private sectors together and (a) delineating the basis for a vate enterprises. national seed system, (b) training personnel, (c) * Inspection and laboratory testing of all seed establishing regional seed associations, and (d) by regional seed committees (performed at creating "seed awareness" among farmers to the request of seed companies, even though promote seed use and to increase demand. A certification is non-mandatory) new government in 1991 further emphasized the advantages of private sector participation. * Financial autonomy was achieved by six of the eight regional seed committees by 1994. Reforms and results In terms of policy reforms, there were three Regional training courses on seed produc- key policies in the form of decrees that were tion and quality control. instrumental in the development of the private Installation of seed drying and processing seed industry: (a) the de-activation of two seed facilities in two key areas, managed by production parastatals; (b) the transfer of the regional seed associations and open to any parastatal seed infrastructure to farmer associa- farmer wishing to establish a seed enterprise. tions; and (c) the creation of regional seed com- mittees responsible for assisting the develop- . Provision of on-site technical assistance to ment of private seed enterprises as well as seed new seed entrepreneurs. certification. The results in the Peruvian seed system * Annual seed meetings among representatives were as follows: of the government, private seed companies, and regional seed committees to identify constraints and propose solutions. Recent developments Joseph Cortes is Coordinator ofInternational Seed As of May, 1995, Peru has increased its Programsfor the Seed Science Center at Iowa number of seed enterprises to 178 (see table State University. 4.11). This is primarily due to the incorpora- 134 REFORMS, REGULATIONS, AND RECENT DEVELOPMENTS IN THE SEED SYSTEM IN PERU Table 4.11 Number of private seed enterprises in Peru, 1988-95 State/department 1988 1991 1995 Arequipa - 5 18 Ica 3 5 7 La Libertad - 6 20 Lima 4 6 8 Lambayeque 2 10 11 Piura 1 6 22 Cusco 1 3 38 San Martin - 3 9 Andahuaylas - - 45 Total 11 44 178 tion of eighty-eight potato seed enterprises in Sources of improved germplasm and seed the highlands. Table 4.12 provides information Table 4.13 presents the origin of improved on number of seed enterprises by size as well as germplasm by cultivar. Cotton, which is a total area and crops under certification for the major commodity in Peru, is mainly dependent Departments of Apurimac, Arequipa, Cusco, on local and US germplasm. Rice, another Ica, La Libertad, Lambayeque, Lima, Piura and major crop, is originally derived from materials San Martin in 1995. spanning the globe. However, these cultivars are Of the 178 existing seed enterprises, 12 bred by IRRI, regionally tested by Centro percent have production areas ranging from 50 Internacional de Agricultura Tropical (CIAT), to 1,700 hectares, 33 percent are producing 10 and introduced by the national research insti- to 50 hectares of seed, and the remaining 56 tute. percent are smallholders with 1-10 has in seed Corn seed germplasm has different avenues. production (table 4.12). As mentioned earlier, Transnationals such as Pioneer have a contract small seed producers are mainly located in the agreement with a Peruvian seed company in highlands and are involved in potato seed pro- which Pioneer provides the parental materials duction, though some are in the coastal areas and seed is produced and marketed by the producing seed of rice, grain legumes, wheat Peruvian enterprise. Cargill, in a similar mode, and maize. The 43 percent of medium and provides their parental material from Brazil and large seed enterprises (by Peruvian standards) a local company produces and markets their are located in the coastal and jungle areas of own seed. Peru, which has the larger commercial produc- The University of La Molina has an active tion areas, and in some cases in the highlands corn breeding program and their materials are where potato seed producers are grouped into provided to a select group of seed producers associations. who produce and market under the auspices of Of importance is to note that all 9,300 an association. There is also a local seed compa- hectares currently under seed production are ny with its own breeding program. The corn being inspected, tested and certified by the breeding program of the national research insti- regional seed committees on a fee basis, tute is the other source of improved cultivars, although seed certification is not mandatory in which is sold in the form of foundation seed to Peru. local seed enterprises. Most of their improved materials emanate from Centro Internacional Table 4.12 Characteristics of seed enterprises by department, 1995 Number of seed enterprises with: Total Total number of less than between 10 more than seed area State/Department seed enterprises 10 hectares and 50 hectares 50 hectares Kinds of seeds (hectares) Arequipa 18 11 7 - mostly rice, some wheat and potato 220 Ica 6 3 1 2 mostly cotton, some beans and maize 2,000 La Libertad 20 5 14 1 rice and potato, with some maize, beans, 380 and wheat Lima 12 6 3 3 mostly cotton, some maize, beans, barley, 1,600 and wheat Lambayeque 11 1 6 4 mostly rice and cotton, some cowpea and maize 660 Piura 22 10 8 4 mostly cotton, some rice, soybean, sunflower, 1,900 cotton, maize, potato, wheat, grain legumes, and fruit trees Cusco 38 30 7 1 mostly potato, some wheat, maize, bean, barley, 340 oat, quinua, peas, and kiwicha San Martin 8 1 5 2 rice and cotton, some maize 240 Andah'ylas 45 33 9 4 1,100 Total 178 98 59 21 Table 4.13 Source of cultivars in use in Peru Area Crop Cultivar Onigin (percent) Cotton Tanguis Peru 60 Pima United States 30 Del Cerro United States 5 Asperos Peru 5 Others Peru Rice Vinflor Philippines-Vietnam 43 Inti China 16 Linea 14 Philippines 14 BG 90-2 Philippines 6 Sican Peru-Africa 5 Amazonas Peru-Philippines 4 Others Maize (hybrid) Marginal 28 CIMMYT-Mexico 30 Cargill Brazil 29 PM CIMMYT-Peru 20 Penta Cuba 15 Poey T-66 United States 6 Maize (OPV) Blanco Uru'ba Peru 10 Pardo Peru 5 Chancayano Peru 5 Puente Peru 5 Others Peru 75 Beans Canario 2000 CIAT 50 Centinela CIAT 30 Others CIAT 20 Lima beans Sol De Ica Carolina Del Norte 20 Iqueno Precoz Carolina Del Norte 20 G De ICA Carolina Del Norte 20 Promesa De Ica Carolina Del Norte 20 Senor De Luren Carolina Del Norte 20 Wheat La Molina 82 CIMMYT 50 Bavilan CIMMYT 40 Others ? 10 Potato Tomasa Tito United States/Peru 10 Condemayta Holland-US/Peru 8 Yungay Ecuador/US-Peru 2 Revolucion Holland/Peru 2 Mariva Peru-Holland 1 Andina India/Peru-Bolivia 8 Cica UK/Colombia 0.5 Canchan-Inia Peru 25 Many others Peru Table 4.14 Volume and value of US seed exports to Peru, 1991-92 to 1993-94 1991/92 1992/93 1993/94 1993194 Kind (volume in kgs) (volume in kgs) (volume In kgs) (value in $1,000) Grasses 10,100 - 50,100 51,900 bermudagrass 5,100 - 17,500 12,400 orchardgrass - - 12,300 17,700 ryograss, perennial 5,000 - 4,000 8,100 sorghum-sudan cross - - 16,300 13,700 Forages 14,700 60,100 153,300 455,100 alfalfa, certified 4,000 45,800 107,500 330,200 alfalfa, uncertified 10,700 14,300 42,400 118,500 clover, red - - 2,300 3,700 other - - 1,100 2,700 Leguminous vegetables 613,000 718,900 1,989,300 791,100 bean, gram - - 2,000 4,700 bean, kidney 26,500 10,000 bean, other - - 1,800 3,000 chickpea 16,400 9,900 lentil - - 63,300 29,200 pea 566,800 708,900 1,855,100 727,000 other legumes 19,700 - 20,600 17,3000 Vegetables 110,800 123,300 125,000 928,400 beet 5,100 580 cantaloupe - 340 1,900 7,200 carrot 6,700 20,800 9,300 77,200 cucumber 32 1,300 540 4,200 lettuce - 580 120 3,800 onion 1,000 490 2,000 37,500 radish 4,700 - 130 5,500 spinach 24,000 4,600 9,800 49,600 tomato 2,100 2,600 2,400 180,500 waterrnelon 2,300 6,700 2,900 35,000 other 64,900 85,200 95,900 527,900 Field crop seeds 35,000 54,000 16,100 22,000 com (except sweet) 25,200 grain sorghum 9,800 54,000 16,100 22,000 Flower seeds - 110 210 8,700 Sugar beet 7,200 - 1,500 11,900 Other seeds 23,800 5,200 15 7,600 Total 814,000 962,000 2,337,000 2,277,000 Source: Foreign Agricultural Service, USDA (United States Department of Agriculture). 138 REFORMS, REGULATIONS, AND RECENT DEVELOPMENTS IN THE SEED SYSTEM IN PERU de Mejoramiento de Maiz y Trigo (CIMMYT). The seed of other crops such as vegetables, Almost 100 percent of the improved bean grain legumes, grasses and forages is obtained seed varieties in Peru have been developed at by direct import of bagged seed mostly from CIAT, tested by the bean program of the NRI, the United States. Table 4.14 shows the annual and released after two years. Parental material of quantities of Peru's seed exports from the lima beans, however, originate in North United States during 1991/92-1993/94 as well Carolina and are made available through the the as their dollar value in 1993/94. Although a few national research institute to seed producers. of the seed enterprises import some of this seed, Wheat varieties are also very much depen- most of it is brought in by importers with dis- dent on CIMMYT research covering 100 per- tributorships in various areas of Peru. It is inter- cent of the varieties offered. As for potatoes, esting to note from table 4.14 that in the three their center of origin is Peru, which explains the years from 1991/92 through 1993/94, Peru's large number of local varieties. Most of the seed imports from the United States have improved varieties are generated through the almost tripled. International Potato Center (CIP), which works closely with the potato research program of the national reserach institute. 1990 Seed Policy Reform in Bangladesh: MovingAway from Variety Lists David Gisselquist This paper describes 1990 seed policy trade had given way to markets. However, by reforms in Bangladesh. While these reforms are the time general economic liberalization had not ideal, features may be recommended for gone far enough to allow private seed compa- other developing countries where farmers cur- nies to operate, the government had already rently have poor access to modern seed technol- introduced restrictive seed policies that specifi- ogy. Key features of 1990 seed reforms in cally blocked private seed production and trade. Bangladesh indude: The Seeds Ordinance 1977 gives the execu- tive branch authority to regulate seed import * government abandoned positive variety lists and trade, and provides for a National Seed for most crops, allowing anyone - private Board (NSB), chaired by the Secretary of companies, NGOs, and public institutions Agriculture, to advise government. The general -to sell seeds of new varieties without intent of the act seems to be to allow govern- prior government variety testing and ment to set some seed quality standards, in approval; and terms of germination and purity, for seeds of specified crops or varieties (sections 5 and 6, * government implemented these reforms The Seeds Ordinance 1977, Bangladesh Gazette, through executive decision, acting with 19 July 1977, p 4): authority from existing seed legislation. If the Government after consultation Bangladeshi seed policies and trade prior to with the [National Seed] Board is of refborm opinion that is is necessary or expedient During the 1970s and early 1980s, to regulate the quality of seed of any Drngl te had sn , f kind or variety to be sold and used for Bsangladesh's economy had strong socialist rea- th pupsso giutr,i a,b tures with limited scope for private seed compa- D p ny activity. Government agencies allocated for- notification in the official Gazette, speci- ny~ ~ ~ ~ ~ ~~~~~~~~f suchly kindnen ageae variette tobear-ife eign exchange and controlled private invest- fy such kind or variety to be a notified ment, whether foreign or domestic. With these kind or varety for the purposes of this controls, the only commercial private seed trade Ordinance... allowed was for vegetable seeds. the Government may, by notification By the end of the 1980s, socialist patterns in the official Gazette, specify ... the and controls on investment and international minimum limits of germination and purity with respect to any seed of any notified kind or variety... David Gissecquist is a consultant with the World Although the Seeds Ordinance seemed to Bank's International Economics Department and deal with quality issues more than varieties, is editor of this volume. 140 1990 SEED POLICY REFORM IN BANGLADESH: MOVING AWAY FROM VARIETY LisTs NSB used the authority of the Seeds Ordinance government obstructed introduction of new to create lists of allowed varieties for virtually all varieties of papayas and bananas, but high farm crops grown in Bangladesh. During a 1985 seed profits stimulated illegal trade in planting mate- workshop in Bangladesh, the Director of the rials; toward the end of the 1980s seed dealers Seed Certification Agency stated that "It is not smuggled papaya seed, and banana cuttings of legal to import any variety without approval of an improved variety from India moved farmer- NSB" (Bangladesh Agricultural Research to-fatmer. Council 1985, p 18). In a pattern that has been common in devel- Stimulus for seed policy reform: intention to oping countries, Bangladesh's list of allowed diversify varieties at the end of the 1980s was short and At the end of the 1980s, Bangladeshi farm- - except for vegetables - limited almost ers took advantage of liberalized market access entirely to varieties that had been identified and to irrigation equipment (primarily low-cost promoted by government scientists in the diesel engines from China) to expand irrigated Bangladesh Rice Research Institute, the area at record rates. Rapid growth in rice pro- Bangladesh Agricultural Research Institute, and duction promised to bring national rice self-suf- other public institutions in the national agricul- ficiency. Seeing success with rice, planners in tural research system (NARS). Many of the the Ministry of Agriculture feared that farmers approved varieties had been bred and released as would suffer weak rice prices and loss of lines by international agricultural research income unless they could be encouraged to centers (IARCs), then subsequendy tested for diversify into crops other than rice. years by NARS scientists before NSB approval The Ministry intended to improve farmer allowed their introduction to Bangladeshi access to inputs for a variety of crops, leaving farmers. farmers to choose the particular crops into For rice, the list at the end of the 1980s which they might want to diversify. In other allowed about twenty improved varieties, words, the Ministry did not want to choose However, many of these have been of no inter- winners, but rather wanted to offer farmers a est to farmers, leaving relatively few useful vari- menu of options. eties for three rice seasons (early and late mon- One episode in the Ministry's effort to soon and winter). Government scientists expand farmer crop options illustrates difficul- through the years generally ignored farmer ties with in-country breeding and restricted interest in short duration varieties; seeds of varieties lists. The Secretary of Agriculture some short duration varieties have come in ille- chaired a 1989 meeting on a-particular minor gally from India. Similarly, government scien- field crop as one option. At the meeting, an tists did not approve pajam, which was for expert from Bangladesh Agricultural Research more than a decade the most widely grown Institute proposed that his institute could breed improved rice variety in Bangladesh. improved varieties. At the time, the government For other field crops, the list of varieties of Bangladesh had approved and listed only two allowed was much slimmer than for rice. For varieties for the crop in question. Following a example, toward the end of the 1980s, the offi- breeding strategy, Bangladesh could look for cial list offered only two varieties of soybeans, introduction of new varieties in ten years. one of sunflower, and two of maize. However, the crop in question was one of the For fruits and vegetables, the government crops covered by the Asian Vegetable Research also limited varieties allowed for import. and Development Center (AVRDC), which is Government controls protected lucrative mar- located in Taiwan, a country with the same lati- kets for seeds of some popular varieties (for tude as Bangladesh and with comparable cli- example, for seedless watermelon). For fruits, mate. Suitable varieties were available from EASING BARRIERS TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 141 AVRDC as well as other regional and interna- strained arguments to justify variety controls for tional sources. major crops such as rice and wheat. Some Through these and other exchanges, argued, for example, that for a major crop there Ministry officials recognized that relying exdu- was risk that farmer choice of a "wrong" variety sively on in-country breeding and requiring could lead to significant regional or national prior government approval for new varieties losses; wide adoption of a disease-susceptible obstructed farmer access to suitable varieties variety could lead to collapse of production in already available elsewhere in the world. If some future year. However, these arguments Bangladeshi farmers were to gain access to a generally do not bear dose scrutiny. Experience menu of crop options in the foreseeable future, in Bangladesh with spread of new varieties sug- then variety lists had to go. gests that farmers and scientists have time to gain experience with a variety before it is widely Reform: doing away with variety lists for grown. Also, if scientists want to discourage a most crops particular variety because it is susceptible to dis- Anticipating opposition, leading officials in ease, the variety can be put on a negative list the Ministry of Agriculture did not want to act (which is inherently less restrictive than a posi- on an important technical matter without gain- tive list). Furthermore, doing away with or ing some support from specialists in the relaxing a variety list is likely to lead to more Ministry. In 1989, the Ministry established a varieties being available and grown, which committee to review and recommend new seed would cut risk of production collapse. policies. As expected, the main opposition to Subsequent to 1990 seed reforms, many liberalizing introduction of new varieties came organizations have gotten involved in identify- from government scientists in the national agri- ing and introducing new varieties. For example: cultural research system, including leaders of the Bangladesh Agricultural Research Council. * Local seed companies have been testing In 1990 the committee reported, recom- hybrids of maize, sorghum, and sunflower mending new seed policies, induding relaxing in collaboration with foreign seed compa- introduction of new varieties to improve farmer nies. access to technology available in world markets. Acting expeditiously to implement the commit- * The Bangladesh Tobacco Company, sub- tee's recommendations, the government used sidiary of a multinational, has been intro- authority from existing seed legislation to do ducing hybrid sunflower in an out-growers away with variety lists for all but five crops. program. One of the key steps was an intra-Ministerial order to the Plant Protection Wing to write * International Development Enterprises, an seed import permits according to species, not international NGO, has been assisting local varieties, for all but five crops. input dealers to introduce new imported The government exempted five major crops hybrids for popular vegetables. - rice, wheat, potatoes, sugar cane, and jute - from this reform. The real reason for contin- * Grameen Krishi Foundation and Bangladesh uing variety controls for these crops seems to Rural Advancement Committee (BRAC), have been that most government scientists dealt two large local NGOs, have organized farm- with these crops; allowing continuation of vari- ers to grow hybrid maize on several hundred ety controls for these crops was designed to buy hectares. their agreement to do away with variety con- trols for other crops. At the time that the new Most important, local seed traders have policy was being designed, there were some expanded domestic sales, increased contacts 142 1990 SEED POLIcY REFORM IN BANGLADESH: MOVING AWAY FROM VARIETY LISTS with foreign companies, and strengthened their of Agriculture drove the reform process through ability to influence future policies. In recent to success in 1990. years, private seed dealers have pressed for relax- In 1990, ADB included key features of seed ing variety controls on potatoes, one of the five reforms as conditions in its Foodcrops crops exempted from 1990 reforms. Development Program Loan (FDPL). ADB Despite liberalizing introduction of new asked the government to "Allow the private sec- varieties, the government has maintained exist- tor to: (i) import seeds and genetic material ing programs for breeding in public research (subject to health requirements)..." with a note institutes. This makes sense, insofar as it takes exempting rice and wheat seeds from this con- time to see how private organizations will dition (ADB 1990, p 72). In late 1990, ADB respond to new policies. Some adjustment in reported this condition as accomplished government research focus may be considered (though variety controls also remained for jute, in the future. sugar cane, and potatoes). Rules governing certification have not been The World Bank assisted in negotiating a major element of reform. The government did conditions for ADB's FDPL and repeated the not require certification before reform, so com- key condition concerning seed imports, but pulsory certification has not been a problem. with a weakening amendment, in Agricultural Certification is available as a service from the Support Services Project (WB 1991, pp 26, Seed Certification Agency. 33): BADC, a parastatal, has organized some seed production, mostly through contract grow- under the normal quarantine controls, ers, but there are also some large BADC seed both public and private sectors seed farms. BADC subsidizes its seed, which upsets companies would be allowed to freely private traders dealing with potato seed but so import all types of crop seeds for testing far has not been a major obstacle for other pri- [emphasis added] by June 30, 1992 vate seed trade. BADC seed has often been of poor quality. More importantly, after 1990 While the government independently liber- reforms, private seed dealers are able to com- alized seed trade in 1990, and ADB confirmed pete by offering seed of new varieties that farm- that policy through conditions in a 1990 loan, ers would prefer to varieties available from World Bank conditions dealing with seeds did BADC. not oppose government control over varieties allowed for commercial trade. On the other Donor involvement in 1990 seed policy hand, the World Bank asked that companies be reform allowed improved access to seeds for all crops The issue of seed policy reform emerged (for testing), including the five crops for which during 1989 discussions between the Asian the government continued to control varieties Development Bank (ADB) and Ministry of allowed for commercial seed sale. Another Agriculture while ADB was designing a World Bank condition in the same project program loan to support the Ministry's ongoing asked for unspecified changes in seed regula- reforms with fertilizer and irrigation. Shortly tions and laws. thereafter, the Ministry of Agriculture asked Subsequent to 1990s reforms, several donor- USAID to provide some technical advice on supported projects have had seed policy compo- seeds; USAID arranged for consultants from nents that have encouraged private seed compa- the Mississippi State Seed Technology nies. Also, from 1993 the World Bank has Laboratory. While donors contributed some repeatedly criticized variety-based restrictions ideas and information, officials in the Ministry for five crops; for example, a 1995 report urges EASING BARRIERs TO MOVEMENT OF PLANT VARIETIES FOR AGRICULTURAL DEVELOPMENT 143 that "to encourage private sector seed develop- it would automatically accept rice and wheat ment, the Government should eliminate the varieties registered by a list of other govern- 'notified' crops distinction for potatoes and ments. Moving toward automatic registration jute" (World Bank 1995, p.61). for varieties tested in other countries is more limiting than doing away with positive lists, but Bangladesh seed policy reform in it can be far less limiting than a single-country international context positive list. With 1990 reforms doing away with variety lists for all but five crops, Bangladesh moves Conclusion toward patterns common in the United States, If reform options are to be considered in India, Chile, and other countries, where private terms of what is best for the farmer, standard companies are allowed to sell seeds of new vari- economic theory supports policies that give eties without prior government approval based farmers the chance to test and choose varieties, on performance (or VCU [value in cultivation except where externalities are a factor. In other and use]) tests. These governments leave variety words, governments can leave decisions about performance evaluation to companies and farm- variety performance to farmers. Where this is ers, who communicate their assessments not possible because of opposition from inter- through the market. ested quarters, partial reform may be possible, Bangladesh's decision to maintain its single- limiting variety lists to a few crops or moving country positive lists for five crops - rice, towards multi-country variety lists. When a wheat, potatoes, sugar cane, and jute - contin- political compromise is possible, expeditious ues patterns that have obstructed variety trans- reform is often legally possible through execu- fer into many developing countries. This pat- tive orders by the minister of agriculture, with tern of single-country positive lists is superfi- authority from existing seed legislation. cially similar to patterns in Europe, but there is an important difference. Most European coun- References tries maintain positive lists of allowed varieties for many crops, but they are multi-country lists Asian Development Bank (ADB). 1990. "Report and Recommendation of the President to the Board of rather than single-country lists. Governments Directors on a Proposed Loan and Technical in the European Union accept varieties tested Assistance to the People"s Republic of Bangladesh for by other governments without any further tests. a Foodcrops Development Program," RAP: BAN These varieties are listed in two Common 22217. Manila: ADB. Catalogues, one for field crops and another for Bangladesh Agricultural Research Council (BARC). 1985. vegetables. Proceedings of the National Seed Technology The government of Bangladesh did away Workshop 1985. Dhaka: BARC. with single-country positive lists for most crops People's Republic of Bangladesh. 1977. 'The Seeds with simple government orders, acting with Ordinance, 1977" in Bangadesh Gazette, 19 July authority from existing seed legislation. This 1977. same pattern of reform is accessible to most World Bank. 1991. "SAR: Bangladesh: Agricultural developing countries with single-country posi- Support Services Project" Document 9071-BD. tive lists. Also, countries could, by executive World Bank. Washington, DC. order and without any formal international _______. 1995. 'Bangladesh: Agricultural Growth with agreements, move to multi-country lists. Diversification: Prospects and Issues," Report 14315- Bangladesh could, for example, announce that BD. World Bank. Washington, DC. Annex 1: Workshop Program Monday, June 12 9:00 to 9:30 am Introduction: Carl Pray, David Gisseiquist, Jitendra Srivastava, and Will Candler 9:30 to 12:00 pm Session 1: Technology Transfer: Conceptual Framework and Special Non-Policy Issues for Agriculture Chair: Jitendra Srivastava Discussant: Brian Fikkert Speakers: Larry Westphal, Phil Pardey, and Don Plucknett 1:00 to 2:45 pm Session 2: International Variety Supply: Public Research Institutions Chair: Derek Byerlee Discussant: Dana Dalrymple Speakers: Mywish Maredia and Doug Gollin 2:45 to 5:30 pm Session 3: International Variety Supply: Private Companies Chair: Mark Condon, ASTA Discussant: Carl Pray Speakers: Bobby Ansaldo, Ray Riley, and Jon Geadelmann Tuesday, June 13 9:00 to 11:00 am Session 4: Policy Obstacles and Incentives to Variety Transfer: Seed Regulations, etc Chair: David Gisselquist Discussant: Robert Tripp Speakers: Jitendra Srivastava, Jim Elgin, Dennis McGee, and Edward Eggers 146 WORKSHOP PROGRAM 11:15 am to 12:30 pm Session 5: Country Focus: Seed Policies, Projects, and Experiences and 1:30 to 3:15 pm Introducing Foreign Varieties through Public Research and Private Companies Chair: Matthew McMahon Discussants: Shiv Singh and Will Candler Speakers: Joe Cortes (Peru), Carl Pray (Turkey), Pramod Agrawal (India), and Venkatachalam Venkatesan (Nigeria and other African countries) 3:30 to 5:30 pm Session 6: Policy Reforms to Increase the Flow of New Varieties to Farmers in Developing Countries Moderator: Douglas Forno Panelists: Frederico Poey, Carl Pray, Robert Tripp, Mark Condon, and Derek Byerlee Annex 2: List of Participants Pramod Agrawal Dely Gapasin Proagro-PGS India, Ltd The World Bank Bobby Ansaldo Jon Geadelmann Pioneer Hi-Bred International, Inc Holdens Foundation Seeds, Inc Sharon Blinco David Gisselquist The World Bank The World Bank Michael Buechting K Glode-Shuh Buechting Agri-Business Rutgers University Derek Bycrlee Francesco Goletti The World Bank IFPRI Will Candler Doug Gollin The World Bank Williams College Mark Condon Oskar Honisch Mark Condon ~~~~~The World Bank American Seed Trade Association Joseph Cortes Carol Jacka Iowa State University Diversity Magazine Dana Dalrymple Rapeepun Jaisard USAIDaryp The World Bank USAID S S Kamwazina Ed Eggers Principal Secretary, Ministry of Agriculture and Dekalb Genetics Co Livestock Development, Malawi Jim Elgin David Lambert USDA American Seed Trade Association Brian Fikkert Jeffrey H Luhanga University of Maryland Chitedze Research Station, Malawi Douglas Forno Denis C McGee The World Bank Iowa State University 148 LIsT OF PARTICIPANTS Matthew McMahon Ed Quisimbing The World Bank The World Bank Mywish Maredia Ray Riley Michigan State University Pioneer Hi-Bred International, Inc Gustavo Nores Jitendra Srivastava CEO Consultants The World Bank Phil Pardey Shiv Singh IFPRI The World Bank Donald Plucknett Agricultural Research and Development Robert Tripp International ~~~~~~Overseas Development Institute Intemational Federico Poey Venkat achal am Venkatesan Agridec The World Bank Carl Pray Larry Westphal Rutgers University Swarthmore College Dennis Purcell The World Bank Annex 3: List of Other Papers Presented and Distributed at the Workshop Basant, Rakesh and Brian Fikkert, "The Effects of R and D, Foreign Technology Purchase, and Domestic and International Spillovers on Productivity in Indian Firms," forthcoming in The Review ofEconomics and Statistics. Byerlce, Derek, "On the Comparative Advantage of International Agricultural Research: Exploiting Economics of Size to Generate Global Spillovers," paper pre- pared for workshop on "Integration of Research Efforts of ICRISAT with NARSs and with other International Institutions," ICRISAT, Hyderabad, 14-16 Dec, 1994. Dalrymple, Dana and Jitendra Srivastava, "Transfer of Plant Cultivars: Seeds, Sectors, and Society." Evenson, Robert and Larry Westphal. "Technological Change and Technology Strategy," forthcoming chapter in T N Srinivasan and Jere Behrman, eds, Handbook ofDevelopment Economics, North-Holland Publishing Co. Fikkert, Brian, "An Open or Closed Technology Policy: The Effects of Technology Licenses, Foreign Direct Investments, and Domestic and International Spillovers on R and D in Indian Firms," unpublished paper. Fikkert, Brian. "Reforming India's Technology Policies: The Impacts of Liberalization on Self-Reliance and Welfare," unpublished draft. Jaffee, Steven and Jitendra Srivastava, "Seed System Development: The Appropriate Roles of the Private and Public Sectors," World Bank Discussion Paper 167. Wood, Stanley and Phillip Pardey, "Agroecological Dimensions of Evaluating and Prioritizing from a Regional Perspective: Latin America and the Caribbean," discus- sion paper, ISNAR. Wood and Pardey, 'Methods for and Limitations of Ecoregional Analysis," paper pre- pared for "IFPRI Ecoregional/2020 Vision Workshop," 7-9 November 1994, Airlie House Conference Center, Virginia. Distributors of COLOMBIA HAlTI ITALY NEW ZEALAND ROMANLA SWITZERLAND Intoenlace Ltda. Cuaure Diffusion Licusa Comnissionaria Sansoni SPA EBSCO NZ Ltd. Compani De Librarii Bucuresti S.A. Librairie Payot Seivice Instnutionnel W orld Bank Carrera6No.51-21 S Rue Capois ViaDucaDiCalabria, 111 Private Mail Bag 99914 Str. Lipsceani no. 26 sector3 Cotes-de-Montbenon30 Apartado Aereo 34270 C.P 257 Caselba Postule 552 New Market Bucharest 1002 Lausanne Publications Santati de Bogota. D.C. Ponrtau-Prince 50125 Firenze Auckland Tel: (40 1)613 9645 Tel: (41 21)341-3229 Prices rind credit terurs osanrfromn Tel: (571)285-2798 Tel: (5091)39260 Tel: (55)645-415 Tel: (64 9) 524-8119 Fox: (40 1) 312 4000 Fax: (41 21)341-3235 ounltl/ to) country. Coiisrdt your Fax: (571) 285-2798 Fax: (55) 641-257 Fax: (64 9) 524-8067 loatl distributor lircfire puaeipig an HONG KONG, MACAO E-mail:lcosa@1tbcc.i RUSSIANFEDERATION ADECO Von Diermen EditionsTechniques COTE D'IVOIRE AsMa 2000 Ltd. Ud: hfp1/www .thbc.it1Fcosa NIGERIA lsdatelstvo