University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln
International Sorghum and Millet Collaborative INTSORMIL Impacts and Bulletins Research Support Program (INTSORMIL CRSP)
4-2002
Proceedings of the West African Hybrid Sorghum and Pearl Millet Seed Workshop
John D. Axtell
Moussa Oumarou
Issoufou Kapran
Anand Kumar
Ouendeba Botorou
See next page for additional authors
Follow this and additional works at: https://digitalcommons.unl.edu/intsormilimpacts
Part of the Agricultural Science Commons, and the Agronomy and Crop Sciences Commons
This Article is brought to you for free and open access by the International Sorghum and Millet Collaborative Research Support Program (INTSORMIL CRSP) at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in INTSORMIL Impacts and Bulletins by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Authors John D. Axtell; Moussa Oumarou; Issoufou Kapran; Anand Kumar; Ouendeba Botorou; Gebisa Ejeta; Lee House; Thomas Crawford; Bruce Hamaker,; and Inoussa Akintayo West Af~can Hybrid Sorghum a dPearl Millet Seed Workshop Proceedings of the West African Hybrid Sorghum and Pearl Millet Seed Workshop
September 28 - October 2, 1998 Palais de Congres, Niamey, Niger
Conference support INRAN INTSORMIL ICRISAT ROCAFREMI ROCARS McKnight Foundation Rockefeller Foundation U.S. Agency for International Development Winrock International
April 2002 Publication number 02-02 Workshop Organizational Committee
Dr. John D. Axtell, INTSORMIL, Purdue University, W. Lafayette, IN, Chair Dr. Moussa Oumarou, INRAN, Niamey, Niger Dr. Issoufou Kapran, INRAN, Niamey, Niger Dr. Anand Kumar, ICRISAT, Niamey, Niger Dr. Ouendeba Botorou, ROCAFREMI, Niamey, Niger Dr. Gebisa Ejeta, INTSORMIL, Purdue University, W.Lafayette, IN Dr. Lee House, INTSORMIL, Bakersville, North Carolina Dr. Thomas Crawford, INTSORMIL, University of Nebraska, Lincoln, NE Dr. Bruce Hamaker, INTSORMIL, Purdue University, W. Lafayette, IN Dr. Inoussa Akintayo, ROCARS, Bamako, Mali
This publication was made possible through support provided by the U.S. Agency for International Development, under the terms of Grant No. LAG-G-OO-96-90009-00. The opinions expressed herein are those of the author(s) and do not necessarily reflect the views of the U.S. Agency for International Development. Dr. John Axtell These proceedings of the West Africa Dr. Axtell and his students demon Hybrid Sorghum and Pearl Millet Seed strated for the first time that tannins had a Workshop are dedicated to the memory of negative effect of protein availability and Dr. John Axtell in remembrance of his hence nutritional quality of sorghum. contribution to INTSORMIL and interna This resulted in a shift in emphasis of sor tional sorghum/millet research. ghum breeding programs around the world towards low tannin sorghums. Dr. John Axtell, an outstanding scien Thus, almost all sorghum varieties and tist, member of the National Academy of hybrids developed and released for hu Sciences and founding member of the man consumption in the last 10 - 15 years INTSORMIL CRSP died unexpectedly have been of the low tannin types. nd on December 2 , 2000. Because his ac complishments are significant and ex • Because of the discovery of the press the commitment which Dr. Axtell negative effects of tannins in hu had for international development and re man nutrition, it has attracted scien search we in INTSORMIL are proud to tists in biochemistry, genetics, and print this list of contributions toward the food science to undertake research improvement of grain sorghum in both on the various aspects of tannin and the U.S. and developing country context. phenolics in sorghum and food products. • Dr. Axtell pioneered a successful sponsible for the decreased digest research program that led to the ibility in sorghum. identification of natural and in • Dr. Axtell made world-wide contri duced high lysine mutants in sor butions in germplasm develop ghum. ment. Varieties and populations of • Dr. Axtell and colleagues con sorghum developed by Dr. Axtell ducted an array of studies which and colleagues have been widely clearly established that methods of distributed to national research pro food preparation can have dramatic grams and international agricul effects on the digestibility of sor tural research centers where they ghum proteins. They developed the have been widely utilized in vari technique for determining the di etal and hybrid development. gestibility value of sorghum-based • During his long and fruitful career, foods, in vitro. Then using their in Dr. Axtell trained many graduate vitro assay, they demonstrated that students and through their work as the proteins of various cereals be scientists, his legacy extends have differently when cooked, and throughout the United States and that the cooking process was re- many foreign countries. Table of Contents
Foreward, John Yohe ...... 9
Monday, September 28, 1998
Session 1 - Potential of Hybrids in West African Agriculture - Opening Remarks Moderator - John D. Axtell, INTSORMIUPurdue University
Welcome Speech - M. Oumarou ...... 13 Opening Speech - Omar Idi Engo ...... 15
Opening Comments - John D. Axtell ...... 17 What INRAN Expects from the Regional Sorghum and Millet Hybrid Seed Workshop - Moustapha Amadou ...... 20
Why Hybrids? - A. Bruce Maunder ...... 21 Rapporteur's Report - Botorou Ouendeba ...... 28
Session 2 - Hybrid Seed Experiences in Developing Countries Moderator - Lee House, Consultant, Seed Production
Hybrid Seed Experiences in Developing Countries - B.R. Barwale ...... 33 Zambian Seed Industry - History and Experience Bho1a Nath Verma and Winter M. Chibasa ...... 43 The·Experience ofWinrock International with Seed Multiplication in West Africa Through: The On-Farm Seed Project (OFSP), 1987-1992 and The On-Farm Productivity Enhancement Program (OFPEP), 1992-1998 Alphonse Faye and Pierre Antoine ...... 56 Rapporteur's Report - S.C. Gupta ...... 62
5 Session 3 - Current Status of Seed Production Moderator - Darrell Rosenow, INSORMIUTexas A&M University
Pearl Millet Seed Systems in Niger Jupiter Ndjeunga and Aboubakar Sidi ...... 67 Review of Sorghum Hybrid Research in Niger Issoufou Kapran, Moussa Adamou, Gebisa Ejeta and John D. Axtell ...... 83
Experience ofNAD-l Seed Production at Djirataoua Irrigated Perimeter, Maradi Region of Niger Laouali Dogo and Oumarou Issa ...... 88 Hybrid Sorghum in West Africa - D.S. Murty ...... 91
Heterosis and Hybrid Sorghum Seed Production in West Africa Aboubacar Toure, H. Frederick W. Rattunde and Inoussa Akintayo ...... 94
Rapporteur's Report - D.S. Murty ...... 104 Current Status of Hybrid Pearl Millet in West Africa K. Anand Kumar, B. Ouendeba, S. Boureima, S.c. Gupta and S. Ouattara .... 105
New Cytoplasmic Male Sterility Systems for Hybrids in Pearl Millet DJ. Andrews, K.N. Rai and J.F. Rajewski ...... 124
Creation of a Seed Industry: Role of the Private Industry - A. Bruce Maunder... 128 Rapporteur's Report - Anand Kumar ...... 136
Tuesday, September 29,1998
Session 1 - Bases for Succesful Hybrid Enterprises Moderator - A. Kumar, lCRISATIISC
Production of Quality Hybrid Seed - Soumana Souley and Magagi Abdou ...... 139 Field Management for Sorghum Seed Production Seyni Sirifi, lW. Maranville and R.K. Pandey ...... 145 Seed Conditioning and Commercialization - A.Bruce Maunder...... 155 Seed Production on Irrigated Perimenters M. Naroua Djibo and M. Jean Jacques Schalbroek ...... 161
6 Farm Level Profitability and Evolution ofInput-Output Markets: Economic Perspective - Tahirou Abdoulaye and lH. Sanders ...... 171
Regional Needs for Training in Seed Production N.D. Yagoua and N.I. Jika ...... 179
Training for a Future Seed Industry in Niger I. Kapran, D.S. Murty and L.R. House ...... 181
Rapporteur's Report - I. Kapran ...... 184
Session 2 - Marketing by Commercial Firms and Cooperatives Moderator - Kore Harouna, Fac. AgrolAMD University, Niamey
Role of the Public and Private Sectors in Seed Production Ismael Mouddour ...... 189
Constraints and Suggested Solutions in the West African Seed Sector - A. Joshua ...... 191
Role of the Government of Niger in the Initiation of a Seed Industry - Paul Buckner ...... 200
CreditiLoanslBanks Serving Rural Users in India Kuldip R. Chopra and Rakesh Chopra ...... 203
Wednesday, September 30, 1998
Session 1 - Sorghum and Millet Utilization Moderator - Bruce Hamaker, INTSORMIL
Evaluation of Sorghum Grain Quality for Traditional Thick Porridge in Africa - G. Fliedel, C. Sambumukama, L.. Songre, and F. Matenncio ...... 213
Senegalese Experience in Initiation and Strengthening Semi-Industrial Millet-Based Food Technology Plants in Urban Areas - Ababacar Ndoye ..... 226
Contribution ofPROCELOS to the Marketing of Food Products in Sahelian Countries - Mamadou Diouf ...... 228
Factors and Recommendations Concerning Seed Production - Lee House ...... 231
7 Pilot Project Report on Sorghum and Pearl Millet Utilization I. Akintayo and J. Ndjeunga ...... 239 Processing Sorghum and Millets into Value-Added Products Adam Aboubacar and Bruce Hamaker...... 241 Panel Discussion - Human Utilization - Bruce Hamaker ...... 246
Panel Discussion - Animal Utilization - Lee House ...... 249
Wrap Up Remarks - John Yohe...... 251 Conference photographs ...... 255
Acronyms ...... 263
Conference participants ...... 273
8 FOREWARD
Sorghum and pearl millet are the most strated the value of the collaborative re important cereal crops in West Africa search model in the agricultural sciences. where they provide 70% of total cereal This workshop has placed on record the production that provide the major com value of collaboration between scientists plements in the diets of people in the re in the national programs, INTSORMIL, gion. Research progress in the improve ICRISAT and the coordination efforts of ment of sorghum and millet is, therefore, the ROCAFREMI and ROCARS re of vital importance to the well being of gional networks in the development and people in West Africa. The Hybrid Sor promotion of new technologies for sor ghum and Pearl Millet Seed Workshop in ghum and pearl millet production and uti West Africa is the culmination of the re lization in West Africa. search and development effort by the Institut National de Recherches Several scientists made significant Agronomiques du Niger (INRAN) , the contributions in this experience. NAD-l International Sorghum and Millet Collab was developed as a collaborative partner orative Research Support Program ship between Dr. Issoufou Kapran (INTSORMIL), and the International (INRAN) and Dr. Gebisa Ejeta Crop Research Institute for the Semi-arid (INTSORMIL). Dr. John Axtell, coordi Tropics (lCRISAT) during the last 10 nator for INTSORMILlNiger led the cru years. cial effort to develop a hybrid seed indus try in Niger. Dr. Issoufou Kapran and Dr. The workshop served as a forum for Lee House (serving as INTSORMIL con documenting experience gained in the de sultant) worked tirelessly to make the pro velopment, release, and diffusion of duction and distribution of hybrid sor NAD-l as the first commercial sorghum ghum seed a reality for Nigerien farmers. hybrid in Niger. It marked a significant They promoted NAD-l and demonstrated milestone in the history of agriculture for the technical feasibility as well as the Niger and possibly for West Africa. The profitability of hybrid sorghum seed in workshop brought together sorghum sci Niger. entists from throughout Africa, the United States, India, and Europe. Mr. The production of hybrid seed requires B.R. Barwale, the 1998 recipient of the a multi-tiered effort toward moving from World Food Prize who is credited for lay a public seed production effort for varietal ing the foundation of the seed industry in seed toward private sector initiatives India also attended the meeting and better organized for producing and mar shared his experience with a paper enti keting of hybrid seed. In Niger, great ef tled "Hybrid Seed Experiences in De fort has been directed toward developing veloping Countries". local capability and experience for hybrid seed production and marketing. Small The research, which was presented and farmers have been organized and trained discussed, during this workshop demon- to become skilled seed producers. The
9 private sector in Niger is gradually as port of international participant travel and suming more responsibility for seed pro conference activities which made this duction and marketing. The process has meeting a special success. As Program been slow but deliberate and is likely to Director of INTSORMIL, it gives me grow setting a good example for the other great pleasure to have our program asso countries in West Africa. ciated with this important milestone in the history of sorghum and millet in Niger On behalf of IN RAN and and West Africa. INTSORMIL, I wish to thank ICRISAT , ROCAFREMI, ROCARS, the McKnight JohnM. Yohe Foundation, the Rockefeller Foundation, INTSORMIL Program Director the U.S. Agency for International Devel opment and Winrock International for co-sponsoring this workshop and for sup-
10 Monday - September 28 - Session 1
Potential of Hybrids in West African Agriculture
Opening Remarks
Moderator - John D. Axtell, INTSORMIL/Purdue University Welcome Speech
Mr. Moussa Oumarou, Director General of INRAN
His Excellency the State Minister, Honorable Director of the INTSORMIL program, Honorable Representative ofICRlSAT in Niger, Honorable guests, Ladies and Gentlemen;
On behalf of INRAN, I congratulate teen improved varieties of sorghum and the organizers of the Regional Sorghum millet that are available in Niger. and Millet Hybrid Seed Workshop. We are grateful to all of you, as representa Honorable guests, ladies and gentle tives of the scientific and agricultural de men, our government and international velopment communities, for taking time donors have made a tremendous effort away from your important tasks to be with and invested large sums of money to pro us. To all participants, we extend a warm mote the use of improved seeds by our welcome to Niamey. farmers. Yet all these initiatives have only had only a limited impact on agricultural INRAN, created in 1975, is a young ag production in the Sahel. This is due - in ricultural research institute. Yet in the part- to the lack of a viable enterprise in Sahelian context, it has the tremendously seed production and marketing. As agri important responsibility of conducting re cultural scientists, the responsibility to search aimed at improving food remedy this situation falls on us. The self-sufficiency in Niger. Very early on, value of an improved variety will remain we understood that if a significant in rather academic as long as good quality crease in agricultural production was to seed is not produced in quantities that can be obtained we needed to promote im cover the substantial needs at the national proved seeds. As an institute, our strategy level. Improving the seed sector should was based largely on collaborative re start with a unanimous recognition by search. This workshop gives me, there producers, consumers, extension agents, fore, the opportunity to recognize and scientists and policymakers that seed and thank INTSORMIL and ICRlSAT for be grain are not synonymous. ing with us from the beginning and stay ing with us as we follow the long, but pro Ladies and gentlemen, INRAN has ductive, path of agricultural research. taken steps, especially through training, to develop and to support a seed sector Dear participants, in terms of agricul where quality is a key word. Infrastruc tural research, varietal development of tures have been set up to insure that spe sorghum and millet has probably been the cialists produce good quality breeder most successful venture. The latest seed. These investments were timely in a INRAN catalog displays more than fif- context where our joint efforts have de-
13 veloped high yielding and well-adapted Ladies and gentlemen, we hope all of hybrids of sorghum and millet. Since hy you will have the opportunity to share brids can significantly improve the food your experience, and to see the results of situation and serve as a better source of our collaborative research through the profit for commercial growers, we be field tours. We hope these efforts will be lieve hybrids can easily serve as vehicle rightly appreciated, so that agricultural for a viable seed industry. The system will research regains its position as the engine be more sustainable if food products like for development and our institutions re couscous are locally made from sorghum ceive the full and timely support they de and millet and are acceptable to consum serve. ers. This is a brief summary of INRAN' s dynamic approach to the question of sor Once more, I wish you a pleasant stay ghum and millet hybrids, an approach among us. well received by many of our partners. Thank you
14 Opening Speech His Excellency the State Minister Omar Idi Engo, Minister of Agriculture and Livestock,
Distinguished INTSORMIL Director, Distinguished Director of the ICRISA T Sahelian Center, Distinguished Director General ofINRAN, Distinguished representatives of international organizations, Distinguished representatives of farmer associations, Honorable guests, Dear participants, Ladies and gentlemen:
Let me start with the pleasant duty of search Institute of Niger (INRAN), aim welcoming among us so many important ing at improving the productivity of scientific personalities; to all participants important crops. We know that useful re I wish, on behalf of Niger's people and search has been conducted in the areas of government, a wonderful stay among us. varietal improvement and seed produc tion. In particular, hybrids have per It is my privilege to give the opening formed better than open pollinated variet speech at this Regional Sorghum and Mil ies in the difficult conditions where millet let Hybrid Seed Workshop. Our people and sorghum are grown. And, we recog and government are grateful for the op nize that improved varieties are - like fer portunity to host such a significant event, tilizers, water, and phytosanitary prod as this workshop is indeed in line with ucts - indispensable for high agricultural Niger's strategic objectives of food productivity. Therefore, we are proud to self-sufficiency. know these results have been accom plished in Niger. I commend the Millet and sorghum are the most im long-term collaboration between portant cereal food crops in this country, INRAN, INTSORMIL, and ICRISAT, with more than 7 million hectares devoted for all they have contributed toward this to these crops. Yet we face almost en success. demic food shortages due - in part - to poor yielding varieties grown under Ladies and gentlemen, dear partici harsher and harsher environmental condi pants, a modem agriculture will signifi tions. The trend is very alarming and it cantly contribute to the fight against pov makes food availability a critical issue, erty. This is a key element for economic since the population is growing annually development in any nation. Experience in by 3%. other parts ofthe world has shown that the availability of improved varieties and We follow closely and fully support the seed marketing are crucial for modem ag efforts of the National Agricultural Re- riculture. The eminent experts, scientists
15 and producers in this room will undoubt Dear participants, I know you are better edly exchange knowledge and experi informed than I am about the issues de ences so that we can develop a pragmatic scribed briefly here. I know you will thor approach for extending sorghum and mil oughly analyze the value of hybrid tech let hybrids to farmers. Your agenda in nology and seed production for us. I am cludes discussion on production, process confident your discussions will produce ing, and marketing of hybrid seed, as well recommendations that we can implement as human and animal utilizations of the in Niger and other Sahelian countries. In grain. It is indeed essential that modem closing, I wish success to you all and de transformation of the grain be developed clare open the Regional Sorghum and to absorb increased grain production, and Millet Hybrid Seed Workshop. make the system sustainable. Thank you
16 Opening Comments
John D. Axtell
Minister of State, Agriculture and Animal Husbandry Director General of INRAN Director of the INTSORMIL Program Director General of the ICRISAT Sahelien Center Honored Invitees
Ladies and gentlemen, welcome to the ROCAFREMI and ROCARS regional first West African Hybrid Sorghum and networks have resulted in the establish Pearl Millet Seed Workshop. This work ment of the first significant foothold for shop represents the culmination of re the development of sorghum and pearl search efforts by INRAN in collaboration millet hybrids in Sahelian West Africa. In with many agricultural scientists and sev addition, important financial support eral research organizations over the past from the McKnight Foundation was re 18 years. ceived by Dr. Kapran and INRAN during the research and development phases I would be remiss, however, if! did not leading to the development ofNAD-l. make special note of the important role played by Dr. Issoufou Kapran and his Weare optimistic that this beachhead colleagues at INRAN and INTSORMIL. that has been established through the ded Their bold vision for the future of hybrid ication and perseverance of Dr. Issoufou sorghum in Niger, and the exuberant Kapran and Dr. Lee House will grow and farmer response to the hybrid NAD-l at spread. A solid beginning has been made, field days throughout Niger, bolstered our but this is only the beginning of what must faith that we could and must overcome the become a sustained effort throughout the issues involving hybrid seed production region. Many challenges must be over in Niger. We are fortunate to have "pio come, and many issues remain unre neer" researchers from India, Sudan, solved. But we are driven by the fact that Zambia, and Nigeria at this meeting who sorghum and millet are essential to the di will discuss their experiences with hy ets of poor people in the semi-arid tropics, brids later this morning. where drought causes frequent failure of other crops. These cereals are most im The research we will be discussing and portant in West Africa, where they pro viewing this week is a monumental trib vide 70% of the total cereal production. ute to the value of the collaborative re search model in the agricultural sciences. In North Africa they provide 8 - 91 % of Research collaboration between national the total staple for over 216 million peo program scientists, INTSORMIL U.S. ple. In East Africa they provide 30% of university scientists, and the coordinated the total cereal production; and in Central research efforts represented by the and Southern Africa they provide 9 - 85%
17 of the total staples for over 126 million declines to the farmers for these cereals at people. In addition, they are important the farm gate. The organizers of this crops for over 2.1 billion people in Asia, Workshop give equal priority to increased mainly India and China. research and development efforts on the utilization of these cereals, which allow Sorghum and pearl millet are, there farmers to benefit from the productivity fore, the two most important cereal crops lllcrease. grown to feed people living in the semi-arid, low input, dry-land agriculture The research and development efforts regions of Africa and South Asia. In the on the cereal technology of sorghum and last 35 years, the area devoted to sorghum millet grains have lagged far behind that in Africa has nearly doubled, but yields of other major cereals. The research on have not increased - still averaging 800 utilization of these cereals must be in kg/ha. A similar trend exists for all millets creased if we are to provide acceptable, in Africa where the area planted has in value-added products to an increasingly creased by 50% but yields, averaging 620 urbanized population. Some encouraging kg ha- 1 and showing more fluctuation than research is in progress in several West Af sorghum, have changed little. By contrast rican laboratories, and we must strive to in India during the same period sorghum ensure that these efforts are encouraged area has declined 37%, but yields have in and enhanced. My opinion, in fact, is that 1 creased 80% (now about 880 kg ha- ). one of the greatest deterrents to increased Likewise area planted to millets has de funding for sorghum and millet research creased 28%, but yields have increased is the lack of delectable, convenient sor 1 75% (now ±720 kg ha- ; USDA, 1997). ghum and millet products in the cities Hybrid use, and a consequent increase in where policymakers live. genetic research has been a major factor in the yield increases recorded in India. There is a whole gamut of issues in We are delighted to have Mr. Barwale at cluding hybrid research; training in seed our workshop to share his experience with production and seed quality; financing, hybrid seed production in India. Mr. marketing and seed distribution; and seed Barwale is founder and chief executive of law and government policies that need to Mahyco Seed Company which has been be discussed. Weare fortunate to have ex instrumental in the development of this perts in each of these areas to help with remarkable story in India. Mr. Barwale is our discussion. But we need your exper the recipient of the 1998 World Food tise and your participation if this work Prize for his pioneering work in develop shop is to be successful. An important ing the private seed sector in India. start has been made, but much work re mains ahead. We know what the benefits of hybrid vigor have been in other parts of the So the objectives of this Workshop in world. We now know that these benefits clude seizing the opportunity provided by can be achieved in West Africa. We also the benefits of the increased productivity know that the benefits of significantly in of sorghum and millet hybrids, and ac creased yields can cause dramatic price cepting the challenges of providing high
18 quality hybrid seed to the fanners of West brid sorghum and pearl millet cultivars in Africa in a timely fashion and at a reason Sahelien Africa. Let us begin this revolu able cost. Experiences throughout the tion here and now, and let historians note world have demonstrated that this can that this Workshop set the stage for signif best be achieved by a private sector seed icant gains in the welfare of people living industry, accompanied by appropriate in the semi-arid tropical regions of the government policies. We will have a road world. map provided by other developing coun tries that have faced and overcome these My home institution, Purdue Univer challenges. How can this road map be al sity, is also the home institution of a very tered and adapted to West African condi famous American, named Neil tions? That is our challenge. Annstrong, who was the first man to walk on the surface of the moon. His comment The huge production areas of sorghum as he stepped out ofhis space capsule was, and pearl millet in Africa have yet to ben and I quote "One small step for man, one efit from the use of heterosis. Weare con giant leap for mankind." Let's make this vinced that the use of heterosis or hybrid Workshop one small step toward great vigor will playa major role in the next strides forward toward freedom from green revolution: a dryland revolution re hunger for all the peoples of the Sahel. sulting from increased productivity ofhy-
19 Introductory remarks
What INRAN expects from the Regional Sorghum and Millet Hybrid Seed Workshop
Dr. Moustapha Amadou Scientific Director of INRAN
INRAN is a young organization whose technical requirements in a country which goals are to find technical solutions to ru has no modem seed company. ral development in the areas of crop pro duction, livestock and fisheries, natural We also have an experimental unit for resource management, and production the processing of millet and sorghum into systems. This is a very difficult task in the couscous (grain decortication, flour pro current economic crisis, especially in the duction, rolling, cooking, and drying). harsh Sahelian environment. Still, with a This will be a useful tool for processing team of well motivated scientists and sup excess grain production resulting from port from government and numerous re improved cultivars. search partners, INRAN keeps improving its impact on agricultural development in Therefore, we are interested in hearing Niger. Production, processing, and mar the comments of the experts in this room keting of millet and sorghum hybrid seed, on how to improve the seed sector in as outlined on the agenda of this work Niger. This is why we set up two working shop, are key elements for agricultural groups, one on seed production and the production. other on utilization, to review and im prove our preproposals. We expect each In Niger highly productive varieties of group to analyze their document thor millet and sorghum (NAD-l) have been oughly and improve so that it may be sub developed. Now we need to adequately mitted to government and other funding produce their seeds in sufficient quantity agencies. and good quality, as well as comply with
20 Why Hybrids?
A. Bruce Maunder
Introduction (intravariety crosses) that produce a hy brid superior in growth, size, yield, or Agriculture and plant improvement general vigor. The value of heterosis de have evolved in rather definitive stages pends upon the ease of hybridization and during the 10,000 years since man's orga cost of seed production to be justified by nized dependence on crop production. increased yield, as well as other agro Assuming we can now safely consider nomic characters being enhanced. ourselves in the Science Revolution, as described by Plucknett (1991), being Advantages of Hybrids from 1950 to the present, then we ad vanced from an agriculture dependent on Increased grain yield is not only a uni manpower followed by animal power; versally recognized manifestation of hy and, of course, during this century mecha brid vigor, but represents the most signifi nization or machine power. At the same cant breeding accomplishment for im time crops moved from improvement proved productivity. Dash, in opening through evolution with natural selection remarks at the 1983 Hybrid Sorghum to farmer selection for the most desirable Seed for Sudan workshop related to the traits. This rather lengthy time frame was first commercial hybrid for Sahelian Af interrupted by the discovery of Mendel's rica, Hageen Dura - 1, indicated field per Laws of Heredity at the beginning of the formance of 158% over the local variety 20th Century. Then plant scientists devel under irrigation and 150% of the local un oped the principles of modern plant der rainfed conditions in Sudan. Ejeta, at breeding and explained the phenomenon the same meeting, explained that test re of hybridization, which has most recently sults over four years favored the elite hy been upgraded through the newest breed brids over open-pollinated local varieties ing tool, biotechnology. in 27 yield trials by a 50% yield advan tage. Quinby (1963) found that RS 610, Hybrid vigor has proven to be by far the the first widely grown U.S. hybrid, pro most productive and economically valu duced 82 percent more grain than the av able tool of this scientific revolution in erage of its parents, but produced only 31 agriculture. Although maize receives percent more stover. most of the attention related to hybrids, certainly sorghum, millet, sunflower, Doggett (1967), working in Uganda, rice, and alfalfa have benefitted signifi evaluated the sorghum hybrids H x 57 and cantly as have numerous vegetable crops, H x 58 in 123 trials in East Africa (Table Maunder (1998). Shull in 1914 suggested 1) and found an average 35.8% yield in the term "heterosis," which may best be crease. Data also presented from Zimba described as hybrid vigor. Heterosis re bwe, India, and the USA, representing an- sults from crossing two unlike plants
21 Table 1. Relationship between hybrids and Table 2. Evolution of Yield of the Principal varieti~s (H. Doggett, 1967, Uganda). Cereals and Oil Crops in Argentina, Country/ % increase No. Of 1910-1994. Region hybrid trials Kg ha-1/year East Africa 35.8 123 Cultiy,!!: __ .____ ~.Years an(Ltylle . Maize 1910-50 varieties 9.00 Zimbabwe 53.7 64 1951-94 hybrids 57.92 India 34.1 76 U.S.A. 28.1 128 Sorghum 1954-65 varieties - 7.52 1966-94 hybrids 58.43 Nature, Vol. 216, No. 5117
Sunflower 1929-80 varieties -0.36 1981-94..hybrids other 268 trials, ranged from 28.1 % to 43.0_0~~ 53.7% in favor of hybrids. Data from Argentine Secretary of Agriculture
Nider et aI., (1995, unpublished), using data from the Argentine Secretary of Ag Besides grain number per inflores riculture 1910-1994, found an annual cence other benefits from heterosis may gain of 57.9 kg ha- I for maize hybrids for be equally important. Often a valuable at 44 years compared to 9.0 kg ha- I for vari tribute is earlier maturity. Work in maize eties for the prior 41 years (Table 2, Fig. (Duvick, personal communication) and 1). Sorghum and sunflower had annual sorghum suggests that hybrids from par . f I gams 0 58.4 and 43.0 kg ha- , respec- ents having similar maturity genes, will tively whereas variety analysis prior to flower in 93-94% of the days required for hybrids showed annual yield reductions the mean of the parents. A DEKALB Ge of-7.5 and-O.4kg ha- I for these two crops netics study with 15 commercial hybrids (Fig. 2,3). and their parents confirmed this advan tage and further showed a hybrid advan In the USA, yield gains ofthe 90's from tage over the best parent of22% and 35% the base 10 year averages of 1930-39 fa of the best parent over the mid-parent av vor the hybrid crops of maize and sor erage. Earlier maturity from hybrids is es ghum over varieties such as wheat, cot pecially valuable for improved water use ton, and soybeans by five fold (5X) com efficiency, insect escape, and avoidance pared to 2.6X for the non-hybrid crops. of temperature extremes. Quinby, (1973) suggested that since the advent of hybrids, a combination of hy Hybrids generally result in signifi brid vigor, irrigation, and better nutrition cantly improved seedling vigor, an im has increased the sorghum grain crop in portant aspect in weed control. They also the United States, from 250 to 800 million frequently have greater height than the bushels (63.6 to 203.6 million quintals). tallest parent as well as more extensive Thus, hybrids not only bring an initial tillering. Although seed number per pani yield gain of 20-40% but show greater cle, as mentioned, is the greatest factor for yield gains once a cultivar is hybridized. increased yield, both panicle size and panicle weight are also favored by hy brids (Quinby, 1973). Certainly the
22 ~~------~
4600 .... --"... - .... - .."- .. "-..... -~ .... -.... '"-.. --..... --.. -.-.... ~ ... ".-.-... ·Hibrides.. 2aa...... --.. ;IK .. .
4000 ----.-.... -'".-... --~-.~-.... -" ... - ...... ~ ..... --... -.g8osracion ...... -~ )II! 3500 •
" 3000 ...~ ~ 2500 ;;: lit--- 1600 .. J!--...... Hlbridos .. '1Ta: - .. -~ ..... ----. -----I-generadGn
o if I i Ji i n it i i it fi f "H t'"I it Ii' Hi" 11" I It' I'Tt"~ • jill, , 11 t rrrl I , n Ii l , I Ii I " U uno 1931 1962 1973 1994
Figure 1. Yield of Maize - Argentina, 1910-1994.
~oo~~~~~~----~----~----~~------=~
3500 ...... _ ..... -_...• - ...... - ...... _...... --...... ~ ...... -*...... - .... --.~ .. - ...... * ....
* )111* i.E 3000 _ ...... _ .. _ .. _ .. __... _ ... _ ... _ .. __...... _ ...... _ .. _ .. .11"-...... _...... - .. ~~ .. - ... -•• -.--..... lIE * ** 2500 .. -...... •.... ·.. "'· .. ~.... ·- ...... ·-~"-· .. - .... ··--.... ··.... ""~~-.... -·i Hibridos 1· .. -·-..... - .. ·-·-· * ._.. }IE *.._...... _ ...... _ ..... _ ...... _...... _ ..... ___ .. __. __ .. )111** ME )IE . 1500 ·...... _·.. ·_... ·_·_.... _··lIr· .. __ .... __ .. ··_· ..· .. •__ ...... _·· ..· .... · ...... ~ ..... - .... --.. --... ME
Figure 2. Yield of Sorghum - Argentina, 1910-1994.
23 2000~------~====~-c . IHlbrldos I 1800 "-'-"""-~"-"-"-"--"-~"-""""-~--""~--''''''''''''-"H· .. · " .. ".-...... " .. - .. --... .,~"-."-.....*- ...
1600
IlK )IE 1400 ."._ ...... _..... _.... - .. __ .-.. _-...... __ ...... - ...... -.~...... -...... _ ...
;.t( )111
1200 M ••_. ___.. _"'~· __•• __·_· ...... _·· __.. ·_· __ .._ ..___ • .. _··_··_·_·... - --"--'-""~-" lIiI!* 1 000 .-.. - .. - .... -"'*.~-.~.-.. -.... -. -_.. -... __ .... _...... -...... _.., ...
* lIE lIE ~ lIE ~ 800 .... _.. "'_ ..... _.... _ .. _ ..I!IIS._ .. _. ___.,_. __...... _ ...... _ .. _.. _~ .. .:. •. 1lIL.. .• '*...... ~ .. - ... - ...-.
* lIE ~ * 600 ... -~. .,.~- ...... -..... --~-... ~ .... -;... ! .. ~ ...... ~.-,.. ... "...... "...... "'."' .. ..
Figure 3. Yield of Sunflower - Argentina, 1929-1994 higher threshing percentage, larger root gest modern maize hybrids (single system, and frequently greater test or crosses vs. double-crosses or 4-way hy grain weight are significant to improved brids) not only yielded more grain than performance. Doggett (1967), suggests did older hybrids in low moisture envi the small increase in height, stem thick ronments, but exhibited genetic capacity ness, leaf area, forage, and tiller number to increase this yield difference in favor could together add up to quite a large in able environments. The same analogy crease in photosynthetic area. With more was found between low- and high-fertility floral buds/hectare than varieties, the conditions, with newer hybrids superior plant factory could always fill the extra to older ones in both fertility levels. Thus, seeds in those florets, for a wide range of with both moisture and fertility, hybrids growing conditions. of maize continue to provide increasing yield stability with improving genetic po Drought and temperature tolerance of tentiaL ten permit hybrids to offer an even greater advantage under stress. By having a From a breeding perspective, hybrids higher yield potential, hybrids illustrate a benefit from non-additive gene action form of improved water use efficiency. such as dominance and epistasis. In fact, Studies by Castleberry, et. aL (1984) sug- both parents need not be superior thus al-
24 lowing for the expression of maize (Hallauer, et. al. 1988) results from overdominance. Hybrid improvement for the accumulation ofloci with alleles hav adaptation has been clearly shown with ing partial or complete dominance. maize when comparing increasing yield Overdominance and the interaction of al gain and higher plant populations. Eastin leles (epistasis) may be significant de (1983) has further added, that crops that pending on the particular germplasm be can be hybridized stand to benefit more ing used. Glenn Burton (1966) specu quickly from a team approach (physiolo lated that sorghum breeders had not gist-geneticist team) that is so critical for begun to use all of the additive gene ef improving abiotic stress tolerance. fects before moving into a hybrid pro Finally, hybrids can be upgraded by im gram, that also allowed them to utilize proving only one of the two parents, since dominance. dominance frequently affects simply in herited or qualitative traits such as biotic Breeders recoglllze that all stress, nutritional characteristics, and intra-variety crosses do not produce the height. same amount of vigor when crossed. I have had opportunity to develop, observe, Explanation of Hybrid Effects and release approximately 150 commer cial hybrids. Of interest, however, is the Heterosis results in increased size and knowledge that considerably more than number ofplant parts, from a faster rate of 150,000 have been evaluated in our cell division and cell activity, thus, giving breeding effort, or less than one in 1000 hybrids a greater number of cells. Quinby ever reach the producer. Thus, not all (1963) reported faster cell division as the combinations exhibit adequate levels of basis of the heterotic manifestations ob heterosis or lack some other essential served in sorghum. Processes not in trait. The breeder becomes more selec volved by rate of cell division, as a result, tive for a useable hybrid as his research show little effect on hybrid vigor. Those program becomes more advanced or ma factors developed more by breeding than ture. heterosis include test weight which is cor related to days from bloom to physiologic Acceptance of Hybrids maturity, seed size, protein composition and relative percent, ear placement in Perhaps the u.s. history, which illus maize, stay green, and biotic resistance. trates the relative acceptance ofmaize and Quantitatively inherited traits such as sorghum hybrids, will have value for new stover, grain, and root growth are most af regions moving toward hybridization. fected by hybrid vigor in the first, second, The time lag for maize was 20 or more and third or final stage ofthe life cycle, re years, whereas sorghum, benefitting from spectively. The heterotic response ap farmer experiences with maize, took only pears to increase the general metabolic ef four years to gain near universal accep ficiency of the hybrid plant. tance (Figure 4). Russell (1991) points out that although maize had a relatively Generally, the preponderance of evi slow acceptance nationwide, in Iowa, the dence supports the view that heterosis in leading area of maize production, accep-
25 100
90
80
70
1If 80
i 50 'I I 4() 1------+----, ------,------
80
iO
10
0 1!l3O 1955 1980
Figure 4. Percentage of maize and sorghum planted with hybrid seed in the United States from 1930 -1960. tance did occur within a few years as with vestment in product improvement which sorghum. may be less critical to the hybrid industry short-term, assuming in-country national A secondary benefit from the accep programs are effective. India illustrates, tance of hybrids, which is critical, relates perhaps as well as any country, that hy to private sector involvement. The extra brids combined with private sector invest value and cost applied to seed not only en ment can allow for self sufficiency in courages better management practices food, having increased production from but allows for a private sector input into 50 million tons in 1950 to 200 million in all aspects of the seed industry. The op recent years. portunity with hybrids for annual seed re placement, as well as proprietary owner In 1966, G.W. Burton speculated that ship, further encourages private invest the next century will see F I hybrids of ev ment. Benefits from private sector ery major crop in general use throughout competition (Maunder, 1998) will (a) the world. The greater yield potential, ef better assure an adequate supply of seed, ficiency, and ability of the outstanding FI (b) lead to improved seed purity and ade hybrids to withstand stress when com quate germination, and (c) increase in- pared with varietal cultivars suggest a real
26 opportunity to increase feed and food pro Eastin, J.D., T.E. Dickinson, D.R. Krieg, and AB. Maunder. 1983. Crop physiology in dryland ag duction. As the world increases in popula riculture.E. Dregne and W.O. Willis (eds.) tion, and standards of living increasingly Dryland Agriculture. Am. Soc. Agron., Madi demand more protein through meat in the son, WI.pp. 333-364. Ejeta, G. 1983. Current status of sorghum improve diet, Burton's thoughts along with current ment research and development in the Sudan. agronomic developments seem all the Pages 17-18 in Proc. of Hybrid Sorghum Seed more prophetic. Hybrids will be used and for Sudan, G. Ejeta (ed.). Hallauer, A.R., W.A Russell, and K.R. Lamkey. accepted in developing countries if hy 1988. Com breeding. In Com and Com Im brids with seed of high purity, adequate provemeft. G.F. Sprague and J.W. Dudley, (eds.) 3r Ed., pp. 469-564. Am. Soc. Agron., germination and yield performance be Madison, WI. come readily available. Maunder, AB. 1997. Role of private sector. Pages 605-612 in Proceedings of the International List of References Conference on the Genetic Improvement of Sorghum and Pearl Millet. . Maunder, AB. 1998. Logistics of seed productIOn Burton, G.W. 1966. Meeting human needs pp. and commercialization. In J. Coors (ed.) 391-408. In K.J. Frey (ed.) Plant Breeding. Uni Heterosis in Crops. CIMMYT Pub. Mexico. versity Press, Ames, Iowa. Plucknett, D.L. 1991. Saving lives through agricul Castleberry, R.M., C.W. Crum, and C.F. I 27 Rapporteur's Report Botorou Ouendeba Opening Remarks Day 1 - Session 1 The Director General of INRAN con erty, agriculture must be modernized. gratulated the organizers. He emphasized Seed activities must be developed to fully the importance of improved seeds and the take advantage of hybrids. We must have need to pursue collaborative research be a sustainable seed sector, and Niger is tween INRAN, INTSORMIL, and willing to consider any proposal to stimu ICRISAT. There are notable results in the late the use of hybrids in the country and area ofvarietal development: INRAN has in the sub-region. produced a catalogue of all its varieties. However, more funding is needed for mil The session president welcomed all let and sorghum research, as well as for participants from Africa, Europe, Asia developing seed production and market and the US. Hybrid development has mg. made significant progress in Niger in re cent years; these achievements must be Two key points must be noted: (I) seed supported so that many farmers in West is different from grain; (2) a seed unit is Africa can take advantage of the phenom necessary, as well as hybrid development enon of heterosis. This workshop was by other NARS. made possible because of the persistence ofIssoufou Kapran, John Axtell, and Lee Following the Director General of House. INRAN, the minister of agriculture gave the opening speech. He mentioned how Millet and sorghum are the two most proud Niger is to have so many leading important crops in Niger and the scientists in attendance at the Workshop sub-region, comprising 70% of all cereal and their importance to the Workshop. To production in West Africa. Although cul reach the goal of food self-sufficiency in tivated land devoted to millet and sor Niger, millet and sorghum must be im ghum increased by 50%, yields remain proved, since they cover more than seven low at 600 kglha on average. The oppo million hectares and provide the staple site was observed in India. Mr. Barwale foods for its population. The government will later on in this workshop provide in of Niger supports INRAN's effort in vari formation on how hybrid development etal development and seed production. evolved there. Collaborative research has permitted the development of high yielding millet and We should insist on the importance of sorghum hybrids. To fight against pov- grain transformation as well as seed pro- 28 duction. This meeting should focus on The second group will discuss transfor ways to increase sorghum production and mation and other uses for millet and sor the use of hybrids. ghum. This group will be organized by Bruce Hamaker, A. Ndoye, Moussa Dr Lee House gave some details about Oumarou, and Adam Aboubacar. the group working on seed production. They will focus on the following: Bruce Maunder made a presentation entitled "Why Hybrids." He clearly Background history of seed pro pointed out several advantages of hy duction in Niger; brids, including high yield, earliness, de Current situation, and sirable morphological traits, and a good Future possibilities in relation to reaction to both drought and soil fertility the availability of hybrids problems. The yield advantage of hybrids versus open-pollinated varieties varies At the end of their work, the seed pro between 28% and 54%. Examples were duction group will present a strategy to given of sorghum, millet and corn in other ward a sustainable production and com countries. mercialization of hybrid seed to the ple nary session. 29 Monday - September 28 - Session 2 Hybrid Seed Experiences in Developing Countries Opening Remarks Moderator - Lee House, Consultant, Seed Production ment and process the seed. We processed "Yes, sir, we would be able to the seed speedily. Ladies and gentlemen, provide that amount to you." that was the best yield of processed seed "We want it tomorrow. of maize I obtained in my life: 33 quintals "OK, sir." (132 bushels) of processed maize seed per acre! Within two hours, we had hired a truck and instructed our representative to de The seed was ready. We stored it in a liver the seed to the officer the next morn warehouse, and we looked for customers. ing. It was delivered and reported. What a We contacted several groups and tried to day it was! Really a day of achievement convince them of the usefulness of plant and celebration. Our credibility as seed ing good quality hybrid seeds. In the producers and progressive farmers was meetings, there were lots of questions and established. lots ofresistance to paying 3 to 4 times the price of the grain for the hybrid seed. The ICAR's All India Sorghum Im provement Programme, with assistance One good thing that we had done prior from The Rockefeller Foundation, also to planting the seed crop was import cer developed new sorghum hybrids. That tain quantities of hybrid seed from same rabi (dry) season, we planted a dem Gujarat. We sold it, then helped people onstration of2 Jowar (sorghum) hybrids, plant that seed to produce a commercial CSHI and CSH2 along with the local crop. Thus, the crop was grown in a num check. The earheads of the local check ber of places and we obtained very en were big and compact and looked very couraging results. These demonstrations impressive. Comparatively, the hybrids created some curiosity and interest among looked unimpressive, even though the the farmers and extension officers. We plant growth was very uniform with ex used to celebrate if we were able to sell cellent plant populations. one sack of seed a day. One sack is equal to 90 kg, enough seed to sow 6 hectares We harvested the demonstration hy worth only Rs. 120 at that time (i.e. 3 USD brids, and everybody was surprised to see today). that the hybrids had tremendously high yields compared to the local check. Later, One afternoon, I received a phone call I had a meeting with Dr. House in Delhi. from an Agricultural Officer from the He informed me that the first sorghum hy neighboring district asking me "Mr. brid CSHI (Kafir 60 x IS84) was likely to Barwale, do you have hybrid maize be released at the next workshop. As a re seed?" sult, The Rockefeller Foundation wanted to see some seed produced that year in or "Yes, sir," der to support good demonstrations dur "How much you have?" ing the next year. "We would be able to fulfill your demand." Dr. Wayne Freeman and Dr. Lee "I want one truckload." House were helping this sorghum pro gram by assisting with field operations. 34 ment and process the seed. We processed "Yes, sir, we would be able to the seed speedily. Ladies and gentlemen, provide that amount to you." that was the best yield of processed seed "We want it tomorrow. ofmaize I obtained in my life: 33 quintals "OK, sir." (132 bushels) of processed maize seed per acre! Within two hours, we had hired a truck and instructed our representative to de The seed was ready. We stored it in a liver the seed to the officer the next morn warehouse, and we looked for customers. ing. It was delivered and reported. What a We contacted several groups and tried to day it was! Really a day of achievement convince them of the usefulness of plant and celebration. Our credibility as seed ing good quality hybrid seeds. In the producers and progressive farmers was meetings, there were lots of questions and established. lots ofresistance to paying 3 to 4 times the price of the grain for the hybrid seed. The ICAR's All India Sorghum Im provement Programme, with assistance One good thing that we had done prior from The Rockefeller Foundation, also to planting the seed crop was import cer developed new sorghum hybrids. That tain quantities of hybrid seed from same rabi (dry) season, we planted a dem Gujarat. We sold it, then helped people onstration of2 Jowar (sorghum) hybrids, plant that seed to produce a commercial CSHI and CSH2 along with the local crop. Thus, the crop was grown in a num check. The earheads of the local check ber of places and we obtained very en were big and compact and looked very couraging results. These demonstrations impressive. Comparatively, the hybrids created some curiosity and interest among looked unimpressive, even though the the farmers and extension officers. We plant growth was very uniform with ex used to celebrate if we were able to sell cellent plant populations. one sack of seed a day. One sack is equal to 90 kg, enough seed to sow 6 hectares We harvested the demonstration hy worth only Rs. 120 at that time (i.e. 3 USD brids, and everybody was surprised to see today). that the hybrids had tremendously high yields compared to the local check. Later, One afternoon, I received a phone call I had a meeting with Dr. House in Delhi. from an Agricultural Officer from the He informed me that the first sorghum hy neighboring district asking me "Mr. brid CSHI (Kafir 60 x IS84) was likely to Barwale, do you have hybrid maize be released at the next workshop. As a re seed?" sult, The Rockefeller Foundation wanted to see some seed produced that year in or "Yes, sir," der to support good demonstrations dur "How much you have?" ing the next year. "We would be able to fulfill your demand." Dr. Wayne Freeman and Dr. Lee "I want one truckload." House were helping this sorghum pro gram by assisting with field operations. 34 The Rockefeller Foundation sent equip excellent crops. Suddenly, it attracted the ment, trained human resources, founda attention of everybody who was con tion seed, and thimet (a systemic insecti cerned with agricultural production. cide with an obnoxious odor). Actually, Dr. Lee House had to travel with this We received tremendous support from chemical from Pune to Jalna, 190 miles the then Chief Minister of Maharashtra, distance, and had to inhale the chemical. the late Shri Vasantroji Naik, who pro moted the seed and the idea in a great way. Four acres of Hybrid CSHI seed pro The same summer we produced the first duction was planted. We did not know pearl millet hybrid, a product of the how the parents would behave in a differ ICAR's All India Coordinated Millet Im ent environment and different season. provement Program. By this time we Therefore, we planted the male 3 times at were experienced growers. We were 4 day intervals before (and after) the fe trained how to find a shedder. We used to male was planted to ensure that there take black slates in our hand to see would be a longer pollen availability in whether a doubtful plant was a shedder or order to achieve a good nick. a male sterile. The purity of the founda tion seed was a little doubtful. However, Ladies and gentlemen, I do not want to the need was urgent so good-quality hy take your valuable time by providing you brid seed had to be produced, which - for minor details. However, I would like to example - involved an extra effort in re say that it was a successful production and moving rogues from the parents before we got 8 quintals per acre of hybrid sor they could have a damaging effect. ghum seed. Thus, for the first year alone, the country had 32 quintals ofthe first sor By this time I was further convinced ghum hybrid to distribute throughout the and committed to the program: a na country. tion-building program, agricultural pro duction program, and a program that Dr. House, I still have on record the would bring prosperity and respectability long letter you had written describing the to the farming community. I realized that I sort of risk I was taking, that it may be a alone might not have been able to achieve total loss, that I may not be able to market the objectives as the task was gigantic and the seed, but promising me all support to we started contracting seed production make ita successful program. Although with small farmers, progressive ones who there was a risk, we decided to move for were prepared to follow instructions, had ward. The extension agencies in some reasonable resources to irrigate and were states were promoting the hybrid sor responsive to the idea and instructions. ghum. The hybrid sorghum grown was qualitatively poorer than the best local va We appointed production assistants, rieties, but it was a shorter duration than generally agricultural graduates or post the local varieties and those were the graduates, and provided them motor years we had severe droughts. While all bikes. These assistants regularly use them local varieties failed, and did not even to visit these growers and report on the produce earheads, the hybrid produced progress to the officers responsible for 35 production. Through this process, we un construction of a small processing plant derstood the problems of the growers and (20' x 120'). USAID provided processing solved them as quickly as possible, if not equipment to the government of India, immediately. which was provided to the National Seeds Corporation and to private seed produc We were growing seed in the kharif ers. We got our share and erected a pro (monsoon) season (June - October) and cessing line while our building was yet at had to face the problem of grain mold if lintel level high, since we needed to pro there was rain at the time of harvest, cess. We also rented warehouse space in a which would generally affect the storage, building close to our previous site. In our germination, and viability of the seed. As processing plant, the storage space was a result, we began looking into dryseason only 900 sq. ft., which we quickly real seed production. After looking into other ized, was very small. production areas, we moved to the rice fallows in Andhra Pradesh. Meanwhile, Ladies and gentlemen, in a develop we had faced problems of crop ment economy, in the beginning there nonuniformity, which we solved by puri were two options: create a large capacity fying the parental material and having di and have an extended period for utilizing rect quality control over this material in all the capacity, or create a small capacity the production fields. Once the growers and - if you have more work - work over understood that it could be a profitable time and do the job. We adopted the sec crop for them, we did not have difficulty ond method and created a small facility. in converting the entire village into a seed Although we still use this facility for pro production village. cessing our foundation seed, we do have newer equipment. This is the story of the beginning of Mahyco. We did not look back. We con As you will have an opportunity to ask tinuously produced as much seed as pos questions, I can then provide more infor sible. While we also tried to establish our mation and details market by selling seed with our brand name, there was government support Fortunately for us, generally we did not available to procure the seed produced as have unsold seed in the first 3 to 4 years, well as extension agencies to distribute but thereafter we did have carryover seed. the seed. In the meantime, The Because of poor storage we lost seed due Rockefeller Foundation assessed our to insect damage, fluctuating humidity, need to have a processing plant with and loss of viability/germination from equipment and storage space. The engi time to time and wrote off good quantities neering officer responsible for processing of seed. This made a dent in our profit and consistently helped and visited in order to loss account. assist us in establishing a processing plant. I referred to the National Seeds Corpo ration (NSC) earlier. NSC was estab We took a one-acre plot in the Jalna lished by the government ofIndia, with an Cooperative Industrial Estate and started objective of providing foundation seed 36 and services ofthe seed industry. It started themselves. They set priorities: 1) them in earnest in 1963. Before that, foundation selves, 2) state seed corporations, 3) state seed production and relevant services governments, and 4) private seed produc were being provided by the ICAR. The ers. This sort of situation created uncer Rockefeller Foundation also became part tainty among the producers about the of the system to provide services. overall program and produced a setback. Even though it was unintentional, NSC Establishment of the NSC was very started claiming that they had a monopoly critical in the early '60s. It produced foun for foundation seed production. Conse dation seeds. It also provided voluntary quently, producers faced problems with certification services and a support opera quantity and quality of foundation seed. tion for seed producers. From my 35 years When the situation became very serious, of experience with the Indian seed indus the state governments responded by de try, I have come to believe that while the claring that NSC did not have a monopoly NSC provided great support to the Indian and started distributing breeders' seed in seed program, it also made mistakes selected crops to competent producers, which should have been avoided. I shall who were then entitled to produce foun comment on this a little later. dation seed themselves. A seed law was enacted in Parliament India is a large country consisting of in 1966. Consequently, official seed cer several states. Agriculture is a state matter tification came into existence in 1968, and the state governments were not happy and NSC was nominated as the official that NSC served as a certification agency. seed certification agency for the entire Thus societies were formed under the country. While it facilitated certification chairmanship of the State Agriculture procedures, it also generated bureaucratic Commissioners to carryout certification systems, which affected the efficiency of work in the respective states. NSC did ev seed operations. A few years later unfor erything to prevent this from happening, tunately, NSC moved from being a sup but the states prevailed. Except for a few port agency for the seed sector to being a union territories, states - basically societ commercial seed marketing agency in ies under the chairmanship of the State competition with the industry they were Agricultural Commissioners - became trying to encourage. NSC started insisting certification agencies. Several sources for that one side of the seed bag should be foundation seed also came into existence. printed the same way as NSC. So it acted This situation improved the quality and as an advertisement for NSC. This went quantity of the foundation seed, and the against the spirit of the establishment of availability of foundation seed was no the NSC. NSC was to act as a foundation longer a constraint for planting seed pro seed agency and a provider ofservices. As duction programs. soon as it became a marketing agency, it established different objectives and thus NSC and The Rockefeller Foundation became partial. It also started practicing provided remarkable services for training discrimination in distribution of founda seedsmen in certification, quality control, tion seed to the producers other than seed testing, seed processing, seed equip- 37 ment maintenance, and demonstration reaucratic hurdles. This also helped make plots, among other topics. USAID also seed available to the farming community supported some of the activities. What on time. you have today is in part a product of this support. I gratefully acknowledge the Subsidy has always been a matter of contributions made in this regard by The great controversy in our mind. In the In Rockefeller Foundation, USAID, the dian program, subsidies were granted for government of India, Indian Council of several kinds of improved or hybrid seed Agricultural Research and state govern varieties. Even though initially it helped ments directly or through their undertak to promote the program, there was no way ing like NSC and the state seed corpora to stop it once it started. Interested per tions. sons became more and more greedy, and every elected representative took it as a Seed certification created general stan part ofhis political exigency to have max dards of quality and a base for independ imum subsidies in his area. Subsidies also ent vouching quality undertaken by the created discrimination between NSC, producer. One thing should not be forgot state seed corporations, and private seed ten: certification provided minimum producers. In most of the cases private standards. Thus quality of seed differed seed producers were not able to get an ad from producer to producer; a number of vantage from the subsidies. If subsidy them observed stricter and better stan was provided by the state governments, dards than required by the certification state seed corporations were the first ben process. They put more emphasis on the eficiaries. If it was by central government, quality of foundation seed, training of the then the first beneficiary was the NSC. company professionals and training seed Therefore, this was an incentive for NSC growers, and rigidly implemented the and state seed corporations. Private pro quality control system. I remember the ducers had to market their seeds on their day Mahyco had to reject a large hybrid reputation and on the quality of the seed. I maize seed plot on its chairman's farm. am glad to tell you that the farmers' first This sort of decision-making process cre preference was always private seed pro ated greater confidence in the quality of ducer's seed that had a reputation for seed produced by independent organiza quality. tions. By disseminating information and con The Indian Seed Law as it was intro ducting comparative demonstrations (us duced and passed by the Parliament is an ing local practices and improved seed vs. ideal seed law. While it provided compul improved practices and improved seed), sory truthful labeling of notified varieties the extension agencies of the government or crops, certification is voluntary. Thus, did do a good job in carrying a message to those producers who were able to estab the farmers of the importance of quality lish their reputation gradually did not seed and improved agronomic practices. bother with seed certification. Conse This did create good demand. quently, they avoided delays in the pro cess of certification that were due to bu- 38 I have always thought that the Indian from all over the world. We hired addi farmer, (and I am sure, the African tional scientists rapidly and also added farmer) is a wise, hard-working and hon pearl millet to our research crops. We est entrepreneur. I feel he has always were investing money by reinvesting the looked not to subsidies but to ultimate profits as well as increasing our capital. profit that he could make from the crop. This was not appreciated by all our part Surely he needs support in terms of credit ners/shareholders. We went through hard (if the money is not available) to purchase times, but we stuck to it. agricultural inputs, such as seed, fertil izer, and insecticides. With the evolution ofhybrid cotton, we added hybrid cottonseed production to I do not have a presentation here of the our program. As soon as we learned about first 15 to 20 years ofMahyco; however, I the availability of genetic male sterile, we do have a graphic presentation of the 14 went for the germplasm and the establish years from 1983-1997 of Mahyco which ment of a hybrid cotton seed program, should give you an idea of how the private based upon genetic male sterile and later enterprise has performed in the Indian en on the CMS/Restorer System. vironment. The public breeding program was quite I was very fortunate to visit the U. S. in comprehensive and was very well funded 1965 and again in 1966: the first time in with the help of The Rockefeller Founda the Farm Leaders' Exchange Program, tion and several other agencies. For quite and the second time as a member of the some time we did not produce anything Seed Review Team appointed by govern from our research programs. I remember ment of India. This team was established that we identified a few good sorghum hy to make recommendations on basic seed brids. Those were really good - good policies. The recommendations of this yielding, good grain quality, and giving committee became a key document for good results in our trials. However, when several decades to come. This also pro we went into commercial marketing we vided me the opportunity to visit different realized that because sorghum had black seed companies, research facilities, re glume it was not accepted by the market search institutions, and universities in and - thus - by the farmer. This was a big Sweden, Holland, England, United setback, but we kept on working and had States, Japan and the Philippines. Thus, good research products that really right from the early years ofMahyco, with changed the entire profit scenario of the my constant contact with the Indian Agri company. We also managed the company cultural Research Institute, Mahyco was economically. Costs were absolutely un to establish its own research program as der control, expenditure within the in rapidly as possible. In 1966, just two come limits, and minimum borrowing. years after Mahyco was established, we started the first research program. We Speaking of banking, we had been suc hired a Ph.D., with the help of The cessful in persuading our bankers to fi Rockefeller Foundation, and started col nance us as well as finance our growers. It lecting maize and sorghum germplasm was not a large operation, and things were 39 going on satisfactorily - as we had been have a policy where every action of the successful in gaining the confidence of seedsman is protected as long as he has our bankers. However, one day our bank operated sincerely and honestly. Even if ers had to withdraw our advances, be he has done wrong, he needs protection - cause the Reserve Bank of India did not unless he has done it with dishonest inten discriminate between seed and food tions. We must trust him at least as much grains. They accused our bankers for as an administrator, a scientist, or a gov making advances against food grains, ernment servant is trusted. which was against the government policy given that food was in short supply. It What I am trying to emphasize is that to took almost six months for us to getthe or start and have a successful program you der changed. The Reserve Bank of India need to have good and committed entre was kind enough to recognize seed as a preneurs. Good understanding and sup different product than grain and advances port from within government administra were made to seed enterprises on the tion and a good political will is necessary strength of seed stocks. The growers who for success even when it is not exactly borrowed were treated to priority ad what they would want. vances. They began receiving preferential treatment. For the seed program in our state, we always had the patronage of our Chief I have always believed that if we have a Minister. He never let us down. We had will it is always possible to find a solution. good administrative support. We had We need to have a will and need to be de good linkages in the central government termined to solve the problem. This is the with the agricultural ministry. We had beginning of the solution to the problem. good support from The Rockefeller F oun As you listen to me, I want you to under dation, who always constructively con stand that over all of these years (or, at tributed problem solutions to the govern least the first 25 years) I felt that a sword ment and enterprises. We had an excellent was constantly hanging over my head, team in Delhi consisting of our then Food and I did not know when it would fall. In and Agriculture Minister, Bharat Ratna spite of that, with great optimism, the In Shri C. Subramaniam, the then Secretary dian seed industry has progressed. ofAgriculture, the late Mr. B. Sivaraman, and Professor M.S. Swaminathan, Direc As I reflect on the past, I would like to tor IAR! and later on Director General, advise those countries who are now trying lCAR. to establish a seed industry: do not keep a sword hanging over the heads of private The fortunes of the Indian seed indus enterprises. They should feel reasonably try have fluctuated many times in the last secure so that they can converge their en three and a half decades. If I am asked to ergies on their enterprises, and through give reasons for this, it would be govern that, on services of the country's agricul ment policies, indecision and lack of clear ture rather than waste their energies and objectives. Surely, as you know, there are valuable time solving problems that slow always several factors that make an enter program development. Thus, we need to prise/industry successful: good planning, 40 good financial management, good peo The current scenario in India has ple, good marketing arrangements, and changed quite a bit in comparison to the overall execution. Several seed enter seed situation in the 80s and before. A prises failed in the late 70s and 80s due to change of policy was promulgated by the government policies and! or poor manage government of India whereby multina ment. Thus, I believe it would have been tional/foreign seed companies were wel possible to have a more growth of the In comed to utilize their expertise and estab dian seed industry if we had very clear lish themselves in India. In the last five goals and strong determination. years, particularly, we see a positive im pact from this. Even though initially these Any developing country is always companies did not have products for In looking for opportunities to earn foreign dian conditions, gradually they improved markets, as it always needs foreign ex and now are providing tough competition change. As the Indian seed industry pro to the existing Indian seed industry. The gressed and increasing emphasis was current seed industry will have difficulty placed on exports, the Indian seed indus facing the competition unless it takes the try tried to get into international markets. situation very seriously and becomes Although they first tried as contract grow quite aggressive in research and market ers, they later tried to collaborate with ing. leading foreign enterprises. As research and development were established and in In the last 10 years, the scientific situa creased, these enterprises tried to market tion also has changed with the emergence their products in the same ecological ar of new scientific tools (e.g., biotechnol eas of the world outside India. From zero ogy: molecular, cellular and other new export level in the 50s, India is now ex scientific methods based upon DNA). We porting reasonable quantities of con are seeing now the products out of this tracted seeds and seeds derived from the new science: products that are benefiting research of Indian seed companies. Cer the farmers in terms of insect resistance, tain multinational companies are also try fungi resistance, virus resistance, etc. An ing to export seeds into African or South other scenario in the global seed industry East Asian countries. is that in the developed world, the seed in dustry is becoming more concentrated. A There is always going to be an opportu number of chemical/pharmaceutical nity to produce and export seeds particu companies have become involved in the larly where it is labor intensive as it is very seed enterprise in the last 25 years, and expensive to produce such seeds in devel particularly so in the last 10 to 12 years. oped parts of the world. Besides this, seed Four or five companies have emerged as companies are willing to share knowledge leaders in biotechnology and have ac and train technical people and growers in quired interests in the seed business as a producing seeds. This gives a direct expo vehicle to deliver the new findings. sure to the seedsmen in developing coun tries. Thus, the situation today is very differ ent than when Mahyco or I got started. Everybody must realize this. However, 41 the basic components still remain for a the Indian seed industry. I do not want to successful enterprise. Good technology, further extend it. However, I welcome honest work, good management, and a questions that you feel would further clar will to serve are the essential aspects. ify the situation. Ladies and gentlemen, I think I have Thank you very much for providing me talked a lot about Barwale, Mahyco, and this opportunity to be with you. 42 Zambian Seed Industry - History and Experience Bhola Nath Verma and Winter M. Chibasa Historical and political developments improve farming conditions for this in southern Africa, including Zambia, group, but it has remained difficult to do have led to an emergence oftwo distinctly so. different farming sectors. To one sector belongs the large-scale commercial farm The number of commercial farmers has ers, mostly of European origin, who own tended to vary from country to country. At large tracts of land, ranging from 100 independence in Zimbabwe in 1980, there hectares to 20,000 ha. These farmers are were about 5,000 such farmers. By 1996, situated in the country's most productive the number was reduced to 4,000 (Eicher agroecological region, often on good fer and Kupfuma, 1997). Zambia, which was tile soils. then regarded primarily a copper produc ing country, had only about 100 commer The facilities for farming, including cial farmers at independence in 1964. road and rail transport networks, electric This number by 1997 had increased to ity, irrigation, and services like credit and 740, of which 20 are corporate entities. marketing, including seed supplies, have been better developed in the area of their According to present estimates in Zam domination. The commercial farmers fol bia, there are approximately 603,400 farm low improved agricultural practices, in households, comprising 459,000 cluding high use of fertilizers, chemicals, small-holder subsistence farmers (of less and other inputs. Some of these farms than 10 ha); 119,200 emergent commer may have thousands of hectares of land cial farmers (ranging between 10 to 20 under center-pivot system of irrigation ha); and 25,200 medium-scale commer and may own airplanes for use in agricul cial farmers (ranging between 20 to 60 ture. ha). The majority (75%) belong to the other The last two categories together consti extreme sector. This sector consists of re tute what is now being termed the inter source-poor traditional farmers whose mediate sector, which is gaining strength. holding sizes are below 10 hectares. These farmers depend on their family for Maize in the Agriculture of labor. Many of them hardly own any Southern Africa equipment, and at most, have a locally de signed hand-hoe. They have limited ac White maize was introduced, pro cess to agricultural credit, transport, mar moted, and cultivated in southern Africa keting facilities and inputs, including the by European settler farmers for its pro availability of seed, fertilizer and chemi cessing ease to feed local people em cals. In the post-independence era, the ployed in the mining sector and/or for ex government in southern Africa tried to port to Europe for the starch industry. 43 However, in less than two generations, it The farmers also persuaded the gov has become the most popular food among ernment in Zimbabwe to establish a net the indigenous people. work of agricultural research stations. This network was further expanded in the After independence, the wealthy gov region after the creation of the Federation ernment based on the rich mining sector, of Rhodesia and Nyasaland in 1953. The introduced policies for promoting maize Federation consisted of Zambia (then by providing subsidies for inputs (seed, Northern Rhodesia), Zimbabwe (then fertilizers and credit) and grain market Southern Rhodesia), and Malawi (then ing, both for the production of maize and called Nyasaland). for its consumption. These policies made maize not only a main national food but As early as 1932 in Zimbabwe, some of also a major cash crop, displacing all the world's leading maize breeding work other crops, even in the areas that were not was underway. The breeders were the late suitable for the cultivation of maize. As a Harry Arnold and Allan Rattray. In 1949, result, in Malawi, for example, maize ac they released SRI, the first maize hybrid. counts for 90% of the cultivated area The hybrid covered 20% of total maize (Smale and Heisey, 1997). In Zambia, at area produced by commercial farmers in one stage, it accounted for 70% of the to the first year of its release. But the real tal cultivated area. breakthrough from their work came with the release of SR52 in 1960. In Zambia, maize was favored by the government. Until the late '70s, it was the SR52 is cited as the world's first single only crop that was purchased by cross maize hybrid released commer NAMBoard (National Agricultural Mar cially. It is a white grain hybrid of long keting Board), the sole agency estab duration with very high yield potential - lished for the marketing of agricultural ideally suited to commercial farmers, liv produce and distribution of inputs on a ing in high-potential farming areas. The pan-territorial price basis. hybrid became the region's most famous crop variety, which is still in production Commercial farmers in southern Af after 37 years after its release. rica formed a strong political and eco nomic force prior to independence. They The release ofSR52 provided political, influenced the government at that time to institutional and technical foundation for help develop and nurture one of Africa's Zimbabwe's first "green revolution" by best national infrastructure of roads, con white farmers in 1960. Five years later, it centrated primarily in the areas where started in Asia. SR52 gave 46% more they had settled. They formed several grain yield than the best contemporary powerful organizations/associations, in open-pollinated variety, Southern Cross. cluding the Seed Maize Association of Zimbabwe, formed in 1940 and the Crops Within eight years of its release, this Seed Association of Zimbabwe in 1955. hybrid covered over 75% of maize area planted by commercial farmers in Zimba bwe (Eicher and Kupfuma 1997). The re- 44 gional Federation further bolstered the Until the formation of the Federation, expansion in neighboring countries. Zambia was dependent on Zimbabwe for its seed needs. After the breakup of the As the commercial farmers' dryland Federation in 1963, Zambia had to start its crop, tobacco, ran into problems, more own industry. Since its establishment, the maize hybrids (mainly three-way crosses) Zambian seed industry has undergone were developed to suit the sandy soils. major changes which are described under Such short duration hybrids also hap the following three separate periods: prior pened to meet the requirements of to the establishment of the Zambia Seed small-scale farmers and for drought prone Company (ZAMSEED); ZAMSEED's areas. domination; and liberalization after 1991. The above development represents an Industry Prior to the excellent example of a technological Establishment ofZamseed spillover from the resourceful commer cial farmers to resource-poor small-scale After the breakup of the Federation, farmers. Subsequently, and especially af Zambia with only three commercial farm ter independence, when the governments ers who had experience in seed produc decided to pass on the benefits of this tion received eight pounds of seed of pa technology to small-scale farmers and ur rental lines of SR52. The government ban dwellers, a much wider adoption of formed a Seed Advisory Committee un maize hybrids occurred throughout the re der the Ministry of Agriculture, Food and gion. Contribution of small-scale farmers Fisheries (MAFF) and set up a Seed Ser to total maize area and production in vice Unit within its Research Branch. The creased tremendously (65% in Zambia unit was given the responsibility of seed and 70% in Zimbabwe). Zambia's maize control and certification and production production increased four-folds from the of basic seed. Commercial farmers, some early 60's to the late 80's. of whom had experience in seed produc tion, formed ZSP A (Zambia Seed Pro Development of the Seed Industry ducers' Association). ZSPA produced certified seed under the supervision of the Technological advances in maize, and Seed Services Unit but was not given ac formation of strong commercial farmers' cess to parental lines of SR52 until 1978 organizations in Zimbabwe led to the de because of the wary relations with the velopment of the seed industry. In the be government at independence. ginning, farmers formed a cooperative NAMBoard (National Agricultural Mar called Seed Co-op. Seed Co-op was later keting Board), created by the govern converted into a public company, Seed ment, was made sole buyer and seller of Co. Zimbabwe now has a separate com the seed. pany for export, Certseed International. Recently, Seed Co. has gone into partner During that time, there was a disruption ship with giants like DeKalb Genetics among maize breeders in the country. Corporation of the United States. This resulted in inconsistency of objec tives and priorities in breeding. Poor 45 maintenance of parental stocks of SR52 SW and Swedfund were invited to par caused deterioration in the Zambian ver ticipate as shareholders in the company sion of SR52, which gave 15-20% less for the purpose of providing technical yield compared to the original hybrid know-how and to ensuring access to inter from Zimbabwe. By the late 70's, several national seed markets and techniques of donors (Sida, USAID and F AO) were in high standard. Within the framework of volved on different aspects of maize re the development cooperation between the search in Zambia. governments of Zambia and Sweden, and working through a technical assistance Period ofZamseed's Domination agreement with MAFF, Sva10f provided the company with technical assistance Zambia Seed Company Limited services. (ZAMSEED) was formed in March 1974 under the then Rural Development Cor The permanent Secretary of MAFF poration (RDC) but remained basically was the chairman of the Board of Direc inactive until1981. tors ofZAMSEED. In line with the politi cal and economic environment, seed ZAMSEED in its present form became prices (especially those of maize seed) operational in January 1981 following a were controlled by the government, al decision by the Government of Zambia though the company was actually allowed through the Ministry of Agriculture and to set its own price to make a certain per Water Development (MAWD) to pro centage of profit. mote the development of the Zambian seed industry. The company was given Seed production, processing and distri exclusive rights to the growing, process bution were ZAMSEED's major respon ing and distribution of all kinds of seeds. sibilities. ZAMSEED purchased a In exchange, it would get - free of charge 1,200-hectare farm for seed production. - breeders' seed from the Government. Certified seed was produced by contract ing large-scale farmers, who had to be The shareholders of the company were members ofZSPA (150-200 growers). the Government (40%) through RDC/ZIMCO and presently the Ministry Most seed growers are located in Zam of Finance and Economic Development bia's maize growing areas. This made the (MFED); Zambia Seed Producers' Asso growing of maize seed easier but other ciation (ZSPA, 20%); Zambia crops, especially sorghum and millets, Co-operative Federation Limited (ZCF, suffered from lack of interest and experi 20%); Sva10f Weibull AB, (SW 10%); ence among farmers. At a later stage, and Swedfund International AB (10%). ZAMSEED started to use small-scale SW is the largest seed company in Swe seed growers to produce seed of some den, while Swedfund International, self-pollinated crops. which falls under Sida, is a venture capital Swedish institution that encourages The geographical expanse and poor in Swedish firms to invest in developing frastructure of the country posed a great countries by providing equity risk capital. challenge to ZAMSEED with regard to 46 the distribution of seeds. ZAMSEED, dertaken by PCUs. Experience over time therefore, acted primarily as a wholesaler, revealed however that even these institu except for its one retail shop at its head tions were beset with the same shortcom quarters. ings that beset PCUs, forcing the govern ment to look elsewhere. In the pre-1991 era, ZAMSEED went through different types ofmarket involve 1995 to 1996: The government turned ment, which was affected to a large extent to Cavmont and SGS Zambia Ltd. in the by existing intermediaries at the time and hope that being well-standing financial their capacities as retail marketing outlets institutions and holding quality inspec for both seed and other agricultural in tion authority of international repute, puts. This can be chronicled as follows: these institutions would finally be the an swer to loan recovery and accountability. 1981 to 1989: Seed was marketed This was not to be, and the same problems through the Provincial Co-operative Un surfaced again forcing even this arrange ions (CPUs) which were quasi-govern ment to be abandoned. Billions of ment institutions charged with providing Kwacha remain unrecovered in input agricultural inputs to peasant farmers. loans. Problems of debt recovery and other inef ficiencies eventually made these interme ZAMSEED sales of different com diaries unsustainable. modities are given in Table 1. It must be noted however, that the sale volumes do 1990 to 1994: In an attempt to improve not necessarily represent market poten the financial discipline and loan recovery tials. Most commodities other than maize rate from peasant farmers, the govern have not been effectively served. During ment turned to CUSA (Credit Union and the periods that have been chronicled Saving Association), LIMA Bank, and above, the pattern of maize seed sale re ZCF IFS to assume the role previously un- veals the following: Table 1. ZAMSEEED Seed Sales (in tons) of Different Crops. Year Maize Soya Wheat Sor- Sun- G/nuts Beans P.millet Rice Cow- Pasture Potato Vege- Flower Paprika ghum Flower peas tables 1981/82 8,655 224 132 412 0,0 16.1 0.0 0.0 o 1982/83 8,350 315 63 15 211 44.0 26.0 0.0 0.0 o 1983/84 6.200 333 179 23 158 0.7 57.0 0.0 0.0 o 1984/85 6,890 749 686 71 192 12.4 16.0 8.0 0.0 1,152 1985/86 7,303 991 267 42 48 12.2 16.0 12.5 0.0 1,354 1986/87 8,864 1,048 282 68 63 8.9 40.0 8.6 0.0 1,561 1987/88 10,404 1,030 690 159 III 1.3 57.0 28.0 0.0 1,757 1988/89 13,604 1,959 1,300 139 167 5.5 64.0 32.0 0.0 1,889 o 1989/90 13,343 2,587 1,212 40 80 7.8 71.0 2.5 o 0.0 1,409 31.5 o 1990/91 7,189 2,474 952 22 66 8.1 18.0 19.2 o 0.0 1,271 39.7 o o 1991192 9,993 1,575 1,031 30 68 8.6 0.0 5.2 2.7 1,620 33.3 o 1992/93 8,896 1,222 628 91 144 18.6 49.0 4.4 12.1 1,111 32.3 1993/94 8,411 1,020 771 681 66 25.0 0.6 12 0.1 13.5 897 36.6 0 1994/95 2,942 1,124 244 534 37 28.6 51.0 16 0.0 0.0 908 37.2 3,799 1995/96 4,702 635 382 880 178 320.7 33.0 100 0.0 234 0.0 328 25.7 1996/97 4,201 299 207 350 60 62.7 2.0 0.0 64 3.5 190 27.3 0.043 1997/98 2,985 76 198 178 83 92.1 1.0 0.0 62 1.5 79 13.6 0.036 4.05 47 • a fairly stable rate of sales up until tion. The situation was good until 1986/87; 1993/94. • a significant increase in the subse quent three years from 1987/88 to The Seed Service Unit in the Research 1990/91; and Branch was upgraded and given the status • reduction in 1991192 and decline of an independent institution (Seed Con from 1992/93 until 1997/98. trol and Certification Institute (SCCI)) with the help of Sida's technical and fi The reasons for the above pattern were nancial support. SCCI was responsible the following: for seed certification and quality control. It also worked as a secretariat to the Vari • readily available finance and credit ety Release Committee. Sida, being a to agriculture in the 1980s, perpe common donor, supporting breeding re trated by the government ofthe day; • a move in the early 1990s, away search, SCCI and ZAMSEED, provided from subsidized agriculture to a strong linkages among the essential com more market driven laissez-faire ponents of the seed industry. economy; and • Competition from other companies Zambian Maize Research as a result of liberalization. The Research Branch of the MAFF During the period between 1981 and carries out plant breeding research in 1990, ZAMSEED had a virtual monopoly Zambia. Maize had been the center of fo on the seed business. It supplied seed to cus of research for the seed industry, al PCU s who placed bulk orders based on though seed research on other crops, (e.g., their own estimates. PCUs had no prob wheat, soybean, and sunflower) had also lem in making payments as they were fi been done for a long time with the support nancially supported by the government of other donors. Pasture grasses and le for their role in crop marketing. PCU's gumes, vegetables, root and tubers and only promoted maize seed which was in sorghum/millets were later added into the line with the then existing policies. Sida support. Seeds of non-maize crops such as The Sida-supported maize breeder re wheat, sunflower, soybean, groundnuts, constituted SR52 after purifying its old cowpeas, rice, sorghum and millets were parents. The reconstituted hybrid was re not available to the majority of farmers. leased in Zambia as MM752 in 1984. Most small-scale farmers used Nine additional hybrids and two farm-retained seed for planting of their open-pollinated varieties (OPVs) were traditional crops. Nonetheless, by the late also released between 1984 and 1992. 80's, Zambia was self-sufficient in maize The main emphasis was on the develop seed. ZAMSEED was profitable and was ment of earlier maturing, three-way and able to pay dividends to its shareholders, double-cross hybrids, which are better with some of the profits being poured suited to small-scale farmers. Also seed back into the company for its consolida- yields of three-way and double-crosses 48 are higher, making seed availability better for the first time, announced an official and cheaper to farmers. producer price for sorghum. In 1983-84, with financial support from Sida, it New maize hybrids became extremely sought ICRISAT's assistance with its re popular among farmers, including search on sorghum and millets. small-scale farmers. An adoption rate of 70% was reported by small-scale farmers. Beginning in 1987, the program re The government policies of supporting leased a series of high yielding sorghum maize - including access to seed, fertiliz varieties. Figure 1 gives average ers, credit; extension services, and crop "on-station" grain yield performance of marketing and pricing policies - all of the old and new sorghum cultivars avail which were heavily subsidized by the able in Zambia. New released cultivars government, contributed to a high rate of included Kuyuma, a drought tolerant adoption. widely adopted white grain variety; Sima, a high yielding, dual purpose variety; A 1992 survey of the maize growing WSH 287, MMSH 1324, and MMSH region found that 88% of small-scale 1257, high potential, white grain hybrids adopters ofZambian hybrids used fertiliz of varying maturity duration; MMSH 375 ers at least once during the growing sea and MMSH 413, high yielding, son. In addition, fertilizer rates for maize brown-grain hybrids of good brewing were the second highest in Africa in the qualities; FSH 22, a forage type hybrid; late 80's. Fertilizer prices were also subsi and WP 13 and ZSV 12, partially dized, which cut the prices by 30-60% photo-period sensitive varieties with during the '70s and early '80s. In addi good tolerance to soil acidity for high tion, 64% of small-scale farmers sold rainfall zones. their maize, 42% received credit for maize, while 47% were visited by an ex In the case of pearl millet, released va tension agent (Howard and Mungoma, rieties included Kaufela, Lubasi and 1997). However, in 1988, the govern Sepo, and varieties of finger millet, in ment's expenditure in support of maize cluded Lima, and Nyika. was only about 17% of the government's total budget. This degree of support was Despite good progress in varietal im obviously not sustainable. provement, the research team met with skepticism especially among the sociolo Sorghum/Millet Improvement gist dominated Adaptive Research Planning Teams (ARPT's) who were re Sorghum and millets were totally ne sponsible for on-farm verification and glected crops in Zambia until 1980. The transfer of technology to farmers. Their decline of copper prices in the mid-70's, reservations primarily stemmed from the and subsequent worsening of the coun perception that resource poor small-scale try's economy, aggravated further by re farmers, who were major producers of current droughts, put pressure on the gov sorghum and millets had no capacity to ernment and public to consider alterna invest in seed and other inputs necessary tives to maize. In 1981, the government, for improved cultivars. 49 .. P------~ 1000 +-Hybrids - - __ OP's A- Reg.III OP's 1000 Old Var. HOG -1000 0 I't ...... ~ .. i'l ~ ...... e ~ ::>.. M ! I ::II: ...... & a I ::II: ;: t ::II:= 1 J I III I i ot I I&. a i a Ii I a II Variety Figure 1. Sorghum variety development Poorly conducted ARPT experiments eluded policy makers, ARPTs and were bringing contradicting results indi farmers. cating that the improved cultivars under farmers' conditions were yielding lower The results of this 2 to 3 year aggres than their traditional varieties. A few sive promotion campaign were very en teams even went to the extent of stating couraging. New cultivars started gaining that farmers do not consider improved popularity among farmers. The types as sorghum because of a different farmer-group pressure drew the attention phenology. of policymakers and local NGOs (Non-Governmental Organizations) op In order to circumvent the problem, the erating in the area, decided to capitalize research team volunteered to conduct a on this opportunity and included a sor series of pre-planting meetings directly ghum/millet variety promotion program with farmers in a strategically selected in their agenda. ZAMSEED, which was district. Seed of new cultivars was sold to otherwise hesitant in getting involved in farmers at a commercial price. A very un bulk production of sorghum/millet seeds, orthodox system involving primary for fear of a lack of demand, started re schools in the rural areas came in very ceiving orders for seed. handy for this exercise. The team also car ried out a series of "on-farm" demonstra Amid mounting popularity of im tions and invited related staff, which in- proved cultivars in the area, the research 50 team further decided to capitalize on the Kuyuma is especially popular in the droughts that occurred in 1992 and 1994 SADC region for its grain quality and by conducting large-scale, off-season wide adaptation. Similarly, high yielding seed production of sorghum and millet for brown-grain hybrids (MMSH 413 and drought relief exercises in the country. It MMSH 375) are in demand in the brew also assisted, guided and trained seed ing industry but often, the demand ex growers and production officers, mainly ceeds the supply. Unfortunately, the new in the area of hybrid seed production. government has left the grain marketing totally in the hands of the private sector, Today sorghum is seen as a viable al which will take some time to develop in ternative to maize in the country's remote areas. crop-diversification program and many NGOs across the country are involved in Munyinda, (1998) summarized the ac its promotion and seed production. How complishments of different breeding pro ever, limited seed production and a lack of grams in Zambia. Although yield gains market for the surplus grain are still major over a period of time were some of the constraints to the wide adoption of sor best with regard to sorghum (Figure 2), ghum. and at least as good as in the case of maize ------_._--._-_._- .. --Hybrids -D-OP's - -Old Var. 1fIIO I -=" .. ~ :5! .. .~ I .;= .. q ...... ~ ,; .. -< .. '-I/' ...... 1_ r I) I i Years Figure 2. Sorghum improvement 51 (Figure 3), the adoption rates remained fertilizer subsidies. As part of this privat far short of maize. The average rate of ization program, ZAMSEED was also adoption for sorghum was 37% compared listed as private even though it was not a to 70% for maize. Policy and organiza typical parastatal company. tional support for maize accounts for the difference. The above process had at least tempo rarily, destabilized the seed sector, which Liberalization of the Seed Industry reflected the situation in the entire econ omy of the country. Crucial financial and The worsening state of the economy credit institutions (ZCF/FS, LIMA Bank, pressured the government of Zambia to and CUSA) that served agriculture had implement economic reforms, which closed. The closure of these institutions started in the late '80s. These reforms resulted not only in huge financial losses (withdrawal of subsidies, liberalization to ZAMSEED, but also in the loss of seed and privatization, and deregulation of marketing services. state controls etc.) gained momentum af ter the change in government in 1991. By The policy of liberalization attracted 1993, the new Movement for Multi-party several multinational and private seed Democracy (MMD) government had companies to the Zambian market. Some withdrawn itself from direct involvement of these companies included P ANNAR, in input supply and removed all crop and Cargill Seed Co., Pioneer Hi-bred Inter- ,... r------, 7IDO... 1------.....- 1_HOI -f~~~~~~~~~~~ ______~ I I I I Iii i &I I I I I I I I I I Years Figure 3. Maize improvement - ____0- ______ 52 national, Carnia, the Maize Research In ZAMSEED responded to these stitute (MRI), and Bio-seed Genetics. changes by diversifying its production Several of these companies, after only a base and by reorganizing its market strat year or two, closed down. Perhaps they egy (e.g. recruiting its own agents and set found the Zambian market too small and ting up company outlets). It offered price difficult, after trying to sell introduced, incentives to its distributors and promoted and perhaps, inappropriate varieties. its seed line through demonstrations and advertisements, but the situation contin No other company has invested in local ues to be difficult. research other than MRI, a local venture, cooperating with the Maize Research In ZAMSEED may find itself stronger if stitute of Yugoslavia. At present, it ap it could expand quickly into non-maize pears unlikely that many will invest. The crops whose areas are increasing and unrestricted entry of insufficiently tested competition from other companies is al maize varieties is being blamed by many most non-existent. Fortunately, Zambia is for the introduction and spread of a new well ahead ofmany neighboring countries devastating disease, gray leaf spot (GLS) in seed research. Prospects for export ap in the country. Many of the existing hy pear good provided ZAMSEED can re brids have been found susceptible to GLS tain the control, ownership and access to and are being withdrawn from the market. government-bred germplasm. The disease will now require new invest ments in research to develop resistant va This agreement of ownership of the rieties. government-bred varieties by ZAMSEED is being challenged. The res The entry of new companies in the olution now appears to be ZAMSEED market has, however, increased farmers' starting its own research and development choices in purchasing maize seed variet (R&D) program. However, after suffer ies; this put complacent ZAMSEED on ing from great financial losses due to the guard and under pressure as its market closure of the LIMA Bank, CUSA and share for maize seed dropped. However, ZCF/FS, the company has no capital to this reduction cannot be blamed on com start its own R&D. petition alone. The area under maize cul tivation, as well as use of fresh seed There is a proposal to re-organize the dropped drastically because of the gov shareholding of ZAMSEED whereby the ernment's reduced support of maize. To government would agree to reduce its day, farmers in Zambia have neither guar holdings from 40% to 37.5%; ZCF and anteed supply of seed, fertilizer, credit nor ZSP A from 20% to 5% each; while the assured market for their maize. By 1995 shares of Svalof Weibull and Swedfund total cultivated area in Zambia declined would be increased to 27.5% and 25.5% by more than 15% from the 1985-1990 respectively. average, due to this reduction, and ZAMSEED sales went down from a peak Seed production is also being decen of 13,600 Mt. in 1989-1990 to 3-4000 Mt. tralized. There has been a proliferation of in recent years (Table 1). NGOs involved in seed distribution and 53 production of self-pollinating crops at the Zambia's experience of the maize sec village level. These NGOs are now being tor development clearly demonstrates encouraged to produce quality declared that the policy and organizational envi seed (QDS) to improve the supply of ronment on top of technological develop non-hybrid and non-maize seed at the lo ment have profound effect on the growth cal community level. The involvement of of the seed industry and agriculture as a NGOs in seed production with whole. Policies also determine who bene small-scale farmers is to establish an in fits from these technological accomplish formal seed sector to complement the for ments. mal sector. The post-independent government's Conclusions liberal policies of heavy and unsustain able subsidies and its direct involvement The Zambian industry represents an in maize credit and input supplies, crop example of a young and emerging seed in marketing and consumption, created an dustry in a small (in terms of agriculture imbalance and non-sustainable develop and local seed market), developing coun ment displacing most other crops. try where 60% of the people depend on agriculture (of which 75% are re Sudden changes in government poli source-poor small scale farmers) and the cies from heavy subsidies to a total with country's infrastructure is poorly devel drawal of all government services threat oped. The industry has gone through ma ens past gains, creates mistrust and uncer jor changes since its establishment - re tainties, hinders the private sector sponding to political and economic investment, and leads to hardships in the changes in the country - and the challenge seed industry. continues. It is hoped, however, that in the long The industry has had mixed success: run, sustainable and balanced agriculture Research produced excellent results. will emerge in which commercial crops Seed production volumes generally in will dominate in the better developed creased, particularly in maize, where the parts of the country, while traditional country accomplished self-sufficiency; crops best suited to each agroecological and the quality of seed has been excellent. zone will dominate the resource-poor There have been some weak areas such as farming sector. And the seed industry will low seed volumes of non-commercial respond accordingly. crops, seed deliveries are generally late, Extension services for non-maize crops References are not up to the mark, and farmers in re mote areas not getting the proper seed va Cromwell, E. 1996. Governments, Farmers and seeds in a Changing Africa. CAB International, rieties or crops of their choice (Cromwell, Overseas Development Institute. 1960). Critics also complain that Eicher, C. L. and B. Kufuma. 1997. Zimbabwe's Emerging Maize Revolution. In Africa's ZAMSEED has been unable to pay for the Emerging Maize Revolution (ed.) Derek royalties on varieties to the Research Byerlee and Carl K. Eicher., Lynne Rienner Branch. Publishers, Boulder, London., pp. 25-44. 54 Howard, 1. A. 1994. The economic Impact ofIm Byerlee and Carl K. Eicher., Lynne Rienner proved Maize Varieties in Zambia. Ph, D. the Publishers, Boulder, London., pp. 45- 62. sis, Michigan State University, Dept. of Muniynda, K. 1998. Agricultural Research in Zam Agricultural Economics. East Lansing. Michi bia - A Historical Perspective. In print. gan. Smale,M. andP. W. Heisey. 1977. Maize Technol Howard, J. A. and C. Mungoma. 1977. Zambia's ogy and Productivity in Malawi. In Africa's Stop and Go Maize revolution. In Africa's Emerging Maize Revolution (ed.) Derek Emerging Maize Revolution (ed.) Derek Byerlee and Carl K Eicher., Lynne Rienner Publishers, Boulder, London., pp. 63-81. 55 The Experience of Winrock International with Seed Multiplication in West Africa Through: The On-Farm Seed Project (OFSP), 1987-1992 and The On-Farm Productivity Enhancement Program (OFPEP), 1992-1998 Alphonse Faye and Pierre Antoine Introduction the introduction of non traditional crops and restrict the range of crops planted. In The focus of seed production in Sene light of the mounting pressures on food gal has traditionally been on the main supplies and the need to generate income, cash crop of peanuts. Other crops have re such problems cannot be ignored. ceived limited attention and funding. The new agriculture policy resulting from the Winrock International, through its recent Economic Recovery Program has On-Farm Seed project (OFSP) and led to reduced financial support from the On-Farm Productivity Enhancement Pro government to the seed sub-sector. It is in gram (OFPEP), has addressed the prob tended that the Directorate of Seed Pro lem of meeting farmer demands for seed duction and Control (DPCS) withdraw by improving their practices of seed se from seed production and restrict its role lection, production and storage. The ratio to quality control. Although privatization nale for the programs arises from esti of marketing of agricultural inputs, in mates by Mississippi State University, cluding seed, is encouraged, it still re (MSU), a world leading institution on ag mains to be seen whether the private sec ricultural seeds, that more than 90% ofthe tor can replace the public system of agri crops in developing countries are sown cultural input supply. In the interim, from seed stocks selected and saved by farmers are increasingly responsible for farmers. This situation still prevails de their own seed supply. spite 30 years of emphasis on national seed programs and industry investment in One of the major weaknesses in the seed production and supply programs. seed system for food crops is the high cost Seed industry development has been im of the distribution and marketing of seed portant and the assistance rendered not at the local level. Another is the limited wasted; but the national seed institutions testing conducted with improved variet still need to develop ways to make tradi ies under farmers' conditions to assess tional seed production and distribution compatibility with their production sys systems more effective and to convince tems and to identify appropriate hus farmers to include improved seed into bandry practices. their traditional seed production and ex change systems. Inadequate methods of selecting and saving seed diminish the potential bene Successful identification (of varieties) fits of improved seed varieties, impede and production of appropriate seed variet- 56 ies clearly requires sustainable research Winrock International's assistance to and development efforts. However, farm NGOs, and the U.S. Peace Corps en ers themselves can multiply many variet hances their existing activities through ies of seed once they have the basic mate training and technical assistance. The rial. In fact, seed production can be a project has also initiated functional link small-scale rural enterprise for income ages between regional and local groups, generation. national agricultural research institutions (such as ISRA in Senegal and the Seed Program Development and Support Technology Unit in The Gambia) and in ternational consortia or programs such as Winrock's programs have assisted the the Collaborative Research Support Pro seed-related programs of (non-govern grams (CRSPs) including INSORMIL. ment organizations (NGOs) and the U.S. Peace Corps in Senegal and The Gambia The Winrock International Seed Pro using a process approach that adapts the gram Approach technical assistance to the needs ex pressed by small farmers. Support and The approach ofWinrock International commitment to the seed program came has two key elements: (1) it is collabora from several sources. tive in that all technology dissemination activities are implemented through coop In The Gambia, Winrock International erating institutions, primarily NGOs, collaborated, often on a cost-sharing ba community-based organizations (CBOs), sis, with many NGOs that had established and in Senegal the U.S. Peace Corps, and seed production activities. The program (2) it is participatory in that traditional received technical assistance from and seed production techniques and problems provided field data to the Seed Technol are examined with farmers before activi ogy Unit in Sapu. USAIDlBanjul also ties are undertaken. The long-term goal of supported seed research in its develop the Winrock International seed program ment strategy for The Gambia. includes the development of "a model or models" for on-farm seed production sys The Government of Senegal actively tems which can be replicated or adapted supports seed research and production. for use in other areas. The approach can Winrock International helped national re be analyzed from several viewpoints: search and production programs by iden tifying and disseminating improved agro Rep/icability nomic practices and appropriate varieties to the extension staff of NGOs, and the Because of the diversity of interest U.S. Peace Corps volunteers implement among NGOs, the Winrock International ing seed improvement activities at the vil seed program adapts services to fulfill the lage level. Winrock International also needs of each organization. This has contributed significantly to the design of proven to be a valid means of assisting the Peace Corps Rice Initiative, World NGO field personnel in the area of seed Vision International and Christian technology. This approach should be rep Children Fund's agricultural programs. licated provided that training and techni- 57 cal assistance has the required flexibility terprises. Winrock's seed program com to tailor the service to each NGO' s unique plements the policy by developing "mod need. The Winrock International seed els of on-farm seed production systems program develops basic training pack through the promotion of village level ages in seed production and in rice agron seed production as reported by omy, which can be modified to fit the Christophersen et al. (1998) relating needs of the particular audience. In some adoption of improved millet seeds as one cases, the Winrock International seed of the most interesting impacts on how program provides technical assistance in Winrock is helping villagers respond to crop production and extension rather than USAID/Senegal presence in their areas, in seed production, at the request of the and fleshing out the synergies between NGO/partner. improved natural resources management (NRM) and the new decentralization and Innovativeness and Leadership private sector strategic objectives (SOs)". Development From Seed Demonstration Plots to Most seed programs in developing On-farm Seed Multiplication countries target the market-oriented farmers as the potential recipients of their Unlike the majority of seed production technology and, as a consequence, large programs in developing countries, which farmers are the main beneficiaries. The focus on developing national seed pro approach of Winrock is to make small grams, Winrock's approach concentrates farmers the priority target of its seed pro primarily on the introduction of seed of gram. The program works at the "grass improved varieties, through demonstra roots level", where NGO field staff and tion plots and organization offield days to volunteers working at village level pro enhance farmer-to-farmer extension and vide the linkage between Winrock and the provision of technical assistance on seed lead farmers. Through its participatory multiplication and seed technology. approach, Winrock encourages the estab lishment of farmers in each village or In most cases, as illustrated in the next CBOs. section, this collaborative approach goes further than the relationship with only Policy Implications NGOs, and extends to CBOs created by farmers and supported by the NGOs. This A critical factor in any project is its enables Winrock to reach one ofthe major conformity with the national policy of the program's outputs: "some communities country in which it is operating. will support local system or seed produc Winrock's seed activity has been consis tion and distribution either through coop tent with government policy in both Sene eratives or by individuals who create pri gal and The Gambia. In Senegal, the agri vate firms". cultural policy attempts to reduce the role of government in the supply of agricul A Winrock study carried out, in 1993, tural inputs, and promote the develop by D. Heinen in northern Senegal, in co ment of private sector in agricultural en- operation with World Vision, confirmed 58 the development of a marketing system of ricultural development programs in Sene cowpea and millet seeds produced by gal and The Gambia. Project some farmers. The study showed that beneficiaries, therefore, includes both sales of seed were indeed taking place at NGOs and smallholder farmers. different levels: among the farmers within the village, between villages and at Smallholder Farmers weekly markets. The programs, directly or indirectly However, World Vision did not record benefited people through seed multiplica the quantities produced through multipli tion, field demonstrations, and training cation, but only took into account the programs. Conservative estimates indi quantities it purchased, conditioned at its cate that more than 100,000 smallholder processing center and then sold as certi farmers were beneficiaries during the fied seeds. This means that part of the 1988-1998 period. At least two-thirds of seed produced, but not recorded, ended up the beneficiaries were women, given the being sold in various containers in the programs emphasis on rice, a crop mostly markets and villages. Such "packaging" grown by women in southern Senegal and raises the question of how such a product The Gambia. can be perceived as a quality product as opposed to what is usually sold on market. Among the most visible NGO benefi ciaries, one can include: The answer lies in the effect improved variety demonstrations has had on farm Save the Children Federation (SCF), ers. Word of mouth advertising and rec The Gambia ognition of the competence of "bush con sultants" (lead farmers) has generated in SCF has a diverse program in health, creased sales. In some instances, "bush food production and education in the consultants" carry out germination tests North Bank division of The Gambia. Ap in small metal basins in their house and proximately 30 % ofSCF's resources are show those to potential buyers. Inter devoted to agricultural activities, and views show evidence that sales in the vil aimed at helping 10,000 women farmers lage and to neighboring villages are tak in 24 villages to increase food production ing place with buyers coming directly to through gardening, introduction of new the bush consultant's compound to buy, rice varieties and seed multiplication. cash and carry, sometimes with orders Winrock International OFSP project as from other farmers from the same village. sisted SCF since 1988 by: (1) training field staff farmers in seed production, Project Beneficiaries post harvest handling, and rice produc tion; (2) providing technical assistance in Winrock International has provided rice agronomy; and (3) assisting in the training, technical assistance, and other planning of seed multiplication and dem services to NGOs and the U.S. Peace onstration activities. SCF rice activities Corps enabling their personnel to more have initially involved 25 contract grow effectively implement their respective ag- ers producing rice seed in 10 villages. In 59 1991, there were 117 contract growers in multiplication involving 60 "bush consul 20 villages producing seed on 21 hect tants" in 57 villages. ares. Christian Children's Fund (CCF), Freedom from Hunger Campaign Senegal (FFHC), the Gambia Collaboration between Winrock Inter FFHC works with women rice growers national and CCF integrates efforts to im in 32 villages of Lower River, North prove farmers' access to good seeds, with Bank, and Mac Carthy Island North Divi activities aimed at improving soil fertility sions in The Gambia. FFHC concentrates and management at the farm level. This on introduction of new rice technologies collaboration relies on the existence of six in tidal swamps and rainfed lowland ar community-based organizations financed eas, promoting water harvesting, control by a system of children's sponsorship. ling salt water intrusion, and strengthen CBOs are playing an important role in the ing village institutions. Winrock's assis adoption and diffusion of the technolo tance to FFHC has included rice training gies set out by OF PEP and represent a real for field staff, field visits to monitor seed link in the training of farmers, the imple multiplication and demonstration plots mentation of demonstration plots and and the testing ofa prototype rice seeder. monitoring of seed activities by assigning the best qualified and dynamic members World Vision International, to serve as part time extensionists for Senegal (WVI) Winrock. WVI has an integrated development Experiences in Senegal show that easy program of water, agriculture and health access to seed is a critical factor in en in the regions ofLouga and Thies serving hancing improved seed use because when over 320 villages with a total population farmers are convinced that the improved of 150,000. These northern regions of variety is significantly better than their Senegal served by WVI have become in traditional varieties, the adoption of that creasingly marginal for agriculture be variety still depends on availability of cause of the reduced rainfall (250-300 seed supplies locally. Only a decentral mm per year) widespread deforestation, ized system of seed production and mar and decline in soil fertility. Winrock In keting can efficiently address the issue of ternational, through the OFSP and providing seed to farmers in the 41 scat OFPEP programs, conducted several tered villages composing the six CBOs. training sessions for WVI staff member Through the CBOs, Winrock Interna and farmers. Twenty-five WVI staffini tional trained 6 seed extension specialists tially received training from OFSP. Forty and 73 seed growers. As a result of the "bush consultants" who are responsible training and field visits, each CBO staffis for demonstrations and seed multiplica monitoring the production of millet seeds tion with farmers also received training. and ensuring marketing of the product. WVI conducts millet and cowpea seed 60 Conclusions seed technology that can be useful to NGO agricultural programs, enhancing Implementation of Winrock Interna their effectiveness, and in turn, benefiting tional's seed activities relied on collabo the farmers. ration with organizations with planned or on-going seed activities and adequate Another conclusion of the program is number offield staff. Through these col that small-holder farmers who can pro laborative programs, Winrock Interna duce quality seeds of various crop variet tional has facilitated the development of a ies can be identified in most rural commu model of on-farm seed production that is nities of Senegal and The Gambia. Given beneficial to all parties involved: the proper incentives and adequate mar keting opportunities, these farmers can • the farmers are more easily access become "contract seed farmers" and con ing improved seed which is pro stitute the first step toward the develop duced locally; ment of a private seed sector production • NGOs have benefited from a seed system. program that is providing useful services to farmers; References • Research institutions' (i.e. ISRA) technologies are being extended to Bragantini, Claudio and Schillinger, William F. farmers, and valuable feedback is 1992, Final Evaluation of the On-Farm Seed Project, Senegal and The Gambia, Winrock In obtained from the farmers. These ternational, September 1992 are useful for future technology de Christopherseb, K., Hadji, A., Rand, B., and velopment. Winterbottom, R. USAID/Senegal NRM Limited Impact Assessment, February 1998 Winrock International Institute for Agricultural Furthermore, it would have been diffi Development On-Farm Seed Project: 1990 An cult to develop these activities without the nual Report training, technical assistance and facili Winrock International Institute for Agricultural Development On-Farm Productivity Enhance tating role of Winrock International. ment Program, 1993 Annual Report NGO managers have many roles to per Heinen, Daniele, 1993. Seed Marketing Survey form in the implementation of the re within the World Vision. International Program Area, On-Farm Productivity Enhancement Pro search activities which include manage gram in Senegal, Winrock International ment ofpersonnel and budget, and logisti cal support and linkages with their Acknowledgements headquarters. Technical aspects and training necessary for their programs are The authors are indebted to Dr. Moses often difficult to manage because of lack Onim and Mr. Niels Hanssens for their of time and internal technical expertise. useful suggestions and comments regard Winrock International has the capability ing this paper. to provide services in a specific area of 61 Rapporteur's Report s.c. Gupta Hybrid Seed Experiences in Developing Countries Day 1- Session 2 All the presentations were very clear Subsidy on seed is not necessary, even and, therefore, there was no discussion on though initially it helps the seed program. the four papers presented during this ses sion. The salient features are summarized Gebisa Ejeta presented the story of in the following paragraphs: Hageen Dura-l (released in 1983 in Su dan) which led to the development of the B.R. Barwale presented the story of seed industry in Sudan. Sorghum is MAHYCO Hybrid Seeds in India. He grown on 5m ha in Sudan or wherever you started producing the first maize hybrid can grow a crop. He described different seed in 1964 and acknowledged the sup systems of agriculture, the seed industry port of L.R. House and the Rockefeller organization, and the ups and downs in Foundation. In 1965, the first sorghum the production ofHageen Dura seed. The hybrid, CSH 1, was released in India and seed industry development is durable, but MAHYCO produced 3.2 tons of seed on is a slow process. He also emphasized 1.6 hectares. In hybrid sorghum seed pro that poor farmers from developing coun duction, MAHYCO faced the problem of tries will buy the seed only if it is profit grain mold and, as a result, started pro able to them. A sustainable seed industry ducing the seed in the dry season. For the is feasible only with favorable govern first 3 to 4 years, there was no unsold seed. ment policies and support. Public aware They established a seed processing plant ness is vital for the success of a seed in with the assistance of USAID, and later dustry. on developed a seed store. Initially, foun dation seed was produced only by the Na Bhola Nath Verma presented the his tional Seed Certification (NCS) agency, tory and experience of the Zambian Seed and thereafter, the monopoly was re Industry. Zambia represents an example moved. Breeder seed was given to com of a young industry in a developing coun petent organizations to produce founda try with poor infrastructure and a small 10- tion seed. These were certified by the cal market. The industry has gone State Seed Service Agencies. The key through major changes as a result of the items for the seed industry are seed regu political and economic climate of the lations, a seed processing plant and ware country. Initially, ZAMSEED was not in housing, truthful labeling, and certifica terested in sorghum seed production, as tion. Farmers always look for ultimate they were multiplying primarily maize profit and they buy good quality seed. seed. Recently, they recognized opportu- 62 nities in the export market for sorghum USAID, IF AD, Cargill, etc. The project seed production. Generally, policymakers is not yet working on hybrid seed produc require education and/or pressure in order tion. Alphonse Faye stated the private to yield to changes. Recently many seed sector is responsible for seed production companies have come to Zambia. and the government is responsible for seed quality control. Pierre Antoine described the Winrock International seed production program, L.R. House commented that every which was started in 1987 in Senegal and body's input - whether it is the govern the Gambia. Their aim is not to create a ment, farmers, seed producers, or re seed industry, but to assist farmers in seed searchers - is necessary for the success of production. The project is supported by a seed program. 63 Monday - September 28 -Session 3 Current Status of Seed Production Opening Remarks Moderator - Darrell Rosenow, INTSORMILI Texas A&M University Pearl Millet Seed Systems in Niger Jupiter Ndjeunga and Aboubakar Sidi Summary Niger. A series of research, extension, ru ral integrated development projects, and This paper reviews the structure and NGOs have been heavily involved in de performance of formal and informal pearl veloping, disseminating varieties, and millet seed systems in Niger. Donors have distributing seed to end-users. These in invested more than US$45 million in seed vestments have largely failed. The seed production projects in Niger during the production units (SPU) encountered past two decades. These investments have many difficulties, largely because of the largely failed. Public seed supply systems project's considerable scope which made operate at a consistent loss while farmers seed multiplication and distribution activ complain about poor seed quality or lack ities unsustainable (Mazuccato and Ly, of seed access. In contrast, informal vil 1993). Research stations were unable to lage seed systems work well. Most farm provide SPU with the quantity of breeder ers consistently derive their pearl millet seed needed for further multiplication and seed from their own harvests, from neigh less than 30% of improved pearl millet bors or from village markets. Seed quality seeds produced were actually sold. is high and a range of varieties remain Farmers, NGOs, and rural integrated de available. Village seed systems offer a velopment projects managers complain cheaper and more efficient means of de of poor access and late delivery of seed livering seed to farmers. Future invest and poor quality of seed provided by the ments in seed systems development public system. The low supply of breeder should target improvements in the capac seed, poor seed quality and poor demand ity ofvillage seed systems to maintain and estimation and distribution systems have distribute seed security stocks for drought been identified as the major constraints to years. More efficient village seed produc seed multiplication and distribution in ers should be encouraged to become en Niger (Ndjeunga, 1997). These combina trepreneurs tasked with the multiplication tion of constraints made the adoption of and distribution of new improved pearl improved varieties very low (less than 1% millet varieties. of the total cultivated areas) and thus the returns to investment in research and seed Introduction multiplication are also very low. During the last twenty years, interna During and after the period of involve tional and national donor agencies I have ment of seed projects (1976-1992), the invested more than US$45 million in seed private sector has shown little interest in multiplication and distribution projects in seed multiplication and distribution, es- IDonoT agencies include the United States Agency fOT International Development (USAID); the "Fond Europeen de Developpernent'· (FED); the "Agence Canadienne de Developpernent International'· (ACDI); the International Bank fOT Reconstruction and Develop ment (IBRD) and government of Niger. 67 pecially for seed of subsistence crops with This paper compares the performance high multiplication ratios such as pearl of formal and informal pearl millet seed millet partially suggesting limited market systems and questions whether informal potential for pearl millet seed. Poor per seed systems might serve the availability formance of the public sector and limited of pearl millet seed in Niger. Findings commercial interest in pearl millet seed may well be applied to other countries in multiplication and distribution by the pri the semi-arid tropics of West and Central vate sector prompts the need to explore al Africa which face similar situations ternative seed supply schemes. One alter (Chad and Burkina Faso). native is the village seed markets. Farmers still draw most of their planting Historical Development of the Formal seed from their own seed stocks, ex Pearl Millet Seed Sector In Niger change seed with neighbors or relatives, or purchase seed from local village mar In Niger, formal pearl millet seed sys kets. tems evolved through two stages ofdevel opment with structural and institutional Recent research in a few less devel changes. Seed multiplication and distri oped countries highlight the efficiency of bution for cereal crops started in 1976 local village seed systems at supplying with the Niger Cereal Research project seed and resolving seed shortage at low (NCR) funded by the United States transaction costs. In 1990, a rural house Agency for International Development hold survey of small-scale farmers in Ma (USAID). The main objectives of the lawi indicated that 68 percent used their NCR were mainly focused on capacity own seed stocks as their primary source of building of seed infrastructure with the maize seed; followed by seed from neigh overriding goal to supplying annually the bors and relatives (44 percent) as a sec quantity of seed needed to cover one-third ondary source, and a few used the local of the cultivated areas. The project was market as a secondary seed source (4 per implemented in two sub-phases. The first cent). A small proportion of survey farm sub-phase was mainly used to build up the ers were chronically seed insecure required human and physical infrastruc (Cromwell, 1996). In Southern Zimba ture for seed multiplication and distribu bwe, despite the relatively well developed tion; and the second sub-phase was used seed markets, a survey of rural house to support agricultural research and agri holds in 1995/96, revealed that about 64 cultural production. At the same time, the and 44 percent of surveyed farmers draw project attempted to revitalize the input their planting pearl millet and sorghum supply sector especially fertilizers seed respectively from previous harvests. through cooperatives. During this phase, Thirty and 32% obtained their planting 6 seed productions units (SPU) were built pearl millet and sorghum seed from and equipped with seed laboratories and neighbors, friends or relatives. Seed large seed storage and conditioning units. shortages are resolved within the village Moreover, 4 seed technologists and 14 seed trade and most transactions are free technicians were trained to conduct field (Rohrbach, 1997). inspections and laboratory tests. 68 Variety development and breeder seed able scope which made them unsustain production were undertaken by the na able. Seed production units complained tional research institute: The Institut Na of insufficient and late delivery ofbreeder tional de la Recherche Agronomique du and basic seed by INRAN and the SPU at Niger (INRAN). Basic seed production Lossa. Farmers had little access to seed of was carried out by the SPU at Lossa. Ba many varieties developed and released. In sic seed was bulked into registered seed fact, out of the 17 varieties developed and by SPUs and certified seed multiplication released, only three varieties were actu was contracted out to farmers under the ally multiplied and distributed (Tables 1 supervision of the production and quality and 2). The project never met its objective control units ofSPU. Seed quality control of supplying seed necessary to plant one for breeder and basic seed; i.e. field in third ofthe total pearl millet cultivated ar spections and post-harvest tests, were car eas. On average during the last 5 years of ried out by the SPU of Lossa, and that of the NCR, only about 10% of the targeted registered and certified seeds was the in seed production was achieved (Table 3). ternal responsibility of seed inspectors Because there were few sale points and and laboratory technicians within SPU transportation difficulties involved in according to the International Seed serving scattered farmers located in Testing Association (ISTA) regulations poorly accessible areas, farmers had little as specified in the Seed Certification Act access to certified seed and seed was often of 1976. Improved seed was distributed delivered late. Due to poor demand esti and sold to farmers by cooperatives, mation, less than 30% of the total seed NGOs, SPU, regional direction of agri production was actually sold (Tassiou, culture and development projects. Seed 1996). Because of the project's large prices for all seed classes were set by the scope, there was less manpower available government. At the end of the project in to properly carry out all seed inspection 1989, despite the massive investment in and post-harvest tests. Thus, only few seed multiplication and distribution activ contract farmers were monitored and ities, seed production units operated at many contract farmers did not often re consistent losses and thus were unsustain spect the required isolation distance nec able. essary to maintain purity. Moreover, seed bought by SPU from contract farmers During this phase, the seed production were not properly inspected. This resulted units encountered many difficulties, in poor seed quality and many farmers, largely because of the project's consider- NGOs, rural development projects Table 1. Pearl millet certified seed production (tons) by the SPU by variety from 1985 to 1989. Year Variety 1985 1986 1987 1988 1989 Average HKP 576 569 76 632 0 370.6 CIVT 2885 1354 334 731 226 JI06 Ba-angourie 0 0 2 10 5 3.4 P3koIlo 50 436 26 273 0 157 Total 35II 2359 438 1646 231 1637 69 Table 2. List of varieties developed and released in Niger. Pearl millet Agroecological zone Crop cycle Average potential yield Year of I varieties (mmrainfalll (days to maturity) (tons ha- ) development I. HKP 350-500 80-90 2.0 1975 2. HKP3 280-250 70-75 1.5 1983 3_ P3KOLLO 500-600 90-95 2.5 1962-1977 4. CIVT 450-600 80-90 2.5 1977 5. %HK 450-600 70-75 2.0 1975 6. MORO-PI 200-300 90-95 1.8 1985 7. GR-PI 450-600 70-75 2.5 1985 8. ANK-PI 300-350 70-75 1.0 1985 9. HKB-Tift <300 70-75 2.0 1982 10. H-80-JO-GR 300-400 80-85 2.3 1980 II. T-18-L >250 85-95 2.0 1982 12. ITMV8304 300-400 80-85 2.5 1983 13. ITMV 8002 300-400 80-85 2.5 1980 14. ITMV 8001 400-500 80-85 2.5 1980 15. SOUNAIII >600 Intermediate 2.0 1982 16. ZATIB 300-600 Intermediate 1.5 1981 17. HKB-PI >300 Intermediate 2.0 Source: INRAN (1994). Catalogue Nigerien des Varietes de Cereales et de Legumineuses_ Ministere de l' Agriculture et de l'Elevage. Edition, 1994. Table 3. Pearl millet area planted (ha), total seed production (tons) by variety and seed production units, and proportion of the total seed produced to total seed planted (%) in Niger, 1991-1995. Year Variety 1991 1992 1993 1994 1995 HKP 27.9 20.2 21.0 166.8 55.5 CIVT 199.7 22.2 25.4 95.2 23.9 ZATIB 0 0 0 1.4 2.2 P3kollo 0 0 7.5 6.1 5.6 Total improved seed (tons) 227.6 42.4 53.9 269.5 87.2 Area planted (000 ha) 4,386 4,939 3,860 4,935 5,229 Estimated total seed planted (tons)l 14620 16463 12867 16450 17430 Ratio of improved over total seed planted ( 1.56 0.26 0.42 1.64 0.50- I This assumes a seeding rate of 10 kg ha as actually practiced by farmers and a renewal rate of 3 years as suggested by the seed division in Niger. complained about the poor germination than the average grain price, 130 Fcfa/kg rate and seed viability. in local markets (Rachmeler, 1991). Therefore, seed prices were highly subsi Price of seed was set by the govern dized by the project and at the end of the ment with little consideration of grain project in 1989, due to lack of external prices in local village grain markets and funding, SPU almost ceased to operate. the cost of seed production. Between 1985 and 1988, the average cost for pro· In 1989, the government of Niger and ducing certified improved millet by SPU donors re-assessed the seed multiplica was estimated to 1,720 Fcfa/kg; whereas tion and distribution strategies and it was sold to farmers at 100 Fcfa/kg less launched another project: the Projet de 70 Developpement de l' Activite Semen tion of both. The foundation seed produc ciere au Niger (PDASN) funded by tion unit at Lossa and the SPU of USAID with ultimate objectives to build Guecheme continue to produce seed in on failure of the previous seed projects government plots without contracting out identified as 1) poor seed production to farmers. The SPUs of Kourougoussao planning, 2) poor seed quality control; 3) and Doukou-Doukou contract seed multi poor seed demand estimation, 4) distribu plication to farmers, and finally the SPU tion and promotion, and finally 5) the lack of Magaria contracted farmers for com of a viable seed security stock scheme. mercial seed and still produces registered Unfortunately, this project never really seed in government plots. While the other took off and in 1992; the project was can SPU still operate with heavy losses, the celed. Once more, the government reas SPU of Kourougoussao incurred rela sessed its strategies by decentralizing all tively little losses in 1996, and could be seed multiplication and distribution activ profitable if one does not account for de ities at the regional level. Price setting and preciation on long-term assets (Table 4). seed production planning and distribution Large reduction of losses in were to be handled at the regional level. Kourougoussao were attributed to the fact that farmers were entirely responsible for The third and current phase started in seed multiplication and incurred the en 1992 with the creation of regional seed tire seed production and distribution risks centers. The start-up capital for these cen since seed is purchased by managers of ters was made from sale of seed and fertil SPU on demand by private individuals, izer stocks left after the PDASN. These farmers or rural development projects. funds were to be administered at the re Seed quality and conditioning are rarely gionallevel by a regional seed committee performed because of lack of staff and composed of the regional administrator funding. Seed is distributed mainly at the (Prefet), regional director of agriculture regional level or seed multiplication cen and representatives from other institu ters and thus involves little cost. Finally tions involved in seed activities. Cur prices are set taking into consideration the rently, three seed multiplication and dis grain price and cost of production. How tribution schemes exist: seed production ever, the SPU ofKourougoussao operates in government plots without involving as traders. Their current role could as well farmers, contract farming, and a comb ina- be played by the most efficient farmers or Table 4. Profitability of seed production units in Niger in 1996. Average cost of seed Seed Production Units Profit or losses (Fcfa) production I (F cfalkg I) SPU of Lossa (515,250) 1633 SPU of Magaria (905,475) 1151 SPU Kourougoussao (385,362) 305 SPU of Guecheme (2,356,675) 880 SPU of Doukou-Doukou n.a. n.a. PU of Hamdallaye Not operational Not operational 1These costs do include depreciation on heavy capital items such as the buildings. Numbers in parentheses are negative and represent losses. 71 groups of farmers who could be trained Methodology and Data Sources and assisted in seed multiplication and distribution. Two sets of data were gathered at the institutional and rural household levels to Currently, many seed multiplication compare the relative efficiency of formal and distribution centers have emerged. and informal pearl millet seed multiplica These schemes are initiated by NGOs, ru tion and distribution schemes at develop ral development projects and local com ing and maintaining varieties, multiply munities. In contrast, the private sector ing seed, maintaining high quality seed, has weakly emerged. About five private storing and distributing seed and finally seed growers have been reported. How providing seed at relatively low costs. ever, these individuals are less interested in producing varieties with a high multi The first set of data was obtained in plication ratio or with no readily available September-October 1996 from an infor markets such as pearl millet and sorghum. mal survey of all institutions and mem The private sector is interested in produc bers of institutions involved in pearl mil ing seed of crops with low multiplication let seed multiplication and distribution. ratio such as cowpea and groundnut or They include: four pearl millet breeders; with high multiplication ratio such on five managers of seed multiplication cen ions, and other horticulture crops with ters; five regional directors of agriculture; ready available markets for the the head of the seed division, and two end-products. They have also expressed NGOs: Care International and interest in hybrid seed where there is a CARITAS; one private seed grower, the higher probability of profit, as seeds have President of the large cooperative union to be purchased every year. Currently, in (UNCA); and the Director of Agriculture. Niger, the hybrid sorghum NAD-l is The second set of data was drawn from a gaining interest in the private sector. survey of302 rural households conducted during June-July 1997 to assess the opera The low commercial interest for tional procedures of the informal seed open-pollinated crops such as pearl millet systems, the sources of farmers' seed, the prompts the need to explore alternative magnitude of seed traded, the type of seed seed supply sources. Currently, less than transactions and the magnitude of seed 1% of pearl millet cultivated areas are stocks in 1996 and 1997. The year 1995 planted to improved seed varieties. Many was a relatively good year in terms of farmers still draw their planting seed from rainfall amount and distribution com previous crops, exchange seed between pared to 1996. This allowed a comparison neighbors or relatives or purchase seed of rural household's seed purchasing be from the local village markets. In order to havior between good and poor rainfall examine and compare the formal and in years. formal seed systems, a survey of institu tions and members of institutions were A stratified random sampling proce carried out. dure based on agroecological zone and accessibility to the main road were used to select villages. The parameter determin- 72 ing the agroecological zone was the context, a review of the pearl millet vari Length of Growing Period (LGP) of 75 eties developed by the national and inter days corresponding approximately to 400 national research institutes in Niger will mm of rainfall. Villages located within be examined relative to their suitability to the range of 5 km from the main road were agroecological zones and farmers' prefer assumed to be accessible and others not ences. accessible. Within stratum, a minimum of 12 villages were chosen according to the Since 1975, national and international village population size. Within the vil research institutes in Niger have per lage, a range of 3 to 10 rural households formed relatively well at developing and were chosen according to village popula releasing varieties adaptable to a wide tion size. Overall, 302 rural households range of agroecological zones (Table 2). were selected of which 20% were located More than 17 pearl millet varieties were 2 in the more drought prone areas, and developed and released • Breeding work about half the sample was positioned in focused on the development ofhigh yield poorly accessible areas. ing, early maturing, pest and disease re sistant varieties adapted to a wide range of Current Status of Pearl Millet Formal agroecological zones using local and Informal Seed Systems in Niger germplasm. More than 50% of the variet ies developed and released are grown in The performance of formal and infor areas with average rainfall between 200 mal seed institutions will be examined and 350 mm and the remainder can be with regard to their relative efficiency at grown in environments of more than 400 performing basic seed activities. They in mm ofaverage rainfall. About one third of clude: variety development and release, the varieties released are short-maturing seed production, seed quality, seed stor (less than 75 days of days to maturity); age, and seed distribution. The overall about 50 percent are intermediate (80-85 economic efficiency will be examined in days to maturity) and the balance are the light of price margins (unit sale price long-maturing varieties. More than 60% minus the unit cost of production). of the varieties released result from the purification oflocal cultivars through S 1 Variety Development and Release testing with selection for pest and disease resistance, high yield, and early maturity Pearl millet varieties developed and re as the basis for purification. Five local va leased in seed systems should be suitable rieties formed the basis of breeding work to agroecological zones and farmers' namely: Hainikire, Moro, Guerguera, preferences. The extent to which this Ankoutess and Zanfaroua. For example, holds determines the potential demand for the series of improved varieties (HKP, varieties by farmers and seed production HKP3, 3/4 HK, HKB-Tift, HKP-P 1) planning by seed production units. In this 2 The varieties released include: HKP, HKP, P, Kollo, CIVT,'/4 HK, MORO·PI, GR-PI, ANK-PI, HKBTift, H-80-IO-GR, T 18-L, ITMV 8001, ITMV 8002, ITMV 8304, SOUNA 3, ZATIB, HKB-Pl. 73 were obtained from the purification ofthe the surveyed rural households select good local variety Hainikire. panicles right after harvest, while only 7 percent select seed based on plant traits. Overall, the formal seed systems pro Survey results also indicate that, collec vide a wide range of varieties that are suit tively, farmers grow up to 33 varieties of able to all agroecological zones. How which 3 are improved varieties (CIVT, P ever, the formal seed system performed KOLLO, HKP). There is a large number poorly at promoting pearl millet varieties. of varieties available within departments Only a few of those varieties released are and agroecological zones. Averages of 15 actually multiplied and distributed. From and 17 pearl millet varieties were found to 1991 to 1995 only 4 varieties: HKP, be grown by departments and within CIVT, P3Kollo and ZA TIB were multi agroecological zones respectively. There plied by seed production units. Between fore, there is a wide range of local variet those, aggregate production during those ies available at the local level. years indicate that about 54% of the seed produced was CIVT and 43% HKP (Ta Overall, the formal seed system con ble 5). During the last 15 years, the formal tributes little to the demand for seed ofva seed system has performed poorly in re rieties or the promotion ofreleased variet leasing varieties. No variety has been re ies. Many farmers still grow their local leased although there is a wide set of varieties. The low interest in currently im promising varieties in the pre-release proved varieties may be explained by lim 3 stage . ited improvements offered by the new va rieties which may, in tum, partially ex For centuries, farmers have been seek plain the failure of the formal seed ing to maintain old varieties or specific system. Mazzucato and Ly (1993) sets of traits. Survey results also indicate claimed that " .. the majority of the seed that large, long and compact panicles, and multiplied were local purified varieties. large grains are the main criteria of seed selection by farmers. About 90 percent of Table 5. Proportion of seed planted from alternative market sources by agroecological zone and road accessibility in Niger, 1996-1997. Agroecological zone road accessibilitv and vear Less than 75 d~s ofLGP More than 75 days ofLGP Poor access Good access Poor access Good access 1996 1997 1996 1997 1996 1997 1996 1997 Number of observations 27 30 124 115 Seed sources Own seed stocks 99.7 81.7 97.4 83.2 92.8 87.0 90.8 77.3 Friends, neighbors, relatives 0.3 2.4 1.3 1.0 2.6 5.9 2.7 10.2 Village markets 0.0 15.9 1.3 3.2 4.4 6.7 5.7 12.3 NGO, Projects, SPU, and RC 0.0 0.0 0.0 12.6 0.1 0.4 0.9 0.3 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 3The varieties in the pre-release stage include: ICMV 89305, ICMV 92222, ICMV 94 206, MTDO 92, MTTY 92. 74 Because these varieties were not hy Fanners draw the largest share of their brids, it was very easy for fanners to planting seed from their own stocks. Sur maintain the improved characteristics by vey results indicate that about 83% of the selecting and retaining the seed from year total seed planted in 1997 was drawn from to year". Efforts to release and promote fanners' own stocks, 7% from relatives, current pearl millet varieties by the fonnal neighbors and friends; 9% purchased seed seed systems are warranted. The consis from the local village markets and 1% tent use of improved varieties may make from rural integrated projects, NGOs, it easier for farmers to maintain this seed production units and research sta gennplasm on their own. tions (Table 6). Fanners' own stocks is the preferred source of seed. As long as Pearl Millet Seed Production by the the climatic conditions are favorable, Formal and Informal Seed Sectors fanners will draw their entire seed needs from their own stocks. In 1996, for exam The relative importance of the quantity ple, which was a relatively good year, on of seed produced by each sector is mea average 93% of the total seed planted was sured with regard to the share of seed pro drawn from fanners' own stocks (Table duced relative to the total quantity of seed 6). This behavior is consistent across planted by fanners in Niger. Despite the agroecological zones and village accessi large investments in seed multiplication bility. and distribution, the fonnal seed sector has consistently produced a negligible Overall, the infonnal village seed sys share of the total seed planted by fanners tems provide the largest share of the total in Niger. Between 1991 and 1995, for ex seed planted by rural households regard ample, less than 1 % of the total seed less of agroecological zone and accessi planted was supplied by the fonnal seed bility to the main road. In contrast the for sector (Table 3). These results are consis mal seed system is a minor contributor to tent with the rural household survey the seed supply. Further improvement in which indicates that about 1 percent ofthe seed production should target local seed total seed planted in 1997 was from the supply systems. fonnal seed sector. (Table 5). Table 6. Cross-tabulation of rural households pearl millet seed stocks' levels by agroecological zones and road accessibility in Niger (1996-1997. Agroecological zone road accessibility and year Less than 75 days of LGP More than 75 days ofLGP Poor access Good access Poor access Good access 1996 1997 1996 1997 1996 1997 1996 1997 Number of observations 27 30 124 115 Percentage of seed stockholdin s Less than 25% 0.0 3.7 6.7 6.7 7.7 14.5 5.2 14.8 Between 25 and 50% 0.0 7.4 0.0 0.0 0.8 3.2 2.6 0.9 Between 50 and 75% 0.0 11.l 0.0 3.3 2.4 3.2 2.6 7.0 Between 75 and 99% 3.7 3.7 0.0 6.7 4.0 8.1 2.6 5.2 More than 100% 96.3 74.1 93.3 83.3 83.9 71.0 87.0 72.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 75 Seed Quality ofPearl Millet Seed Pro ject, the entire quality personnel left the duced by the Formal Seed Sector and seed production units due to lack of fund Farmers' Seed Stocks ing and incentives. Therefore, during the last five years, no quality control has been A seed ofhigh quality as defined by the performed by the SPU. Consequently, formal sector in Niger follows the stan NGOS, rural development projects and dards of the International Seed Testing farmers purchasing certified seed com Association (1ST A). They include a high plain of their poor germination rate; and percentage of germination rate (97-98%), are reluctant to purchase seed from state low moisture content (12%) and a high seed projects. Even in Lossa, where basic level of physical purity (98%). In addi seed is produced, the germination rates tion, genetic purity of about 98% is re are low, averaging 70 percent, far below quired through field inspections. In the in the ISTA standards. Formal seed systems formal sector, observable grain character are unable to meet the required standards istics such as: inert matter content, set by 1ST A and it is unlikely that they uniformity of color, uniformity and large will ever contribute significantly to seed grain size, shape, texture and low in quality with continuing funding con sect/disease attacks are the main criteria straints. of quality preferred by farmers. However, farmers did not point out germination rate In contrast, local village seed systems as one of the main seed quality criteria. offer high quality seed to end-users ac This may occur because farmers encoun cording to their set of standards. Labora ter few problems of germination. Though tory tests indicate that farmers' seed it may be an important standard in the in stocks have acceptable levels of germina formal market, no one perceived it to be tion, moisture content and low incidence problematic. Seed quality criteria per of seed borne fungus diseases consistent ceived by farmers differ from to those of with their set of standards. Farmers' pearl breeders. While the formal seed sector millet seed stocks were found to have ac emphasizes on some genetic characteris ceptable levels of germination and mois tics of purity or germination, farmers do ture content. Individual results on viabil not necessarily care. ity testing offarmers' pearl millet seed in dicate that the germination rate ranges During the period of involvement of from 71 to 98% with an average of 88%. seed projects (1975-1993), seed quality Fifty ofthe seed lots were above 87.5 per was low. Numerous complaints of poor cent germination. There was no differ seed quality from NGOs, rural develop ence in germination rate based on ment project managers and farmers were agroecological zone and accessibility but recorded. "Few visits were made by qual differences were found in moisture con ified personnel to inspect the conditions tent based on agroecological zones. The under which the seed was grown. Often, moisture content ranges between 7 to farmers did not respect the isolation dis 17% with an average of 9%. Only 10 per tance from other fields required to main cent of the seed lots were above the 12 tain purity in the grain grown for seed" percent moisture content. These rates (Couvillon, 1985). At the end of the pro- were higher in relatively more humid ar- 76 eas (more than 75 days ofLGP), but were lower end of granaries. These granaries kept at a reasonable level of moisture con are separated from their homes to avoid tent so that germination was not affected. potential fire hazards. Other traditional The results may explain why farmers do storage techniques such as the use of ash, not perceive germination as a problem tree bark, and leaves were not found. The since they do not encounter large germi average storage time is about 7 months nation problems. Moreover, at the current which corresponds to the time span be average high seeding rate of 10 kg/ha as tween harvest and planting: Few farmers practiced by farmers, it is unlikely that were found to store seed more than a year. they encounter germination problems Due to large risk associated with seed (Ndjeunga, 1998). drawn from other sources than farmers' own seed, exchange with friends and rela Pearl Millet Seed Storage tives or purchase from village markets, there is a high incentive for farmers to Due to the perishability of seeds as liv store their own seed. Failure to hold seed ing organisms and the seasonality of both stocks reflect farmers' inability to main seed production and use, seed storage is a tain and manage seed stocks. Therefore very important function. Seed storage is investment in training farmers to better undertaken for several purposes and takes store their seed may be warranted different forms, including long-term stor (Ndjeunga, 1998). age ofbreeder seed and other seed classes, the maintenance of varieties, and seed se Pearl Millet Seed Security Stocks curity stocks, especially useful during drought years. Poor storage is often asso In 1996 and 1997, national estimates ciated with poor seed quality. In this sec point to 20 and 39% of villages which tion, the relative efficiency of seed sys were short on cereal grain. The villages tems at storing seed, and ability to hold short on cereal grain are de facto consid and manage seed security stocks will be ered as being short on seed to plant. There examined. is no estimation of seed shortages per se, rather inferences drawn from cereal short Formal state seed production units are ages. From the government's viewpoint, a all equipped with large storage facilities. village which is short on cereal grain is A survey ofthose facilities indicated that automatically short on seed. Estimates of they were poorly ventilated, exposing 6,000 and 12,000 tons respectively in seed to a host of disease and pests hazards. 1996 and 1997 were reported for pearl This, in turn, partially explained the poor millet seed shortages of which 60% of seed quality reported by farmers, NGOs these are pearl millet seed (MAG/EL, and rural development project managers. 1998a). These figures are largely On the other hand, farmers store seed in over-estimated given the actual potential well-ventilated granaries. Pearl millet seed needs of about 17,000 tons of pearl seeds are often stored in the form of single millet seed. It is likely that seed provided or bulked panicles and rarely in the form to farmers through drought relief pro of grains and are always located out of the grams are merely consumed as grains. immediate reach of people, often at the 77 In contrast, survey results indicated ated with wholesale and retailing and the that very few villages were completely related functions of risk-bearing and fi short of seed to plant. In the survey, only nancing. Seed distribution can be very one village Garbey Gommande reported costly in environments where farmers are being completely short on seed because of scattered and located in poorly accessible very poor yields. In 1997, survey results areas. Seed production units in Niger indicate that 27 percent of rural house have encountered many difficulties dis holds held seed stocks less than their tributing seed to largely scattered farmers planting seed (Table 6). Paradoxically, a located in poorly accessible areas. Less high proportion of seed short households than 30% of the total seed produced was were found in favorable environments. actually planted. Farmers in poor environments tend to hold some seed stocks no matter how In contrast, the local pearl millet vil small it may be, because of fear to loose lage system is efficient at supplying seed adaptable germplasm to their harsh envi to end-users. Survey results indicate that ronment. Survey results indicated that 3 about 41 percent of farmers surveyed sup and 12% of rural households did not hold plied seed to other farmers, 19 percent any seed stocks in drought prone and were seed recipients and 40 percent were more favorable environments respec neither recipients nor seed suppliers. No tively. No significant differences in seed association was found between seed sup stockholdings by households was found pliers/recipients and agroecological based on agroecological zone and acces zones or accessibility. Seed suppliers are sibility. However, seed shortages are well often large grain producers. On average, a resolved through exchange between seed supplier produces 2,194 kg of grain friends and relatives or purchase through against 1,420 kg for seed recipients local village markets. There did not ap (Ndjeunga, 1998). These types of farmers pear to be a serious problem with seed should be targeted and assisted to under shortage resolution through the informal take seed distribution. seed system. However, there may be need to invest in training farmers to better seed Village market seed transactions are selection and storage methods. flexible and dominated by cash, followed by barter, exchange and free transactions, Pearl Millet Seed Distribution regardless of the agroecological zone and accessibility. In 1997, 50 percent of seed Seed distribution is a multi-faceted traded within the rural community was for process involving both the physical cash, 30 percent was free gifts, and the re movement and handling of seeds and the maining was barter transactions. A simi actual exchange of seed and money be lar trend was observed in 1996. Cash tween and among producers, traders and transactions are likely to be observed in farmers. It involves different yet closely large communities. Because of imperfect related functions including: physical lo information between households in large gistic operations (handling and transport), populated areas, farmers may not know market research, promotional activities, each other's seed stock holdings. This the buying and selling functions associ- problem is lessened in smaller communi- 78 ties which are made of farmers with fam from Nigeria also offer grain to Zinder ily ties, who interact frequently and share which in tum supply grain to Agadez, risks. Gift transactions are found in favor Tanout, and Diffa. Also, pearl millet able environments where more farmers grains are supplied by Mali and Nigeria to hold large seed stocks. Survey results in Niamey. Imported grain is solely used for dicated that 97 percent of gift transactions domestic consumption. Farmers claimed were located in more favorable environ that those grains are of poor quality and ments. Non-cash transactions are less are not adapted to their agroecological ar known. The level of reciprocal obliga eas. Thus, grain purchased as seed are do tions which sustain such transactions mestically produced. Seed purchased are needs further consideration. Limited data well known varieties. Grain traders in suggest that often farmers trade seed of Niamey differentiate imported grain from different varieties. domestic grain. There is often a margin of 10 cfa/kg between domestic and imported The average quantity of seed traded grain. Farmers as well as grain traders varied according to the type of transac may be used as seed distributors. tions and agroecological zone. Cash transactions are higher on average, 28 kg Seed Pricing and Costs ofPearl against 7 kg for gifts and 8 kg for barter. In Millet Seed Production the poor environment, survey results indi cate that in 1997 and 1996, there were no Seed pricing is an important compo barter or gift transactions. While, farmers nent of seed marketing that determines travel an average 11 km to purchase seed, the sustainability of seed multiplication the average distances for barter and gift and distribution activities along with transactions were estimated to 0.625 km costs of seed production. Sustainable seed and 0 km respectively. multiplication and distribution requires the average sale price to be higher than av Pearl millet seed channels are centered erage cost of production. Seed pricing is around the grain as many farmers pur dependent on the type of product. For va chase high quality grain as seed from vil rieties or composites, one cannot expect lage markets. However, farmer-to-farmer seed price to be far higher than grain price seed exchange remains the main distribu as grain is a perfect substitute for seed. In tion channel followed by farmer-to-grain this study, the margins over the cost of traders. Due to insufficient offer of grains production will be used as a criteria for fi through domestic production, there is an nancial efficiency of seed systems. Com active grain import from neighboring parisons of seed to grain prices will also countries such as Nigeria and Mali via be used to ascertain the potential demand Burkina Faso. In general, grain is chan for seed by farmers. neled through 4 distribution networks centered around Niamey, Maradi and During and after the period when seed Zinder. Large suppliers from Nigeria of projects were active, seed production fer grain to Maradi which in tum supplies units operated at heavy losses if we ac Tahoua, Arlit, and Niamey along with its count for depreciation on long -term assets domestic production. Large suppliers inherited by the Niger Cereal Research 79 Project (Table 3). Between 1985 and at 14 Fcfa/kg. This profitability occurred 1988, the average cost for producing cer because farmers are entirely responsible tified improved millet seed by SPU was for producing seed and assume all the pro estimated to 1,720 Fcfaikg, whereas it duction and distribution risks. These seed was sold to farmers at 100 Fcfa/kg less production units are simple seed traders than the average grain price, 130 Fcfa/kg who purchase seed as the demand is ex in local markets (Rachmeler, 1991). In pressed by N GOs, rural development pro 1996, the average cost of seed production jects and farmers. These seed multiplica was estimated to vary between 305 tion units neither perform quality control F cfa/kg in Kourougoussao to 1633 nor do they incur significant transporta Fcfa/kg in Lossa. Therefore, price of tion costs since seed is supplied to seed seeds was largely subsidized with regard production centers or regional offices of to seed sale prices which vary between agriculture. These examples show that 130 Fcfa/ha in Magaria to 250 Fcfa/ha in farmers could undertake seed production Lossa. To be financially sustainable, seed on their own if they are properly assisted. production units have to sell seed at prices The current intermediary role played by varying between 6 to 8 times the average SPU could well be undertaken by the grain price depending on SPU (Table 7). most efficient farmers or groups of farm At these sale prices, the margin (sale price ers assisted and tasked with multiplica minus average cost of production) over tion and distribution of seed. the cost of production per kg of seed sold by seed production units are all negative Appropriate Roles of the Formal and and varied between -0.49 and -0.90. For Informal Pearl Millet Seed Sectors every dollar invested in producing seed, on average seed production units incur Formal pearl millet seed systems are losses between 50 and 90 cents. There unreliable sources of seed supply which fore, this scheme is unsustainable. How offer poor access to improved varieties ever, if one does not account for deprecia and operate at high transaction costs. tion on long-term assets inherited from Therefore, it is unlikely that they will ever the NCR project, assuming that the cur be financially or economically sustain rent seed multiplication schemes do not able. Formal seed systems perform as need this heavy infrastructure, the SPU well as local village systems at develop could be profitable with profits estimated ing and maintaining varieties, offer vari- Table 7. Pearl millet seed and grain prices in seed production units in Niger, 1996. Seed Production Units Price (Fcfa/kg)2 Lossa1 Kourougoussao Guecheme Magaria Seed 250 150 175 130 Grain 157 97 131 151 Margins 93[59 [%] 53[52%] 44[34%] -21[-14%] 1 Basic seed production and the other centers certified seed production. 2 Nominal consumer grain price at planting period in the nearest market (average June & July) Source: Seed Production Units and SIM/OPVN (1998) 80 eties of comparable value with local vari could be encouraged and assisted in eties, and are weak in performing all other small-scale seed multiplication schemes. activities (Table 8). In contrast, local vil Even if this activity may not be economi lage seed systems in Niger are working cally profitable for them, it infers some fairly well. They perform well at supply social status. A large share of investment ing seed to end-users, maintaining and in seed systems should target the develop disseminating varieties, distributing seed ment of local village seed systems. at relatively low costs and maintaining ac ceptable levels of seed viability, purity, This situation could change if new va and seed health. This gives no reason to rieties or hybrids which may yield signifi believe that there is a structural problem cantly more than local or current varieties with pearl millet seed quality. Overall, the are made available in the market. Then, formal seed systems do not contribute the role of the formal seed systems could much in the development of seed systems be limited to providing breeder seed to ef as a whole. However, the informal system ficient farmers, groups of farmers or co as well, is unable to cope with contin operatives to produce seed of other gency seed needs. Other means to deliver classes (basic, registered and certified contingency seed needs to end-users seed) and facilitate the interaction be should target the development of local tween farmers, research institutes, rural village seed systems. development projects and NGOs. This will require additional investments in Local village systems have a number of training efficient farmers and groups to strengths that can be utilized to resolve seed production techniques via NGOs or contingency seed needs. A large number well-established local organizations. of varieties easily accessible by farmers are available at low transaction costs. References Seed quality of farmers' stocks is of ac Couvillon, w.e. (1985). Etude economique du ceptable quality. The local village seed programme de multiplication des semences au systems are endowed with potential seed Niger: Consultancy Report. Niamey, Niger: producers who are large farmers and United States Agency for International Devel opment Department. Table 8. Current performance assessment of seed activities performed by both the formal and in formal pearl millet seed sectors in Niger. Seed sector Activity Fonnal Infonnal Variety development/selection Strong Weak Varietal maintenance Strong Moderate Variety transfer/dissemination Weak Strong Quantity of seed supplied Weak Strong Seed quality (gennination rate) Moderate Moderate Seed storage Weak Moderate Seed promotion/distribution Weak Strong Seed security stocks Weak Moderate Cost of seed production Weak Strong 81 Cromwell, E. (1996). Governments, Farmers and Ndjeunga, J. (1998). Local village seed systems Seeds in a Changing Africa. Overseas Develop and pearl millet seed quality in Niger. Unpub ment Institute. CAN International. lished Report. ICRISAT Niamey. Wallingford. Oxon OX 10 8DE UK. Rachmeler, D. (1991). Les semences au Niger: une INRAN (1994). Catalogue Nigerien des Varietes nouvelle orientation politique. Unpublished de Cereales et de Legumineuses. Ministere de Report. Republic of Niger. I 'Agriculture et de I 'Elevage. Edition, 1994. Rohrbach, et al. (eds) (1997). Alternative strategies Mazzucato, V. and S. Ly. (1993). An economic for smallholder and seed supply. Proceedings Analysis of Research and Technology transfer of an International Conference on Options for of Millet, Sorghum and Cowpeas in Niger". Strengthening National and Regional Seed Sys Discussion Paper No 93-06 ISNAR. tems in Africa and West Asia, 10-14 Mar 1997, Ndjeunga, J. (1997). Constraints to variety release, Harare, Zimbabwe (Rohrbach, D.D., Bishaw, seed multiplication and distribution of sor Z., and van Gastel, A.J.G., eds.). Patancheru ghum, pearl millet and groundnut in Western 502 324, Andhra Pradesh, India: International and Central Africa. Pages 34-46. In Alterna Crop Research Institute for the Semi-Arid tives strategies for smallholder and seed supply Tropics. proceedings of an International Conference on SIM/OPVN (1998). Annuaire des Prix des Options for Strengthening national and Re Cereales au Niger. Systeme d' Information sur gional Seed Systems in Africa and West Asia, les Marches Cerealiers. Office des Produits 10-14 Mar 1997, Harare, Zimbabwe (Rohrbach, Vivriers du Niger. Republique du Niger. Avec D.D., Bishaw, Z., and van Gastel, A.J.G., eds.). la Collaboration du Projet Diaper III/CILSS. Patancheru 502 324, Andhra Pradesh, India: Inter Tassiou, A (1996). Contribution it l' Atelier de national Crop Research Institute for the rMlexion sur la Problematique Semenciere. Semi-Arid Tropics. ROCAFREMIIWCAMRN. 82 Review of Sorghum Hybrid Research in Niger Issoufou Kapran, Moussa Adamou, Gebisa Ejeta, and John D. Axtell Introduction nization. Current responsibility for agri cultural research in Niger is held by the Sorghum is the second most important Institut National de Recherches food crop in Niger following pearl millet. Agronomiques du Niger (INRAN) cre It is grown in rainfed regions ofthe coun ated in 1975. INRAN has a broad mandate try that receive approximately 400-800 to contribute to national food mm of annual precipitation. The annual self-sufficiency through genetic improve sorghum acreage has increased from less ment of crops like sorghum and pearl mil than 500,000 hectares in 1961 to two mil let. Sorghum breeding goals include the lion hectares in 1996. In contrast, national development and release of genotypes average yields declined from 0.6/t ha in with high and stable yields, resistance to 1961 to 0.2 t ha in 1996 (FAO/SMIAR, biotic and abiotic stresses, and possessing 1997). In a country where only 12% ofthe good food quality. In the 1980s, sorghum land is available for rainfed agriculture, breeders at INRAN in collaboration with cultivation in marginal lands with declin the International Sorghum and Millet Pro ing fertility and the wide use of gram (INTSORMIL) added a hybrid test poor-yielding local cultivars are largely ing component to the selection of im responsible for the limited agricultural proved cultivars. It was felt that the use of producti vi ty. heterosis is in line with INRAN's man date. Research on sorghum hybrids in It is widely acknowledged that yield in Niger has moved steadily forward and is creases offer the best approach to solve expanding gradually from experimental increasing food demands in developing activity towards a private commercial op countries (Pinstrup-Andersen and eration. The purpose of this paper is to re Pandya-Lorch, 1995). Hybrids are known view progress in breeding sorghum hy to capitalize on the phenomenon of brids in Niger, including the initial work heterosis to produce higher yields than by IRA T, evaluation of experimental hy open-pollinated varieties, even under brids by INRAN, and efforts by INRAN stress conditions. Adoption of hybrids to promote production and marketing of would also be beneficial to the environ hybrid seed in the private sector. Experi ment since high and stable productivity ence elsewhere shows that hybrids have may restrain the desperate use ofmarginal the potential of modernizing the agricul soils (Kapran et aI, 1995). tural sector of a country. Crop improvement research in Niger Sorghum improvement by IRAT was started by the Institut de Recherches Agricoles Tropicales et des cultures During the 1950s and 1960s, emphasis vivrieres (IRAT), a French research orga- was placed on the breeding and/or intro- 83 duction of open-pollinated varieties. Ini came more systematic with two tially, local land races were collected and INTSORMIL supervised - M.S. research evaluated at several research stations. Af projects (Kapran, 1988; Tyler, 1988). Se ter going through mass selection, some of lected hybrids gradually moved from ob the best adapted locals including BDF, servation nurseries to advanced yield tri JanJare, and Bagoba were released back als across several locations in Niger. Elite to farmers. Some of the land races were hybrids were also evaluated in the West also crossed to exotics in pedigree breed African Sorghum Hybrid Adaptation ing, which led to the development of im Trial (WASHAT). Specifically, the sin proved varieties like L30 (137-62 x gle cross between TX623A and MR732, JanJare) released in 1974. Although com later named NAD-l, emerged as the most mercial sorghum hybrids were intro promising based on the summary of the duced, tested and found superior to cumulative data collected over the years open-pollinated varieties during the both on-station and on-farm. IRAT program, they were rejected on the basis of grain quality problems, and the Results perception that they were excessively in put-dependent (Chantereau and Adamou, A summary of annual analyses of data 1977). reported by INRAN sorghum breeders is presented in Table 1. Experimental hy INRAN/INTSORMIL Evaluation of brid yields averaged approximately 2 Experimental Sorghum Hybrids tons/ha. The best hybrids yielded up to 6.5 tons ha whereas best variety checks never With the advent of collaborative re exceeded 3.8 tons ha. These results con search between INRAN and firm the findings of the two graduate the INTSORMIL, the sorghum breeding pro sis projects conducted at Purdue Univer gram in Niger got access to germplasm sity and initiated the trust in sorghum hy and experiences from other parts of the brid research in Niger. Kapran (1988) world. Sorghum hybrids were synthe reported heterosis values of45% under ir sized by crossing adapted pollinators (R rigation and 66% in rainfed conditions of lines) and seed parents (A lines) of di Niger. In the same tests, hybrids also out verse origins that were provided essen yielded local checks by 61 % with irriga tially by the INTSORMILlPurdue Uni tion and 49% under rainfed conditions. versity sorghum program. The objective Tyler (1988) found that hybrids with par was to evaluate the superiority of hybrid ents grouped as exotic, intermediate, or sorghum cultivars under the marginal local, were higher yielding than their re growing conditions prevalent in much of spective parents by an average of 127%, Niger (Kapran et aI, 1997). 83%, and 66%. Respectively from the re sults of these early thesis projects and Approach subsequent national tests conducted through INRAN and INTSORMIL col Starting in 1983, experimental sor laboration, the agronomic value of ghum hybrids were tested at several heterosis for Niger's agriculture was INRAN research stations. Testing be- demonstrated. 84 Table 1. Yields of sorghum hybrids at research stations in Niger. Grain yield Hybrids tested Trial mean Best hybrid Local variety Year (number) t ha- 1 t ha- I tha-1 1984 22 2.3 5.6 1985 17 4.0 1986 149 2.5 2.7 1.1 1987 147 2.6 5.5 3.2 1988 81 1.9 4.9 2.7 1989 67 2.8 6.5 3.8 1990 49 234 5.0 3.0 1991 78 1.8 2.0 1.0 1992 88 1.9 4.6 Source: INRAN sorghum breeding annual reports, 1984-92. Table 2. Yield of elite hybrid NAD-l under farmer conditions in Niger. Grain yield Year! Local variety NAD-1 % NAD-1 over local Activity t ha- 1 t ha- I t ha- I 1993 Demonstrations 2.4 1994 Demonstrations 3.3 1995 on-farm trial 1.0 1.6 60 1996 on-farm trial (INRAN) 0.7 1.7 143 1996 ON-FARM TRIAL (ROCARS) 1.0 1.8 80 1997 on-farm trial 1.6 3.1 87 Source: INRAN sorghum breeding annual reports, 1993-97; ROCARS 1997. NAD-1, a medium-maturing, white regional sorghum network (ROCARS), seeded hybrid with good tolerance to the yield ofNAD-l was 80% higher than mid~season drought, was the most attrac the average offarmer checks. This kind of tive cultivarto breeders and farmers alike. field performance created a strong farmer In national trials, its yield average was 3.1 interest in NAD-l over the years and tons/ha between 1986 and 1992. Also in stirred up the issue of reliable seed pro the 1989 W ASHAT trial, it ranked third duction to satisfy fast-growing requests. among 20 entries for grain yield at nine locations across West and Central Africa Towards Marketing of Sorghum Hy (ICRISATIWASIP, 1989). Estimates of brid Seed in the Private Sector NAD-1 yield potential with farmer man agement were obtained starting in 1993 Experimental Seed Production (Table 2). Averages vary between 1.7 tonlha and 3.3 tons/ha. Despite severe Starting in 1989, NAD-1 seed has been drought conditions, the hybrid was supe regularly produced in isolated fields at rior to Mota Maradi (MM), an INRAN research stations, to evaluate the early-maturing and widely adapted local feasibility ofcommercial hybrid seed pro variety. Also, from on-farm trials of the duction in Niger. Best locations and other 85 environmental effects, as well as the im supporting new seed producers. As the portance of field management on N AD-1 latter gain confidence, INRAN will grad seed production, have been closely evalu ually abandon the production of commer ated. Some of the most crucial factors in cial seed. Beginning seed producers will clude the availability of irrigation water be encouraged to limit their field size to for a more reliable seed production, a high no more than 0.5 ha, to receive good pa level of soil fertility to counter frequent rental seed and advice on field manage soil variability, and in the case ofNAD-1, ment, seed storage, etc. A manual on hy the need for staggered planting since pa brid seed production was developed and rental lines have different maturities. We used in 1997 and 1998 to provide practi have estimated that under good manage cal training to seed producers and their ment, up to 1.5 tonslha of NAD-1 seed technicians. This will be discussed in a may be produced. It has been repeatedly separate paper during this workshop. demonstrated that hybrid seed production is possible in Niger. Table 3 shows that a modest increase in hybrid seed production has been Initiation ofHybrid Seed achieved, and that private producers are Production in the Private Sector gradually taking a bigger role in quanti ties/produced. Production is still far be Experience elsewhere in the world low needs (Table 3), but it also shows the shows that the production of good quality existence of opportunity for commercial seed in adequate quantity is best realized seed activity. Based on a planting rate of by the private rather than the public sector 10 kg/ha, private involvement will be (L.R. House, 1995; pers. comm). Having needed to produce the 2,000 tons of seed a hybrid that appeals to farmers and can be which will be required even if only 10% reliably produced are key elements for of the sorghum acreage is planted with seed marketing. In 1995, INRAN and hybrid seed. This is especially true since INTSORMIL took definite steps to facili the public sector does not have the more tate the production and marketing of than 1,000 hectares of good land which NAD-1 seed by interested private parties. would be for seed production ofall crops. Since there are no local seed compa Conclusions and Prospects nies interested in seed of hybrid cereals, it was decided to sell the seed at an average Research on sorghum hybrids in Niger of eight times the price of grain on the has demonstrated conclusively the supe market or CF A 800. Selling seed and pro rior agronomic potential of hybrids in viding a profit margin encourages the pri comparison with open-pollinated variet vate producer, whereas subsidizing the ies. Based on the example of the NAD-l availability of seed by the government hybrid, INRAN and INTSORMIL are creates a situation with which the private now collaborating in helping hybrid seed sector cannot compete. production and marketing to become a vi able activity in the private sector. In the INRAN continues to collaborate with last two years, more hybrid seed than ever INTSORMIL in training and technically has been produced and easily sold in 86 Table 3. NAD-l seed production by public (INRAN) and private sectors in Niger and estimates of total needs. Quantity (kg) Total Year Producer by producer (kg) 1995 INRAN 700 Private -none 1996 INRAN 1,400 1,400 Private negligible 1997 INRAN _2,813 7,101 Private 4.288 1998 INRAN 8,000 26,000 Private 18,000 I. Total sorghum acreage II. Total seed needed III. Seed needed for 10% of IV. Area required to In 1996 (hectares) for I. (Kg) sorghum acreage (kg) produce seed for III. 2,000,000 20,000,000 2,000,000 (Hectars)1 1333 1 Asswning a planting rate of 10 kglha in II; a seed yield of 1500 kgiha in IV. Sourrce: INRAN/INTSORMIL sorghum breeding, consultant reports 1998. Niger. This should move forward quickly Kapran, L, M. Adamou, M. Abdou, S. Souley, N. Kondo, J.D. Axtell, G. Ejeta, T. Tyler. 1997. since there is interest on the part of farm Expression of heterosis and prospects for mar ers to purchase seed, on the part of gov keting of sorghum hybrids in Niger. Poster pre ernment to pull out of seed production, sented at the International Symposium on the Genetics and Exploitation of Heterosis in and on the part of private individuals to Crops. August 17-22,1997. CIMMYT, Mexico invest in seed production. D.F., Mexico Kapran, L, J.D. Axtell, and G. Ejeta. 1995. Sor ghum hybrids: a viable technology for sustain There are still important questions to ing Sahelian agriculture. Workshop on be answered, including the need for capi Technology Development and Transfer to im tal to help beginning seed producers, for prove Natural Resource Management in West Africa. Niamey, Niger, 18-22 September 1995 farm equipment to improve production Kapran,1. 1988. Evaluation of the agronomic per and conditioning of seed, and for adapta formance and food quality characteristics of ex tion of marketing techniques. Last but not perimental sorghum hybrids in Niger, West Africa. M.S. thesis, Purdue University, West least, quality control must be addressed in Lafayette, Indiana a planned way through a seed law and cer Pinstrup-Andersen P. and R. Pandya-Lorch. 1995. tification program. It is however felt at The future food and agricultural situation in de veloping countries and the role of research and this time, that getting production to a level training. Twenty-first J. C. Snyder Memorial of routine activity is the most crucial, so lecture in Agricultural Economics, Purdue Uni that farmers and seed producers take ad versity, Feb. 6, 1995. Tyler, T. A. 1988. Heterotic pattern and combining vantage of improved technologies like ability for agronomic and food grain quality NAD-l. traits in Exotic x Exotic, Exotic x Intermediate and Exotic x Local sorghum hybrids in Niger. M.S. thesis, Purdue University, West Lafayette, Literature Cited Indiana Chantereau, J., and M. Adamou. 1977. Principaux resultats et orientation sur sorgho au Niger. Agronomie Tropicale xxx-3:299-303 FAO/SMIAR. 1997. FAO, Rome, Italy 87 Experience of NAD-l Seed Production at Djirataoua Irrigated Perimeter, Maradi Region of Niger Laouali Dogo and Oumarou Issa Background (cotton, pepper, onion, tomatoes, wheat and others). We will begin with a short description of the Djirataoua irrigated perimeter (IP). For food crops like sorghum, yield av Geographically, it is located 10 km south erages are 2.55 t/ha for local varieties, 3.1 of Maradi and 700 km east of Niamey. t/ha for improved varieties, 4.7 t/ha for The IP was constructed in the more fertile NAD-l (on real size plots since 1997). central basin of the Goulbi 'n Maradi River, with funding from the former Experience with NAD-l Seed Maradi Rural Development Project Production at Djirataoua (1982). The IP covers 512 hectares ofirri gable land, out of 7300 hectares initially INRANIINTSORMIL involvement: in available for improvement. Water is ob 1995, after several demonstrations of the tained using electric pumps in deep wells, NAD-l cultivar on the perimeter, INRAN 10 to 35 meters depth. Forty of the origi and INTSORMIL started a new approach na143 wells are still in use. Water is taken of technology transfer: farmers them to the plots in cement canals. Soils are selves got involved in producing the hy sandy, easy to drain, and flat. A 2400 m brid seed. These farmers were contracted protection wall separates the perimeter to produce the seed for INRAN during the from river floods. The socio-economic two crop seasons. In exchange for mone objectives of the IP are to help rural fami tary support, farmers were supervised by lies reach food self-sufficiency through INRAN staff to produce, thresh, and bag intensive agriculture. It is estimated that NAD-l seed for INRAN. However, farm 876 heads of household own a plot, ie. ers got to keep the grain produced on the more than 10,000 people live off the IP. A male parent. This deal was arranged in farmer plot is usually about 0.32 hectare. collaboration with the co-op manage Crop production is year round, with two ment. One important learning experience main seasons of activity. During the was the failure of this activity during the rainfed season (May to September), crops dry season due to excessive stemborer grown are sorghum (50%), millet (2%), damage, and the success achieved during maize (0.5%), hot pepper (15%), toma the main season in part due to the skills of toes (5%), cotton (25%), and others (2%). the Djirataoua IP farmers. During the dry season (November to March), wheat (68%), onion (15%), green ONARA seed project: this is a support and hot peppers (15%), tomatoes (3%), project intended to bring technical and fi and others (2%) are grown. Net estimated nancial backing to producers on the IPs. It monetary returns to the IP farmers are of is funded by the Belgian government and Fcfa 200 million from cash crops alone located within ONAHA, the government 88 agency that manages IPs in Niger. The kg/ha of 15-15-15). Ridges were made in project's goal is to increase the yield of the field following the first rain or irriga the most prevalent crops grown on the IPs tion. Staggered planting was used to en through the use of improved seeds pro sure good nick between flowering of the duced by the farmers themselves. A yield two parents; the female parent was trial was organized with several improved planted 7 to 8 days later than the male par varieties including NAD-l, while two ent which was also planted in five border farmers were selected to produce the rows. Planting was done in hills and seed. Following a successful field day, plants were thinned to three per hill for the NAD-l seed was bought by the project to male parent; the female parent required sell to other farmers. transplanting because of poorer germina tion so hills were thinned to two plants Training: in 1997 the ONAHA seed and the third was transplanted. Irrigation project contacted INRAN/INTSORMIL was done as needed. Urea was applied on for a training session on techniques ofhy young seedlings and later at the boot brid seed production. A second training stage. Transplanted plants received fur session was organized by INRAN in 1998 ther urea because INRAN scientists where both farmers and technicians were checked the floral initial and realized the invited. These training sessions provided transplants were lagging behind. Other hands-on experience on management of measures included the shaking of sor seed production and ensuring seed quality ghum panicles to improve pollen dis both in the field and during storage. persal, getting rid of off-types in male rows and shedders in female rows, and Selecting farmers for hybrid seed pro spraying against sorghum midge with duction: the co-op management, CYPERCAL. At maturity, the male par ONAHA staff and project leaders met to ent was harvested first and the panicles set guidelines: farmers must have some taken away from the field. Unfortunately, previous experience with seed produc one late rain fell on the panicles of the fe tion, be hard workers, ready to learn, and male parent before harvest. All harvested ready to work with other farmers. They panicles were dried under the sun over a must have neighbors that are available, five-day period before threshing. well organized and open to innovation. The first farmer selected was Oumarou Considerations: technically it was very Issa, who had some formal training, expe important to shield seed fields from other rience in producing wheat seed, and sorghums. All neighboring fields were worked for the pilot farm. His neighbors thus planting with the male parent of are mostly young, dynamic farmers. NAD-l. The male parent, MR732 is rec ognized and accepted by neighboring Actual NAD-J seed production: the se farmers as a good food grain variety. All lected field was in the middle of the coop technical recommendations were fol erative 'block' and has homogenous soil lowed including spacing and thinning, leveled for ease of irrigation. The field fertilizer application, off-type clearing, was cleaned and spread with animal ma seed cleaning and storing. Seed yields nure (1.75 tlha) and NPK fertilizer (156 recommended by INRAN and 89 INTSORMIL were between 1.7 and 2.0 the four co-ops is now interested in their tlha under optimum conditions, and our own seed production 'block'. It may be farmers averaged 2.4 t/ha. The activity said that this was a successful case of was supported financially by the co-op technology transfer, since a good quantity management, ONAHA seed project and of NAD-l hybrid seed was actually pro the farmers themselves. The money was duced by farmers with no major prob used for purchase of parental seed from lems. Some largely unsuspected con INRAN, land preparation, fertilizers, straints included poor germination of the field treatments, harvest, processing and female parent seed, its susceptibility to storing the seed. All processing was done many pests requiring several treatments, by the co-op to ensure uniform quality late season rain on mature unharvested seed. Farmers received Fcfa350 per kg of seed, and limited drying and storage ca hybrid seed, based on the cost of sorghum pacity. grain in 1997 and taking their investment and labor into account. The co-op decided It is expected that land allocation for to sell processed seed at Fcfa650 per kg to hybrid seed production will rise to 2.56 ha its members. Overall the coop bought 859 from the currentO.8 ha, and it may be even kg of seed from farmers at a cost ofFcfa greater with new demand for seed. We 304,150; after cleaning, the coop sold 834 therefore conclude that the effort should kg to farmers at Fcfa 542,100. be pursued since the experience was posi tive, seed producers are willing, the desire Perspectives to transfer technology is there, and there is technical and financial support. With hu Positive achievements from this activ man conviction and irrigation facilities ity included returns to the farmers actu available, it remains for Djirataoua ally involved in seed production as well as co-ops to improve their facilities for dry to the local community. The two farmers ing and storing the seed, and for research made net monetary gains ofFcfa 184,000 to improve on the reddi~h pigmentation and harvested 1.15 tons of grain on the on the grain ofNAD-1. Finally we thank male parental line for their family con all those that contributed to our experi sumption; there was a bonus in stover ence, particularly the ONAHA Seed Pro since the two parental lines produced an ject for its technical and financial support, estimated 1.25 tons of stover. Learning and INRAN and INTSORMIL for their new techniques and making a profit inter technical support. ested neighboring farmers so that each of 90 Hybrid Sorghum in West Africa D.S.Murty Cultivation of hybrid crops in western 623A MR 732), led to its successful re Africa started about ten years ago with lease to farmers. maize. Research on hybrid sorghum was initiated in the early seventies and cur Other national programs in western Af rently sorghum hybrids are formally re rica, namely, Burkina Paso, Cameroon, leased in Nigeria and Niger. However, Cote d'Ivoire, and Mali evinced interest large scale production and distribution of in sorghum hybrids but did not have suffi hybrid seed is lagging far behind. While cient scientific data and experience to there is considerable infrastructure in proceed further. The suitability of hybrids countries such as Nigeria to support oper to subsistence farmers in traditional West ation of private/public sector seed indus African farming systems was also de try, there is little, if any, in most of the bated. The lack of organized seed produc francophone countries. In this context, the tion infrastructure, either public or pri current workshop on sorghum and millet vate, was a major deterrent to experimen hybrid seed production is timely and rele tation with hybrids, and still remains so. vant. A brief summary of the progress In 1986 the West and Central African Sor made in sorghum hybrid research, pro ghum Research Network (WCASRN) au duction and extension in the West African thorized ICRISA T to carry out breeding region and the problems encountered are and trials with sorghum hybrids in coop presented below. eration with national programs. Early research in Nigeria indicated that In the West African Sorghum Hybrid experimental sorghum hybrids out Adaptation Trials (W ASHAT) conducted yielded variety controls by 40% (An during 1986 -1995, at 6 to 13 locations per drews, 1975). Subsequently a few hybrids year, top ranking hybrids yielded 3.6 t ha- 1 based on photosensitive parents were re compared to 2.6 t ha- 1 for early maturing leased by the Institute for Agricultural variety controls. The local controls which Research (IAR), but these did not reach were generally late maturing were the the production stage (Obilana, 1981). In poorest in all these years. Grain yields of Senegal, early maturing hybrids based on the best hybrids in individual tests were as American female parents yielded> 6 t ha· 1 high as 6 t ha- 1 in some years. Regression under irrigation, compared to 4.5 t ha· 1 analyses of the multilocation data each under rainfed conditions (Chantereau, year showed that the selected hybrids 1983). In Niger, Kapran (1988) and Tyler maintained their yield superiority consis (1988) obtained highly encouraging re tently across low- and high-yielding envi sults with sorghum hybrids across loca ronments in all the years. Stability param tions. Continued efforts in Niger with one eters of hybrids for grain yield were gen promising hybrid, namely, NAD-I (Tx erally equal to 1 or sometimes > 1, indicating that the hybrids were positively 91 responding to improved production envi as early as 1990-1991. Parental seeds ronments. were also supplied to Premier Seeds Co (formerly Agricultural Seeds Co and later Agronomic research at ICRISAT, Pioneer Hi-Bred Seed Nigeria Ltd.) and Kano, Nigeria showed that an early ma Birma Agricultural Seeds and other inter turing hybrid was superior to local land ested public or private agencies. Assis races in grain yield at any planting date. tance in AxR seed production was given Superiority of hybrids for grain yield was by ICRISAT to Premier Seeds Co, Birma also observed at varied population density Agricultural Seeds and the NSS. Premier and nitrogen levels. The harvest index of Seeds Co continued to produce and mar hybrids was about 0.40 compared to ket seeds of ICSH 89002 NG on a small 1 <0.20 for the local control and 0.34 for the scale (1 to 10 t ha- ) since 1993. During early maturing variety (Murty and Tabo, each of the years 1990-1992, ICRISAT, 1997). The hybrids also exhibited favor Kano conducted short-term training pro able land equivalent ratios (LER) com grams in hybrid seed production for staff pared to early maturing varieties and local of public and private agencies involved in controls in intercropping experiments seed multiplication. with groundnut, cowpea, soybean and pi geon pea. Hybrids such as ICSH 89002 In Burkina Faso, sorghum hybrid trials NG (ICSA 38 x ICSV 247) and ICSH were conducted by the IN.E.R.A. (Insti 89009 NG (ICSA 39 x MR 906) were tute National d'Etudes et de Recherche tested in farmer managed on-farm trials in Agricoles) scientists at the Farako Bfr re the Kano and Jigawa states of northern search station for several years. Hybrid Nigeria in 1991 and 1992 and average ICSH 89002 NG was demonstrated by grain yields from 21 such trials grown in them to farmers in their own fields in the these two states ranged from 1.24 to 1.3 5 t adjoining villages. In cooperation with ha- 1 for hybrids compared to 1.07 t ha- 1 for ICRISAT, Mali, hybrid seed was pro the farmers' variety. duced under irrigation in the summer sea sons of 1996 and 1997 and the seeds were Two early maturing hybrids from the used for demonstrations nationwide. Ef Institute for Agricultural Research (IAR) forts are underway by the INERA to en and two from ICRISAT (ICSH 89002 NG courage the establishment ofa seed indus and ICSH 89009 NG) were successfully try. tested in multilocation trials of IAR and the Nationally Coordinated Sorghum Re In Mali, the CMDT (la Compagnie search Project (NCSRP) of Nigeria. They Malienne pour Ie Developpement des were recommended for release in 1994 Textiles) has been interested in promoting and the National Committee on the improved sorghum cultivars in the cot Naming, Registration and Release of ton-farming systems. In view of some en Crop Varieties and Livestock Breeds for couraging on-farm trials with ICSH mally released them to farmers in early 89002 NG in the San area, CMDT started 1997. Meanwhile, ICRISAT supplied the producing hybrid seeds on their own male and female parental seeds of the hy farms (in the rainy season) at Sougoula in brids to the National Seed Service (NSS) cooperation with ICRISAT, Mali, since 92 1997. The seeds are currently being used References for on-farm trials in all the cotton produc Andrews,D.1., 1975. SorghumgrainhybridsinNi ing zones of Mali. geria. Expl.Agric. 11:119-127. Chantereau, J., 1983. Selection d'hybrides de The establishment of a seed industry in sorgho pour Ie centre-nord et Ie nord du Sene gal. L'Agronomie Tropicale 38 (4) 295-302. the region, either private or public, will go Kapran,l., 1988. Evaluation of the agronomic per a long way in extending hybrid sorghum formance and food quality characteristics of ex and other improved crops to the farmers. perimental sorghum hybrids in Niger, West Africa. M.S. Thesis, Purdue University, West In areas where irrigation is not feasible or Lafayette, Indiana. economically viable sorghum hybrid seed Murty, D.S., and R.Tabo, 1997. Potential of sor production can be safely done in July ghum hybrids in West and Central Africa. Pages 211-214 In Actes de I 'atelier de restitution du plantings (in the Sudanian zones). In programme conjoint sur Ie sorgho some parts of western Africa, head bugs ICRISAT-ClRAD, 17-20 Mar. 1997, Bamako, (Eurystylus oldi) could cause seed and Mali. Collection Colloques, Montpellier, CIRAD- CA, 315p grain quality problems on sorghum hy Obilana, A. T., 1981. Sorghum hybrids for Nigeria, brid production plots, and require inte 510 NAFPP National SorghumlMilletiWheat grated pest management methods for an Workshop. ABU, Zaria, Nigeria. effective, safe and economic control. Er Tyler, T.A., 1988. Heterotic pattern and combining ability for agronomic and food grain quality got (Sphacelia sorghi) infection has also traits in exotic x exotic, exotic x intermediate been noticed on male sterile lines flower and exotic x local sorghum hybrids in Niger. ing in cool weather. Minor problems re M.S. Thesis, Purdue University, West Lafay ette, Indiana lated to the hybrids could eventually be solved by breeding hybrid parents derived from exotic local crosses. 93 Heterosis and Hybrid Sorghum Seed Production in West Africa Aboubacar Toure, H. Frederick W. Rattunde, and Inoussa Akintayo West African landrace varieties of sor der these conditions. F 1 sorghum hybrids, ghum are well adapted to traditional with their vigorous early growth, fast rate methods of cultivation. However, they of growth, and ability to efficiently utilize are not highly responsive to increased lev limited moisture, can produce higher els ofinputs and thus have a relatively low yields under stress conditions than inbred grain yield potential. Conner and Karper varieties. first pointed out the advantages of hybrid sorghum in 1927. Stephens has discussed A study of sorghum hybrids in Kenya the first possibilities of economic hybrid showed that heterosis could improve pro seed production in sorghum since the dis ductivity in moderate and extreme covery of the genetic male-sterility in Su drought conditions as well as in more fa dan grass in 1929, which was transferred vorable moisture environments to grain sorghum. (Haussmann et aI., 1998). The hybrids out-yielded the two local varieties by 12% The greater advantage of hybrid yields across the eight environments. as compared to cultivars, has been dem onstrated throughout the world. Many The beneficial effects of heterosis in studies showed that under stress condi sorghum in the West Central African tions, the yields of both hybrids and vari (WCA) region could possibly have con eties decline, but frequently the magni tributed indirectly to the selection of tude of difference, percent wise, is greater Guinea type sorghums with relatively for hybrids compared to the situation high outcrossing rates. A study in when growing conditions are favorable. Burkina F aso found an outcrossing rate of By comparing the performance of hy 20% in traditional Guinea varieties brids, improved varieties, and landrace (Ollitrault et aI., 1997). The challenge for cultivars House et al. (1996) showed a F 1 hybrids in the region is to surpass the yield increase for the hybrids over the im performance of Guinea landrace varieties proved varieties from 15 to 66%, and over that possess significant heterozygosity locals from 7 to 131 %. rather than highly inbred varieties. Guinea varieties account for over 70% of Rainfed agriculture is subjected to un the sorghum in the area. predictable and unreliable rainfall that can cause yield reductions and even total The economic viability of F 1 sorghum crop failures. It is believed that superior hybrids in the WCA region would depend hybrids identified under local conditions on: can increase and stabilize yield levels un- 94 1) Economic value ofF\ hybrids (mar research on millet and sorghum started in ket or subsistence value) 1931. In 1938,Bambeybecamethehead quarters of agricultural research in French • Extent of grain yield superiority Sudan with a regional vocation of the over local varieties, sahelo-soudanian zone of West Africa • Quality of grain for milling, stor from Senegal to Niger including countries age, of the coast from Guinea to Benin. • Desirability of grain for food prepa rations, In 1950, the French West Africa gov • Value of stover for livestock feed ernment established the Agricultural Re and other uses. search Center of Bambey (C.R.A.). In 1960 this center was transferred to the Re 2) Cost of hybrid seed relative to eco public of Senegal. nomic returns. Before 1951 and until 1960 the Agri 3) Stability ofF \ hybrid production be cultural Research Center of Bambey was ing adequate: the only organization mandated for sor ghum improvement in most West African • Resistance to insect pests, e.g. countries. The main sorghum breeding headbug, objective was the improvement of local • Resistance to diseases, e.g. cultivars through mass selection. anthracnose, grain mold, • Ability to regulate date of maturity (through photoperiod sensitivity or In the 1960's after independence, most controlled date of sowing.) West African countries established their own National Agricultural Research In 4) Adequate production and distribu stitutes (NAR). In addition to the im tion of good quality seed. provement oflocal cultivars, NARs initi ated breeding programs through hybrid The development and use of hybrids ization, introduction of exotic material, has been generally slow in developing and exploration of hybrid vigor. countries. This is particularly true in West and Central Africa. However, there West African Local Germplasm-Based Hybrids is experience in sorghum hybrids in West Africa, which provides a base for further In the late 70's an attempt to cross each developments. of the 800 entries representing the differ Early History of Sorghum Breeding in ent sorghum races of the Malian Sorghum West Africa Collection (Guinea, Durra, Caudatum) on to the cytoplasm male sterile line Tx623A In 1921 an experiment station was cre was made. The objectives of these ated at Bambey (Senegal) to conduct re crosses were to monitor the potential search on peanuts, using the infrastruc combining ability of Malian landraces ture of a farm created in 1913. The first with currently available introduced male sterile line. 95 Heterotic effects were clear in all ous geographic pattern (Toure and stages of plant growth. Seedling vigor Scheuring, 1982). and seedling drought tolerance were much better than the Malian parents In 1982 at Sotuba a backcrossing (Toure, 1980). Generally, plant growth program was used to convert Malian local was much more rapid and lush than that of B lines to A lines. In 1985 a total 22 parents. At all pre-floral stages of plant female plants issue from 6 different growth the F 1 hybrids demonstrated a tol recurrent parents were selected through 5 erance to drought. generations of backcrossing. The progenies should have 97% of Malian Panicle shape was intermediate be recurrent parent genes with the cytoplasm tween the parents. The female parent, a male sterility of Tx623A. Unfortunately Caudatum-Zerazera, has a long upright these Malian A lines were lost between panicle with numerous short seed 1986 and 1990 when work on hybrid branches. Hybridization of lax panicle development was halted. Keninke and Kende with the Caudatum-Zerazera resulted in upright Significant and positive heterotic ef semi-open panicles with the same num fects were recorded for grain yield per bers of seed branches as the Guinean par panicle and for panicle yield components; ent but with increased seed numbers. The seed number and seed weight (Table 2). dense panicle Gadiaba (Durra) and However, we see no scope for direct ex Hegari (Caudatum) had long and dense ploitation of F 1 hybrids involving the panicles in hybrids combination with the Tx623A and Malian variety parents. The Tx623A parent (Toure, 1980). turtle-back seed shape (typical of all Caudatum) was dominant in hybrid com Cultivars were evaluated for B/R fertil binations with all Malian races. That seed ityrestorerreactiononAI cytoplasm. The shape made the grain more difficult to frequency of fertility restorer lines among dehull than Malian Keninke grain that is the different sorghum race groups in Mali oval, rounded and symmetrical. The is presented in Table 1. The distribution grain of F 1 hybrids undercoat carries an ofmaintainer lines did not show any obvi- astringent tannin, which renders the grain Table 1. Rand B reaction to Al cytoplasm of cultivars among sorghum races in Mali Races Durral Membranaceu Guinea! Guinea! Durra! Keninkel m Caudatuml Kende Keninke Gadiaba Gadiaba (Niofionto) Feterita Total Total Cultivars 138 436 151 29 14 7 775 FJ hybrids 52 116 34 16 5 4 227 B lines 19 41 16 8 2 0 86 Rlines 33 75 18 8 3 4 141 96 Table 2. Heterosis for grain weight per panicle, seed number per panicle and lOOO-grain weight in Tx623A *CSMhvbnds. . Grain weight Seed per panicle 1000 seed per panicle Race of male Mean %het. Number %het. Mean %het. parent Entry (g) (g) CSM277 67.5 143 3386 93 19.9 25 Keninke A623*CSM277 142.5 115 5636 93 25.3 11 Keninke CSM400 30.0 261 1466 191 20.5 18 Keninke A623*CSM400 86.3 127 3777 73 22.8 22 Keninke CSM432 15.0 17 768 34 19.5 8 Keninke A623*CSM432 117.5 148 4972 150 23.6 9 Kende CSM433 22.5 245 833 153 27.0 45 Kende A623*CSM433 82.5 14 2848 6 29.0 10 Kende CSM440 113.3 58 5841 96 19.4 7 Gadiaba A623*CSM440 120.5 28 5569 33 21.6 27 Gadiaba CSM396 11.7 142 868 97 13.4 7 Gadiaba A623*CSM396 76.7 4143 18.5 CSM427 16.7 1214 13.7 A623*CSM427 115.0 4618 24.9 CSM 134 28.3 1997 14.2 A623*134 45.0 2354 19.1 CSM40 148.3 4091 36.2 A623*CSM40 156.7 5934 26.4 CSM46 90.0 2654 33.9 A623*CSM46 117.5 3394 34.6 CSM59 43.3 1065 40.6 A623*CSM59 113.3 3462 32.7 B623 50.0 2442 20.5 Het.: heterosis undesirable for major food uses. The F 1 efficient approach. Sterilizing and dwarf hybrids with Gadiaba males did not have ing Malian maintainer lines would open a seed undercoat and made excellent up a new range of possibilities for devel plants. However, their growth was so lux oping hybrids for the region. uriant that they were unusually suscepti ble to post-floral drought and charcoal rot Current Hybrid Research Activities caused by Macrophomina Phaseolina U sing Local Germplasm (Tassi) Goid. Characterization of Guinea and partial The development of acceptable hy guinea test crosses was initiated jointly by brids using local germplasm can be ap IER and ICRlSAT in 1998. Two male proached from a number of angles. Traits steriles were used (SP 47562A and SP involving relatively few genes such as 52161A) that were derived in the seed subcoat and peri carp thickness could ICRISA T grain-mold breeding work in be addressed with a single cycle of selec India. Four Guinea varieties and four par tion involving currently available local tial guinea lines (Bimbiri Soumale!S-34 varieties and introduced male sterile derivatives) were used as pollinators. lines. More complexly, inherited traits Further test crossing on Malisor 84-7 A and improvement of multiple traits would (with A2 cytoplasm) with guinea and par require more cycles of breeding. Popula tial guinea lines will be initiated this year. tion breeding approaches would offer an 97 The development of guinea A lines is plants, relatively late, and had large underway at Farako, Burkina Faso (Da spear-shaped panicles with large seed. Sansan, personal, communication), with one line in advanced stage of sterilization. A total of 59 hybrids and their male A combining ability study within the parents were planted in 1980 in an obser Guinea Race was initiated by vation trial in Mali (Sotuba, Cinzana, CIRAD/IN.E.R.A. in Burkina Faso Babougou and Baramandougou). The which could provide useful insights. Tx632A hybrids grew very well at all 4 locations. At flowering there were 3 hy Introduced Hybrids (Caudatum Hy brids, which became badly infected by brids) sooty stripe. During grain filling, 15 of the hybrids were infected by charcoal rot Mali and lodged. The average panicle grain weight for the male parents was 52.6g, for From 1979 to 1989 many experimental the Tx623A female 69g, and for F 1 hy and established F 1 hybrids from various brids 92.4g. breeders (ICRISAT, USA) were tested in Mali. Those hybrids represented a bal The West African Sorghum Hybrids ance of the common Indian females Adaptation Trial (W ASHAT) from (296A, 2219A, 2077A, A1296, A378) ICRISAT was initiated in the mid-1980s and the Texas female Tx623A as well as a and was conducted at different locations good representation of the best experi in Mali. In 1987,21 hybrids were com mental males from the Indian national pared to Naga white, Hageen dura, program and the ICRISA T population Framida and Malisor 84-3 for grain yield and grain-mold resistance breeding pro and adaptation. Hybrids generally grams. Those hybrids were planted out as showed good levels of grain yield. The an observation trial at different locations. grain yield ofICSH 232 was double ofthe local check Malisor 84-3 and resistant to During 1979 and 1980, hybrids from leaf diseases (Table 3). The 1988 evalua ICRISAT/India using 296A, 2219A, and tion of 19 hybrids at Cinzana and 2077 A as female parent were tested in dif Samanko showed significant differences ferent locations in Mali. Plat growth and compared to the local check. The hybrids drought tolerance were very good ICSH-88042, ICSH-330, ICSH-507 and throughout all locations. Every plant of ICSH-88038 yielded more than 3000 every hybrid made a normal panicle with kg/ha in both locations (Table 4). Eigh no blasting and set normal seed at allioca teen hybrids were tested at Sotuba in 1989 tions. Grain weights were determined for adaptation. The grain yields were and food trials were conducted to deter good in general. Most of the hybrids mine food quality traits. No differences yielded more than the check Malisor 84-7 were detected between hybrids for food (Table 5). quality. In general, the 2219A hybrids were the best of the lot. They were earli est and had small, cylindrical panicles. The 296A and 2077 A hybrids were sturdy 98 Table 3. West African SOrj~hum Hybrids Adaptation Trial (WASHA T), Sotuba, 1987 Hybrids Grain yield 50% Flowering Plant height (Kg/ha) (days) jem) ICSH 232 6244 52 223 ICSH 526 5777 56 248 ICSH 647 5730 54 155 ICSH 336 5728 52 183 ICSH 644 5409 57 232 ICSH 233 5196 52 233 ICSH 646 5128 57 225 ICSH 643 4929 52 232 ICSH 569 4925 52 233 ICSH 479 4925 53 228 ICSH 648 4713 53 208 ICSH 229 4693 51 230 ICSH401 4567 50 227 ICSH230 4473 52 225 ICSH 231 4340 51 245 ICSH 109 4280 53 228 ICSH 507 4249 53 228 ICSH 641 4216 58 192 ICSH 645 4077 54 253 ICSH 331 3958 55 212 ICSH 642 3788 50 240 Naga White 5253 51 248 Hageen durra 4091 50 187 Framida 4063 56 285 Malisor 84-3 3353 54 265 Mean 4724 53 227 CV(%) 19 2 12 Table 4. West African Sor~hum Hybrids Adaptation Trial (WASHA T), Cinzana Samanko, 1988. Grain yield (kg/ha) 50% Flowering Plant height Head bug (days) (em) damage rating * Hybrids CZ SK Mean ICSH 88042 3060 3690 3379 72 190 8.5 ICSH 330 3000 3630 3315 70 195 8.5 ICSH 507 3160 3443 3302 69 200 7.5 ICSH 88038 3380 3195 3288 70 205 8.5 ICSH 88040 2750 3608 3179 72 195 7.5 ICSH 88045 2410 3698 3054 70 190 7 ICSH 643 2320 3698 3009 77 205 8 ICSH 88039 2960 2738 2849 71 200 8 ICSH 369 2760 2738 2749 73 200 8 ICSH 336 2430 3008 2719 70 150 8 ICSH230 2650 2738 2694 70 190 8.5 ICSH 232 2160 3218 2689 68 190 7.5 ICSH 88043 2160 3113 2637 73 185 7 ICSH 88041 2000 3233 2617 71 205 8 ICSH 88044 2340 2798 2569 72 195 8.5 ICSH III 2500 2318 2409 66 220 6.5 ICSH 231 2530 2190 2360 66 215 8.5 ICSH 780 2240 2460 2350 66 200 8.5 ICSH 527 1920 1568 1744 70 200 8.5 Local (check) 2140 938 1539 74 390 3 Mean 2550 2910 2723 71 206 7.7 CV(%) 23 14 CZ: Cinzana SK: Samanko * Head bug damage scoring. scale based 1 to 10 with 6 ~ average: 7 to 10 - susceptible. 99 Table 5. West African Sorghum Hybrids Adaptation Trail (WASHAT), Sotuba, 1989. 50% Hybrids Grain yield Flowering Weathering (k~/hal (days) Scoring* Tx623Ax MR732 3700 66 4 ICSH 479 3500 68 5 ICSH 89003 3500 69 5 ICSH 507 3400 69 5 ICSH 89002 NG 3400 67 5 ICSH 88038 3350 69 5 ICSH 780 3300 62 4 ICSH 642 3290 62 4 ICSH 330 3180 69 5 ICSH 88039 2900 68 4 Tx631A x SUC36 2730 68 5 ICSH 109 2680 67 5 ICSH 88042 2500 70 5 ICSH III 2430 66 4 ICSH 89001 NG 2300 63 4 ICSH 646 1820 72 5 ICSH 230 3600 64 5 89004NG 3550 69 4 Malisor 84-7 1500 66 2 Mean 2190 66 • Weathering scoring, scale based I to 5 with I ~ resistant; ~ susceptible. Nigeria An action committee for a hybrid sor ghum project encouraged the interest in From 1968 to 1970 experiments were hybrid sorghum in Nigeria prior to 1980 conducted in Nigeria (Samaru) to study (Obilana, 1982). Five sorghum hybrids plant densities and nitrogen fertilizer lev were released in 1985, and breeder seed els. Three density levels (35,880, 73,810 was provided to the National Seed Ser and 155,000 plants per ha), three levels of vice and the private company AGSEED. nitrogen fertilizer (53, 132,211 kg N per Seed production slowed down and finally ha) and four genotypes (three varieties stopped. The only interest was in the vari and on hybrid) were included in the study. ety SK 5912 which was used in commer In the 1968 experiment there was no sig cial malting for the production of lager nificant response to the nitrogen fertilizer. beer. In all but variety 2123 the higher densities were better than the lowest. The hybrids Burkina Faso gave significantly better yields than vari eties, but at the lowest density, they were In 1986, about 454 preliminary hybrids only superior to varieties 2141 and 5921. were observed along with their parents at The hybrids achieved this extra yield by Farako-Ba. The hybrids produced from means of heavier rather than more heads. crosses involving disease-resistant par Grain weight changed little or decreased ents identified in 1985 were generally free slightly as plant density increased (An from leaf diseases and showed no symp drews, 1977). toms oflodging. Female parents ICSA 1, 100 ICSA 2, ICSA 11, ICSA 26, ICSA 38, and food quality and was released as NAD-l ICSA 40, which were introduced by the in 1992. The average yield on-farm was ICRISA T Center, combined good levels 3000 kg/ha. Currently INRAN is devel of leaf disease resistance with high grain oping the production and marketing of yield. seed ofNAD-I. A project was initiated to train private entrepreneurs to take up the Niger hybrid production from INRAN to a smooth transition. INRAN is also in Hybrids were introduced in Niger in creasing its capacity of saturating the ex the 1960s, but were discontinued because isting demand for NAD-l seed. of grain quality problems and perception that hybrids required good level of inputs. Limitations of the First Sorghum Hy In 1986 the hybrid experimentation brids in the WCA region re-started under the INTSORMIL INRAN collaborative program. Kapran Out of the 17 sorghum growing coun (1988) reported the testing of 90 hybrids tries in West and Central Africa, only Ni at several sites in Niger. Hybrids sur geria and Niger have formally released passed the locals by 61 % with irrigation sorghum hybrids (Murty, 1996, personal and 49% under rainfed conditions. Yield communication). However, the area cur advantage of hybrids over parents rently planted in hybrids does not exceed (heterosis) was 45% higher under irri 1000 hectares (Table 6). gated conditions, and even higher in the rainfed experiments (66%). Early hybrid evaluations involved only a limited number of closely related hy Recent Research on Caudatum-based brids and, quite often, inadapted male Sorghum Hybrids sterile genotypes were used as the female parents of the hybrids. Quite often the They hybrid ICSH 89002 (ICSA 38 x male parents were lines derived in variety ICSV 247) was planted in on-farm tests in development programs that did not spe the 600-900mm rainfall area ofMali (Bla, cifically target hybrids. Fana) from 1996 to 1998. Farmers showed much interest in this hybrid for its early maturity, desirable stover qualities Table 6. Area (ha) of sorghum and percent for feeding livestock, and acceptable hybrid seed for specific countries. grain quality for preparation of To. The Sorghum % hybrid susceptibility of ICSH 89002 to head sorghum India 12,800 40 bugs was the most serious constraint for Australia 1,000 100 this hybrid. The CMDT Seed Production Nigeria 4,000 1 Farm successfullyproauced seed ofICSH Niger 1,000 0 Sudan 6234 6 89002, although the germinability of the Mexico 1,650 100 seed still needs to be improved. Argentina 674 100 U.S. 4,940 100 In Niger, Tx623A x MR732 was found to be a high yielding hybrid with good Source: Bruce Maunder, 1996. 101 Constraints to adoption of the first hy tant issues for hybrid sorghums such as brids included: lodging, severe leaf dis identifying the agronomic/quality traits eases, head bug damage, grain mold, required for acceptance and economically striga and grain quality. In areas where viable production, and addressing issues yield levels were high, lodging quite often of seed production and distribution. canceled any hybrid yield advantage. In areas where lodging would not have been The network's second strategic objec so serious, head bugs and grain mold pre tive deals with enhancing market oppor vented the use of hybrids. It is not know tunities and strengthening demand for what impact hybrid sorghums would have sorghum. The successful development of on the disease and insect situation if hy new sorghum-based products, alternative brids were grown over large areas. uses of sorghum, and processing equip ment envisaged by this objective will be The photoperiod response of most lo of great relevance for sorghum hybrids. cal varieties provided an advantage over Stronger market demand for sorghum will photoperiod insensitive hybrids, as the provide a favorable environment for high photoperiod sensitivity ensures flowering yielding hybrids. The quality require at the end of the rainy season with the ments of future hybrids will be deter grain maturing under dry conditions mined by the end uses demanded. (Toure et aI., 1996). Guinea varieties were able to produce vitreous grain under Conclusion post anthesis drought stress while under identical conditions, exotic-based materi The exploitation of heterosis is one of als failed to adequately fill the grain. The the best strategies available for sorghum grain of local guineas gave higher improvement. Major impact could be decortication yields compared to the early achieved in WCA when sorghum hybrids exotic hybrids and as well as possessing are developed that dramatically exceed good grain quality, pest resistance and the yield potential oftraditional varieties, storage qualities. possess adequate resistance to diseases and insects, and provide acceptable grain West and Central Africa Sorghum quality for processing and food prepara Research Network (WCASRN) tion. These hybrids would generate great farmer enthusiasm and interest on the part The recently developed strategic plan of government officials and seed produc ofWCASRN provides the framework for ers. the networks activities for the next five years. Hybrids can play an important role Improved sorghum seed production in addressing the network's strategic ob and distribution to farmers is required in jective of improving sorghum production most West African countries. The exist and productivity. Collaborative breeding ing seed production structures are not activities are envisaged for development well organized to handle F 1 hybrid seeds. of novel hybrids as well as varieties. Ac The success of hybrid production will tivities aimed at enhancing stakeholder therefore depend on organizing such participation can address several impor- structures. 102 References Obi lana, A.B. 1982. A report ofthe action commit tee on hybrid sorghum project for the technical Andrews, D.J. 1977. Plant density and grain yield sub-com~ittee of the national sorghum project ofNigerian sorghums. Samaru Research Bulle for commIttee on green revolution, Mar. 1982. tin 279. Institute for Agricultural Research, Federal Dept. of Agricultural Lagos Nigeria Samaru Ahmadu Bello University PMB 1044 Ollitrault, P., J.L. Noyer, J. Ch~nterea~, and J.e. Zaria, Nigeria. Pp 55-60. Glasz~ann. 1997. Structure genetique et Haussmann, B.I.G., A.B. Obilana, A. Blum, P.O. dynamlque des varietes traditionnelles de Ayiecho, W. Schipprack and H.H. Geiger. sorgho au Burkina Faso. In: Gestion des 1998. Hybrid performance of sorghum and its Ressources Genetiques des plantes en Afrique relationship to morphological and physiologi des Savanes. Actes du colloque, Bamako, Mali cal traits under variable drought stress in Kenya. 24-28 February, 1997. pp 231-240. Plant Breeding 117:223-229. Toure, A., F.R. Miller, and D.T. Rosenow. 1996. House L.R., B.N. Verma, G. Ejeta, B.S. Rana, I. Heter?sis and combining ability for grain yield Kapran, A.B. Obilana, B.V.S. Reddy. De and YIeld components in guinea sorghums. Af veloping Countries Breeding and Potential of rican Crop Science Journal, vol. 4 No. 4 pp. 383-39l. ' ~ybrid Sorghum. In Proceedings of the Interna tIOnal Conference on Genetic Improvement of Toure, A. 1980. Etude de l'effect heterosis chez Ie Sorghum and Pearl Millet. September 23-27, sor~h? au~ali. ~emoire de find'etudes. Cycle 1996-Lubbock, Texas, USA. Ingemeur d Agnculture. Institut Polytechnique Kapran, I. 1988. Evaluation ofthe Agronomic Per Rural de Katibougou, Mali. formance and Food Quality Characteristics of Toure, A.B. and J.F. Scheuring. 1982. Presence des Ex~erimental Sorghum Hybrids in Niger, West genes mainteneurs d' androsterilite Afnca. M.S. thesis, Purdue University, West cytoplasmique parmi les varietes locales de Lafayette, Indiana. sorghos au Mali. Agronomie tropicale Tome XXXVII. 103 Rapporteur's Report D.S. Murty Current Status of Hybrid Sorghum Seed Production Day 1 - Session 3 Presentations were made by B.Sidi, J. food products with the least "off colors" Ndeunga, LKapran, L. Dogo and O. Issa, and therefore tan colored guinea plant D.S. Murty, A. Toure and A. Joshua types would find better application. In the SADC region, early maturing tan plant fe Discussion Summary male and male parents have been bred and can be evaluated for use in the Sahelian The hybrid sorghum being produced is zones. It was also suggested that photo mainly for human consumption by the sensitive female parents bred at IAR, Ni farmer but eventually could be used either geria during the seventies could also be for food or feed. It was clarified by the looked into for their potential use. In INRAN staff that NAD-l grain is suitable Egypt, hybrid sorghum grain is being for producing acceptable t6 and couscous. used in composite bread formulation However sometimes, particularly when (80% wheat: 20% sorghum) to save im the grain is exposed to rain before harvest, portation costs for wheat. There was con the flour color was red and not appealing. cern about competition of sorghum with The identification of sorghum hybrids maize in the wet zones of Ghana. There with superior grain and food quality is a was some discussion on what could be a matter of time. In Farako Ba, fair price of sorghum hybrid seed in com LN.E.R.A.lBurkina Faso scientists have parison with grain. Currently in India, the bred new traits. These are yet to undergo hybrid seed is sold at grain price x 10 ra advanced tests. Tan plant color results in tio. 104 Current Status of Hybrid Pearl Millet in West Africa K. Anand Kumar, B. Ouendeba, S. Boureima, S.c. Gupta and S. Ouattara Abstract The discovery of cytoplasmic male-sterility (CMS) in pearl millet (pennisetum glaucum) and the breeding and release ofseed parent Tift 23A in the early sixties, made the production ofsingle cross hybrids (SCa male-sterile x inbred restorer) possible in India and contributed to spectacular grain yield increases. Male-sterile Tift 23A was found not to be adapted to West Africa. Early work on variety x variety hybrids and in bred x variety hybrids demonstrated that increases in yield potential are obtainable without the use of CMS. However, such hybrids were not put into production. In 1986, exploratory research on pearl millet hybrids was initiated at the ICRISAT Sahelian Center. Our initial efforts were devoted to the development ofinbred x variety hybrids (IVH), because adapted male-sterile lines were not available. IVHs offer oppor tunities of using new normal, non-male sterile inbred lines as "seed parents" thereby avoiding the time consuming process of developing of adapted male-sterile seed par ents. Extensive evaluation ofIVHs indicated that compared to the Open-Pollinated Va rieties (OPVs), IVHs were early by two days, exhibited a wider range ofplant height and earhead length, produced more tillers, and showed around 10% yield advantage. Al though there were exceptions, higher yielding OPVs tended to produce higher yielding IVHs. A study using 12 environments created using different levels of moisture, nutri ents, and plant densities indicated that IVH hybrids gave a stable performance and high grain yields across all the environments. Thus, the investment in variety breeding also contributes to hybrids. Studies on pro portion ofIVH hybrid seed produced showed the presence of up to 20% inbred parent seed had little effect on hybrid grain yields. Evaluation of established male-steriles from India and the USA indicated a need to breed CMS seed parents in situ in West Africa. A first series of 18 male-sterile and maintainer lines was developed and used in the breeding and evaluation ofSCH and top cross hybrids (TCH-the CMS based equivalent of IVH hybrids). A second series of male-steriles is now under development. A study using TCH hybrids made from crossingfive landraces withfour male-sterile lines indicated that the TCHs gave better grain yields and were more stable than the Approved as Conference Paper Number 1367 ofICRISAT. 105 landraces themselves, offering opportunities to combine local adaptation with high yield potential. TCH seed yields obtained in seed production plots show that with a female:male ratio of1:2 it is possible to harvest 0.31 t ha- I ofhybrid seed and 1.1 t ha- I of pollinator variety. Back-of-the-envelope' calculations show that profits using more op timal4:2 ratio would be between $650 and $850 from a hectare ofhybrid seed produc tion. Based on experience gained thus far, criteria for breeding IVH and TCH are given. Encouraged by the results obtained, a collaborative ICRISAT-WCAMRN Regional Pearl Millet Hybrid Trial was initiated. In order to capitalize on the yieldpotentialofhy brids, some general improvements in soil fertility are essential. Problems ofinfrastruc ture related to seed production, assured food grain markets, and diversification ofend uses must also be overcome to allow hybrids to be profitably used by smallerfarmers. Introduction recognized: (i) to initiate exploratory work on pearl millet hybrids in West Af The discovery of cytoplasmic rica, (ii) to study the disease reactions of male-sterility (CMS) in pearl millet hybrids and varieties, and (iii) to gather (Pennisetum glaucum) and the breeding information on performance of hybrids and release of CMS Tift 23A seed parent under moisture stress. Based on these in the early sixties made the production of conclusions, exploratory work on the de single cross hybrids (SCH, male-sterile x velopment of pearl millet hybrids was ini inbred restorer lines) possible in India tiated at the ICRISAT Sahelian Center in which contributed to spectacular yield in the mid-eighties. Thus, a major propor creases. Although Tift 23A was made tion of research on pearl millet seed par available to breeding programs in West ent and hybrid development in West Af Africa at the same time, pearl millet hy rica was conducted at .ICRISAT -Niger brids were not extensively tested. Tift (ICRISAT Sahelian Center), and collabo 23A was not adapted to the environments rative exploratory research with of West Africa for the following reasons: INTSORMIL was conducted in Mali and an early flowering time; susceptibility to Senegal. Experience gathered and prog downy mildew; the absence of early vigor ress made to date are summarized in the resulting in poor emergence and stands; following sections. and occasional breakdown ofmale-steril ity. Therefore, extensive resources were Types of Hybrids Possible not devoted to breeding SCHs, but only in Pearl Millet exploratory research was directed to wards inter-varietal or top cross hybrids Because pearl millet has a protogynous (TCH, male-sterile x open-pollinated va reproductive system, five types ofhybrids riety). At the last International Pearl Mil are possible (Table 1). Of these, the most let Workshop held in 1986 in Hyderabad, popular are the Single Cross Hybrid India, major conclusions related to hybrid (SCH, Male-sterile seed parent - A x In development in Africa (ICRISAT 1987) bred restorer male parent - R) and the Top were drawn. The following needs were 106 Table 1. Types of I!Ybrids Dossible in Dearl millet. Designation Developed by crossing: Fertile inbred x Open-pollinated variety Inbred x variety hybrid (IVH) or Pro-hybrids Using protogyny as a mechanism to facilitate pollina tion between the parents Cytoplasmic male sterile x fertile inbred Single cross hybrid (SCH) Restoration of fertililic in the F IJybrid is critical. Top cross hybrid Cytoplasmic male sterile x Open-pollinated variety* (TCH) Restoration of fertility in the F 1 hybrid is critical. Pollinator could be a local landrace or an improved variety. F l( A x B) x male fertile restorer. Three-way hybrid (TWH) Seed parent is an F 1 between an A line of one geno type and a genetically different maintainer line. Possi ble to use when the use of a seed parent of low pro ductivijyhas to be circumvented Open-pollinated variety x Open-pollinated variety Variety x Variety hybrid Open pollinated varieties, whether improved or farm ers' local cultivars. Rlmfoducibili!y...I2oor. *SCH hybrids made from two male-fertile inbreds are not practical due to nicking difficulties. cross Hybrid (TCH, Male-sterile - A x portunities of using new inbred lines as Open-pollinated variety, OPV). seed parents, (ii) if cytoplasmic male-sterility is not involved, the F I hy In the mid-'80s, as adapted male-ster brid is always fully male-fertile, and there ile lines were not available to develop and is no requirement for the male parent to be test SCH and TCH hybrids in West Af a CMS restorer, (iii) multiplication of a rica, much of the research at ICRISA T straight inbred is relatively easier than Sahelian Center was concentrated on In multiplication of the A line (AxB), (iv) bred x Variety Hybrids (lVH). Here, a the tedious process of searching for per normally male-fertile inbred line is used fect maintainer lines, and the backcross as a female parent, as protogyny in pearl ing involved to produce male-sterile lines millet leads to a situation where the inflo is no longer needed, and (v) fewer nicking rescence is initially only female-fertile for problems, because of the naturally long a period of one to several days before flowering duration of the pollinator vari anthesis occurs - during this period it is ety (desirable to have the pollinator reach functionally a male-sterile, so, if pollen 1/3 pollen shed before seed parent only from a chosen male parent is present, protogyny). hybrid seed will result. IVHs could be rapidly bred because of (i) the greater op- 107 Historical Background A trial of 28 SCH hybrids (developed from four male-steriles introduced from Between 1967 and 1975 IRAT scien India x four inbreds developed at tists in West Africa investigated the levels Bambey) was conducted at Bambey dur of heterosis possible from inter-varietal ing the 1983 rainy season. The hybrid - 81 crosses (between West African land A x IBMI 8108 produced the highest races) and top cross hybrids. The latter grain yield (46% superior to the control I were made with lines of good combining variety Souna III which yielded 1.3 t ha- ) ability derived from one population followed by 81 A x IBMI 8207, 111 A x crossed, using protogyny, with another PS 90-2, and 111 A x IBMI 8108. All the open-pollinated variety and included hy hybrids were significantly less suscepti brids between dwarf lines and tall ble to downy mildew than Souna III, but pollinator parents where the seed natu more susceptible to grain smut. All the rally produced was shown to be 95% hy hybrids were shorter both in stature and brid seed (Lambert 1982). Inter-varietal earhead length compared to Souna III. hybrids gave yields up to 59% higher than Two male-sterile lines (lIlA and 81A) the best local cultivar. Though these hy and two testers, (IBMI 8108 and IBMI brids were not put in to production, the re 8206) were the best general combiners in sults are very relevant today as they dem respect to grain yield production. onstrated the increases in yield potential that are obtainable without the use of In 1984, a trial of 24 hybrids was con CMS, even within the adapted genetic di ducted at Nioro and Bambey. The trial versity of West Africa. was composed of two inbred variety hy brids, seven single cross hybrids, (se In the early '80s research at the Insti lected from a 1983 yield trial), and 15 hy tute of Agricultural Research, Samaru, brids that were selected in 1983 from an Nigeria showed that the then available observation nursery. Based on the mean male-sterile lines (5141 A, 5054 A bred in over two locations, eight hybrids pro India as downy mildew resistant versions duced significantly more (43-62%) grain from Tift 23 A) and hybrids developed yield than adapted variety Souna III (Ta from them were highly susceptible to ble 2). The highest yielding hybrid was downy mildew (Nwasike, 1987). ICMH 8407 followed by ICMH 8413, ICRISAT, in collaboration with the ICMH 8411, and ICMH 8403. The hy Institut Senegalais de Recherches brids, in general, produced a higher num Agricoles (ISRA), started a pearl millet ber of earheads but had shorter earheads improvement program in Senegal in than Souna III. Until very recently, the 1977, funded by UNDP. In 1982 this col limiting factor for production of hybrids laborative program evaluated eight in continues to be the availability of adapted bred x variety (lVH) hybrids at Bambey. male-sterile lines. Our observations indi Only one IVH hybrid, IBMI 8108 x cate that the frequency of Al maintainer Souna III, gave significantly higher grain genes in West Africa material is low. yield than the control Souna III. A second Though male-steriles 81A, and IlIA, hybrid IBMI 8108 x IBV 8401 was eye were retained, there was little follow-up catching, and retained for further testing. on hybrid development and testing due to 108 Table 2. Summary performance of pearl millet single cross hybrids for six characters aver aged over two locations (Nioro and Bambey, Senegal), ICRISAT-ISRA Collaborative p ro ram, ralIlL . season 1984 Entry Pedigree Grain Time to Plant Ear Downy Smut yield 50% height length mildew (%) 1 (t ha· ) bloom (em) (em) (%) (davs) ICMH 8403** III A x IBMI 8\08 2.3 53 208 47 11 15 ICMH 8407 11IAxF4195 2.6 52 214 42 5 29 ICMH 8411 III A x IBMI 8108·58·3 2.4 52 210 49 II 16 ICMH 8412** 81 A x IBMI 8206 2.3 49 217 35 4 21 ICMH 8413*' 81 A x IBMI 8207 2.4 52 217 36 5 20 ICMH 8414*' 81 A x IBMI 8108 2.3 51 152 36 \0 23 Control Sauna III variel~ 1.6 61 264 62 12 13 Mean 1.9 55 196 44 9 18 SE± 0.16 0.6 5.7 1.3 1.8 3.9 •• , Second year testing, promoted from 1983 trial. the termination of the ICRISAT-ISRA and produced two times more grain yield collaborative program in 1985. Later than local and improved varieties (90-110 1 studies (e.g., Ouendeba et al. 1996) have days, 0.7 t ha· ). Because the season was also suggested that use of pearl millet hy dry, all the testcrosses were free from brids can contribute to both increased downy mildew. However, our observa grain and fodder production. tions between 1983 and 1987 indicated that male-sterile 81 A, though resistant in Breeding of Seed Parents India, was extremely susceptible to downy mildew in most of West Africa When the pearl millet breeding pro (95-100% both on 81 A, and 81 B). gram at the ICRISAT Sahelian Center was initiated in 1983-84, a small propor Evaluation of 21 male-sterile and tion of time was devoted to explore the maintainer pairs in 1987 (including possibilities of hybrid development. In male-sterile lines incorporating 1984, using 81 A (lCMA 1 from Pennisetum violaceum cytoplasm from ICRISAT, India), 92 testcross hybrids ORSTOM, Niger and Tifton, GA, U.S.) were generated and evaluated during the indicated that all ofthem were susceptible rainy season. During the 1984 growing to downy mildew. Reselection within es season, there was a severe drought. Total tablished A and B lines showed that rainfall was only 260 mm, the lowest re downy mildew susceptibility continued corded in the previous 78 years, with an to be a major constraint. This indicated August rainfall of only 56 mm (vs. the the need to breed seed parents for West long-term average of204 mm). In parts of Africa in situ, using adapted inbred lines. western Niger the local landrace (Haini-Kirei, 110 days to maturity) failed Our initial attempts to incorporate ster to produce any grain. However, the ile cytoplasm in to adapted genetic back testcrosses performed extremely well; grounds of (Ex Bornu x Togo), Souna (ex they were early (85-90 days to maturity) Mali) were not encouraging. The inci- 109 dence of downy mildew on plant x plant pairs that are currently being evaluated pairs was as much as 70% and sterility for downy mildew resistance and stability was not stable. Male-sterile ICMA 90111 of sterility. from ICRISAT in India was found to be free from downy mildew. In addition, fur All our first and second series of ther selections were made within plant x male-sterile lines have Al cytoplasm. plant crosses in the downy mildew nurs This cytoplasm appears to have ery in male-sterile lines ICMA 90111, constraints related to complete expres DSA 134, PMC 23, and one derived from sion of male sterility and complete resto the cross (ITMV 8001 x INMV 8220, in ration of male fertility. It is known that corporating 81 A sterile cytoplasm). male-sterile Al lines can revert to pollen These were used in the development of shedders at a low rate but sufficient to ne thefirst series ofmale-sterile lines as well cessitate careful rogueing of both A line as in the breeding and evaluation of SCH increase and hybrid seed production and TCH hybrids to determine the type of plots. We have just begun to use the ~ cy hybrids suitable for West Africa. Cur toplasm source (Hanna 1989) and have 20 rently 18 male-sterile and maintainer AlB pairs derived from related genetic pairs derived from these four genetic backgrounds. This cytoplasm does not backgrounds are available. These exhibit seem to have modifiers for sterility genes a wide variation of several traits including and has fewer environmental effects on time to flowering, plant stature, and the complete expression of male sterility earhead characteristics. These and fertility, thus simplifying the devel male-steriles have been bred using the opment of male-sterile lines. The sterile downy mildew nursery at Sadore to en anthers look different from Al steriles in sure downy mildew resistance but there is terms of their size, shriveling and translu a need to extensively test these in the re cence but are readily distinguishable from gion to determine stability of this resis fertile anthers. tance. In Nigeria male-sterile development Male-sterile lines under development was initiated in 1996 by introducing A4 in the second series are derived from and B populations derived from the Nige inbreds from the cross (vr W, a genetic rian Composite d2 (NCd2BC3 - in ~ cyto stock from University of Nebraska plasm) from I CRISA T - India. During the [UNL], Lincoln x Sadore local); 8lBHT 1997 off-season, these ~ and B popula (cross between 81 B x Tiotande, ex IER, tions were planted at Dadin-Kowa, pro Mali); (ARD 125 accession from Togo x ducing 86 plant-to-plant crosses. These [90PV 6 x 13] inbred from the UNL, Lin 86 NCd2BC4 pairs were evaluated for coln); (81 BHT x Sadore Local); and male-sterility and downy mildew resis ([Togo-2 x INMG-3] cross between tance at Bagauda during the 1997 main Iniari an early landrace from Togo and a season. The male-sterility was confirmed gene pool developed by ICRISA T in all the ~-plants. Fifty-one downy mil -INRAN x IP 8056, a dwarf accession). dew resistant AlB pairs were selected and We have just completed the fourth 129 AlB plant-to-plant crosses were pro- backcross and have selected 55 CMS 110 duced. These BC5 progenies were sown at a female parent, the dwarf population Kadawa and Dadin-Kowa during the 3/4HK-B78 was used. Initial evaluation 1998 off-season. Seventy-five A-lines with 24 testcrosses indicated that the TCH were crossed by corresponding B-lines. and IVH hybrids had similar yield levels l We plan to retain 20 morphologically di (a mean of 1.9 t ha- ), out-yielded the con verse pairs after evaluating these 75 AlB trol variety (CIVT) by 23%, and took sim pairs (BC6) for downy mildew reaction, ilar time to 50% bloom (a mean of 56 stability of sterility, and uniformity at days). TCHs with male-sterile 81 A Bagauda and Maiduguri during the cur showed 25% downy mildew whereas rent rainy season. Studies on combining IVHs showed only 7%. ability characteristics of these lines have been initiated. Between 1988-1994, over 350 IVHs were developed (by crossing male-fertile Inbred x Variety Hybrids inbreds as females using protogyny with 24 improved OPVs as male parents) and In 1985, to circumvent the lack of were evaluated in 30 trials at three loca male-sterile lines, we initiated work on tions. Compared to the OPV s, the IVHs adapted IVH and continued work on were generally early by two days; exhib TCH. Instead of a male-fertile' inbred' as ited a wider range of plant height and Table 3. Comparison of mean values for nine agronomic characters in pearl millet inbred x variety hybrids (IVHs) and open-pollinated varieties, ICRISAT Sadore and INRAN Station Ben ou Ni!!er. IER C"mzana M'ali ramv seasons 1988- 1994 Character Mean value for Observation( s) IVH OPV Time to hloom (d) 59° 61 a IVHs earlv hv two days Range exhibited by IVHs Plant height (cm) 214b 223 a was from 150 to 270 cm vs. OPVs 165 to 250 cm. IVH range form 25-65 Earhead length (cm) 41 b 45 a cm vs. OPVs 20-60 cm. Tiller number planf1 4a 3b 33% more tillers by IVHs Earheads plof1 60a 48b 25% more earheads by IVHs Threshing (%) 65 a 66a No difference Empty earhead plof 21 a liD More empty earheads in (%) IVHs Downy mildew (%) 78% of IVHs showed 7b 9a <10% downy mildew 1 1.30' 1.17b Individual IVH yield ad- Grain yield (t ha- ) vantages recorded up to 202% over correspond- ing pollinator OPV (=100%) Means followed by a different letter in the same row are statistically different at 0.05 probability level. 111 earhead length; produced more tillers and showed grain yield advantages from 22% earheads; and showed around 10% yield to 102% over their respective variety par advantage (Table 3). The range for grain ents. The yield advantages observed here yield of IVHs was from 0.2 to 2.4 t ha- 1 are more than those reported by Bono and 1 and for OPVs from 0.2 to 1.7 t ha- . Dur Leclerq (1963) in Senegal. They reported ing six years of evaluation, except in that inbred x landrace hybrids showed a 1993, IVHs out-yielded OPVs. Grain yield increase of 47% over their parents yield advantage of IVHs was 18% in and hybrids, with male-sterile lines show 1988; 23% in 1989; 13% in 1990; 20% in ing a 59% grain yield advantage. The 1992; -5% in 1993; and 14% in 1994. grain yields of IVHs also represented a More IVHs were tested in 1993 than any 20-25% increase over control varieties. other year, which may have resulted in The better OPVs that gave superior IVHs inclusion of poor entries with lower yield included IBMV 8001, DGP1, ICMV IS potential and may have effected the over 86330, and ICMV IS 90311. A significant all mean. Over the years, individual IVHs positive correlation was observed be- + + 1.5 '7 ..c.. ;t::, '0 Ol ';:" .5 ..... lOll I '" 1.0 ~ ..c =:... i 0.5 0.5 1.0 1.5 2.0 Varieties grain yield t ha-1 Figure 1. Relationship between grain yield of pearl millet Inbred x Variety hybrids and corre sponding pollinator varieties in yield trials conducted in the rainy seasons between 1988-1994, ICRISAT, Niger. 112 tween mean grain yield ofOPVs and their plants and the seed from deliberate cross IVHs (Figure 1, r = 0.75** ). Therefore, ing produced 82% hybrid plants. It ap OPVs with higher yield potential tended pears that in deliberate crossing either to produce IVHs with higher grain yield pollen viability and or stigma receptivity though as yield levels rise heterosis per were not optimal for producing 100% hy centages may be less. In several of our hy brid seed. However, hybrid grain yields of brid trials, a strong negative correlation the two seed lots in either hybrid were not was observed between time to 50% bloom significantly different. These results were and hybrid grain yield, and time to 50% encouraging; they show that the presence bloom and the number of empty earheads, ofnearly 20% inbred parent seed had little especially with early hybrids of late effect on grain yields, probably because OPVs. The empty earheads were a result the smaller inbred plants offered little of scarabid beetle and Dysdercus volkeri competition to the taller vigorous hybrid infestations. It became clear that by se plants. This is in agreement with earlier lecting mid-late to late-maturity hybrids research by Andrews et al. (1993) which (around 100 days) losses caused by these also showed up to 20% female selfing had insect pest infestations can largely be little effect on hybrid performance. These avoided. Using this information, we have results also demonstrated that use of begun testing of hybrids developed with protogyny in pearl millet appears to have landrace material by choosing those best potential for making hybrids. It is possi adapted to the stresses ofthe target envi ble to undertake large-scale multiplica ronment. tion of IVHs (with the pollinator variety flowering at least 5-7 days earlier than the Proportion of IVH Seed inbred parent), thus permitting a wider Produced in Isolation choice of parental combinations and more rapid hybrid development compared to A comparison was made of the fre those dependant on CMS systems. Cur quency of hybrid and inbred plants in the rently, we have eight such IVHs in ad seed lots derived from IVH multiplication vanced tests. These are particularly suit both in isolation and by deliberate cross able for programs in West Africa, where ing to determine the feasibility of the attributes of local landraces and im large-scale seed multiplication of IVHs. proved varieties can be directly utilized as Two IVHs - where the inbred 'seed par pollinators in easily produced adapted hy ent' and the hybrid plants have distinctly brids. different characteristics and can be easily identified - were used. Only the seed from Performance of IVH Hybrids the fertile inbred (used as female) was under Moisture Stress harvested. With the first hybrid (458 x GB 8735), the frequency of hybrid plants was In 1995, we evaluated 50 hybrids and 92% (8% inbred plants) in the lot multi OPVs using a factorial design with two plied in isolation and 78% in the lot de replications. This design included three rived by deliberate crossing. For the sec moisture levels (200, 300, 600 mm of ap ond hybrid (410 x GRPl), the seed lot plied irrigation), two planting densities 1 from isolation produced 78% hybrid (10,000 and 62,000 plants ha- ) and two 113 levels offertility (500 kg FYM ha- I or 10 t ies. Stability analysis indicated (Table 4 of FYM) plus 200 kg single super phos and Figure 2) that IVHs gave stable per I phate ha- . The plot size used was 16 sq m. formance and high grain yield across all Only the results that relate to the perfor the environments (b = 1.10) and were re mance of IVHs (10), their corresponding sponding to better environments in a simi- pollinators (6) and 2 control varieties 1ar manner to pollinator varieties which (HKP and Sadore local) are presented also showed stable performance. here. Top Cross Hybrids Table 4 summarizes the overall grain yield performance of IVHs, correspond So far, (excluding hybrids which are ing pollinators and control varieties. The too early in relation to their OPV male mean grain yield of IVHs across the 12 parent) our results have shown that OPVs I I environments was 1.10 t ha- , 0.98 t ha- with higher yield potential tended to pro for the pollinator varieties, and 0.70 t ha- I duce IVHs with higher grain yield. Using for the two control varieties. Across all this information, we have begun testing of the 12 environments, the IVHs yielded up hybrids developed with landrace material to 28% more grain yield than the control which are adapted to the stresses of the HKP and between -6% and 25% more target environment. than the corresponding pollinator variet- Table 4. Pearl millet Top cross hybrids, pollinators and controls mean grain yields across 12 test enVIronmen t s DrOU!!l htN ursery. ICRISAT- N'l!!er Sa d'Sore ummer 1995 Grain yield Change (%) over 'b' 1 Entrv Desilffiation (kQ ha· ) Dollinator vield stabiliti s2d¥ Hybrids 10 1347-52 x CIVT 1062 108 1.08 -404 20 1591 x PV SSZ 926 975 94 0.88 -322 9 1317-28 x P3KOLLO 951 III 0.93 -389 6 1411-18 x SOSAT-C88 1136 115 1.09 209 7 1317-28 x SOSAT-C88 1186 119 1.06 -109 11 1317-28 x GB 8735 1165 116 1.16 323 12 1353 x GB 8735 1168 116 0.90 492 13 1419 x GB 8735 1255 125 1.19 372 16 1529-38 x PV SSZ 925 1118 109 0.92 792 17 1583-86 x PV SSZ 925 1069 104 1.05 -187 Mean Hybrids 1109 1.10 28 Pollinators 2 SOSAT-C88 996 100 1.12 -316 3 CIVT 987 100 1.01 -164 5 GB 8735 1005 100 1.08 -258 8 P3KOLLO 859 100 0.91 -60 14 PV SSZ 925 1027 100 1.12 -256 19 PV SSZ 926 1041 100 1.08 207 986 1.12 -81 ControlsM~.p.rn·"i I Sadore local 536 - 0.68 -313 4 HKP 863 0.76 -246 Control Meal 699 - 0.78 45 Overall mean 1022 - - cv(%) 35 - LSD 10.05) 202 -- ¥ Mean squares equal 102 times the column values. 114 4~ ~------ 3000 "';' -=01 JJ ~ ~ c 2000 .OJ... toll C ...01 :;: 1000 Controls baO.7' R.O.ts Hyorlds b.1.10 R.O.t. Parents 0.. 1.12 R.O." o ·1000 o 1000 Environmental index Figure 2. Regression between mean grain yield and environmental index of pearl millet Inbred x variety hybrids, corresponding parent pollinator varieties and controls, Drought screen nursery, ICRISAT, Sadore, Niger, Summer, 1995. Time to maturity in relation to the dura Landrace varieties developed over cen tion of the growing season at that locality turies of creative and conscious human is often critical to maximizing other adap and environmental selection have accu tive traits. These locally developed vari mulated high levels of either genetic re ety characteristics confer good stability of sistance or tolerance to both the prevail production. However, these landraces are ing physical stress conditions (i.e., locally specialized and adapted only to drought, low nutrient supply, unfavorable their particular ecogeographic zone. The soil pH, and high soil surface tempera use oflandraces as hybrid parents assures tures at emergence) and to many biologi good adaptation, and the resulting TCH cal stresses (i.e., pests, diseases, weeds). (or IVH) hybrids will elicit responsive- 115 ness ofthese populations to improved en flowering only three days earlier than the vironmental conditions. Landrace popu landrace. The TCH derived from the con lations do not possess sufficient variation trol showed the same mean time to flow for yield potential or disease resistance to ering (61 days) as the control variety. For permit appreciably better varieties to be downy mildew, of the 20 TCHs derived made from them by direct selection. from the landraces, 13 had less DM than Though several varieties have been de the parents, three showed more DM inci rived from landraces and recommended dence than the parental variety, and four for cultivation, they do not show signifi had incidences equal to the parent. Over cantly higher grain yield over the original all, male-steriles 712 A and 680 A pro population. Nevertheless, they do possess duced TCHs that recorded less downy many farmer- preferred traits, such as mildew. Those TCHs that were derived moderate earliness and a reduced fre from the male-sterile 726 A were most quency of shibras (re-occurring wild type susceptible to downy mildew. On an aver plants). With the availability of age, there was nearly 50% more tillering male-sterile lines there is now an opportu in the TCH compared to the parents; the nity to combine the adaptation ofthe land highest increase in tillering was in TCHs races to the prevailing low moisture, with Bazagome (87%), and lowest in low-soil fertility conditions with hybrid crease in tillering was in TCHs with vigor and - as a result - improve yield po Boudouma (25%). Nevertheless, it has to tential. TCH (or IVH) hybrids appear to be noted that Boudouma itself is a be the easiest route for a rapid and effec high-tillering landrace. tive incorporation of tolerances to stress factors with increments in yield potential Mean grain yield oflandraces was 0.67 and better use of both limited resources or t ha- I and ofTCH derived from them was improved management (Bidinger et al. 1.16 t ha-I, an average increase of 73 % 1994). over the parental pollinators. Increases over landrace pollinators ranged from We evaluated 20 TCH derived from 53% (TCH ofBoudouma) to 113% (TCH crosses between four male-sterile lines with Guerguera). The control CIVT mean I and five landraces from Niger (and eight grain yield was 1.02 t ha- , with mean of improved varieties). In addition, crosses its TCHs ranging from 1.15 t ha- I to 1.35 t I I were also made with these four male-ster ha- (mean of 1.32 t ha- ) with a mean ile lines and eight improved varieties. yield increase of 30%. The highest yield These along with pollinator parents (land ing TCH (726 A x Dan Tehama) gave races themselves) and TCHs derived from 35% more grain yield than the control va a control variety (CIVT) and the control riety CIVT (Table 5). A stability analysis variety itself were tested over two years at indicated (Figure 3) that the TCH made four locations (Table 5). with landraces gave better grain yields in all the four environments and were more The TCHs derived from the landraces stable than the landraces themselves. Bazagome, Guerguera, and Dan Tehama, were 7 to 8 days earlier to flower, while the TCHs derived from the landrace Moro 116 Table 5. Mean values for four characters of 20 Top cross hybrids, their pollen parents and con- trol from tests conducted in four environments, ICRISAT-Niger, rainy seasons, 1996 and 1997. Time to 50% DM Tillering Grain yield l En!!y Desig!)ation bloom (d) (% s9r) (#) (tha· ) Hybrids from landraces 7 680 x Bazagome 62 1.03 8 698 x Bazagome 66 2 1.12 9 712 x Bazagome 66 2 1.16 726 x Bazagome 60 4 4 1.21 11 680 x Guerguera 62 2 4 1.03 12 698 x Guerguera 67 2 3 1.08 13 742 x Guerguera 67 2 2 1.01 14 726 x Guergyera 59 3 4 1.24 15 680 x Mora 58 1.05 16 698 x Moro 62 4 1.33 17 712 x Moro 63 0 1.21 18 726 x Moro 57 3 6 1.35 19 680 x Dan Tehama 59 6 1.31 20 698 x Dan Tehama 63 2 U4 21 712 x Dan Tehama 65 1.32 22 726 x Dan Tehama 58 1.35 23 680 x Boudouma 56 5 0.73 24 698 x Boudouma 5' 4 U9 25 712 x Boudouma 60 2 4 U5 26 726 x Boudouma 55 7 1.08 Landraces Bazagome 72 2 0.55 2 Guerguera 73 2 0.51 3 Moro 63 3 0.79 4 Dan Tehama 68 2 0.82 Boudouma 58 3 4 0.68 Hybrids from control variety 27 680 x CIVT 59 4 0.55 28 698 x CIVT 63 0.45 29 712 x CIVT 65 I 0.50 30 726 x CIVT 58 2 0.30 Control variety 6 CIVT 60 2 1.02 Overall mean 62 2 4 UO CV(%) 4 50 25 26 LSD (0.05) 2 0.20 Hybrid type Hybrids from Landraee 61bl 2b 4a U6a Hybrids from VarielY 61b 2b 4a 1.33b Pollinator parents Landraees 67a 3a 3b 0.67e VarielY 60b 2b 3b 1.02d Male steriles, mean effects 680A 59b 2b 5a 1.05b 698A 63a 2b 3b 1.20a 712A 64a Ie 3b 1.23a 726A 58e 3a 5a 1.26a I Means followed by a different letter are significantly different at 0.05% probability. 117 2.00 -.------, 1.80 "7 ..c-= 1.20 ...:ell :S! .~ = .;... ell 0.80 =-= ~" ~ Variety hybrids b. 1.3 R2=97 0.40 -+- L.andracas b,. 1.2 R2=95 -+- L.andrace hybrids b= 1.0 R2=99 -e-- Control b= 0.5 R=45 0.00 -+---..... --...,---..----,---.-----,.---.----1 -0.40 -0.20 0.00 0.20 0.40 Environmental index Figure 3. Regression between mean grain yield of pearl millet Top cross hybrids, corresponding pollinator parents, control variety (CIVT), and environmental index, ICRISAT, Sadore, Niger, 1996-97. Top Cross Hybrid Tchama), ICMH 9802 (712 A x CIVT), Seed Multiplication ICMH 9803 (698 A x ICMV IS 89305), and ICMH 9804 (712 A x ICMV IS As mentioned previously, the evalua 90309). We have gathered data on three tion ofTCHs was undertaken along with a hybrid seed multiplication plots. The plot set of TCHs produced by crossing size used was 492m2 (or approximately male-steriles x improved OPVs. From 0.05 ha), with 20 ridges 40m long and this set, we selected four hybrids for seed 0.75cm apart. Within ridge spacing was multiplication in isolation. To our knowl 0.80m ( 41 hills). Except for the first three edge, this is the first time that multiplica and last three rows the pollinator parent tion of CMS pearl millet hybrids was un was planted in two rows (pollinator parent dertaken in West Africa. These four hy was planted in the first three and the last brids were ICMH 9801 (680 A x Dan 118 Table 6. Summary of observations recorded in two pearl millet Top cross hybrid seed production plots, ICRISAT-Sadore, Niger, Jan. - May, 1998. Trait measured Male-sterile line in TCH plot of ICMH 9802 ICMH 9804 Male-sterile 712Al 712Al Pollinator variety CIVT ICMV IS 90309 Male-sterile Number of hills planted 240 240 Number of hills emerged 236 239 Number of non-flowering plants 10 Number of effective plants 216 224 Plant height (cm) 113±11.2 117±9.5 Mean tiller number (per plant) 3.8 ± 0.9 3.8± 1.16 Mean earhead length (cm) 45 ± 5.2 44 ± 4.0 Mean earhead circumference (cm) 7.3 ± 0.6 7.6 ± 0.59 Mean head exsertion (cm) 0.9 ± 3.9 1.6±3.13 Fertile plant frequency (%) 4.4 (10) 5.0 (12) Pollinator Variety Number of hills planted 540 520 Number of hills germinated 524 515 Number of non-flowering plants 29 4 Number of effective plants 485 511 Plant height (cm) 184 ± 11.5 180 ± 16.7 Mean tiller number (per plant) 3.8 ± 1.47 5.5 ± 1.70 Mean earhead length (cm) 56 ± 5.3 47 ± 6.0 Mean earhead circumference 7.6 ± 0.67 8.1 ±0.81 Mean head exsertion -8.8 ± 4.7 -4.0 ± 6.8 Shibra frequency (%) 2.03 0 Ratio Female:male Plants 1.0:2.2 1.0:2.3 30 26.8 25 • Pollinator --~ f3 Male-sterile '-' .....en 20 fa -...... 0.- 0 15 :>-. u 1:1 (l) ::s 10 0" (l) I-< ~ 0.4 o 3 4 5 6 Tiller number per plant Figure 4. Frequency distribution of tiller number plant-l of male-sterile 712 A and pollinator parent CIVT in a hybrid seed production plot, ICRISAT, Sadore, Niger, Sum mer 1998. 119 three rows) and alternated with two rows quency) or the effect of high tempera of male-sterile line. To ensure that ample tures. In April, the ambient maximum pollen is available, the first four hills and temperatures recorded were very high; the last five hills of every row were the mean maximum temperature was planted to the pollinator. 42.7C. In comparison, the mean maxi mum temperature recorded in the last 15 A summary of observations recorded years at Sadore (1983-1997) was 4l.4 C. on two of the multiplication plots is given There were 24 days that recorded a maxi in Table 6. In all three hybrid multiplica mum temperature of 43C in April 1998 tion plots, the nick (synchrony in flower vs. a mean of 5 days with 43C recorded ing between the male-sterile and for the month of April over the last 15 pollinator parents) was optimum, with years. These high temperatures may have pollinator parents beginning flowering influenced the observed break-down of 7-10 days before the male-sterile line. In male-sterility. all cases the pollinator parents flowered in 50-55 days. Tillering was asynchronous Based on the seed yields obtained in (with a mean of3.7 tillers and 3.8-5.5 till these isolation plots estimates of hybrid ers per plant in the female line, and and pollinator yields for different fe pollinators ICMV IS 90309 and CIVT, re male:male row ratios were made (Table spectively) both in the male parent and the 7). A female:male ratio of 1:2 was used to male-sterile line with a 10-20 day interval ensure success of the seed production between stigma emergence on the main plots. At this ratio a hybrid seed yield of tiller and secondary tillers. The tiller 0.31 t ha- I and pollinator yield of l.1 t ha- I number planrl (Figure 4) assured that was recorded, for a total yield of 1.40 t I enough pollen was available for all the fe ha- . At a female:male ratio of 4:2 (ap male earheads. The frequency of fertile pears optimum for larger plots) the esti plants (5%) in the female rows was high; mated hybrid seed yield is 620 kg ha- I and I these fertile plants were removed soon af of the pollinator 700 kg ha- . Our ter anthesis. We are unable to determine at . back-of-the-envelope' calculations this time if this occurrence offertile plants show that profits between $ 650 and $ 850 was due to a mechanical mixture (subse can be expected from a hectare of hybrid quent grow-out from the same seed lot seed production (assuming 200 Fcfa for a will be used to confirm this observed fre- kg of grain, 800 Fcfa for a kg of hybrid Table 7. Estimates of seed yields of pearl millet hybrid and pollinator parents in TCH seed pro duction fields 1 Ratio Ridges Plantsiha Estimated seed yield (t ha" Total yield % of seed yield I Male Female Male Female Male Female Hybrid Pollinator (tha- ) H brid Male 2 I 88 44 11 100 4400 0.31 1.10 lAO 22 78 2 2 66 66 9900 6600 0047 0.90 1.36 34 66 2 4 44 88 7700 8800 0.62 0.70 1.32 47 53 2 8 26 106 5900 10600 0.75 0.54 1.28 58 42 I Estimated based on one ha as 132 ridges with a spacing of 0.75 m, and within ridge spacing ofO.8m, The ridge length is 100 m (or 125 hills with single plants per ridge of 100 m). There will be 125 plants in all male rows, and in female rows 25 hills at either end will be planted to pollinator parent (13 + 12). Hence all female rows will only have a plant population of 100 per ridge. Plant population per ha is assumed at 16500 hal. 120 seed and 400 Fcfa for the pollinator OPV tinue pollen shed throughout flowering seed). It has to be noted that production of period of the female parent to ensure pu TCHs permits harvest of two seed crops rity and good seed set in multiplication that of the hybrid and of the pollinator va plots. riety - from one field. Seed producers can sell the seed of pollinator variety which Grain Quality Criteria will generally have better seed quality. This assures an increase in the value of Grain quality criteria are determined produce from the male parent rows. by end uses. Pearl millet in West Africa is grown mainly for food, and only occa Criteria For Parental Lines sionally used for brewing. The grain qual ity criteria for pearl millet are less Based on our experience thus far, the well-defined than for sorghum (Andrews following are the criteria for the develop and Anand Kumar 1992). Usually gray to ment of parental lines for successful hy yellowish white plump grain is widely ac brid seed production and good hybrids cepted, though in parts of WCA light (Andrews et al. 1997). Efforts are under brown or creamy white grains are pre way to develop complementary informa ferred. There is currently no incentive in tion through phenology studies of elite fe either food or feed markets to demand male and male parents. higher nutritional quality. Thus, at this time, there is no motivation to put much Female Parents. Female parents emphasis on selection based on nutri should possess a high degree of resistance tional quality in breeding millet hybrids. to downy mildew, stable and perfect male-sterility, high yield, and moderate to Future Outlook good synchronous tillering. It is impor tant to have good earhead exsertion, good Requirements in TCHs and IVHs. seed set, bold seed size and seedling vigor Both IVHs and TCHs are practical in in addition traits that have good correla pearl millet and have real advantages for tion with hybrid performance, such as use in Africa in terms of durability of dis plant stature, maturity, and disease and ease resistance and adaptation, which are pest resistance. difficult to obtain in SCHs. Using IVHs, ergot susceptibility is reduced, as CMS is Male Parents. In addition to excellent not involved. For TCHs that are based on combining ability with the seed parent, landraces an adequate supply of breeders male parents should completely restore seed must be available to ensure repeat male-fertility and should have profuse able performance of the hybrid. Hybrids and early pollen shed. High tillering, to should be tailored closely to fit the exist spread pollen shed over time, is desirable. ing growing season lengths; both to maxi The parents should have a good and early mize environmental resources available "nick". The "nick" should be early for growth and to minimize pest and dis enough for the male parent to start shed ease infestations/incidence. Hybrids ding pollen before the stigmas of the seed should have similar crop establishment parent become receptive and should con- features as the local landraces. Cost of 121 seed should not be very high. Unlike sor Regional Pearl Millet Hybrid Trial. ghum, which is grown under more as Encouraged by the results obtained thus sured rainfall situations, pearl millet far (with IVH, TCHs, and availability of growing environments often necessitate adapted male-sterile lines) participants of replanting (up to 3-4 times). There will be the project P 7 (Farmer participatory vari a need to investigate seed production in ety development and seed multiplication) selected areas during the rainy season in of the West and Central African Millet order to reduce the costs of summer sea Research Network (ROCAFREMI) have son production (e.g., irrigation, bird at decided that a collaborative ICRISAT tacks, etc). Other factors that would need ROCAFREMI Regional Pearl Millet Hy attention include the following: 1) pro brid Trial be initiated during the 1998 duction of hybrid seed in environments rainy season. ICRISAT and where a successful crop is relatively as ROCAFREMI have put together such a sured, 2) development of productive seed trial with 10 TCHs (9 TCHs with im parents, and 3) assessment of hybrids not proved OPVs, and one with a landrace), only for grain yield, but also for stover eight corresponding pollinator parents yield. There is a need to evaluate the opti and two controls to be conducted at 11 lo mum isolation distances required, and the cations in seven countries. Some of these effect of pollen shedders in seed parents TCHs were also selected for on-farm in hybrid production plots. demonstrations in Niger in collaboration with INRAN. It is hoped that by the end of Lessons from Experience in India. In this year information will be available on India, adoption of pearl millet hybrids general adaptation and grain yield poten (now estimated to cover 45% of the 10.5 tial of these top cross hybrids. Obviously, m ha crop area) resulted in dramatic in positive performance will trigger a range creases both in production and productiv of developmental activities, including hy ity. It has stimulated the development of a brid seed production training courses. large private sector (estimated to produce up to 55% of the total seed made) and Conclusion availability of quality seed to the farmers. The adverse aspect relates to the suscepti Yield potential is important at all levels bility of single cross hybrids to downy of production, but perhaps more so in mildew epidemics accentuated by repeat stress conditions where limited resources edly growing a single genetically uniform must be utilized more efficiently. Yield hybrid over large planted areas. It is now improvement in pearl millet is at a rela clear that the risk of downy mildew sus tively early stage in West Africa. Yield in ceptibility can be greatly reduced by se creases in the range of20-25%, through a lecting for durable resistance and adopt change from varieties to hybrids, has yet ing appropriate breeding strategies, such to be realized/utilized at the farmers' level as the development of TCHs and IVHs, and is likely to take at least 4-7 years. Hy which exploit the stability and durability brids have a higher yield potential, even characteristics of genetically variable under the conditions of moderate produc adapted male parent varieties. tivity, and use resources more effectively. We have demonstrated through trials of 122 adapted hybrids that this yield increase is of the International Conference on Genetic Improvement of Sorghum and Pearl Millet, available for farmers to use in WestAf Lubbock, Texas, USA, 22-27 September 1996. nca. Lincoln, Nebraska, USA. Collaborative Re search support Program on Sorghum and Pearl Millet (INTSORMIL Publication No. 97-5). It is important to recognize that the va Bidinger, F.R., E. Weitzien, R.V. Mahalakshmi, riety or hybrid component is only one fac S.D. Singh, and K.P. Rao. 1994. Evaluation of tor (though probably the catalytic factor) landrace top cross hybrids of pearl millet for arid zone environments. Euphytica 76: in the total agronomic, institutional, and 215-226. social package needed to sustain im Bono, M., and P. Lec1erq. 1963. Methodes proved production. In all pearl millet d'amelioration varietale des mils et sorghos utilises au CRA (Centre de Recherche growing regions of the Sahel, a general Agronomique), Bambey. Agronomie Tropicale improvement in plant nutrient supply is 18: 33-52. critical to complement genetic improve Hanna, W.W.1989. Characteristics and stability of a new nuclear-cytoplasmic male-sterile source ment. Such improvements in soil fertility in pearl millet. Crop Science 29:1457-1459. are vital to preserving and enhancing the ICRISAT(lnternational Crops Research Institute crop production resource base for the fu for the Semi-Arid Tropics).1987. Proceedings ofthe International Pearl Millet workshop, 7-11 ture. Problems of infrastructure related to April 1986, ICRISAT Center, India. seed production, assured food grain mar Patancheru, A.P. 502324, India:ICRISAT. Lambert, C. 1982. IRAT et l'amelioration de mil. kets, and diversification of end uses are Agronomie Tropicale 38:78-88. among the many areas that need attention Ndoye, A.T. and S.C. Gupta. 1987. Research on to insure profitable use of hybrids - or pearl millet hybrids in Senegal. Page 285 In ICRISAT(lnternational Crops Research Insti even OPVs - by smaller farmers. tute for the Semi-Arid Tropics). 1987. Proceed ings ofthe International Pearl Millet workshop, References 7 -11 April 1986, ICRISAT Center, India. Patancheru, A.P. 502324, India:ICRISAT. Andrews, D.l, B. Kiula and J.F. Rajewski. 1993. Nwasike, C.C. 1987. Pearl millet male-steriles and The use of protogyny to make hybrids in pearl hybrids in Nigeria, Page 284 In ICRISAT (In millet. P. 208-210 in: J. Janick and J. Simon ternational Crops Research Institute for the (eds.) New Crops. John Wiley and sons. New Semi-Arid Tropics). 1987. Proceedings of the York. International Pearl Millet workshop, 7-11 April Andrews, D.J. and K.A. Kumar. 1992. Pearl millet 1986, ICRISAT Center, India. Patancheru, A.P. for food, feed and forage. Advances in Agron 502324,India:ICRISAT. omy 48: 89-139. Ouendeba, B., W.W. Hanna, G. Ejeta, W.W. Andrews, D.J., G. Ejeta., M. Gilbert, P. Goswami, Nyquist, and lB. Santini. 1996. Forage yield K. Anand Kumar, A.B. Maunder, K. Porter, and digestibility of African pearl millet land K.N. Rai, J.F. Rajewski, V.S. Belum Reddy, W. races in diallel with missing cross. Crop Science Stegmeier, and B.S. Talukdar. 1997. Breeding 36: 1517-1520. hybrid parents. Pages 173-187. In Proceedings 123 New Cytoplasmic Male Sterility Systems for Hybrids in Pearl Millet D. J. Andrews, K. N. Rai, and J. F. Rajewski Generally, alternative cytoplas the minimum criteria, and since they were mic-nuclearmale sterility (CMS) systems the first available systems to allow com are thought of as "useful" because they mercial access to the tremendous advan reduce "vulnerability." However they are tages of hybrids, they became widely much more valuable for other reasons. used. Only now, when other CMS sys Our research in INTSORMIL and tems have been discovered and compared ICRISAT over the last few years has be to the Al system, can we characterize the gun to describe these and, though the re deficiencies and constraints of the Al sys search has been on CMS systems in pearl tem, which could not previously be seen millet, similar situations are to be ex clearly. pected in any crop that depends on CMS for commercial hybrid seed production. Many of the constraints to the Al sys tems in sorghum or pearl millet can be il CMS systems can differ in a number of lustrated by examining the F 2 of any com aspects and these differences can have mon hybrid. The large class of partial profound effects on breeding success, steriles (or weak fertiles) that occur are in seed production and hybrid performance. dicative of the difficulties in breeding For a character to be most easily used in good seed and pollen parents with these breeding it should be clearly identifiable, systems. In pearl millet, we have com simply inherited, and controlled by major pared the ~ (Hanna, 1989) with the Al genes,its expression being unaffected by (Burton, 1958) system. More recently, we other genes or the environment. The Al have been examining the A5 system (Rai, CMS system in pearl millet and the Al 1995). The important differences are (kafir-milo) system in sorghum did meet summarized in Table 1. (The A2 and A3 Table 1. Attribute comparisons in AI, A4 and AS cytoplasmic-nuclear male sterile systems in pearl millet AJ A4 As Germplasm Access V. Limited Wide V.Wide % B lines in germplasm ±10 ±85 99 % Perfect B lines from B x B crosses ±50 100 100 % R lines in germplasm <15 ±IO% ±l% <5' % Germplasm giving partial male sterility ±75 <5 , G x E effects on restoration Present Low V.Low Appearance of sterility Good Good V. Good. R Line breeding in sterile cytoplasm V. Difficult Easy Easy, Cytoplasmic reversions to male fertility Yes No No 'Initial results 124 systems (Burton and Athwal, 1967) were • The expression ofmale sterility and never sufficiently male sterile to be of fertility in the Al system is subject commercial use). The most significant to environmental influence. In high differences include the following: temperatures, for example, Al seed parents can break sterility. In cooler • Very few partial fertiles occur with conditions, and Al hybrids can be A4 and As This makes breeding come significantly less male fertile many times easier, especially in giving poor seed set and reduction terms of developing A-lines and in yield. generating new B-lines. • Finally, Al sterile cytoplasm is • High frequency of B lines in slightly unstable and mutates to fer germplasm. Producing new seed tility at a low rate compared to fer parents, which is key to advancing tile (normal) cytoplasm. This phe hybrid yields, was a major con nomenon has so far not been noted straint with the Al eMS system. in A4 or As. Almost all cultivated germplasms are B lines in the in the A4 and As The sum of these differences can result systems. in a huge effect for breeders. With the Al • Though the natural frequency of re system, the primary breeding effort is to storers (R4 in the A4 system, Rs in develop new male sterile lines and re the As system) in cultivated storer lines, and then (within those limits) germplasm is low and very low re look for heterotic combinations. With A4 spectively, the conversion of any and A5, however, the primary effort can line to an R4 or Rs pollen parent is a be to look first for heterotic combinations, straightforward dominant back and only then convert the superior crossing procedure when utilizing combiners to male sterile (A lines) or sterile cytoplasm. The use of sterile male (pollen, restorer) R lines. cytoplasm in breeding R4 or Rs re storers is possible in these systems, The A4 and A5 eMS systems, thus, pro because-in contrast to the Al sys vide access to a much wider range of ge tem-there is little interaction with other nuclear or cytoplasmic genes, netic diversity for hybrid breeding and which cause variable partial steril present a rapid, certain method of devel ity expression. oping both male and female hybrid par When the concept ofheterotic pools ents, which substantially improves the ef is considered, the amount of diver ficiency of breeding. A similar situation sity available for seed parents in A4 may exist in sorghum: where the Al sys and As systems is much larger than tem has similar constraints and many for AI. The diversity available for eMS systems are known to exist R4 or Rs pollen parents is poten (Schertz, 1998) but have not been exten tially unlimited, as any line can be sively explored. rapidly converted to R4 or Rs. By the same token, any Al or RI line One aspect that needs further investi (or existing Al hybrid) can be con gation is whether the different cytoplasms verted to A4 or As lines, or hybrids. cause differences in hybrid performance. 125 While only Al and At have been com sterility or poor pollen shed in the Al sys pared, initial results suggest that there tem. Farmers will benefit in several other may be some nuclear cytoplasmic interac ways from the use of A4 or As hybrids. tions. In some cases, the same nuclear hy First, hybrids will have less off-types in brid was better when made in Al cyto them. Second, there will be a high level of plasm, while in other combinations the At male fertility that will be unaffected by hybrid was better. It also seems possible cool conditions. Third, because hybrid that nuclear cytoplasmic interaction can breeding becomes quicker with At or As contribute to heterosis, since eMS hy yet draws on more diversity, superior new brids can be slightly better than the same hybrids will reach farmers more quickly. hybrid made without sterile cytoplasm (possible in pearl millet and maize). Re Summary search is continuing to confirm these re sults. One consistent difference was that New cytoplasmic-nuclear male steril the protogyny period in At was shorter ity systems A4 and As in pearl mil than in AI, but stigma longevity is longer let-compared to the original Al sys in At which leads to better seed set under tem-offer advantages to breeders, seed stressful conditions. producers, and farmers. These advan tages are based on clearer expression of Seed producers will find At and As hy male sterility or male fertility, with less brids easier to produce. In A4 and As hy environmental effects, greater access to brid seed production, most off-types that genetic diversity, and the rapid conver result from outcrosses in seed parent in sion of any line to ~ or Rs restorer lines. creases will be male sterile, because most Additionally, A4 and As sterile sources of stray pollen are maintainers. If cytoplasms appear not to spontaneously ~ and Rs germplasm becomes more revert to the normal (fertile) condition, as common, this advantage will decline. in AI. The At and As eMS systems now allow breeders to concentrate first and No reversion of A line plants to B line foremost on hybrid breeding (finding the plants, which occurs in the Al eMS sys best heterotic combinations), then on con tem, will also mean cleaner seed parents verting the parents using eMS sterile and and, hence, purer hybrids. Such rever restorer lines, leading to a large increase sions are very difficult to find, because in efficiency in hybrid breeding. (except for pollen shed) they are identical to male sterile plants. These two charac References teristics will greatly reduce the need for Burton, G.W. 1958. Cytoplasmic male sterility in roguing seed production fields, reducing pearl millet (Pennisetum glaucum (L.) R. Br. production costs and increasing hybrid Burton, G.W., and D.S. Athwal. 1967. Two addi purity. tional sources of cytoplasmic male-sterility in pearl millet and their relationship to Tift 23A. Crop Sci. 7:209-211 Seed producers will not have to worry Hanna, W.W. 1989. Characteristics and stability of about unusually hot or cool weather when a new cytoplasmic-nuclear source in pearl mil let. Crop Sci. 29:1457-1459. producing A4 or As hybrids. Such condi tions can either cause breakdown of male 126 Rai, K.N. 1995. A new cytoplasmic-nuclear male Conference on Genetic Improvement of Sor sterility system in pearl millet. Plant Breeding ghum and Pearl Millet. INTSORMIL, Univ. of 114:448-450. Nebraska-Lincoln. Shertz, K.F., S. Sivaramakrishnan, W.W. Hanna, J. Mullet, Yi Sun, U.R. Murty, D.R. Pring, K.N. Rai, and B.V.S. Reddy. 1997. Alternate cytoplasms and apomixis of sorghum and pearl millet. pp. 213- 223 In Proc. ofIntemational 127 Creation of a Seed Industry: Role of the Private Industry A. Bruce Maunder Introduction specialize in basic research and applied research on subsistence crops while the Seed, currently a 55 billion dollar mar private sector increasingly focuses on re ket, worldwide, has evolved since 13,000 search, production, and marketing of im BC through several phases. These proved seed having a high multiplication changes, while well documented in devel factor and a relatively low seeding rate. oped countries, provide likely projections NGO's conversely, concentrate on multi for the future role ofprivate industry in in plication and distribution of seed not tar ternational agricultural development: geted by the private sector. Hybrids in es sence widened the agricultural evolution • In phase I, farmers save their own ary gap between developed and seed or obtain it from nearby farm developing countries. ers with the low rate of varietal de velopment limited to farmer selec The private sector became a major tion within landraces. Adoption of player in seed in Europe after the Second new types is low. World War. Agroceres, a Brazilian na • During phase 2, varieties devel tional company began in 1947; the private oped by the public sector begin, al sector became active in Argentina and though slowly, to replace tradi Mexico in the 1950's and 1960's; and tional seeds and inputs such as fer Thailand with Suwan-l in the 1970's. Of tilizer, while limited, are increasing, with the likelihood ofan ten the private sector got its start with emerging indigenous private sector public varieties or hybrids and depended involved in multiplication and dis on a consistent supply of high quality seed tribution of public varieties. to develop a market. Also, the first hy • With phase 3, the private sector be brids, 4-way doublecrosses, as with the comes active with research and de U.S., were better adapted to low input ag velopment, particularly in develop riculture. These were followed by 3-way ing hybrids as well as seeds for spe and then single crosses which required cialized crops such as vegetables. adequate inputs for maximum yield ex Commercial seed production and pression. Morphologically they were of marketing are common, effective shorter plant stature, with better seed laws are in place, and the use standability, and often responded to in of improved seed is high. creasingly higher plant populations. Whereas phase 3 seems essential for a Parastatal seed systems, while supply green revolution in maize comparable to ing only 10% of total seed annually in Af that in rice or wheat, public organizations rica occupy a position late in phase 2 or will and must continue to play an impor early in phase 3. The public sector may tant role in training, basic research and 128 technology transfer. Crop improvement quate quantities of high quality seed and for specific markets, including subsis to provide sufficient fertilizer. Current tence farming, and germplasm develop world agricultural statistics alarmingly ment for smaller, indigenous seed firms tell us that grain consumption per person also requires a strong public component. has dropped 7 percent since 1984; grain The current U.S. evolution from public to harvested area per person dropped 48% private seed improvement suggests an ad since 1950; irrigated area since 1978 by vanced stage of phase 3: 6%; and of perhaps most concern, annual yield gains during the 1990's only in u.s private sector in plant breeding - creased by 1.1 % compared to 2.1 % from % of breeders and number 1950 to 1990. With little more than of companies. one-third of the world's crop area in im Breeders Companies proved planting seed, an obvious need of (No.) Crop (%) developing country agriculture is tech Com 91 94 nology transfer. To encourage develop Soybeans 64 38 ment of the private seed industry, govern Sorghum 74 19 ments should first survey the state of agri Sunflowers 89 14 culture, by crop and by socio-economic Wheat 42 27 region, to determine which crops and ar Cotton 77 35 eas oftheir country can benefit from a pri vate seed industry. Current figures suggest an annual gain of 32 breeders per year for industry and a Success ofthe private seed industry de loss of 2.5 for the public sector with the pends on its ability to provide: (1) a de private plant breeding effort approaching pendable supply, (2) acceptable quality 2/3' s of the total monetary input. Many and purity, and (3) hopefully, but not nec developing countries, in comparison to essarily, an improved level of perfor the U.S. seed industry evolution, would mance. Any of these advantages would range from 50 to more than 75 years be provide for profit to farmers, allowing a hind. higher price for good quality seed. In brief, commercial seeds are best suited to Contributions of Private Sector profitable crops in favorable farming re gions. The private sector companies must A commonly recognized logo is "first concentrate on activities they do best, the seed" yet we've seen the private sector considering funding limits. Their infra as "last to enter." Krull et. al. indicate structures support applied breeding of seed to be the single-most important cata lines and hybrids with increasing lyst for achieving significant increases in introgression of biotechnology, followed agricultural productivity. Sanders et. al. by extensive testing over wide environ working in Sudan found from farmer in mental/geographical areas. Private Indus terviews that the principal constraint to a try can best accomplish these activities more rapid introduction of the first sor by: ghum hybrid, Hageen-Dura-l, was the in ability of input suppliers to produce ade- • being more efficient and flexible; 129 • better understanding market re perienced seedsmen who may be retired quirements; but willing to impart their expertise in an • interacting with new agronomic advisory capacity. Not to be un practices; der-emphasized, marketing and distribu • concentrating on a holistic ap tion playa vital role in determining the proach; s~cce~s or failure of any hybrid seed orga • working large numbers in selection nIZatIOn, whether public or private. and testing; Finally, the absence or ineffectiveness of • being able to assume a level ofrisk the seed industry can pose a major prob • applying seed cost economics to lem in the spread of new crop cultivars, potential pedigrees; • emphasizing short or medium term (Pray and Ramaswami). breeding programs; and, • reducing crop vulnerability. Business Considerations of Private Sector The government ofIndia, for example, The obvious business requirement be has recognized the capabilities of the pri fore investing in a private sector seed ven vate sector to breed, produce, and market ture, whether by a national or interna proprietary and publicly bred materials tional effort, relates to an expected return more efficiently. With control of 60% of on investment. This concern will be af the coarse grain market now, the Indian fected, however, by a host of other criteria private sector unanimously suggests such as strength and stability of the econ things will be better business-wise. Note omy; freedom from disincentives result that the 1998 recipient of the World Food ing in a level playing field; seed price to Prize, Mr. B.R. Barwale and his Mahyco grain ratio; intellectual property rights; Seed Company contributed to the Green need for product; its infrastructure, partic R~volution in India by establishing mech ularly transportation, adequate for deliv amsms of production and distribution of ery of goods and services to the farming quality seeds of major cereals and vegeta community; and certainly political stabil bles plus generated job opportunities in ity. Since private companies are seldom seed production for women belonging to subsidized or given incentives, their en the socially and economically underprivi trance into a seed venture must be con leged sections of the society. cerned with generating a return on share holder equity. They therefore choose an Multi-nationals characteristically de environment where demand is likely to be pend .o~ i~-country personnel preferring strong. Public seed organizations to the to mInImIZe ex-patriots in overseas as contrary are often motivated by signments. This is not to minimize the non-economic considerations such as considerable interaction between the base with a mandate to serve all farmers. Also, location and any other participating coun vertical integration is commonplace in the try. Certainly the private sector offers a private sector but uncommon in the public logical place for short term training of sector. those desiring to establish a seed busi ness. Another opportunity rests with ex- 130 Improved technologies, readily avail In most developing countries in which able in developed countries but which hybrid seed has been widely adopted by have not reached the farm level in the de small-scale farmers, seed-to-grain price veloping world, will be needed for the full ratios are usually less than 10: 1. Later expression of the genetic potential. Nev during the maturity phase of the seed in ertheless, ample evidence exists for the dustry, according to Heisey et. al., superiority of high-tech seeds over con seed-to-grain price ratios rise sharply, of ventional landraces even under less than ten stabilizing in the range of25: 1 to 30: 1. favorable growing conditions. A common China, the world's largest producer and suggestion on crop yield relates to a 50% consumer of maize seed, plants 740,000 yield increase for consideration of im tons annually of which 90% are hybrids. proved seed compared to traditional land Price controls here keep the seed-to-grain races. Relate this, however, to developing price ratio unreasonably low at 4 com country yields, often less than a ton per pared to the rest of Asia at 24. This pricing hectare, where such an increase becomes structure brought about widespread adop somewhat less significant related to seed tion of hybrid maize but has likely dis and other input costs. In transition to hy couraged investment in agricultural re brids from varieties increases have been search and in-fact the average age of commonly expected to range from maize hybrids used by Chinese farmers in 20-40% from heterosis with other inputs 1993 was about 20 years and of low qual additionally significant. In developing ity compared to six years and high quality countries, management practices and cer in the U.S. Without controlled pricing, tainly availability of water may often be producers in Sudan have indicated a will more critical for increased productivity. ingness to pay as much as double for seed from a multi-national whose reputation With maize, adaptation allowed tem and track record assured appropriate pu perate hybrids of the private sector to rity and quality even for a locally devel move into Europe in the late 1940's, and oped product. Actually, higher seed costs in fact 90% of industrialized countries are are often a key factor in getting the farmer temperate. Only 25% ofdeveloping coun to improve management practices. Better tries, however, can use temperate seed doesn't cost more, it ~ more. germplasm. Therefore, most of the high tech seed developed for the U.S., Europe, Non-hybrid crops, except for vegeta and thus the vast majority of hybrids de bles, have been less likely as products for veloped by private seed companies are of opportunity in overseas markets. little direct use to maize farmers in devel DEKALB, using CIMMYT spring wheat oping countries. With Argentina there materials, found an acceptance of more was a flint versus dent grain type require than 50% in Argentina which was good ment. Therefore, the multi-national pri for the country but not profitable com vate sector at times found hybrid sorghum pared to the marketing of small grain vari a more likely first crop to enter into mar eties in Europe. Soybeans and alfalfa may kets such as Mexico and Argentina, per also be examples, especially with trans haps erroneously seeing it to be more genic traits. In September, 1998, widely adapted. Monsanto, the world's largest supplier of 131 genetically modified crops, announced it additional and significant business con would invest $550 million in Brazil to siderations. produce its herbicide, Roundup. Shortly afterwards, the Brazilian government Benefits for Private Investment made Monsanto's Roundup resistant soy beans that country's first legally ap Examples of successful seed ventures proved, genetically engineered crop. by the private sector are numerous such as maize in Argentina, Brazil, Mexico, Thai Whereas a return on investment and land and Europe; sorghum in Mexico, protection of germplasm seems consis South Africa, Argentina; sunflowers in tent across company philosophies for Argentina and Europe; and wheat in Eu overseas expansion, a more conservative rope and Argentina. Pioneer's 1998 an and an approach of much less risk may in nual report shows 27% of their com in volve exports of product or a licensing ar come was outside North America rangement with an established indige whereas DEKALB shows 34% of earn nous business. Small markets as well as ings from international seed operations in government restrictions on ownership fiscal '97 with '98 an aberration at 83%. may lead to these approaches. Zimbabwe is an example of public seed activity go Often public agricultural funding is ing to the semi-private Seed Co-op which sacrificed when monetary constraints of to survive against multi-nationals formed government shift funds to alternate in technical collaboration agreements with vestments. The private sector then will be U.S., South African, Zambian, and Ken more encouraged to increase its invest yan seed companies. ment. Such occurred as both Mexico and Argentina moved out of the developing The promotion ofprivate agri-business country category. Argentina provides a activity in sub-Saharan Africa has not good example of changes brought about been successful even when incentives by a heavy commercial emphasis with hy were given for private entrepreneurs. brids ofmaize, sorghum, and sunflowers: Pray and Tripp indicate the most com mon strategy for encouraging increased Evolution of yield of the principal cereals participation in the seed industry would and oil crops in Argentina, 1910-1994. be for governments to insure free and Cultivar Years and type Kg ha-1/year open access to products of public breed Maize 1910-50 varieties 9.00 ing programs. In India, where a supply of 1951-94 hybrids 57.92 trained agriculturalists to operate seed firms is not limiting, an absence of breed Sorghum 1954-65 varieties -7.52 ers rights legislation caused reluctance to 1966-94 hybrids 58.43 develop full fledged breeding programs or introduce proprietary inbred lines, de Sunflower 1929-80 varieties -0.36 nying producers access to the best 1981-94 hybrids 43.00 germplasm and technology. Certainly Data from Argentine Secretary of Agriculture open market policies between countries and fair and uniform seed regulations are 132 Such results greatly enhance private Limitations/Restrictions/Concerns sector relations with the producer and will Affecting the Private Sector most often lead to greater private sector support by the government. In Thailand, The unpredictable climatic character Michael Morris of CIMMYT, points out istics of many developing countries have that seed prices, as in Europe and North discouraged investment in yield enhanc America, increase with the yield poten ing technology. Farmers purchasing im tial, leading to single-cross hybrids hav proved seed and fertilizer risk losing their ing a seed-to-grain price ratio of 27-30 investment causing growth in agricultural compared to OPV's of 4-5. production to depend more on land ex pansion than yield increase, even then Numerous secondary advantages for lagging in comparison to population. foreign investment include extra genera Also, the private sector, while able to op tions such as with the southern hemi erate more efficiently, faces lower com sphere, early screening, germplasm ex petitive pricing since public seed organi change, grain quality differences, plant zations can include some type of direct or pest screening, temperature, moisture and indirect subsidy. Pronase of Mexico sets soil toxicity stress tolerance. A significant prices to only cover its operating costs benefit, perhaps utilized more frequently and does not have to recover research, with the rapid introduction of packaging, or promotional costs as it at transgenically modified hybrids, relates tempts to provide seed for low-income to seed multiplication of a foundation or farmers. commercial nature by use of the opposite hemisphere. Since multi-nationals require a mini mum time frame of 8-1 0 years to develop Also, expanding multi-nationals gen adapted cultivars to a new-entry market erally operate in a growing market. For they must be prepared to give a new ven example, Brazil has moved maize from a ture a reasonably long term trial for suc food to feed crop but only 50% of the area cess. The availability of public breeding is in hybrids. A similar trend is occurring material during this phase of establish in Africa with sorghum and very little is in ment becomes all the more critical. With hybrid. In fact, in Ethiopia, where the crop either public or private improved seed, is thought to have originated, only 5% of cultural preference can become signifi the planted area is even in improved vari cant to product acceptance. Hageen eties. Naturally, the larger seed concerns Dura-1 grain price dropped to 35-50% of can better justify their costs ofbiotechnol the local varieties three years after its re ogy by applying this science to appropri lease in part due to farmer and millers' ate benefits in the developing countries as complaints about the smaller, harder seed well. and a blander kisra. By 1990 the price dif ferential disappeared and 90% of farmers interviewed reported that HD-1 kisra was equal to or better than traditional sorghums. Sanders et. al. conclude that the narrowing of price over time may in- 133 dicate tastes to be dynamic, responding to an appreciation of intellectual property higher yield potential (lower costs) of the rights including plant variety protection. new cultivar. In Latin America and sub-Saharan Africa consumers strongly The emerging private sector must be prefer white maize but there is no scien prepared to face real concerns as they add tific support associated with digestibility a new dimension to a previously less com or nutritional quality. Thus the introduc plex agriculture. The movement, mainly tion of improved seed must overcome in developed countries toward stricter farmer preference for traditional variet protection of commercial plant varieties, ies, especially if they command a pre has been accompanied by a parallel initia mium in the market. tive by the developing countries to protect local varieties and landraces, requiring Pray and Tripp point out that during the payment for their use. Another concern, early stages of seed industry development as well stated by Morris, suggests that the regulatory policies may be designed to re drive for increased efficiency leads to an strict participation. When decisions about increasing concentration and vertical in which varieties may be released are made tegration in the private seed industry. by government plant breeders, the result Such trends may prove undesirable if a can be an effective government monopoly small number of very large companies of new varieties and virtually no private dominate the global seed industry. An ex sector involvement in plant breeding. ample of the current impact of one such Regulation should address breeder's re company, Monsanto, only in relation to quirements for variety protection and ade the Western Hemisphere: quate rewards; seed producers' concerns about bureaucratic interference and un 1998 Monssanto Global Seed fair competition; and farmers' expecta Market Share (SOM) tions ofaccess to a wide range of varieties Country SOM% Country SOM% from good quality seed. u.s. Mexico Com 15/50' Com 60 Finally, the private sector must operate Soybeans 33 Sorghum 64 in a partnership with public sector Rand Sorghum 29 Argentina D both worthy ofrecognition and support, Cotton 84 Com 65 not regarded as competition to be met Brazil Sorghum 60 with suspicion. Plant genetic resources Com 60 Sunflowers 30 can best be exploited if they remain fully 'withHolden accessible to all users and if passport in formation and technology on their use is Another example could very well have widely disseminated. Not only must the been Empresas La Moderna SA, a Mexi seed industry educate farmers about the can vegetable seed company whose seed advantages of adopting improved accounts for 40% of all vegetables sold in germplasm and the benefits of using pur U.S. supermarkets. ELM recently ac chased seed but all parties, which in quired some 11 companies that sell vege cludes scientists and policy makers, need table seed. 134 Predictions heard in many developing much needed contribution of the private and industrialized countries that sector to countries ready for such a transi profit-motivated seed companies will in tion. With the excessive costs associated evitably engage in socially undesirable with products of biotechnology the larger behavior in the absence of strict govern multi-nationals offer the best opportunity ment controls have proved largely un to provide this science in problem solving founded. Optimistically, such a consoli or improved products. As developing dation as shown by Monsanto, with large country's seed systems evolve so will market share, will in fact stimulate the their agricultural productivity, a mission larger player to increasingly search out of urgent importance. new markets and in so doing bring the 135 Rapporteur's Report Anand Kumar Current Status of Hybrid Pearl Millet Seed Production Day 1 - Session 3 General introductions were made by have to appreciate that any errors on their Dr. Botorou Ouendeba. part will lead to erosion of confidence. Dr. David Andrews discussed pearl I. Mouddour - the Multiplication Cen millet research on new cytoplasmic tre in Niger clearly demonstrated the need male-sterility hybrids. for the private sector to be involved. We, in Niger, did not get any impact from the B.R. Barwale, of Mahyco, India, dis multiplication centers. Efficiency and ef cussed the critical role of the private sec fectiveness for the private sector are criti tor in less developed countries where re cal. sources are limited. He stated that initial subsidies might create an impact - the B. Maunder - Reasons for success - De farmer is always interested in getting a pendable supply of seed, acceptable vari better price for his product than getting eties/purity, improved levels of perfor subsidies such as seed at a reduced rate. mance, and public sector plant breeding Industry has to be aggressive. Seedsmen programs are essential for the success of the private industry. 136 Tuesday - September 29 -Session 1 Bases for Successful Hybrid Enterprises Opening Remarks Moderator - A. Kumar, ICRISATIISC Production of Quality Hybrid Seed Soumana Souley and Magagi Abdou Introduction be clean. These measures will favor good irrigation and normal flowering ofthe two Experience elsewhere has shown that parents of the hybrid, especially a good the existence of a viable seed industry de grainfill in the seed parent. Abundant pol pends on the production of quality seeds. len in the male parent will also minimize Seed certification provides standards of unwanted fertilization from sorghum in quality but there are other measures that neighboring fields. need to be taken during the production process. The field should not be smaller than one quarter of a hectare to obtain a mass In this presentation we will discuss pollen effect. Isolation is important to critical factors for the production of qual avoid foreign pollen from causing fertil ity NAD-l seeds as per our experience in ization. Where there are a large number Niger; they relate to choice of field, crop of farmers in the same area it may be diffi management, and postharvest handling of cult to realize the recommended isolation the seed. distance. This can be compensated to a degree by increasing the number of bor Field selection der rows between the seed field and a field of a different variety within the isolation During the flowering period a lot of recommendation. A distance of 200 me pollen is required. For adequate pollen ters is recommended for the production of distribution, a square field is better than a seed of hybrid sorghum. If the potential long narrow one. contaminating sorghum is similar in grain color, plant height, and the head is not of Irrigation is important for seed produc an open grassy type, the isolation distance tion. A good seedbed must be available can be modified as indicated in the fol with no bumps, and irrigation canals must lowing table. Table 1. The minimal number of border rows needed for the male parent. Minimum distance from Field size up Field size greater contaminant (meters) to 8 hectares than 8 hectares 125 o o 115 2 1 110 4 2 87 6 3 72 8 4 60 10 5 50 12 6 38 14 7 25 16 8 o 10 139 If a neighbor will sow the pollinator Land Preparation parent this will also help avoid problems of isolation. When there are a number of Land preparation should involve farmers each farming relatively small bringing the field to good tilth and being hectarages that are contiguous with completely free of weeds. It may be nec neighbors it may be possible for the group essary to wait for rain or pre-irrigate to of farmers to all produce seed. have adequate moisture for land prepara tion. The field chosen should not have been sown to the same crop the previous year or Fertilization the field should be pre-irrigated to germi nate any volunteer plants that can then be Some farmers broadcast a basal dose of destroyed prior to sowing the seed crop. fertilizer that is worked into the soil at land preparation. This is not a bad tech Crop Management nique if the field will be subsequently fur rowed for irrigation. If the field is flat or It is important to appreciate that there is the irrigated units are flat it is probably a difference in growing a crop for seed best to band fertilizer close to the seed. and growing a crop for grain. A producer This can be done at the time of sowing but wants to maximize seed yield and because is frequently done when the plants are of the increased value of seed may use 10-15 cm tall. It is recommended to fertil more fertilizer and other inputs than he ize with 200 kg Nlha, that can be in the would for a grain crop. form of 150 kg of15-15-15 at land prepa ration and 100 kg of urea as side dressing. Experience in Niger indicates that the Some producers have used manure and best seed yields come from fields that lower rates of inorganic fertilizer and have been well managed over the years. have realized excellent yields. Fields that have been exhausted by con tinuous cropping over many years with Irrigation low or no inputs not only result in crops with lower yields but are less uniform If the field is furrowed, seeds can be even if high rates of fertilizer are used on sown on the top ofthe bed or on the side of the seed crop. Continuously cropped the bed which many prefer because the fields to the same crop may harbor pest soil is more readily wet by irrigation wa problems that are not uniformly distrib ter. This can be a problem in heavy clay uted in the field. In the case of sorghum, soils with poor water penetration or very nematodes, and acrimonium wilt, are ex sandy soils where water primarily moves amples of such problems. Crop rotation, downward and the soil can dry rapidly in the growing of green manure crops and the hot sun. the incorporation of crop residue will all help to keep fields productive. Sowing If the parents of the hybrids do not flower at the same time (nick), the parent 140 that takes longer to flower needs to be can delay the time of flowering of one par sown first. In the case of NAD-1 seed ent more than the other. In higher rainfall production, it is established that the pol and longer season areas it may be best to len parent (male parent) must be planted 7 delay sowing until the danger oflate rains days before the seed parent. Sowing in at crop maturity has passed and good structions are given in a training manual quality seed can be harvested (free of published by INRAN and INTSORMIL mold and discoloring). A late sowing at in 1997. A comment on border rows is ap Bengou, however, resulted in a severe at propriate here. In keeping with certifica tack of midge resulting in substantial tion standards, border rows may not ap yield loss. Possibly late sowing could re pear as necessary but they are so. A good sult in increased stemborer attack. These yield on the seed parent is necessary and problems differ with each location so ex for those rows of the seed parent on the perience is valuable. A well-managed outside they may depend on border rows. seed field will produce 1,800 to 2,000 kg In that case border rows should be planted of seed per hectare, not counting the grain seven days before the female parent. This produced on the male parent. way there will be abundant pollen when the female parent starts flowering, and re Weeding duced contamination from foreign pollen. Sowing rate and spacing should be used It is important to keep seed production so that final plant stand is 180,000 to fields free of weeds. Two to three 200,000 plants per hectare. In rows 80 cm weedings should be done as early as pos apart there should be a plant about every sible for the crop to get ahead. 10-15 cm. Thinning Isolation distance can also be manipu lated by sowing the seed crop 3 or 4 weeks Thinning to the right plant stand before or after the neighbor has sown pro (80,000-120,000 plants per hectare) is im vided both crops take about the same portant. This is particularly true if 10-20 amount of time to flower. If the neigh seeds have been dropped in pockets. bor's crop flowers 3 or 4 weeks before or Thinning should take place when plants after the seed crop isolation by time might are 10-15 cm tall and it is best if the soil is be adequate. It would be worthwhile for wet to allow roots to be pulled and to per producers to become familiar with the be mit remaining plants to recover from root havior of other varieties of the same crop pruning. in their areas of seed production. Transplanting Experience needs to be gained about the best time for sowing. It appears that in Ifthere are gaps in the stands in parts of much of the cropping area in southern the field and thinning is required in other Niger the best period is the last week or parts it is possible to use the thinned two of June up to the middle of July. plants for transplanting. Experience indi Sowing as late as August can result in cates that very young plants do not trans plants of less stature and lower yield, and plant very well. Some farmers are even 141 transplanting plants 25-30 cm tall. It is sible to identify an off-type until seed be best if plants are 15-20 cm tall. Most of gins to form and a difference in grain the leafblades are removed from the plant color becomes apparent. It is important to leaving a fairly stiff plant for transplant make a periodic inspection of the field to ing. The transplanted plant should be identify off-types as soon as possible. placed well into the ground and irrigated as soon as transplanting is finished. There are certification standards that indicate a permissible level of off-type If transplanting is undertaken in a por contamination. For production of hybrid tion of the field, plants of the seed parent sorghum, the number of heads or plants of and the pollinator parent should be trans another variety is one in 2,000. If it is planted as this process adds a week or doubtful whether it is an off-type plant more to the time of flowering and can re (similar to other plants in the field), the sult in failure of a good nick. level of contamination is one plant in 1,000. For breeder and foundation seed Transplanting can be a useful correc the standards are stricter. Breeder seed is tive measure but it is not as good as a crop one off-type in 50,000 plants and for from seed. foundation seed, one off-type in 35,000 plants. While there is not yet an active Crickets at times occur in large num seed certification program, these figures bers and damage seedlings to the point of do present limits to the producer who is affecting stand. The use of an insecticide concerned about seed quality. to control this would be worthwhile. Pollination Roguing Pollination on a warm dry morning Roguing is the removal of off-type pollen may begin to shed soon after the plants and should be practice throughout sun reaches the field. If there is humidity the life of the crop. At times an off-type or dew the shedding of pollen may be de plant is vigorous with a good-sized pani layed until 8 :00-9:00 am. And if the plants cle and there is temptation to leave it. are wet from rain it could be 10:00-11 :00 Again, one must remember the difference a.m. between seed and grain production and good quality seed cannot be full of seeds If nicking is good, natural pollination that will be off-type in the farmers' field. may be sufficient to result in a good seed It is important that off-type plants are re set on the seed parent. It may be useful to moved before they shed pollen. If they walk between the pollinator rows with a shed pollen they can fertilize seed parent stick in each hand and hit the panicles plants resulting in seed that will produce knocking the pollen free. On a day with off-type plants for the farmer. no or little wind much of the pollen will fall to the ground. If a motorized duster is It may not be possible to identify an available it can be used to places where off-type plant during the grand growth pe seed parent heads are in flower with few riod of the crop. Sometimes it is not pos- pollinator heads nearby, pollen can be 142 collected in bags. The bags are slipped moisture and if not dried again and they over the seed parent panicle and vigor go into storage at high moisture, germina ously shaken. Pollen from one bag gener tion will go down and seed quality lost. ally will pollinate one seed parent head. A The germination of seed sold to farmers crew of 4-5 people can handle many should be 80% or better. pollinations in a day. When harvesting, it is recommended Crop maturity and harvest: these are that all pollinator plants are harvested and important considerations. The seed in the the panicles removed from the field be tip of the sorghum head matures about 4 fore harvest of the seed parent begins. days before seed at the base. As the grain This will help prevent mixtures. Also, the matures, it begins to color and when fully threshing of pollinator and seed parent colored (no tinge of green), it is approach heads should be sufficiently separated ing what is known as physiological matu that mixture is not possible. If a threshing rity. At this point in the seeds develop machine is used, it needs to be carefully ment it has reached maximum dry weight. cleaned so there is no seed of one type in A good way of knowing is to take a few the machine when another type is to be seeds from the base ofthe panicle. Exam threshed. ine the bottom of the seed and if there is a dark black spot where the seed was at Seed should be stored in bags, prefera tached to the plant, physiological matu bly cloth, in an area that can be kept rela rity has been reached. Moisture percent tively cool. Thick walls, high ceilings, age is still high- over 20%; however, if the and insulated roof even of thatch or stover weather is rainy and conducive to grain will help reduce temperature. If air condi damage, harvest can be undertaken. The tioning is available, this is a big help. harvested heads must be hung under a There should be air movement and the shelter where there is air circulation, even floor and walls should be smooth so that ifby a fan or they must be dried in a grain they can be swept to reduce a build up of drier when available. storage insects that will live in cracks. Steps should also be taken to control ro Normally, the crop is left in the field for dents. several more weeks to dry. For harvest and storage the moisture percentage Conclusion should be 12 percent or less. It is advis able, following harvest, to dry the heads To produce N AD-l seed, a seed grower in the sun for a few days to be sure that the must follow certain practices including: grain moisture is low, at 10-12% moisture the seed is hard to bite. • the choice of a uniform, flat and well-drained field; Harvested heads need to be protected • avoidance of a field with sorghum from rain otherwise they will pick up as previous crop; • good isolation of the field; 143 • good seedbed to promote vigorous • follow adequate agronomIc prac germination and growth of the tices; seedlings; • harvest and thresh the parental • sufficient quantity of water; heads separately; • high level of inorganic fertilizer • treat the seed and store it in a dry like 15-15-15 in basal application cool place; and urea in sidedressing; • check germination before the seed • follow staggered planting dates for is distributed for commercial plant the parents; ing. 144 Field Management for Sorghum Seed Production Seyni Sirifi, J.W. Maranville and R.K. Pandey Summary ity condition is generally required. The distribution and quantity of rainfall deter Successful sorghum seed production in mines the water availability for crop Niger requires careful adoption of crop growth and crop response to nutrients. management practices. Lack of appropri Poor distribution and low amount of rain ate techniques of crop management can fall often demands supplementary irriga lead to poor crop stand, reduced seed tion to ensure stable production and mini yield and sometimes failure of the crop. mize the adverse effect of water stress The most important crop management during grain filling. criteria are site selection and land prepa ration, crop establishment - row spacing In the early stage of crop growth, and density, nutrient management (fertil weeds compete with crop for water, nutri izer rate, application method and time), ir ent and light. Failure to have good weed rigation requirement, plant protection, control results in yield reduction. When harvesting and threshing. the crop is tall (60-80 cm), it can smother weeds. Generally one or sometime two Results from field studies on several weedings are required to eliminate weeds soil types in Niger suggest that the crop during the season. performs better on heavy fine textured soil compared to sandy soil; however, Harvesting and threshing are also cru with improved nutrient and water man cial factors that determine the quantity agement, a satisfactory seed crop can be and quality of seed yield. Grain moisture raised on a sandy soil. Field studies on content at harvest and method of thresh crop establishment suggest that the ridge ing may determine seed grown breakage. and furrow system is superior to flat bed Field losses due to these factors must be as it allows greater water use by the crop minimized. and facilitates intercultivation. Studies on row (60-70 cm) and plant spacing (20-25 Introduction cm) provides sufficient plants (80,000 to 100,000 plants/ha) to produce high yield. Hybrid seed technology offers poten tial for enhancing crop productivity Nutrient management is critical for in through heterosis both in rainfed and irri creased seed production. Under low soil gated culture. However, adoption of this fertility, increased amounts of nutrient is technology without appropriate nutrient required compared to moderate to high management in Niger on nutrient de fertility soils. To achieve a target yield of pleted soils would further deteriorate soil 2 t seed per hectare, 120-30-30 kg N-P-K fertility. Thus, proper crop and soil man per ha in low soil fertility conditions and agement under a well coordinated gov 90-30-30 kg N-P-K under moderate fertil- ernment policy on fertilizer availability 145 and pricing would be required to enhance Soil Type and Its Fertility sorghum crop yield from the present low productivity. Developing a seed produc Soil type and its fertility level are key tion system to support improved seed, fer factors affecting sorghum productivity in tilizer and other input supplies to farmers the Sahel. Sorghum performs well on is a necessary step to achieve the goal of finer textured soil, and most of Nigerien increased food production. farmers prefer to grow the crop in finer textured soil than sandy soil since they Managing a sorghum crop for seed re use very low inputs. However, sorghum quires understanding of crop needs for can produce well on sandy soil provided plant nutrients, water, plant protection the constraint of water and nutrients is re and environments in which the crop is moved or significantly reduced. Most grown. Use of improved crop production soils in rainfed areas are poor in fertility practices and inputs determine the yield and have low organic matter and water re level. Depending on target seed yield, tention; thus, soil based management production practices and inputs must be practices is crucial for seed production to adjusted to achieve the goal. The sorghum attain target yields. seed crop is usually planted in a ration of two male rows and four female rows and A field study was conducted to com both have very similar cultural require pare the effect of soil types on nitrogen ments except possibly an adjustment of and phosphorus requirement of sorghum planting date to synchronize flowering. in the Konni Perimeter. This area is char Otherwise, that of field management of a acterized by mean rainfall of 477 mm seed crop is fairly similar to that of com with potential evaporation of 679 mm mercial open pollinated varieties regard from July to October. Four nitrogen levels ing nutrient and water requirement to pro were tested with and without phosphorus duce similar yields. on nine farmer fields involving three soil types: silty clay loam (vertisol), sandy In Niger, successful seed production loam and sandy (inceptisol). Results pre must have a satisfactory seed yield level, sented in Table 1 show a differential re and improving current seed yield from 1 t sponse and requirement of sorghum to ni ha- I to 2.5 t ha- I would require adoption of trogen and phosphorus in the three soil improved crop management practices. types. The silty clay soil had higher yield There is obviously a need to examine crop than sandy soil at similar nutrient applica management practices using relevant tion as it has higher nutrient supplying ca field data that determines crop productiv pacity. A differential genotypic response ity in Niger. to nutrients was observed and the hybrid NAD-l had slightly higher yield than the The purpose of this paper is to review improved line Sepon-82 (Table 2). Sor crop management practices based on the ghum requires more nutrient when grown results of field experiments conducted in in a low fertility sandy soil than a moder Niger. ate fertile silty loam to produce a similar yield. In the Maradi perimeter, where two types of soil, sandy and sandy loam are 146 Table 1. Response of sorghum to nitrogen on three soil types in the Konni perimeter in 1996. Nitrogen Soil type Rates Sandy Loam Siltv clay loam K~ha-l PO P50 PO P50 PO P50 N45 2035 2084 1993 2132 2229 2453 N90 2174 2381 2111 2285 2458 2658 N135 2521 2403 2528 2740 2653 2757 N180 2799 2917 2958 3417 3278 3201 Mean 2382 2446 2397 2643 2654 2767 SE+ 256 Silty clay soil (vertisol), sandy soil (inceptisol), crop had water stress at several growth stages. Table 2. Nitrogen response ofNAD-l and Sepon-82 to nitrogen with P and without P on grain yield on the Konni perimeter in 1997. Nitrogen rate Sepon-82 NAD-l iI~~ha-l) PO P50 PO P50 N45 2030 2218 2199 2324 N90 2190 2318 2241 2574 N135 2370 2669 2606 2754 N180 2874 3079 3097 3329 Mean 2366 2571 2536 2745 SE+ 209 Silty clay soil (vertisol), sandy soil (inceptisol), crop had water stress at several growth stages. found, grain yield was superior in the Good seedbed preparation is necessary sandy loam compared to sandy soil at the for seed emergence and subsequent crop same level of nutrient application (Table growth as well as water retention. Soil 3). Maradi is characterized by a mean moisture at planting is critical for germi rainfall of 491 mm and potential nation and good field emergence. Warm evapotranspiration of 701 mm. To opti air temperatures and high atmospheric mize sorghum productivity, nutrient ap evaporative demand allows the soil sur plication and use in different soil types face to dry fast. Lack of sufficient mois must be based on soil tests if available. ture in the upper soil layer and a shallow planting depth often results in poor Crop Establishment stands. Seeding depth and number of seeds per hill often determines field emer Crop establishment is a crucial factor gence. in realizing good seed yield in Niger. 147 T a bl e 3 0 n-f arm per ormance ofNAD-l in comparison to sepon- 82 at M arad" lIn 1997 Soil type On farmer's fields Sandy Sandy loam N rates kg ha- 1 St:pon-82 NAD-I Sepon-82 NAD-l 27 1833 2100 2300 2566 54 2000 2333 2400 3633 81 2233 2620 2566 3700 108 2488 3200 3400 3866 LSD5% 438 CV 19.9 Sandy soil (>70% sand), Sandy loam (45 sand, 35 clay, 20 silt). Crop had mild water stress three times during crop season. Sorghum is generally planted on a flat study showed that use of the ridge system bed in many areas for grain production. had advantage in improving grain yield Use of furrow and ridge systems increases over a flat bed system when rainfall is low rainwater harvesting in rainfed areas, thus by improving water use efficiency (Table higher productivity. In the irrigated per 4). imeter, these systems facilitate irrigation and inter-cultivation for weed control; In another field study, effects of alter thus, they deserve greater attention by ing plant density from 40,000 to 120,000 seed growers. Research results generally plants were evaluated at the Maradi re show improved productivity when using search station in 1997. Results of this the ridge-furrow system where rainfall study showed that increasing plant den conditions are limited. sity from 40,000 to 100,000 plants per hectare increased grain yield (Table 5). A crop planted on ridges 60-75 cm As the application of nitrogen was re- apart and a hill to hill spacing of25-30 cm with 2 seeds per hill generally results in an Table 4. Effect of method of planting and ni estimated plant density of trogen rate on grain yield of sorghum 80,000-100,000 plants per hectare if in Mardi in 1997. Nitrogen Flat bed RidgelFurrow seeds have 75-80% field germination. 1 (kg ha- ) Method Method Thus, crop establishment becomes the 0 1087 1583 first step in crop management for en 45 1584 1950 hanced productivity (Pal et aI., 1996; 90 2108 2375 Patel et aI., 1992). 135 2775 3117 180 2590 3095 A field study was conducted to test the effect of the ridge and furrow system of Mean 2029 2424 planting on grain yield of sorghum com pared to the use offlat beds in Maradi dur LSDO.05 324 ing the 1997 rainy season. Results of this 148 duced, grain yield was also reduced at suggests that sorghum responds to nitro each level of plant stand. Under field con gen up to 180 kg ha when soil moisture is ditions, to attain high grain yield, optimal not limiting; however, under rainfed con plant density with optimal nutrient use ditions, lower rates should be used. should be targeted (Clegg et aI., 1974; Patel et aI., 1992; Singh et aI., 1987; Villar A field study was conducted at Lossa in et aI., 1989). 1996 to evaluate the effects of nitrogen phosphorus and potassium on grain yield Nutrient Management of sorghum on sandy soil where organic matter and available nutrients were low. Among plant nutrients, nitrogen is the Rainfall was 350-400 mm in the 1996 most limiting nutrient followed by phos- rainy season. NAD-l hybrid and the open pollinated variety Sepon-82 were evalu Table 5. Effect of plant density at three nitro ated at five nitrogen levels: 0,45, 90, 135 gen levels on grain yield with supple and 180 kg N per hectare with and without mentary irrigation (150mm) in Maradi in 1997. phosphorus and potassium. Results Plant den- showed increasing yields up to 180 kg N sity N45 N90 N135 per hectare, and NAD-l out yielded plants/ha Sepon 82 (Table 6). 40,000 2095 2980 3413 60,000 2360 3170 4667 80,000 2780 4250 5306 Another experiment was conducted at 100,00 3884 4917 5534 Lossa in 1997 to evaluate the effects of ni 120,000 3467 4580 5417 trogen, phosphorus and potassium on grain yield of sorghum on a sandy soil Mean with low organic matter and available nu LSD 0.05 310 trients. NAD-l hybrid was evaluated at Seasonal rainfall - 465 mm five nitrogen levels: 0,45,90, 135 and 180 kg N per hectare with and without phos phorus (Maranville et aI., 1980; Singh et phorus and potassium. A split plot design aI., 1987). Sorghum response to nitrogen with phosphorus and potassium levels as is very dramatic, and 2-3 fold increases in the main plot and nitrogen as the subplot grain yield are not uncommon. Phospho was used with four replications. Supple rus effect on yield increase may be 10-25 mentary irrigation (300 mm) was applied percent depending on the level of avail to meet the crop water requirement. Under able phosphorus in soil (Cripps and supplementary irrigation, sorghum re Matocha, 1987; Pawar et aI., 1987; Pawar sponse to nitrogen on sandy soil was dra et aI., 1988). Response of the crop to po matic and an increase of nearly three fold tassium is lower (Pawar et aI., 1987; was observed. Response to phosphorus Pawar et aI., 1988). Removal ofpotassium was smaller but significant (Table 7). Ap by the crop from the soil deserves consid plication of potassium did not signifi eration if long-term productivity is to be cantly increase grain yield. Addition of realized. Thus, a basal application of com organic residue with phosphorus and po pound fertilizer (15-15-15) is generally tassium increased yield over phosphorus recommended. Recent research in Niger and potassium application. Application of 149 Table 6. Effect of nitrogen, phosphorus and potassium on grain yield (kglha'l) of rain fed grain sor- ghum in Lossa during rain~ season in 1996. Nrate (kg ha'l) P40 P40 + K75 Mean P40 P40 + K75 Mean NO 625 679 652 481 539 460 N45 951 969 960 839 791 815 N90 1055 1360 1210 902 993 947 N135 1665 1922 1793 1581 1525 1553 N180 2012 2387 2199 1757 1812 1734 Mean 1262 1463 1363 1112 1132 1102 SE 173 Seasonal rainfall 343 mm, sandy soil (86% sand), inceptisol, Low organic matter (0.3%) Potential evapotranspiration (lOOdays) 693 mm. Table 7. Effect of nitrogen, phosphorus and potassium on grain yield (kg ha'l) ofsorghum with sup- plementary irrigation 300 mm) in Lossa during the rainy season in 1997. Nrate P40 K75 + Mean grain (kg ha'l) POKO P40KO P40 K 75 2tha'l CR* yield NO 2220 2390 2520 2690 2455 N45 3080 3490 3590 3720 3465 N90 5080 5320 4350 5430 5050 N135 5610 6500 5430 7770 6170 N180 6580 7280 6420 8250 7230 Mean 4510 4996 4462 5450 4854 SE 390 CV% 14 CR' ,crop residue (2tha'l) Seasonal rainfall 343 mm, sandy soil (86% sand), inceptisol, Low organic matter (0.3%) Potential evapotranspiration (100 days) 693 mm. nitrogen resulted in a linear grain yield in weeks or so. This results in frequent water crease to the highest rare applied (180 kg stress. In contrast to this, crop per ha). evapotranspiration demand on station was met by frequent irrigation. Under Nitrogen Use Efficiency rainfed conditions, Pfp was much lower than with supplementary irrigation. Be Nitrogen use efficiency as determined tween the two genotypes, NAD-l had a by partial factor productivity (kg grain/kg higher Pfp than Sepon-82 (Table 8). N) is a measure of return on the invest ment made by farmers, Partial factor pro Crop Water Requirement ductivity (Pfp) under farmer management was lower than on-station (Table 8). One The water requirement of sorghum in reason was crop water stress due to de the Sahelian environment is not often met layed irrigation. Normally irrigation is by rainfall where there is erratic distribu provided when there is no rain for three tion and inadequate amounts. Supple- 150 Table 8. Partial factor productivity (Pfp, kg grain/kg N) ofNAD-l and Sepon-82 under supplemen tary irrigation and different nitrogen rates. On-fann Station Nitrogen I (kg ha- ) Konni Maradi Lossa Maradi N45 42.7 56.8 77.0 69.5 N90 29.8 47.2 56.1 57.1 N135 23.2 33.9 45.7 41.5 N180 22.2 27.0 40.2 34.1 Rainfed Conditions Nitrogen NAD-l Sepon-82 I (kg ha- ) Pfp N45 2l.3 18.1 N90 13.4 10.5 N135 13.3 11.5 N180 12.2 9.6 mentary irrigation is necessary to realize need to be made. Ifweed intensity is low, full yield potential. Evapotranspiration one weeding may be sufficient in some lo demand of sorghum ranges from 450-550 cations. Due to lack of availability, use of mm if grain yield of4000 kg or above is to herbicides is generally not feasible. be realized (Brun et aI., 1972; Lewis et aI., 1974; Zweifel et aI., 1987). In an Harvesting and Threshing agro-ecological zone with 300-400 mm rainfall, supplementary irrigation is Harvesting of the seed crop can take needed to produce good seed yield. In the place anytime after the crop has reached 450-550 mm Rainfall zone, supplemen physiological maturity (when moisture is tary irrigation is needed only if rainfall is between 20-25 percent). Generally, the inadequate or not well distributed. Three crop is kept in the field until grain mois to four supplementary irrigations may be ture falls to 9-11 %. Ifweather conditions needed in many Sahelian areas to produce favor grain deterioration, harvest can oc a good seed crop. A comparison of a cur earlier but the heads must be dried un rainfed and supplementary irrigated crop der shelter. If sorghum heads with high showed a large increase (nearly two fold) grain moisture are piled, heat from respi in grain yield when irrigation was pro ration will occur and germination lost. vided at booting, heading and grain filling Generally, all pollinator heads should be (Table 9). harvested and removed from the field be fore harvest of seed parents begins. Plant Protection Growth Stage Based Crop Manage Among plant protection measures, ment to Get High Seed Yield weed control deserves the greatest atten tion. Normally, two manual weedings at Crop growth and development stages three sand six weeks after crop emergence can be broadly divided into three phases. 151 1 Table 9. Effect of supplementary irrigation on grain yield (kg ha- ) of sorghum in 1997 grown at dif ferent nitrogen rates. Rainfed Rainfall (402 mm)+ Nitrogen rate rainfall supplementary 1 (kg ha- ) (402 mm) yield irrigation (200 mm) yield 1 ------kg ha------o 848 1964 45 1248 2772 90 1573 4040 135 2331 4936 180 2834 5784 Mean 1777 3899 LSD 0.05 412 Long tenn rainfall (491 mm with standard deviation 143 mm) Evapotranspirationn potential-557 mm. Crop management specific to each stage physiological maturity of the should be followed to achieve the best grain. Water stress can be a major seed yield. limiting factor during this stage and the crop water requirement must be • Growth stage 1 (GS-l) is the pe met to ensure high seed yield an riod from germination to conver quality. sion of the apical meristem from a vegetative to a reproductive one. Any stress (water or nutrient) which Here attention to weed control, wa hinders the normal progression of events ter and nutrient are key components in GS-2 can have marked effects on yield of good management. The applica through reduction in plant size (root and tion of N-P-K is normally done at leaf area) and the number of seeds per time of planting to avoid nutrient panicle (Lewis et aI., 1974). Water stress stress. has the most adverse effect on yield when • Growth stage 2 (GS-2) is the de it occurs during booting or heading stages velopment ofthe plant from panicle (Musick and Grimes, 1961). Among the initiation to flowering. Attention to three stages, GS-2 is the most sensitive nutrient and water supply at GS-2 is followed by GS-3 and GS-l. To maxi a key for determining seed yield as mize seed production, efforts should be stress will lower panicle grain num made to reduce water and nutrient ber. A second application of nitro gen is generally done at this stage to stresses. avoid stress. • Growth stage 3 (GS-3) is the grain filling period from flowering to 152 Conclusions fect quantity and quality of seed produc tion. Since management of a seed crop is High seed yield requires proper crop very similar to that of a commercial grain management at all stages of crop growth. crop, suitable modification in crop man Raising a good seed crop requires atten agement practices should be tailored to tion from field selection to crop harvest meet the needs of the region. Raising the ing and threshing. Neglect at any stage, current average yield level from 1 t ha- 1 to site selection and land preparation, crop 2.5 t ha- 1 will require fine tuning of crop establishment-row spacing and density, management practices. nutrient management (fertilizer amount, application method and time), irrigation requirement, plant protection and har vesting and threshing will adversely af- Target Yield Based Nutrient Management Low fertility sandy soil Seed yield Nutrient level Fertilizer 1 1 (t ha- ) N-PKkgha- 1-1.5 t ha- 1 30-30-30 basal 4 bags (15-15-15)-200 kg 46 top dress 2 bags (urea)-IOO kg 2-2.5 t ha- 1 30-30-30 basal 4 bags (15-15-15)-200 kg 92 top dress 4 bags (urea)-200 kg Moderate fertility level (irrigated KonnilMaradi perimeter) 1-1.5 t ha- 1 30-30-30 basal 4 bags (15-15-15)-200 kg 22.5 top dress 2 bags (urea)- \00 kg 2-2.5 t ha- 1 30-30-30 basal 4 bags (15-15-15)-200 kg 67.5 top dress 4 bags (urea)-200 kg Target yield (2.0t or above) based on crop water requirement Rainfall zone Potential ET Supplementary irrigation need 300-400 Tilleberry 400-500 2-3 irrigations(150-200 mm) Lossa 450-550 Konni 400-500 1-2 irrigation(50-100 mm) Maradi 600-700 Bengou 400-500 None adequate Tara Rainfall expected 153 Summary of management practices for successful growing of a sorghum seed crop Consideration Specific Requirement 1. Field selection Sandy/clay loam, fine texture soil and land preparation Plowing and cross harrowing 2. Crop establishment Ridge and furrow system row spacing and 60-75 cm, 20-25 cm hill to hill plant density 80,000 to 100,000 plantslha 3. Nutrient management Based on soil test - N, P and K fertilizer amount, 90-120 kg N, 30 kg P and 30 kg application method I. Basal at planting, side band (15-15-15) and time 2. Side dress urea 4-6 weeks after planting. 4. Irrigation requirement 2-3 irrigations or more depending on rainfall and its distribution 5. Plant protection Floral initiation and flowering are important times for moisture availability Weeding -1-2 hand weedings Fungicide/insecticide use - sometimes necessary. 6. Harvesting and threshing After physiological maturity Seed breakage to be avoided. 7. Storage Store dry (10% seed moisture) and prevent infestation of storage insects 8. Field uniformity Either by crop rotation or by an off-season crop grow a green manure crop at a good level of fertilitv that can be plowed into the soil. References influenced by fertility levels under rainfed con ditions. Trop. Agric. (Trinidad) Vol 73:6-9. Brun, L., E.T. Cinemas, and W.L. Powers. 1972. Pawar, V.S., A.C. Joshi, and N.K. Umrani. 1987. Evapotranspiration from soybean and sorghum Response ofKharif sorghum to N, P and K. Sor fields. Agron. 1 64: 145-48. ghum News]. 30:46. Clegg, M.D., W.W. Biggs, J.D. Eastin, J.W. Pawar, R.S., H.C. Sharma, H. Singh, and R.S. Maranville, and C.Y. Sullivan. 1974. Light and Malik. 1988. Effect of irrigation, nitrogen and transmission in field communities of sorghum. phosphorus levels on N P and K content and up take of hybrid sorghum. Haryana J. Agron. Agron. 1 66:471-476. Cripps, R.W. and J.E. Matocha. 1987. Compara 4(1):36-41. !ive grain yield response of two sorghum variet Patel A. I. , K.N. Agawal, and A.R. Parel. 1992. Ef Ies to Nand P fertilizaton. Sorghum Newl. fects of rowspacing, seed rate and fertilizers on 30:54 the yield of sorghum. Indian J. Agric. Sci. Lewis, R.B., E.A. Hiller, and W.R. Jordan 1974. 38.(5):805-810. Susceptibility of grain sorghum to water deficit Singh, T., H. Singh, and M. Singh. 1987. Effect of at three growth stages. Agron. J. 66:589-590. plant ~eomet~ and nitrogen on yield and N up Maranville, lW., R.B. Clark and W.M. Ross 1980. take III hybrId sorghum. Indian J. Agron. Nitrogen efficiency in grain sorghum J. Plant 32(4):407-408. Nutr.2:577-589 Villar, lL., J.W. Maranville, and J.C. Gardner. Musick, J.T. and D.W. Grimes. 1961. Water man 1989. High density sorghum production for late agement and consumptive use of irrigated grain planting in the central great plains. J. Prod. sorghum in Western Kansas. Tech. Bull. Kans. Agric.2:33-338. Agric. Exp. Stat. 113. Zweifel, T.R., J.W. Maranville, W.M. Ross, and Pal M.S., O.P. Singh, and H.P.S. Malik. 1996. Nu ~.B .. Clark. 1987. Nitrogen fertility and irriga trient uptake parrten and quality of sorghum tIOn Illfluence on grain sorghum nitrogen use ef [Sorghum bicolor (L.) Moench] genotypes as ficiency. Agron. J. 79:419-422. 154 Seed Conditioning and Commercialization A. Bruce Maunder Introduction not been effective or efficient as providers of hybrid seed nor have they been struc An obvious need of developing coun tured to do production and marketing. To try agriculture is technology transfer such the contrary, on-going technical support as with improved planting seed, from na and encouragement by the public breeder tional or international research improve until the seed producer becomes ade ment programs to the producer, both sub quately experienced is highly desirable. sistence as well as the large operator. Un Those new to the hybrid business must fortunately, major improvements at the anticipate a more complex procedure with experiment station level often only show outside training early on but every coun up in annual reports and producer yields try possesses the type of individuals capa remain low. Lack ofthe proper infrastruc ble of such activity. A supply of trained ture for effective multiplication and dis agriculturalists will be indispensable in tribution of improved cultivars, such as operating the seed firms. from hybridization, which requires input from the private sector, most likely ex Over the past seventy years, since the plains this constraint. For any crop the ex hybridization of maize, numerous field istence of dependable markets, relatively and vegetable crops have benefitted from large areas of cultivation, and a desire on the phenomenon coined by Shull in 1917 the farmers' part to increase yields at "heterosis" (Crabb, 1947). Examples through hybridization indicate that farm include plants, characterized by both per ers might benefit from the presence of fect and imperfect flowers, with genetic commercial seed firms. Advantages and and cytoplasmic-genetic-sterility, likely reasons for success would be: (1) protogyny, incompatibility, and genetic dependable supply, (2) acceptable qual markers allowing for wind, insect, of ity/purity, and (3) improved level of per hand pollination (Table 1). Even the oil formance (Maunder et aI., 1994). palm, Elaeis guineensis, which accounts for the major cooking oil in much of the To develop a seed industry means pro world, is in fact now derived from a wide viding a reasonable return on investment F 1 hand-cross of dura x psifera which without undue government restrictions brings not only higher yield of oil but since capital risk will be required. Private better seed characteristics into the end but indigenous seed enterprises are a logi product. cal first step but their inability to cope with monetary fluctuations and lack of Location of Production Control sufficient funds for research investment often has put them at a serious disadvan Success in seed production often re tage. Historically, public institutions have lates to the specific ecogeographic area 155 Table 1. Specific field and vegetable crops currently hybridized and the method of pollination. Field Cro s Ve etables Maize w Cotton ih Onions Squash Sorghum w Alfalfa Broccoli Beets w Wheat w Oil Palm h Cauliflower Cantaloupe h Canola wi Millet w Cabbage Cucumber ih Sunflower Castor w Tomatoes h Eggplant h Rice wh Sugarbeet w Watermelon h Pepper h Carrots S inach w W=by wind, i=insect transmission, h=hand pollination. chosen. Location should be determined adapt to the same area, production will by region of crop adaptation which will take place reasonably near the greatest de determine parental line performance in mand and will become a component of relation to frost-free days, temperature more intense growing areas of a specific means and extremes, relative humidity, crop to reduce freight costs. adequate and timely moisture, often by ir rigation, well drained soils, and minimal Equipment Required abiotic and biotic stresses. Certainly free dom from weedy relatives of the specific Field equipment required for seed pro crop will be essential. Heat units to enable duction will be dependent on that avail normal growth and pollen viability are able by the contract growers with some just as critical as a relatively dry harvest more expensive or complex items at times period to provide adequate purity as well furnished by the seedsman, examples be as quality ofthe seed crop. Areas of heavy ing a pesticide applicator and a stationary, disease pressures on the seed crop will not or preferably mobile, harvester/combine. only reduce yield but may very well pre Depending on the level of inputs, even vent movement of seed into commercial some conventional farming equipment channels. Weedy relatives not only cause may of necessity have to be provided. high costs of roguing but may cause Seed conditioning equipment would be whole lots of segments of the production the principal investment along with a rea to be eliminated from the market. sonably useable building for this equip ment as well as seed storage. Cost of such Also, the availability of temporary la equipment based on U.S. prices for such bor for field activity such as roguing, crops as com, sorghum, and sunflower detasseling, and harvesting can affect cost may approach $75,000 U.S. (Table 2) of the finished product as well as quality. where season long conditioning would This "unskilled labor" still requires a rea readily handle 1000 tons. Harvest bags, sonable amount of training. Proximity of holding bins, scales, and finished product the seed production and conditioning to sacks ofpaper or cloth as well as chemical its intended market further relates to unit treatments would be additional as would cost. Most likely because the parental ma be the need for a dryer. Some level of use terial as well as the end product generally able equipment could very well already be in place as with Niger's effort to pro- 156 duce NAD-1, a first sorghum hybrid for 20-100% for hybrids over varieties are the country (L.R. House, 1997, pers. common in crops like sorghum, Sorghum Comm.). Also developed seed operations hieolor (L.) Moench, along with enhance constantly upgrade, putting much useful ment of seedling vigor, 5-10 days quicker conditioning equipment on the market at flowering, and improved test weight, not more favorable prices. Finally, consulta to mention the greater ease of handing tion with experienced seedsmen will abiotic and biotic stress, so often associ prove invaluable at the time of putting to ated with this crop. gether an efficient and cost effective seed conditioning plant. Foundation Table 2. Equipment to condition 1000 tons of Ready availability of the parental seed maize, sorghum or sunflower seed stocks will determine the need for founda based on U.S. cost estimates. tion maintenance. Should foundation Component for Estimated cost seed be a component of the seed industry U.S. export* conditioning smaller blocks for inbred increase with Seed cleaners and sizer $ 32,000 even greater isolation to conform to cer Ideal length grades 4,500 Pre-mix tank 2,700 tification/purity restrictions must be es Seed treater 3,300 tablished. These may require increase of Elevator 8,000 all inbreds used in large scale production Conveyors (2) 8,000 as well as single cross production where $ 58,500 hybrids are 3-way or double cross types. Total Male sterility will require A-line or fe *FOB prices plus packing male production by the single cross ofA x B where the B-line is identical genetically Germplasm, Foundation, Hybrid Pro but in normal cytoplasm allowing for pol duction len production. This requirement to main tain seed stocks in sufficient quality and Without preliminary germplasm re quantity when no other source is available search to develop and determine best par becomes a significant component of a ents for hybrid production, parental lines successful seed venture. Where the devel may be obtained from rARC's, NGO's, oping or research entity has personnel and NARS, or USArD CRSPs (Collaborative equipment to handle foundation, new Research Support Programs). As reiter seed ventures will have a higher probabil ated often by L.R. House from experi ity for success. ences in India, Sudan, and Southern Af rica, the importance of a significant yield Hybrid Seed Production increase for the first hybrid or varietal re lease cannot be over emphasized. In fact The process of production requires Norman Borlaug believes that under many variables, some controllable, others some cultures yield increases of as much not, making it impossible to be crop spe as 50% may be required to achieve move cific within the framework of this paper. ment away from traditional landrace Reference to Table 1 made earlier is criti cultivars (Maunder, 1997). Yield gains of cal in that the transmission of pollen dur- 157 ing hybridization significantly effects sterile plants to escape such problems as many of these variables. sorghum ergot. Planting date studies to predict most likely splits are encouraged First, growers, often on a contract ar for all parental or most likely parental rangement, must be located for the actual lines. Roguing before, during, and fol crop production. These individuals lowing pollination is essential to maxi should be the more progressive farmers in mize purity and to meet inspection re their community. The more interested and quirements. Another complexity of deal prepared they are to cooperate, the less the ing with hybridization relates to adequate risk for a successful crop. If at all possible isolation from non-male pollen such as they should have access to irrigation and commercial cultivars and or weedy spe be able to provide reasonably uniform cies some of which may result in off-types soils without toxicity or water logging of an extremely objectionable nature such problems. Some agreement within the as rhizomes or seed shattering with sor contract may be necessary to insure suffi ghum. cient chemical inputs for a successful seed crop. These seed growers may be the Harvest must stress seed purity as well supplier oflabor necessary to handle such as quality. Generally, the male parent field activities as detasseling or roguing. should be removed first to avoid contami Payment can vary but is often based on nation. Often a pre-harvest roguing will market price plus premium per unit take place depending on season-long weight of harvested seed in an amount challenges with the specific field. Where sufficient to attract the best possible and when possible harvest will take place growers. may avoid use of a dryer or need for a desiccant. High moisture at harvest with Successful hybrid production depends too much heat from artificial drying can on high purity and quality parental (foun induce dormancy, reduce germination, or dation) seed planted in the correct se lead to rapid aging, a condition often asso quence to assure pollination nicking and ciated with light frost damage. Every ef in the correct ratio female to male for an fort at this stage should be made to care adequate volume of pollen. Examples of fully label and store separately by seed ratio by crop would be sorghum 3: 1, com lots, often quantities which represent a 2: 1 to 4: 1, sunflower from 2: 1 to 7: 1, and specific area of the field. These lots then wheat from 1:1 to 3:1 (Wright, 1980). allow for purity checks affecting only a Vegetable crop ratios also range from 1: 1 part of a field and can avoid excessive dis to 4:1 (D. Homes, 1997, pers. comm.). card should a problem with grow-outs oc Row width, ease and volume of pollen cur. movement are other factors. Where nick ing is environmentally affected, even With the overview of production re from year to year, two or more planting quirements the complexity of hybrid pro dates of male are practiced to spread the duction suggests outside technical input pollen over more days to assure a good at the beginning of a hybrid seed industry. seed set. Good nicking best reduces Many experienced with the crop could off-types in the final product and allows serve as short term consultants. Where 158 government or international center re pensive approaches is through demon search has developed the specific hybrid stration or show plots either done inde cultivar, the breeder or related agronomist pendently or in conjunction with govern would greatly enhance success by being ment extension efforts. Producer meet available for guidance and season-long ings are often held at these plots to allow problem solving. Additionally, the risks for discussion of the specific hybrid at of low seed parent yields, inclement tributes or limitations with harvest data weather, and failure to meet purity stan helpful in convincing growers to make the dards all favor input from the private sec transition to high technology seeds from tor where investors are willing to take on their common varieties. These meetings risk for a potential profit, an undertaking may also take place between cropping not common to the public sector. seasons where an update on agronomic practices, insect, disease, and weed prob Marketing, Distribution, and Service lems and their control as well as nutrient requirements can be discussed. Once the seed is produced, the cultivar's attributes must be communi Pricing of the seed must also be given cated to the commercial producer and the high priority along with collection of seed be made readily available for pur sales. No specific figures can be given chase. This may take place (1) directly here but cost of seed from the contract from the seed operation itself, (2) through grower, sales and advertising expense, a store or other retail type operation, or (3) and where applicable research cost are through farmer dealers, the approach critical to determine price which still must most commonly used in the early history take into account a reasonable return on of the U.S. hybrid industry. Larger dis the investment by the seed industry and a tributors, capable of providing more ser competitive range to other companies of vices, may become dominant in some fering the same or similar performing markets. This trend has developed to products. Remember the grower will do better provide financing and service after his own calculation and a critical piece of the sale. The larger seed companies, how information will be the opportunity for re ever, also are expanding their efforts to turn on the potentially extra cost of his provide agronomic support to customers. seed. In most countries utilizing hybrids These services attempt to provide product the seed investment is relatively small in information as well as information on relation to the increased potential for both likely crop production problems and their higher yield and extra earnings. control such as from abiotic and biotic stress. At the time of purchase the grower Conclusions should additionally receive information on germination, purity, seed size, and Success as a seedsman, once an infra even specific lot numbers. structure is in place, will relate to many variables among which will be the vol Product advertising is especially es ume to produce to adequately meet de sential in a competitive environment. Per mand. This may require a 25-30% addi haps one of the most effective and inex- tional production to account for distribu- 159 tion logistics, change in cropping improved and proprietary product. India systems, possible quality discards, and has no doubt been the recent best example hopefully competitive gains. As stated where in sorghum and millet the private earlier, likely reasons for success not only sector, value-wise, has captured 60% of include a dependable supply but also the formal seed market (Maunder, 1997). credibility and conformance to pu Now, according to a survey by Pray et aI., rity/quality standards. In developing (1991) some seventeen firms had Re countries this assurance of a true search and Development programs on ness-to-type or expectation can be an ade these crops, spent an average four percent quate advantage to develop a successful of seed sales on research and employed 31 business. The third and most significant Ph.D. graduates and 45 with M.S. de reason for a commercial seed venture grees. would be to produce and distribute to an adequate customer base a product of suf To realize benefit from the billions of ficient yield, quality or other agronomic dollars of agricultural investment for im advantage to provide the consumer or provement of developing country agricul producer a more profitable business ture will require a strong private sector whether it be selling the seed or growing a seed industry. Only with this transition superior crop. The profit incentive pro joined by those providing other critical in motes improved quality and performance, puts such as chemicals and machinery can a characteristic not natural to the public the basic principles of agronomic re sector. search, already developed, return the nec essary impact to mankind. To start up a successful operation re quires the ability to secure adequate fi References nancing by an individual or organization, Crabb, R.A. 1947. The hybrid-com makers, proph be willing to study and learn the business ets of plenty. Rutgers University Press, New and then become a primary component of Brunswick, N.J. 331 p. the agribusiness community. At times the Maunder, A.B., F. Bidinger, L. Busch, J. Husle, and M. Shepard. 1994. INTSORMIL five year best ofplanning will still result in too little EEPreview. US/AID, Washington,D.C.116p. or a carry over of supply which will re Maunder, A.B. 1997. Role of private sector. p. quire storage facilities to maintain an ade 605-612. In Proceedings of the International Conference on Genetic Improvement of Sor quate quality to the seed. There will al ghum and Pearl Millet. 22-27 September 1996. ways be challenges with collection on Lubbock, Texas. INTSORMIL Publication sales but this is a business and these chal 97-5. Pray, C. E., S. Ribeiro, R.A.E. Moeller, and P.P. lenges are part of the risk. Depending on Rao. 1991. Private research and public benefit: the quantity and quality of competition, the private seed industry for sorghum and millet the product performance must develop a in India. p. 315-324. Research Policy 20, Elsener Science Publishers B.V. customer base adequate in size to support Wright, H. 1980. Commercial hybrid seed produc the available seed industry. Success at tion. p. 161-176. In Fehr W.R. and Hadley H. this entry level can then lead to a form of (ed.) Hybridization of crop plants. Am. Soc. Of Agron. Madison, WI. in-house research activity to provide an 160 Seed Production on Irrigated Perimeters Naroua Djibo and Jean Jacques Schalbroek The Irrigated Perimeters (IPs) Niamey, Gaya, Konni Maradi, and Diffa. These regional offices support local IPs in ONAHA's Mission and collaboration with cooperatives. Organization ONAHA is being reorganized to focus more on its role of public service (techni ONAHA was created in 1978 to super cal management of IPs, extension, seed vise the creation and management of the production, research and development) Irrigated Perimeters (IPs). Today, it is a and make better use of its human re critical element in the national strategy of sources and equipment in new projects. rural development in the area of irrigated agriculture. Initially, ONAHA directly The Cooperatives (Co-ops) managed the IPs. It gradually shifted to a support organization that helps farmer co A national workshop held in Zinder operatives toward self-management. (1982) decided that farmers would be in Since 1982, all IPs have had coops that are volved in coop management and control in charge of planning and managing their in order to make the IPs more sustainable. own activities. In 1984, ONAHA's status ONAHA presently supports 47 coops was changed to a public office with indus composed 0[306 production groups for a trial and commercial activities, which total of approximately 30,000 farmers. gave it more flexibility for management of its human and financial resources. Its Description ofthe IPs present mission is threefold: The organization of the Irrigated Per • create IPs at the request ofthe gov imeters is described in Table 1. Thirty Six ernment; IPs are along the Niger River where irri • technically manage the IPs, includ gation water is pumped (Tillabery, ing training coop members; and Niamey, and Gaya). The two Konni IPs • implement and/or evaluate irriga are irrigated from reservoirs that collect tion projects. water flowing from the hills during the rainy season. The Maradi IPs are irrigated ONAHA has its headquarter offices in from deep wells. The Diffa IPs are wa Saga (10 km from Niamey). Its staff in tered from the Komadougou River and cludes a director general, an administra also from deep wells. In total, 13, 000 tive and financial director, a director of hectares are under irrigation. equipment t and maintenance, a director of operations, and two units which pro Rice is grown mainly along the river vide evaluation and management/control. under the regional services of Tillabery, Under the director of operations, there are Niamey, and Gaya. Multiple cropping is five regional offices located at Tillabery, done at Konni and Diffa, with such crops 161 Table 1. Irrigated Perimeters supported by ONAHA. Number Irrigated area Number of Number of Irrigated area of coops (ha) farmers farmer groups under crops (ha) Niamey regional office 16 3,605 9,242 74 Rice 3,551 Vegetables 54 Tillabery regional office 16 4,373 9,771 87 Rice 4.194 Vegetables 140 Others 39 Gaya regional office 4 672 840 28 Rice Konni* regional office 7 4,104 5,514 109 Multiple crops Diffa regional office 4 304 733 8 Multiple crops 208 Summer Rice 96 Grand Total 47 13,058 26,100 306 • Includes Maradi IPs Source: ONAHA 1994: presentation as wheat, onion, peppers in the dry sea Irrigated rice son, and cotton, sorghum, and millet dur ing the rainy season (Tables 2 and 3), In This area of production has five strong Niamey and Tillabery, 200 hectares are achievements in Niger: organization of devoted to vegetables, including green farmers into coops; two cropping seasons, beans which are grown for export, with an average yield of 4,5 t/ha; use of animal traction for land preparation; use Production of Improved Seeds of improved seeds and seed treatment; and use of high-level fertilization, The The program started with the two varieties grown, IR1529 (99% ofland SAADIA-A V AL seed farm in area) and BG90-2 on 380 hectares at 1984-1990 with funding from the Bel Saga, are susceptible to RYMV, In gian Agency for Cooperation (AGCD) to 1996-1997, the ONAHA seed project produce rice seed, This support has now collaborated with INRAN in the identifi expanded to cover sorghum, wheat, and cation of two new varieties (WITA8, cotton, The current phase covers the WITA) from the West African Rice De 1998-2000 period, but the project started velopment Association (WARDA) pro in 1995, Activities have included the gram that are resistant to RYMV, The two identification of adapted varieties, their new varieties are currently under seed in extension, and seed production. crease at the seed farm (breeder seed in the off-season, and foundation seed in the rainfed season during 1998) to increase seed of improved rice (Table 4), Produc- 162 Table 2. Surface area, yield and actual production of multiple crops during the rainy season under management of ONAHA regional offices (1994-1995 average). Area Yield Production Area in % (ha) (t/ha) (t) of total Cotton Konni regional office 1,605 1,819 2,920 42.5 Diffa regional office 0 0 0 0 Yield average 1.819 Total 1,605 2,920 42.5 Sorghum Konni regional office 1,411 1,769 2,495 37.4 Diffa regional office 0 0 0 0 Yield average 1.769 Total 1,411 2.495 37.4 Millet Konni regional office 703 1,129 794 18.6 Diffa regional office 0 0 0 Yield average 1.129 Total 703 794 18.6 Other (maize, groundnuts, cowpea) Konni regional office 52 1.4 Diffa regional office 3 0.1 Total 55 1.5 Grand Total 3,774 100 Table 3. Surface area, yield and actual production of multiple crops in the dry season under man agement ofthe ONAHA regional offices (1994-1995 average). Area Yield Production Area in % (ba) (1.Iha) (t) (of total) Wheat Konni regional office 879 3.08 2,705 82.0 Diffa regional office 13 1.69 22 1.2 Yield average 306 Total 892 2,727 83.2 Onion Konni regional office 116 38.4 4,457 10.8 Diffa regional office 4 22.5 90 0.4 Yield average Total 120 379 4,547 11.2 Green Peppers Konni regional office 0 0 0 0 . Diffa regional office 47 1.64 77 4.4 Yield average 164 Total 47 77 4.4 Other (tomatoes, peppers, cabbage) Konni regional office 11 1.0 Diffa regional office 2 02 Total 13 1.2 Grand Total 1,072 100 163 tion of rice seed is done atthe Saadia-A val 500 added to rent and service dues each seed farm, the farmer-seed producers' season) in a seed fund. The seed project coop responsible for primary production also contributes by providing advance who are also in charge of secondary seed money to the coops. The seed fund allows production at 35 other rice coops. This the following: a guarantee to farmers in farmer activity has been self-managed volved in seed production (30% objec since 1990 and has proven its viability. tive); advances for fertilizer and other The production and diffusion scheme de specific needs; -purchase of foundation vised in 1992 indicates seed replacement seed from INRAN (wheat and sorghum); every five seasons on all perimeters. Each paying for post-harvest seed handling, cooperative must then produce its own treatment, packaging, storage and diffu seeds for four seasons out of five. This is sion; and farmer compensation, particu done by farmer-seed producers in two larly neighboring farmers in sorghum hy seasons. A new seed production mecha brid seed production. nism is now in place whereby INRAN produces breeder seed and the seed farm The same farmers produce wheat and produces only the quantity needed by sorghum seed, which gives them the po farmer-seed producers. In addition, previ tential to become professional seed pro ous-year seed will be used instead of pre ducers. vious-season seed (Table 4). Approach Multiple crops Sorghum and wheat Production of improved seeds of sor ghum, wheat, and cotton is of concern. Multiplication of sorghum and wheat The program is conducted by three coop seed is carried out together with a compo eratives (Konni 1, Konni 2, and nent of adaptive research, which allows Djiratawa) where other farmers could get further changes in the seed program. The their seeds. This program, by the year main components are the following: de 2000, should cover 90% of the needs for termination by ONAHA technicians and sorghum and wheat seed, and 100% of farmers of the relative importance of all cotton seeds, which represents 30% per varieties grown on the IPs; identification year increase in seed production. The of seed blocks for sorghum hybrid seed unique characteristic of this program is production (farmer producers and their the farmer's financial participation (Fcfa neighbors); delivery and pre-financing of T a ble 4 P roce d ure ~or nee see dI pro d uctlOn. Maximum area Maximum production Producer Seed tvoe (ha) (t) INRAN Breeder 0.25 1.0 Seed farm Foundation 25 120 Farmers- seed producers Commercial 300 360 Farmers-non seed producers Grain 6900 - 164 T a ble 5. Wheat seed produced in dry season 1996 -1997. GrandTotal COQPerative Variety EI Madaoua* Variety EI koderawa** seed oroduced Area Yield Produced Area Yield Produced Konni I and II 0.74 ha 4.0 tlha 2960 kg - -- 2960 kg Diiratawa - - - 0.72 ha 4.03 tlha 2900 kiz 2900 kll Total 0.74 ha 4.0 tlha 2960 kl!. 0.72 ha 4.03 tlha 2900 kll 5860 kv • EI Madaoua ~ Yekora Rojo ** EI Koderawa ~ Jerusalem Table 6. Wheat seed produced during the dry season. Variety Variety Total Cooperative EI Madaoua EI Koderawa seed produced Konni I 7,000 kg 1,800 kg 8,800 kg Konni II 7,000 kg 3,600 kg 10,600 kg Djiratawa 6,300 kg 1,500 kg 7,800 kg Total 20,300 kg 6,900 kg 27.200 kg inputs (seeds, fertilizers, and Niger (SCN), which provides advance of phytosanitary products); pre-financing funds for inputs and assists with field land preparation; removal of off-types in management. One hectare of pure seed of seed fields; harvesting, transportation, variety STAMP from INRAB (National and threshing; seed packaging; germina Agricultural Research Institute of Benin), tion tests; diffusion (sale) of seed by the was planted, coop; and survey of planted area. Results and on-going activities It should be noted that seed production is handled directly by farmers or coops. Wheat The coops may pre-finance some activi ties using funds provided by the seed pro Seed produced in 1999 will cover the ject, but this would be recuperated from needs for the dry season which represents the seed cost. The project input is de 225 hectares or 30% of the wheat area on signed to be phased out in two years so IPs (800 hectare). In the trials, the variet that the coops can themselves fund a prof ies El Madaoua and El Koderawa are itable seed production. The ONAHA 20-25% higher yielding than other variet staff, who have been trained in hybrid Ies. seed production by INRAN, supervise all activities. In addition, INRAN provides Sorghum foundation seed each year. In 1997, three varieties were selected Cotton by farmers: SEPON82, 112 MSB, and NAD-l. Production of NAD-1 hybrid Cotton seed production began in 1998 seed was handled by one farmer in each of with the help ofthe Societe Cotonniere du the Konni I and Konni II coops, and two 165 neighboring farmers at Djiratawa. Pro requires staggered planting, and supple duction of SEPON82 seed was handled mental irrigation, and fertilizer. In addi by one farmer at Djiratawa and 112 MSB tion, to prevent pollen contamination, by one farmer at Konni 2. It should be neighboring fields must be planted with noted that hybrid seed production by the male parent. At flowering, off-types farmers is a first in Niger. NAD-I seed and shedders must be taken out of the production requires close monitoring and male and female rows, respectively. At special training. Care must be taken to en maturity, any panicle that is off-type is sure nick in the flowering of the two par also taken out of the field. Results of the ents. NAD-l female parent flowers about 1997 and 1998 seasons are shown in Ta a week earlier than the male parent, which bles 7-12. Table 7. Sorghum varieties grown on Konni and Maradi IPs in rainy season 1997. Area cultivated Konni I Konni I + Konni II Djiratawa Konni + Djiratawa Varie ha ha % ha % ha % Local varieties E1 Tsedawa 270 393 663 72 0 0 663 62 Janjare 88 87 175 19 0 0 175 16 Mota Rouge 0 0 0 0 38 26 38 4 Others 50 0 50 5 2 1 52 5 Total 408 480 888 96 40 27 928 87 INRAN recommended varieties YzMSB 0 0 0 0 57 39 57 5 Sepon 82 2 10 12 1 37 25 49 5 IRAT204 0,2 12 12 1 0 0 12 1 NAD-l 0 2 2 0.2 7 5 9 1 Total 2 24 26 2 101 69 127 12 Farmer seed roducers Diverse varieties 3 9 5 3 14 Total Diverse varieties 413 510 923 100 I 146 100 1.069 100 166 Table S. Summary of results from demonstration trials at Birni N'Konni and Djiratwa (Trials without midge attack. Grainxield Plant Panicles 50% 80% % of mota height par hill flowering maturity Variety tlha blanc (cm) (nbr) (ii) (ii) Local varieties EI Tsedawa 3.14 98 295 3.1 57 73 Janjare 2.93 92 328 3.0 62 80 Mota Blanc 3.19 400 302 3.3 72 90 Mota Rouge 2.81 88 287 3.1 51 80 INRAN varieties V,MSB 3.10 97 258 2.7 64 89 Sepon 82 3.95 124 176 3.1 66 90 lRAT204 2.61 82 159 3.0 60 84 90 SN7 2.82 88 244 2.89 63 87 NAD--1 4.20 132 191 3.1 66 89 Male parent ofNAD-1 hybrid MR 732 2,40 75 161 2.9 70 92 Number oftrials 6 6 6 5 6 Yield: CV = 28.2% Isd (0.05) = 1.02 tlha Table Sa. Results ofsorghum varietal survey: farmer choice of best {o(!en-(!ollinated} varieties. Konni (I and II) Variety Konni (l and II) Djiratawa and Djiratawa Total % Total % Total % Open pollinated variety EI Tsedawa 28 13 5 5 33 11 V,MSB 47 23 25 24 72 23 Sepon 82 18 9 25 24 43 14 90 SN 7 31 15 23 22 54 17 lRAT 204 19 9 0 0 19 6 Janjara 9 4 0 0 9 3 Mota blanc 10 5 0 0 10 3 Mota rouge 141 5 0 0 11 4 Hybrid sorghum NAD-I and its male parent NAD-l 32 15 26 25 58 19 MR 732 3 1 0 0 3 Total 208 100 104 100 312 100 NB: farmer sample size: Konni (I and II): 52, Djiratawa: 26 167 Table 8b. Results 0 f sorghum varietal survey: the best two varieties selected by farmer. Konni (I and II Variety Konni (I and II) Diiratawa and Diiratawar Total % Total % Total % El Tsedawa 10 10 4 8 14 9 Y2MSB 34 33 18 35 52 33 Sepon 82 36 35 17 33 53 34 90 SN7 20 19 13 24 33 21 1RAT 204 4 4 0 0 4 3 Total 104 100 52 100 156 100 NB: farmer sample size: Konni (I and II): 52, Djiratawa: 26. T able 9 Results of sorghum varietal survey: attributes preferred or disliked by armers. Variety Positive attribute Ne!!ative attribute El Tsedawa -earliness -low yield -midge resistance -poor storage capacity -wide adaptation to soils -bad cooking quality -!!ood taste Janjare -good storage capacity (germinates -poor yield after 5-6 years) -susceptible to lodging -good taste -susceptible to midge -gOOd_Ilroduct yield at cooking -food darkens after long conservation Mota blanc -good yield -susceptible to Striga -good tillering -difficult to thresh -good taste -susceptible to lodging and breakage -Door storage caoacity Mota rouge - earliness -low yield -good storage capacity -small panicles -good threshing -susceptible to long smut -!!ood taste Y2MSB -high yield -susceptible to midge -good germination potential -late maturing -!!ood food quality -!!rain shattering if drv in the field Sepon 82 -earliness -susceptible to grain molds -high yield -needs high soil moisture for germination -good food quality (good for tuwo and fura -good taste 1RAT 204 -high yield -hard grain: high soil moisture needed for germination -earliness -susceptible to Striga -good storage capacity -susceptible to grain molds -good taste 90 SN 7 -high yield -Striga resistant -good taste -!!ood oroduct yield at cooking NAD-l -high yield -need to renew seeds each season -wide adaptation to soils -Striga resistant -!!ood taste MR 732 -low yield -late maturity 168 Table 10. Results of sorghum varietal survey: farmer intent on seed purchase. Question: of which two varieties would the farmer request good quality seed next season. Konni (I + II) Variety KonniG+rn Oiiratawa and O'irata Total % Total % Total % EI Tsedawa 0 0 0 0 0 0 YzMSB 39 38 0 0 39 25 Sepon 82 21 20 10 19 31 20 90 SN 7 10 18 13 25 32 21 IRAT204 0 0 0 0 0 0 NAD-I 18 17 29 56 47 30 None 7 7 0 0 7 4 Total 104 100 52 100 156 100 NB: fanner sample size: Konni (I and II): 52; Djiratawa: 26. Table 11. Seed of ol!en I!ollinated varieties I!roduced {rain~ season 1997}. Seed block Vari!): Area (ha) Prod. (t) Yield (tlhal Konni II Y2 MSB 0.19 0.64 3.38 Ojiratawa Segon 82 0.32 0.77 2.40 Totauxlmoyenne 0.51 1.41 2.76 Table 12. Diffusion of iml!roved sorghum seeds during rain~ season 1998. Total Percent Area under sorhum Seed Area under sowing improved Cooperatives area Seed quaotity used improved seeds success seeds (ha) source (kg) (hal (%) (%) Forecasted Actual Sorghum hybrid NAD-l Konni I 420 Konni I 780 73.1 24.1 33 6 Konni2 593 Konni 2 835 78.3 47.6 61 Djiratawa 148 Djiratawa 322 25.8 8.6 33 6 Ibohamane 229 Konni I et2 200 16.7 10.9 65 5 MoulHa 6 Konni 2 20 1.7 1.4 82 23 Galmi 15 Konni2 100 8.3 4.9 59 33 Total 1410 2.257 203,9 97.5 48 7 Sorghum open pollinated variety Sepon 82 Konni I 420 Djiratawa 174 16.3 1.9 12 0.5 Konni2 593 Djiratawa et INRAN 312 29.2 23.0 79 4 Djiratawa 148 Djiratawa 199 15.9 15.0 94 10 Ibohamaoe 229 Djiratawa 170 14.2 12.3 37 5 Moulela 6 Djiratawa et INRAN 28 2.3 1.4 61 23 Galmi 15 Djiratawa et INRAN 85 7.1 2.4 34 16 Total 1410 968 85.0 56.0 66 4 Sorghum open pollinated variety Y, MSB Konni I 420 Konni2 220 20.6 1.5 7 0.4 Konni2 593 Konni 2 400 37.5 14 . .2 38 2.4 Autres 397 0 0 0 Total 1410 620 58.1 15.7 27 l.l All three varieties (NAD-l, Sellon 82 and Y, MSB) Total 1410 3.845 347.0 162.2 49 12 169 Conclusions • less cost for seed blocks; • better supervision from technical Production of hybrid seed is very risky staff; for fanners given the unpredictable start • better diffusion program; and of the growing season,which may impact • possibility to produce under con plant growth for both female and male tract for other institutions or pro parents. This activity is intenseand re jects. quires continuous field presence by farm ers and the ONAHA staff. Although the price of seed (650-800 Fcfa for a kg ofNAD-l) does not limit dif Overall, supervision of the program by fusion, Nigerien farmers are not used to the co-op has many advantages: purchasing seed every year. This is espe cially problematic, because seed supply is • a seed fund that guarantees sale of small, and seed may not be available for seed produced by individual farm replanting. ers; 170 Farm Level Profitability and Evolution of Input-Output Markets: Economic Perspective Tahirou Abdoulaye and J. H. Sanders Introduction ing the rainy season for both dryland (rainfed) and perimeter land (where sup New cultivars with high yield potential plemental irrigation is available). On the have been developed to increase agricul irrigated perimeter of Birni N'Konni, tural output and to improve farmers' in crops are grown during the rainy season come. Among those is NAD-1, the first with supplemental irrigation. sorghum hybrid released in Niger. Exper iment station and on-farm trials con This paper examines the effects of the ducted since 1990 show that NAD-1 pro economic environment (prices of inputs duced 4 to 5 times more grain per hectare and outputs) and the ability of the farmer than local varieties (Kapran, 1998). Ef to finance input purchases on the adoption forts are now being made to develop a decision ofthe new hybrid. These are con seed industry for this hybrid, based upon sidered to be critical conditions to ensure farmer seed production. A consistent seed profitability and thereby adoption. In the supply is necessary for a sustainable use following sections, we present an over of hybrid technology. For farmers to start view ofthe region, then the model, and, fi producing hybrid seed, assured demand nally, results and conclusions. for hybrid seed is necessary. So there is an interaction between the evolution of a lo Region and Background cal seed industry and adoption of the new cultivar. Here, we are concerned with the Sorghum is the second staple crop (af conditions under which the introduction ter millet) in Niger. In addition to the of the hybrid will be profitable to farmers grain, sorghum stalks are used for con and the constraints at the farm level to the struction and animal feed. Birni N'Konni rapid introduction of the hybrid and asso is representative of the low lands, where ciated technology. sorghum is concentrated in Niger!. In 1995, 12,000 tons of sorghum were pro Programming results showed that hy duced in the Birni N'Konni region. This brid sorghums under irrigation and inor represents about 5% oftotal sorghum pro ganic fertilizer use are more profitable duction in the country for the same year than local and improved sorghum variet (Republique du Niger, 1997). ies in Birni N'Konni (Ens ink, 1989). Un der rainfed conditions, farm level eco Population density in Birni N'Konni nomic evaluation of hybrid sorghum in has increased from 33 person/km2 in 1977 2 Niger has not been conducted. NAD-l to 48 personlkm in 1988 (Republique du with inorganic fertilizer is proposed dur- Niger, 1991). The farm households in the I With the exception of the Gaya region which has higher rainfall. 171 study area averaged 6 persons, with 3.9 In contrast to local sorghum, NAD-l hectares of dry land. The main crops offers impressive yield increases when grown include millet, sorghum, and cow combined with moderate fertilization in pea on dryland; cotton and sorghum on both rainfed and irrigated zones. On-farm perimeter land during the rainy season; trials showed that with fertilization levels and wheat during the off-season. Inor of 100 kg/ha of super simple phosphate ganic fertilizer is not applied on tradi and 50 kg/ha of urea, NAD-l yields are tional crops (millet, sorghum, cowpeas). 1500 kg on dryland and 2500 kg on irri However it is used on cotton and wheat gated perimeter, where supplemental irri grown on the irrigated perimeter. gation is available (Kapran, 1998). One of the largest irrigated perimeters Constraints to sorghum production in in Niger where sorghum production is this area include inadequate and irregular promoted is also located in this region. Ir rainfall, poor soil fertility, low and vari rigated land on publicly managed perime able output prices and cash availability ters is allocated to farmers from surround for input purchase. In Bimi N'Konni, wa ing villages. The average area per house ter quantity in the reservoir depends on hold on the irrigated perimeter is 0.75 rainfall, so that in bad rainfall years, water hectares (Abdoulaye, 1995). is insufficient to irrigate the whole perim eter. Soil fertility especially deficiency in On drylands, sorghum is intercropped phosphorous and nitrogen has been a ma with millet and/or cowpea. It is also jor problem for agriculture in Niger grown as a sole crop on lowlands with (NAARP,1992). heavier soils. Sorghum yields are gener ally low on the drylands because of low On the economic side, the main con rainfall, use of traditional cultivars and straint has been low output prices relative poor soil fertility. For Bimi N'Konni the to input prices, which prevent farmers 1991-1995 average yields were from purchasing the inputs necessary for 370-kg/ha (Republique du Niger, 1997). increased production. Formal credit sys Almost no purchased inputs are used in tems do not exist in most rural areas in the production process, with the excep Niger. There are several ways farmers can tion of hired labor for weeding. get the funds for input purchases: off-season work in the city by members of On the irrigated perimeter, sorghum is the household, livestock sales, non-farm grown with supplemental irrigation in ro activities, or work on another farm while tation with cotton during the rainy season. still living on the farm, and informal Sorghum yields here are relatively higher credit systems (Abdoulaye et aI., 1991, than on dryland due to the better availabil Reardon et aI., 1994). Non-farm activi ity of soil moisture and residual fertilizer ties generally are the most important from the cotton crop. In 1990, local sor source of income. Non-farm activities are ghum yields on the irrigated perimeter very common as many trade opportunities were estimated at 690 kg/ha exist in the region due to proximity to Ni (Lowenberg-DeBoer, 1990). geria. 172 Methods limits the land available to 4.65 hectares, of which 3.9 hectares is on dryland. The general framework for this analy sis is a linear programming model, which Hybrids require investment in pur maximizes farm income subject to re chased inputs. To determine available source constraints. Single activity evalua cash, average investment in crops includ tion, such as partial budgeting, does not ing seeds, fertilizers, and hired labor was account for the economic context under used. Survey data from Bimi N'Konni, which production takes place. There are indicated that the average investment was important system interactions in farm de 7,000 Fcfalha (1$ U.S. = 272 FCFA in cision-making. Programming is suitable 1989,IMF 1997). When the sample was for farm level analysis, because it incor restricted to farmers with irrigated perim porates these and considers constraints of eter land, that average was 13,000 F cfalha available farm resources. Farm program (Abdoulaye, 1995). In this model, we as ming models enable scientists to project sumed that 65,000 Fcfa (13,000 x 5 hect potential farm income and to identify ares) of cash was available at the begin farm-level constraints delaying adoption. ning of the cropping season for invest ment in crops. With sensitivity analysis, An important aspect of Nigerien agri we varied the amount of capital to evalu culture not included here is risk associ ate the effect of cash availability on the ated with agricultural production. Risk is adoption decision. an important determinant of new technol ogy adoption in rainfed systems, espe The technology being evaluated is cially if these technologies involve high NAD-l with associated practices. In the use of purchase inputs, such as inorganic model, a hectare ofNAD-l requires 10 kg fertilizers. Given the importance ofrisk in of hybrid seed along with 50 kg of urea this type of economic environment, future (46% N) and 100 kg NPK (15-15-15). modeling work on NAD-1 will include NAD-l technology is introduced in the risk as more data on the distribution of model as a sole crop on both dryland and crop yields becomes available. irrigated perimeter and in rotation with cotton on the perimeter. The model determines optimal land and resource allocation among cropping Breeders' estimates of NAD-l yields activities. Crop production activities on are discounted by 35% and 40% on dry land include sole crop (millet and sor dryland and irrigated land, respectively, ghum) and intercrop (millet/cowpea, mil in the modeling. These values are given in let/sorghum, and millet/sorghum/cow Table 1. Differences in farm management pea). Activities on the irrigated perimeter account for this. Usually, under trial set include sorghum-cotton in the rainy sea tings, farmers devote more attention to son and wheat for the dry season. Three crops than they do normally. Still, these principal constraints (land, labor and cap yields for the new hybrid and associated ital) are imposed in the model. Land con practices are much higher than presently straint (dryland and irrigated perimeter) observed among cereal yields on either dryland or irrigated areas. 173 Data for this analysis come from sur dominant cropping system (Lowenberg veys conducted in the area by Institut Na Deboer, 1990). tional pour la Recherche Agronomique du Niger (INRAN) (farm resources and crop When NAD-l is introduced in the sys yields), National Statistics (crop yields tem, land is allocated to it mainly because and prices) and Systeme d'information of the increased yields. According to the sur les Marches (SIM) (crop prices). In- model results, if farmers knew about the technology and could get the seeds and Table 1. NAD-l grain yields assumed in the fertilizer, then they would adopt it. model. Farmers plant NAD-l on all their irri Perimeter gated land (0.375 hectare) replacing local NAD-1 1500 sorghum cultivar. Investment in irrigated land has a higher return than investing in dryland area. But even on the drylands, 0.55 hectare of dryland is allocated to the put and output prices are summarized in new hybrid sorghum (Table 3). Table 2. Fertilizer prices also include the cost of search and transportation. These Another reduction in input costs were assumed to be 10 Fcfa/kg or 500 can be achieved by increasing the effi Fcfa per 50-kg bag. Sensitivity of model ciency ofthe seed industry. Currently hy results to a lower level of transportation brid seed price at the seed co-op is about cost, (5 Fcfalkg), and to higher costs (20 seven times the price of grain or 650 Fcfa/kg) was also determined. Fcfa/kg (House, 1997). The Co-op pro cures seeds from seed producers for 350 Results Fcfa/kg. With increased efficiency of the seed industry, hybrid seed prices are ex Before the introduction of NAD-l, pected to decline. A decrease in seed price land is allocated to millet/cowpea to 350 Fcfa/kg leads to an increase of27% intercropped and millet/sorghUm/cowpea in dryland area allocated to NAD-l. intercropped and to sorghum cotton rota tion and wheat in the off-season accord On the other hand, if infrastructures ing to model results. These results are worsen so that search and transportation consistent with surveys from the region, costs of fertilizer increase to 20 Fcfalkg, which showed that intercropping is the the required investment for NAD-l in creases, and its profitability is reduced. NAD Seed Price k -1 650 Source: 1. SIM for sorghum (average of Nov 95,96, and Oct 97), for Millet (average of Sept 95,96, 97), for cowpea (average of Oct 95.96,97). Source 2: Centrale d 'approvisionnement for fertilizer (official price + transport cost, 1997/1998) Source 3: House (1997) for NAD seed price 174 T a ble 3 L an d area IIoca t e d t o crOPI~In2 ac f'fIVI les an d'Incomee f~ec t s In . mo d e I resu Its. Without WithNAD-l Activities NAD-l Base Model Increase in liquidity Output price increase l. Dryland - Millet/Cowpea 3.5 ha 2.949 ha 2.261 ha 3.321 ha - Millet/Sorghum/Cowpea 0.4 ha 0.4 ha 0.4 ha - -NAD-l na2 0.551 ha 1.239 0.579 ha 2. Perimeter land 0.75 ha - _ Local Sorghum-Cotton I na - - - NAD-1- Cotton I 0.375 0.75 ha 0.75 ha 0.75 ha Wheat (off-season) ha 0.375 ha 0.375 ha 0.375 ha 3. Total NAD-l area na 0.926 ha 1.614 ha 0.954 ha 4. Farm Income Increase na 6% 7% 16% Note: I. Rotation modeled as 50% of the area for each crop. 2. Na: not available. Consequently, dryland area allocated to show that dryland area allocated to hybrid NAD-l is reduced by 19%. Basically, all technology on dry land increases by 5%. cost-reducing situations lead to an in crease in the dryland area of adoption of Thus, reducing the effects of price col hybrid sorghum. Given the current eco lapse has a short-term effect on the nomic situation in Niger, it is very un dryland area. For the long-term, the farm likely that the government will engage in income increase would allow capital ac subsidy programs. In the long-run, there cumulation over time as well as increase will be improvement in the country's in future investments in purchasing more frastructure. In the short-run (next de fertilizer and seeds. cade), however, it is probably better to concentrate on improving prices received The liquidity available for investment by farmers to stimulate the diffusion of in agriculture is another major constraint. hybrid sorghum and associated improved The importance of liquidity was tested practices. through an increase in cash available for investment in agriculture at the beginning On the output side, improvement in the of the cropping season from 65,000 Fcfa economic environment may lead to to 85,000 Fcfa (1$ U.S. = 583.67 Fcfa, higher output prices for farmers. Cereal IMF 1998). This 20,000 Fcfa increase in prices in Niger are very low at harvest cash availability corresponds to the sale time. However, prices rise gradually af of a sheep or an increase in non-farm in ter. In Bimi N'Konni, the average sor come. Moreover, farmers often engage in ghum price was 116 Fcfa/kg (May off-farm activities and receive remit 1995-1997) compared to 89 Fcfa/kg at tances, which can be used to increase cap harvest (Computed from OPVN data, ital availability to the household. This in 1995-1997). If we assume storage of the crease in liquidity leads to more than dou hybrid until Mayor if hybrid grain can re bling dryland area allocated to hybrid ceive a price premium, for example, for sorghum (Table 3). better processing quality, model results 175 Cash availability determines how quidity available (Levels 1 and 2) This li much the farmer can afford to invest in in quidity available for investment is set to put purchases. The quantity of purchased 65,000 Fcfa for Levelland 85,000 Fcfa inputs in tum determines the area, which for Level 2 as explained previously. will be allocated to hybrid sorghum. These levels could also represent farmers However, if the yield levels in the model with different resource availability in the are possible, farmers would adopt. In that same region. case, the diffusion process would acceler ate as adopting farmers would accumulate If this price ratio is changed from 3 to capital for higher input purchases the fol 2.5, this corresponds to a 19% increase in lowing year. Their neighbors would see output price (for fixed input costs). In or the possible gains and imitate them. der for the relative price ratio to fall to 2, given fixed input costs, output prices will The combined effect of increased out have to increase by 48% from their cur put price and higher liquidity are analyzed rent levels. This level of increase is not by decreasing input-output price ratio for unusual in Niger, where price increases of two levels of liquidity (Figure 1). The this magnitude or more are frequently ob weighted2 costs of purchased inputs served between harvest time and the next (seed, fertilizers) are divided by the price planting period. However, it is at harvest of sorghum to obtain the input-output time (when prices are low) that most price ratio. Change in input/output ratio farmers sell their output. was considered for the two levels of li- :g 3.5 0.. Le\€11 Le\€1 2 g-:::J 3 5. 2.5 c: ;::.. 0 2 :;::; co 0:: 1.5 CD () ·c 1 a.. 0 1 2 3 4 5 NAD-1 area (ha) Figure 1. Simulation results of hybrid sorghum area per farm as a function of relative input price for two levels of liquidity availability. 2weighted by the relative importance of both expenditures by farmers. 176 These downward-sloping relationships (storage) or increased demand for sor between area devoted to NAD-l and the ghum through research on new uses (pro improved economic environment (with a cessing, industry) can give farmers better higher output price for sorghum) indicate prices for their outputs. Storage opportu the importance of storage and other pro nity costs may be very high for most farm grams in order to increase the demand for ers, if they are pressed to repay loans or to sorghum as part of organized efforts to make urgent purchases immediately after more rapidly introduce NAD-l into the harvest. Nigerien agriculture. The response curve shifts to the right with more capital avail Decision-makers often fear increases ability. Therefore, response to improved in basic food prices including cereals be economic environment is larger when cause of the burden this imposes on urban more initial capital is available to the consumers, however, in the long run household (Figure 1). Finding ways to in prices decline because production costs crease the cash available for purchasing per output unit decreases with technologi inputs also accelerates the introduction of cal change thus, both farmers and con NAD-l. This is more difficult since pub sumers will benefit. Because ofthe power lic credit programs are unlikely; however, of urban consumers, decision-makers of over time, farmers will develop financing ten adopt policy measures that reduce the mechanisms with the fertilizer dealers. prices of cereals. This type of action hin ders the introduction of technological Conclusions change in agriculture. Policymakers need to recognize the importance of a favorable Hybrid sorghum technology is profit economic environment for agriculture to able and expected to be adopted by farm hasten farm level-investment and techno ers on rainfed and irrigated areas. Im logical change. provements in economic conditions will increase diffusion of hybrid sorghum For diffusion of hybrid technology to technology. In years with good cereal be sustained, input markets need to be production, output prices collapse and well-developed. Farmers often complain without financing, farmers frequently feel about the availability of necessary inputs. pressure to sell at low harvest prices to re Farmers need access to fertilizers and pay debts or to obtain cash. So, strategies seed for production of NAD-l. The pri to offset seasonal price collapses after vate marketing sector needs to make seed harvest and to moderate the price declines and fertilizers available at the village resulting from good weather or improve level. ment in input markets (seed and fertiliz ers) are critical. Improvements in the in References frastructure also can have similar effects. Abdoulaye, Tahirou (1995), "Impact of Opportu nity Cost of Capital on Adoption of New Tech Accelerating the adoption of NAD-l nologies in the Birni N'Konni region of Niger", can be achieved by increasing the output M.Sc thesis, Department Of Agricultural Eco price farmers receive. Better access to nomics, Purdue University, West Lafayette, IN, USA. markets, farmer management practices 177 Abdoulaye, T., J. Lowenberg-DeBoer, K.C. NAARP (Niger Applied Agricultural Research Reddy, and P. Abbott (1991), "Les Contraintes Project), (1992), Final Technical Report, part Socio-Economiques au niveau de II, Purdue University, Alabama A&M Univer I'Exploitation pour I'Intensification de la Cul sity, Winrock International and the University ture de Niebe au Niger," INRANIDECOR, Doc. of Nebraska. 24F. OPVN (office des Produits Vivriers du Niger), Ensink, Peter (1989), "Sorghum technology As SIM, Marches des cereales et du Niebe, Publi sessment for Southern Niger Using a Target cations Mensuelles, Janvier 1995- Octobre MOTAD Approach," M.Sc thesis, Department 1997. of Agricultural Economics, Purdue University, Reardon, Thomas, Eric Crawford, and Valerie West Lafayette, IN, USA. Kelly (1994), "Links between Nonfarm Income House, Leland (1997).Trip Report, Production of and Farm Investment in African Households: hybrid seed, Oct-Nov, 1997. Adding the Capital Market Perspective, Ameri IMF (1998), "International Financial Statistics", can Journal of Agricultural Economics, 76 (De July 1998. cember 1994): 1172-1776. IMF (1997), "International Financial Statistics", Republique du Niger (1991), "Annuaire September 1997. Statistique: Series Longues", Ministere du Plan, Kapran, Issoufou (1998), Personnal Communica Edition 1991. tion Republique du Niger (1997), "Annuaire des Lowenberg-DeBoer J. (1990), "Rapport sur Statistiques de I'agriculture et de l'elevage," I'Enquete-Diagnostic dans al zone de Birni Ministere de l' Agriculture et de l' elevage, N'Konni, Campagne 89," INRAN/DECOR. Janvier 1997. 178 Regional Needs for Training in Seed Production N.D. Yagoua and N. I. Jika Introduction Training Seed is one of the most important agri Breeders cultural inputs. As such, it must be avail able in adequate quantity and quality, and They need to learn the techniques of at the right time to those who need it. hybrid variety selection. This includes pa rental identification and maintenance as Current Status in West well as hybrid seed production. A breeder and Central Africa must be familiar with certification re quirements for different types of seeds In most countries, research efforts have (breeder, foundation, commercial). led to the release of improved varieties Breeders in this region should aim at hy that are wanted by farmers. However, the brid seed production in collaboration with results of collaborative activities at the farmers. Finally breeders must be able to national, regional, and international lev train others, who will eventually become els do not quite meet the needs of farmers trainers. and food processors in terms of seeds. Technicians Seed production and distribution by the public sector has usually been too expen In addition to trial management, tech sive and non viable. This was primarily nicians must know the requirements for a due to the lack of an organized seed sec seed production field, recognize the dif tor. A sustainable seed activity will be ference between parents and hybrids, based on the following: off-types, and regulations for seed pro cessing. Technicians should be able to • concerted activity involving all train farmers in seed production. partners from breeders to end users; • sufficient number of well-trained Farmers contributors at each level. As the backbone for a sustainable seed The above is valid for open pollinated activity, farmers must be able to produce varieties but even more so for hybrids. seeds that have the required qualities. For this, they must: know the criteria for a seed production field; distinguish be tween parents, hybrids, and off-types; know the needs for quality harvest and storage. 179 A farmer that follows these rules will Regional training should be organized produce quality seed and have a competi by various networks including the sor tive price on it. ghum network (ROCARS) and the millet network (ROCAFREMI) in West and Suggested Process Central Africa, and the SMIP in East Af rica. Some countries have benefited from collaboration between national, regional References and international organizations and ROCAFREMI workshop on the seed issue, started a well functioning seed activity Niamey, Niger 26-29 August 1996. (Nigeria, Niger, Burkina). These coun Food Grain production in semi-arid Africa. tries have accumulated a wealth ofex peri OAU/STRC-SAFGRAD publication (Interna tional Drought Symposium on Food Grain Pro ence in the areas of hybrid synthesis, pa duction in the Semi-Arid Regions of rental identification and maintenance by Sub-Saharian Africa. Nairobi, Kenya 19-23 breeders, training farmers and technicians 1986). Seminar-Workshop on the reviving of sorghum re in hybrid seed production. The region search in West and Central Africa. should capitalize on these results and IERlICRISAT/INSAH. Bamako, Mali 21-25 make training possible for other coun February 1994. Strategic Plan of the West and Central Africa Sor tries. There should be specific training ghum Research Network, 1998-2003. sessions for breeders, technicians, and WCASRNIUSAID/INTSORMILlICRISAT/C farmers. lRAD. June 1998. Reports of ROCAFREMI planning meetings. 1997. There should also be regional training in seed management (private enterprises, cooperatives). 180 Training for a Future Seed Industry in Niger I. Kapran, D.S. Murty, and L.R. House Introduction The availability of the sorghum hybrid, irrigated perimeters near Birni NAD-1, and farmer interest in it has laid N'Konni.The focus of the one-day train the foundation for the establishment of a ing session was on seed production, and a seed industry in Niger. The interest in training manual was developed. This seeds is not new to the region, and there training was appreciated, and its effec are some successful seed activities in the tiveness was evident, especially during country, particularly with rice and vegeta the growing season. ble seeds. However, the idea to develop a seed industry in the private sector to ad In 1998, four similar training programs dress the issue of seeds for millions of were conducted near Niamey, Say, Birni hectares has its own uniqueness and re N'Konni, and Maradi. Farmers and quirements. There are many components INRAN staff participated in these training to consider in this effort: seed production, programs. In instances where the owner processing, quality control, marketing, was not the farmer, an effort was made to credit and financial considerations. All include the farm managers. Again, the these aspects have their own requirements training was well received and the de that must be met if they are to be effec mand for the training manual high. More tively carried out. For most ofthe people than 250 individuals benefited from this who will be involved, these are new expe training. riences, and the opportunity to gain these experiences is complicated by the lack of A few things have been learned to infrastructure in the country. Education strengthen the training: and training are the means of providing opportunities to learn the necessary skills. • The training should not only cover the processes, but also the reasons Background behind those processes. • The program should include more To date, the effort to produce seed of training on crop management and NAD-1 in the private sector has met with pest control. In the future an agron modest success and is encouraging. As omist, pathologist, and an entomol private producers really began to be in ogist should participate in the train volved, a training program was begun in ing. Niamey in 1997. This was followed by a • Training should not only be in the classroom, but should also be con second training activity for interested ducted in the field during the grow farmers, the staff of Office Nation et des ing season. These could be one-day Amenagements Hydro-Agricoles training events given at four or five (ONAHA) working via the co-ops on the locations in order to reach many 181 farmers. Farmer training could take It has been recognized for some time place during the growing season. that a training center would provide a fo • Training should be given in the 10- cus and a facility for a training activity. cal language. This was much appre The Lossa station, some 70 kilometers ciated by the participants who re West of Niamey, is a possible location quested that the training manual be since it already has a seed processing unit. translated into local languages, and The period of training would increase to specifically Hausa. permit hands-on experience with process • The training should lead to consid ing as well as greater exposure to crop erable discussion among farmers who produce seed. This was the management and pest control. Training of case as they were able to contribute inspectors in seed certification and seed from their experiences. law would also be important, requiring exposure to production and processing In the beginning, INRAN technicians activities. Training for market develop were able to assist producers on an indi ment would be crucial as well. Having vidual basis; however, the number of in consultants to help formulate these pro dividuals involved now (and the fact that grams and develop the training compo they are spread across the southern por nent would be very beneficial. tion ofthe country) makes it impossible to deal individually with producers. As a re The training center should have a mod sult, training becomes an important est relevant library and include informa means of interaction. Focus on hybrid tion on farming, irrigation, and seed pro seeds so far has been the identification of cessing equipment. This would contribute individuals interested in seed production to making the training center useful to es and the stabilization of production. The tablished seed producers who are seeking training activity to date, therefore, has fo information. cused only on considerations relevant to seed production. INRAN is developing a seed unit, and a strong training component should be part The Future of it. As the existing input expands, a training officer will be required to man Several seed producers are already age both the array of training topics and a commenting on the importance of seed training facility. The training center could quality, which implies a capability to pro not only serve Niger, but also other West cess seed with proper treatment, packag African countries. ing, and quality control via a seed certifi cation program and seed law. In 1997, 7.5 Conclusion tons ofNAD-l was produced. It is antici pated that this will increase in 1998 and Training in seed production has taken 1999. The situation is expanding so that place for the past two years, and some 200 reasonably soon these functions should people have participated. A training man begin. Training will be important to effec ual has been developed. The training has tively carry them out. been well received with useful comments for expanding topic coverage. Effective- 182 ness of training has been realized while including training during a full crop sea working with individuals in the field. The son. It is doubtful that many interested in number of individuals involved has in seed production in Niger can afford this creased to the point where training is the much time. If a training center is estab most feasible-and, in some cases, the lished, it would be possible to make se only way to interact with producers. quential sowings so that planting, seed ling management (thinning, transplant Looking to the future, there is need to ing, fertilization, weeding), cross expand training to include seed process pollination techniques, harvesting, and ing, quality control, and marketing. In ad seed processing could all be part of a dition, there is a need to have a center short-term training activity. In a where relevant reference material is avail three-week training program, aspects of able to any interested producer. Having crop management and seed processing consultants can also help strengthen the could be considered. There is a place for training base. It is proposed that there be a the one-day course for those who cannot training center that is regional in scope afford to set aside more time. However, and can be managed by a member of during the off-season it may be possible to INRAN's Seed Unit. more easily get people for a longer period of time. Additional Thoughts on Training If this training opportunity would be Six training programs have now been available, a reasonable assumption is to conducted in which approximately 200 have 15 people for the three-week period people have participated. A training man and twenty for the one-day period. If ual has been organized that is being ex funding is sufficient, training could be panded. The training has been 1 or 1.5 done twice (three-week program) and days in duration and has been based on three to five times (one-day program) an follow-up observations in the field. How nually. ever, these observations have also indi cated shortfalls, primarily practical expe To manage this level of activity, it rience in managing a seed crop. would be necessary to have a training of ficer, and a budget would need to be de For years, ICRISAT has conducted a veloped. six-month training program at its center, 183 Rapporteur's Report I. Kapran Bases for Successful Hybrid Enterprises Day 2 - Session 1 Presentations Eight presentations were made cover prise is one with dependable seed supply, ing topics from seed production tech acceptable quality/purity, and improved niques to training needs and economics of cultivars. seed production. Naroua's presentation of seed produc Souley presented the field manage tion on irrigated perimeters of Niger indi ment and seed production techniques im cated that rice seed has been regularly plemented for optimum production of produced on 25 hectares since 1984, and NAD-1 seed in Niger. Specific instruc only recently they have started with hy tions were given to beginning seed pro brid sorghum. Individual farmers and co ducers, which they must follow closely if operatives contribute to a start-up fund they are to profit from this activity. and decide on seed production and mar keting procedures. To date, improved va Sirifi discussed cultural practices ap rieties (including SEPON82 and NAD-1) plied to hybrid sorghum. Results of agro occupy 12% of the irrigated perimeters. nomic trials showed a linear response of Farmers' seed production yields varied yield with N fertilizer: an increase in yield between 1.5 t/ha and 2 t/ha, with one due to supplemental irrigation. NAD-1 farmer producing 2.4 t/ha ofNAD-1 seed. responded better to N application and had higher yields than the improved, Salifou described his experience with open-pollinated variety SEPON82 on seed production in the public sector: dur both sandy and heavy soils. Recommen ing the USAID-funded Projet Cerealier dations were presented for critical fertility National du Niger. One major problem and supplemental irrigation levels. was that demand was poorly assessed. As a private producer, Salifou learned to Maunder discussed the importance of know the market well and sees the impor seed processing and quality control, in tance of technical support. The presenter cluding need to produce the best possible recognized the need to continue with re foundation seed while considering the search, urged quality control, and favored breeder's responsibility, harvest instruc the training of a select group of seed pro tions, and availability of units for seed ducers, who would constitute an elite cleaning. Overall, a successful seed enter- group of competitors. 184 Tahirou presented hypotheses on re for groundnuts, where yields are higher. turns to fanners who grow hybrid variet For NAD-l under irrigation, the cost was ies. He concluded that NAD-l can be CF A 380 per kilogram of seed produced. adopted, but he said there is need for good It was later sold at CF A 750 per kilogram. agricultural policy if fanners are to bene fit from their activity. Ndoye (to Tahirou): Is there a possi bility that the price of millet could ever be Yagoua stressed the urgency of re as low as CF A 60 per kilogram under con gional training in seed production for ditions of higher productivity (3 times to breeders, technicians, and fanners. day's yields)? House described his experience in Tahirou: Yes, that would be the Niger with fonnal training of technicians long-tenn objective of technology adop and fanners in hybrid seed production and tion. However, in the beginning, rela strongly supported its continuation. tively high prices are needed to support the process. Questions and Comments Following the Presentations Gupta (comment): For the adoption of hybrids, we should also consider cur Andrews (to Tahirou): What solution rent fanning systems. In Nigeria and do economists have for price fluctuation? Niger, the majority of the fanners grow sorghum and millet together - Tahirou: Economists do not necessar intercropped with either cowpea or ily have the solution. Possibilities include groundnut or both. There will always be price-fixing (although difficult to imple volunteers in hybrid fields. There is a ment in Niger), or improving value of need for education. Fanners grow many crops by use of new technologies (eg. crops to minimize the risk. Hybrids couscous). should be evaluated under intercropping also. De Vries (to Salifou): Who are your clients (wealthy or small fanners)? Andrews (comment): 1. To Tahirou: For many years, the biological scientists Salifou: I was surprised that it was the have produced good yield-producing small fanners that came first to buy seed technologies (new varieties, hybrids, fer ofNAD-I. I have also had NGOs, and de tilizers, cultivators, etc). However, the velopment projects buy seed. Fanners main blockade to their adoption is the will buy seed; the price is not the biggest fluctuation and weakness of the cereal issue. grain market. The market economists have known of this problem for many Jada (to Salifou): How much does it years. What is their solution? (Expanding cost to produce on one hectare? end-use possibilities is interesting, but it should not be considered the main solu Salifou: CF A 150,000 to 180,000 for tion). 2. Reference to Gupta's comment millet-but more for cowpea-and less about Nigerian production systems: I am 185 sure (I have had some experience with appear in Niger. This observation com intercropping in Nigeria.) that improved bined the concerns of S.C. Gupta and clar cereal genotypes of any kind can be effec ifications by D.J. Andrews on the tively used (perhaps more cost effec usefulness of hybrid sorghum by tively) in intercropping. Perhaps some ad small-scale farmers. However, the useful justments need to be made (e.g., spacing), ness and impact of hybrid sorghum culti but farmers are experts at making these vation will be better accomplished if the adjustments. hybrids target particular end uses (food, feed, malt, or a combination of any of Obilana (comment): I believe there these end uses). The hybrids should, are opportunities for the use of hybrids of therefore, be developed and released with sorghum in both the small-scale mixed particular and specific quality traits: for farming system (where sorghum and productivity, high-milling quality, good pearl millet are cropped together with feedstock quality, or high diastatic power groundnut and cowpeas) and the mono (for good malting in brewing and malt culture farming and single cropping sys food/drink processing). Increased use and tems in large-scale agriculture. These sit production of hybrid sorghums will result uations exist in Nigeria, and are starting to in more cultivar-specific cultivation. 186 Tuesday - September 29 - Session 2 Marketing by Commercial Firms and Cooperatives Opening Remarks Moderator - Kore Harouna, Fac. Agro/AMD University, Niamey Role of the Public and Private Sectors in Seed Production Ismael Mouddour Introduction Historically, modem seed production ested probably because seed had no in Niger was launched by the National greater value than grain and the seed sec Cereals Project in 1975 following the tor was not organized. 1968-73 drought. The seed industry is vir tually unknown to businessmen in Niger, Role of the Private Sector for reasons that include the following: in Seed Production • seed and grain are not differentiated The idea of a seed industry implies an by users of one or the other; economic approach to seed production. • presently seed activity is not well Therefore, as with any enterprise, seed structured to attract investment; production must follow some guidelines • seed producers have no guarantee in order to be profitable and attractive to for a market and price that is attrac private investors. These include: tive. • who does what and what is the role The Government of Niger thought of of the state? initiating a seed industry by investing • what demand is there for seed? heavily in one national seed farm and five • what quantity of seed is available? regional seed centers, designed to in • what seed should be produced and crease and distribute the best seeds from at what cost? research to farmers. Unfortunately the • what kind of marketing is appropri cost of production of MI and M2 seeds ate? was much higher than the sale price to farmers. This system quickly showed its The most important role of the private limitations and was abandoned. The pub sector is to be able to produce and sell "a lic sector was responsible at each step of quality product at the right time in the seed production, including research on right place, in the right form and price", improved cultivars, seed increase, seed the consumer being the one deciding what processing, packaging and storage, seed is 'right'. In terms of seeds, right is that marketing, training of technicians and seed produced by research that has the re farmers, seed law and regulations, price quired attributes. The private sector has fixing, evaluation of needs and planning an important role here too in new varietal seed increases. A small number of farm research and funding to research centers. ers (farmers-producers) located in the vi In summary, in the current situation of cinity of seed centers were contracted to ~iger, the private sector may have a role produce but there was no real involve Ill: ment of private investors. As suggested above, private investors were not inter- 189 • increasing seed of varieties pro ing, the seed store must be as near as pos vided by research; sible to vegetable growers. • seed marketing; • extension (demonstration plots). Vegetable seeds are mostly imported from Europe; however there is also local Private seed producers should organize production for some crops including on themselves in a professional association ions and amaranth. that could be the ideal partner of the pub lic sector. (Editor's note: The private seed Role of the Public Sector producers in Niger did form a profes sional seed association in early 1999. The experience of the National Cereals INRAN will be an honorary member.) Project indicated that cost of production was too high for seed production in Marketing of Vegetable state-owned farms. The public sector Seeds in Niger should pull out to allow the emergence of the private sector. The role of the public In Niger, large-scale production of sector would be as follows: vegetables followed the 1984 bad year for cereal production. The idea was to com • abquality control and setting up plement food production and help cattle rules for seed production and pro owners settle around water ponds after cessing. There is no efficient pro they lost their animals to drought. With duction system without quality con time farmers became interested in vegeta trol. Since control will differentiate ble production especially since vegeta the good from the bad, this will cre bles had become a regular part ofthe diet. ate a momentum for production of Indeed professional vegetable producers quality seeds. around large cities could now make a • Role ofINRAN. As a public sector good living from this activity. Onion pro agency involved in genetic im ducers are now able to export their prod provement of crops, the Institute uct to other countries. will have important activities in cluding development of new variet However, vegetable production is not ies with superior performance in without problems, chiefly in the produc comparison with farmers' varieties; tion ofquality seeds. Farmers tend to keep production of breeder seed; planting the same grain as seed leading to germplasm conservation; develop the loss of purity. Since they were used to ment and distribution of materials free starter seed by the state or in some describing the various varieties. cases from emergency relief programs. • Security stocks. The public sector could set up seed security stocks so The private sector is attempting to pro that in bad years farmer producers vide quality seeds to vegetable growers could rely on it as their source of through companies like Agrimex which seed. In tum they will increase the was created in 1993. Agrimex has three seed to cover national needs the fol retail stores in Niamey, four in other cities lowing year. and a network of small retailers. This is important because for vegetable market- 190 Constraints and Suggested Solutions in the West African Seed Sector A. Joshua Introduction do this (as it has done elsewhere) but for seed enterprises to be commercially suc Generally, though there have been cessfully, a change in government atti many attempts by the public sectors or tude and legislation is needed to allow the parastatal bodies, supply of improved private sector to operate profitably. seeds has been inadequate and unsus tained in the Economic Community of Governments have generally not real West African States (ECOWAS) region. ized that the best service they can do for The ECOWAS region is composed of the farmers is to foster the private sector to following states: Benin, Burkina Faso, provide the farming community with in Cape Verde, Cote d'Ivoire, The Gambia, puts that make farming profitable - not Ghana, Guinea, Guinea-Bissau, Liberia, only quality seeds, but fertilizers and Mali, Mauritania, Niger, Nigeria, Sene agrochemicals. Concerns are often cited gal, Sierra Leone, and Togo. Government about exploitation of farmers - but the regulations, subsidies and seed pricing principal goal of a private seed company controls have inhibited emergence of the (or a self-financing parastatal institution) private seed sector. is sustainability, and this can only be achieved by providing farmers with seeds Additionally government breeding and that are profitable for them. Farmers are variety testing programs have been poorly the ultimate judges of seed companies, funded, and farmers have not been in and only those that satisfy them are suc volved in variety/hybrid development. cessful. Releases have been slow and seed avail ability and quality have been low. The General Constraints of Improved source of most of the improved varieties Seed Production and Marketing in the has been the IARCs. ECOWAS/Central African Countries In many countries there is neither a Na • Limited funding for seed improve ment programs. tional Seed Policy or a coordinated seed • Inadequate infrastructures. system, so there are no clear certification • Insufficient focus on breeder seed procedures, or independent certification production agencies, and no funds for quality control. • Insufficient trained staff for seed Finally there is an acute shortage of production. trained seedsmen. • Insufficient focus on seed improved seeds in general. The public sector and parastatal orga • Limited fund allocation for the seed nizations have failed to provide quality sub-sector. seeds to Farmers. The private sector can 191 • Even bilateral projects have limited cion. (Although IDA concepts are legal status for commercial opera changing). tions. Market is often too small to attract • Non-existence or not functional sufficient private investment. National seed committee. • Inputs (fertilizers/herbicides/seed • Lack of information on seed needs. treatment/bags) are often in sort • Cumbersome procedures/non-exis supply. tent variety release procedures. • Facilities and equipments for seed • Inadequate equipments for produc treatment are limited in local mar tion and storage. kets. • Inadequate facilities for commer • Marketing and input delivery sys cial-oriented seed production and tems are often in Government con marketing. trol. • Rigid/cumbersome roles with a • In many cases, poor hybrid seed number of phytosanitary opera adoption is due to poor extension. tions. Plant breeders rights legislation ei • End users of seeds are re ther does not exist or is poorly en source-poor, rural based farmers. forced. • Insufficient trained staff for effec • Many regulations and cumbersome tive supervision of public sector process on importing or exporting seed programs. seed. • Delayed and limited traveling and • Major problem of emerging seed operational funding for field seed industry - Lack of development activities. seed policies and seed rules. • Difficult and lack of capital and • Lack of organization like National credit facilities at high rates. seed council/State seed coordinat Difficult situations to guarantee ing committees. loan acquisition. • Limited seed retail outlets and weak Suggested Activities to Support development of seed distribution. Regional Seed Program Development • Inadequate availability of breeder and foundation seed for certified Regional Seed Networking seed production. • Current low profitability and re Proposed Activities for Regional turns to private seed investment. Seed Network • Limited fund/credit/awareness/in volvement of public firms in seed • Regional seed newsletter. trade. • Regional seed director. • Catalogue of National seed stan Constraints to Private Sector Partici dards. pation In ECOWAS Seed System • Data base on morphological variety descriptions. • Governments and International De • Vital current technical seed tech velopment Agencies (IDA) view nology information. profits of private firms with suspi- • Develop and compliment Univer sity M.Sc. seed technology courses. 192 • Training courses outlined. • Who should pay for the economics of structured adjustment changes. Regional Variety Network • Studies on existing seed produc tion/processing/marketing/ distri • Regional variety evaluation. bution system. • Regional seed security. • Regional regulation for establish • Regional variety catalogue. ment of authentic private seed com pames. Regional Seed Standards Suggested Cooperative Activities in • Standard seed rules and regulations the West Africa/Central Africa and standards (AUSCA, IITA). Sub-region with Regard to Seed • Uniform variety registration/re Production and Marketing lease/certification procedures. • Regional variety data collection Strengthening National Seed Systems andinformation - IITAlICRISAT/ And Regional Seed Centers WARDA. • Ensure barriers to seed movement • Develop National seed systems ca between ECOW AS region mem pacities and regional seed centers. bers. • Encourage regional seed trade. • Standard see policy. • Use of private seed companies for Regional plant quarantine servIce seed production. and procedures. • Adoption of cost-conscious, • Seed trade liberalization and seed cost-effective, cost-recovery policy trade promotion. by public sector seed agencies. • National Agricultural Research Regional Seed System Systems (NARS) Analysis/Research • Coordination of National variety trials. • Cost/benefits of seed produc • Coordinate variety trials (public tion/marketing and use. and private developed) lines. • Institutional and management pol • Facilitation of National variety re icy studies. lease mechanism. • Presentation and discussions of • Ensure adequate production of model National seed policy; Na breeder seeds for National seed sys tional seed legislation; model stan tem dards and marketing network. • Social and environmental policy International Agriculture Research and environment studies. Centers (IARCs) • Inter-play between seed science technology and policy economics. • Enhance availability of greater • What options can take in privatiza breeder seed production. tion. Continuous development of supe • What options can take in seed secu rior improved varieties. rity for the country region. 193 • Increase availability of useful ge Seed Importers/Exporters netic lines within the West Afri can/CA Regions. • Trade in truthfully labeled commer cial and certified seeds. National Variety Release Mechanism • Facilitate export seed trade process. (NVRM) • Enter seed business and promote use of quality seeds. • Timely variety registration and re lease based on performance data. MANRIGovernment • Develop network between National variety release mechanism and re • Establishment of National seed gional seed trade. council. • Support National seed program de Private Seed Companies (PSC) velopment. • Regulatory Barriers to Regional • Provide best-produced, best Seed Trade priced, best-quality seeds. • Summarized (Table 1.) • Develop vibrant effective market ing network. National Seed Council • Assist in production of foundation seed of publicly bred seeds for • Provide forum for seed stake- NARS and IARCs. holders participation. • Use wholesales/seed distribu • Develop seed sub-sector policy. tors/seed producer/sellers. • Coordinate National seed activities. • National Seed Development Orga • Enact National seed legislation and nization provide guidelines for see industry • Coordination of National seed ef development. forts. • Promotion of seed industry devel National Seed Growers Association opment. • Ensuring cost-recovery, pricing • Link marketerslNational seed asso policy by public sector agencies. ciations/regional seed associa • Develop National seed quality con tion/stakeholders. trol agency. • Forum for meeting of formal and • Effective National seed policy. informal seed systems. • Provide meeting forum for regional Phytosanitary Regulations seed trade development in West Af rica/CA. • Test and issue phytosanitary certifi • Enable effective dissemination and cates. exchange of seed information tech • Clarify requirements for re nology. gional/international seed trade. • Improve economics of seed distri bution. 194 T abile e ommon R ejfU atory B arrlers t 0 S ee dMovemen t an dRecommen ddRfi e e orms. Rel!Ulatorv barriers Recommended reforms 1 Compulsory variety registration. Preferred solution: *Voluntary variety registration. If there is too much opposition then: *Compulsory variety registration limited, to a few major crops, leaving all other crops with voluntary registration; or *Automatic registration for varieties registered in a list of other countries within a region creating a multi-county variety list as in EU countries. 2 Compulsory seed certification and other quality *Voluntary seed certification. control. *Truthfullabeling: Mandate what companies put on labels, then prosecute if customers complain. *If a company repeatedly mis-labels seeds, force closinl! of the comoanv. 3 Lack of private access to public germplasm and *Allow private companies to get materials direct breeders' seeds, including materials fro IARCs. from IARCs. *Establish standard commercial arrangements to sell other public germplasm and seeds at cost. *F or some valuable products, sell or license for a orofit. 4 Inadequate protection for intellectual property rights *Do not require companies to give up physical con- in seed technology. trol of parent hybrid lines. *Adopt and implement laws allowing companies to rel!ister ownershins of varieties. 5 Unreasonable phytosanitary rules and other non-tar- *Work with international agencies and private orga- iff barriers on seed imports and exports. nizations to identify realistic pest and disease threats, to standardize seed tests, and to arrange transparent procedures to resolve disposals. *Do away with all other non-tariff barriers of seed imoort and exoort. 6 Over-centralized administration of regulations: *Deregulate as above, particularly doing away with compulsory variety registration and compulsory seed certification; and *As far as possible, decentralized administration of remaining regulations to provincial government or local offices of central agencies. Fragmented markets: i.e., small countries with ex- *For regional country does away with compulsory cessively Nationalistic seed systems. variety registration; (or automatically accepted vari- eties approved list of other countries). *Each regional country removes non-tariff barriers on seed imports. *Governments cooperate to study seed borne pests and diseases in the region. *Govemments harmonize phytosanitary and seed quality standards for internally traded seeds. Agric. Technology Notes (1995) AGRN Publication No. 12. World Bank Washington, D.C. (Sept. 1995). GovernmentlUNINGO Agencies (Re • Enhance regional seed produc gional Seed Centers) tion/marketing network. • Analyze seed demand and seed • Develop programs to strengthen production potentials in selected National and Regional seed supply countries. systems. 195 • Enhance development and dissemi • Increase availability of useful ge nation of seed professional knowl netic times within the W AlCA re edge. gion. • Support regional seed center and • Collate existing seed produc training activities for more tion/regional seed trade data/make seedsmen. known to states. • Estimate potential, effective and actual seed demands. Donor Agencies • Assistance on overcoming seed production/marketing data con • Sensitize governments on impor straints. tance of effective National seed • Carry out needed research on seed systems. technology on ECOW AS crop • Prepare draft seed policy/seed leg seeds. islation/seed rules and regulations • Ease barriers to seed trade and seed for modifications. movement in the Region. • Highlight measures for easing bar • Sensitize National and Regional rier to regional seed trade/supply programs to support adequate im systems. proved seed supply. • Develop Inter-regional program for Key Issues for Strengthening West security stock of foundation seed. African and Central African Seed • Strengthen base seed and founda Systems tion seed production at National and International Research Centers. Vital Areas where Improvements and • Promotion of certified improved Interventions are Needed Can be Sum seed production and marketing sys marized as Follows tem. • Analyze country seed demand/sup • Development of appropriate seed ply - crop-wise and variety-wise. policies and regulations. • Strengthen information collection • Improving dissemination of infor and dissemination (Training/ mation on available improved Newsletters). seeds. • Sufficient emphasis on timely vari Roles ofInternational Research Cen ety release, breeder and foundation ters seed production. • Strengthening seed marketing net • Continuous development of supe work for farmers as and when rior improved varieties. needed. • Support integration of regional and • Estimation of potential, effective national variety trials and release. and actual seed demand and feed • Make more breeder and foundation back. see production available. • To better understand and incorpo • Accept promising entries from pri rate farmers needs and priorities vate seed companies into Interna into research activities. tional trials. 196 • Appropriate research priorities/re Global Classification of Types of Seed search/extension/farmer participa Production/Marketing Enterprises tory breeding for needed seeds. • Timely variety testing/release and Variety Maintenance and Breeder Seed process of nation seed centers. Production • Improved seed campaigns coupled with training for different catego Public Sector ries of seedsmens/stakeholders. Improving seed production skills at Breeding lines gIven to private farm levels/seed processing quality companies. control development. • Released varieties multiplied by • Strengthening National seed sys foundation seed organization and tems. sold to seed producers. • Promoting privatization and com • Foundation seeds sold to seed pro mercialization (cost recovery) of ducers. both public and private sector at National seed system for Private Sector sustainability and quality seed trade. • Company controls all steps of ini • Use seed price support program by tial seed increases. government for both public and pri • Control either on own land or very vate sector cost-recovery produced closely supervised contract grow seeds to promote improved seed ers. adoptions, for some short period. • Strengthening regional seed trade Commercial Seed Production Agencies and easing barriers to seed move ment in region. Public Sector • Provide potential list of useful pro jects for support by donor agencies. • Ideally none. • Private firms to have sole rights to materials they develop. Private Sector • Private firms when viable to pro vide increment support and private • Contract seed producer - formal funded cooperation program for contract made between seed pro public section applied seed research ducer and company. programs and seed developmental • Most seeds produced for commer activities. cialization is produced under con • Institutional roles and responsibili tract. ties. • Independent seed producer - seed produced certified/not certified. • The farmer takes the risk for pro duction and marketing. • Production on company land - some companies make their land available and contract with farmers to increase production. 197 • Farmer-saved seeds - Among Producer/Wholesaler/Retailer self-pollinated species. • Most frequent system encountered. Seed Processing Organization contracts for large volume of seeds for wholesaler Public Sector and/or retail seed dealers. • Will also provide retail sales to lo • Limited to foundation seed organi cal farmers + other inputs. The zations. wholesaler also sells seed to other wholesalers or retailers in several Private Sector states. Some organization may have several operational sites. • Company owned facilities are used only for conditioning "company" Wholesaler seed. Company owned facilities which These are companies which pur are used to condition "contracted" chase processed seeds. Sell primar seed/or provide customer seed ily to retail seed stores and some cleaning services. other wholesalers. Generally their • Custom cleaners - stationary or operations cover a multi-state re mobile cleaners that move from gion. farm to farm cleaning seed for shelling, etc. Retail Only Classes ofSeed Production and Mar • The greatest number of retail keting Agencies "seed" stores handle all kinds of seeds and other inputs. Retail stores Producer/Retailer mayor may not be affiliated with a wholesale company. • Typically small organizations where the seed company produces Enabling Actions for Strengthening and conditions seed of two or three and Coordinating Both the Formal species and may retail seed of other and Informal Seed Systems crops purchased from other compa meso Enabling Actions Producer/Wholesaler Crop Research Primarily restricted to companies • Develop a timely system for the re which operate on a National and/or lease and introduction of new vari International basis. Seed are pro eties. duced under contract and sold to • Provide descriptive agronomic fea wholesalers after conditioning. tures of released varieties. • Have concern for who multiplies the seed and make available to farmers. 198 • Provide data from regional trials, • Leasing/availability of rural public which cultivars are most suited for rural agriculture input sales outlets each country. to private sector. • Provide cooperation for partnership • Custom seed processing. between research and private sec • Active improved seed campaigns, tor. demonstration plots/MTP /Field Days on economic benefits of im Foundation Seed Neededfor National proved seed. Seed Production System • Import privileges. • Strengthening for effective, reliable • A consistent supply of basic seed is quality seed control. essential. • Encouragement through availabil- • National foundation seed produc ity of basic seed production. tion agency and reliable private • Good technical support. seed companies should provide the • Good National seed policy. commercial seed sector of continu • Proper funding for the seed sector. ous supply of high quality seeds for commercial seed production. Quality Control Improved Seed Production and Mar • Need enabling actions for improv keting ing quality of seeds in seed trade of both formal and informal seed sys • Need to provide answers to who tem. grows, process and sells seeds. • How the quality of seed saved by • Put increasing emphasis on devel the farmer can be improved. oping private seed growers, enter • Seed association to push for entire prises seed producer/sellers. seed sector to be of higher quality. • Policies that encourage the dy • Internal quality control scheme to namic development of private com be integral part of the formal seed mercial seed industry. sector seed companies/enterprises. • Good cooperation and adherence to Promoting Growth and Development the laid down guidelines for seed of Private Seed Enterprises Producers/ certification. Sellers • There are at least two (2) types of external quality control programs: • Favorable credit for capital invest • Seed certification and Quality Con ment. trol at marketing. • Good credit to seed producers/mar • Quality declared quality seed pro keters/farmers. gram developed by FAOIUNDP. 199 Role of the Government of Niger in the Initiation of a Seed Industry Paul Buckner Background Niger is currently going through nu • Unfavorable agronomic and clima merous institutional changes including tic environments (periodic drought, economic liberalization, government parasites, soil erosion, abusive use pullout of production, democratization of the soil; and decentralization. • Poor economic environment due to poverty (per capita income has de Agriculture still remains an important creased by 33% between 1960 and contributor to rural economy. Despite its 1990 and is currently CF A 8,000); lower impact on the country's gross in • Extensive genetic heterogeneity of the material grown by farmers due come (66% in 1960 vs. 41 % in 1995), ag to the use of poor quality grain as ricultural improvement has been accepted seed. as the engine of the national economy un der the Economic Renewal Program. The use of improved seeds is indis pensable for intensive agriculture to in Cereal crops are the most prevalent un crease and stabilize production and food der dryland agriculture, occupying three security. Indeed improved seeds allow a fourth of the total agricultural land. Over better valorization of other agricultural the past ten years (1987-1995), the culti inputs. vated area under millet went from 3 mil lion hectares to 5.2 million, and that for Current Status of the Seed Sector sorghum from 1.3 to 1.9 million. Yields are generally variable and have declined There are two important institutional in the same period. The average millet components: yield is 390 kg/ha and that for sorghum is 190 kg/ha. Overall agricultural produc 1) The public sector which is responsi tion is variable, from 1.37 to 2.4 million ble for research, seed multiplication, leg tons between 1987 and 1997; still there is islation, and extension. In this context a a slight 1.4% increase in production for National Seed Committee supervises millet and sorghum. inter-ministerial and inter-professional activities aimed at advising the govern The most critical reasons for a lower ment on issues pertaining to seed activi productivity are: ties. So far the committee has essentially dealt with price fixation, and at the local • Low soil fertility and mono-crop level, it has promoted planning and coor ping; dination of activities. A National Seed 200 Service was created in 1985 but was re ture, grassroot organizations and partici placed in 1990 with an office in charge of pation, and a new role for the State. The inputs, control, processing, and agricul Economic Renewal Program emphasizes turallegislation. There is presently no ag the role and value of the private sector in riculturallegislation enforced. The 1990 cluding individual businessmen and co government decree (No. 9055 PRN operatives on one hand, and a reinforce MAG/EL) dealt with the requirements for ment of economic ties among private en seed production, processing, and quality trepreneurs on the other hand. control, but it was not implemented. Seed activity is only protected by phytosanitary In terms of seed activities, the strategy regulations. A research organization, the is to facilitate the emergence of profes Institut National de Recherche sionals in seed production and marketing. Agronomique (INRAN) is in charge of The state would have a role in defining na crop improvement as well as production tional policies, in research, in supporting of breeder seed. Regional and local ex and advising different partners, in quality tension offices provide training, exten control, in defining and controlling seed sion and diffusion of technologies. legislation. This role may be imple mented at three different levels: 2) The private sector deals with com mercial seed production and marketing. Coordination The above role distribution was opera The current policy aims at a progres tional between 1976 and 1989 under the sive withdrawal of the state from seed ac Cereals Project. The public sector be tivity and its gradual replacement with the came a lone actor after 1989 while the private sector. The state will remain as a economic situation was deteriorating. regulator, producer, and lawmaker. Seed centers were assigned to local com munity governments, and seed increased For a successful seed production, the in part by farmer producers and a few in different regions should make their own dividuals. Seed production in the private decisions and plan their activities. On the sector is still emerging, being conducted national level, it will be responsible only by a few farmers in the vicinity of seed for general orientations. However breed centers and a few private producers. ing and production ofbreeder and founda tion seed will remain in the public sector. What Role Government Can Play in Certified seed will be produced by farmer Initiating Seed Enterprises? groups with technical support from ap propriate government agencies. The gov In July 1997 a government act defined ernment role may be summarized as: sectorial policies under the Economic Re newal Program. This program includes • farmer support by extension agents; the 'strategic principles for rural develop • breeder seed from research agen ment' that were developed in ordinance cies; no. 92-030 of July 8, 1992. The policy • an economic environment leading goals are food security, intensive agricul- to opportunities for credit to seed 201 producers and other input suppliers; Quality Control • more dynamic regional seed com This is a critical aspect and government mittees. has a key role to play: Marketing of improved seed should be • prepare a seed law; left to farmer organizations and private • create a variety release committee; producers in different zones. Govern • manage genetic resources. ment will help through: Perspectives • promoting liberal pricing; • circulating the information on price The overall strategy would anticipate a of commodities; national seed program to organize all the • supporting cooperative; partners, define the objectives, and decide • improving techniques for process on specific programs and projects. This ing and storage; strategy is not functional at this point, so • initiating banking and credit orga presently only a minimum level of funds nizations to facilitate activities of are provided by the state and local gov private entrepreneurs. ernments to support seed activity. Training and Information Exchange A project for promotion of improved seeds has been submitted to the European Tthrough its own agencies and devel Union under the 8th FED for ECU2, opment partners, the government will: 208,108. The main objective of this pro ject aims at ensuring food security • train in seed production techniques; through a realistic approach to seed man • inform producers and farmers on is agement, involving the private sector sues relevant to improved seeds; through organized groups, along with • demonstrate improved varieties through extension programs. other partners including institutes, devel opment projects, and NGOs. 202 Credit/LoanslBanks Serving Rural Users in India Dr. Kuldip R. Chopra and Dr. Rakesh Chopra Introduction straints operate more severely in underde veloped regions than in the more India is a vast country with varying developed regions of the country. agro-climatic conditions, tenancy laws and land holdings ranging from 0.4 ha Rural credit in India has been an area of (marginal) to 17.33 ha (large) with an av policy interventions for nearly five de erage of 1.57 ha per family. Only 33% of cades (since the country's independence). the cultivated land is irrigated. Policy objectives were: Agriculture is the mainstay ofthe econ • To institutionalize credit; omy. About 76 percent of the estimated • To enlarge its coverage; 900 millions inhabitants lives in rural ar • To ensure provision of timely and eas. About 60 percent of the total popula adequate finance to as large a seg tion is directly dependent on agriculture. ment of rural population as possi The share of agriculture in the net domes ble; and tic product was about 32 percent in • To achieve an anticipated growth rate in agricultural production and 1990-91 and is projected to decline to rural employment. about 25 percent by the year 2000. Small and marginal farmers and high population The Indian rural credit system, as it has density characterize Indian agriculture. In emerged, is a product of both evolution a good agricultural year, food is adequate and intervention. Until the late sixties in relation to effective demand, price re policy interventions in the area of rural mains stable, agro business and industries credit were mostly confined to coopera lookup, employment improves and rural tives, which were considered as most suit poverty recedes somewhat. able this work. However, the "All India Rural Credit Review Committee" pointed The main constraints to increase agri out that, despite all support extended, co cultural productivity are the convenient operatives had begun to exhibit several availability of rural credit on time, in weaknesses in the form of overdues and needed amounts and at reasonable interest organizational ineffectiveness, and are rates and the declining size of farms re not geared for meeting the growing needs sulting from the continuous process of ofthe fast emerging technology and inno subdivision of holdings, overexploitation vations relating to production, processing of water in green revolution areas where and marketing in the agricultural sector. by rotation paddy and wheat are grown With the nationalization of major com year after year, and the energy constraint mercial banks in 1969 and introduction of persisting as a critical factor in spite of regional rural banks (RRB) in 1975, large public investment. These con- 203 branch banking in the rural areas acquired Middle Tier a new momentum. The outreach of the commercial banking system was consid • Public and private Sector Commer- erably enlarged within a relatively short cial Banks period of time, resulting in certain real • Regional Rural Banks (RRBs) gains such as: • Land Development Banks (LDBs) • State Cooperative Banks (SCBs) • Broadening of the rural infrastruc • Central Cooperative Banks (CCBs) ture for credit delivery; • Increase in outreach; Third Tier (in direct contact with rural • Reduction in the influence of infor users) mal segments; and • Substantially stepped-up credit • Primary Agricultural Credit Soci flow for agriculture and allied ac eties (P ACS) tivities. • Central Banks and State Govern ments mostly for subsidized credit However, on the debit side, quality of • Commercial Banks Rural service suffered, overdues increased, re- Branches; and payment ethics was diluted, and profit • Regional Rural Bank Branches ability and viability of the institutional structure was seriously eroded. RBI is responsible for overall mone tary policy and provides accommodation Present Situation to NABARD and IDBI for agriculture and rural industries respectively. These insti Rural users in India have access to tutions in tum provide refinance to the af credit through either an organized or for filiated commercial banks, including mal segment, which functions within pro RRBs, SCBs and State Land Develop vision of the 1956 "Indian Banking Com ment Banks (SLDBs). The refinance from panies Act", or an unorganized or infor NABARD is distributed to rural entrepre mal segment, which operates outside this neurs through two and/or three tier coop Act. erative structure respectively for long, short and medium term lending. How Organized or Formal Segment ever, in case of commercial banks and RRBs, the refinance flows directly to the The formal segment is organized as users. Subsidized credit flows from State follows: Governments through their district or block offices or the lead bank in the area. At the Top Despite an appropriate and • Reserve Bank of India (RBI) well-funded structure and support from • National Bank for Agriculture and policy interventions, the formal segment Rural Development (NABARD) shared about 65 percent of the rural credit • Industrial Development Bank ofIn market in 1996. Likely reasons for failure dia (lDBI) 204 to cover 90 percent or more of the market Informal segment continues to domi are: nate the rural credit scene primarily be cause its lenders combine money lending • Political interventions especially to another rural economic activity, e.g. when borrowing from cooperative grocery store, agro-product processing, banks; poultry, etc., in order to ensure • Delays in disbursement of loan in sustainability of their own business. Fur stallments; thermore, their loan operations are quite • Previous overdues making the bor flexible in limits, end use of loan, recov rower ineligible; ery schedule, repayment in cash or kind • Cumbersome and bureaucratic loan and thus are more geared to meeting the processing procedures. needs of borrowers. Records are simple and not subject to audit. Finally, these ca As a result, small and marginal farmers sual lenders have the advantages of per fail to benefit from institutional credit. sonal contact, secrecy of loan and terms and better relations with the users. Unorganized or Informal Segment District Credit Plan The share of the informal segment in the total rural debt even today is estimated The efficiency of credit depends on the at about 35 percent. The unorganized sec quality of integration between in tor consists of: fra-structural facilities, technological in novation, research and extension support. • Professional Money Lenders: their Such linkages were generally weak be primary business is rural financing. cause different agencies tend to operate in • Nonprofessional Money Lenders: casually involved in financing, they a compartmentalized fashion with field focus on other rural businesses. level interaction being minimal. There • Indigenous Banks: native or re was, thus, a need to strengthen such link gional private banks. ages through a positive effort at the vari • Merchants: they provide seeds, pes ous fora available at the district level. The ticides, fertilizers and farm equip Lead bank concept was, therefore, started ment on credit. for each service area. The Lead bank is ei • NIDHIS: Hundi instrument letter ther a cooperative, or a Regional Rural for loan Bank (RRB) or a commercial bank. It acts • Chit funds: a voucher of collective as a consortium leader for coordinating fund that could be utilized by any the efforts of all credit institutions in the member of fund as and when, re district granting credit for the purposes of quired. branch expansion and meeting credit • Landlords or large-scale farmers: needs of the rural economy. Credit plan they own large land holding and fi ning for a village or for that matter any nance their tenant farmers only. area, is more meaningful if all the credit • Friends and relatives: they offer and development agencies in the area ef short-term loans. fectively coalesce in the preparation of the credit plan and each agency is given a 205 task to perfonn in its implementation. It is RBI from time to time issues guide necessary to have better collaboration be lines on the crop loan system. Appropri tween bankers and development planners ates scales of finance for each crop are in this effort. To tone up the quality of ru fixed by the field workers in consultation rallending, RBI from April 1989 initiated with agricultural experts of the lead bank a new approach known as "service area of the district, for both cash credit and in approach" . put credit. Inputs are sometimes supplied in kind. These scales vary for irrigated or The crucial element of the "service non-irrigated cultivation and according to area approach" is the interaction of each the level of development of the area (de bank branch's credit plan with the "Dis veloped versus non developed). trict Credit Plan", taking in account the fi nancial support and infra-structural facili Investment Credit ties available in the area. Recently the "service area approach" has been modi Investment credit needs are met fied into a development linked credit through various "product specific" plan. schemes, i.e., the loan is specific to the ac tivity. It is usually sanctioned on the eco Rural Credit Segmentation nomics of the activity and the asset/liabil ity and income/expenditure pattern of the Under the present arrangement, agri borrowers. Investment loans are available cultural credit for all purposes relating to for: agricultural production and activities al lied to agriculture, whether of short, me • land development; dium or long duration, is provided by an • fann mechanization extensive network of cooperatives and/or • minor irrigation such as digging a designated "service area" of commercial well/tube well; banks or Regional Rural Bank (RRBs) • production and processing of seeds; branches in the rural and semi urban ar • horticulture and plantations; eas. • post harvest operations; • construction of cold storage; On Farm Production Credit • fann forestry; or • setting up a biogas plant, etc. Crop production requirements of the While production loan requirements on fann sector are met through: are recurring in nature, fanners ordinarily do not make investments on above prod • loans granted on a crop, • cash credit granted on a crop ucts on an annual basis, hence depending • overdraft facilities granted on a on the activity, tenn loan periods range crop between 3-9 ans. Before extending credit • variations of the agricultural credit facility, bankers make comprehensive as card system. sessment of the regions cropping pattern and the likely production income. They 206 also take into account the borrower's pas tion. Despite a significant expansion in repayment record. the flow of credit to the agricultural sec tor, there is a perception that investment Scales of Finance in agriculture, particularly non-farm in vestment, has not kept pace with demand. The credit limit fixed for production It is believed that the greatest stimulant credit is generally much less than the ac for the growth of the non-agricultural tual cultivation costs and expected reve economy in the rural areas is sustained ag nue. For example, the cultivation costs for ricultural growth. Apart from agriculture, grain sorghum under irrigated and rainfed the primary activities that serve as an im conditions following recommended culti portant basis for the non-farm sector vation packages are approximately INR (NFS) livelihood include livestock rear 2,570 and INR 1,740 per acre respec ing, fisheries, forestry and mining. tively. Similarly for pearl millet, they amount respectively to INR 1,965 and Rural households generally pursue a INR 1,445. However, the banks have number of activities both agricultural and fixed a uniform production credit limit of non-agricultural for supplementing their INR 1,200 for grain sorghum and INR cash flow. These activities which include 900 for pearl mil per acre irrespective of small scale trading, agro-based micro in irrigated or rainfed conditions. On ac dustries, local transport, etc., have modest count of the large difference on the actual working capital requirements and lend expenditure to be incurred and the limits themselves admirably to bank finance. fixed by banks, farmers are compelled to Therefore, rural banks design specific borrow from the non-organized sector at loan products for this sector to provide very high rates. The seed companies ar short-term loans up to 3 months with range production credit for their contract weakly or fortnightly repayment of inter sorghum and pearl millet seed multipliers est/principal installments. Borrowers' re under a group warranty scheme. They quirements are reassessed at frequent in provide cross guarantee to the lead bank tervals and repeat loans of increasingly of that area, on behalf of all contract mul higher amounts for longer periods are of tipliers, to first pay the principal and inter fered. There are built-in incentives of in est amount of all accounts (borrowers) terest rebate in case of prompt payment. from their seed income and pay the bal Credit to non-farm sector has a high turn ance to each multiplier. Banks prefer this over on account of rapid recycling of system as it ensures almost 100 percent funds leading to increase in credit flow. loan recovery and the user gets timely, ad equate and convenient availability of Informal Sector credit at a prevailing subsidized rate of in terest. The problems faced by the poor in dealing with formal rural financial institu Non-farm Sector tions are many. The factors that constrain the capacity of the poor to borrow from The rural non-farm sector covers ap the organized credit institutions have proximately 16% ofthe country's popula- been: 207 • lack of skills; • employ experts to evaluate eco • ignorance of economic opportuni nomic feasibility of projects pro ties; and posals, • their inability to comply with the • sanction loans, and procedural requirements. • monitor project progress under 100 percent NABARD refinance. State governments have encouraged formation of "Self Help Groups" (SHGs) Some examples of hi-tech capital-in by "Non-Governmental Organizations tensive projects for rural credit financing (NGOs) to perform intermediary func schemes are: tions between SHGs and the financial in stitutions. Reserve Bank of India (RBI) seed production, processing and has advised rural lending institutions to marketing; consider SHGs needs as their regular • floriculture; lending operations. Banks lend to SHGs • mushroom cultivation; who in tum lend to their members. Alter • aquaculture natively banks work through NGOs, • tissue culture; and which provide credit to SHGs to be • value addition to agricultural com modities for urban centers and for passed on to their members. SHGs, espe exports. cially those formed by rural women have been very aggressive and successful. Under World Bank funded National Seed Project's phases I, II and III, credit to The main advantage to banks is exter nascent seed companies was made avail nalization of a part of the work of the able at subsidized interest rates and partly credit cycle such as: in foreign exchange to import seed pro cessing equipment and technology. • assessment of credit needs; • appraisal, disbursal, supervision, repayment; Subsidy Linked Credit • reduction in paper work involved, hence consequent reduction in the To promote a new product or a process transaction cost; and in the rural sector, both the Central and • recovery of about 95 percent dues. State Governments often resort to subsidy linked credit schemes. The emphasis of Hi-tech Credit Needs of the Rural these credit linked subsidy schemes is on Sector achievement and implementation of the district wise physical targets. The subsidy To efficiently respond to the emerging component is usually channeled through a agricultural technologies leading to com lead bank under the "Area Development mercially oriented Hi-tech agriculture, Approach". It is disbursed through the with capital intensive credit needs, RBI lead bank of the area to which the bor have encouraged banks to: rower belongs. The extent of subsidy may range from 25 to 75 percent of the total • open Hi-tech branches in rural ar costs. Some subsidies are recurring; oth- eas, 208 ers are a one-time disbursement only. Re credit (GLC) to NABARD at a fixed in cent examples of subsidy linked credit in terestrate of5.5 % per year. NABARD in clude setting up a fruit orchard, setting up tum extends Seasonal Agricultural Oper a tissue culture lab, setting up a fertilizer ations (SAO) refinance credit to coopera mixing plant, purchase of sprinkler or tives and lead banks at 6.5% for short drip irrigation system, construction of term, e.g. crop production credits. The cold storage. The impact of such schemes banks in tum are required to fix inter in bringing families above the poverty est/lending rates in a manner that, glob line or providing gainful employment is ally, the interest earning by the ground usually weak. level institutions namely PACs, PLDBs, RRBs, and commercial banks gross mar Interest Rates gins ranges between 5 and 7%. From the viewpoint of banks, the regulated rates of Organized Sector interest operate as a barrier to the sanction of small loan as managers tend to look for The Government ofIndia believes that comparatively larger borrowers. This is in present state of development of the In one of the reasons why the share of com dian money market and capital market, mercial banks in crop production loans is determination of interest rates cannot be substantially smaller than that of the co entirely left to market forces and a fair de operatives. For crop loans, there is a sub gree of regulation is necessary to provide stantial unfulfilled demand that is being for an orderly mobilization of financial met either by the moneylender or by re savings for the purpose of planned eco ducing the use of inputs, either strategy nomic development and in the interest of resulting in a loss of income to the farmer. the viability of bank operations. It also Therefore, RBI has recently deregulated recognizes the need to make the adminis the interest rate for rural credit; hence the tered rate system less complex, more flex lending rate for short-term credit now ible and whenever possible to reduce the ranges between 12.5 and 16.5 percent and extent of concessions and categories of is negotiable. The government is obliged concessionary lending. to reimbursed the lead bank if they direct them to provide credit to certain catego For the rural credit, the category of ries of rural borrowers at lesser than the concessionary financing is comprised of: recommended minimum rate of 12.5 per cent. • small and marginal farmers; and • other weaker sections, only. In principle, during the preparation of a district plan, the targets fixed for financ The "other" category comprising the ing the weaker section and the rural poor rest is free from any regulation except a are such that the banks can bear the losses maximum ceiling on interest rates of 15.5 on this score through earnings from their %. commercial lending. Mandatory lending has accordingly to be within limits that Effective July 1,1996, the Reserve cross-subsidization permits so that the vi- Bank ofIndia sanctioned a special line of 209 ability of the institution is not adversely rate of interest. The maximum interest affected. rates can go up to 50% depending on the credibility of the borrower. This is gener For commercial actIvItIes of the ally available to the non-farm sector and non-farm sector, NABARD/IDBI issues those who can't provide any security. The loans to lead banks at 9.5%, and the later rest of interest is negotiable and is gener have the flexibility of lending at 12.5% ally charged on a monthly basis. The first and above. Loans to tribal and other installment of interest is deducted from weaker sections of rural society are given the initial payment (loan amount). The ru at 1% lesser interest rates. ral users ofthis kind ofloan are poor farm ers and artisans who find difficult to get a Unorganized Sector loan from the organized sector. Due to the high interest payable every month, once Interest rates charged by the unorga caught in the clutches of such moneylend nized sector vary a great deal. Minimum ers, it becomes difficult to come out. interest rate is usually 24 % against hy pothecation of standing crop, some prop Loan repayment erty, land or security of gold, farm equip ment or even other household goods. This Experience has shown: is usually available to the farming com munity. Those farmers who do not have • repayment rate of loans sanctioned access to organized sector for finance, by cooperatives and RBI is low, usually prefer this channel. Many times, ranging between 50 and 60% in case of urgency, they fulfill their finan • repayment rate of loans sanctioned cial needs by taking loans at such a high by commercial bank rural branches is usually 80% or more. 210 Wednesday - September 30 - Session 1 Sorghum and Millet Utilization Opening Remarks Moderator - Bruce Hamaker, INTSORMIL Evaluation of Sorghum Grain Quality for Traditional Thick Porridge in Africa G. Fliedel, C. Sambumukama, L. Songre, and F. Matenncio Introduction gradually replaced with maize but most of the countries suffer from drought and A major objective of the Cereal Tech maize cannot be grown under this condi nology Laboratory is to study sorghum tion. Therefore, it is essential to identify grain quality for specific uses so as to help sorghum varieties with good adaptation breeders in the identification of varieties but more suitable to consumer taste. Con that are not only productive but also have sumer surveys have shown that the most potential for utilization. Therefore the common food use of sorghum in Africa is laboratory developed mini procedures to a thick porridge usually called 'bowl' or predict varietal attributes that are addi 'paste'. Hence under breeders' request, tional breeding tools and to understand three tests were developed in our lab to the physico-chemical basis of quality. predict sorghum porridge quality: one test for West African to, the second for South The present document summarizes ern African bogobe (Botswana), and the several research projects funded by the third for Eastern African ugali (Tanza European Community in collaboration nia). with research organizations in Europe and Africa: Vrije Universiteit (Belgium), Three tests were necessary because the Hamburg University (Germany), NRI three thick porridges were different in (England), nCT (Portugal), IER (Mali), texture, flour particles, flour concentra IPR (Mali), CNRST (Burkina Faso), BTC tion, and cooking procedure. The West (Botswana), and Sokoine University African to starts with a light flour-water (Tanzania). This work would not have mixture to which flour is continuously been possible without the help of col added and stirred during cooking; the leagues and technicians in the laboratory. flour/water ratio is 20% on dry flour weight basis. The Southern African This report describes three tests that bogobe is very thick and is prepared from can be used to predict sorghum grain a mixture of coarse and fine flour parti quality for making traditional thick por cles; the flour/water ratio is 40% on dry ridge in Africa. The work was based on flour basis. Cooking method is simpler: two observations: in West Africa sor flour is poured directly in boiling water, ghum is produced and consumed in many the pot is covered and the mix is stirred countries ofthe semi-arid tropics. In these only at the beginning and at the end of countries only local varieties are grown cooking. The East African ugali is similar and most improved varieties have not to to in texture, flour particles and cook been adopted by farmers and end-users. ing procedure, however the flour/water In Southern Africa, sorghum has been ratio is intermediate at 30%. 213 The tests were developed using the fol The Particle Size Index (PSI) was de lowing steps: 1) field surveys were con termined according to a method devel ducted to understand the criteria for a oped by Fliedel et aI., (1989). After grind good quality porridge; 2) learning the tra ing 20 grams of grain in a Falling Number ditional cooking procedures and transfer KT30 type of grinder (using the setting ring them in small scale to the lab; 3) set for coarse semolina, level 4), the product ting-up the porridge quality test; and 4) was sieved for one minute using a 250 f.1m studying the physico-chemical basis for sieve (Alpine 200LS air sieve). The PSI is quality. the amount of flour passing through the sieve, and is inversely proportional to Materials and Methods grain hardness. The dehulling yield was obtained after decorticating the grain for Materials five minutes in the T ADD (Tangential Abrasive Dehulling Device). This T ADD To quality was evaluated using several is made of a carborundum abrasive disk West African varieties (IER Mali, that turns horizontally under a steel box CNRST/INERA, Burkina Faso; made of 8 separate containers so that 8 ICRISAT/CIRAD, Mali). The to equa different varieties may be decorticated at tion was checked using 32 varieties from once. The dehulling yield is the ratio of the ICRISAT/CIRAD station at grain weight before and after Samanko, Mali (1995 harvest). decortication. Bogobe quality test was developed us Flour Analyses ing three commercial flours from Bot swana (February 1996), five varieties Flours were obtained from 75 to 80% from the Botswana Department of Agro decorticated grains. Three types of mills nomic Research (1995 harvest), and four were used: Perten hammer mill and varieties from ICRISAT/Zimbabwe. The Cyclotec helix mill for fine flour, and test was conducted on 31 varieties includ Cemotec 1090 mill for coarse flour. ing 24 from ICRISAT/Zimbabwe (one from 1995,23 from 1996 harvest), and 8 Flour particle size was determined by from the Botswana Department of Agro successive sieving using sieves with di nomic Research (1995 harvest). ameters varying from 75 f.1m to 1000 f.1m, depending on the flour. Thirty grams of Methods flour at constant moisture content (10-11 %) were sieved on an Alpine sieve Whole Grain Analyses for 3 min. The amount of flour remaining in each sieve was measured. The average Prior to analysis, grains were cleaned particle size G50 was measured as sieve to remove dust, glumes, and broken ker openings that retains 50% of the flour. nels. The cleaned grains were condi tioned to 11-12% moisture content to Flour solubility and volume were de minimize variability in grain hardness. termined on suspensions made of flour solution (7. 1 dry flour weight in water, 214 I Rupture 20 / 16 -b S-o u 4 2 4 6 8 10 Distance (mm) Figure 1. Compression curve for a to sample a e o consls. t ency score Iior 32 sorglh urn cu It·Ivars ongma.. f mg f rom Mral Breaking force Homogeneous groups Number of Cultivars IN) (Newman-Keuls test) T6quality 2 >18 A 4 18-15 B Very hard 10 15-12 C Hard 5 12-10 D Medium 9 10-7 E Soft 2 <7 1 ~nLSOft i.e. 2 grams of flour in 28 grams ofwater), to 90°C at a rate of 6°C/min, held at 90°C using the RVA (Rapid Visco Analyzer for 5 minutes, and cooled to 36°C at type 3D, Newport Scientific, Narrabeer, 6°C/min. The porridge was immediately Australia). The temperature cycle in the transferred to a 50 m!. tube and centri RVA was as follows: the suspension was fuged for 10 min at 5000 g and a tempera maintained at 36°C for 2 minutes, heated ture of 25°C. The supernatant (Ps) and 215 pellet (Pc) were weighed separately and acid in the presence of a catalyzer (99.5% dried at 100°C for 24 hrs and 48 hrs. re potassium sulfate + 0.5% selenium). spectively to determine their dry weight (MSs and MSc). Three parameters were Amylose content was established by calculated: the amount of soluble materi differential enthalpic analysis or differen als in the supernatant phase (SM), the tial scanning Calorimetry (DSC) using swelling power (G) and the volume of dis hermetic stainless steel pans (Mestres et persed matter ( ) according to the formu aI., 1996) on a Perkin Limer DSC7 lae: (Perkin Elmer, Norwalk, CT). The princi ple is to measure enthalpy differences be SM( mg/ml )=MSs(Ps-MSs) tween a sample and a reference that were G(g/g)=(Pc- MSc )IMSc submitted to linear series of heating and =(27 .25-(Ps-MSs »/27.25 cooling cycles. Enthalpy variation was determined from complexes between where 27.25 is total volume (cm3) of the amylose and added monoglyceride (ex suspension assuming specific flour den tracted from egg yoke). The pan contain sity of 1.5 g/cm3. ing the sample (10-11 mg of flour in 50 /-ll ofL- -lysophosphatidylcholine 2%) and Chemical Analyses on Whole Grains the pan containing the reference (50 /-lml And Flours of purified water) were heated from 35°C and 160°C at 8°C/min, held at 160°C for 2 Water content was determined follow min and cooled from 160°C to 60°C at ing NF V03.707, using a 5 gram ground 10°C/min. The exothermic heat for sample that was weighed before and after complexing amylose and phospholipid drying in an oven at 130°C for 2 hrs. during the cooling is proportional to the amylose content in the sample and can be Ash content was determined following determined in comparison with a standard NF V03. 702, using a 5 gram ground sam (100% pure amylose from potatoes). ple that is weighed before and after it was ignited at 900°C for 3 hrs. in a muffle fur Description and Discussion of Tests to nace. Evaluate Quality Lipid content (free lipids) of the dried Test/or Evaluating To Quality samples was measured by the Soxhlet ex traction method using the Tecator Soxtec Traditionally, women sieve the flour HT type distillator. Lipid extraction was obtained after pounding the grain in a done at 100°C after immersion ofthe sam mortar and grinding it at the neighbor ple in diethyl ether. Total protein content hood mill. Then a 'flour milk' is prepared (%N x 6.25) was determined by the by diluting a small quantity of the fine Kjeldahl method using an automatic flour in acidic water (using lemon or Tecator device (Kjeltec). The sample was tamarin juice), then alkalized (with pot mineralized with concentrated sulfuric ash from wood ash) or not depending on the ethnic group or country. The flour milk is poured in a cooking pot containing 216 Maximun ...... P"'I"'., - _._ •• _ •• _. __ • __ •• __ • Minimun Force of extrusion 10 1 4 (; 8 Distance (mm) Figure 2. Extrusion curve to evaluate the consistency of bogobe or ugali. Table 2. Bogobe consistency for nine varieties from Botswana and Zimbabwe Extrusion force Homogeneous groups Cultivars (N) (Newman-Keuls,5%) BSHI B,H 16.2 Mmabai'tse B,H 18.5 2 Mahube B,L 21.7 3 SDSL2593 Z,I 23.9 4 Pato Z,I 24.2 4 Larsvyt Z,I 24.8 4 Segaolane B,L 26.0 4 5 Phofu B,I 27.5 5 Kuyuma Z,I 30 5 6 B = Botswana; Z = Zimbabwe H = Hybrid; L = Improved local variety; I = Improved introduced variety boiling water and cooked for approxi sistency may be adjusted to a desired level mately 8 min to obtain a thin porridge. To by adding more of the thin porridge saved part of this porridge was added the rest of earlier. After 20 min the to is ready. To is the flour by handfuls while vigorously consumed with a sauce made of legumes stirring with a wooden stick during the (onion, tomatoes, okra), meat or fish. To cooking. The porridge thickens, becomes is eaten as a principal meal, and if it re- homogeneous and smooth. Porridge con- 217 Table 3. Ugali consistency of13 sorghum varieties originating from Tanzania, Botswana, and Zim babwe Extrusion force Homogeneous groups Origin Cultivars (N) (Newman-Keuls test, 5%) T,L Langa-Ianga 54.1 T,L Mbangala 50.2 2 B,r Phofu 43.0 3 T,r Pato 42.0 3 z,r Kuyuma 41.9 3 z,r Macia 41.8 3 T,L Udo 39.8 3 B,L Segaolane 38.9 3 T,r Tegemeo 1995 37.8 3 4 T,r Tegemeo 1997 33.7 4 B,L Mahube 33.3 4 B,L Mmabaltse 25.6 5 B,H BSH-I 23.1 5 T~Tanzania; B~Botswana; Z~Zimbabwe L~Improved local variety; I~Improved introduced variety; H~Hybrid. mains it will be used the next day for under mechanical shaking (400 rpm) and breakfast with sour milk or sauce. pOUTed into two small steel cylinders (4 cm diameter and 2 cm height) that were The most important criteria for t6 qual closed with adhesive band and placed on a ity by decreasing order are consistency or glass platform. After 20 hrs in a stove at firmness, overnight keeping quality with 35°C saturated with water, the adhesive out syneresis, taste, and color. A good t6 band was removed, and a sample of t6 is hard, with good overnight keeping with homogeneous texture and constant quality, is light colored, and tastes good size (3 cm diameter, 3 cm height) was ob (no bran taste). tained. In the laboratory the test was devel T6 consistency was measured with an oped as follows, largely based on consis Instron (Universal Food Testing Machine tency since surveys have shown that a model 4300) by placing the sample on a thick porridge keeps its quality dUTing the platform and compressing it with a probe night: samples were decorticated in the moving at a speed of 5 mm/min. A resis TADD to yield 75% decorticated grains tance force was measured in Newtons. that were ground in a Perten mill to a fine The force needed to fracture the sample flour (particle size of 150 ~m). Flour was selected as a meaSUTe of t6 consis (16.8 grams at 11.5% moisture content) tency (Figure 1) because it allowed a was mixed with 90 ml of water to obtain a better differentiation of the varieties. The flOUT: water ratio of 16.5%. The suspen greater the force, the harder the t6. Four sion was cooked for 20 min in a water bath measurements were taken for each variety 218 (two samples for each of two cookings). Test for Evaluating Bogobe Quality Under these conditions, two varieties with hardness values differing by one N The traditional method for making will be statistically different at the 5% bogobe is very simple: flour is directly level of significance. poured in boiling water and stirred using a small wooden stick at the beginning and The test was applied to a wide range of at the end of cooking. Porridge is cooked cultivars including local and improved for 20 to 40 minutes. varieties, new cultivars and breeding lines. A scale for t6 quality evaluation was Flour may be obtained by pounding established (Table 1). Varieties with t6 grain in a mortar but often commercial consistency score greater than 12 flour is bought in a neighborhood store Newtons are considered acceptable and and distributed by some 70 mills existing those with a score greater than 15 now in Botswana. The mills have differ Newtons had excellent quality; these ent capacities and are all made of a series were usually local varieties. Below 12 of RIle decorticators (PRL type, locally Newtons, t6 was soft and unacceptable. fabricated). In all cases, the flour is a mix ture of coarse and fine particles. Surveys indicated that important quality criteria were bogobe hardness, grainy texture due to the presence of coarse and fine parti- Table 4. Grouping of 13 sorghum varieties in two quality tests: for to by compression and for ugali by extrusIOn. Consistency Homogeneous Consistency Homogeneous Cultivars (N) groups Cultivars (N) grouns Kuyuma 13.9 A Langa-Ianga 54.1 A Mbangala 13.0 AB Mbangala 50.2 B Pato 12.8 AB Phofu 43.0 C Langa-Ianga 12.4 AB Pato 42.0 C Macia 12.0 ABC Kuyuma 41.9 C Phofu 11.3 BCD Macia 41.8 C Tegemeo 95 10.6 BCD Udo 39.8 C Tegemeo 97 10.1 CD Segaolane 38.9 C Segaolane 9.8 D Tegemeo 95 37.8 CD Udo 8.0 E Tegemeo 97 33.7 D Mahube 7.1 EF Mahube 33.3 D BSH-J 5.8 FG Mmabaitse 25.6 E Mmabaitse 4.5 G BSH-I 23.1 E T a bl e 5 G roupmg 0 f Botswana sor!!:hum varieties in three laboratory tests. Cultivars T6IN) Ugali (N) Bogobe IN) Kuyurna 13.9 41.9 30.5 Phofu 11.3 43.0 27.5 Segaolane 9.8 38.9 26.0 Mahube 7.1 33.3 21.7 Mmabaitse 4.5 25.6 18.5 BSH-I 5.8 23.1 219 60~------• 50 r=0.84 • 40 • • 30 • 20 4 6 8 10 12 14 To consistency (N) Figure 3. Comparison between two laboratory tests for evaluating to and ugali consistency using 13 sorghum varieties. cles, bland taste, and light color. In the stirred twice at 5 minutes after the begin laboratory, grains were decorticated to a ning and before the end of cooking. 75% yield on the TADD and ground into flour using two different mills so that a It was not possible to make small cylin flour with particle size close to traditional drical samples of bogobe for measuring or commercial flours were obtained: 30% consistency through a compression test as was ground in a Cyclotec helical mill it was in the case of to, since bogobe is (fine flour) and 70% was ground in a very thick and has a grainy texture. Cemotec mill (coarse particles). Therefore, an Instron extrusion test of the Twenty-seven grams of this flour mixture type used for rice or green peas was ap were poured in 60 ml distilled boiling wa plied to bogobe. Immediately after cook ter in a stainless steel container and placed ing the bogobe sample was transferred to on a hot plate (1500 watts, set to level 4 for an extrusion cell. As the probe was moved boiling and 2 for cooking). The mixture at a speed of 100 mmlmin, a resistance was stirred with a fork to avoid formation force was measured. The force increased oflumps. The container was covered with until the porridge went through the hole at a watch glass and the porridge was the bottom of the cell. Then the force re cooked for 20 minutes. The mixture was mained constant and this was the extru- 220 Table 6. Correlations between consistency of thick porridges and some physico-chemical character- IStICS 0 f t h e gram an d fIour 0 f 13 sorglh urn varieties Grain Flour Consistency Amylose Amylose Ash Lipids Proteins To +0.85** +0.84** -0.80* -0.84** -0.76* Ugali +0.84** +0.92** -0.80* • significant at the 5% level •• significant at the I % level sion force used as a measure of the consis Test for Evaluating Ugali Quality tency of the product (Figure 2): a thick and consistent product will have a higher The traditional method for making extrusion force. Three separate measure ugali is similar to that of to; however, ments were taken for each variety. ugali uses about 30% more flour. As with to, ugali is consumed during the two main This test was conducted on 9 varieties meals with legume sauce, meat or fish. (1995 harvest) obtained from the Bot Again the most important criterion for swana Department of Research, Ministry quality is product consistency. As a con of Agriculture and from ICRISAT/ sequence the laboratory method for mak BULAW A YO, Zimbabwe (Table 2). Ex ing ugali was similar to that for to but the trusion force varied from 16.2 N for hy flour/water ratio was changed and the ex brid BSHI to 30.5 N for the improved va trusion test was used as for bogobe since riety Kuyuma. The Newman-Keuls test the product is too thick to produce sam resulted in 6 groupings of the 9 varieties, ples for compression test. showing that the test is very repeatable and effectively distinguishes between va Sorghum grains decorticated at a 75% rieties. The test was then conducted on 30 yield on TADD were finely ground in a other varieties (1996 harvest) including Perten hammer mill. Twenty-seven 23 breeding lines from ICRISAT/Zimba grams of fine flour at 11.5% moisture bwe and the remaining 7 were local or im content were dispersed in 80 ml of cold proved varieties from Botswana. Two ex distilled water at 30% flour/water ratio, in cellent varieties had extrusion force a stainless steel container. The mixture greater than 26 N: SDSL89426 (26.2N) was cooked for 20 minutes in a water bath and 65D (30.1N) while two other poor at constant shaking speed (400 rpm). Fol quality varieties had forces smaller than lowing cooking ugali was immediately 16N: SDSH94012 (14.4N) and placed in an extrusion cell and left to cool Mmabai:tse (l5.7N). Varieties with force for one hour before the extrusion test was greater than 22 N were well accepted and conducted as described for bogobe. made a consistent bogobe. This test was conducted on 13 varieties from Sokoine Agricultural University of Tanzania, from the Department of Agri cultural Research of Botswana, and from 221 T a bi e 7 r o consis. t ency an d amYlose con t en t 0f26 sorglh um cu IfIvars f rom Mra I. T6 consistency Grain Flour amylose Homogeneous (N) amylose(%dw) (%dw) groups T6 qualitv >18 >22.5 >26 A Very hard 18-15 22.5-21.5 26-25 B Hard 15-12 21.5-20.5 25-24 C Medium 12-10 20.5-19.5 24-3 D Soft 10-7 19.5-18.5 23-21 E Very soft <7 <18.5 <21 F ICRISAT/Zimbabwe (Table 3). The ex larly the compression test which had a trusion force varied from 23.1 N for higher F-value and more groups by the BSH-I to 54.1 N for the local variety Newman-Keuls analysis. Langa-Ianga. The Newman-Keuls test classified the 13 varieties into 5 signifi By comparison, the two tests do not cantly different groups. Above 38 N ugali group the varieties the same way (Table was acceptable to consumers. Two 4); however the group of varieties that cultivars made a poor quality ugali oflow made a good to (higher than 12 N) also consistency: hybrid BSH-l and variety made a very good ugali (higher than 40 Mmabai:tse. N); on the other hand varieties that make a bad to (smaller than 8 N) also made a bad Comparisons Among the ugali (smaller than 35 N), while there was Three Laboratory Tests an intermediate group poorly defined (8-12 N for to, 35-40 N for ugali). The three laboratory tests on traditional thick porridges were compared for repeat Figure 3 shows the correlation ability, effectiveness in distinguishing (r=+0.84**) between to consistency mea among varieties and in grouping them. sured by compression and ugali consis The 13 varieties mentioned above were tency measured by extrusion for the 13 decorticated, milled into flour and used varieties. The result indicates that the two the same day to produce to and ugali. In tests are relatively similar even though both cases, porridge consistency was they evaluated products that are different measured after one hour. Analysis ofvari from a rheological standpoint. ance on consistency values at a 5% level of significance gave an F-value of 75.2 Table 5 combines to and ugali results with 7 homogeneous groups in the com from Table 4 as well as bogobe results pression test for to, and an F -value of 53.7 from Table 2, for varieties originating with 5 groups in the extrusion test for from Botswana. It appeared that these va ugali. Coefficients of variation were 6.7% rieties were grouped the same way except and 4.2%, respectively. These results for two varieties (Phofu for ugali and show that the two tests are repeatable Mmabai:tse for to). We therefore con since the coefficient of variation were cluded that varieties that make a good to small and close to 5% especially for the may also make a good ugali or bogobe, in extrusion test; the two tests effectively other words, varieties selected for good distinguished among varieties, particu- thick porridge in West Africa may have 222 the same porridge-making quality in East For porridges that are smooth and homo and Southern Africa. Thus, genetic mate geneous and are made from less concen rial with good culinary properties in West trated flours, like ugali and to, the Africa could be transferred to East or presence of peripheral parts like bran and Southern Africa and vice-versa provided germ that are rich in minerals, lipids and it is adapted to the local agro-climatic proteins, reduces porridge quality. In this conditions. These results may help breed case, a good ugali and particularly a good ers who are interested in varieties pos to will be obtained from grains that are not sessing agronomic and food quality at only rich in amylose but also have a good tributes that are acceptable to farmers and potential for decortication (to eliminate end-users. peripheral parts), and/or that are rich in starch (less minerals, lipids and proteins Physico-chemical Basis for Thick Por to start with). For porridges that are made ridge Quality with highly concentrated flour, like bogobe, the presence of other flour com To understand which characteristics of ponents does not appear to paste cohe the grain are related to quality of thick siveness and firmness. Only a high grain porridges, we conducted correlation anal amylose content is important to obtain a ysis between the physico-chemical attrib good bogobe quality. utes of grain and porridge consistency (Table 6). For 13 varieties originating These results were confirmed by run from Botswana, Zimbabwe, and Tanza ning multiple regression analyses on all nia, to and ugali consistency was posi variables and thick porridges. Using 13 tively and highly significantly correlated varieties, 96% of to consistency could be with amylose content of whole grain explained by two variables: grain (+0.85 and +0.84, respectively) and flour amylose (r2=0.72) and flour protein (neg (+0.84 and +0.92, respectively); for atively) which could as well be lipid or bogobe, with 9 varieties, the correlation ash content (same probability). Ugali and coefficients were +0.75 with whole grain bogobe consistency for 9 varieties could amylose, and +0.94 with flour amylose be explained by flour amylose content content. alone (84% and 89% respectively). Anal ysis of31 additional varieties showed that Whereas bogobe consistency was only amylose is the single most important vari correlated with amylose content, ugali able for bogobe consistency. For to con consistency was also correlated (nega sistency, the following equation that ex tively) with flour ash content, and to con plains 80% of the variability was devel sistency was negatively correlated with oped using 32 varieties from Mali: ash, lipids, and protein content. To consistency = + 11.28 + (0.33 x Hence the consistency ofthick porridg starch solubility)-(5.12 x ash in es may be linked to grain amylose, and flour) therefore to flour amylose content. The higher the grain is in amylose the better is If starch solubility was not included, its thick porridge (hard and consistent). amylose content was the next most impor- 223 tant variable that determines to consis amylose content, make excellent to and tency. Similarly, the variable ash in flour are not easy to replace with new varieties which has a negative effect on to quality that do not possess all these attributes. could be replaced with the variables lipids or proteins. Hence to consistency would On the other hand these results also ex be primarily determined by the quantity plain why commercial flours produced by of starch molecules dissolved during some 70 mills in Botswana are surpris cooking - notably amylose molecules - ingly so successful despite their lower which increase viscosity of the liquid quality compared to traditionally milled phase and therefore gel consistency. The flour. These commercial flours sold in su other component is flour purity (free of permarkets are acceptable to women who peripheral parts), which is directly related like the convenience, cheap price and to decortication yield and starch content bogobe quality. Brands that use local va of the grain. riety Segaolane or newly introduced vari ety Phofu, both high in amylose are pre Based on these results, our recommen ferred to brands that use mixture based on dations to breeders are: select varieties South African hybrids. Contamination by with amylose content greater than bran or germ which are unavoidable with 20-21 % which is equivalent to flour industrial or semi-industrial mills do not amylose content greater than 24% (Table constitute a constraint for the use of these 7); among these varieties, select those that flour by local people. Since the introduc decorticate better, are hard enough (PSI tion of these mills and the large diffusion test) to go through abrasive decortication of the produced flours from, women no or pounding and yield 'pure' flour; or se longer do not want to pound grain unless lect varieties that are low in ash (thin they live on a farm. pericarp), lipids (small germ) or proteins. These varieties should be acceptable to Conclusion consumers and make a good to. Three laboratory tests were developed These results may explain why women to predict the potential of sorghum breed in West Africa who prepare to on a daily ing lines in making traditional thick por basis are very selective in the flour qual ridges in Africa: to in West Africa (Mali, ity, why they prefer to wash the grain Burkina Faso), ugali in East Africa (Tan themselves, decorticate it with a mortar zania), and bogobe in Southern Africa and pestle, wash it again to make sure they (Botswana). These tests use small quanti get rid of bran and all of the germ, and fi ties of grain (20-35 grams at 11.5% mois nally why the majority of small-scale ture), are highly repeatable and efficiently mills that are less hygienic, less strict on distinguish among varieties. cleaning, on the extent of decortication and on storage, have very little success in These tests allowed the identification meeting local demand for good quality of groups of good, bad, and intermediate flour. Local guinea varieties which have varieties. Varieties that make good to may small vitreous and hard grain, thin make good ugali and bogobe and vice peri carp, good decortication yield, high versa. 224 Grain amylose content (>20-21 % dwb) Mestres C., Matencio F., Pens B., Yajid M., Fliedel G., 1996. A rapid method for the determination is the single most important variable for of amylose content by using differential scan thick porridge quality. For West African ning calorimetry. Starch/starke. 48, no. 1, ~-6. to, another characteristic is flour purity, or Fliedel G., Yaji d M., Matencio F., FrancalancI P., 1998. Mise au point de tests de laboratoire pour high starch content. This purity factor evaluer la qualite du sorgho - Relations avec may explain why West African women certaines caracteristiques physico-chimiques are more demanding for flour quality than des grains. In: Amelioration du sorgho et de sa culture en Afrique de l'Ouest et du Centre. women in Southern Africa. Ratnadass A., Chantereau 1, GIGOU J., 315 p. Actes de l'atelier de restitution du programme References conjoint sur Ie sorgho ICRISA T -CIRAD, 17-20 mars 1997, Bamako, Mali. Eds CORAD=CA. MontpelIier, France, p. 55-59. Fleidel G., Grenet c., Gontard N., Pens B., 1989. Fliedel G., Faure J., 1998. Development oflabora Durete, caracteristiques physico-chimiques et tory quality tests for sorghum and millet. Final aptitude au decorticage des grains de sorg~o.In: report. A component of the Collaborative pro Cereales en regions chaudes: conservatIOn et ject to investigate consumer preferences for se transformation. AUPELF-UREF, Eds. John lected sorghum and millet products in th.e Libbey Eurotext, Paris, France, p. l87-2~1. SADC region of Africa. Supported by the SCI Fleidel G., 1995. Appraisal of sorghum quality for ence and Technology of Development, making to. Agriculture et developpement. Spe Programme of the European Community, Con cial issue. December. Pp. 34-42 tract N TS3*-CST94-0267. 225 Senegalese Experience in Initiation and Strengthening Semi-Industrial Millet-Based Food Technology Plants in Urban Areas Ababacar N doye Background consumption in urban centers. One rea son rice and wheat are popular is that their Local cereals, including millet, are food products are readily available in ur grown on about half of the agricultural ban markets (i.e., processed grain, semo land and produce 800,000 to one million lina, flour, couscous, etc). tons of grain. However, half of the na tional consumption of cereals consists of Millet must be processed in order to be imported rice (400,000 tons) and wheat come an urban consumer product. Be (200,000 tons) at a cost of Fcfa 80-100 cause of this, the strategy in Senegal is to billion, depending on world prices and create economically viable, semi-indus exchange rates. This dependence on for trial processing units that give millet the eign food is very difficult for the Senega necessary momentum which would im lese economy and represents a loss to ru prove economic returns and agricultural ral producers and local food processors, marketing. since imported rice does not need addi tional processing. On the other hand, the Senegalese Experience in national objective is for the agricultural Semi-industrial Units for sector to grow by 4% yearly, improve Local Cereal Technology food security and increase income in rural areas, as well as create jobs. Towards this To start these units, it was recognized end, the government decided a strategy that technological development could not that includes supporting local cereals. As be limited to a simple transfer oftechnol oflast year, the government decided on an ogies, products, or equipments. Instead, agricultural improvement program to fa an aggressive policy of knowledge or cilitate farmer access to inputs, including technology introduction was decided and quality seeds, fertilizers, pesticides and reinforced in a joint program with farm equipment through small-scale CIRAD. The results led policymakers to credit to farmers. In practice, import of start a program named 'Dugub Sunu agricultural equipment is not taxed. Ad Cosaan' in 1989 on millet promotion. ditionally, several programs were started Products like 'Soungouf Jaboot' and to stimulate the consumption of local ce 'Sankhal Jaboot' were developed. The reals - such as millet - in urban centers. program was implemented by Moulins Sentenac, with support from ITA, to en Although urban consumers tend to pre sure quality. During this time, urban con fer rice and wheat, millet is by far the ce sumers were educated on accepting and real of choice for cultural or religious oc integrating these products in their diet. casions (lakh for birth ceremony; nakk for funerals; cakes for weddings, etc). As is In 1986 a different program was devel evident, millet and its derivatives are a oped by policymakers and conducted by main part of the culture for Senegalese the Commissariat it la Securite civilizations. This cereal is, therefore, the Alimentaire (CSA) to promote local cere best option to replace some of the rice als through semi- industrial processing 226 plants. CSA selected local businessmen It should be noted the development of and provided them with equipment (i.e., these semi-industrial units for millet pro decorticators, cleaners, grinders, sealers, cessing is occurring in a supportive envi etc), as well as basic products and training ronment with the existence of two pro (technical and marketing). The business grams that support local cereals, with men signed contracts with CSA to build funding from CILSS (PROCELOS) and the plants to house the equipment. The the European Union (PPCL). The two processing plants, some of which are still programs are designed to help with micro operational, produce flour (soungouf) and units, semi-industrial units, industrial semolina (sankha1), which are packaged units (candy, bread, cookie factories) and in polyethylene bags. The CSA program with food processors that are interested in resulted in the expansion of these local cereals (millet, maize, sorghum) and semi-industrial plants in urban centers, cowpea. which then challenged research for diver sification of their products. Conclusion In 1994, CRDI funded ITA to create Senegal has accomplished significant mechanical equipments for the process achievements in local technologies, ing of secondary products. Specifically adapted technologies, and promotion of the program was aimed at modernizing small and medium processing units in the production of arraw (particles made of volved with millet. This was possible millet flour) through the following objec with the help of research and develop tives: ment centers, the private sector, industry, and donors. • knowledge of the market size for millet arraw; The aggressive approach has resulted • development and testing a lab pro in positive developments including the cedure for preparing millet arraw; initiation of small and medium millet pro • conducting a pilot study of an arraw cessing enterprises that are technically making technology and optimizing and economically viable. The number of production; and, units and the quantity of products they market, as well as their financial returns • marketing before transferring it to have steadily increased through the years. private entrepreneurs. To sustain this tendency and reinforce This last stage is currently being imple the results, the semi-industrial entrepre mented with five enterprises, four of neurs decided to organize in a which are in a six-month testing period. twenty-member for profit association These enterprises have agreed to reim 'Groupement d'interet economique'. burse the cost of setting-up the technol ogy (particle maker, equalizer, gas dryer, Much has been accomplished in Sene sealer, balance) in five years. gal to initiate small and medium scale processing plants for local cereals but The need for a better quality product much is left to do. Scientists and entrepre led to research for other equipment to neurs face the challenge of using their ini clean the starting material, including tiative and good will to make contribu stone removers and decorticators. Since tions to food security. 1998, this is being done with the assis tance of ROCAFREMI-P5 project. 227 Contribution of PROCELOS to the Marketing of Food Products in Sahelian Countries Mamadou Diouf Introduction Greetings from Madame the Secretary • support new product development General of CILSS and myself to the orga in cereals, fruits and vegetables, an nizing committee of the workshop, espe imal and fish products cially to INTSORMIL for understanding • provide information on nutritive that millet/sorghum production and utili value and product quality zation are extremely important in Sub-Saharan African, and more so in the These activities are preferable to vari Sahel. ous restrictions on imports, by creating conditions under which local processors It is indeed clear to the organizers that can offer quality products that are attrac objectives can be met only through the in tive to consumers. volvement of all partners from production to utilization, in such a way that consumer Sahelian countries recognized as early needs are satisfied. This approach is cer as 1980 that the rules of international tainly one of the best if results of agro trade and solidarity on one hand, rapid ur nomic research are to be useful, so that va banization and demand for alternative rieties should be perceived as raw rather products on the other hand, require that than finished material. The area of food food products target city dwellers. Thus quality and increased utilization is hence CILSS decided to do something in the uti one for which CILSS/PROCELOS and lization area. The strategic objective is INTSORMIL can join their efforts. defined as 'a significant share for local food products in the consumer habits' Objectives and Approach while the short term objective of the past ofPROCELOS three years was 'the increase of certain processed foods and their utilization'. CILSS in its major program on food se The following expected results were set: curity started the regional program called PROCELOS to support local food prod • the most competitive processed ucts and the utilization sector. Activities foods which are better known and include: more attractive to consumers • the socioeconomic and legal envi • promotion of increased consump ronments for improved food prod tion ucts • improved knowledge of product • the investors in food products are utilizations better organized to defend their • transfer of improved technologies own interests • an improved product quality 228 Major achievements by PROCELOS est for food processors and other partners in food security. The project acts mainly as a facilitator • a higher awareness of the impor between interested parties but a few times tance of quality. Past results and it has worked alone. Some key results are those being evaluated in some as follows: countries will lead to adoption of sub-regional standards, quality la • food products developed by re bels that will be presented to re search and other partners are better gionalleaders and national govern known. Semolina, flour, couscous, ments so that quality requirements infant foods, all in better presenta of urban dwellers are satisfied. tion are now more appreciated by consumers as can be seen in super Problems markets, neighborhood shops, drugstores, and even in some devel The principal ones are: oped countries. • improved market for food product • need to develop better entrepre (increased sales and sometimes in neurs through training and informa capacity to satisfy demand). tion exchange; • development and distribution of cu • produce concise information on re linary books whose positive effects search results that are readily usable on consumption have been docu by food processors in collaboration mented in Senegal. F AO has re with researchers; quested some ofthe recipes for their • a better management of quality in INPHO program on the Internet. processing plants; • discussions on packaging have led • a better institutional environment to the finding of solutions being for food processing enterprises to currently implemented in some expand and improve their products; countries while long term solutions • a system of credit more favorable to are been evaluated for the food processing specialists; sub-region. • the institutional environment, eco Conclusions and Recommendations nomic and monetary, was studied before the project started and in Food deficits have been reduced in part some countries suggestions for im due to imports and food aid, but modified provement have been made. food habits have come to exist because of • National PROCELOS offices have the joint impact of colonization, drought, helped food specialists to organize rules of international trade, and increas themselves around cereal process ingly because of urban living conditions. ing and others. • more professionalism in the activi Food security issues will be more and ties of various partners has been more applicable to urban centers and con promoted by the advisory offices sumer preferences in rural areas will be set up by the project. modernized. This gives a high priority to • at the national level, bulletins are the development of local food products. published to discuss topics of inter- 229 This may be the best way to increase pro though not always known by food spe duction, because a breeder will target cialists. This is an important development each variety to a specific use. This use because at PROCELOS we believe that a will correspond to an identified need, or long-term solution to the problems of better yet to what consumers demand. food security has to involve local prod ucts, because one can hardly win a war Researchers have interesting results with someone else's weapons. Thank you that may satisfy food requirements al- for your attention. 230 Factors and Recommendations Concerning Seed Production Lee House Introduction evaluated at several locations in the coun try. Hybrids averaged 61 % increase over Pearl millet is the most important crop locals with irrigation and 49% when in the country being grown on an esti rainfed. The advantage of hybrids over mated 4.8 million hectares. Sorghum is parents was 45% with irrigation and 66 % the second most important crop being when rainfed. Additional testing contin grown on just under two million hectares. ued to conform the yield advantage ofhy Yields are low, ranging in the area of350 brids across locations and under both, irri to 400 kgihectare, but the crops are virtu gation and rainfed conditions. In 1989, ally always produced in hot, dry rainfed the hybrid Tx623A x MR732 was entered conditions and on soils depleted ofnutri in regional sorghum trials. It ranked third ents. Agriculture is confined to the south among twenty entries including local ern end of the country primarily below checks at nine loctions in Cote d'Ivoire, 12° latitude. A very small percentage of Burkina Faso, Cameroon, Nigeria and the sorghum crop is irrigated. Niger. Experimental hybrid yield (1984-1992) averaged approximately 2 Sorghum Research tonsiha. The best hybrids yielded up to 6.5 tons/ha where the best varietal checks Varietal development in Niger was un never exceeded 3.5 tons/ha. dertaken in the 1950's. Several varieties were released. Hybrids were introduced In 1991, the hybrids Tx623A and but discontinued because of poor seed MR732 were released as NAD-l. Since quality. its release, the hybrid has been grown by numerous farmers where it has received About 15 years ago, INRAN in collab an enthusiast response but has also raised oration with INTSORMIL, developed ex concern about the availability of seed. As perimental hybrids for evaluation in the a hybrid, NAD-l could be used to facili country. Promising hybrids were also tate the initiation of a seed industry in the evaluated in the West Africa Sorghum private sector. Establishing a seed indus Hybrids Adaptation Trial (WASHA T). try will ensure the possibility that farmers Hybrids are still being developed from can benefit from hybrids. this cooperation and are being evaluated in the country and the region as an on go Pearl Millet ing research activity. Pearl millet is the most important ce In 1986 systematic studies on cultivars real crop in Niger. It is grown on an esti including hybrids, their parents, and mated4.8 million ha with average yield of adapted local checks were initiated and 380 kg/ha and 1.86 million tons of pro- 231 duction. All of the improved and recom ICRISAT-Niger is a major breakthrough mended millet varieties are and opens up enormous opportunities for open-pollinated varieties bred by breeding of superior hybrids. ICRISAT, INRAN, and more recently by the in collaboration with national programs ICRISAT Sahelien Center. However, the and the millet network (ROCAFREMI), area occupied by these improved varieties has been able to demonstrate progress in is very limited. the following: The discovery of cytoplasmic • Development of adapted seed par male-sterility (CMS) in pearl millet and ents (male-sterile line) for use in the breeding and release of Tift 23A in the hybrid development. early sixties made the production of sin • Demonstration of hybrid superior gle cross hybrids (SCH, male-sterile x in ity for grain yield and disease resis bred) possible in India and has contrib tance using top-cross hybrids (A uted to spectacular yield increases. Al line x open-pollinated varieties). though the male-sterile Tift 23A was • Hybrids performed better than open-pollinated varieties under made available to breeding programs in moisture and fertility stress, the two West Africa in the early sixties, pearl mil principal contraints for millet pro let hybrids were not extensively tested. duction in Niger. Tift 23A was not adapted to the environ • With the availability of male-sterile ments of West Africa: early flowering, lines there is now an opportunity to high susceptibility to downy mildew, ab combine adaptation of the land sence of early vigor resulting in poor races to the prevailing low mois emergence and stands, and occasional ture, low soil fertility conditions breakdown of sterility. Because of these with an improved yield potential. problems, extensive resources were not Hybrids appear to be the easiest devoted to breeding SCHs, but only ex route for a rapid and effective incor ploratory research was directed towards poration of tolerances to stress fac inter-varietal or top-cross hybrids (TCH, tors and increase in yield potential. male sterile x open-pollinated variety). The landrace variability in pearl Limited amount of research on peral mil millet within Niger captures about let hybrids was caried out in the 80% of the variability present in the mid-eighties in Senegal, Mali, and Nige World Collection. na. • Studies on hybrid multiplication plots indicated that the nick (syn chrony in flowering between the Without exception, it was concluded male sterile and pollinator parents) by every national program in W. Africa was optimum, with pollinator par that unless adapted male-sterile lines ents flowering one week to 10 days were available, it would be difficult to de before the male-sterile line. Tiller velop and assess the potential of pearl ing was asynchronous in the female millet hybrids in the region. Therefore, line, with a 10-20 days interval be the breeding of adapted male-sterile lines tween stigma emergence on the and development of top-cross hybrids (A main tiller and secondary tiller. The line x open-pollinated variety) at 232 high tiller number plant assured that Opportunities in Niger for Commer enough pollen was available for all cial Production of Hybrid Seed the female earheads. • At a female to male ratio of 4:2 (ap Experience to date indicates the impor pears optimum), the estimated hy tance of irrigation to ensure simultaneous brid seed yield is 620 kg ha-I and of flowering of both parents and of good I the pollinator 700 kg ha- . Our quality seed yield. One can rationalize 'back-of- the-envelope' calcula three categories of seed producers.: tions show that profits between $650 to $850 can be expected from One group consists of individuals with a hectare of hybrid seed production reasonably large land holdings ranging (assuming 200 Fcfa for a kg of from 50 to 300 hectares. Opportunities for grain, 800 Fcfa for a kg of hybrid irrigation generally do not exceed 30 seed and 400 Fcfa for the pollinator hectares and are usually less. Owners of OPV seed). these farms almost always have employ • It has to be noted that production of topcross hybrids permits harvest of ment with the Government or private two seed crops from one field- that business. of the hybrid and of the pollinator variety. Seed producers can sell the The second group consists of coopera seed of the pollinator variety which tives in the irrigated perimeter. There are will generally have better seed numerous irrigation schemes in the south quality assuring an increase in the ern part of the country mostly along the value of product from the male par Niger River (Birni N'Konni, Maradi, ent rows. Diffa) which are referred to as irrigated • ICRISAT has gathered extensive perimeters. These are managed by pri data on the requirements of vate cooperatives with technical supervi topcross hybrids, and the need for sion from ONAHA (Office National des durable disease resistances for Amdriagements Hydro Agricoles). Hold downy mildew. A Regional Pearl ings by individual farmer members of the Millet Hybrid Trial has been initiate cooperatives are small, ranging from one in collaboration with the West Afri third to just over one hectare. Farmers re can Millet Research Network. It is ceive support for input such as fertilizer hoped that results from this trial will help in the identification of su and irrigation water and they sell their perior hybrid combinations. produce through a cooperative, which en • There will be a need for extensive ables the co-operative to recover the cost training in seed production. Stu of the inputs. Hybrid sorghum seed has dents and technicians from Niger been very successfully produced by farm have been trained in the production ers on several irrigation perimeters. To and evaluation of male-sterile lines date, seed has been sold within the perim and hybrids. eter but the opportunity to sell outside of the perimeters is there. The third group is small farmers who farm on dry land. Initial efforts with them 233 failed but there may be an opportunity to ganization for production of hybrid seed. produce hybrid seed, particularly in the The initial effort has been to stabilize seed rice growing areas where irrigation is pos production and identify individuals inter sible. Presently, these farmers are produc ested in seed production. This exercise ing vegetables and onion seeds. This has been supported by several INRAN group could be considered for contract staff and consultants. In 1995, 750 kg of seed production. NAD-l seed was produced, which in creased to 1400 kg in 1996, 7100 kg in Overall, the capacity to produce hybrid 1997. Based on visual observation, it is seed of sorghum and pearl millet is there. estimated that approximately 26,000 kg Also to be considered is the farmers' en of NAD-l seed was produced in 1998. thusiasm to produce seed, given the good The number of producers also increased monetary earning of farmers who pro from 6 in 1995 to 43 in 1998 with private duced seed ofNAD-l. producers increasing from 3 to 15, and in the cooperatives from none to 24. INRAN has and is involved in the pro duction of seed of NAD- 1. This is ac New producers have been encouraged cepted as an interim contribution to pro to limit their first effort to no more than vide seed and demonstrate techniques to one halfhectare because oflack of experi farmers and potentially interested seed ence. There are concerns that the seed in producers. Already the area under N AD- I dustry is not adequately understood. The seed production is greater in the private current interest in the development of a sector than in INRAN. Even though seed unit at INRAN, and formation of a INRAN is taking responsibility to provide seed producers association might provide foundation seed, three private individuals capital to start a business. Banks and some are producing seed ofTx623A this year. agencies are reluctant to lend, not recog nizing the difference between seed and There appears to be no constraint on gram. the opportunity for private production of breeder and foundation seed of Training INRAN/INTSORMILlICRISAT devel oped hybrids. Two one day training program were or ganized in 1997 involving 50 individuals. At the present time demand for seed of A training manual was made available by NAD-I greatly exceeds supply and since INRAN/INTSORMIL. Prior to the plant 1995 NAD- I seed has sold for 700-800 ing season, six training courses were or Fcfa/kg which is 4 to 8 times the grain ganized involving 250 individuals. Be pnce. cause of the high demand for the training manual, there are now plans to expand it Getting Started by including more information on man agement of a crop for seed, pathology and INRAN is a government reseach entomology. agency with limited capacity to produce and distribute seeds. There is really no or- 234 The possibility of establishing a train ment of a seed industry in the private sec ing center has been discussed and plans tor. made but these have yet to be imple mented. Also under discussion are We should also anticipate that with the expandsion of the training program to in increased production of sorghum and clude more hands on contact, farmer pearl millet grain, we must find means on training and field demonstrations, and to how this grain can be utilized for food and establish a library with information rele feed. vant to seed production, standards and certification, seed processing equipment, Objectives seed storage and processing plant require ments. • To continue research to develop sorghum and pearl millet hybrids The Near Future with a range of maturities to satisfy short growing seasons such as Currently new sorghum hybrids are be Diffa, and longer growing seasons in the South around the Bengou ing developed by the IN RANI area. It is also important to have hy INTSORMIL collaboration, and pearl brids with better grain quality, and millet hybrids are being developed at the for sorghum to have hybrids with a ICRISAT Sahelien Center. lower susceptibility to long smut. • To expand experimental produc The establishment of a varietal release tion of pearl millet hybrids so that committee should be considered to evalu several tons of seed are available ate cultivar superiority and/or adaptation for on farm demonstration. The in different environments. Such a com purpose would be to gain mittee could also give balance to the num enthusiatic response from farmers bers of cultivars released. to create a demand for seed indus try. To date, the effort has been to stablilize • To process seed which has been seed production and to identify and train certified through a voluntary certif seed producers. The program is or soon ication program. will be at a point when other consider • To expand training and educational ations such as seed processing, imple opportunities to seed producers and menting existing seed laws, establishing a their managerial staff and to the program of voluntary seed certification, farmers. The training would in and training in the marketing of seed will clude production, processing, qual ity control, marketing, and crop uti become important. It is to be noted that lization. the Ministry of Agriculture is only casu • To establish a training center that ally aware of the effort to establish a seed could serve Niger and other West industry. Now that this opportunity ap African countries, which would pears to be moving forward, it would be serve as a resource to seed produc useful for the Ministry to develop a policy ers in becoming established and that recognizes and encourages develop- maintaining their facilities. This 235 could be a function of the Seed • Two INRAN breeders (other scien Unit. tists could be invited to meetings • To establish an independent Seed where they could contribute) Unit or national seed service within • INRAN research station directors IN RAN which would handle a • Two extension agents range of functions. The seed unit • Two private seed producers, one should have funding to effectively should represent the Seed Pro support and encourage the future ducers Association establishment and growth of private • ONAHA representative sector. • ICRISAT representative • To conduct training programs for • Two or three farmers seed producers and their staff, pro vide technical support to new pro Presently, INRAN has approved the ducers, contribute to the implemen seed unit to be part of its organization, but tation of quality control programs it is yet to be approved by the Ministry. (serve as the operational unit in ex This needs to be resolved so that develop ercising the stipulation of the seed ment of the unit can proceed. law), help reactivate the seed lab and mini seed labs associated with • To encourage the establishment of a seed processing facilities estab Seed Producers' Association. Once lished by the previous program established, the Association would (1976-1986), and assist with the es implement the seed certification tablishment of a Seed Producers' program, manage funds both from Association. The seed unit could be participants and donors to provide a resource of technical information capital on a loan basis to help new relevant to the seed industry and at producers get started, and conduct tract outside expertise (consul seminars involving banks, NGOs, tants), interact with donors to sup Ministries, and others on topics rel port seed activities in Niger, and in evant to the seed industry, market teract with the government on seed ing, and storage. The Association related issues. could also encourage good business practices with strong emphasis on It is suggested that the seed unit have an seed quality, interact not only with oversight committee, comprised of the government on issues related to arrpoximately 20 members, that would the seed industry but also with meet twice a year. The composition of the internaitonal seed companies with committee would be as follows: interest in Niger. It should also pro vide opportunities for short-term • INRAN Scientific Director overseas training with existing seed • Senior staff of the seed unit (the agencies. Another proposal would head of the seed unit would chair be to support an 8 to 12 member the committee) study tour of 3 to 5 countries with a • Member of the liaison unit between well-established seed industry. The INRAN and Extension study tour would include several private seed producers, the Minis- 236 try, a member of a financial institu • Crop management should be tion, and INRAN specialists. strengthened - hybrids can be used • On-farm demonstrations should be as a vehicle for this purpose. Dem encouraged to show gains in yield onstrations are useful for farmers to from better crop management since gain appreciation for better crop hybrids are generally more respon management. sive to management inputs. • The government needs to establish • While the effort to produce hybrid a seed policy indicating their goal is sorghum seed is still modest, sev to establish a seed industry in the eral problems have been encoun private sector with profit and com tered. Rain on harvested panicles petition. increased grain moisture which re • The establishment of a seed unit at duced germination while in storage. IN RAN was appreciated and Quality seed drying facilities at sev should be kept broad based. The es eral sites in the country are urgently tablishment of a Seed Producers needed. Association is encouraged. • There is need to explore opportuni • There is need for continuous train ties for small-holder farmers on the ing activities for scientists, exten irrigated perimeters to produce hy sion agents and farmers. It is also brid seed. important to provide opportunities for prospective seed producers, pol Resolutions of the Working Group on icy makers, and financial managers Seed Production to attend training sessions, semi nars and topic meetings on aspects • It is suggested that a project be of the hybrid seed industry. funded for at least a 10-year period • Expand the training program to in which could be divided into two clude more hands-on opportunity. 5-year units while establishing a Expansion of the training manual seed industry in the private sector. was also recommended. Such a project would provide the • Expand Lossa as a Seed Production opportunity to focus staff on the and Training Center under various aspects of the seed industry INRAN's management. and to capitalize on the expertise of • The seed law should be imple individuals with experiences in mented to regulate seed quality. other countries. Statues ofthe seed law should be at • There should be a continuing input tainable to protect the honest into research from which there will seedsmen. be a useful flow of genetic material • Seed certification should be volun to farmers through demonstrations tary to provide the opportunity for and involvement of Extension brand names. Agents. • Financing seed producers is a key • Concern was expressed that oppor issue to be addressed by govern tunities for scientists should be im ment through the banks. The differ proved to encourage them and re ence between seed and grain must tain them in their positions. be clearly understood. 237 • Certification for pearl millet top • Seed law and seed certification offi- cross hybrids require their own reg cer ulations to accommodate greater • Engineer variability. Certification for single • Training officer cross pearl millet hybrids would • Grain utilization scientist follow normal requirements. • Up to 20 technicians to support the • Good seed storage is recognized as growing seed industry across the a necessity. This storage could in southern region of the country clude keeping several hundred kg • Consultants in seed quality, mar of seeds of good adapted cultivars keting, and utilization. It is ex that could be used for off-season pected that these consultants would seed production following a year of be hired on a recurrent basis. crop failure. • Secretariat and other support staff • A variety release committee to con • It is also suggested that up to eight sider both varieties and hybrids was individuals should be trained over recommended. seas, and approximately 70 individ uals to be provided short-term train Staff ing in-country which would include study tours. Staff which could be part of the seed • Provide start-up capital for new unit would include: seed producers through the Associ ation. Seed production, processing and overall project manager 238 Pilot Project Report on Sorghum and Pearl Millet Utilization Rapporteurs: I. Akintayo and J. Ndjeunga The Utilization group led by M. Diouf, Constraint 2 first reviewed the constraints to the ex pansion of a pearl millet and sorghum The inconsistency of grain supply as food processing industry. Then, made well as the high variability in both the recommendations on lifting those con supply and grain prices could discourage straints. The following constraints were the investors from entering the food pro reviewed: cessing industry. • Low grain quality Constraint 3 • Inconsistency of grain supply and price variability Equipment currently available in the • Inadequate food processing equip West and Central African region is not ment adapted to sorghum and pearl millet pro • Lack of training and poor informa cessing. In a recent WCAMRN meeting, tion flow this lack was identified as a primary con straint. Poor packaging of the final prod Rationale for Identifying Those Constraints uct was also found to attract less consum ers. Constraint 1 Constraint 4 A successful pearl millet and sorghum food industry is contingent on high qual Processors are not well trained in ity grain both for its physical purity as state-of-the-art food processing technolo well as its suitability to making the gies. Therefore the products are often of end-products. Grain cleaned of physical poor quality. Also, lack of organization impurities, uniform in size, etc., is re among food processors reduces their abil quired because it is easy to process. Oth ity to collaborate with their partners such erwise processing costs will be high and as farmers, researchers, and deci the quality of the final product may be af sion-makers. Sharing information be fected (sand, ash content and high lipid tween the networks and INTSORMIL is content). Moreover, the consistency of also necessary to avoid duplication. the grain varieties with food processors and consumers' preference are important. Recommendations There is also a need for standardization of grain quality in the sub-region. In order to ensure good quality grain, optimal contractual arrangements be tween producers and food processors are 239 required whereby producers are offered and/or pearl millet. However, they should premium for grain quality and food pro be made available to food processors at cessors are being supplied grain of high low price and at the same time improve quality. Moreover, breeders should de the performance of existing equipment. velop varieties that are suitable to food processors' needs as well as consumer's On the one hand, facilitate and ex preferences. Breeders should be involved change information between networks at the early stages of the variety develop (WCASRN and WCAMRN) and ment process. Grain quality in the INTSORMIL, and on the other, provide sub-region must be standardized (this is support to food processors and promote currently being done through the good collaboration with researchers. WCAMRN for millet). Finally, as an ex ample, the hybrid seed offers the opportu Train food processors in better food nity of increasing production of uniform processing techniques, small-business grain (size and color) that are necessary to management, and marketing promotion facilitate processing and to develop high techniques. quality food products. Provide support in promoting pro Existing equipment used for process cessed products via advertisements so as ing wheat should be adapted to sorghum to be perceived as "new products". 240 Processing Sorghum and Millets into Value-Added Products Adam Aboubacar and Bruce Hamaker Introduction Table 1. Products with Commercial Poten tial. Although sorghum and millets are the Primary processing products major staple foods in many African coun Decorticated grain tries, commercial products made from Flour them are rarely available in urban and ru ral markets. In contrast, imported wheat Grits and com products are readily available, Secondary processing products particularly in urban centers. In many Af Couscous rican countries, several attempts are now Snacks being made towards processing sorghum Composite flour for and millets into commercial products. Weaning foods Long term goals of such processing pro Bread and biscuit jects are to reduce food imports in sor Pasta and noodles ghum-and millets- producing countries, Parboiled grain to benefit local entrepreneurs by giving Flat bread them value-added products, and to give Popped grains farmers another outlet to sell their surplus Beverages grain. are used in the manufacture of weaning Table 1 lists some sorghum and millets foods and baked products such as bread products that have commercial potential. and biscuits. New and improved commer Because of its many different uses, flour cial products that can be made from sor constitutes the single most important ghum and millets include pasta, noodles, product that can be made from sorghum chips, parboiled products, and beverages. and millets. Flour is most commonly con Studies at Texas A& M University and at sumed in the form of thick and thin por CIRAD (Centre de Cooperation Inter ridges. Traditional processing of flour nationale en Recherche Agronomique into couscous is a very common practice, pour Ie Developpement) have shown that particularly in the Sahel region. Later in it is possible to make acceptable pasta and this report, we will present our experience noodles using sorghum/wheat composite in setting-up a small-scale cous flours. cous-processing unit at INRAN. Other products with commercial potential in Small-Scale Couscous Processing Unit clude composite flours prepared by sub stituting parts of sorghum/millets flours Couscous is a steamed product made with legumes (cowpea, soybean, peanut) from cereal flours. Millet, sorghum, or wheat flours. These composite flours maize, and fonio couscous are common in 241 West Africa, whereas durum wheat cous processed into couscous using a labora cous is found predominantly in North Af tory procedure (Galiba et. aI. 1987) and rica. Couscous production is a laborious compared to a laboratory-prepared and time-consuming process. Production (Debbouz et. aI. 1994) and a commercial steps include milling grain into flour, durum wheat couscous. mixing flour with water, agglomerating the flour-water mixture into couscous Couscous Yield granules, steaming, and drying. The most critical step in the processing of couscous Sidibe et. aI. (1981) reported that, in is the flour agglomeration step. Tradi Mali, the most important couscous qual tionally, agglomeration is done by hand ity criterion is the yield of the final prod rolling ofthe flour and water mixture until uct compared to the original flour. Cous couscous granules are formed. Another cous yield is highly dependent upon the traditional procedure is to force the ability of the agglomerated flour particles flour-water mixture through a sieve to to retain their integrity when steamed produce couscous granules. This step has (Galiba et aI, 1987). In our study, we been partially mechanized for small-scale found that hard grain produced flours processing in West Africa with the fabri containing high proportions of coarse par cation oftwo types of flour agglomerators ticles and damaged starch. In addition, (Richard, 1993; Anonymous, 1995). we found it yielded a higher proportion of desirable sorghum couscous granules. With funding from INTSORMIL, a Cultivars P721 Q and P851171 yielded small-scale couscous-processing unit has higher proportion of small undesirable been recently installed at INRAN for the couscous granules. These two cultivars purpose of research and demonstration to also had the highest ash content and high entrepreneurs, NGOs, and cooperatives. est proportion of fine flour fraction. Since The unit can also be used to produce flour ash is often used as an indicator of the and grits from locally grown cereals. The presence of bran in flour, the low yield of unit is comprised of several main pieces desirable couscous particles from these of equipment: a grain decorticator, a flour cultivars may be due to competition for mill, a flour agglomerator and a solar water between starch and bran, resulting dryer. All this equipment can be found or in less water available for starch fabricated in the West African region. gelatinization during couscous manufac ture. Studies on Sorghum Couscous Couscous color Research was undertaken at Purdue University to determine the fundamental Couscous made from sorghum vary grain and flour properties that determine widely in color from dark brown to sorghum couscous quality. Grains from creamy-white depending on the type and eight sorghum cultivars grown in West amount of pigments present in the flour Lafayette, IN were decorticated in order (Galiba et. aI., 1987). Millet couscous to remove 20% of the outer layers of the was reported to have a darker color than kernel then milled into flours. Flours were sorghum couscous (Galiba et. aI., 1987), 242 though this too depends on the cultivar lighter and had more yellow color than used and flour preparation. All sorghum sorghum couscous. Among the sorghum flours used in this study appeared white cultivars used in this study, couscous after milling. A variety of colors ranging made from SEPON82, IRA T -204, and from brown to yellow were obtained NAD-l gave color comparable to that of when the flours were processed into cous durum wheat couscous. cous Changes in color became obvious as soon as flour was mixed with water. Couscous texture These changes were accentuated during steaming and drying. Wet couscous ap Couscous of good quality should ab peared lighter in color than dry couscous. sorb sauce well and must have uniform These changes in color are likely due to particles that neither stick to one another, the presence in the flours of various phe nor lose their integrity when steam or nolic compounds which have been re sauce is applied (Sidibe et. ai. 1981; Kaup ported to be responsible for the colors and Walker, 1986). Significant differ found in sorghum-based foods (Hoseney ences in couscous stickiness were found et. aI., 1981). A sensory study conducted among the sorghum cultivars (Figure 1). by our group in Niger revealed that con Cooked sorghum couscous stickiness was sumers accepted a wide range of sorghum positively correlated (r = 0.89, P < 0.01) couscous color, although there was a clear with the amount of damaged starch in preference for white- and yellow-colored flour. A low molecular weight wa couscous. Durum wheat couscous was ter-soluble branched starch molecule ex- 5 -'Ii! ~3 I ~2 iI!, IJ , a Sl ;: ...... ~ iii' ~ «co N ",,p 1'0 e « 173 z ~ a... i .. III .. .. !Ii"" l!I a- - e '"III 0 ~i S 8 :; e. CultiYars Figure 1. Stickiness of sorghum and durum wheat couscous with 0 and without • addition of oil. 243 tracted from sorghum couscous was also Effects of decortication rate on sor found to significantly (r=0.86, P < 0.01) ghum couscous yield and color correlate with couscous stickiness. Sor ghum couscous was stickier than durum The sorghum grains were decorticated wheat couscous. There was less variation to remove 10, 20, 30, and 40% of the ker in couscous hardness than in stickiness nel and milled into flours, which were among the sorghum cultivars (Figure 2). used to produce couscous. Significant ef Cooked couscous hardness correlated fects of decortication rate (percent kernel positively (r = 0.79, P < 0.05) with appar removed) on couscous particle size distri ent amylose content of flour. Addition of bution (Figure 3) and color were ob 2% oil to the cooking water considerably served. For example, at 10% kernel re decreased the stickiness and hardness of moval, high amounts (30-55 %) offine « some sorghum couscous to a level compa 1 mm) couscous particles were produced; rable to durum wheat couscous (Figures I whereas at 40 % kernel removal, greater and 2). amounts (10-40 %) of large chunks (>2 mm) of couscous particles were obtained. Couscous yield of particles (1-2 mm) was highest (60-90%) at 20 and 30 % kernel Figure 2. Hardness of sorghum and durum wheat couscous with 0 and without. addition of oil. 244 1.5 l~: ~ 1~4J ~ o •• 0.5 -= • 20 -< ~ 0 0 10 20 3J 4J % hemel removed 1_<1mm c:::::J1·2 mm c:::::J>2mm -o-.Ash I Figure 3. Effects of decortication rate on couscous particle size distribution and flour ash. removal, depending on the cultivar. Flour and couscous lightness significantly im proved as decortication rate increased. Debbouz, Dick, 1., and Donneley, B. J. 1994. Influ ence of raw material on couscous quality. Ce References real Foods World 39:231-236. Sidibe, S., Diarra, M., and Scheuring, J. F. 1982. Richard, M. 1993. Fabrication des produits roules it Sorghum couscous: Quality considerations. base des cereales tropicales: Mecanisation du Pages 110-112 in: Proc. Int. Symp. Sorghum malaxage et du roulage. Memoire, DESS. Grain Quality. L.W. Rooney and D. S. Murty, Universitede Montpellier II: Montpellier, eds. ICRISAT: Hyderabad. France. Hoseney, R. C., Varriano-Marston, E., and Dendy, Anonymous. 1995. Projet transformation du mil, D. A. V. 1981. Sorghum and millets. Pages rapport n02. Institut de Technologie 71-144 in: Advances in Cereal Science and Alimentaire (ITA), Dakar, Senegal. Technology, Vol. 4. Y. Pomeranz, ed. Am. Galiba, M., Waniska, R. D., Rooney, L. W., and Assoc. Cereal Chern.: St. Paul, MN. Miller, F. R. 1987. The preparation of sorghum Kaup, S. M., and Walker, C. E. 1986. Couscous in and millet couscous in West Africa. Cereal North Africa. Cereal Foods World 3 I : 179-182. Foods World 32:878-884. 245 Panel Discussion Bruce Hamaker Human Utilization Day 3 - Session 1 Hamaker to panel: What does it take to Wolf: The biggest constraint is contin make a successful commercial product? uous supply of good quality grain to the What are the limitations, constraints and industry. successes from the producer, technology and marketing approach. Hamaker to Biago: In your experi ence, do you need to have a contract with N'Doye: A major constraint is the cost your suppliers or do you get your grain of raw material. It should be low enough supply from the market. in order to produce a product that is af fordable to consumers. Second important Biago: At the beginning, we used to point is that both researchers and proces buy our grain from the market. We did not sors need to work together to define spe have a particular contract with farmers. cific uses for each sorghum or millet vari First we tested our product to determine if ety. We need to have specification of raw it is acceptable to consumers. We had to materials, similar to the one used in wheat work with local women because they utilization where soft and hard wheat va knew the type of flour or products con rieties are used in baking, whereas very sumers want. We did not have pure variet hard varieties are used in pasta produc ies available. We used whatever we found tion. These are approaches currently be on the market. It was not until recently ing used by ROCAFREMI and that I had a contract with a farmer who PROCELOS. produced on 75 ha land, which is still not enough raw material for me. Although I Jupiter: First we need to have varieties am satisfied with this producer, I still need suitable for each end use. Second there is to get the grain with my former suppliers a need to have a consistent supply of good who themselves buy the grain on the mar quality sorghum and millets grain to sus ket. I think that many farmers will be mo tain a food industry. The other problem is tivated and produce more grain if they are that high grain price may discourage in assured of this type of contract. Farmers vestment. These are the points we need to do not produce a lot of grain because there look at from the production side. is not an outlet for the surplus. We as pro cessors need to give confidence to farm Hamaker to Wolf: Looking at it from ers. There is also a need to do quality and the entrepreneur (processor) side, what do nutritional tests on the final product. West you feel as a breeder and processor the African consumers are very selective limitations are ? What do you need to when it comes to quality. It is true that have to get the product you want? color plays a role but it is not the deter mining factor. Texture is more important. 246 Berthe: I want to add that grain quality sand. We also need good equipment for is very important. In Mali we had contract small-scale processing. There is a cereal with farmers specifying the type of grain processing company called F AMA that we want for a particular product. We also uses good equipment to make rice, maize made sure that we had continuous supply and sorghum couscous. In contrast, we of good quality grain, which made it pos use traditional methods to produce cous sible to operate the processing units year cous. Despite these problems, we produce round. and market about 100 kg sorghum cous cous per day. Another major problem Fliedel: I agree that grain quality and with our operation is drying. It takes three supply are important. At CIRAD, we days to sun-dry the home-made couscous came up with a laboratory procedure to and because of that we face fermentation manufacture pasta using 100% maize. We problem. And to avoid fermentation, we did some food quality trials in Senegal are obliged to reduce production to about and people liked the product. A major 25 kg couscous per day. With this level of problem was that of getting the supply of production, we cannot meet the increas maize from Casamance where it is pro ing demand for couscous which I should duced to the capital Dakar. The price of say is now sold in several supermarkets. maize was so high that the project had to be abandoned. One last thing I wanted to Oumarou: We need to have specifica mention is flour quality. Many of the tion at the breeding step. We need to de decorticators and mills that are commer velop varieties suitable for each specific cially available do not produce flour of product and easy to process using existing good quality. This is the reason why some mills. The smale-scale commercial mills women still prefer to use that are currently used in Niger do not pro hand-decortication. We need to develop duce good quality flour because most of new techniques for processing grain into the time, they are used for both good quality flour. There is one type of decortication and grinding into flour. mill tried by Dr. Gomez ofICRISAT in There is also the problem of quality con Bulawayo which gives flour ofgood qual trol. Because the mills are used to process ity. different types of cereal grains, flour from a new batch ofgrain is often contaminated Fo/ana: I am going to talk about cous with flour from previous batch. Another cous. In Niger, urban dwellers prefer rice problem is that of communication or a or maize couscous. We have a big prob lack of it between research institutions lem with sorghum couscous. This is due and food industries. We know that to shortage of good quality sorghum NAD-l is now accepted by farmers and is grains. However, last year due to food being cultivated throughout Niger despite shortage there was an increase in con the problem associated with its color. The sumption of sorghum couscous among ur grain ofNAD-1 has a thin pericarp which ban population. And we now have a big makes it easy to decorticate. However demand for sorghum couscous. Our main since the grain has a hard endosperm, it problem as small-scale processors is that requires longer decortication time. The of supply. We have problem getting good quality of NAD-l flour depends on the quality grain from the markets because growing locations. When grown in the major grain buyers purchase most of the dryer, more arid Tillabery region the good quality grains immediately after grain of NAD-I produces flour as white harvest leaving us, the small buyers, with as wheat flour. However, when it is grown grains of low quality. In addition, the in the higher rainfall region of Gaya, it is grains we get are often contaminated with susceptible to attack by insects and grain 247 mould. This in tum discolors the grain and can be used to produce acceptable biscuit. yields flour of low quality. Maybe by Several people in the panel have raised changing the planting date a solution can the question of grain supply and surplus. I be found to this problem. I hope and wish would like to mention that at SNP A, de that with the adoption ofNAD-l, in a few spite our current difficult situation, we years Niger would have surplus in grain have an annual need of about 70 to 80 tons production. This surplus grain will have of wheat. This is not a negligible quantity to find an outlet for farmers to continue to of imported grain. We could have dou produce NAD-l and adopt it irreversibly. bled this quantity if locally grown grain such as sorghum and millet could be used Sekou: I represent SNPA, a biscuit and to replace part ofthe imported wheat. This candy company. First I would like to means that local producers could easily thank INRAN for inviting us to this find market for 70 to 80 tons of their crop timely meeting. A major problem we are and we in tum would have a much lower facing at SNP A is supply of our raw mate production cost. rial. Ninety to ninety five percent of our raw material and packaging comes from Diouf One of the main utilization con outside Niger. Our production cost is very straints is lack of consumption of local high. In addition, due to recent market lib products. We all know that consumption eralization, we are facing high competi is the driving force for processing which tion with similar but much cheaper prod in tum motivates farmers to produce ucts entering Niger from countries like more. Another problem is that Nigeria, Ivory Coast, Ghana, Saudi Ara small-scale processors do not have access bia, and Lebanon. We are therefore very to credit. Also there is a need to organize interested in using locally grown cereals entrepreneurs to be good processors. and that is why this workshop is of the ut That is why PROCELOS decided to have most importance to us. Of course we can in each country a group of entrepreneurs not use these cereals indiscriminately. and processors that meets on a regular ba That is why I was very happy to see the sis to ~iscuss problems related to food question of grain quality being discussed processmg. here. We have previously done some tests using sorghum flour in the manufacture of Obi/ana: Food security involves all biscuit. However, the quality ofthe prod what you are talking about. Utilization of uct was very poor due to the low quality of technologies that have been generated that particular sorghum flour. We are leads to commercialization which leads to therefore asking the researchers to try to generation of income. provide us with good quality grain that 248 Panel Discussion Lee House Animal Utilization Day 3 - Session 1 House to panel: Are feed for animals Fernandez: Most of imported maize is from within Niger or imported? If im not used as animal feed. It is used for hu ported, what type of feed is imported? man consumption. Daouda: For poultry feed, we use im Kolo: Using stover as animal feed leads ported maize more than sorghum or mil to poor soil and desertification. Because let. We also import fish, synthetic amino ofthat soils become less productive. This acids such as lysine and methionine, is a problem that needs some solution. wheat bran, peanut pellets and cotton seeds. We import almost everything ex Idi: It is a problem that both agrono cept blood that we obtain from the slaugh mists and animal scientists need to ad ter house, chalk, and bones. dress. We take nutrients from soil without restituting. There seem to be some type of House: How much maize grain is im collaboration between crop farmers and ported for use as feed? Are stem, stover, livestock farmers. Crop farmers let live and hay sufficient in Niger? stock feed on crop residues. They in tum use the manure left by the livestock as fer Daouda: Ninety percent of the feed tilizer. produced are used to feed poultry. The other 10% are used for small ruminants. House: Agronomists have to do their part to solve the problem by creating new Douma: The annual production of sor hybrids. ghum and millet stem can only last to feed the whole livestock for 120 days. We Abdoulaye: In Niger, there is no differ therefore have to import considerable ence between crop and livestock farmers. quantity of cotton seed and wheat bran. Most farmers who grow crop also raise livestock. Idi: According to a survey conducted by INRAN in 1996, 65% of farmers are Maiga: Dr. Kapran has started devel not satisfied with stover and forage pro oping hybrids for forage. This will help duction. There is also a lot ofloss due to a livestock farmers. lack of storage facility. 249 Wrap Up Remarks from the Niger Workshop By John Yohe I wish to express my appreciation to tries. We have also discussed the current Dr. Moussa Oumarou, Director General status of hybrid sorghum and pearl millet ofINRAN, Dr. Anand Kumar, Executive production in Niger and other West Afri Director, ICRISAT Sahelian Center, hon can countries, as well as the required ored guests, and all our colleagues from bases for successful hybrid seed enter across Africa who have made this work prises, seed production, training, and seed shop a resounding success. I wish to also marketing. We have gone into the field acknowledge our sponsoring organiza and observed first-hand, farmer-man tions, INRAN, INTSORMIL, ICRISAT, aged, hybrid-seed production fields ofthe ROCAFREMI, ROCARS, the McKnight NAD-l sorghum hybrid here in Niger. Foundation, Rockefeller Foundation, Finally, we provided an overview of what USAID, and Winrock International. I is needed in value-added treatment of sor have been impressed by the collegiality, ghum and millet grain in order to increase the cooperation of the sponsoring organi local consumption and utilization of these zations, the participating institutions, and two crops. the scientists who have come to discuss and observe the potential for developing This has been an excellent experience. and marketing sorghum and pearl millet Let us go back to our homes and our re hybrids in West Africa. search laboratories and work on making hybrid sorghums and millets a reality in We have learned about the hybrid seed West Africa. experiences in other developing coun- 251 Conference Photographs ~;. I' • l \._...... '-" Participants N V> V> Dr. Moussa Oumraou, Acting Director, TNRAN ,viewing Dr. John Axtell , Dr. Lee House, and Dr. B.R. Barwale (1998 Recipient farmers hybrid sorghum production fie ld, Konni, Niger, 1998 of the World Food Prize). Pioneers in the promotion and development of hybrid sorghum and for developing countries Niger, 1998 N Dr. Larry Busch , TNSORMIL EEP and Dr. John Axtell, Dr. B.R. Barwale, and Dr. Bruce Maunder, V> 0-- Dr. John Axtell, Purdue University Niger Workshop, Ni amey, Niger, 1998 Dr. 1. Mouddour and Dr. B.R . Barwale NAD- I Hybrid Sorghum Harvested hybrid pearl millet, 1998 Village processing of hybrid pearl millet, 1998 tv v. -...J Production ofNAD-l Hybrid Sorghum Seed Farmer's field of hybrid NAD-2 sorghum, Konni, Niger, 1998 Fanner's fi eld of Hybrid NAD-2 Sorghum, Lossa, Niger, 1998 Solar dryer used in couscous preparation; INRAN Cereals Lab, 1998 N V> 00 \ \ Cereal Lab couscous roller, INRAN, 1998 Issofou Kapran and Adam Aboubacar demonstrating couscous made from Hybrid NA D-2 Sorghum 259 Acronyms Acronyms ADRAO Association pour Ie Developpement du Riz en Afrique de I'Quest AED Academy for Educational Development (F AFSA) AGCD Administration Generale de la Cooperation au Developpement AHA Le Amenagements Hydro-Agricoles ARDC Agricultural Refinance and Development Corporation ARPT Adaptive Research Planning Teams BTC Basic Training Center CA Central Africa or Central America CBO Community-based Organization CCB Central Cooperative Bank CCF Christian Children's Fund CEE Communaute Economique Europeene CERRA Centre Regional de Recherches Agronomiques (INRAN Regional Research Center) CFA Communaute Financiere Africain CILSS Comite Permenente Inter-etats de la Lutte Contre la Secheresse dans Ie Sahel (Interstate Committee to Combat Drought in the Sahel) CIMMYT Centro Internacional de Mejoramiento de Maiz y Trigo (International Mais and Wheat Improvement Center) CIRAD Centre de Cooperation International en Recherche Agronomique pour Le Developpement CLDB Central Land Development Bank CMDT la Compagnie Malienne pour Ie Developpement des Textiles CMS Cytoplasmic male-sterility CNRST Centre National de Recherche Scientifique et Technologique CRR Cash Reserve Ratio CRSP Collaborative Research Support Program 263 CSA Commissariat it la Securite Alimentaire CUSA Credit Union and Saving Association DAF la Direction Administrative et Financiere DAP Development Action Plans DCCB District Co-operative Credit Bank DG Diretor General DMM la Direction de la Maintenance et du Materiel DMV la Direction de la Mise en Valeur DNA Deoxyribonucleic Acid DPCS Directorate of Seed Production and Control DSC Differential Scanning Calorimetry ECOWAS Economic Community of West African States ECU European Currency Unit ELM Empresas La Moderna SA EPIC Etablissement Public it Caratere Industriel et Commercial FAO Food and Agricultural Organization (United Nations) FCCA Fond Commun de Contrepartie it l' Aide Alimentaiere FED Fonds Europeen de developpement (European Common Market Fund) FFHC Freedom from Hunger Campaign FS Farming Services FSS Farmer's Service Societies GIE Groupement d'Interet Economique GLC Special Line of Credit GLS Gray Leaf Spot GO! Government of India IAR Institute of Agricultural Research 264 IARCS International Agricultural Research Centers IBRD International Bank for Reconstruction and Development ICAR Indian Council of Agricultural Research ICRISAT International Crops Research Institute for the Semi-Arid Tropics IDA International Development Agency IDBI Industrial Development Bank of India IER Institut d'Economie Rurale, Mali (University of Rural Economy) IFAD International Fund for Agricultural Development IICT Indian Institute of Chemical Technology IITA International Institute of Tropical Agriculture INERA Institut D'Etudes et de Recherches Agricoles INPHO Information Network on Post-Harvest Operations INR Indian Ruppee INRAB Institut National de Recherche Agronomique du Benin INRAN Institut National de Recherches Agricoles du Niger (Nigerien National Institute of Agricultural Research) INSAH Institut du Sahel (Institute of the Sahel) INTSORMIL International Sorghum And Millet - Collaborative Research Support Program IP Irrigated Perimeter IPR Institut Polytechnique Rural, Katibouqou lRAT Institut de Recherches Agricoles Tropicales IRDP Integrated Rural Development Project ISRA Institute Senegalais de Recherche Agricole ISTA International Seed Testing Association IVH Inbred x variety hybrids LAMPS Large Sized Adivasi Multi-Purpose Societies 265 LDB Land Development Bank LDC Less Developed Country-Defined in Economic Terms (AID) LER Land Equivalent Ratios LGP Length of Growing Period LlC Life Insurance Corporation MAFF Ministry of Agriculture, Food and Fisheries MAG/EL Ministere de l'Agriculture et de l'Elevage (Ministry of Agriculture and Livestock) MANR Ministry of Agriculture and Natural Resources (Mauritius) MAWD Ministry of Agriculture and Water Development MFED Ministry of Finance and Economic Development MM Mota Maradi MMD Movement for Multi-party Democracy MRI Maize Research Institute MSU Mississippi State University MTP Mid-Term Planning NABARD National Bank for Agriculture and Rural Development NAMBoard National Agricultural Marketing Board NARS National Agricultural Research Systems NCB! National Cooperative Bank of India NCP Niger Cereal Project NCSRP Nationally Coordinated Sorghum Research Project NDC No Dues Certificate NFS Non Farm Sector NGO Non Government Organization NIDHIS Hindi Instrument Letter of Loan NPK Nitrogen Phosphorous Potassium Fertilizer 266 NRI National Research Institute NRM Natural Resources Management NSC National Seeds Corporation NSP National Seed Project NSS National Seed Service NVRM National Variety Release Mechanism OFPEP On-Farm Productivity Enhancement Program OFSP On-Farm Seed Project ONAHA Office National des Amenagements Hydro Agricoles ONG Organisations Non Gouvernementales OPVs Open-pollinated varieties OUA Organization for African Unity PAAT Produits Agroalimentaires Transformes PACS Primary Agricultural Credit Societies PCN Projet Cerealieer National (National Cereals Project) PCU Provincial Co-operative Union PDASN Projet de Developpement de l' Activite Semenciere au Niger PIB Produit Interieur Brut (Gross Internal Product) PLDB Primary Land Development Bank PNACS Primary Non-Credit (Multi-Purpose) Activities Credit Societies PPCL Projet de Promotion de Cereales Locales (Project on Local Cereals) PROCELOS Projet de Promotion des Cen!ales Locales au Sahel PSC Private Seed Companies PSI Particle Size Index PVO Private Voluntary Organization QDS Quality Declared Seed 267 R&D Research and Development RBI Reserve Bank of India RDC Rural Development Corporation ROCAFREMI Reseau Quest et Centre Africain de Recherche sur Ie Mil (Based at the ICRISAT Sahelian Center) ROCARS Reseau Quest et Centre Africain de Recherche sur Ie Sorgho RRBs Regional Rural Banks RVA Rapid Visco Analyser RYMV Rice Yellow Mosaic Virus SADC Southern Africa Development Community SAFGRAD Semi-Arid Food Grain Research and Development SAO Seasonal Agricultural Operations SB State Bank SC Scheduled Cost SCB State Cooperative Bank SCCI Seed Control and Certification Institute SCF Save the Children Federation SCH Single Cross Hybrids (male sterile x inbred) SCN Societe Cotoniere du Niger SFCI State Farm Corporation ofIndia SHG Self Help Group SICCLA Service des Intrants du Contr6le, du conditionnement et de la Legislation Agricole SIM Systeme d'Information sur les Marches SLBC State Level Banks Committee SLDB State Land Development Bank SMIP Sorghum and Millet Improvement Program 268 SNRA Systeme National de Recherche Agronomique (National Agronomic Research System) so Strategic Objectives SPU Seed Production Unit SSCs State Seeds Corporations SSCI Seed Control and Certification Institute ST Scheduled Tribes STRC Scientific and Technical Research Commission SW SvalofWeibull AB TADD Tangential Abrasive Dehulling Device TCH Topcross hybrids (male sterile x open-pollinated variety) UN United Nations UNDP United Nations Development Programme USAID United States Agency for International Development WA West Africa WASHAT West Africa Sorghum Hybrids Adaptation Trial WCASRN West and Central African Sorghum Research Network WVI Wodd Vision International ZAMSEED Zambia Seed Company Limited ZCF Zambia Co-operative Federation Limited ZIMCO Zambia Mining and Industrial Corporation ZSPA Zambia Seed Producers Association 269 Conference Participants Participants Burkina Faso Germany Samsonna Biego W.R. Lechner Director General Mittel STR 288477 C. TRA. P.A. 06 BP 10100 Schwendi, Germany Ouagadougou, 06, Burkina Faso Ghana Sansan Da Centre Regional de Recherche Agricole Harry Addor 01 B.P. 910 World Vision International Bobo-Dioulasso 01, Burkina Faso P.O. Box 31 ER PMB Accra-North, Ghana Mamadou Diouf PROCELOS-CILSS I.D.K. Atokple 03 B.P. 7049 Savanna Agric. Res. Institute Ouagadougou 03, Burkina Faso P.O. Box 52 Tamale M.R., Ghana Chad Peter Dasaah Yagoua N. Djool World Vision International ITRA PBM Accra-North B.P.441 Accra, Ghana Ndjemena, Chad Andrew Kuyipwa Cote D'Ivoire Ministry of Food and Agriculture Box 14 Francois Anoh Tamale, Ghana IDES SA Box 121 India Ferkessedougou, Cote D'Ivoire B.R. Barwale Egypt Mahyco Seed company Ldt. 19 Rajmahel 78 Veer Nariman Rd. Osman EI-Nagouly Mumbai 4000 020 India ARCIFCRI 9 Gamaa Str. Rakesh Chopra Giza, Egypt Mahindra Hybrid Seed Company A-10 Old Midc - Post Box 52 France Jalna 431203, India Genevieve Fliedel D.S. Murty CIRAD-CA R&D, Mahyco 73 Rue J.F. Breton, B.P. 5035 Kallakal, Medak (DT) Montpellier 34032, France Medak, India 502334 273 Kenya Mamaoudou Diourte IERlCRRAISotuba A. Babatunde B.P. 262 - Sotuba ICRISATIKenya Bamako, Mali P.O. Box 39063 Nairobi, Kenya Ndjeunga Jupiter ICRISAT Joseph DeVries B.P.320 Rockefeller Foundation Bamako, Mali Box 47543 Nairobi, Kenya Seydou Keita Service Semencier National Mali B.P. 167 Segou, Mali Innoussa Akintayo WCASRN/ROCARS Karamoko Sacko B.P.320 WinrocklIER Bamako, Mali B.P.258 Bamako, Mali Berthe Aissata Bengaly IER Abdoul W. Toure CRRSAISotuba - B.P. 438 IER Bamako, Mali Sotuba B.P. 438 Bamako, Mali Bourema Dembele IERlCRRAlSotuba Aboubacar Toure B.P.438 IERlCRRA Sotuba Bamako, Mali B.P.438 Bamako, Mali Dabere Dembele World Vision International Niger B.P. 06 504 Bamako, Mali Magagi Abdou INRAN/CERRA Philippe Dembele B.P.240 World Vision International Maradi, Niger B.P.26 Bamako, Mali Mahaman Sani Abdou INRAN Siaka Dembele B.P.204-Tahoua IER Tahoua, Niger B.P.258 Bamako, Mali Garba Abdoulaye INRAN/CERRAIKollo Niamoye Yaro Diarisso B.P.60 IERlCRRAISotuba Kollo, Niger B.P.438 Bamako, Mali 274 Mohamadou Abdoulaye Paul Buckner INRAN Direction de I' Agriculture B.P.429 B.P.323 Niamey, Niger Niamey, Niger Sidi Aboubacar Souley Daouda DEPIMAG/EI UAB B.P.12091 B.P. 11041 Niamey, Niger Niamey, Niger Elhadji Ada Middah Darius Producteur SIC Onaha CARE International B.P.34 Maradi Niger Konni Niger Garbo Annou Naroua Djibo INRAN ONAHA B.P.429 B.P.10697 Niamey, Niger Niamey, Niger Sanoussi Atta Laouali Dogo B.P.13723 B.P.120 Niamey, Niger Maradi, Niger Souley Attahirou Amadou Douma Chef Service Agricole de Madaoua INRAN B.P.42 B.P.429 Madaoua, Niger Niamey, Niger Elhadji Illa Baoua M. Fernandez INRANICERRAITAHOUA ILRIIICRISAT Sahelian Center B.P.78 B.P.12404 Konni, Niger Niamey, Niger Amadou Barmou Aissa Maiga Fofana Producteur SIC OPVN ONGWEYBI B.P. 12 B.P. 13211 Diffa, Niger Niamey, Niger Ouendeba Botorou Adamou Habou ROCAFREMI INRAN I CRISAT ISahelian Center B.P.l1 B.P.12404 Diffa, Niger Niamey, Niger Sidibe Hachimou Alain Bruneteau INRAN Basse Vallee B.P.429 B.P.42 Niamey, Niger Madaoua, Niger 275 Boukary Hama Gonda Jada INRAN INRAN/CERRAIMaradi B.P.429 B.P.60 Niamey, Niger Kollo, Niger Seyni Hamadou Naino Jika INRANIDECOR INRAN/CERRAIKollo B.P.429 B.P.429 Niamey, Niger Niamey, Niger Issa Harouna Boureima Kanfideni INRAN/CERRAIMaradi CLRV/INRAN B.P.240 B.P.429 Maradi, Niger Niamey, Niger Kore Harouna HabouKano Universite de Niamey INRAN B.P. 10960 B.P.60 Niamey, Niger Kollo, Niger Mahamadou Ibrahim Issoufou Kapran INRAN INRAN B.P.204 B.P.429 Tahoua, Niger Niamey, Niger Assoumane Idi Issoufou Kollo INRAN INRAN/Kollo B.P.429 B.P.60 Niamey, Niger Kollo, Niger Laouali lro Boureima Korombe INRAN ONAHA B.P.60 B.P.54 Kollo, Niger Konni, Niger Oumarou Issa K. Anand Kumar B.P. 120 Djirataoua ICRISAT Sahelian Center Maradi, Niger B.P.12404 Niamey, Niger Mahaman Issaka Sarna Laouali INRAN PASP B.P.240 Niamey, Niger Maradi, Niger Bickou Issoufou Abba Diou Lawan DDEIAICUN CDA de GAYI S/CIDDA B.P. 10262 Zinder, Niger Niamey, Niger 276 Salifou Mahaman Moussa Moustapha Producteur Semences INRAN B.P.12402 B.P.429 Niamey, Niger Niamey, Niger Seyni Maiga Amadou Moustapha INRAN/CERRAIKollo INRAN BP.60 B.P.429 Kollo, Niger Niamey, Niger Sani Maman Mbaye Ndiaye INRAN DFPVI AGRHYMET B.P.429 BP.12625 Niamey, Niger Niamey, Niger Illya Miko ZakouNoma INRAN/CERRAIKollo SAA BOURZ SIC PBVT B.P.60 B.P.42 Kollo, Niger Madaoua, Niger Mossi Mohamadou Abdoulaye Nomaou INRAN/CERRAIKollo PROCELOS BP.429 B.P. 12091 Niamey, Niger Niamey, Niger Ismail Mouddour Kondo Nouhou AGRIMEX INRAN/CERRAIKollo B.P. 10091 B.P.60 Niamey, Niger Kollo, Niger Siraji Moumouni Moussa Oumarou INRAN INRAN B.P.83 B.P.429 Tillaberry, Niger Niamey, Niger Chetima Mai Moussa R.K. Pandey INRAN PNRAlINRAN B.P.429 B.P.429 Niamey, Niger Niamey, Niger Manzo Moussa Jean Jacques Schalbroek INRAN ONAHAIProjet Semencier B.P.78 B.P.I0192 Konni, Niger Niamey, Niger Salou Moussa Alain Kasse Sekou INRAN SNPA B.P.429 B.P.10423 Niamey, Niger Niamey, Niger 277 Boureima Siaka Amadou Fofana ICRISAT Sahelian Center ISRA B.P.12404 CRZB.P.53 Niamey, Niger Kolda, Senegal Seyni Sirifi Ababacar Ndoye INRAN/CERRAIKollo ITA-Route des Peres Maristes B.P.60 B.P.2765 Niamey, Niger Dakar, Senegal Abdoulaye Souley United States INRAN/CERRAIKollo B.P.60 Tahirou Abdoulaye Kollo, Niger Purdue University Agricultural Economics Maliki Tambari West Lafayette, IN 47907 ONAHA A.H.A. B.P.54 Adam Aboubacar Konni, Niger Purdue University 1160 Food Science Bldg. Aminou Tassiou W. Lafayette, IN 47907-1160 INRAN B.P.429 David Andrews Niamey, Niger 6501 Darlington Ct. Lincoln, NE 68510 Ousmane Y oum ICRISA T Sahelian Center Pierre Antoine B.P.12404 Winrock International Niamey, Niger 38 Winrock Dr., Petit Jean Mountain Morrilton, AR 72110-9370 Nigeria John D. Axtell S.c. Gupta Purdue University ICRISAT/WCA, UTA Office-SABO Department of Agronomy - Lilly Bldg. PMB-3491 W. Lafayette, IN 47907 Kano, Nigeria Lawrence Busch A. Joshua Michigan State University Premier Seed Nigeria LTD 210 Berkey Hall P.O. Box 1673 E. Lansing, MI 48824-1111 Zaria Kaduna State, Nigeria Thomas Crawford, Jr. Senegal INTSORMIL - University of Nebraska 113 Biochemistry Hall Alphonse Faye Lincoln, NE 68583-0748 Winrock International B.P.16278 Dakar, Senegal 278 David Drake A. Bruce Maunder Purdue University 4511 Ninth St 402 E. 2nd St., Box 361 Lubbock, TX 79416 Otterbein, IN 47970 Gebisa Ejeta John Mullenax Purdue University DCHAlPVC Department of Agronomy - Lilly Bldg. RRB 7.06-071 W. Lafayette, IN 47907 Washington, D.C. 20523-2052 Joan Frederick Gary C. Peterson INTSORMIL - University of Nebraska Texas A&M University 113 Biochemistry Hall Rt. 3, Box 219 Lincoln, NE 68583-0748 Lubbock, TX 79403-9803 Richard Frederiksen Yves-Coffi Prudencio 8248 Raintree Dr. NE W orId Bank - J Bldg. Albuquerque, NM 87122 Room 6-165, 1818 H St., N.W. Washington, DC 20433 Richard Hahn Darrell T. Rosenow 1200 Sharingbrook Texas A&M University Manhattan, KS 66503 Rt. 3, Box 219 Lubbock, TX 79403-9803 Bruce Hamaker John Swanson Purdue University USAID/G/EG/AFS/ST Dept. Of Food Science, Smith Hall RRB, Rm 2.11-006 W. Lafayette, IN 47907 Washington, D.C. 20523-2110 Leland House JohnM. Yohe 547 Broadbranch Rd. INTSORMILlUniversity of Nebraska Bakersville, NC 28705 113 Biochemistry Hall Lincoln, NE 68583-0748 Katy Ibrahim Zambia Purdue University 1168 Agric. Adm. Bldg. Room 26 Bhola Nath Verma W. Lafayette, IN 47907-1168 ZAMSEED P.O. Box 35441 NouriMaman Lusaka, Zambia Dept. Of Agronomy and Horticulture University of Nebraska Zimbabwe Lincoln, NE 68583-0915 Emmanuel S. Monyo Jerry Maranville SADC/ICRISAT 37345 Broadway Avenue Box 776 Matheson, CO 80830 Bulawayo, Zimbabwe 279