IDRC -MR171e

SMALL MILLETS -- RECO...ENOATIONS FOR A NETWORK

Proceedings of the Small Millets Steering COl'llDittee Meeting, Addi s Ababa, Ethiopia, 7-9 October 1987

Material contained in this report is reproduced as submitted and has not been subjected to peer review or rigorous editing by IDRC Communications Division staff. Unless otherwise stated, copyright for material in t,his report is he ld by the authors . Mention of a proprietary name does not constitute endorsement of the product and is given only for information . FOREWORD

The S;nall Millets, which include finger millet (Eleusine ccraca..~a), fm,"tail millet ( Setaria i taiica) , proso millet ( Panicum m1liaceum) , kodo millet (Paspalum scrobiculaturn), little millet (Pan1cu~ surnatrense), barnyard millet (Echinochloa .§B2.), fomo (Digitaria exilis) and teff (Eragrotis tef) are important to global agriculture and are m=.jor cereal crops, grcwn in fairly large areas of South l<..s1a, China, USSR and Africa. They are also found in areas cf the Ur11 ted States and Europe on a iinu. ted scale. Al though precise estimates on tl1eir area .:J1d product1cn are not ava11ao1e, these crops may occupy between 18-20 mil.lion hectares, prxucing 15-18 million tonnes cf gra:::..n. The reg1cnwise distr1bucic:1 cf area is 6 . 5 m. ha in South Asia , 5 m. ha in Chim. , 4 rr,. ha 111 USSR arni 2 m. ha 111 Africa. Finger millet is the principal small millet species grown in South Asia, followed. by kcx:io millet, f m.."tail millet, little millet, proso millet and barnyard mlllet in that order. Fcxta:::..l millet, and proso millet are important in China and proso nuiiet is grown extensively in Soutb Western USSR. In Africa, finger millet, teff and ·fonio have local l.Iliportance.

In spite of the large areas ·chey OCC'J.PY and their importance in providin; fo:>d for mi ii ions of n.tral people, t..1ese c:::-op.s have re~e1ve:. verf lit~le attention at e1 ther r.at1cm~l or 1:1temanonal level. :n 1nc.r:y count.nes .. small millet research pro?"ranmes ge we~ er non-e::uste:.t..

Ali small millets are grown lil arid, serr..i-arid er montane :c:nes as rainf ed crops, i.l:lder marginal a.~i submarginal conC.1 tlons c: scii .fertility and misture. Even so, it should be appreciated that t.he average global proouctivi ty cf sm?.11 rr.illets is almost. a t.onrie per ha. There ha.s been a trend in the last tw decades to repiace these crops by major cereals like maize and wheat which has been c. factor in the reduction of area under these crops. Preser.tly 1 soall millets ge cultivated in areas ~nere they produce a more dependable harvest compared with an;.: other crop. This has been largely resp;:::r...sible -:or thei= continued presence and cul tivat ion in m:my parts cf the w:-rld. T:iere ls now an increasing reailzatic-::1 cf this fact, and a greater awareness that these crcps merit rrcre research and development.

As a response to this nee-:i, the Indian Council cf A;ricult:i..l.rai Research, New Del.l'u, India, Ur.1 versi ty cf Agricu.i tur5.l Sciences, Bangalcre, India, and the Interriational De\~elop11€r~:r: ?--e~.e5.rc~1 Cer:Lre, Canada, Jointly crgan1zed the first Interr1at1onal Wcrkehop c-:; S:nall Millets, lil Octcber 1986, at 3angal0re, In:.ia. V.ore details en t.he obJect1 ve.s, and o:Jtcome cf th:..s meenng are fo;_;_;"ld :rn :.hs I:i.t.:::-OO'Jctcr;· Re:narY.s by K ::\iley, whiie the Feco!':'IT"lendatonE cf this. f:.!:"2L: ·... •::-rtshcp are fo'..lnd in 1'.;ipenci.J..x I::: c: t:ri:..s rep::.r::.

One ·..•. ::=:..:: :~.·=- formation of a Small Millets Steering Conmittee, ccnsisting of members from India, China, Ethiopia, Zimbabwe and canada. The St:eering Cornnittee was charged w~th the responsibilities of establishing a small millets network and identifying areas of mutual co-operation and collal:x:>ration. The Steering Cornnittee was given one year to come out with specific proposals on the nature of a srr~ll millets netw;)rk. The International Development Research Centre, Canada, was requested to take the init:1ative in organizing a meeting of Steering Corrrnlttee members at a suitable place.

This report contains the proceedings of the first Steering Cormuttee meeting at the International Livestock Centre for Africa in Addis P.baba from Cctober 7-9 1987, under the auspices of the Institute of Agricultural Research. The proceedings contain the preparej papers presented during the first day cf the meeting, and a repcrt of the outcome of the discussions which took place during the second day.

This report is being sent to the participants who at:tended the first workshop, other small millet scientists, research managers and directors of National Progranrnes, Universities and International Centres, as well as to potential donors. The Steering Cormuttee requests those who receive the report to carefully review the proposed network structure and activities, and provide carrnents to one of the Steering Cornnittee members.

We look forward to your pa.."l.icipation and supp:;,rt in developing the International Sm5ll }'f.illets Networ).:. A Seetharam K Riley C"nainnan Secretary Small Millets Steering Cornnittee November, 1987

ll PARTICIPANTS ATI'ENDING FIRST SMALL MILLETS STEERING ca+il'I'I'EE MEETING

A. MEMBERS

1. Mr Chen Jiaju Professor and Head Institute of Plant Germplasm Resources Laboratory of Sorghum and Pd.llets Chinese Academy of Agricul~ural Sciences 30 Bal Shi Quoa Lu West SUburbs, BelJing The People's Republic of China 2. Dr Seyf u Ketema {Co-Chairman) Leader, Teff Prograrrrne Institute of Agricltural Research Box 2003, Addis Ababa, Ethiopia 3. Mr Samuel E O:ielle Millets Breeder Sorghum and Millets Unit Uganda Agriculture and Forestry Research Organisation Serere Research Station PO Seroti, Uganda 4. Dr K Riley Senior Progrc.rn Officer Interna~1onal Develoµnent Research Centre 11, Jorbagh New Delhi - 110003, India 5 . Dr A Seetharam (Chairman) Project Coordinator All India Coordinated Millets Improvement Project GKVK campus, University of Agricultural Sciences Bangalore - 560065, India

iii B. INVITED PARTICIPANTS

1. Dr Seine Debela General .Manager Institute of Agricultural Research Box 2003 Addis Ababa, Ethiopia

2 . Dr Helak Werede Director Plant Genetic Resources Centre P o Box 30726 Addis Ababa, Ethiopia

3. Dr Yilma Kebede Leader, Sorghum and Millets Progranme Institu~e of Agricultural Research Box 2003 Addis Ababa, Ethiopia

4. Mr K E Prasada Rao Senior Botanist, Genetic Resources unit International crops Research Institute for the Semi-Arid Tropics (ICRISAT) Patancheru P O Andhra Pradesh 502324, India

lV CONTENTS

Page

PRESENTED PAPERS

Dr. Seme Debele...... Opening Remarks 1 Dr. Ken Riley ...... Background to the Steering conmitte meeting 2 Dr. Sayfu Ketel13...... Status of Small Millets in Ethiopia 6 Dr Yilma Kebede ...... Improvement of Finger Millet in (Eleusine coracana) Ethiopia 16 Mr. Fassli Kebebew...... ~cti vi ties of the Plant Genetic Centre;Ethiop1a 21 Mr. Sam Odelle ...... Small Millet Improvement in Uganda 23 Dr. A. Seethara~ ...... Srrall Millets - Their Importance, Research and Developnental Status in India and South Asia 28 Mr. K.E. Prasada Rao ...... Yi.lnor Millets Germplasm Resources at ICRISAT 38 Mr. Chen Jiaju ...... The Present Status of Small Millets in China 46

DISCUSSIONS AND RECOMMENDATIONS

PROPOSED ORGANIZATION OF THE ~RK 51

Background 51 Network Name 51 Netw:>rk Participants 51 Crops Covered in the Netw:>rk 51 Objectives of the Netw:>rk 52 Financial Suppcn: 52 Steering Corrrr~t~ee 53

PROPOSED NE'IWJRK ACTIVITIES 53

Scientific Interaction 54 Broader Awareness of Small Millets 59 Genetic Resources of Srr~ll Y.illets 59 Collaborabve Programs 61 · Utilization of Technolo;y 63 Generation and Dissemination of Inforrration 63 Training 64

PRIORITY NEIWJRK ACTIVITIES REQUIRING EXTERNAL FUNDING 64

Jl.ppen1i:-: I Table - Collectio!"lS Of Small Millets ~£ Reported By Partic1pa~ing Prograi11S 66

Appendix II ...... Reconmendations Cf The First ~!"°J"Ce=r.:.c:::.onal ·vJcr~:.:h:.:~ C'n S:r.all ~tilletE E9 l

Seme Debela

I feel deeply honoured to be invited to make an opening statement on the ocassion of your first meeting here in .Addis Ababa. I thank you very ITR.lch for giving me this honour.

The millets as a group are not widely grown in Ethiopia, although they possess attributes that could make them quite attractive particularly to the semi-arid zones of the country. About the only millet of the group that is grown by our mid-altitude fanners is finger millet. It is mostly used in beverage making.

I am very happy to note that Teff, our staple food crop, is included in the small millets. This is very i.rrprtant to us, since it \oOJ.l.d allow international focus on the crop. SO far, there has been very little intellectual effort to i.nprove the crop at the international level, as it is only in Ethiopia that the crop is used as a major food item. Teff has been on the research agenda in Ethiopia for many years. Ho"'1ever, progress has been slow, mainly because of lack of focus and concentration. The lack of information en, and international support for the crop has been airong the major limiting factor.

The Institute of Agricultural Research is now putting 91ll>hasis on the crop. We are in the process of creating a national research centre responsible for teff breeding, teff agronany and teff utilization. The centre will have a rm.llti-disciplinary team to guide as "'1ell as execute the research and developnent (R & D) programne on teff. We have already acquired technical and financial support fran IDA and the ODA/ United Kingdan for our project. The Small Millets Programne will be an additional support for the project.

Once again, I 'lo.Ould like to "'1elcane you to Ethiopia and I hope your meeting will be fruitful 2

K. Riley

Small Millets With Big Potential

Small millets are generally grown in traditional agricultural systems, with low inputs and low productivity. There is a trend to replace millets with m::>re productive crops such as wheat maize or rice in m::>re intensive system.s. For example, in Sri Lanka, the finger millet area which was associated with slash and burn agriculture, has decreased as traditional agricultural methods are replaced by more settled systems. Similarly, in Bangladesh the increased area now devoted to wheat and rice has ·resulted in a decrease in millet area. There are several other countries where an increase in the major cereals has resulted in a decrease in the small millet acreage. We rrust recognize that there are situations in which major cereals do have an advantage and should be encouraged. Nevertheless, there are many unique traits possessed by the different millet species, which should make millets an important component of improved agricultural system.s. I will try and give a few examples of sane of these unique traits possessed by millets.

* Millets are generally fast maturing, which should enable them to fit into more intensive cropping system.s. A fast maturing millet could be used as a catch or relay crop in association with other, slower maturing crops.

* The name millet canes fran the 'WOrd mil. or thousand, referring to the large nunber of grains that can be grown frcxn a single seed. Rapid rm.11.tiplication of seed is generally a relatively simple matter and seed costs are low. The small seeds of millet generally store well for long periods, ensuring a continued foo:l supply during dry season or when there is a crop failure. * The small millet seeds of ten require less cooking or preparation time. this can be an increasingly important factor when ....anen beccxne involved in more productive farming systems and have less time to devote to foo:l preparation.

* There are a large number of ways of processing milletE in traditional and novel preparations. This can be a factcr in increaEin? the market demand for millets. 3

* There are many difficult or marginal farming situations in which specific millets species perfonn well. Teff, for example is tolerant of water-logged and acid soils. Proso millet can tolerate both drought and saline soil conditions. foxtail millet p:Jssesses adaptation to low fertility soils. Because many millets are fast maturing, they can prcx:iuce a crop quickly and escape the onset of stress conditions such as drought.

* Many varieties of millets have excellent nutritional properties, containing high levels of essential minerals such as Iron and calcium. Finger millet is especially known for its characteristic of providing energy for a long time after it is consumed. This is an i.rrp:>rtant trait for people who have jobs that require hard manual v.ork.

* Millets are not necessarily low yielding crops. Grain yields of finger millet in field conditions in India and Uganda frequently exceed six tons per hectare, and foxtail millet can prcx:iuce similar yields in China. The lower yields of small millets compared wi tJ1 yields of the major cereals may be due to selection by fanners, over thousands of years, for tolerance to difficult conditions rather than for high grain yield~ se.

* Millets are generally highly valued for their fodder. As indicated in the 1986 Workshop, a new Indian variety of foxtail millet, called SIA 326, is proving to be extremely popular with the fanners in Andhra Pradesh. In addition to its high grain yield, this variety has straw which is highly palatable as livestock feed. The econanic value of foxtail millet straw to these farmers, is almost equal to the value of the grain. Recent work at the Dryland Agricultural centre in Bangalore has found that little millet and barnyard millet produced rrore forage yield per day under dry conditions canpared to any other forage crops tested.

* We are painfully aware that agriculture is still vulnerable to crop failure, often due to flooding or drought, or due to mismanagement of soil and water resources. Although millets cannot prevent these catastrophes, millets are known as famine crops, that can ensure a quick food crop when other crops have failed. In Bangladesh, for example, the millet area this year is expected to substantially increase, following the worst flooding in perhaps 100 years.

* In rrountain, or hill areas of Nepal the millet area is increasing, from 123000 ha in 1983 to 151000 ha in 1985. Recent estirrates place the area planted at 235, OOC ha. L~1i rc:c:es cf finger millet in Nepal are adap'ted tc the e:;.:':re:7F- v.=-.r1anc:-i and harsh ccn:.::.. t1on2 in mountain regions. 4

The above examples indicate that millets have an important role to play as a component of m:>re sustainable and productive agricultural technology.

First International W:>rkshop.on Small Millets

The Indian Council of Agricultural Research, the University of Agricultural Sciences, at Bangalore, and IDRC jointly hosted the First International Worh:.::!1"."!' on Small Millets in Bangalore, September 29 to October 3, 1986. out of approximately 14 species of millet, there were seven species chosen for consideration at the "WOrkshop.

* Finger millet or ragi Eleusine coracana * Foxtail millet Setaria italica * Prose millet Panicum miliaceum * Tef f Eragrostis tef * Kodo millet Pasw.lurn scrobiculatum * Barnyard millet Echinochloa spp * Little millet PaniCLDll sumatrense * Fonio Di

Bullrush millet (Penniseturn americanurn) is being well supported by ICRISAT, while the other small millets, al though important locally, are limited in area and were not considered at the "WOrkshop.

The purpose of the "WOrY.shop was to bring together scientists "WOrking on these millets, from countries where these crops are important; to assess the importance, production, and the place of these crops in improved agricultural systems; to discuss the status of research on these crops and explore ways to collaborate in strengthening millet research.

over 50 scientists from India, Bangladesh, Nepal, Sri Lanka, China, Russia, Ethiopia, Kenya, Zimbabwe, the ICRISAT SADCC Program, Tanzania, Uganda and IDRC attended. Sessions on production trends, genetic resources, breeding, cropping systems and production technology, physiology, food and forage uses were held. There was discussicn following each session, and a great deal of information was exchanged, ooth fsm.::.lly and informally, and some direct arrange.'Tlents made f~r excha.~ge of seed material. 5

A set of recornnendations arising from the workshop were developed and approved by the participants. These recorrmendations are found in Appendix II of this report.

Small Millets Steering Ccmnittee

The Steering COlllnittee meeting, in which we are now participating, is a response to general recorrmendation 4. The objectives of this Steering carrnittee are generally to reconmend how a small millet netw:::>rk is to be developed. The specific objectives are to :

1) Review the progress to date in exchanging small millets genetic resources, and plan specific action for future exchange.

2) Review the information sources available on small millets and w:::>rk out how information on small millet might be effectively transferred.

3) Develop priorities for research and suggest mechanisms for effective collaborative research on small millets.

4) Consider training needs and opportunities.

Of the five nominated steering conmittee members, we are happy to have four with us today. Mr. Fighur M.lza, who was to represent Eastern and Southern Africa, is away on study leave. We are glad that Mr. sarnuel OOelle was able to participate in his place. I suggest he could be considered as an acting Steering Ccmnittee member.

We are very fortunate to have three other participants at this meeting. Mr. K.E. Prasada Rao has great experience with the small millets and is able to discuss ICRISAT's involvement in this area. We also welcome the participation of Dr. Melaku Worede, Director, Plant Genetic Resources Centre, and Dr. Yilma Kebede, leader of the Ethiopian Sorghum Irri>rovernent Program. 6

STATUS OF SHALL MILLETS IN ETHIOPIA AND AFRICA

Seyfu Ketema

I. Introduction

It is stated that "in a wide belt across parts of South .Asia, India, Africa, the Middle East, and in other areas from northern Argentina and north-east Brazil to Mexico, food production for millions of people in the semi-arid tropics is limited primarily by the erratic nature of the rains. Some of the countries of major concern within the seasonally dry and semi-arid tropics include : Senegal, Mali, Guinea (North), Burkina­ Fasso, Niger, Ivory Coast (North) , Sudan, Ethiopia, Kenya, Uganda, SOmalia, Tanzania, Pakistan and India. Other countries in southern Africa, Central and South America, and an important area of Australia also fall within this Ecological zone", (Gupta, 1975). It has been quite a number of years now since it has been realized that in many of these countries drought was, is and will remain a recurrent problem. Recently, the conditions in Ethiopa and surrounding regions have derronstrated that these regions are prone to recurrent droughts. This also implies that rain-fed agriculture in these regions is unreliable. Unless a solution is sought in a systematic v.-ay for dealing with the situation, recurrent drought with the accompanying food shortages will continue to have a devastating effect on the lives of millions of people in these regions.

Several countries have attempted to introduce various measures, such as developing food security systems, food policy systems and agricultural policies in order to combat the hazards of recurrrent drought and its negative effects on the social and economic conditions of their countries. Agriculturally, attempts have been made to shift fran rain­ fed to irrigated agriculture wherever the potential exists. Also, development of water-harvesting mechanisms, agronomic practices that help to conserve and utilize available rroisture, development of early maturing and drought tolerant crops is being attempted in different countries. Internationally the establishment of institutes such as International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) are a response to help semi-arid regions to improve and stabilize agriculture menaced by rroisture stress. The resources allocated by many international centers to develop varieties of major cereals that cope with various stress conditions that exist in the semi-arid tropics is comnendable.

~ This paper contai.'15 information presented by participants f=om Tanzania, Kenya, Uganda, Zimbabwe, and the S1'.DCC/ICRIS.Z..T Program at the First Interr.c.:2.cnal S;n.s.::..i P.:.. Hlets WQrJr~hop, Ba.'1galore. ::.e~:. 29 - O:::t 3, 1%6. 7

However, it might also be wise to question sane of the approaches. It is one thing to improve the productivity of the major cereals in order to eA-ploit their full potential under ideal conditions and another to develop varieties that cope with stress conditions under marginal soil and rroisture regimes. Should researchers try to adapt maize, ..meat or other major cereals that are unadapted to marginal areas, or should they try to develop the productivity of such crops as millets, 'n11ich are already adapted to semi-arid stress conditions? The question is put m resposnse to the present fact that no internationally coordinated effort is being made to exploit the potentil as of the small millets. Would it not be worth spending at least part of the funds allocated to International Agricultural Research Centre for carrying out internationally coo-ordinated research in small millets, if we want to help people in regions prone to drought and surviving under marginal agricultural conditions? What advantages have the small millets to offer? One can em.unerate a number of points. The following are a few of the merits of small millets:

1 . M:>st of the small millets are hardy and low risk crops and do relatively better than the major cereals under low m::>isture stress.

2. These crops do relatively better than the major cereals under IM.rginal soil conditions.

3. In developing countries, the increasing costs for inputs, and low prices for grain makes it difficult for the fanners to apply sufficient inputs for the major cereals to achieve their full genetic potential. Hence there is a need to shift to small millets. 4. Even in countries that have well developed irrigated agriculture, there are often problems associated in sustaining these intensive systems, also due to the constantly increasing demand for food and. the necessity to expand agriculture to marginal land, attention is being given to the role of millets to help solve the problem.

In recognition of these and other facts, the first International Workshop on small Millets, which took place in Bangalore, India Sept. 29- 0ct. 3, 1986, has pointed out in its reconmendation that an international centre be established to conduct research in order to improve the productivity of small millets and expliot their potential. The present food shortage in Africa, associated with drought leads to the need t::: explore all possible options that could help end the hazards of f coj shortages and the misery it causes to lives of millions of people. One such option which has not been fully exploited is the option of srrall millets.

II. Area and Production

Srrall r.-.1llets are i::rcC.uced under rain-fed condi tior..s 1 ard are ·:;r8·.. ;:-, in many CO'Jlitries. of Africa. The fact they are hardy crops rn the serJEe of being adapted tc low rainfall areas that they have less disease anC. pest probl~ both under production in the field and storage, makes them fX)pule:r with small fa..."T.lers living in marginal areas. 8

Table 1. shows the area and production of sane of the millets grown in ~.frica.

Table 1. Area and Production of Small Millets in sane African COlUltries.

COlUltry Year Area Production Crop(s) (Hectares) (tons)

Ethiopia 1984/85 1,317,950 1,090,238 Tef Ethiopia 1984/85 220,000 200,000 Finger Millet Kenya 1981 154,400 98,816 sorghum/Millet

Tanzania 100,000 NA Finger Millet Uganda 1985 310,000 210,000 Finger Millet Zambia 1969/70 130,500 84,900 Fingar Millet Zimbabwe 1986 200,000 100,000 Finger Millet

III. Importance ~ Research EtbioPia Teff (Eragr05tis tef), finger millet (Eleucilla corocana), and prose millet (Panicum miliaceum) are known to be under production in Ethiopia. Tef f is not only the nost widely grown millet but also the staple food and rrost irTix:>rtant cereal crop in the country, occupying 28 percent of the cultivated land lUlder cereals. It occupies nore cultivated land than wheat, barley, maize or sorghum. Finger millet occupies about 5 percent of the cultivated land. No statistical data is available for prose millet, but its production is marginal. The national average yield for tef is 1 ton per hectare (t/ha),and for finger millet is less than 1 t/ha. The grain yield for improved teff cultivars with improved rnanage.rnent practices is 2 t/ha on the farmers field.

Finger and prose millet production is confined to a few areas mainly in t.~e northern part and on the western part of the country. On the other hand, teff is produced in almost every part of the country but concentrated production is in the central highland .

.t.ll the tt..ree rr.iliets are use:i to make a ieavened bread called "inJera'', pcrridge c.nc alcohclic G.r1r.Y_s. In r..o.J.:ing the::e prepc.ratior.s, the rm.llets may be mixed with other cereals. Injera made f rorn pure tef is, however, rrost preferred. The straw of tef is c. special and i.mp::JJ:taYJt souce of feed for cattle during the dry season and that of the other two 9 millets is also used for the same purpose.

Tef f grain fetches a higher price in the market than any other cereal produced in the country. Hence, it offers an extra attraction to farmers. This is in addition to the advantages it shares with the other millets that makes them popular with small farmers.

Research in teff started around the late 1950's and that for finger millet in 1965. However, a systematic and sustained \-.Clrk towards the improvement of finger millet did not start until 1985. Apart from some adaptation trial which started in 1979 for proso millet no serious and sustained effort has been put tow-ards its improvement. Tef is getting the highest available research attention both in tenns of manpower a~d other resource allocation, while finger millet ranks second and proso r~ last and alm::>st gets no attention at all. Generally the objective is to develop high yielding varieties and improved cultural practices.

Several improved cultivars of tef for both grain yield and quality (white seeded) have been developed and released to farmers. Also improved management practices that help to increase grain yield have been identified and recon111ended to farmers through agricultural extension workers. These include appropriate sowing date, seed rate, fertilizer rate, cultural and chernical weed control methods, and chemical pest control measures.

So far, no improved cultivar or improved management practice has been developed or recomnended for finger or proso millets.

In Kenya, finger millet and pearl millet are the tv.u rrost important millets. Foxtain and proso millet are of less importance. The average yields of millets in the country are around one ton per hectare. Finger millet, which is mainly grown in the western part of the country is also found distributed in other regions. Pearl millet whose production is declining due to its susceptibility to bird damage, is mainly gro"Wn in the eastern pc.._rt cf the country.

The increased production of maize in the Rift Valley, in Western Kenya, and to some extent the growing of rice, wheat and industrial crops in the production areas of small mlllets, has resulted in the decline in millet area. However, the replacement is not canplete, since millet proj~ction continues in areas where maize does no~ do well. For example, miliets are gyown in the eastern pa._rt of Kenya, which is hot and dry with ma.rgir.al ra1!1fall, and in the Victoria region which is humid with p::or drainage and strik.en with the parasitic weed striga. 10

In Kenya, millets are generally used in pure form or in mixture with other cereals to make their taste appealing for preparing porridge and a cake-like substance (Ugali). Finger millet is also used to make alcoholic drinks. Pearl millet is sometimes consumed by itself or in mixture with beans, pigeon peas or peas after boiling.

The stalks are used as fire w:::x:xi, or left in the field. They are rarely used as feed, except those of finger millet which are so used only when they are green.

Research to improve the productivity of millets started in 1978 with the objectives of developing improved cultivars and production technology especially suitable for areas that are marginal for maize and wheat. The areas considered marginal are :

a. Water logged and striga striken areas around Lake Victoria.

b. The hot, dry meditun and low altitude areas of the Eastern Province. c. The coastal strip. d. The cold highlands of the Rift Valley region.

So far some i.nproved cultivars of finger, pearl and foxtail millets which perform better than the land races have been identified. Efforts are being made to develop i.nproved management practices.

~Reqion The Southern African DeveloPT19nt Coordination Conference (SADCC) region consists of nine countries. Five of these, Botswona, Malawi, Tanzania, Zimbabwe and Zambia, grow finger milelt widely. '!'¥.0 others M:>zarnbique and Angola, grow it in small quantities whiie much is not known about finger millet production in Lesotho and Swaziland. In general, finger millet area and production ranks fourth next to maize, sorghun and pearl millet in the SADCC region. It is mainly used for making porridge and local beer. Also, especially in Zimbabwe, in combination with maize or pollards it was used for fattening pigs. Very little research work has been done on this crop. The little that has been done was mainly in Zambia starting in the early 1960's and Zimbabwe beginning in 1968. In 1985 an improvement program for yield and quality of finger millet for the whole region was started at Matopas (Zimbabwe) by SADCC/ICRISAT Sorghl.D'n and Millet Improvernent Project. T'r.:::-o'Jgh earlier research work some improved cultivars ~t.ich give a mean y:.e.:..G. c.£ t'W:) t:::--~ per hectare have been identified. Agronomic pra::t1ces suer. c..~ SJW-.:_ng date and method and. on the use cf fertilizer rate and type h5.-,'e also been worked out. 11

Tanzania Finger millet is the only impcrtant small millet produced on large scale in Tanzania. The main growing areas include the Southern Highlands of Tanzania, and the eastern side of Lake Victoria. However, it is grown to sane extent in different parts of the country. It is used for making porridge and local beer. Yields vary from 0.5 to 2 tons per hectare. I.ow yield is partly attributed to be due to the comron practice of producing finger millet under an intercropping production system, mainly with maize.

A concentrated research effort putting high priority to semi-arid areas of central and western Tanzania W""QS started at Ilonga Agricultural Research Institute (TARO-Ilonga). Apart from this from 1970-71to1978-79 a research program centered at Uyole was run giving errg_:>hasis to develop high altitude finger millet, mainly to the Southern Highlands of Tanzania. The general objectives of the program were :

1. Breeding of millets and sorghum and providing a continuous source of gennplasm adapted to different agroecological zones.

2. Developnent of packages of agronomic practices which ~uld help to increase the productivity of sorghum and millets.

Although improved cultivars have been developed mainly through introduction from Uganda rrost traditional finger millet growers have preferred to grow local finger millet types for the preparation of both porridge and local beer. An improvement program using local cultivars

instead of introduction which began 1976-77 1 has identified cultivars that have high grain yield and acceptable focxi value.

Agronomic practices on the use of fertilizers, chemical weed control, appropriate plant density and stand establishment have been investigated and reconmendations made. Uganda

Finger millet is the rrost irnp::rtant cerel crop in Uganda, where both its area and production exceeds that of maize and sorghum.

It is grown throughout the country. However, the major product:.o:; areas are the intericr plateau betwen Mt. Elgin and Lake Kyoga and Northern and Westerr1 Ug.=.nC.a. Es~ially in the highlcn::S. cf Y.igezi, Tes(::o aY)~ B·J..t:eii ~·:-:.:.:.~..:::tio:-. !.""e=.·:hes its 17.?..Xi.JTilJ.ITt.

The main finger rrillet prcrluct:c:-i ::ones are the areas which are net. suitable fer the production of other crops such as bar.aria, sweet potatoes and maize. In some places, becc.use cf it.s higher yield, sorghm is 12 repacing finger millet.

Finger millet is mainly used for making porridge and local beer. Sometimes it is used as feed for chicken and the by-product of the beer is fed to pigs, goats and chickens. The straw is not usually cut and fed to cattle. However, remaining stalks in the field are grazed by cattle, sheep and goats. For unknown reasons this practice is noted for causing abortions in pregnant cows.

Due to its good keeping quality in storage, and since no major storage pests attack it, finger millet is often kept aside as an insurance crop against drought or lean years to prevent famine.

Although some v.crk has been done before, it was in 1965, at serere Research Station, that a systematic finger millet improvement v.crk started.

Although improved cultivars and cultural practices have been developed, the weakening of the extension setvice has prevented their wide dessimination to farmers. This problem has been recognized and since 1985, Serere researchers have started a Farming Systems Research Program aimed at extending the developed materials and technologies to the farmers.

Zimbabwe In Zimbabwe, finger millet is the rrost important small millet, c.nd the second largest nost important cereal after maize. Pearl millet and other millets are of lesser importance. In areas which are favourable for growing maize, finger millet suffers from competition. Therefore, it is usually grown in areas which have low fertility, resulting in lower productivity. Yields are estimated to be 0.5 t/ha for small farmers and 2 t/ha for semi-corrrnercial farmers.

Finger millet is mainly used for making beer and porridge. After harvest, cattle feed on the stalks in the field. Otherwise it is not considered as an important souce of feed.

Improvement work was conducted in 1968-69 and later in 1980. Recomnendations on the use of f etilizer, -weed and pest control, sowi_ng methx.s and. h:Lrvesting have been made. 13

DJ. Gennplasrn Resources Ethiopia

Ethiopia is considered to be both the center of origin and diversity for tef. Since tef is neither produced on a large scale nor distributed to many other countries, Ethiopia is the rrost important source of germplasm, possessing a tremendous wealth of diversity. Ethiopia and the ¥.Orld corrmu.nity therefore should give emphasis to the colletion and donservation of as much germplasm of tef as possible since we have only one source and no other. It will be a great pity if we lose the variability without conserving it here or introducing it to other countries. Finger millet is also considered to be indigenous to Ethiopia and there is ITR..lch diversity within the local land races.

At the plant Genetic Resources Center in Ethiopia, there are 2650 accessions of tef and 1956 accessions of other small millets. A systematic collection, conservation and utilization ¥.Ork is in progress with the cooperation of the gene bank and the various research institutions in the country. These efforts ¥.Ould be sustained with international cooperation.

Kenya lies within the area of origin and domestication of rr~llets that extends from the North-Eastern quadrant of Africa to the Central and Eastern parts of the continent. Hence, there is a great diversity of finger and pearl millet. Mlch of the diversity in finger millets exists in the western part of the country while variability in pearl millet is found in the extremely dry areas of the eastern provinces. Collections and introductions have been made at different times and some of these are available at Kitale, KatU1roni and M.lguga in Nairobi. However, rrore collection ~rk is needed. SADCC Region In the Southern African Developnent Coordination Conference Region (SADCC) germplasm was collected from certain regions such as Zimbabwe, Zambia, Y.ialawi, M'::)sambique, Tanzania and Botsw-ana.

In collaboration ~ith IBPGR/ICRISP.T and the respective agencies within the S1'.DCC countries germplasm collections have been made. The finger millet collections included 190 from M:tlawi in 1979, t~ fran M:>zambique in 1981, five from Tanzania in 1978-79, 273 fran Zambia in 1980 to 1984, and 474 from Zimbabwe in 1982 and 1985. These collections are small and fro.11 few places a'r'ld t.11ere is need for rrore collections. Uga'r'lda

Uganda ~as a grec.t d~vern":y of finger milelt germplasm. W:)rk on collection, corJ.Ser..-at1cr., character:;..:::c.:::ioL arid utilization is proceeding. The following are some of the local and introduced collections maintained m Uganda. The lry-...,al collect1or.s includes 778 accessions of finger millet, while 927 collections have been introduced. from various 14

countries. Apart from these 83 foxtail, 93 conm::>n, 84 kodo and 88 little millet collections have also been introduced. M:>re collections in the future are recannended. V. E'Uture Research Priorities and Needs

Small millets research was not carried out for a long period in a sustained way in many African countries. Therefore, there are many topics that need to be addressed. However, the following are some of the resarch topics that can be considered as priorities and of conm::>n interest to rrost African countries. 1. GermPlasm : There is a tremendous diversity in different small millets in Africa. This has not been well explored, collected and conserved. There is a great need to coordinate and systematically collect and conserve this germplasm as this ferns the basis for rrost of the future improvement w:::>rk. Apart from collection and utilization in individual countries there is also a great need to coordinate and facilitate a germpla.sm exchange mechanism between the different countries. 2. Drought avoiding or toleraDt cultivars Drought has been a recurrent problern in Africa. The major cereals are sometimes completely destroyed. Small millets are the ones that survive and give at least some yield. Therefore, early maturing an:. high yielding drought avoiders or other cultivars that are high yielding and drought tolerant will be of tremendous value. T'nis area of research is of high priority. 3. Low input cultivars : Developnent of cultivars that are adapted to low fertility and other inputs will be of great value. 4. DIYland aqriculture : JI.long side the developnent of improved cultivars, a whole package of agronomic practices is needed including seed bed preparation methods, water harvesting techniques. Cropping systems to increase the productivity of small milelts in low moisture stress areas is also of high priority.

5. Alterntive ~ for soil fertility restoration and maintenance: 1'.n agricultural system that has not been based on conservation of the land, high population pressure and other­ factors have led to the degradation of the agricultural land. Top soils are lost and fertility status is very low in nany African countries. Srna.li farmers are finding it difficult to meet the ever increasing price of conmercial fertilizers. Hence, alternative emthods for restoring and maintaining the fertility of the soil in smail millet growing areas is anothe~ essential research priority. 15

One can also note that there is a great need for training, and coordinating information exchange on research developments and sharing experiences on how to solve problems like dessimination of research information to fazmers as well as getting feed back for researchers.

Reference 1. Physiological Aspects of Diyland Farming, edited by U.S. Gupta (New Delhi 1982), P. vii. 16

IMPROVEMFNl' OF FINGER MILLET (ELEUSINE CQRACANA) IN fil'HIOPIA

Yilma Kebede and Abebe Menkir*

In Ethiopia, despite fluctuations between years and regions finger millet comprises about 5% of the total cultivated area under cereals, <'ln~ makes up a similar percentage of the total cereal production in the country (Table 1). The cultivation of finger millet is concentrated in the mid and lower altitude regions of Eritrea, Tigrai, Gojjarn, Gonder and Wellega. In these regions, finger millet constitutes 10% to 20% of the total cereal production. From an area of just over 220,000 ha less than 200,000 tons of finger millet is produced, resulting in a national average of below 1 ton/ha.

TABLE 1. Estimated Production of ~Cereal Crops in Ethiopia '000 Tons

1976/77 77/78 78/79 79/80 84/85

Tef f 994.5 1022.4 1083.8 1144.0 873.8 Barley 894.5 689.9 696.8 772.0 749.2 Wheat 605.2 428.9 448.8 469.0 561.3 Maize 947.8 929.1 981.6 977.2 923.9 Sorghum 755.7 708.0 679.9 689.0 489.2 Millet 172.4 207.2 189.6 193.0 187.2

Source Land use planning and regulatory department (1976-80 data) 1982 data book on land use and agriculture in Ethiopia Vol.I, Ministry of Agriculture Addis .Ababa PM3E, ONCCP, cso (1984/85 data) Agricultural Sample Survey 1984/85 (1977 EC) Results of area and production 1985

Finger millet is p:>pular for making local beer and distilled spirit (Areki). Tne grain is also used for bread (injera), although ot."ler grains are rrcre preferred f c.r- t.'ris purp::;se. I:-:. t.11.e majcr pro:iucl.:-;g re~ionE , i 't does relatively well in drier years an:i becomes a fa'Tli.ne crop for tanners (Asrat, 1965).

*Research Officer, I~.R, Nazareth, Box 2003, Addis Ababa, Et'..hiopia l7

The rrost important advantage of this crop is that the seed can be stored for a longer period without the use of insecticides. It is also considered to be free of the major pests and diseases, and unlike sorghum is not favored by birds (Cloutier, 1984).

Millet Improvement Historical Sketch

The earliest reported w::irk on finger millet was from Debre Zeit experiment station, where preliminary results indicated that as a group white seeded types had higher yields compared to red and dark seed types (Asrat, 1965).

Later on, the Ethiopian Sorghum Improvement Program considered the possibility of introducing millets as alternate crops for noisture stress areas of the country. In 1979, sixty millet accessions, ten each of finger, proso, little, kodo, foxtail and barnyard millets were received from ICRISAT and evaluated at the research station at Melkassa (Nazareth). ~Y of a these millets were not as early as expected. Only proso and f O:il."tail appeared promising.

/.J1other consignment of 247 foxtail millet and 35 barnyard millet accessions was received and evaluated in 1980. A trial consisting of so.re foxtail selections was carried out in subsequent years but the entries did not perform as expected.

Further "-Ork

Recognizing the importance of finger millet as a potential dryland crop in the country, a program was initiated in 1985, aimed at identifying high yielding, lodging and disease resistant lines for the major millet areas.

Finger millet is considered indigenous to Ethiopia (Huffnagel, 1961) and occupies diverse agro-ecological situations. There is a vast ra.11ge of genetic variability in indigenous Ethiopian germplasm. Taking into accolli1t that assembly of the genetic resources and identification o: important traits is essential for attaining breeding objectives, the Plant Genetic Resources Center/Ethiopia has assembled over 700 accessions of finger millet both from indigenous and foreign locations. The majority of the indigenous collections are from Gojja11, Gonder and. Wellegs. administrative districts. Cl".aracterization data indicated the range cf variation in seedling of vigor, days to maturity ( 102-163) , rr.l'7t...1>er o: fingers (6-11.) and ears (16-72) length of finge:::-s (5-14 cm) ::.rid r:;:a:-,t height ( 40-142 cm) as well as seed color (light to dark). So far, our :research efforts have focw:ed on isolating fTCT:"ising genotypes f ran indigenous collections grow:1 at Melkassa an:: further evaluating them in possible areas of product1ci~. In conjW1ct1c!"l wi t.'1 the 18

Plant Genetic Resource Center/Ethiopia, a total of 524 finger millet accessions were evaluated for various agronanic and rrorphologic characteristics in 1985 and 1986 (Table 2). Just under 50% of the collections were indigenous. The rest were introduction from Zimbabwe, Burundi and Nepal. out of these accessions, 44 early and 69 medium-late maturing types were selected and included in advanced observation nurseries (Table 3). Nurseries consisting of early maturing millets were grown and evaluated at low elevation locations (Melkassa, Mi.esso and Kobo). Medium-late maturing types were tested at intennediate (Pawe and high {Adet) elevation sites. Based on agronomic excellence and adaptation as well as tolerance to lodging and head blast, some pronusing entries have been identified and included in variety trials.

Table 2. Selection from 1985 and 1986 Grow ~of Finger Millet Accessions EJ.anted at Melkassa

Total Source Accessions Percent evaluated Advanced Selected

186 IND. 40 22 1985 25 INTRO. 22 88

63 IND. 36 57 1986 250 INTRO. 15 6

IND - Indigenous INTRO - Introduction

Table 3. Classification of 1985 and 1986 Finger Millet Selections ~ on Maturity snQ. ~ion for Testing

Days of Maturity 1985 1986 Test location

Early < 110 29 15 MK, MI, KB

Interm. 110-130 17 36 PW, AfJ Late > 130 16

MK - Melkassa, MI - Mi.eso, PW - Pawe, AfJ - Met

In a finger millet prelimir.ary var::.ety trial gro;..n at MelY.assa a..'1d Kobo, yield levels averaged. over locations ·.,raried from 15 tc 34 tor...s/lli:.. About 50% of the accessions produced yields of 25 tons/ha or better 19

(Table 4). lt>re variety and observation trials have been planned for low (Kobo, Miesse, Melkassa), intermediate (Pawe) and high (Met) elevation testing sites during the coming season.

Table 4. Grain Yield~ Days to Flowerinq of Selected Finger Millet Entries (Based on Yield and Earliness) frm g_ Preliminary Variety Trial Grown at Melkassa J.M!U and Kobo Jrn, 1986

Accession No. Grain Yield Flowering ------MK KB x MK KB x tons/ha days 203253 1.8 4.9 3.4 79 74 76 203261 2.0 4.4 3.2 81 78 80 100008X 2.5 3.3 2.9 75 73 74 100007X 1. 9 2.8 2.4 74 73 74 100058 1.4 3.3 2.4 81 76 78 76T1=23+ 1.8 2.0 1. 9 63 68 66

- X indigenous - + early sorghum Conclusion

Sane progress has been made in the area of finger millet ~rovement in Ethiopia. Nevertheless, there is a lot to be done in identifying production constraints, refining reconmendation danains, defining the requirements of the small farmers, and establishing several testing sites representing the major production zones. We have yet to initiate 'WOrk on developing appropriate management systems and crop protection measures. Furthernore, 'WOrk on nutritional characteristic of finger millet gennplasm as well as for processing and utilization of the crop need due attention.

REFERENCES

1. P£rat Feleke, 1965. A progress report on cereal and oil seed research, 1955-1963. Branch Experiment Station, Debare Zeit, Ethiopia.

2. C2.out1er / P. 1984. AgrJ.cJlture-strategies for reclamation ar1d developnent. Land use Planning and Regulatory Department of the Ministry of Agriculture and FMJ, Addis Ababa. 20

3. Huffnagel, H.P. 1961. Agriculture in Ethiopia FAO, Rane.

DISCU$SICli

1) The relationship between maturity and altitude is the opposite in South Asia compared with Ethiopia : In India and Nepal, shorter maturity millets are required for higher elevation areas because of low temperature at the end of the growing season, while in Ethiopia, shorter maturing millets are needed for low elevation areas because of a limited rainy season and rroisture stress.

2) It was suggested that diverse early types of finger millet frcxn Zimbabwe and Uganda might have potential in Ethiopia.

3) It was mentioned that very early types of finger millet have been collected recently in Tigrai (Northern Ethiopia). These collections are now being evaluated by PGRC/E.

4) Other species of small millets still need to be evaluated in Ethiopia, but the present millet program will concentrate on finger millet. 21

llm. ACTMTIES Qf. 'lHE flMI_ GENEI'IC RE&URCE Cml'RE/ E'IHIOl'IA CPGRG/El 00 'l'EFF' HiQ. MINOR MILLETS Fassil KebebeW*

Ethiopia is believed to be an important gene centre for various crops including teff, wheat, barley, sorghum and millets.

Since the inception of the PGRC/E in 1976, its objectives focus on exploration and collection, multiplication/characterization, documentation, distribution and utilisation of indigenous crop species available in the country.

A total of 1602 accessions of tef f and 601 accessions of minor millets have been collected from different agroecological zones in the country. Rational planning for future collections based on diversity, survey and study is being undertaken. Rescue missions during the drought crisis and a systematic collection of large samples of indigenous landraces in drought prone areas are being carried out.

Characterization and evaluation of gennplasm is undertaken using appropriate descriptors that are based on breeders need.

There are about 2650 accessions of teff and 1056 accessions of miner millets in the holdings of PGRC/E. A rrore diversified conservation is being developed in the centre that include in situ conservation to back­ up the existing ex situ conservation.

The Centre is distributing gennplasm materials to breeders and scientists working in and outside the country for identification and subsequent utilisation of gennplasm. Excess seed from each accession of tef f and minor millets bulked together crop by crop are also distributed to breeding stations , settlement areas and the Ministrf of Agriculture.

* Research Officer PGRC/E, Box 30726, Addis Ababa, Ethiopia. 22 Discussion 1) It was clarified that the bulk of the minor millet holdings in FGRC are finger millet. Teff is listed separately.

2) The IBPGR descriptors should be used when exchanging seed. However, a general plant aspect score, which gives an indication of a plants general perfonnance in a particular environment is also very useful.

3) The excess seed from the germplasm evaluations are bulked together by PGRC and sent to individuals who have agricultural training in breeding station settlement areas and the Ministry of Agriculture. Single plants which perfonn well in these locations are selected using the knowledge of farmers and ull.i1~~~lied as potential varieties. 23

SMALL MITJ,ET IMPRO\TEMENl' IN txil\NDA

Sam E. Odelle

Introduction

Finger millet (Eleusine coracana) has been grown in Uganda for over 5,000 years. Its improvement in the country is perhaps as old as the crop itself . The practice of selecting and preserving phenotypically attractive panicles to form the next years seeds form the first step towards the improvement of the local varieties, and it is till conm::in to date.

Prior to 1965, however, breeders had introduced and tested a number of local varieties at Serere Research Station for yield, disease resistance and other desirable characateristics.

Early 1965, serious breeding v.ork was started at the station. It is today twenty one years since that time, yet the problems the breeders face today remain al.rrost unchanged. the problerns of developing high yielding varieties, early maturing varieties for rrarginal rainfall areas of the country, lodging resistant varieties, blast and virus disease resistance and good grain quality among other things, pose a big challage to breeders in the face of continuous need to increase the national finger millet grain production to meet the ever growing food demands. Improved varieties like Engeny, Gulu E, Serere 1 and P 224, consistently outyield the local varieties, but have not been fully accepted by the farmers or the consumers, with the result that the technology developed, has so far made little impact on the people for whom it is meant. Coupled to these, the political unrests and insecurity followed by a number of liberation wars during the last twenty years of independence have had very negative impacts on research.

Finger Millet Research Objectives

T'ne broad objectives of finger millet improvement progranme are

1) To continue to evaluate the performance and adaptibility of improved entries screed from existing prograrrrne in as many locations or environments as possible. 24

2) To study the biology and epidemiology of blast disease and develop effective blast resistant varieties for utilisation in resistance breeding.

3) To assemble and establish a large finger millet germplasrn and their ~~ld relatives and populations from as many locations as possible particularly within Uganda.

4) To continue to select for lodging resistance.

5) To select for high yielding early maturing varieties for low rainfall areas of the country.

Small nullet qermolasrn resources

At the end of 1984 and the beginning of 1985, three collecting missions were undertaken by Sorghum and Millets unit, Serere Research Station with the International Board of Plant Genetic Resources (IBPGR} sponsorship and assistance. Collections were made in the Northern, F.astern and Western regions. The objective of the mission was to sample and assess variability of gennplasrn available, to estimate the present degree of genetic erosion and thus propose strategies to prcmote conservation and use of the gennplasrn collected in active breeding prc·;rarrrnes .

At the end of the mission, 684 accessions of finger millet and their wild relatives were collected. This has boosted the number of our w:irld collection of finger millet to 1958. We also have 422 accessions cf rrinor millets, consisting of foxtail millet ( Setaria i talica) , Ciorm'on miliet (Panicurn miliaceum), barnyard millet (Echinochola furnentaceae) and teff (Eragrostis tef l 'Which are showing prolific growth in the observation plots. All the 1058 accession of ~rld collection of finger millet were planted for observation and maintenance. Imnediate selection was made for crossing with our elite lines. Crosses were made using hct water methods and male steriies from populations. :t-bst of these crosses are geared towards blast resis~ance a.~d basal tillering. It is proposed that aii the gennplasm will undergo characterisation and preliminary evaluation to enable us to identify any genotypes both in the breeding progranme and in the field.

Selection of high yielding varieties

In our ef f cns to improve our existing entries in the pipeline for release to farmers, on the spot selections are made from the collection on the basis of yield and agronomic characters. Seventy one such selectio::-..s "1-P-re rr.:ide, and grown together WJ.. th 29 p::1pular local varieties in a 10 x 10 ~riple lattice design. Grain yield, days to flower an:5. plant 25

heights were the observations first taken. Further tests on the same materials are planned.

Emasculation and hybridisation techniques

One of the problems of v.orking with finger millet is the difficulty of making crosses. A number of methods have been tried and developed at Serere Research Station and either used singly or in combinations with another.

The male sterility developed at the station is being used in mating populations composited from several good varieties.

Hot water emasculation using 47 degrees centigrade for 10 minutes and 48 degrees centigrade for 7 minutes respectively have continued giving between 10%-15% seed set. We have however noticed varietal responses to hot water treatment which sometimes gives unreliable results.

Where high accuracy is required and the parents do not have any visible gene markers, hand emasculation using fine needle and magnifocuser is normally applied. It is tedious, slow and expensive but gives about 70%-90% success, especially when applied i:', combination with contact method of pollination. We are using this method in the study of blast disease.

Hand pollination using special paper bags for collecting and dusting pollen on the f ernale panicles is carried out four days after flowering started and between 7 and 9 a.rn.

No single clear cut hybridization method can be reccmnended so far for this crop.

Male Sterility and Population lnm'ovement

From the male sterile populations developed at the station, cross:::.ng with good local and exotic materials is possible.

Good seed set have been obtained as a result of having identified the correct time for pollination.

Fl crosses made with tmknown testers were yield tested ·..:ith tbei.:­ parents for hybrid perfcrmance in 7 x 7 triple lattice design. Six hybric 26 populations outy1elded their parents, but by very small margins.

Population 3 C3-C7 consisting of disease resistant material was yield tested in 8 x 8 lattice design during first rains 1985. The trial is being repeated this year. The material shows some level of field resistance. It is hoped that in 1987, it should be tested in a hot spot for blast. In another population (4 C3-C5) consisting of selections for earliness and lodging resistance, progress has been slow in lowering the number of days to flower and reducing plant height. A total of 471 crosses, made between populations and local and exotic parents, are in F2, F3 and F4 generations, for further selections in these segretating progenies are being made.

Develoµnent of White Seeded Finger Millet Varieties

The white seeded finger millet are attractive for food and have good malting quality for brew. The white seeded introductions WC 717 and WC 65 from Ethiopia and India are poorly adapted and are susceptible to blast and virus diseases. The Indian variety, with shorter and thicker straw, is resistant to lodging. A number of crosses were made on to WC 717 and WC 65 using locally adapted varieties, with a view to developing high yielding white seeded varieties resistant to lodging, blast and virus diseases. Following a number cf bad~crosses and reselect ions thirteen white see.:!ed varieties have been developed, and are being multiplied for yield trials.

Grain ooulds, leaf diseases and blast still persist on green pigmented varieties and they are generally late flowering comapred to their purple pigmented counter parts.

Early maturing f inqer millet

Early maturing finger millet varieties have high potentials in areas of low and uncertain rainfall. They can also be grown during the shorter, lighter second rains in certain parts of the country.

Attempts are being made to develop varieties ~ith short maturity periods (below ninety days) and high basal tillering capacities. Twenty six early maturing populations were planted with 95 selections in 11 x 11 triple lattice design in a screening trial. M::>st varieties flowered under 70 days. Yields, however, were lower than for the longer mat:.:.ring varieties. Four entries produced over 25 q/ha.

It is expected that the re~ently assembled local germplasm will prcvide a good base for selecting fer earliness. 27

Finger Millet Variety Trials

The finger millet variety consisted of a 5 x 5 balanced lattice design, and the seed sent to sixteen centres in Uganda. Data for 1985 came only from four centres. Ten of the entries in the trial were crosses made in the station while the rest are elite lines screened from the collection for yield and blast resistance. Significant differences in grain yield were observed in all locations. Grain yields at Masafu and Serere variety trial centres were low and the reason was attributed to poor management. P 211 yielded highest with a mean of 2.11 tons/ha though not significantly higher than the check varieties P 224, Serere 1, Gulu E and Engeny. A majority of the entries in this trial show high levels of blast and lodging resistance.

Release of finger millet varieties

Despite considerable difficulties in the country at large, the unit has forged ahead and released a number of varieties to farmers. Recently P 224 has officially been added to the list of Engeny, Gulu E and Serere 1. Another corss i.e. Serere x 10 is being considered arrong other sin the pipeline, for the next release.

Discussion

1) It was agreed that the Uganda program had developed a sound and systematic approach for finger millet breeding and improvement.

2) It was suggested that the male sterile source v.ould be very usefully tested in programs of other countries. 28

SMALL MILLETS - THEIR IMPORTANCE, RESEAROI AND DEVELOPMENTAL STATUS IN INDIA AND SOOTH ASIA*

A. Seetharam

Area and Ilrp:>rtance

The term small millets refers to a group of small seeded cereals finger millet, little millet, proso millet, barnyard millet, kcxio millet and foxtail millet. In South Asia, including India, these crops are grown in large areas, rrostly as rainfed crops under marginal and sub-marginal conditions of soil fertility and rroisture. Obviously, the grain yields harvested from these crops are relatively lower when compared with major cereals grown under good management.Nevertheless, these crops have become indispensable in the areas of their cultivation, as they :i;::x:>ssess high drought tolerance and adaptation to diverse agro-climatic conditions. Besides, these crops ~ not attacked by any major pests and. diseases. The grains are consumed in a variety of ways, and f onn a major comp::ment in the daily diets of a vast section of rural people. The grains are nutritionally rich and comparable to major cereal grains in protein content, but superior in mineral corrp:>sition and content. the dry and fresh forage from these crops are a source of good quality fodder for bovines.

ATJOng the countries of South Asia, India, has the largest acreage under small millets (Table 1). Finger millet is the oost i.rrq;x:>rtant small millet in India, Bhutan, Nepal and Sri Lanka. Foxtail millet and p~oso millet are important in both India and Bangladesh. Kodo, little and barnyard millets are grown only in Indian sub-continent. Small millets area in India is around 6.0 rn. ha, producing 4.45 million metric ton."1es of food grain. In India, finger millet alone accounts fer 35% area and 80% of total production of small millets. After India, finger rr~llet area is highest in Nepal.

* This paper contains information presented by participants from India, Sri Lanka, Nepal and Bangladesh at the First International Small Millets \tk>rkshop, Bangalore. Sept 29 - Oct 3, 1986. 29

Table 1. Area and production of small millets in South Asia

Production Country and Year Area (ha) (metric tonnes) Crop(s)

India (1984) 2,387,500 2,532,000 Finger millet India (1984) 3,393,000 1,319,400 Kodo, little fOA"tail proso and barnyard millet

Bhutan (1985) 10,620 11,880 Finger, foxtail and proso millet

Sri Lanka (1983) 20,593 11,662 Finger millet

Bangladesh (1983) 37,600 35,298 Proso and f oA"tail millet Nepal (1985) 151,000 137,900 Finger millet

Total 6,000,313 4,048,140

Res~~ch and Develoµnent

Though crop improvement v.ork on small millets has been going on for several years, these crops have received less attention carpared to ma)or food crops. It is only after 1969, that millet crops started receiving sane attention with the launching of Co-ordinated Millets ~rovement Project. Small millets improvanent received a boost during 1978-79 with the establishment of 5 lead centres under IDRC assistance. The research on these crops included rrore centres during the 6th Plan, enveloping major discipli.11es of crop production. These efforts have started bearing fruit and many new technologies and varieties are now undergoing critical evaluation in different parts of the country. During the 7th Plan (1985- 90), a separate All India Coordinated Small :t'd.llets Improvement Project has been launched with a netv.ork of 12 research centres and rrore than 25 co-operating test centres. The major objectives are :

(1) Diversification of varietal base by evolving high yielding disease resistant and widely adaptable genotypes of finger, kcx5.o, little, foxtail, proso and barnyard millets.

(2) Develop.rent of efficient production techniques using low rnonitary inputs, identification of suitable syste"llS for JTO~~, rr~xej and relay crcp~~~;. 30

(3) Identification of ideal crop mixtures and evolving production systems involving pulses and oilseeds as canponent crops.

(4) Intensification of research on plant health and evolution of cheap and efficient plant protection methods.

(5) Identification of alternate uses for grains in poultry, dairy and in agro-based industries to enhance the economic value of these crops.

Yield ~ovement through creneticallY superior varieties M:my varieties of different maturity durations have been developed and released for general cultivation in different parts of India in all small millets (Table 2 and 3) . The improved varieties of ragi are very popular am:mg the farming corrrnunity in the Southern States of Karnataka, Tamil Nadu and Andhra Pradesh. The enhancement of coverage under improved varieties of ragi and adoption of improved cultivation practices has helped. in al.rrost doubling ragi productivity in the last 15 years. However, in other small millets the irrQ?act of the developnent and release of new varieties has not been significant as coverage under improved varieties has been very slow and poor.

Inu:?roved m management gOO, agro-techni®es : Improved pac}r..age c: production techniques for finger and other small millets have ?::'6Cr: developed in different states. The improved agro-techniques for drylan:i crop includes advance land preparation for m:Jisture conservation a::~ efficient rroisture utilization, balanced inorganic fertiliza~io~; nitrogen saving devices through biof ertilization, rernunerati ve sequence, mixed and relay cropping systems, cultural methods for weed control and pest and disease management by a combination of chemical, cultural and use of resistant varieties. Similariy, separate agronomic packages have been developed for irrigated finger millet in different states. 31

Table 2. Varieties of Ragi reccmnended in various states in India for different seasons

State Season Variety Sowing 1-t:mth

Andhra Pradesh Early kharif AKP 2 May Kharif VZM 1, Sharada, Godavari, July Kalyani Rabi Simhadri, Kalyani Aug-Sept.

Bihar Kharif BR2 I BR407 I RAUS I HR 374 June SUrrmer BR2, BR407 February Gujarat Kharif Gujarat Nagli 1 June-July Himachal Pradesh - High altitude Kharif \TL 204, Local June - I.ow altitude Kharif Local May

Karnataka Kar HR 374, Indaf 9 April-May Kharif Indaf 1, Indaf 3, Indaf 8 June-July HR 911 and HR 919 Late Kharif PR 202, Indaf S, Indaf 9, August and HR 374 Rabi Inda£ 7 and Indaf 9 Sept--Oct.

SUntner Indaf 5 t Indaf 9 and Jan-Feb. HR 911 Maharshtra Kharif B 11, A 16, E 11, June 28-1, 50-1

Madhya Pradesh Kharif JNR 8S2,IE 28, EC4840, June-July JNR 1008 and JNR 981 Orissa Kharif Dibyasingha, B-4-10-S6, June Bhudai Local Rabi Dibyasingha, B-4-10-56, January Bhudai Local Tamil Nadu Nargalipattam CO 7, CO 10, KS, K 6, May-June and Indaf S. Chitaraipattam PR 202, CO 11, K 6, K 7 April-May and co 12 Chitaraipattam KS, K6, CO 11, CO 12, K 7 May-June Adipattam PR 202, KS and K7 June-July Purattasipattam K6 and CO 11 sept--Oct. Uttar Pradesh - Plains Kharif Ni.nnal, T 36-b, PES 176 1'2y-Ju.11e - Hills Kharif \TL 204 and !Deal June 32

Table 3. Varieties of small millets recommended for different states in India

State Crops

Foxtail lllllet Kade millet Little millet Prose millet Barnyard millet

Andhra Pradesh ArJuna, SIA 326 PSC 10 CO 1, CO 2 Varada Local, COl Chitra

Bihar Local, SIA 326 - V 15, V17 BR 7, Ral cheena RAU 2, RAU 3 Shyu cheena RAU 9

Karnataka K 221-1, RS 118 PSC 1, PSC 2 C01,C02 C03 m 364 PRCJ

Maharashtra ARJUllA, SIC 3 Varada, lo.11

Madhya Pradesh - IPS 147-1, RPS 41 Dindori 1 Kebarpur, Dindori 2 RPS 123, m 364 PRC 3I PSC 1 PSC 2, RPS 62-3

Orissa Xoraput, Local - SS 81-1 Bodosat.'an PRC J

Tamil Madu C04,C05,K.I: co 3, Y.l CO 1, C02, YJ Xl, CO 3, CO 3 f.2, co ! rJttar Pradesh PSC 1, PSC 2 v 15, 'l 17 BR 7 'lL 8, VL 11 Ram Cheena Anurag Shyam Cheena 33

Bhutc.r:

Millets with 10,620 ha are the third major food comrodity ne>.1: only to rice and maize. Though finger, foxtail and prose millets are grown, finger millet occupies the major area and is chiefly used for making country liquor. The increasing food demands for human consumption require diversion of finger millet for direct food. Millets are grown under rainfed conditions and planted in April (in Upper Pradesh hills) , in May (in Mid hills) and in July (in foot hills ) . The land is ploughed manually before finger millet is transplanted. Weeding is done by manual labor. Earheads are harvested and the threshed grain is stored in bamboo bins.

So far very little attention has been given to these crops, which are grown using primitive agricultural practices. There are no improved varieties, and only local cultivars are grown. Since local cultivars are grow!:, there may be considerable diversity existing and this needs to be collected and conserved.

Finger millet with rrore than 150,000 hectares is a very important food crop of Nepal ranking fourth in area and production after rice, maize and wheat. The grains are not only a source of food for people livin3' in hills but also a source of dry straw for forage.

Only limited attention has been given to finger millet .improvement in Nepal. Since the early 1970s, about 700 exotic (rrostly Indian) and 100 local land races have been evaluated. this resulted in the identificatio:i of tw~ improved varieties - Dalle-1 and OJr.l'lale-1, released in 1980. However, these varieties have limited range of adaptability.

There is need for extensive collection and testing of both lQC4l and exotic germplasm for identification of premising lines suitable for different elevations, water logged conditions, relay cropping an1 pos.sessing disease resistance.

In recent years the importance of finger millet to Nepal 's agriculture has been reco;nized and prograrnne under 'Hill Crops Improvement' for the develop.nent of finger millet has been drawn for the overall improvement of the crop.

Finger, foxtail and prose millets are grown in Sri Lanka but finger millet is most widely grown. It is grown in dry and intermediate zones where ar..nual mean rainfall range from 875-1875 rrm. Finger millet is grcwn in 'Y..::.ha ' season from October to Januar/ and it ranked third arrongst cer~ls with an area of 20,593 ha and 11,662 tonnes of production duri~; 1983. Introduced varieties particularly from India gave 15-25% rrore grain yield than the reconrnended variety M 1312 of finger millet. Some of the pro.T~sing varieties of finger rr~llet fro.~ the intro:iuctions are C0-10, 34

KM-1, JNR 3b, HR 23-1 and PPR 1091.

seed is generally broadcasted under shifting cultivation which :s still prevalent in some areas and transplanted under irrigation. 16 kg N, 22 kg P205 and 18 kg K20/ha are recomnended for finger millet in all areas. Important diseases include blast disease (neck, finger and leaf) and helminthosporitnn leaf spot. The important insect pests are borer and earhead caterpillers. Work is in progress for the control of these pests and diseases through resistance breeding.

Bangladesh In Bangladesh, proso and fm1:ail.. millets are the important srrell millets grown, on an area of 40,000 ha, producing 35,000 tonnes of grain. Millets research in Bangladesh has been receiving attention only recently after the initiation of millets research project supported by !DRC. ~ith the launching of the project, an attempt has been made to collect 10"'--=.l and exotic gennplasm. Some high yielding varieties from local and exotic collections have been isolated and are undergoing mu.ltilocaticnal evaluation and in yield trials. An exotic white grain proso millet. has been identified and reconmended for the national seed board for release.

Besides evaluation of varieties in different agro-climatic belts, production technology in tenns of seed rate (8 ~:g/ha) and inorganic m3Iluring at the rate of 40:30:30 kg NPK/ha have been recomnended.

The promising cultivars identified are BSi-1 and BSi-5 in foxtail millet and BPm 14, BPm 21, BPm 34 and PBm 40 in proso millet. In fo>..1:a1l millet the improved varieties Bogra 1, Parameshpur, :2.hitnagar ~-:d. Lakshmipur 2. These varieties have recorded 30-40% rrore seed yield than local types, and are undergoing denonstration in large plots.

Conservation of Genetic Resources

During the last 8-10 years intensive efforts have been made to augment germplasm of all small millets . The small millets gerrnplasm unit at Bangalore has 9,595 collections which is probably the largest base collection maintained at one place in the world. About 4,500 collections of finger millet, 2,000 accessions of foxtail millet and rrore than 3,000 collections of other small millets are available for use. M:my promising genotypes of breeding value including sources of resistance to varia..is diseases have been identified and are being extensively involved in the recombination breeding in foxtail, finger, barnyard and proso millets. Banaladesh

T.ie rrQllet research progranme in Bangladesh has 518 collections c: foxtail millet and 206 collectior.s cf proso millet. The collectio:-1 w:d: is cc~tinuing and the ccllected germplasm are at different stages of evaluation. Prorrising germplasm identified are already undergoi.~g 35 evaluation for direct utilization. Bhutan There is no systematic ccllection and conservation of germplasm in Bhutan. Bhutan could be a good source for collection of local germplasm particularly in finger millet.

Recently about 350 local land races and 150 exotic introductions mainly from East Africa have been made. there is still scope for collection and conservation of finger millet germplasm in Nepal.

A limited number of local millet germplasm accession are maintained in dry zone research stations - M3ha Illuppallai~, Karadianaru, Kilinochchi and Angunukolapelessa. As area under millets are declining, greater efforts are needed to collect and conserve the local land races of all small millets. ICRISAT. India The International Crops Research Institute for the Semi Arid Tropics, Patancheru, Hyderabad, India, is maintaining one of the m:>st cc:mprehensive collections of smsll millets germplasm. Presently 5,563 accessions are maintained. In addition another 1,060 new exo~ic accessions have been added during 1986.

All germplasms are conserved under medium storage conditions at 4 degrees centigrade and 20% RH and available for use by scientists. Future Research Priorities The following cOllll"Cn future research priorities in small millets irrq:>rovement are identified for South Asia region :

1) Conservation Qi genetic resources : There is unexplored reservoir of genetic resources of millets in parts of South Asia. It is important to collect and conserve the precious germplasm from these unexplored areas.

2) Productivity improvement through imProved varieties : Except India, in other regions not much crop improvement work has been done. Still local cultivars are grown in large areas. The irrq:>roved cultivars identified in different small millets need to be populariseC.. F\l.rther there is also a need to diversify the varietal base ty developing not only location specific but also widely adaptable varieties. 36

3) Stabilization of yields in rainfed cro~ : M.Jre than 95% of the area under small millets are rainfed and grown in marginal and sub­ marginal conditions. This has resulted in wide fluctuations in production between regions and years. Appropriate technology generation and transfer is needed to impart stability in the production of millets.

4) Recombination breeding : Except in India, no systematic breeding ~rk has been taken up in other countries of South Asia. Even in India except ragi, in other small millets only limit~ l:"~ombination breeding ~rk has been undertaken. One of the major reasons for restricted hybridization in small millets has been its small flower si::;e. Nevertheless, this problem could be surrrounted by adapting either contact method of hybridization or hot water emasculation or hand emasculation itself. There is vast scope for initiating massive hybridization progranme in all small millets.

5) Bio-fertilizers Finger millet and other small millets do respond to biofertilization. The imnediate need is to standardize effective methods of bio-fertilization and identification of more potent strains specific to these crops.

6) ~ and relay cropping mtems : }i:)re remunerative sequences of mixed and relay cropping systems are to be identified to maximize income.

7) Resistance breed,ing Resistance breeding progranmes are to be initiated for blast in ragi, helminthosporiurn leaf spot in both ragi and foxtail millet, smut in foxtail millet, barnyard millet and kodo millet. Similarly among insect pests, shootfly is important in prose millet and little millet; leaf and root aphids in ragi, and stem borer in foxtail, finger and barnyard millets.

8) I.ow ~ technology : The small millets respond well to low inputs and the low cost cultivation practices. This area of research needs priority attention. 9) Alternate ™ fQr. grains The small millet grains are nutritionally rich. Limited studies have shown that the grains could be used with advantage in poultry, in animal and feed industry and also in the manufacture of starch, alcoholic beverages and in confe~tionary. This area needs further critical examination as this enhances the value of these crops.

Discussion

1) In response to a question on how widely improved minor rr~liet tech.'1ol09:i' was beJ.ng used by the f arrners, Dr. Seethara'11 indicated th=.:. in S:>u~h Ind.ic. improve:. finge:!:" rri.llet varie~ies are being gro;.m 37

on about 50% of the total area. A large portion of the fo>."tail millet area in the low rainfall black soil areas of Andhra Pradesh is now growing improved varieties. Improved technology for the other small millets did not yet have as rm.lch impact.

2) A strong seed rm.lltiplication program is essential if improved varieties are to have an impact.

3) Cooperative small millet trials are, for the first time this year, being carried out under the South Asia Association for Regional Cooperation ( SAARC) .

4) Blast disease is prevalent in finger millet in South Asia. It may account for a 15 to 20% crop loss. MINOR MILLETS GERMPLASM RESOURCES AT ICRISAT K.E. Prasad.a Rao and M.H. Mengesha *

Minor millets (small millets) are important in the Semi-Arid Tropics. Presently the total number of minor millets accessions assembled and conserved at International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) is 5563 as detailed in Table 1.

Table 1. Accessions of Minor Millets at ICRISAT

Crops . No. of Holdings Countries

Finger millet (Eleusine coracana) 1993 17 Foxtail millet (Setaria italica) 1328 20 Prose millet (Panicum miliaceum) 757 25 Little millet (Panicum sumatrense 377 1 Barnyard millet (Echinochloa fil2l2.:...) 582 8 Kodo millet (Paspalurn scrobiculatum) 526 2 Total 5563

During 1986, we acquired 1060 new accessions; from Bunmdi (11), Korea ( 126) , Uganda ( 596) , USSR ( 143) and Zimbabwe ( 184) through National Bureau of Plant Genetic Resources (NBPGR), New Delhi, India. T'nese were planted in the Post Entry Quarantine Isolation Area (PEQIA) fer inspection and release. A detailed listing of holdings and countries of origin is presented in Table 1A at the end of this paper.

Assembly and Conservation

One collection mission for Eleusine gerrrrplasrn in Ethiopia W-QS organised in collaboration in Indian Council of Agricultural Research (ICAR) and a scientist from ICAR was deputed for the purpose. 217 samples of Eleusine have been collected which are yet to be received fran Ethiopia.

* Pro;ram Leader, Genetic Resources, ICRISAT 39

M:>st of the minor millets germplasm was obtained by correspondence and some accessions were collected during collection missions launched primarily for other mandate crops.

In the year 1986, a collaborative project between ICRISAT and NBPGR/ICAR has been initiated and we received 466 new accessions of six crop species from NBPGR Regional Research Station, Akola which is a gcx:xi addition to our present collection. We have transferred one complete set of minor millets germplasm available at ICRISAT comprising of 5053 accessions to NBPGR for evaluation at P.kola 1 Maharashtra, India.

All the minor millets germplasm are conserved under medium term storage conditions at 4 degrees centigrade and 20% RH.

In accordance with the reconrnendation of the IBPGR Mvisory Conmittee en Sorghum and Millets, ICRISAT, started assigning accession numbers to minor millets as follows :

Eleusine coracana IE Panicum miliaceum !Pm Panicum sumatrense !Pmr Paspalum scrobiculatum !Ps Echinochloa crusgalli !Ee Setaria italica !Se

Maintenance All minor millets germplasm accessions are planted for seed increase nonnally in the rainy season to meet the seed requests from institutions from India and abroad.

Since all the six minor millets crop species are reported to be mostly self pollinated, they are maintained without any pollination control. Evaluation IEPGR has assigned special responsibility to ICRISAT by providing funds to characterize the available germplasm at ICRISAT. Accordingly the following accessions as shown in Table 2 have been characterized for important morphoagronomic descriptors. 40

Table 2. Evaluation of Minor Millets at ICRISAT No. of accessions No. of Crop evaluated descriptors

Echinochloa ~ 517 39 Pas"Palum scrobiculatum 308 38 Setaria italica 1260 34 Panicum miliaceum 753 37 Eleusine coracana 1863 32 Panicum surnatrense 291 33

As a part of evaluation exercise, the six minor millets species have been classified in collaboration with t.~e Crop Evaluation Laboratory, University of Illinois, Urbana, USA, as shown in Table 3.

Table 3. sumnarv of Classification Qf. Minor Millets

Crop Species SUb species Races SUb-races

Fo».'tail millet 2 2 7 9 Prose millet 1 3 5 Little millet 1 2 2 4 Finger millet 2 2 7 10 Barnyard millet 3 4 6 Kodo millet 1 4

The classification of these minor millets is presented in Figure 1A at the end of this paper. PocLUnentation gng, Cgnputerization of Evaluation Data All the evaluation data have been documented and the data of the following minor millets have been ccmputerized using IDMRS (ICRISAT Data Management Retrieval System) program.

Evaluation data is now available for

1. Finger millet IE 1 to IE 3567 2. Foxtail millet I Se 1 to ISe 1409 3. Little millet IPrnr 1 to IPrnr 852

Evaluation data of other crops are yet to be co:::pu:::er-ized. 41

Distribution of Germplasm for Utilization Since ICRISAT accepted the responsibility to become a major world repository for minor millets germplasm, seed indents have increased from Indian Institutions as well as other institutions outside India. So far 25169 seed samples have been distributed from the ICRISAT gene bank. The yearwise distribution to India and Institutions abroad is presented in Table 3.

Table 3. Distribution of Minor Millet Seed Samples Indian Institutions Year Institutes Abroad Total

1978 2721 973 3694 1979 4004 608 4612 1980 1345 1840 3185 1981 3144 1791 4935 1982 2080 911 2991 1983 212 540 752 1984 200 200 1985 165 165 1986 2138 427 1565 1987 1196 874 2070

Total 16840 8329 25169

Details of this distribution are fou.11d in Table 2A at the end of this paper.

Future Work Plan

The importance of minor millets is increasing in SAT countries. Since no other international center is actively involved in the conservation of world collection of minor millets, ICRISAT will continue its genetic resources activities with minor millets. Presently the following regions and countries are tentatively identified as priority areas for minor millets germplasm collection. Suggestions for new areas of collection will be appreciated.

1. India 5. West Africa 2. Eastern Africa 6. Japan ..,,,':I Southern Africa 7. C"nina 4. Central Africa 8. Bhutan

Recently ICAR/NBPGR accepted collaborative research ~1th GRU, ICES.!..:' in miner rr:.illets germ;:;las:n both in collection and evaluation. 42

Collection missions have to be organized in collaboration with NBroR/ICAR in India and National Governments in other countries.

ICRlSAT will continue to distribute minor millets germplasm to interested scientists in India and other countries through NBPGR, New Delhi, India.

Discussion

The work of the ICRISAT Genetic Resources Unit was appreciated. It was felt that such work, sup:ported by IBPGR, is useful and should continue. 43

Figure Classification of Slall Millet Species

POACEAE IGRAMIKEAE)

Sub Farnly PAKICOIDEAE Sub-faaily CHLORIDOIDEAE

tnbe:I Pamceae tnbe: Eraqrostideae

BARNYARD MILLET FIHGER MILLET Echinochloa colona IL. I L111X Eleusine coracana (L.) Gaertn. Ech1nochloa crusqall1 IL.) P. Beauv. sub sp. coracana sub sp. af ncana (Kennedy O•Byrne) Hilu & deWet PROSO MILLET Pan1cum 11111laceum L.

TEFF LITTLE MILLET Eraqrost1s tef (Zucc.) Trotter Pan1cum sumatrense Roth. ex. Roem. & Schult.

Y.ODO MILLET Paspalum scrob1culatum L.

FOXTAIL MILLET ~ itallca IL.) P. Beauv. sub sp. ltalica sub sp. vmd1s (L.) Thell. Setana PJ!!ili IPair.) Roem. & Schult.

FON IO Dmtana mlis IKippist) Stapf Dmtana iburua Stapf 44

Table lA. Assembly of SrMll Millets Gernt>lasm at ICRISAT (as of August 1987)

COUntry Finger Foxtail Proso Barnyard Kodo Little Afghanistan 20 17 Argentina 1 Australia 2 Bangladesh 2 Burundi 1 Canada 1 China 60 1 F.gypt 1 Ethiopia 27 1 Hungary 10 9 India 877 881 106 394 524 377 Iran 4 9 Iraq 2 Italy 7 Japan 1 164 Kenya 317 8 1 Lebanon 33 Malawi 245 1 2 Mexico 2 13 M:>zambique 1 Nepal 1 6 Pakistan 1 28 34 ' 1... S. Africa 4 Spain 1 1 Sri Lanka 18 14 2 2 Sudan 2 Syria 119 290 2 Switzerland 1 Tanzania 20 Uganda 40 Turkey 23 48 U.K. 20 4 4 U.S.A. 7 38 65 U.S.S.R. 67 128 16 Yugoslavia 1 West Germany 1 12 Za."Tlbia 117 Zimbabwe 233 Origirl Not Known 49 Q

1993 1328 757 582 526 377 TOTAL 5563 45

Table 2A. Minor Millets Gennplasm Distribution From 1978 To 1987

S.No COUNTRY NO. OF ACC S.No COUNI'RY NO. OF ACC 1 .?\rgentina 25 19 Niger 21 2 Australia 10 20 Nigeria 444

3 Bangladesh 5.....?? 21 Romania 2 4 Brazil 180 22 Rwanda 125

5 Burkina Faso 20 23 Saudi Arabia 60 6 Burundi 25 24 Senegal 29 7 China 266 25 Sri Lanka 200

8 Denmark 30 26 Soolalia 10 9 Ethiopia 432 27 SUdan 30 10 India 16840 28 Thailand 92 11 Indonesia 12 29 Uganda 800

12 Italy 33 30 U.K. 107 13 Kenya 1928 31 U.S.A. 576

14 Korea 292 32 U.S.S.R. 206 15 Mali 89 33 Venezuela 20

16 Mexico 288 34 w. Germany 11 17 MJzambique 30 35 Zambia 185 18 Nepal 816 36 Zimbabwe 413 46

THE PRE.SEN!' STATUS OF SMALL MILLEI'S IN CHINA Chen Jiaju

The species of millets in China, in order of importance, include : 1) foAtail millet (Setaria italica Beauv.), 2) proso millet (Panictnn miliaceum, Linn.), 3) finger millet (Eleusine coracana Gaertn.), 4) japanese or barnyard millet ( Echinochloa f rurnentacea Link) . Foxtail millet ranks first while proso millet is also relatively important. During recent years an agricultural reformation has occurred in the countryside. This has affected millet prcx:luction as well as the millet research program.

The Production of Foxtail Millet

According to the "Yearbook of Chinese Agriculture", the area prcx:lucing foxtail millet is at present approximately 4 million hectares. It was expected that the millet area might recover to a higher level, but it did not occur. The increase soon stopped, and in certain regions became stable or even declined slightly, especially in 1985. After a careful examination of new information, we recognize that foxtail millet prcx:luction is in unstable balance caused by a number of factors, which may be stated as follows :

1) The seeding area of foxtail millet is decided by numerous farmers without government direction, and the proouct is not listed in the state purchase and sale program. When production meets the farmer's own consumption needs, and excess prcx:luction cannot find a good market, the area planted cannot increase. This situation has occurred in the rrost important growing areas, such as Shanxi, In.11er 1-bngolia and Heilongjiang.

2) Tne selling price plays an important role. In the past, foxtail millet was much rrore costly than com, but recently when com became an export ite.11, its price became higher than that of millet. Tne planting of com was encouraged which resulted in the reduction of millet growing area. In Jilin Province, the millet growing area decreased by 100,000 hectares.

3) In regions which are marginal because of arid conditions, sandy or erosicn-prone soils, planting trees and grasses is encouraged instead of millet. In fact, it is not feasible to develo~ fcxtail millet on very ariC. land. Only regions of very short growing season in north Heilongjiang have developed some very early rnaturin~ varieties. 47

4) As the comp::>sition of people's diet improves, the consumption 2: millet changes in a different way in the city and countryside. The supply of millet nearly meets the demand in cities and toWP.s in regions where millets are grown. In Beijing, millet is on sale again after being out of stock for many years. But in the :ountrfside people want rrore rice and wheat for food. Studies of food processin; have developed a series of products made of foxtail millet, such as pies, ice cream, millet wine, and oii extracted from millet husk. In the Chinese opinion, fo>..-tail millet is not a coarse food grain. Also, consumption of foxtail millet will probably be increased ty publicizing its good nutritive value.

5) The development of sound cultural practices prO!TQtes production. Where effective technical aspects are developed, demor.Etrations established, and the income cf grower increased, the area planted to millet will be maintained and increased. such is the case in Hebei and Shandong provinces. In derronstrations of improved practices, chemical fertilizers, agricultural chemicals and rrore labor are required. An improved variety should have good yield potential, resistance to diseases and pests, and good quality in the grain and stalk. At present, the improvement of varieties is mainly carried out using local land races, and characters of new varieties are net very desirable. All these problems are being investigated in our research work.

The Priorities of Research Programs on Foxtail Millet

Foxtail millet is an important cereal crop in China. In 1981-1985, the priorities of research work of foxtail millet were : 1) breeding of new varieties, 2) evaluations of genetic resources from the Northern Plain of China. In 1986-1990, the first priorities are : 1) breeding cf new varieties and utilization of heterosis, 2) evaluation of genetic resources collected from the whole country. The second priorities are to carry out regional tests and develop improved processing of the product. All these programs are supported by the government, and the budget has been increased several times so as to ensure the accomplishment of these tasJr..s. Studies of cultural practices are carried out at provincial an:5. regional institutes, such as in Shandong and Henan provinces and Chifeng Region of Inner M:mgolia. The programs o: t.11e top priori ties are stateC. as follows :

1) Breeding Qf. ~varieties and utilization Qf. heterosis A. Breeding for filsh yield; illab. quality, multiPle resistance : The present breeding goal is to develop varieties t~t p:::-::ses:: high yield potential, good eating quality and high prote:~ content, resistance to lodging and diseases. Specific var:ete:: are bred fer the food product industrf for nutrient foc

B. Studies on the utilization of heterosis : Breeding for good quality and high yielding hybrids through the two-line method is going to increase with emphasis on developing cytoplasmic male sterile lines for the three-line method. This is a very hard task.

C. Studies on theoretical Problems in breeding Subjects including the inheritence of nutritive factors, the types, formation and phase of pollen sterility will be studied. 2) Documentation, conservation ,eng, evaluation of genetic re.sources In 1986-1990,the main task is to conserve all the accessions collected in our country (about 15,000 - 18,000) in a new gene barJ: for long-term stcrage. All the accessions are to be catalo;ued and characterized. The primary evaluations are for protein and fat contents, resistance to the main diseases and pests (green ears, blast and rust, stem maggot etc.) and drought tolerance. M:>re than twenty institutes are involved in this program. Data obtained in these evaluations are plarmed to input to the computerized millet ge~lasm data bank. 3) Regional test for imJID)ved varieties Three major locations for regional test were established. In 1986- 1990, abo~t seventy varieties or lines v.ould be under examination to detect whether they are adaptable for production. The newly release:. varieties are e:h."t>ected to cover 2 million hectares of the seeding area. In order to pronote this plan, la..rge area of demonstration and seed multiplication will be set up. A "technical package of seed" will prm·e to be useful for e:h.'tension, this is something like "The Minikit Demonstration for propagation of new tecrnology" in India.

Production and Research with Proso Millet The growing area of proso millet declined in the late fifties. After 1978, it recovered a little. At present, 1.33 million hectares are grown armually. The glutinous endospenn type is used to rrake cake, pie and other sweet foods, and the price is favorable. The non-glutinous endosperm type is used to make fried grain, which is a traditional M:>ngolian food. The market demand for both the glutinous and non­ glutinous types are increasing. The yield of proso millet is generally 0.7 - 1.5 tonne per hectare, while under intensive cultivation the tonnage may increase to 4.5. In irrigated land, the potential yield can reach 7.5 tonnes per hectare. Lodging often occurs under highly ferti2.e irrigated conditions, therefore, lodging resistant varieties are available.

The prOC.uction of proso is ~rtant in Northwest and Northeast China, but it is a minor crop in the nation-wide scale. Top priority w-~ not given tc t::rosc millet befcre 1980. In 1982 a conference c::i miner 49 crops was convened by The Chinese Association of Crop Sciences in Shenyang to set up a cooperation group to coordinate the regional tests and genetic resources research. Twelve provincial and regionai institutes joined in the activities, and research programs on genetic resources are sponsored by the Institute of Crop Germplasrn Resources, Chinese Academy of Agricultural Sciences. Since 1986, we have been giving the first priority to prose genetic resources programs, including cataloging and conservation of all accessions collected. The evaluations of smut resistance, salt endurance, contents of protein and fat are the ID3.in items. Although the first priorities of plant breeding do not include prose, some improved lines have already been released. They are Longsu Nc.16, Neirni No.4, Yi-7619 and Gu-771-2. The cultural practices are alsc studied in various places such as in Guyuan of Ningxia and Yikzho of Inner 1-bngcli.:i. Other Small Millets After the count..":i"Wide large scale collection of landraces , the collection of crop resources is carried on chiefly through rational investigation. In 1986-1990, the priority areas for collection will be Hainan Island and the rrountainous region of Western Hubei. Both regions are much warmer and have rrore rroisture than the main growing region in North China. Landraces of finger millet and Japanese barnyard millet may be found useful for feeding. Finger millet is also cultivated in the saline-alkali dry region in the north. For example, rrore than tv.o thousand hectares are grown in cangzhou, Hebei Province, even m::::re could be grown if the grains had a good market. Outlook for International Cooperation and COllaborative Research

Exchange of genetic ~esources It is necessary to intensify international exchange of millets f cr our country. Since primarily native materials are used in breeding, the genetic basis for inheritance is limited. The progenies are not superior when crossed with each other, and in breeding for sterile lines, the maintainer line is deficient in foxtail millet. Up to now we have introduced m.:;re than three hundred accessions of foxtail millet but few of them performed well in breeding. We wish a primary evaluation of genetic resources be carried out first by the host country for disease evaluation as well as for other traits. This primary evaluation will certainly raise the efficiency of introduction. Our genetic resources may also make a contribution to other countries. Landraces from China are productive and good in quality.

Institute of Crop Germplasm Resources, CAAS, is responsible for the exchange of crcp germplasm. At this institute, a rrodeITLiZed gene bank for long-term stcrage is in function. It is qualified to participate in the netw.'.)rk of the world gerrnplasrn conservation. In fact, a number of crops are granted as international co!'lServation basic points. 50

In China, the program of plant breeding is going to cover rrore and rrore crops. The Institute of Fo>.."tail Yi.illet was established under Academr of Agricultural Sciences of Hebei Province last year. It is sponsored by the Ministry of Agriculture, Animal Husbandry and Fisheries and is responsible for coordination on breeding of f o>.."tail illet. Now the Section of Foxtail Millet (a specialized corrmittee) of Chinese Association of crop Sciences handles official business at the institute, so as to facilitate the organization of the activities on millet crops. A publication "Journal of Millets" is edited for internal exchange. Papers on foxtail millet are issued in our country, but few of them are known by foreigners. Language is a proble.'11. A selection of theses on foxtail millet is published, with an abstract in English for each paper.

Tw:> institutions in China re rec0111T1ended as agencies for international coordination. These are : 1) Institute of Crop Germplasm Resources, C"ninese Academy of Agricultural Sciences, Beijing and 2) Institute of Foxtail Millet, Hebei Provincial Academy of Agricultural Sciences. These institutions will organize the nation-wide activities on millet crops.

Discussion

1) The t"-0-line program for hybrid develoJ;tllent uses a line segregating for genetic male sterility as the ferrale, with a different fe!:"t.ile line as the male parent. The three line method requires develop.ner:.: of a cytoplasmic male sterile line, a maintainer, and a restorer ro produce the hybrid.

2) Work on evaluating fo>.."tail millet for salinity tolerance has been discontinued as this evaluation was not producing the expected results.

3) Evaluation of 8,000 millet lines has now been completed at the Institute cf Crop Germplasm Resources.

4) The suggestion of requesting originating countries to "pre-screen_ their germplasm was felt to be a useful idea to follow-up. This "-OUld enable requesting countries to receive germplasm that "-Ould be rrore likely to be useful.

5) The rrost important diseases of foxtail millet in China are greer:. ears (dow::iey mildew), blast and rust. 51

PROPOSED ORGANIZATICN OF THE NE'OORK

Background

The Steering Conrnittee recognized that small millets are important sources of food and fodder for millions of people. Furtherrrore, various small millets possess unique traits, including tolerance to drought, salinity, acidity and other stress conditions, rapid maturity as well as specific nutritional and processing characteristics. These traits indicate that small millets can play an important role in increasing stability of production income, and food security in improved agricultural systems. Research and developrent of these crops has been neglected or given low priority in many national programs. Scientists v.orking on these crops are constrained by isolation 1 1.=.:k of genetic resources, non-availability of infonnation, lack of training and shortage of funds. Netw;)rk name It was recornnended that a netv.ork of scientists v.orking on small millets in national programs be fonned, with input from concerned international institutions and doncrs. This netw;)rk is to be na.'7led "International Small Y.illets Netv.ork ( ISMN) . 11 Netw;:;rk oart:icipants The ISMN encourages participation of national programs, institutions and scientists concerned with the improvement of small millets, frQ1l countries that grow these crops. suggested countries are Bhutan, Bangladesh, China, Ethiopia, India, Japan, Kenya, P..alawi, Mali, Mexico, Nepal, Nigeria, Sri Lanka, Uganda, USSR, Zambia, Zimbabwe. Collaboration with research programs in Britan USA and ICRISAT as well as with interested donors is also encouraged. Crops covered in the netv.ork It is recomnended that the netv.ork cover the following crops Finger raillet Eleusine coracana Foxtail millet Setaria italica Proso millet Panicum mileaciwn Kodo millet Paspalurn scrobici...uatur:i Little millet Panicum stnnatrense Br-~rnyard millet Echinochloa ssp Tef f Eragrostis tef Fonio Digitaria exilis 52

Goal of the netv.ork The goal of ISMN is to strengthen national programs and increase awareness of the potential of small millets, so that these crops can contribute to the development of stable agriculture and the well being of rural people.

Objectives Qf. the ISMN 1) To bring together scientists and others v.orking on small millets and promote a cOITITOn platfonn for scientific exchange and interaction.

2) To encourage a broader awareness of the importance of small millets and their potential in improved agricultural systems.

3) To promote the effective collection, conservation, exchange a.11d utilization of genetic resources of small millets.

4) To identify cOITITOn research priorities and develop collaborative activities that will strengthen national small millets research programs.

5) To encourage mechanisms that effectively test and develop improved small millet technology with farmers so that this technolo;y is utilized.

6) To assist in the generation and dissemination of infer.nation on small millets.

7) To identify and help organize appropriate training. Financial support The possibility of funding for a small millets institute as recarrnended by the First Small Piillets Workshop had been followed up. There appears to be no possibility for the large arrount of funding needed for physically building or staffing an International Small Millets. Institute.

Several donors, however, have expressed interest in the propose1 International Small Millets Netw::>rk. It was decided that specific priorities/activities should be identified as part of the ISHN program. The activities requiring funding w::>uld be developed as proposals for consideration by doncrs. It w::>uld be eA'P9Cted that existing national/international programs w::>Uld be receipients of funds, and v.ould carry out acti•Ii ties on behalf of the network. 53

Steering conmittee There is need to follow up to ensure the Steering Comnittee recomnendations are considered and approved by the ISMN members and to implement the approved recornnendations. It was recornnended that :

The present Steering Comnittee continue until the neh"'t International Small Millets Workshop.

Due to the non-availability of Mr. Fighur M...tza from Zimbabwe, his position on the Steering Comnittee is to be taken by Mr. Sann.le! O:ielle from Uganda.

An additional member be included from the ICRISAT Genetic Resources Unit.

Dr. Seetharam, Dr. Sayfu and Dr. Riley take the positions as Steering Comnittee Chairman, Co-chairman and Secretary respectively.

PROPOSED NE'l'\-ORK ACTIVITIES

A large number of specific activities for strengthening small millets research and developnent were discussed under four main areas - Genetic Resources, Collaborative Programs, Infonnation Exchange and Scientific Interaction. A sumnary of the discussions on these topics is presented in the following table. The proposed activities, as they relate to the net'W:)rk objectives, are then discussed in greater detail, follo\.led by a list of activities that will require additional funding. 54

Table: Priority Areas For Activities In International Small lllllets Network

Topic Level Of Crops Country Suggested Action Importance

I Genetic Resources

a) Collection *** All Small East Afnca 1. Action plan to be inltiated'for millets Bhutan Collection by respective countries India 2. IBPGR to be requested to assist in collections

b) Hamtenance *** All Small All Small 1. ICRISAT to contmue to take global responsibility millets millet growmg countries 2. IBPGR to assist individual countries in mamtenance

3. Gene Banks having long term storage facility to be identified at a later stage at the time of Second International Small Millets Workshop.

c) Evaluation *** All Small All National 1. All local and exotic germplasm maintained millets Programs to be evaluated on priority. maitaining small millets 2. Standard descriptors or subsets published by IBPGR to be used in evaluation

di Documentation All Small All countries 1. 1BK compatible software recommended to millets allow computer networking.

e) Exchange of Germp lasm **** All Small All network 1. Encourage free exchange of germplasm on millets countries bilateral basis.

2. Germplasm available at ICRISAT can be requested by Network countries.

3. Gene Banks 10 National Programs should identify useful gene resources for each crop.

Level of Importance ****Very High *** High u Kedlllll * Low 55

Table: Priority Areas For Activities In International Small Millets Hetwork

i·opic Level Of Crops Country Suggested Action Importance

~. National Programs to make germplasm available to ICRim for exchange.

f) Exhange of advanced ** All Small All network 1. Wherever possible such exchange to breedrng materials and mlllets countries encouraged between countries. segregating populations 2. Countries 111sh1nq to make crosses may request countries having strong national programs like India to make crosses for others; ISNM may act as nodal agency in this matter.

II Collaborative Programmes

aJ Screenrng for drought *** All Small All Countries 1. Screening of drought and other forms of and other forms of stress, mlllets stress to be given high priority. ISHH to particularly salinity c1rcuiate information on screening techniques initially for drought screening.

bJ Disease

11 Blast xtt Finger millet All Finger 1. Resistance sources to be identified. millet growing areas 2. International Blast screening nursery to be organized by pooling resistant sources fann various countries. ISMH to request India to organize such blast screenrng nursery.

111 Green ear Hx Foxtail millet China 1. National programme to strengthen resistance breeding.

m I Rust Foxtail millet China 1. National programme to develop resistance

Level of Importance tttt Very High ttt High u Medium * Low 56

Table: Priority Areas For Actmties In International Small Klllets Het11m

Topic Level Of Crops Country Suggested Action Importance

iv) Smut Foxtail millet India 1. National programme to strengthen Koda millet resistance breeding. Barnyard millet Tef f Ethiopia

v) Viruses Finger millet India, Uganda, 1. Rational programme to strengthen East Africa resistance breeding.

vi) Others - Smudge on Tef f Ethiopia 1. National programme to strengthen Te ff resistance breeding.

c) Pests

l l Stem Borer tt Finger millet Africa 1. Resistanc sources to be identified

2. A separate screening trial to be organized for African region. ISHH to request Uganda to organize such screening nursery.

ii) Shoot fly tt Tef f Ethiopia 1. National programme to develop resistant Prose millet I India varieties. Little 11illetl

ml Aphids Finger millet India 1. national programme to develop suitable control.

d) Breeding methodologies *** All small All countries 1. As there is paucity of information in and handllnq of segregating millets this area, breeders are advised to initiate materials separate studies in different small millets. ISMJf may consider funding national programmes.

Level of Importance tttt Very High ttt High tt HediWI * Low 57

Table: Pnomy Areas For Activities In International Small Millets Network

Topic Level Of Crops Country Suqqested Action Importance

e} Food uses

i) Traditional and novel *** All small All countries 1. ISKN to compile a publication on food food preparations millets preparations as made in different countries.

2. Donor support to be souqht for this.

ii} Nutri t1onal tt All small All countries 1. National proqraues are to be encouraqed evaluation of millet millets 10 this reqard. varieties

f) Production practices w All small All countries 1. Location and req1on specific production millets technoloqies to be developed by national proqrammes.

q} Transfer of Technoloqy tt All small All counmes 1. National proqramaes are required to millets identify extension methods for effective transfer of technoloqy.

hl Foraqe evaluation tt * All small All countries 1. llat1onal proqrmes to be encouraqed to mlllets take up this aspect.

III Information Transfer

l I lle11sletter *** All small All countries 1. Need for publication of a newsletter at millets periodic intervals was felt. IDRC Canada asked to consider providinq funds for this.

Level of Importance tttt Very Hiqh *** H1qh tt Ked1Ull * Low 58

Table: Priority Areas For Activities In International Small Millets Network ------... ------Topic Level Of Crops Country Suqqested Action Importance ------.... ------... -... ------... ------b) Directory of small ttt All small All countries 1. ISHN to collect upto date list of small millet research workers millets millet scientists form vanous countries.

cI Statistics on production ttt All small All countries 1. ISKH recommends better statistics millets collection on area and production of small millets 10 different countries.

di Published articles ttt All small All countries 1.-SATCRIS is requested to include small millets millets selected references to be sent to each country in the network. IV. Scientific Visits

a) Workshops ttt All small All countries 1. It is proposed to hold the second small 111llets millets workshop in 1989. The venue is Zimbabwe, or Uganda or China or Kenya, Donor support for this is to be explored.

b) Training ttt All small All countries 1. Heed for organizing short training millets courses fer scientists and technicians vas felt. Donor support for this area of activity to be explored.

c) Consultancies, tt All small All countries 1. To be considered at a later stage, when monltonng tours and study millets network established. visits

------.. ------... ----... ------

Leve 1 of Importance **** Very High *** High u Medium * Loli 59

Objective 1 Scientific Interaction

International Small Millets Workshop : It was agreed that a wor}:.shop be held at regular intervals to bring together small millets workers from different countries. Such a wor}:.shop should be a valuable instrument of scientific interaction, and would help sustain a viable network.

The next workshop should be held in 1989. Potential locations are Zimbabwe, Uganda, China and Kenya in that order. Participants from all suggested net'l-Klrk cou.'1tries are to be invited to attend. The sessions should include 1) Presentation of short country reports, 2) In-depth keynote presentation and discussicn en 2-3 relevant themes and 3) a business session to review the organization of the Network and the progress of network activities.

Scientific visits : MJnitoring tours; study visits and consultancies of scientists to each other programs are useful forms of scientific exchange, and could be considered where there is a specific need.

Objective 2 : Broader Awareness of Small Millets

Although farmers and consumers in millet growing areas are familiar with these crops, small millets are not widely known or appreciated. the IS.MN should seek to make policy makers, extension personnel, donors and these international conmunity rrore familiar about small millets by developing information which can be targeted at these audiences. Such information could include accurate production and research information compiled in a popular form.

Objective 3 Genetic Resources of Small Millets Exchange The Small Millets Steering Carrnittee places the very highest priority on the effective exchange of small millets genetic resources among network participants. In order to facilitate exchange, a compilation of the holdings in each national program is being made (Appendix I). This list will be expanded as rrore information becomes available. Bilateral exchange Exchange is encouraged first of all on a bilateral basis between national programs, following the correct procedures for quarantine and national exchange policies. M.lltilateral exchange : The capability of the Genetic Resources Unit (GRU) at ICRISJ:..T for collection, conservation and exchan9e of 6 small millet species was greatly appreciated by the Steering Corrrr~ttee, It W-c.S recorrrnended that ICRISAT pa..--ticipate in the Network, with a ITPJnber from ICRISF.T on the steering comni ttee. The Steering Corrrni ttee reconmended that ICRIS.ll.T include teff and fcr~ic in ~11e Unit's activities, so that all 60 crops in the ISMN v.Duld be the same as those in the ICRISAT GRU.

Net¥.Ork members, who wish to receive germplasm from ICRISAT can request a computer printout of descriptions for a set or sub-set of the ICRISAT holdings. ICRISAT can then despatch the germplasm accessions as requested. It was recOl11!lended by the Steering Cornnittee that funds continue to be made available for this service. Networ}: participants, in turn, are requested to provide local genetic resources to the ICRISAT GRU, with passport description information if available.

National programs and gene banks in member countries are also encouraged to identify genetic stocks with specific traits such sources of pest resistance and male sterility for exchange. Exchange of advanced materials : Exchange of advanced breeding lines and finished varieties was strongly encouraged by the Steering Cornnittee. This exchange if possible is to be carried out on a bilateral exchange basis between member countries of the Network.

The Steering Cornnittee recomnended that countries having strong small millet breeding programs such as India, could make specific crosses on behalf of ether national programs, retu..rning the segregating material for selection to those national programs.

A cooperative adaptation trial was considered. However, it was felt that this activity could be delayed until the network was m:::>re strongly established. Gepnplasm collection The Steering Cornnittee recognized the i.Jti:>ortance of continued collection of small millets, where germplasm may be in danger of being lost. The Steering Comnittee encouraged these acti·vities to be carried out by national programs, with support from ICRISAT and the International Board of Plant Genetic Resources (IBPGR).

The Steering CO.'lllri.ttee agreed that areas where small millets were not previously collected in East Africa, Bhutan and India be given priority in future collecting expeditions. It was mentioned that finger millet should also be recollected in Sri Lc.r"IJ:a. GermPlasm Maintenance The Steering COm'Tlittee recorrmended that ICRISAT continue to rraint:s.in small millets genetic stocks. Improved facilities such as gene bariY.s fer maintaining small millets are neede:i ir, r:r5.l1".l r.5.tio::-1al. pro;:?:"c.:1'..=. 61

Back-up storage fac::.lities fer genetic stocks are also needed. '!'his could be discussed at the next International Small Millets Worv.shop. Germplasm Evaluation National programs are encouraged to evaluate and document germplasm holdings of small millets in their countries. The standard descriptions as published by IBPGR should be followed whenever possible. It is encouraged that evaluation for important useful traits, including agronomic characters and disease resistance be carried out by national programs and made available to Network participants. Germolasm Documentation

It was reconmended that national programs use IBM canpati~le software when considering a canputerized documentation system. The ICRISAT Data Management and Retrieval System (IMRS) program appeared to be a useful format which could allow national programs to readily retrieve and exchange documented information on germplasm accessions. Objective 4: Collaborative Programs Careful consideration was given to a large number of potential collaborative activities. The Steering Conmittee felt that national programs could derive greatest benefit if collaborative programs are developed in the following areas.

Resistance to Stresses Since many small millet species have particular value under stress conditions, particularly roisture and salinity stress, it is important that these traits be maintained or enhanced in improved varieties. It was reconmended that high priority be given to develop suitable methods of screening for stresses, and information on appropriate methods be circulated to network countries. The experience gained by ICRISAT in evaluating useful drought resistance in pearl millet and other crops should be considered for the small rrdllet species. Funding for developing suitable techniques in national programs may be necessary.

Diseases ~ Pests

Blast .in Finger Millet: Is by far the rost important disease of a."1y small millet crop, for the greatest number of Network countries. There are good sources of blast resistance in m:my national programs in both Asia and Africa. It ·was reconrnended that the Indian Coordinated Srr.all Millets Project be requested to crganize a cooperative finger rrdllet blast screening nursery, ~it.i the objective of enabling national progr~Ts to identify and use additional sources of resistance rore effectively. All countries in which finger millet blast is a probla~ are requested ~o submit entries for this nursery. .?... uniform screening procedure should be adopted. 62

Other Diseases: There are many diseases which ITl<3Y be important in specific countries, but are not be widespread enough to merit collaborative activities. National programs were encouraged to develop suitable control. Diseases include green ear (downey mildew) and rust in fox'tail millets in China, smut on several small millets in India, and on teff in Ethiopia, and viruses in finger millet in India and East Africa, and smudge on teff in Ethiopia. Stem Borer in Finger Millet: is the most serious pest of finger millet in Africa. National Cotaitries are encouraged to identify resistance sources. Uganda is requested to organize a collaborative screening nursery, with contributions of resistant sources from countries where this pest is a problem. A uniform screening tech.~ique should also be developed. Other Pests: Shoot fly and aphids were noted as occasionally serious pests on specific small millets crops and in specific countries. National programs were encouraged to develop suitable control measures. Production Practices The corrmittee recognized that high priority should be given in national programs to developnent of improved production practices for small millets in rrore intensive and stable farming systems. As this ~rk is largely location specific, the corrrni.ttee did not recomnend ar!y collaborative program in this area, at this time. Breeding Meth::xJ.ologies Genetic improvement of small millets has been constrained by the difficulty of controlled hybridization in rrost of these species. National programs are encouraged to develop useable hybridization and breeding methods, and to report these tecr.. ru.ques in the Netv.ork. National programs in India and Uganda have developed useable hybridization techniques for several small millets. There is need to identify sources of male sterility in all small millet species, so that population improvement methods of breeding can be utilized. If appropriate, and at a later date, an intensive breeding program could be considered by a national program to incorporate desirable traits into improved progeny, which could be tested by net~rk countries. It was suggested that a review of breeding methodologies, hybridization techniques and methods of handling segregating material could be one of the in-depth topics for the next ~rJr•.shop.

Traditicnal food preparation for small millets vary greatly with different small millet crops, and in different regions. Traditional processing of ten makes use of inherent properties cf the grain to develop an optir:u,1\ product for h1..llTSl consu'Tlp"c.icn. The Steering Corrrni ttee recorrrnended that national progra.r:rnes make an effort to collect the metl1:.:7.E UEec. :r. prep~ing traditional f ood.s fran sm=-ill millets. T:::.2 ir-.f c:::-.=.1::cr.: ~r=··.:.2.C. t:ier: t.e ~e!'lt tc :·:- Seeth?Xa7!1 er- ~'r s.s.:{f,.,;. to be co:n;::2.e'3. 63

by the Netw::>rk and put out as a booklet. Such inforrnation is important in developing acceptable food preparation methods when small millets are introduced into new areas. Existing nutritional information on each of the small millets should be ccmpiled in the same booklet. Fodder Evaluation Many small millets are valued for the value of the straw as a livestock feed in addition to the value of the grain. National programs are encouraged to evaluate the importance of small millet straw as fodder, and to take this use into account when developing new varieties. Objective 5: Utilization of Technology It was agreed that effective interaction with extensicn personnel and a knowledge of farmer needs are essential if the improved technology for small millets is to be effectively transferred to, and utilized by t.."le farmers. It was agreed that the different approaches used by the Training and Visit System of extension, and Fanning Systems Research methodology should be presented and discussed at the neAt w::irkshop, so that national programs might develop the IT()St effective systems for the adoption of new technology. Objective 6: Generation and Dissemination of Information In addition to information aimed at informing a wider audience ai:out the importance of small millets, the Steering Comrd.ttee gave high priority to making relevant information available to small millet scientists in the netw::>r}:. Several specific activities were recOITTT\9nded.

First Small Millets Wor¥'..shop Proceedings : This proceedings of the October 1986 w::>rkshop should be published as soon a~ possible. A publishing date of Mid 1988 is anticipated. Newsletter: A newsletter on small millets should be produced, initially on an annual basis. It is suggested that all network countries or others carrying out research on small millets be requested to contribute short (1-2 page) notes that could be of interest to other netw::irk members. These contributions could include brief descriptions of national program activities, informal findings from surveys or research projects. The contributions must be kept informal, and must be typed in the final fcrm for inclusion in the newsletter, no editing 'WOUld be done. Dr Seetharam was asked to receive contributions and compile the first newsletter. IDRC was requested to contribute financial supi;::.c!:'t. Directory of Small Millet Workers: It was suggested that each netw::>rk country compile a list of active small millet research w::>rkers. This list could be f orw-arded to Dr Seethararn to be included in the newslei:ter.

~':.:itist:..cs on Production of Small Millets: It is. reca.rrnended ':!".=.t infcm.:iticn on areas, yields c.nd production cf sm=.11 millet= bE' se:-:-: ::·: 64 participating member countries for incorporation into the newsletter. Pub-1ished Articles: The Semi Arid Tropics Crop Research Information Service (SATCRIS) at ICRISAT can provide selective disemination of information (SDI) profiles and searches of published information on small millets. The Sorghum and Millets Abstracts also provides current abstracts of published information. It is recomnended that the ISMN insure that these sources of published information are available to members.

Ob iective L.. Traininq

It was felt that the net"'°rk should be particularly concerned wit..'l providing short course training specifically focussed on small millets. such training could include a short course for mid level scientists covering all aspects of small millet improvement, and short course for technicians focussed on practical training for methods of hybridization and breeding in small millets. It was suggested that the Coordinated Small Millets Improvement Program, at Bangalore India be requested to organize such training.

PRIORITY NE'ntl)RK ACTIVITIFS REQUIRING EXTERNAL FUNDING

1) Holding of the Second International Small Millets Net'nDrk WorJr.shop in 1989. 2) Funds to ICRISAT GRU to increase activities on small millets. 3) -Developing screening techniques for drought or salinity may require funds to national prograrn.5. 4) Blast resistance nursery in finger millet may require funding to the Indian Coordinated Small Millets Project, and other national programs. 5) Development of the collaborative stem borer nursery may require funding to the Uganda national program.

6) Funds will be required for producing the booklet on small millets food preparation methods.

7) Funding for the annual newsletter will be required.

8) Funds for holding the training courses in India may be required.

Prop,:isal:: for fur.~ing these activities will be developed =.:te~ ::·£ Netw:)rk me:rbe:-s arid otters w:·,c :re::::ei ve t..l-:iis report, evalu=.";:e a:-1:1 ccn!:.:7· 65 the reccmnendations. The proposals are to be developed jointly by the national programs, Steering Comnittee, and donors. 66

Appendix I

Collections of Small Millets as Reported by Participating Programs.

Ro. of Accessions

Kame and address of Mature of Gene Bank/Institute Finqer Foxtail Kodo Little Barnyard Pro so Tef f Total collection lillet lillet millet aillet millet millet

1. B.A.R.I. Joydebpur Gmpur Banqladesh 510 217 727 Land races and exotics 2. Plant Genetics Unit Dept. of Aqnculture YJ11111al tar Lalltpur Nepal 350 68 58 481 Land races and exotics 3. Botanist , CARI Gannoruwa Peradeniya Sri Lanka 90 5 100 Land races and emtics t Director of Aqnculture Dept. of Aqnculture Peradeniya Sn Lanka 22 22 5. Geraplasa Unit AICRP on Slall lllllets Univ. of Aqric. Sci. Gm, Banqalore India mt 1925 969 712 882 573 9595 Land races and exotics 67 Appendix I

Collections of Small Millets as Reported by Participatinq Proqrams.

Ro. of Accessions

Hame and address of Nature of Gene Bank/Institute Finqer Foxtail Kodo Little Barnyard Pro so Teff Total collection millet millet millet millet millet millet

6. Genetic Resources Unit ICRISAT Pa tancheru, A. P. India 1993 1328 526 377 582 757 5563 World Collection

7. Central Gene Bank of Kenya P.O. Ho.14733 Nairobi, Kenya KA NA NA

8. UAF RO Sere re Sorghum and Millets Unit, PO So rot i Uganda 1057 100 100 100 100 100 1559 Land races and exotics 9. Laboratory of Sorqhum and Millets Institute of Crop Germplasm Resources Chinese Academy of Aqnl. Sciences 30, Bmhi ciao Road Beijinq, PRC 10 3226 16 3255 All Local land races 68

Appendix I

Collections ot Small Millets as Reported by Participating Programs.

No. of Accessions

Name and address of Nature of Gene Bank/Institute Frnqer Foxtail Ko do Little Barnyard Pro so Tef f Total collection millet millet 11llet millet millet millet

10. Crop Breeding Institute Dept of Research and Specialist Services Box 8100 Causeway Harare, Zimbabwe m m

11. Plant Genetic Research Centre, Box 30726 Addis Ababa Ethiopia 1056 2650 3706 Land races and

12. Director, Upland Crop Division II, Crop Experiment Station, RDA Suwon 170 Republic of Korea m 419 - Local and exotics

exotics APPENDIX - II 69

RECOMMENDATIOOS OF THE

FIRST INTERNATIOOAL ~RKSHOP 00 SMALL MILLETS Dates: Septenber 29 to October 03, 1986 Venue Bangalore, India

I. GENERAL RECOMMENDATIONS 1. The papers presented at the International Workshop alongwith the discussions and the reconmendations be published for wide circulation. The International Development Research Centre is requested to fund the publication of the proceedings of the International Workshop on Small Millets. 2. Among the small millets, finger, foxtail and proso millets appear tc have wider clientele and their importance and developrnent aroused considerable interest. The reasonable consistency in performance, the ecological range and the production potential of ko::lo and barnyard millets have been noted. The role of teff in the Ethiopian economy wit..'1-i its possible extensions elsewhere has been recognized. Little millet with limited production potential has its areas specific adaptation. The growing demand for food and a variety cf food products calls for intere~t and investment in their developrnent.

3. The need for an international Small Millets Research Institute W-C1.S widely felt. The CGIAR system may consider this proposal. 4. A Steering Cornnittee of 5 members has been elected and charged with the responsibilities of identifying a net~rk for exchanging germplasm, transferring information and look into other service facilities etc. Tne !'n'='Jnbers of the Steering Conmittee are: Dr. A. Seetharama, India; Mr. Chen Jiaju, China; Dr. Sayfu Keterna, Ethiopia; Mr. Fighur M..iza, Zirnbab-we; and Dr. K. Riley, IDRC.

The Governments of China 1 ru.iopia, India and Zimbab-we are requested to kindly allow the nonrinated members to participate in the Steerin3 Conmittee meetings as and when they are scheduled. The International Developrnent Research Centre is requested to organize the Steering Co.'Tmittee meetings. 70

5 . The needs for obtaining definite figures of area, production and productivity fonn different countries has been pointed out. This may help the planners and the policy makers for deciding priorities. This may also attract the attention of the international camiunity of nations. 6. It is proposed to organize international small millets adaptation trials. 7. The need for exchanging published literature on small millets through a nodal agency has been recognized. 8. The need for encouraging scientific vis.its and providing e::-..-pertise where necessary has been felt. 9. The need for providing training facilities to scientific and technical personnel has been expressed. II. GWETIC RESOURCES 1. The need for updating the list of available small millets germplasm accessions has been recognized. 2. The participating scientists felt the need for free exchange of small millets genetic resources. 3. F.ach of the participating countries is requested to identify the gaps in collecting small millets in their agroclimatic regions and/ or provinces and arrange for collecting missions. It is desirable to associate breeders in collecting e::-..-peditions. 4. Besides collecting fran farmer's fields and markets, it is desirable to collect land races of cultivated small millets species as well as wild relatives of cultivated species. 5. Recognizing the significance of environmental diversity, it is proposed to evaluate the geIT!i)lasm in mul.tilocational National/ International nurseries. 6. To obtain unifonnity in evaluation, the participating countries are requested to use descriptors for various small millets as published by IBPGR, Rome. 7. Biochemical and processing characterization of small millets germplasm appears "WOrthwhile. Identifying sources of useful genes f rcr., the already available and conserved germplas1n should receive top priority since utilization of germplasm is very limited in small millet crops.

8. 'As the descriptor lists are likely to be voluminous and may have limited circulation, the utilization of the germplasm could be accelerated by circulating pocket size editions of descriptor lists with limited characters like plant height, maturity, pest and disease resistance, grain yield etc. 71

III . BREEDING AND VARIETAL IMPROVEMENT

1. The small floret size in all millets has limited the artificial hybridization and recombination breeding. The contact method and hot water emascula'Lion method used to some extent have their own lii:ritations. So there is need to overcome this problem by possibly studying induced male sterility using gametocides and their is need to standardize these methods. Genetic male sterile systeli\S, and mechanisms like protogyny may also be investigated and confirmed. We may also for out crossing systexr.s wherever available.

2. All small millets are inbreeders . There is not nruch ~rk done on the application cf various breeding procedures and assessing their relative efficiency.

3 . M.ltation breeding could be cne of the methods thought of in small millets since artificial hybridization is difficult. T'ne possible application of this breeding procedure in different areas of varietal improvement needs to be fully exploited.

4. Single plant selections fran gennplasm accessions could be one of the simplest and effective means of obtaining superior genotypes. 5. The application of biotechnology particularly anther culture and ovule culture techniques and exploitation of scmaclonal variability in callus cult;.;res could be thought of as a method for evolving varieties rapidly.

6. Quality treeding in small millets is also important although it is a very difficult area to make any headway in short period if time. Seed protein, mineral content and malting quality are some of the areas for consideration. Screening of available gennplasrn for consumer, nutritional and processing quality characters may be the first step in tJris direction.

7. There is need to understand rrore and rrore about the genetic control of various yield and yield contributing characters. 8. Identification of varieties with wide adaptation, high yield stability, and differing photo period sensitivity ie irrq;lortant particularly in finger millet, foxtail millet and proso millet as these crops are grown in varying rainfall areas and temperature regimes and day lengths. Besides evolution of varieties for different cropping systems, both relay and mixed cropping should also be attempted.

9. As all small millets are essentially rainfed crops confining the.~selves to semi-arid trcpics, breeding of drought tolerant varieties is important. The initial step in this direction will be screening of all available gennplasrn and identifying the useful lines. 10. All small millets are low input crops and grown by :poor and marginal farmers. Under such situations use of pesticides for the control cf pests, diseases and we~.....s is neither feasible nor practicable. Inbui2.t resistance :.£ the best wc..y of tackiin; this situation an:. tr.is should recei\'e high pri·~riti'. So:ne c: ~l-ie imp:::rtant dis.ea~es ar1d pe~ts ~r~=.:. 72 could be considered for resistance breeding are blast in finger millet, shootfly in proso millet and little1 smut in kodo millet, foxtail millet and proso millet.

11. Following identification of new practices like seeds, fertilizers, weedicides, fungicides, pesticides, testing on farmer's fields for obtaining his acceptance, laying of large scale demonstrations for exhibiting the production as well as economic potential and of scientists involvement in lab-to-land progrwmies for mutual information transfer require attention.

12. 1he concern for supplying certified seed of high quality small millets including seed treatment with pesticides and fungicides has been expressed. 1he mechanisms of seed production and distribution were discussed. Seed village concept and distribution of seeds through exchanging "new" for "old" varieties to farmers can be e:\.-plored.

IV. PRODUCI'ION TFDINOI..OOY AND CROPPING SYS'l»tS

1. Intercropping: 1here are benefits from intercropping for small farmers, the use of legumes may help yields. It represented a practical way of growing a variety of crops needed for the household.

2. Line sowing: It could be an efficient improvement. In Africa, this may require mechanization using animal draught. 1here is a whole technology developed in India which needs to be transferred to Africa, including the village maintenance services provided by blacksmith and carpenter.

3. Sowing time: In Africa, time of planting is often of critical importance for yield: working up a fine seed-bed could take too much time, especially without animal draught power.

4. Weed control: In Nepal, transplanting is always used and it is useful in filling gaps in the stand. 1his deserves a lot more emphasis, there is evidence of useful yield benefits. Transplanting provides a way of seeding in time into a nursery, and of a weed control by thorough cleaning of the remaining (5/6 th) of the area being cultivated. Wild finger millet in Africa presents a daunting weed problem for other types of control. In Africa, Striga can become a serious problem, especially in pure stands of finger millet.

5. Organic Manure: 1he use farmyard manure was mentioned by several delegates, and this important traditional practice deserves study as the least labour intensive method for using F.Y.M. Results in India suggest benefits from incorporating legume residues in situ.

6. Transfer of Technolosnr: Where there are no recommended practices, these should be developed in close consultation with the farmers, giving special attention to what are the limiting factors in their existing systems. Where reconunended practices have been developed, surveys should be made to determine which factors are being adopted, and which are not being adopted. 1he various constraints preventing the adoption of the latter should be carefully evaluated and alleviated as far as possible. 73

V. PESTS AND DISFASES

A. Diseases

1. Blast is the most serious disease of finger millet. It appears at different stages of plant growth. Different biotypes of blast fungus also could be present as in other crops and this area of racial differentiation of blast is worth investigating.

2. Viral diseases on finger and foxtail millets have been reported from many countries. Insects are generally associated with viral transmissions. There is need to understand the biology of vectors and their relationship with viruses in the transmission of viral diseases.

3. Leaf spot diseases are important in several African countries on finger millet. Resistance breeding is the best and the most effective in preventing yield losses.

B. Pests

1. Birds and aphids have been identified as serious pests of finger millet in Africa. Stem borers occur on finger millet in Asia. Finger millet earhead pests deserve attention in Africa and Asia.

2. Shootfly has been identified as the most predominant on small millets in Asia and the USSR. Shootfly escape mechanisms by manipulating sowing time and other cultural practices, if identified for various regions of the production, would be very helpful.

3. The biology and dynamics of most important pests of small millets, the occurrence of alternate and I or collateral hosts and screening of the available germplasm in pest sick plots deserve investigations.

4. Besides inbuilt resistance and cultural control of pest populations, critical studies on the occurrence of natural enemies, their breeding and release in farmer's fields are required. The occurrence of pests and losses in mixed, inter and relay cropping systems is to be investigated.

C. Pest and disease control

1. Considering the economic value and status of small millets, it can be said that integrated disease/ pest control measures are to be adopted with emphasis on inbuilt resistance, cultural control, biological control and chemical control in that order.

1. Screening of germplasm for malting and popping characteristics and breeding varieties with improved malting and popping characteristics.

2. Develoµnei.~ of simple milling machinery and a ma.king it available in millet growing areas. 74 in millet growing areas. 3. Diversification of uses of small millets and developnent of heal:h or speciality foods from millets: diabetic foods, high fibre fcxxis, weaning foods, flakes, quid;: coo~:ing cereals etc.

4. While breeding varieties, attention should be paid to retain t.rie desirable qualities of millets such as good storage quality, h:gh mineral content etc.

5. .~alysis of varieties grown under different agroclirnatic conditic~.a.S for nutrient content and polyphenols. 6. Studies on fine structure of millet grains .

7. Setting up of a permanent millet quality laboratory, if it is n:·: possible to establish collaboration or association with F::.od Technological Research Institutes or other such institutes for analys:....r3" or testing the varieties and taking research v.iork on millet product formulations. 8. Improving status of millets by substituting rice and other cereals in prestigious foods.

9. Publication of booJr.lets on millet products and giving wide publicity for these.

10. consume!'" preference checJr.s for different millet · products :.:-i.d improving the~ .... quality and consumer acceptability.

VII. SMALL MU.Im M FORlGE

1. Millet straw especially f rorn small millets is highly relished by c:o.ttle. The cost of straw in several millets canpensates the cost :f inputs in cultivation. lt)reover, the peasant fanners depend on this by­ product to large extent in maintaining their milch and draught animals. In many cou.~tries the farmers are willing to sacrifice grain but ~ct fodder. This point has to be kept in mind in future breeding programs. 2. Efforts are essential to identify suitable small millet varieties for fodder purposes. Millets as green forage crops assume trore irrp:Jrtan:e especially in scarce rainfall years. Selected varieties should be able to produce large am::runt of green fodder in a very short time that they may be available for growth.

3. Types suitable for nrulti-cuttings have to be identified which can ::-e grov.':1 even u.~der irrigated conditions.

4. Suitable millet varieties hc.ve to be identified for establishing in w-aSte lands and eroded lands . Methods of establishing such crops als:; nee:. to be w:'.)r~:ed out .

5 . The drf straw of rreny millets may not be nutritious . Methods 75

fortify such low value fodder by use of urea 1 m:>lasses etc. 1 have to worked out and popularized. The other possibility of improvement of the quality of the straw would be to include a legume alongwith the millet crop itself.