ICRISATANNUAL REPORT

1976-1977

I n t e r n a t i o n a l Crops Research I n s t i t u t e for t h e S e m i - A r i d Tropics

1 — 1 1 - 2 5 6 B e g u m p e t

Hyderabad 5 0 0 0 1 6 (A .P . ) Ind ia

ICRISAT's ObjectivesTo serve as world center to improve the geneticpotential for grain yield and nutritional qaaiity ofsorghum, pearl millet, pigeonpea, chickpea, andgroundnut.

To develop farming systems which will help toincrease and stabilize agricultural productionthrough better use of natural and human resourcesin the seasonally dry semi-arid tropics.

To identify socio-economic and other constraintsto agricultural development in the semi-arid trop-ics and to evaluate alternative means of alleviatingthem through technological and institutionalchanges.

To assist national and regional research programsthrough cooperation and support and to contributefurther by sponsoring conferences, operating in-ternational training programs, and assisting ex-tension activities.

About This ReportThis is the fourth annual report published by theInternational Crops Research Institute for theSemi-Arid Tropics. Printed for worldwide distri-bution, the report covers ICRISAT's developmentand activities from 1 June 1976 to 31 May 1977.

Detailed reporting of the extensive activities ofICRISAT's many research support units is beyondthe scope of this volume, but a comprehensivecoverage of ICRISAT's core research programs isincluded. Detailed annual reports have been pre-pared for limited distribution by each unit.

ICRISAT receives support from the ConsultativeGroup on International Agricultural Research.The responsibility for all aspects of this publi-cation rests with ICRISAT. Mention of particularpesticides, fungicides, herbicides, and other chemi-cals does not necessarily imply endorsement of ordiscrimination against any product by the In-stitute. The correct citation for this report isInternational Crops Research Institute for theSemi-Arid Tropics, 1977. ICRISAT Annual Re-port 1976-1977, Hyderabad, India.

C O N T E N T SI C R I S A T Govern ing Boa rd vii

I C R I S A T Personnel i x

A c r o n y m s and Abbrev ia t ions xii

D i rec tor 's I n t r oduc t i on 1

Research H igh l igh ts 11

Sorghum 11Pearl Millet 12Pigeonpea 12Chickpea 13Groundnut 14Farming Systems 14Economics Program 16

T h e C e r e a l s 2 1

S o r g h u m 25

Germplasm 25Breeding 25Physiology 34Entomology 38Pathology 41Nutritional Quality 42Microbiology 43Looking Ahead 44

Pear l M i l l e t 49

Germplasm 49Breeding 49International Cooperation 58Physiology 60Entomology 64Pathology 64Nutritional Quality 68Microbiology 68Intercropping 69Looking Ahead 70

T h e P u l s e s 7 3

P i g e o n p e a 77

Germplasm Collection andEvaluation 77

Elite Lines from Germplasm 78Evaluation of Hybrid Populations 78Progress with Early Maturing

Selections 78Genetic Male Sterility 78Rhizobium Populations in Soil 80Nodulation and Nitrogen Fixation 81Screening for Disease and Insect

Reaction 81Intergeneric Hybridization 83Screening for Soil Factors 83Ratooning 84Pigeonpea as a Winter Crop 85Residual Effects of Pigeonpea on

Subsequent Crops of Pigeonpea 87Effects of Nitrogen Fertilizer on

Pigeonpea Growth and Yield 87Seed Size and Seedling Size 88Ovule and Seed Abortion 88Pod-Set 88Source-sink Relationships 88Diseases and Pests 89Nutritional Quality 90International Cooperation 92Looking Ahead 93

C h i c k p e a 97

Germplasm Resources 97Breeding 98Diseases 99Rhizobium Nodulating Chickpea 100Comparative Growth and

Development 102Insect Pest Damage 107Nutritional Quality 109International Cooperation 110Looking Ahead 111

O i l s e e d 113

G r o u n d n u t 117

Germplasm 117Breeding 118Pathology 121Microbiology 125Looking Ahead 126

F a r m i n g sys tems 129,133 Credit and RiskEffect of "Green Revolution" in

208

Research in S u b p r o g r a m s 133 Wheat on Pulse and Nutrient Pro-

Agroclimatology 134 duction in India 210

HydrologySoil Physics

141143

M a r k e t i n g Economics 211

Soil Fertility and Chemistry 146 Market Surpluses of Food Grains 211Farm Power and Equipment 149 World Sorghum and Millet Trends 211

Land and Water Management 153 Sorghum and Millet Trends in India 213

Cropping Systems 158 Interregional Trade and Welfare 216

Agronomy and Weed ScienceLooking Ahead

167177

L o o k i n g A h e a d 216

Watershed-based Resource C o o p e r a t i v e P r o g r a m s 219

U t i l i z a t i o n Research 178 West Africa 219

Watershed Operations and East Africa 222

Development 181 Southern Africa 222

Water-balance Studies 182 South America 222

Crop Production and Economic C I M M Y T 223

Data 189 Asia 223

Implications and Future Plans 194 Other organizations 224

C o o p e r a t i v e Research 196 Fe l l owsh ips a n d T r a i n i n g 224

E c o n o m i c s P r o g r a m 201W o r k s h o p s , Conferences,a n d Seminars 226

P r o d u c t i o n E c o n o m i c s 201

Resource Base and CroppingP l a n t Q u a r a n t i n e 229

Patterns 201 C o m p u t e r S e r v i c e s 235Weed-control Economics 202Group Action and Organization 206 L i b r a r y 237

I C R I S A T Governing Board

Dr. Ewert Aberg (Member)Professor and HeadDepartment of Plant HusbandryAgricultural College of SwedenS-75007 Uppsala 7 Sweden

Dr. C.F. Bentley (Chairman)Department of Soil ScienceUniversity of AlbertaEdmonton, Alberta T6G 2E3Canada

Dr. N. Bhagwandas (Member)- u n t i l March 1977

Chief SecretaryGovernment of Andhra PradeshSecretariatHyderabad-500022India

Mr. Francis Bour (Member)Director GeneralInstitut de Recherches AgronomiquesTropicales et des Cultures Vivrieres ( IRAT)110 rue de l'UniversiteParis 7eFrance

Dr. Ralph W. Cummings (Ex-officio Member)- u n t i l 4 Mar 1977

DirectorInternational Crops Research Institute

for the Semi-Arid Tropics1-11-256, BegumpetHyderabad-500016India

Dr. Iwao Kobori (Member)Department of GeographyFaculty of ScienceUniversity of TokyoHongo 7-3-1Bunkyo-ku, TokyoJapan

Mr. A. Krishnaswamy (Member)- a s of Apri l 1977

Chief SecretaryGovernment of Andhra PradeshSecretariatHyderabad-500022India

Dr. Klaus Lampe (Member)Deutsche Gesellschaft fur TechnischeZusammenarbeit GmbHPostfach 5180, Stuttgarter Str. 106236 Eschborn 1 Federal Republic of Germany

Mr. A.R. Melville (Member)Chief Natural Resources AdvisorMinistry of Overseas DevelopmentEland House, Stag PlaceLondon SW1E 5DHEngland

Dr. Melak H. Mengesha (Member)Special Assistant to the PresidentAddis Ababa UniversityP.O. Box 1176Addis AbabaEthiopia

Mr. N.K. Mukarji (Member)-as of Apr i l 1977Cabinet SecretaryGovernment of IndiaRashtrapathi BhavanNew Delhi-110004India

Mr. B.D. Pande (Member) - unti l March 1977Cabinet SecretaryGovernment of IndiaRashtrapathi BhavanNew Delhi-110004India

vii

Dr. Djibri l Sene (Member)Delegation Generate a la Recherche

Scientifique et Technique61, Boulevard Pinet-LapradeB.P. 3218DakarSenegal

Dr. M.S. Swaminathan (Vice-Chairman)Director General, ICAR, andSecretary to Government of IndiaDept. of Agricultural Research & EducationKrishi BhavanNew Delhi-110001India

Dr. L.D. Swindale (Ex-officio Member)- a s of 4 Mar 1977

DirectorInternational Crops Research Institute

for the Semi-Arid Tropics1-11-256, Begumpet, Hyderabad-500016India

Dr. D. Wynne Thome (Member)365 East Third NorthLogan, Utah 84321U.S.A.

Dr. D.L. Umali (Member)Assistant Director General, andRegional Representative for Asia

and the Far EastFood and Agricultural Organization

of the U N OMaliwan MansionPhra At i t RoadBangkok-2Thailand

Mr. Rubens Vaz da Costa (Member)Editora Abri l Ltda.Av. Otaviano Alves de Lima, 80001390 Sao Paulo (SP)Brazil

vii i

ICRISAT Personnel -May 31, 1977

Administration

R.W. Cummings, Ph.D., director(until 4 Mar 1977)

L.D. Swindale, Ph.D., director(as of 4 Mar 1977)

J.S. Kanwar, Ph.D., associate directorR.C. McGinnis, Ph.D., associate directorS.K. Sahgal, IAS, principal administratorV. Balasubramanian, executive assistant

to the directorS.K. Mukherjee, B.Sc. (Eng'g.),

personnel managerO.P. Shori, B.Sc., B.L., fiscal managerR. Vaidyanathan, purchase & stores managerR.G. Rao, records managerCol. P.W. Curtis (Grad. I M A , Dehradun),

security officerJ. Pearson, transport officer (until Feb 1977)R. Narsing Reddy, transport officer

(as of Mar 1977)S.B.C.M. Rao, B.Com., travel officerB. Diwakar, Ph.D., scientific liaison officer

(visitors service)N. Rajamani, liaison officer, New Delhi office

Crop Improvement

Cereals

H. Doggett, Ph.D., plant breeder & leader(until Dec 1976)

L.R. House, Ph.D., principal plant breeder(sorghum) (since Nov 1976)

P.K.. Lawrence, Ph.D., assistant plant breeder(sorghum)

D.J. Andrews, B.Sc., (Hons.), D.A.S., D.T.A.,principal plant breeder (millets)

B.W. Hare, Ph.D., assistant plant breeder(millets)

J.C. Davies, Ph.D., principal entomologist(leader as of Jan 1977)

J.D. Skinner, Ph.D., assistant entomologist(until Dec 1976)

R.J. Williams, Ph.D., principal pathologistA .H. Kassam, Ph.D., principal physiologist

(until Oct 1976)F.R. Bidinger, Ph.D., principal physiologist

(as of Oct 1976)Claude Charreau, Ph.D., project leader,

West Africa projectA. Lambert, Ph.D., principal plant breeder,

Senegal (as of Sep 1976)R.T. Gahukar, Ph.D., principal entomologist,

Senegal (as of Dec 1976)C M . Pattanayak, Ph.D., principal plant breeder

(sorghum), Upper VoltaW.A. Stoop, Ph.D., principal agronomist,

Upper Volta (as of Jan 1977)J.A. Frowd, Ph.D., principal plant pathologist.

Upper VoltaJ.P. van Staveren, M.Sc., assistant agronomist,

Upper Volta (as of Mar 1977)S.A. Clarke, M.S., field trials officer, Mal iB.B. Singh, Ph.D., principal plant breeder, Niger

(as of Apr 1977)R.P. Jain, Ph.D., principal plant breeder

(millets), Sudan (as of Apr 1977)S.O. Okiror, Ph.D., principal plant breeder,

NigeriaJ.V. Majmudar, Ph.D., senior plant breeder

(millets)K.V. Ramaiah, Ph.D., plant breeder (sorghum)Bholanath Varma, M.Sc, plant breeder

(sorghum)D.S. Murthy, Ph.D., plant breeder (sorghum)B.L. Agrawal, Ph.D., plant breeder (sorghum),

(as of Nov 1976)S.C. Gupta, Ph.D., plant breeder (millets)K. Anand Kumar, Ph.D., plant breeder (millets)K.E. Prasada Rao, M.Sc. (Ag.), botanist

(genetic resources)S. Appa Rao, Ph.D., botanist

(genetic resources), (as of Nov 1976)

ix

N. Seetharama, Ph.D., plant physiologistG. Alagarswamy, Ph.D., plant physiologistR.K. Mait i , Ph.D., D.Sc., plant physiologistK.V. Seshu Reddy, Ph.D., entomologistV.S. Bhatnagar, Ph.D., entomologistK.N. Rao, Ph.D., plant pathologistS.D. Singh, Ph.D., plant pathologistR.P. Thakur, Ph.D., plant pathologistR.V. Subba Rao, Ph.D., microbiologist

Pulses

J.M. Green, Ph.D., principal plant breeder(pigeonpea) and leader

A.K. Auckland, Ph.D., principal plant breeder(chickpea)

L.J.G. van der Maesen, Ph.D., principal botanistW. Reed, Ph.D., principal entomologist

(as of Feb 1977)Y.L. Nene, Ph.D., principal pathologistA.R. Sheldrake, Ph.D., principal physiologistP.J. Dart, Ph.D., principal microbiologistD. Sharma, Ph.D., senior plant breeder

(pigeonpea)K.B. Singh, Ph.D., senior plant breeder

(chickpea)K.B. Saxena, Ph.D., plant breeder (pigeonpea)L.J. Reddy, Ph.D., plant breeder (pigeonpea)B.V.S. Reddy, Ph.D., plant breeder (pigeonpea)K.C. Jain, M.Sc., plant breeder (chickpea)Onkar Singh, M.Sc. (Ag.) plant breeder

(chickpea)S.C. Sethi, Ph.D., plant breeder (chickpea)C.L.L. Gowda, Ph.D., plant breeder (chickpea)R.P.S. Pundir, M.Sc. (Ag.), botanist

(genetic resources)A .N . Murthi , Ph.D., botanist (genetic resources)A. Narayanan, Ph.D., plant physiologist

(until Nov 1976)N.P. Saxena, Ph.D., plant physiologistI.V. Subbarao, Ph.D., plant physiologist

(as of Mar 1977)S.S. Lateef, Ph.D., entomologistM.P. Haware, Ph.D., plant pathologistM.V. Reddy, Ph.D., plant pathologistJ. Kannaiyan, Ph.D., plant pathologistO.P. Rupela, Ph.D., microbiologistJ.V.D.K. Kumar Rao, Ph.D., microbiologist

Groundnuts

R.W. Gibbons, B.Sc. (Hons.), D.A.S., D.T.A.,principal plant breeder

W.C. Gregory, Ph.D., consultantD.V.R. Reddy, Ph.D., senior plant pathologistA . M . Ghanekar, Ph.D., plant pathologist

(as of Nov 1976)P. Subramanyam, Ph.D., plant pathologist

(as of Nov 1976)S.N. Nigam, Ph.D., plant breederV.R. Rao, Ph.D., botanist (genetic resources)P.T.C. Nambiar, Ph.D., microbiologist

Farming Systems

B.A. Krantz, Ph.D., principal agronomist andleader

J. Kampen, Ph.D., principal agriculturalengineer (soil and water management)

M.B. Russell, Ph.D., consultant (soil physics)S.M. Virmani, Ph.D., principal

agroclimatologist (as of Jan 1977)R.W. Willey, Ph.D., principal agronomist

(as of Aug 1976)L.P.A. Oyen, M.Sc., assistant agronomist

(as of Oct 1976)G.E. Thierstein, M.S., principal agricultural

engineer, small implements development(as of Nov 1976)

M.C. Kla i j , M.Sc., assistant agriculturalengineer, small implements development(as of Jan 1977)

F.P. Huibers, M.Sc., assistant agriculturalengineer, soil and water management(as of Mar 1977)

S.V.R. Shetty, Ph.D., agronomistM.R. Rao, Ph.D., agronomistPiara Singh, M.Sc, soil scientistSardar Singh, Ph.D., soil scientistT.J. Rego, Ph.D., soil scientistS.J. Reddy, Ph.D., agroclimatologistM.V.K. Siva Kumar, Ph.D., agroclimatologist

(as of Mar 1977)J. Harikrishna, M.Sc., agricultural engineerR.C. Sachan, M.Tech. (Agr. Eng'g), agricultural

engineerHarbans Lal, M.Tech., agricultural engineer

x

P. Pathak, M.Tech. (Agr. Eng'g), agriculturalengineer

P.N. Sharma, M.Tech., agricultural engineerS.K. Sharma, B.A., senior research technicianS.N. Kapoor, M.Sc., (Ag.), senior research

technician

Economics Program

J.G. Ryan, Ph.D., principal economistM. von Oppen, Ph.D., principal economistH.P. Binswanger, Ph.D., principal economist

(ADC)V.S. Doherty, Ph.D., principal social

anthropologistN.S. Jodha, Ph.D., senior economistV.T. Raju, Ph.D., economistS.L. Bapna, Ph.D., economistB.C. Barah, Ph.D., economist

Biochemistry andCommon Lab Services

R. Jambunathan, Ph.D., principal biochemistUmaid Singh, Ph.D., biochemist

F a r m Services

E.W. Nunn, B.S. (Agr. Eng'g), principalagricultural engineer

D.S. Bisht, B.Sc. (Ag. & AH) (Hons.), seniorengineer

B.K. Sharma, B.Sc. (Ag. Eng'g.), senior engineerD.N. Sharma, B.Sc. (Ag.) Eng'g & Tech., senior

engineerS.K.V.K. Chari, M.Sc, electronics engineerV. Lakshmanan, executive assistantP.M. Menon, B.A., L.LB., executive assistant

Library andDocumentat ion Services

T.C. Jain, M.Lib.Sci., librarian

Informat ion Services

J.W. Spaven, B.S., head, information services (asof Aug 1976)

G.D. Bengtson, B.S., research editor (as of Oct1976)

S.M. Sinha, N.Dip. Com.Art, senior artist andprintshop supervisor

H.S. Duggal, head photographerMira Shah, M.A. (French), French

translator/interpreter (as of Nov 1976)

Computer Services andStatistics

J.W. Estes, M.S., computer services officerJ.A. Warren, Ph.D., consultantT.B.R.N. Gupta, B.Sc. (Jr. Cert. Agrl. Stat.),

computer programmerB.K. Chakraborty, M. Statistics, computer

programmer

Plant Quarant ine

K.K. Nirula, Ph.D., plant quarantine officer

Fellowships and Train ing

D.L. Oswalt, Ph.D., principal training officerA.S. Murthy, Ph.D., senior training officer

Physical Plant Services

A.D. Leach, Ph.D., engineer (until Dec 1976)N.N. Shah, Ph.D., project manager

(as of Jul 1976)F.J. Bonhage, construction supervisorSudhir Rakhra, B.E., senior engineerB.H. Alurkar, B.E. (Civil), M.I.C.E. (Ind.),

senior engineerD. Subramanyam, B.E. (Elec), senior engineerS.K. Tul i , Dip. (Civil Eng'g.), senior engineer

(as of Sep 1976)T.J. Choksi, B.E., L.LB., senior engineer

(as of Jan 1977)

xi

Acronyms and selected abbreviationsused in this Annual Report:

AICMIP All-India Coordinated MilletsImprovement Project

AICORPO All-India Coordinated ResearchProject on Oilseeds

A P A U Andhra Pradesh AgriculturalUniversity

BARC Bhabha Atomic Research CentreCIBC Commonwealth Institute of

Biological ControlC I M M Y T International Maize and Wheat

Improvement CenterCNRA Centre National de Recherches

AgronomiquesCOPR Centre for Overseas Pest

ResearchCPPTI Central Plant Protection

Training InstituteDPAP Drought-Prone Area ProgrammeEAAFRO East African Agriculture and

Forestry Research OrganizationFAO Food and Agricultural

Organization of the UnitedNations

IBPGR International Board for PlantGenetic Resources

ICAR Indian Council of AgriculturalResearch

ICARDA International Centre forAgricultural Research in the DryAreas

IDRC International DevelopmentResearch Centre

I ITA International Institute ofTropical Agriculture

ILCA International Livestock Centrefor Africa

IPMAT International Pearl MilletAdaptation Trials

I P M D M N International Pearl MilletDowny Mildew Nursery

IPMDRTP International Pearl MilletDisease Resistance TestingProgram

IRAT Institute for Tropical CropsResearch

I S D M N International Sorghum DownyMildew Nursery

ISDRTP International Sorghum DiseaseResistance Testing Program

ISGMN International Sorghum GrainMold Nursery

ISLDN International Sorghum LeafDiseases Nursery

SMIC Sorghums and MilletsInformation Center

UNDP United Nations DevelopmentProgramme

UNEP United Nations EnvironmentProgramme

USAID United States Agency forInternational Development

Abbreviations:

BNV bud necrosis virusCRISP Crop Research Integrated

Statistical PackageCSV chlorotic spot virusDBC dye-binding capacityET evapotranspirationDEP dilution end pointDM downy mildewFSPS full-sib progeny selectionF Y M farm-yard manureGMS gridded mass selectionHYV high-yielding varietyHI harvest indexIP Indian pennisetumLDC less-developed countryLER land-equivalent ratioM A I moisture availability indexMKJ micro-KjeldahlPE potential evaporationPMHT pearl millet hybrid trialsPPMRN preliminary pearl millet rust

nurseryPMV peanut mottle virusPSV peanut stunt virusRRPS recurrent restricted phenotypic

selectionRUE rainfall-use efficiencySAT semi-arid tropicsSCU sulphur-coated ureaTAA Technicon auto analyzerTIP thermal inactivation pointTSWV tomato spotted wilt virusUDY UDY-instrument reading

representing percenttransmission

VB vein-banding diseaseW M I weighted mean index

xii

ICRISA T is dedicated to helping the small SA T farmer of limited means . . .

Director's Introduction

The year described in this annual report findsICRISAT in the busiest stages of its initial develop-

ment. The earliest programs in Farming Systems,Cereals, and Pulses have established their major

organizational lines, appointed most of their seniorstaff, and formulated and implemented sound research

programs. Even some early results are evident infarming systems technology and new millet cultivars.Village level studies in the Economics Program have

progressed far enough to give some clear indications ofthe types of new technology that will be usable by the

Orientation. Retiring Director Cummings (left) explains rationale of field layout at ICRISAT Center to incoming Director Swindale.

1

Overall direction of lCRISAT is vested in its Governing Board, consisting of 13 members representing 12 nations. The Board, under the chairmanship of Dr. C. F. Bentley of the University of Alberta in Canada, met in Hyderabad in May.

small farmer in the semi-arid tropics. The newestprogram, in groundnuts, has completed its first fullyear of operations, although only one quarter of itsprincipal staff has been appointed.

The development of the ICRISAT research farm haskept pace with the growing volume of research. Ad-ditional precision fields for crop breeding research andadditional watersheds for operational research infarming systems have been developed. Because Hyde-rabad is not adequately representative of all the semi-arid tropics, small substations in different agroclima-tological zones in India are being developed with thecooperation of the state agricultural universities. Thesubstations extend the range of climates and seasonsfor which cultivars can be developed, and increase thenumbers of diseases and pests against which we candevelop resistance.

2

A high point of the ICRISAT year was the November visit of Robert F. McNamara, President of the World Bank. Mr. McNamara and his party toured the major research plots at ICRISA T Center, and then participated in a seminar presented by program leaders.

The construction of ICRISAT Center got off to a slow start in 1975 because of the prolonged monsoon

period which greatly interfered with excavations. It hastaken a while to overcome the problems created by

that very wet season, but construction is now well inhand, although the buildings are not likely to be

completed for another 12 months or more.

At its 1977 meeting, the Governing Board ofICRISAT made an important addition to the man-

date. ICRISAT agreed to become a primary repositoryof the five crops included in its mandate and of three

minor millets - Proso millet (Panicum miliaceum),

3

finger millet (Eleusine coracana), and foxtail millet(Setaria italica). ICRISAT expects in the future tomake major contributions to the collection, descrip-tion, preservation, and distribution of these geneticresources. The intention is to complement the work ofnational genetic resources programs and to provide a service to them and to plant breeders throughout theworld. This new initiative will require new programs,staff, and facilities in the future.

After several years of testing, we are growing con-fident that farming systems technologies developed atICRISAT may enable much more effective use of thedeep Vertisols of India during the rainy season.

As ICRISA T's crop-improvement programs grow, thousands of seed samples are shipped to and received from all areas of the world. To prevent accidental introduction of dangerous plant pathogens or other pests and to reduce the amount of time required for quarantine clearance, close cooperation between ICRISAT scientists and quarantine officials of the Government of India is essential. Both parties are working very hard to achieve a rapid and safe flow of material through import and export channels.

4

Research has reached the stage where it can now betested on farmers' fields in the villages of rural India.

ICRISAT workers in pearl millet have successfullydeveloped some new cultivars which have been in-

cluded in the All-India Coordinated Millet Improve-ment Project trials. In addition to greater yields, the

cultivars have improved resistance to downy mildew,the most prevalent and devastating disease of the crop.

Pearl millet will probably be the first improved cropdeveloped at ICRISAT to be incorporated into com-

mercial use in India.

The International Sorghum Workshop brought together 52 sorghum specialists from 26 nations who spent a week discussing sorghum-improvement programs, visiting ICRISAT plots, selecting seed material for their own programs, and exchanging ideas. The country papers presented by the participants provided an up-to-date assessment of the sorghum situation in SAT countries, and is used in the planning of international experiments and programs.

5

Adama Moussa (left), S.M.Bonzi (center), and A.B.M.Salahuddin (right), former ICRISAT trainees, returned to Hyderabad to participate in the Sorghum Workshop in March. Moussa is doing sorghum breeding work in Niger; Bonzi is serving as entomologist in Upper Volta, and Salahuddin is doing breeding work in Bangladesh. They enjoyed a reunion with Dr. R. C. McGinnis, Associate Director for International Programs (second from left), and Training Officer Dr. Dallas Oswalt.

This year saw the Training Program of ICRISATreach its peak in numbers. Twenty-seven in-servicetrainees, most of them from Africa and about half ofthem from French-speaking Africa, are currentlyundergoing intensive field-oriented studies in cropproduction, crop improvement, and farming systems.Eleven research scholars studying for advanced de-grees, six research fellows, and three Dutch AssociateExperts are also learning-by-doing at ICRISAT. Thecapacity of our training facilities has been reached,perhaps over-reached, and no larger numbers of train-ees can be accommodated until ICRISAT Center hasbeen completed. More than one-hundred trainees intotal have already passed through ICRISAT's pro-grams and are now contributing to their nationalprograms, and in some cases to ICRISAT's programs,in their own countries. We wish them well in theirfuture careers and professions.

6

Work continued on the ICRISAT water tower, which will become the Center's most distinguishing landmark.

In 1974 and 1975 ICRISAT, with the financialassistance of UNDP and the USAID, placed 14 scien-

tists in the Sahelian Zone of West Africa to work atnational research stations on sorghum and millet

improvement. This year we were able to complementthat team by additional researchers in two Sahelian

countries and in the Sudan. Our scientists add strengthto on-going work in the countries where indigenous

scientists are few in relation to the needs.

7

As the result of the work done during its first 5 years, ICRISAT has reached the stage, where it will beincreasingly in a position to offer results which can beapplied in farmers' fields. However, the data fromwhich extension recommendations are made to farmersmust be carefully analyzed with reference to socio-economic and sociocultural factors which influence thefeasibility of adoption of the recommended practices indifferent areas. Recognizing this situation, the Govern-

Construction began on the ICRISAT auditorium in the general administrative complex. Occupation of the new quarters is expected to be possible near the end of 1978.

8

ing Board has established a standing Extension Ad-visory Committee which will judge the value of newtechnology and information produced by ICRISAT

research and indicate to our government and farmerclientele what new things are considered ready for use.

After 5 years of service, four members of theGoverning Board are retiring from its membership this

year. Mr. Francis Bour, Director General of IRAT,France, and Mr. Rubens Vaz da Costa, businessman of

Brazil, have completed their terms. They have assistedICRISAT in formulating and implementing its initial

policies and programs, and made a special contri-bution to the development of programs in French-

speaking Africa and in Latin America. Mr. B. D.Pande, Cabinet Secretary to the Government of India,

and Dr. N. Bhagwandas, Chief Secretary of AndhraPradesh, are retiring from these positions and con-

sequently from the ICRISAT Board. These gentlemenhave made exceptionally important contributions to

the establishment of ICRISAT in India and developeda commitment by government that will stand to our

advantage for manyyears to come. We wish them well

in their retirement.

Dr. Ralph W. Cummings, after 20 years in in-ternational agriculture and 5 years as the first Director

of ICRISAT, has retired from ICRISAT and itsGoverning Board. Words cannot adequately describeall that he has done for us. To put it simply, he wasjust the best man that could have been found to put

ICRISAT on the right road. His combined knowledgeof international agricultural research and of India wereunique. Dr. Cummings has assumed the chairmanship

of the Technical Advisory Committee to theConsultative Group on International Agricultural

Research. From this vantage point he will continue toassist and advise upon the future development of

ICRISAT. His vast contributions are gratefullyacknowledged.

ICRISAT again acknowledges the cooperation andassistance received from the Governments of India andAndhra Pradesh throughout the year. Wc acknowledge

also the growing and fruitful cooperation with scien-tific institutions around the world.

10

R e s e a r c h

H i g h l i g h t s

S o r g h u m

A major highlight in the sorghum-improvementprogram during the year was the success ob-tained in screening for Striga resistance in thelaboratory and subsequent confirmation of re-sistance in many lines in field screening. Thecollaborative work between the breeders and

physiologists on mechanisms of mechanical re-sistance is providing extremely useful data.

Preliminary results from the population-breeding program have identified several lineswith yields in excess of those obtained withreleased cultivars; results of this nature in such a short time is very encouraging.

The identification of a range of germplasmwith resistance to several diseases and pestsaugurs well for production of lines from thebreeders' programs incorporating better yieldstability. Some of these lines appear to haveresistance to Striga.

The very useful links forged with cooperatorsoverseas and the extremely encouraging start

Many agricultural officials of cooperating nations visited ICRISAT breeding plots during the year.

11

made by the ICRISAT programs in Africa arealready paying dividends in production of basedata, identification of stable lines, and exchangeof germplasm.

The collection of sorghum germplasm is in-creasing and a useful catalog is being built up.Considerable progress has been made on com-puterization of data.

Pearl M i l l e t

An effective technique was developed for creat-ing uniformly high disease pressure for downymildew in both the rainy and postrainy seasons.Breeders now screen up to 6000 entries fordowny mildew resistance per season on 5 hec-tares, and conduct breeding operations in thisdisease nursery.

Valuable results were obtained from 1976International Cooperative Nurseries. The In-ternational Pearl Millet Downy Mildew Nursery( I P M D M N ) identified 10 lines (mostly of Nige-rian origin) with across-location DM resistance.The best entry in the Second International PearlMillet Adaptation Trial ( IPMAT-2), anICRISAT hybrid ( ICH 105-5054 x B 282), aver-aged 27 percent more yield than the local vari-eties (1934 kg/ha) over 30 locations. The bestvariety WC-C75, also from ICRISAT, gave 10percent more than the local variety average.

The All-India Coordinated Millet Improve-ment Project (AICMIP) selected, from the resultsof country-wide trials, ICH 105 and WC-C75 foradvanced testing. Variety WC-C75 is a first-cycleproduct of the ICRISAT population-improve-ment program. The parents of hybrid I C H 105were released to the National Seeds Corporationfor increase.

Tests for progress in the population-improvement program indicated a slight butnonsignificant yield gain of about 5 percent dueto the first cycle of multilocation selection inseven composites with no indication of reducedgenetic variability. Experimental varieties pro-duced from these populations have equalled theyields of commercial hybrids, and the best pro-

genies have given up to 25 percent more yieldthan the parent population.

Seed requests from cooperators increased.During the year a total of 5 153 germplasm andbreeding lines (excluding cooperative trials) weresupplied to scientists. in 18 countries.

P i g e o n p e a

Collection of Atylosia cajanifolia, apparently theclosest wild relative of pigeonpea, provided ad-ditional genetic resources for breeding.

Differential response of cultivars in pure cropand in cereal/pigeonpea intercrops emphasizedthe importance of selecting under actual pro-duction conditions; we have planted breedingmateria] in intercrops with maize and withsorghum.

A laboratory-screening technique for Phy-tophthora stem blight was developed, and re-sistant cultivars have been identified.

A field-screening method for sterility mosaicdisease utilizing infector rows was developed.

Antibiosis in Atylosia scarabaeoides to Heli-othis armigera was found. This, plus anatomicaldifferences in pods of the two genera, are the first

Germplasm accessions at ICRISAT.

12

1973 1974 1975 1976 1977

ChickpeaPigeonpeaSorghumMilletGroundnut

161514131211109876543210

indications of possible specific characters thatmight confer resistance to Heliothis.

Postrainy season planting of medium-duration cultivars at high population levels gavegrain yields of 1 500 to 1 700 kg/ha.

Chickpea

Among the first advanced breeding lines fur-nished to cooperators, one early and five latelines yielded significantly more than the averageof the checks over all locations, and severaloutstanding lines were selected for increase andfurther testing by local breeders.

Differences among cultivars in germinability

under low levels of soil moisture were found inpreliminary experiments. This indicates an im-portant area of research aimed at improvedstands of a crop planted on receding moisture.

Among 450 lines grown in a multiple-diseasesick plot, 60 with good levels of resistance to soil-borne disease were identified.

Resistance to Ascochyta blight was found intwo wild species of Cicer, C. reticulatum, and C.anatolicum. The former has been successfullycrossed with the cultivated chickpea.

Seed-borne Fusarium oxysporum was elim-inated by treatment with 0.15 % Benlate-T, pro-viding a safe means for controlling this fungus inseed exchanged among scientists.

One of the many sorghum breeding nurseries at ICR1SA T.

13

The efficiency of small plots in screening trialsfor insect damage was established. This willpermit screening many more genotypes in a givenarea.

Groundnu t

In groundnut improvement, we have achieved a very rapid buildup of germplasm to about 6000cultivars in spite of strict quarantineconsiderations.

Cultivars with resistance to rust and to yellowmold and interspecific hybrids with resistance toleafspots have gone through quarantine restric-tions and are now being used in the breedingprogram.

Rapid advances have been made in the identi-fication and purification of groundnut viruses,using advanced laboratory techniques. Simpleand reliable germplasm screening methods havebeen evolved for the major viruses being studied.

A rapid and efficient screening technique,

using detached leaves for the identification ofrust resistance on a large scale under controlledconditions, is being perfected.

Differences in nodulating capacity were notedbetween germplasm lines. A diurnal periodicityin nitrogenase activity suggests a close link withthis activity and photosynthesis.

Farming Systems

Deep Vertisols are normally fallowed during therainy season in the semi-arid tropics (SAT).Present annual yields of cereal food grains onthese soils range from 200 to 600 kg/ha.ICRISAT scientists have developed soil-, water-,and crop-management systems which permitcropping of deep Vertisols during the rainy aswell as the postrainy season. These improvedfarming systems have been tested onoperational-scale watersheds for 5 years, usinganimal draft power and improved equipment.Annual food production has been consistently

A valuable facet of the International Sorghum Workshop was the opportunity for participants to inspect the material growing at ICRISA T Center and select lines that may be useful in their own programs at home. Seed of the selected plants was despatched to the participants after harvest and quarantine clearance.

14

increased to levels severalfold those of presentyields with traditional management. Cooper-ative research on improved resource manage-ment and utilization has been initiated withnational programs of the Government of India.

By timely tillage, deep Vertisols can be pre-pared with improved bullock-drawn implementsduring the dry season, permitting "dry planting"of crops such as sorghum, pigeonpea and maizejust prior to monsoon rains. Because of the lowwater-retention capacity of Alfisols, dry plantingis hazardous and not practiced. By "dry plant-ing" on Vertisols, the first rainfall - which mois-tens the soil to seed level - is utilized for cropgrowth and no water is wasted in the landpreparation process. "Dry planting" facilitatesearly growth to provide a plant canopy for lightand raindrop interception and facilitates plantestablishment ahead of the insect buildup whichstarts at the onset of the rainy season.

Contour bunding is the most common tech-

nique used to conserve soil and water on rainfedlands. Relatively little research has been con-ducted in India on the effect of contour bundingupon crop yields. Two recent reports on deepVertisols actually indicate reduced crop yieldsdue to stagnant water and reduced fertility abovethe contour bunds. Likewise, results fromICRISAT's watershed-based investigationsshow no positive yield effects from contourbunds even in relatively low rainfall years, as in1976 and 1977. There were, however, negativeeffects in many cases due to water stagnationabove the bund, particularly in high rainfallyears such as 1975. Thus, it appears that nopositive moisture-conservation effect upon cropyields can be expected from contour bundingunder most conditions in the SAT.

The watershed-based resource utilization sys-tem using a graded (150 cm) bed-and-furrowsystem at 0.4 to 0.6 percent slopes with grassedwaterways and runoff collection in small tanks

December planting ofpigeonpea results in greatly reduced growth as a result of short-day induction of early flowering. Entire plants are covered with muslin bags or nylon stockings to prevent cross pollination by wild bees.

15

shows great potential for reduced soil erosion,more uniform distribution of rain water, im-proved crop drainage, possibilities for supple-mental irrigation particularly on Alfisols, re-duced risk, and greatly increased crop yields.

On deep Vertisols the yield of postrainy seasoncrops was little affected by growing rainy seasonmaize instead of the traditional practice offallowing. Cropping in both seasons gave sub-stantially better returns than a single postrainyseason crop. Sorghum and pigeonpea sown assecond crops responded to early sowing; there-fore these crops benefitted from "relay" planting24 days before maize harvest. Chickpea andsafflower responded to late sowing, so theseperformed better as sequential crops after maize.

An experiment conducted on an Alfisol withsorghum compared "tradit ional" with " im-proved" levels of three "steps" in technology -variety, fertilization, and soil and cropmanagement—singly and combined. Yield in-creases from the three steps combined wasdouble that of the sum of the increases due to thesame three factors applied singly, thus illustrat-ing the large synergistic effect of the three stepswhen applied together in a system. Similar resultswere obtained with sorghum in 1975 and withmaize/pigeonpea intercrop on a Vertisol in 1976.

ICRISAT scientists believe that the principlesinvolved in improved farming systems, alongwith the introduction of high-yielding varieties,wil l have a great impact on the quantity andstability of food production in the SAT of thedeveloping world.

Economics P rog ram

Village Level Studies have revealed that inregions with less assured rainfall and a moreheterogeneous resource base, the proportion ofintercrops grown is higher than in the moreassured and homogeneous areas. Small farmersin such areas grow a higher proportion of mixedcrops than do large fanners. Successful in-tercropping research may therefore benefit lesswell-endowed regions and farms. The fact that

virtually all high-yielding varieties were found tobe grown as sole crops suggests, though, that theimportance of intercrops declines when HYVsare introduced. Improvements in land and watermanagement which enhance the resource baseand make it more homogeneous may encouragemore sole cropping.

Small farmers in the deep Vertisol villagestudied were found (when compared with largefarmers) to leave a higher proportion of theirland fallow during the rainy season. To theextent that this is true generally in the 18 millionhectares of such fallow land in SAT India,technology which enables more cropping in therainy (as well as the postrainy) season may havesubstantial benefits for small farmers.

Farmers in the village level studies practicedweed control more intensely and timely on thehigher-valued crops and where weeds were morevigorously growing, such as in rainy seasoncrops. With HYVs, farmers would be expected toimprove their weed-control efforts. Herbicideresearch in SAT India does not seem to be a high-payoff proposition; with abundant labor andanimal power and low wage rates weed controlwith herbicides is at least SO percent moreexpensive than with animal and human methods.Herbicides can also have adverse effects onincome-earning opportunities for one of themost disadvantaged groups in SAT India -female labor (particularly hired female labor).Hence, both on grounds of productivity andequity, herbicide research seems a low priorityfor SAT India unless it can provide substantialyield gains not possible with human and animalmethods. The situation in SAT Africa may bequite different.

During drought periods farmers were found toreduce consumption and the value of productivefarm assets. Formal credit agencies providedonly 13 percent of sustenance needs; the bulk ofborrowing was from private lending sources.Government relief works were the major sourceof sustenance in droughts. It was concluded thatmore-flexible credit policies in high-risk SATareas would be of great value in terms ofadjusting to meet the requirements of fannersduring droughts-particularly in maintaining

16

the value of productive assets. Credit linked withthe development and introduction of risk-diffusing technology would be beneficial.

The net nutritional impact of the new HYVs ofwheat introduced in India in the mid-1960's was

positive and substantial. This success clearlyillustrates how a plant-breeding strategy whichemphasizes increased yield potential can lead tosignificant improvements in aggregate nutri-tional well-being.

17

18

19

20

P e a r l M i l l e t ( Pennisetum americanum L . )

Sorghum and millet are the main cereal crops grown under the erratic climatic conditions of the SAT.The two crops occupy in excess of 70 million hectares, and form the staple cereals for many millionsof people.

The quantum leaps seen in production of rice and wheat, brought about by high-yielding varietiesand research-based technology, have in no way been matched by sorghum and pearl millet. Farmpractices remain essentially as they have been for thousands of years. A worldwide effort in breedingsuperior varieties is just now getting under way, and much remains to be learned about the role ofcultural practices in increasing yields of these cereals in the SAT. The two cereals can play majorroles in feeding the underfed of developing nations if the total benefits of the agricultural sciences canbe brought to bear.

I C R I S A T Goals

Long-term benefits expected through ICRISAT's cereal programs include:Consistent improvement in performance of sorghum and pearl millet over a wide range of environ-

ments throughout the SAT.

21

T H E C E R E A L S

Sorghum ( Sorghum bicolor)

Genotypes that have higher yield potential than now found in cultivated varieties.Improved resistance to insects, diseases, and other parasites.Improved nutritional value, cooking quality, and palatability.ICRISAT scientists are pursuing these goals at ICRISAT Center near Patancheru Village in

Andhra Pradesh and at a number of other locations throughout India and the world. They have estab-lished communications with cereals breeders throughout the SAT so that exchange and evaluation ofgenetic materials can be thoroughly accomplished, and so that all may share in progress and ap-proaches to problems. The emphasis is always on cooperation with scientists working in nationalprograms.

ICRISAT CenterICRISAT Center provides a wide range of environments for the development and testing of plantmaterials for the SAT. It has Alfisols and Vertisols and three growing-season environments. Therainy season, known locally as monsoon or kharif, occurs from June through September, and haslong days until the September equinox. The postrainy season, known locally as postmonsoon or rabiand occupying the months of October through January, has little rainfall and is cooler and days areshorter. Crops grown during the postrainy season rely on residual soil moisture or on irrigation.From February until the rains begin in June is known as the hot dry season. Temperatures at flower-ing time are very high during this season, and short-season crops may be grown if irrigation is pro-vided. Certain areas of the Center are saline and others have impeded drainage, making it possible toevaluate plant performance under a variety of conditions as they exist in many areas of the SAT.

22

S o r g h u m

Germplasm

Collection and Maintenance

The total number of accessions in our collectionincreased, with the addition of 797 lines, to15037. We rejuvenated 7930 accessions byselfing five to ten representative heads. Ninety-seven primitive sorghum types and landracematerial were collected in the hilly areas ofAndhra Pradesh and Orissa (Eastern Ghats).Additional material was obtained from cooper-ators in West Africa, Bangladesh, USA, Sudan,Sri Lanka, and the Lebanon. A valuable col-lection of 414 farmers' types was sent to us fromEthiopia.

Evaluation and Classification

The collection was evaluated for morphologicaland agronomic characters including plantheight, days to 50-percent flowering, awning,mid-rib color, plant pigmentation, tillering, pe-duncle exsertion, earhead length and breadth,panicle type, glume color, 100-seed weight,

1973 1974 1975 1976 1977

threshability, luster, presence of subcoat, andendosperm texture. In addition more than 2 700lines were assessed for resistance to diseases,insects, and Striga. Considerable material wassupplied for use in the breeding, biochemistry,physiology, and microbiology programs. Weprovided nearly 6000 germplasm samples tocooperating agencies in India and abroad(Table 1).

A pilot catalog of 300 accessions grown inpostrainy season 1974 was prepared in col-laboration with the Information Sciences/Genetic Resources Program, University of Colo-rado. We have sent additional information on7 215 lines to its laboratory for preparation of anexpanded catalog. The collection has been com-pletely classified according to Harlan and deWet's Scheme, and panicle branches have beenpreserved for future reference.

Introgression

We initiated research on introgression of genesfrom the germplasm into adapted breeding ma-terial; diploids and tetraploids are included in thestudy. Desirable segregants with large heads,bold grains, dwarf plants and photoinsensitivitywere obtained from F2 and (BC1) F2 pop-ulations utilizing CSV5 as the adapted cultivarand landrace material from Ethiopia, Ghana,and Cameroon. Work continues.

A broader-based project, involving parentalmaterial from selected landraces from the kaol-iangs, good-grain dochnas, yellow and corneousendosperm kafirs, caudatums, conspicuums, anddura kauras, zera zeras, and good guineas andusing the adapted cultivars CSV5, M 3 5 - 1 , Nl3Nandyal, 954063 (T x 3927-4), and 954062(T x 3925-2) is under way. Already we havemade 808 fresh crosses.

Breed ing

The dominant objective in the breeding programhas continued to be increased sustained yields inthe highly variable semi-arid environment.

Figure 3. Sorghum germplasm accessions at ICRISAT Center.

25

Sorghum161514

13121110

987654

3210

Table 1. Sorghum germplasm lines supplied to researc h agencies in India and other nations during1976-1977.

Institution Location Entries

I N D I A :All-India Coordinated Sorghum Improvement Project Hyderabad, A.P. 39Andhra Pradesh Agricultural University Hyderabad, A.P. 114Millet Research Station Madhira, A.P. 40Cotton Research Station Nandyal, A.P. 13Millet Research Station Podalkur, A.P. 11

Andhra University Visakhapatnam, A.P. 9Haryana Agricultural University Hissar, Haryana 840Kashmir University Srinagar, J&K 10College of Agriculture Dharwar, Karnataka 260Punjabrao Krishi Vidyapeeth Akola, Maharashtra 678

Agricultural Research Station Mohol, Maharashtra 19Marathwada Agricultural University Parbhani, Maharashtra 45Deccan College Poona, Maharashtra 20Mahatma Phule Krishi Vidyapeeth Rahuri, Maharashtra 24Punjab Agricultural University Ludhiana, Punjab 20

Rajasthan College of Agriculture Udaipur, Rajasthan 39Sugarcane Breeding Institute Coimbatore, Tamil Nadu 5Tamil Nadu Agricultural University Coimbatore, Tamil Nadu 48Regional Station for Forage Madras, Tamil Nadu 37

Production & DemonstrationIndian Grassland & Fodder Jhansi, U.P. 331

Research InstituteG. B. Pant University of Agriculture & Technology Pantnagar, U.P. 53

OTHER NATIONS:Agricultural Research Institute Bangladesh 384Mennonite Central Committee Maijdi Court, Bangladesh 10Dr. Mario Lira . Brazil 9IDRC Canada 4Mr. Guitierrez Palmira, Colombia 10

Chinese Academy of Agricultural & Forestry Sciences People's Republic of China 36Asian Vegetable Research & Development Centre Republic of China 25Ministerio de Agricultura y Ganaderia El Salvador 36Ethiopian Sorghum Improvement Project Ethiopia 1387EAAFRO Kenya 470

Agricultural Research Centre, U N D P Libya 20Shire Valley Agricultural Development Project Malawi 98School of Biological Sciences Malaysia 10University of Agriculture Malaysia 62University of Malaya Malaysia 25

continued

26

Table 1 continued

Institution Location Entries

C I M M Y T Mexico 142Centre Regional de Formation et d'Application en Niger 28

Agrometeorologie et Hydrologie operationnelleChinese Agricultural Technical Mission Saudi Arabia 27Hofuf Agricultural Research Centre Saudi Arabia 14University of Khartoum Sudan 25

University College of North Wales Bangor, United Kingdom 5Tropical Products Institute United Kingdom 100Purdue University Indiana, USA 12Texas A & M University USA 12N . I . Vavilov Institute of Plant Industry Moscow, USSR 239

CIARCO-Venezuela Venezuela 3United States of America Embassy Sana'a, Yemen Arab Republic 20UNDP Yemen 50Eglise du Christ au Zaire Zaire 4

Figure 4. ICRISA T sorghum breeder selects promising plants from composite populations.

Sorghum Population Improvement

Progress in improving populations is very en-couraging. After one or two cycles of selection,six of eight populations show an increase of 21 to36 percent over the base yield (Table 2). Therewas a reduction in yield of 8 and 11 percent in theother two populations. These results wereachieved even though selection was primarily forwhite pearly grain, elimination of photosensi-tivity, and stabilization of populations for plantheight. Plant heights were reduced, and there wasalmost no change in maturity.

Select lines from the composites were eval-uated under high fertility (130 kg N, 60 kgP2O5/ha on Vertisols) and low fertility (20 kg N,20 kg P2O5/ha on Alfisols) in the rainy season.The performance of selected lines is presented inTable 3. Four lines gave yields not significantlydifferent from the hybrid check (CSH-6) in highfertility; and likewise with 15 lines in low fertility.Three lines yielded significantly more than thebest varietal check, 370, in high-fertility Vertisolsand 18 lines were significantly superior to CS3541, the best varietal check on low-fertility

27

28

Table 3. Performance of promising entries in popula tion line evaluation trials on Vertisols and Alfiso ls atICRISAT Center during 1976.

Grain yieldMeanheight

Meantime to

flowering100-seedweight

PlantaspectaEntries Vertisol Alfisol Mean

Meanheight

Meantime to

flowering100-seedweight

Plantaspecta

(kg/ha)

3370322532773173

2385328731113101

352615041867

71617.0

4406433942464056

3 846397534423250

480029022607

67917.6

(cm)

137206196132

158137165152

18317814710

4.1

(days)

58585958

61565756

586064

2.93.0

(g)

2.332.362.092.24

2.732.622.232.99

2.692.692.45

Fast Lane 'R' 101Rs1 x VGC'R' Composite-RedFast Lane 'R ' 101-Red

Fast Lane 'R ' 266Fast Lane 'B ' 117Fast Lane 'B ' 100Rs/R 20

CSH-6CSV-3CSV-4

L.S.D. (0.05)C.V.(%)

5442545252154938

5 30746623 7733 398

607443013 348

96115.8

(kg/ha)

3370322532773173

2385328731113101

352615041867

71617.0

4406433942464056

3 846397534423250

480029022607

67917.6

(cm)

137206196132

158137165152

18317814710

4.1

(days)

58585958

61565756

586064

2.93.0

(g)

2.332.362.092.24

2.732.622.232.99

2.692.692.45

2.32.33.02.3

2.23.33.32.7

1.72.22.2

aPlant aspect is a visual score of general plant attractiveness: 1 = good to 5 =poor.

Alfisols. The experimental lines in this trial weredeveloped with evaluation across seasons andunder high- and low-fertility conditions. Theindications are that this may be a useful breedingprocedure for the selection of stable types. Yieldsof some of these same lines in the postrainyseason were: Fast Lane 'R' 274 (4600 kg/ha);Rs x VGC (4 500 kg/ha); Fast Lane 'R' 101(4100 kg/ha); and Fast Lane 266 (4000 kg/ha).These are good yields and comparable with thoseof the rainy season.

Lines from two of the advanced populationswere tested during the rainy season at twolocations each in India and Ethiopia and at singlelocations in Upper Volta and Senegal. Theselines were simultaneously evaluated for theirreaction to shoot fly, grain mold, Striga, charcoalrot, and downy mildew. About 40 lines wereselected to continue the next cycle.

Preliminary trials were conducted in otherpopulations to identify entries for multilocation

evaluation in 1977. Some 600 lines derived fromexisting populations were evaluated during therainy season on a multilocation basis; parents toform new populations will be selected from these.

The progress of composite formation, de-velopment, and multilocation evaluation is ap-parent. Potentially useful varieties are nowforthcoming.

Breeding for Grain-mold Resistance andImproved Grain Quality

Earlier-maturing types with resistance to grainmold and weathering are important with regardto yield stability. The scope of this project hasbeen broadened from a concentration on earlytypes to include a range of maturities (acrossvarious ecological zones) which would be earlierthan the locals. A large number of cultivars ofdiverse origin, maturity, grain-mold reaction,and grain quality have been crossed, and proge-

29

ny are being evaluated jointly with the cerealspathologists. Selections are listed in Table 4.

Seventy-two elite selections from F4 progenieswere distributed to cooperators in 12 countriesoutside of lndia for evaluation in the 1977 rainyseason.

In the 1976 "off-season" (postrainy season)nursery, 1 350 new single crosses were made and400 F1'sg were advanced; of these, 200 have beenselected for evaluation as F2's in 1977.

Grain quality is being improved by crossinggood early lines with others having desiredtraits - e.g., heavy bold seed, yellow endosperm,pearly white lustrous grain. The process ofevaluation, stabilization of a trait, and recrossingis continuous.

Table 4. Selections made in various generationsfor early good-grain, grain mold-resistant sorghum types at ICRISATCenter, 1976.

Generation Selections

(no)

Single cross F4 748Single cross F3 1030Single cross F2 1474Double cross F3 546Three-way cross F2 645

Total 4443

Figure 5. Discussing a sorghum breeding problem at ICRISA T Center.

30

A mold-resistant composite is being developedto increase levels of mold resistance and toimprove grain quality in an agronomically goodbackground through recurrent selection pro-cedures. Currently, male sterility is being in-corporated into selected entries.

Mold resistance is being incorporated intoadvanced populations and an attempt is beingmade to evaluate the effect of extra-long free-threshing papery straw-colored glumes as bar-riers to grain mold organisms.

A program initiated to develop cultivars suit-able for the postrainy season - incorporatinglines with 90 to 105 days maturity, excellent grainquality, substantial fodder value, resistance toshoot fly, and tolerance to drought - has reachedthe F2 stage. Seed from 500 crosses is availablefor evaluation in postrainy season 1977. Yieldsof some of the photoinsensitive lines tested in therainy and postrainy seasons 1976 are presentedin Table 5.

Attempts to recover sorghum in a plump grainwith high lysine content are frustrated by in-stability in the lysine concentration - with a reversion to lower (normal) lysine levels. Worktowards a better understanding of the reasons forthis is part of a Ph.D. thesis study now under wayat ICRISAT Center.

Across the breeding program, selection forlocally desirable grain qual i ty- i .e. , bold, rea-sonably hard, pearly white seed - is practiced. A program to attempt to evaluate breeding stocksfor food characteristics is being initiated.

Breeding for Striga Resistance

Screening sorghum in the laboratory to identifyplants with low or no strigol production is nowunder way. So far, 3053 cultivars have beenscreened and 115 were found to have low stimu-lant production. Some of these have at leastpartial resistance to shoot fly, midge, grain mold,and other yield reducers, thus enhancing theirvalue to the breeding program. Performance ofthe top 15 cultivars is compared to that of checksin Table 6. Difficulty in uniform field screeningled to large coefficients of variation.

Striga asiatica and S. densiflora are found in

India, while S. hermonthica is found primarily inSavannah Africa. Five Nigerian lines with re-sistance to S. hermonthica were also found to below-stimulant types for S. asiatica. Preliminarydata were obtained suggesting that many of thelines which are low-stimulant types for S. asiatica were positive for 5. densiflora; however there didnot appear to be much difference in reaction to S.asiatica seed collected from several sites in India(Table 7). These data are being checked.

It is now well established that some sorghumcultivars, though strigol positive, have resistancewhich operates after Striga seed has attempted toparasitize them. In cooperation with theICRISAT cereal physiology group, studies of themode of attachment - using root sections ofsusceptible and resistant genotypes - have beenmade. The mechanism of resistance appears to beassociated with thick-walled endodermal cellsand the presence of silica crystals in these cells(see Sorghum Physiology, page 35).

Our field-screening studies at Akola andICRISAT Center showed that 15 of the 275entries tested had high levels of resistance (Table6). These will be incorporated into agronomi-cally elite material.

Inheritance studies indicate that low strigolproduction is controlled by a single recessivegene. Further studies on the nature of inheri-tance of other resistance mechanisms will beundertaken with material currently beingassembled.

Breeding for Pest Resistance

This program was initiated with the objective ofstrengthening resistance to the three major pestgroups present in the SAT—stem borer, shootfly, and midge. Available resistance is beingincorporated into desirable plant types withgood yield and quality traits; populations arebeing bred to combine higher levels of resistancewith agronomically good plants. Resistancesfound in Indian and African sources were in-tercrossed, using male sterility, for a second timein the rainy season; progeny of these crosses willbe further evaluated. Selections following cross-ing, backcrossing, and double crossing have been

31

Table 5. Grain-yield dsta on some elite sorghum sele ctions under high-andthe rainy season and medium-fertility conditions du ring the postrainy season at ICRISATCenter, 1976.

Season

PedigreeRainy Rainy Postrainy

(high fertility)1 (low fertility)a (medium fertility)a

(kg/ha)

(NP x A L A D ) Sib 12-1(Bulk Y x IS 5 1 1 - 1 ) - 1 - 1 - 2(WABC x 954062b) Sib 23-1Good-grainGood-grain Pop. I I

521457035 1045 3634 985

3 7771 8511 3931 8962 578

26982727

70726112698

(Bulk Y x IS 511 -1 ) -1 -1 -3(WABC x Entomology) -7- l(Bulk Y x 954062)-10-2- l(Bulk Y x 954062)-3-2- l(Bulk Y x 954062)-10-2-3

49334 8004 6824 3854 311

235620151 7031 7781 659

2857207729873 7222727

(Bulk Y x 954062)-10-2-2Downes Bulk-1(WABC x Entomology)-10- l(Bulk Y x GPR 165)-2-4(Bulk Y x 954062)

41934037401540073644

1 8821 4372 8301 6302 444

35492712369430302712

Dekalb 15-4 (Yellow)(Bulk Y x GPR 165)-4-2(Bulk Y x GPR 165)-4-1(Bulk Y x GPR 165)-4-3CSH-1

3 1413 015298520445239

19111 5701 304

7702670

32753766346339683304

CSH-6CSH-7CSH-8UChV2

CS 3541GPR 148

6133

41263 8222148

2 304

2 037608

1007

178931023116

aHigh fertility reflects 130 kg N and 60 kgP per ha; medium fertility, 40 kgN, 20 kgP/ha; low fertility, 20 kgN, 20 kgP/ha.b954062 is the Texas Agricultural Experiment Station No. Tx 3925-2.

made to incorporate greater levels of toleranceinto 26 elite agronomic cultivars from South-eastern Asia and West and East Africa.

Initially a modest number of elite selections

from advanced populations were crossed toselected shoot fly-resistant parents to developpopulations combining pest resistance, yield,and quality. The population wil l be ready for the

32

-- -

----

first yield test in postrainy season 1979. It isproposed to incorporate stem-borer and midgeresistance into the same population.

Grain-grass Sorghums

Grain-grass sorghums are shorter, earlier, highertillering, and smaller in the head than normal

sorghums, and are intended for use in drought-prone areas where total rainfall is around 350 to500 mm.

Improved grain-grass types are available fromselections from crosses; a population is alsobeing formed.

Testing under low fertility and high pop-ulation density (450 000 plants/ha) took place

33

Table 6. Striga plants per sorghum plant, Striga plants per plot, and laboratory germination index(Swarna = 100) of the 15 best Striga-resistant line s selected in field testing at Akola duringrainy season 1976.

Striga Strigaplants per plants Strigasorghum per germination

Entry Source plant plot indexa Remarks

(no) (no) (%)

SRN 4841 Samaru, Nigeria 0.00 0.00 0.00 Resistant to S. hermonthica IS 5603 India (farmers' type) 0.00 0.00 180.1723-4 Mohol, India

(selection fromMatdani x Bonganhilo) 0.00 0.00 88.46

IS 2643 India 0.00 0.00 0.00 Resistant to 5. asiatica (BH4-1-4)

IS 6942 Sudan 0.00 0.00 97.31 Resistant to S. hermonthica SRN 1352 Samaru, Nigeria 0.01 0.25 6.92 Resistant to 5. hermonthica IS 1464 India (farmers' type) 0.02 0.25 162.50N-13 Nandyal, India 0.02 0.25 100.10 Yellow pericarpSRN 4310A Samaru, Nigeria 0.02 0.25 5.98 Resistant to S. hermonthica

SRN 6788A Samaru, Nigeria 0.02 0.50 106.68 Resistant to S. hermonthica 555 India (IS3687 x Aispuri) 0.03 0.50 0.00SRN 6838B Samaru, Nigeria 0.05 0.75 64.65 Resistant to S. hermonthica IS 5218 India (farmers' type) 0.05 1.00 98.38IS 8264 Uganda 0.06 1.25 104.56 Resistant to S. hermonthica

IS 1506 India (farmers' type) 0.08 1.50 93.80CSH-1 8.28 128.27 - Check (field tests)Swarna 100.00 Check (laboratory

germination test)

aReflects the number of seeds of S. asiatica germinated by the cultivar in the laboratory, expressed as a percentage of the numbergerminated by the check cultivar Swarna.

- -

Table 7. The relationship between cultivar and germ ination, expressed as laboratory germination index(Swama = 100), of Striga spp. from five locations in India.

Source ofSource of S. asiatica S. densiflora

ICRISATCultivar Center Parbhani Surat Akola Coimbatore Surat Coimbatore

IS 8315 x WABC 4072 0.00 0.00 0.00 0.00 0.87 26.41 0.00IS 7191 1.43 0.00 1.53 0.00 6.33 191.24 0.00IS 3962 x W A B C 1022 0.27 1.34 0.00 0.00 0.00 32.03 0.00IS 2404 0.00 0.00 0.00 0.00 0.00 21.44 0.00S 1137 2.62 1.68 2.18 0.00 0.00 73.73 0.00

IS 7415 0.00 3.47 0.00 0.00 2.05 56.45 0.00SRN-4310A 7.76 4.24 8.15 0.00 0.00 64.11 0.00IS 7227 2.78 0.00 0.00 0.00 0.00 12.83 0.00IS 2646 - 0.00 0.00 0.00 0.00 32.89 0.0076K3 -97 1.33 0.00 0.00 0.00 0.00 11.23 0.00

Framida 0.00 0.00 0.00 0.00 0.00 0.00 137.44IS 7408 3.24 0.00 0.00 0.00 0.00 99.65 0.00IS 3898 1.37 0.00 0.00 1.18 0.00 46.54 0.00SRN-5192 (IS 4997) 0.00 8.16 2.26 - - 0.00 -SRN-6496 (IS 9674) 0.00 4.84 0.00 - - 0.00 -Swarnaa 100.00 100.00 100.00 100.00 100.00 100.00 100.00

a The number of Striga seeds germinated by each cultivar is expressed as a percentage of the number germinated by Swarna. Withthe S. asiatica seed collected from Coimbatore, for example, the number of seeds germinated by IS 7191 was 6.33 percent of thenumber germinated by Swarna.

during postrainy season 1976. One selectionsignificantly outyielded the hybrid check, andeight significantly outyielded the better varietalcheck (Table 8). However, the hybrid and vari-etal checks were sown at the high populationlevel, which inhibited their performance. Trialsare now being conducted with grain-grass andgrain sorghum types, each at their optimumpopulation levels.

Identified sources of resistance to Striga, grainmold, shoot fly, stem borer, and midge have beencrossed into grain-grass types. These crosseswere evaluated in collaboration with entomol-ogists and pathologists; 22 were identified as lesssusceptible to shoot fly and 8 less susceptible toStriga. Further crosses have been made to iden-tify resistant sources.

Collaborative studies with the physiologistsindicate that some grain-grass lines have droughttolerance and good recovery after release ofstress. At ICRISAT Center, a plant populationof 400000 to 700000 plants per hectare isoptimum. Thirteen lines are being further eval-uated for ratooning capability under dry-landconditions.

Physiology

Mechanical Resistance to Striga

In order to better understand the nature ofmechanical resistance to Striga, we have under-taken studies of the mode of infection or invasion

34

of the Striga haustorium and of the anatomy ofthe roots of sorghum cultivars showing mechani-cal resistance.

After the Striga seed germinates, the haus-torium makes contact with the sorghum root andpenetrates its cortex, distorting or rupturing thecortical cells. The haustorium pierces the en-dodermis and pericycle. The elongated cells ofthe haustorium rupture and the differentiatingprotoxylem strands establish vascular connec-tion with the vascular bundle of the host root(Fig 7).

Resistance to Striga is correlated with the

presence of thick-walled endodermal cells (andassociated sclerenchyma tissue) in the pericycleand also the presence of silica crystals in thesecells. Haustoria were observed to penetrate thecortex of resistant cultivars, but failed to advancepast the endodermis into the vascular bundles,suggesting that the combined endodermis/pericycle formed a mechanical barrier (Fig 8).We are carrying out more-detailed studies incollaboration with the sorghum breeders andhope to further examine the role of the mechani-cal barrier in resistance, and to investigate itsinheritance.

35

Figure 6. An ICRISAT breeding nursery.

Table 8. Yield data of selected grain-grass linesin postrainy season 1976 low-fertilitytrials at ICRISAT Center.

Pedigree Grain yield

(kg/ha)

GC 497 1638GC 198 1319GC 421 1304GC 522 1304GC 318 1275

GC 50 1261GC 489 1232GC 478 1 188GC 441 1 150GC 529 1 145

GC 227 1 116GC 152 1101CSH-6 (Check) 1232CSV-3 (Check) 913CSV-4 (Check) 768

C.V. (%) = 28.85L.S.D. (0.05) 2.41

Drought Resistance

We have completed two sets of experimentsdesigned to evaluate the potential of controlledirrigation during the dry season for screening fordrought resistance, one in the hot dry summerwhen atmospheric demand is most severe and theother in the cooler postrainy season. In both sets,irrigation was withheld during the panicle-formation stage (approximately 30 days beforeflowering) when the effects of stress on grainyields are greatest.

The severe stress resulted in a far largerreduction in total crop dry weight, grain yield,and grain numbers than did mild stress (Table 9).Additionally, growth was interrupted in a largenumber of cultivars during the period of severestress, as indicated by differences in days tomaturity between control (irrigated) and the

Figure 7. Infection of the root of a Striga-susceptible cultivar by three Strigahaustoria.

stress treatments. This situation, observed in thehot dry season, was not present in the postrainyseason. Independent of the stress effects, weobserved significant seasonal differences in cropgrowth and yield components.

Grain yield in the presence of stress can bethought of as a function of the genetic yieldpotential in the absence of stress, a generalizedeffect of stress, and a specific genotype x stress interaction. These effects can be estimated in a simple linear regression model of yield in thepresence of stress as a function of yield in theabsence of stress, using the deviation of actualyield from predicted yield under stress to repre-sent the interaction term. This deviation termprovides a measure of the (relative) resistance orsusceptibility of the individual cultivars to stress,i.e. a high positive deviation suggesting relativeresistance and a large negative deviation indicat-ing susceptibility.

Yields under the moderate stress conditionswere observed to be closely related to yields in theabsence of stress (Fig 9A), with the mean absolute

36

deviation from predicted yield being only 9 per-cent of the predicted yield. This suggests that themild stress conditions were not sufficient for theexpression of genotype differences in response tostress. In the severe stress, however, deviationsfrom predicted yield were much larger (Fig 9B)with the mean absolute deviation equal to 26percent of the predicted yield and a range indeviations from —59 to +63 percent This pro-vided the opportunity to tentatively identify, foruse in future studies, a number of genotypes withpossible different responses to stress.

We attempted unsuccessfully to correlate theresults of genotype performance under mild andsevere stress. The reasons for this failure areundergoing study, as there are several potential

Figure 8. Failure of Striga haustoria to penetrate the endodermis of the root of a Striga-resistant sorghum cultivar.

explanations, including the possible lack of ex-pression of significant genotype x stress interac-tion in mild stress and the probability (as ex-perienced with other crops) that genotype x stress interactions are very specific to a given set of environmental conditions. A sol-ution is necessary before routine screening fordrought-stress resistance can be confidentlyundertaken.

Studies on Crop Growth and Yield

We analyzed data from a major study on growthand yield of 49 diverse genotypes. Overall cor-relations were established between grain yieldand the following growth characteristics: seednumber per panicle, seed weight, length of pan-icle formation (GS2), and harvest index.

We selected, for more-detailed analysis, 15genotypes with contrasting GS2 and GS3 (grain-filling) stages. Genotypes with a long GS2 pro-duced greater total dry weight, seed number, leafnumber, and leaf area (Table 10). Types with a long GS3 produced less vegetative growth, buthad a much improved harvest index and grainweight to leaf-area ratio. Our most significantfinding was that yield differences between thetwo groups of cultivars were not statisticallydifferent, indicating that both are capable ofproducing satisfactory yields. The long GS2 typehowever, should have a greater yield potentialbased on its ability to produce greater seednumbers and greater leaf area to support grain

Table 9. Mean effects of artificial drought stress dur ing the panicle-formation stage on sorghum yield andyield components.

Yield component ControlSeverestress

Mi ldControl stress

Time to maturity (days)Total dry matter yield (ton/ha)Grain yield (kg/ha)Harvest index (%)Grain number (1000/m2)100-seed weight (g)

10311.321500.24

12.922.7

1187.9

12300.177.2

19.1

97 976.8 5.6

2510 19400.37 0.35

10.3 8.333.1 30.0

37

Figure 9. Relationships of sorghum yields under stress with yields in absence of stress.

filling. There is also a much better possibility toimprove yields via improvements in harvestindex (converting a greater percentage of thetotal dry matter produced into grain) in the longGS2 types.

Entomology

Pest Assessment and Pest Populations

For the third season three standard sorghumcult ivars-CSH-1, "Swarna," and Local(Pachajonna)-were planted at the break of therainy season on a pesticide-free Vertisol area anddetailed insect observations recorded through-out the season. Shoot-fly infestation was low inthe early sown crop and 95 percent of the malesbred from the crop were Atherigona soccata. A potentially important predatory mite, identifiedas Abrolophus sp., was discovered feeding oneggs and first-instar larvae of this pest. Attack bystem borer, Chilo partellus, was heavier thanusual, and also earlier; however, little differencein damage was recorded between cultivars. Atharvest time more larvae were present in CSH-1than in the other cultivars, and larvae tended tobe more numerous in harvested stalks, to persistlonger, and to survive better in this cultivar.Midge attack was low.

We prepared a detailed list of insects and their

parasites recorded from sorghum at ICRISATCenter; the list will be published.

Shoot-fly Taxonomy and Biology

Of seven species of shoot fly bred from sorghum,only A. soccata is important as a pest. Threeadditional alternative grass hosts of shoot flywere identified - Cymbopogon caesius (Nees)Stapf, Echinocloa crusgalli (Linn.) P. Beauv., andEragrostis japonica (Thunb) Trin. Three othergrass hosts, one of which appears to be ofsignificance, remain unidentified. Shoot flieswere also bred from wheat and Sorghum hal-epense (Linn.) Pers.; the latter is a potentiallyimportant carry-over host. In a study to de-termine if diapause occurs in the pupal stage,most pupae emerged at the expected time - butone took 10 weeks, indicating that extendeddevelopment is possible.

Studies on the various shoot-fly species andalternative hosts confirmed prior observationsthat flies are highly discriminative in selection ofa host plant. Although A. oryzae and A.falcata were dominant on grasses, the former clearlypreferred E. japonica and Digitaria adscendens (H.B.K.) Henr. and the latter Echinocloa col-onum (Linn.) Link and E. crusgalli (Reddy & Davies 1977). A total of 18 different species ofshoot fly has been recovered from fields atICRISAT Center- four of which are new toscience.

38

A. M i l d Stress

r = . 9 4 * * *

3000

2 500

2000

1500

1000

1000 1500 2000 2500 3000 3 500 4000

Yield in control ( kg /ha )

1000 1500 2000 2 500 3000 3500

Yield in control ( kg /ha )

2500

2000

1500

1000

B. Severe Stress

r = . 5 7 * * *

Table 10. Comparison of characteristics of genotypes with long GS 3 and long GS 2.

Mean(Sample size= 15) Difference

Character Long GS3 Long GS2in means*

( G S 3 - G S 2 )

Grain weight (g/plant) 40.9 51.4 -10.5 NSGS1 (days) 32.1 34.0 - 1 . 9 NSGS2 (days) 33.0 53.9 -20 .9 *GS3 (days) 53.6 40.3 13.3*Seed (number/panicle) 1030 1402 - 3 7 2 *100-seed weight (g) 4.13 3.77 0.36 NS

Seed (no/GS2 day) 31.2 26.3 4.9 NSSeed weight (g/GS3 day) 0.77 0.96 -0 .19*Plant height (cm) 183.0 229 -46 .0 *Total dry weight (g/plant) 96.6 167.8 -71 .2 *Harvest index (%) 42.5 32.2 10.3*Grain/leaf ratio (g/cm2) 5.7 3.5 2.2*

Stem weight/total weight (%) 24.4 34.2 - 9 . 8 *Total leaves at flowering 13.7 17.5 - 3 . 8 *Leaves produced in GS2 4.30 8.3 - 4 . 0 *Position of largest leaf (flag leaf = 1) 2.82 4.33 - 1 . 5 1 *Area of upper leaves (cm2)b 506 952 - 4 4 6 *

aNS, Not significant;*, Significant at 5 % level, according to Cochran-Cox t-test.bLeaves above largest leaf, largest leaf included.

Use of fish-meal traps to assess the seasonalincidence of shoot fly revealed two peaks ofactivity - one in August/September and the otherin December. Over the year approximately 12times as many females as males were caught inthe traps at 20 sites. Of a total of 21108 malescaught, 51 percent were A. soccata. We recordedlow catches in the summer months (Apr-Jun),and since they were usually female, identificationby species was impossible.

Stem-borer Biology

In our study of alternative hosts of Chilo par-tellus, the insect was bred from cultivated sorg-hum, pearl millet, maize, and Setaria. Grass

alternative host records included Dinebra re-troftexa (Vahl) Panz., Hetropogon contortus (Linn.) P. Beauv, Cymbopogon caesius, Chloris barbata Sw., and Sorghum halepense.

As part of our intensified program on thestem-borer pheromone, observations were madeon virgin females as attractant sources. Day-oldfemales were highly attractive-up to 173 malesbeing attracted to one female in a single night.The ability of virgin females to attract appearedto decline after about 3 days, but catches wereobtained up to 10 days from exposure. Irrigationof fields boosted trap catches and this obser-vation was used in screening studies.

We carried out several experiments usingfemale pheromones synthesized at the Tropical

39

Products Institute, London. The major phero-mone caught significantly more moths than allother treatments, and there was an indicationthat the minor pheromone was an inhibitor -since fewer moths were caught than with thevirgin female and even fewer than in the "con-t ro l " water traps (Fig 10). We recorded signi-ficant differences in catch over nights but this wasnot solely due to ageing of the pheromone, sincepeaks were recorded in the water traps also. Thepeaks appeared to be associated with irrigationof the experimental blocks or adjacent areas orwith rain (Fig 11). In subsequent experiments inwhich we compared the ratios of major to minorpheromone with virgin females, six parts ofmajor to one part minor was optimal for catch,

Figure 10. Mean daily catch of Chilo partellusmales on five treatments over 15 nights at ICRISAT Center (March 1977).

but all pheromone traps caught fewer malemoths than did the virgin females. The syntheticpheromones were active for up to 49 days.

Screening for Pest Resistance

Shoot fly. We screened a large amount of thesorghum germplasm for shoot-fly resistance,using the fish meal and interlard technique. Twogroups of elite material which had previouslyperformed well were submitted to high levels ofattack using a grid system. IS 1054, 2146, 2162,2312, 5604, 5604 x 23/2, 5613, 5622, and 5648were confirmed as nonpreferred for oviposition.Lines recently collected from farmers' fields bygermplasm botanists and untested IS lines weregrown twice, and IS 1119, 5296, 5113, 2195,5480, and 8311 were nonpreferred for ovi-position. Taking subsequent dead heart andhead production counts into consideration, IS1119, 5480, 2176, and 2195 were selected forfurther investigation.

Lines showing some shoot-fly tolerance weregrown in larger blocks at three scattered sites atICRISAT Center. IS 1082, 2312, 4002, 4036,4829, 5642 and derivatives IS 1082 x WABC4062, IS 5642 x R 960, EN 3308, and EN 3362-1

Figure 11. Mean nightly catch of Chilo partellusover 15 nights in 15 traps at ICRISATCenter (March 1977).

40

Nigh ts

1 2 3 4 5 6 7 8 9 10 1112 13 14 15

55

50

40

30

20

10

0

Rain

50

45

40

35

30

25

20

15

10

5

Major Major Minor Virgin WaterPhero- plus Phero- Femalemone minor mone

all performed well for shoot fly-none havingmore than 1 percent dead hearts. However theywere moderately susceptible to Chilo partellus -the most resistant line was IS 4829, with a meanof 49 percent of the stems affected. Maximumhead production per unit area was obtained fromIS 4002, 4036, EN 3362-1, and EN 3308-thelatter being extremely productive. The best grainyields were obtained from IS 4036 and EN 3308.The lowest number of larvae in stems at harvestwas on IS 4002 (9 % of stems) and the highest wason IS 4829 (23 %). IS 1082 x WABC 4062 had 20percent. More than 5 800 breeders' lines werescored for shoot-fly attack.

Stem Borer. Production of Chilo larvae, usingan artificial diet, was sufficient to initiate artificialinfestation studies on germplasm lines. Linesshowing promise in two seasons of testing were E 302, E 303, BP 53, IS 1044, 1056, 1151, 2122,2205, 4747, 4776, 4799, 4866, 5030, 5470, V-2-1-1-1, and V-2-1-1-2. In "grow out" trials of muchof this material under high natural infestationconditions, E 303, 1044, 1151, and 5470 all hadless than 30 percent damage compared to CSH-1with 69 percent.

Midge. Studies on the damage, biology, andresistance to midge were intensified, but workwas hampered by the low levels of attack atICRISAT Center. Seventy-six lines suspected ofhaving some resistance to midge were sown atSillod and Selsoor with the cooperation of theMaharashtra Department of Agriculture. Thesepreliminary trials indicated that SC 173 (IS12664 C), SC 175 (IS 12666 C), SC 423 (IS 2579C), SC 329 (IS 3574 C), SC 63 (IS 12573 C), andSGIRL-MR-1 should be further tested.

Other Studies

Some work was initiated on storage pests ofsorghum as a preliminary to work on storagecharacteristics of sorghum cultivars. There weredistinct differences in Sitophilus oryzae multipli-cation rates in CSH-1, "Swarna," and Localsorghums.

Our insecticide trials failed to elucidate

reasons for the apparent failure to control Athe-rigona soccata with carbofuran, but did confirmthat the insecticide was less effective than inprevious use.

Pathology

The main objective of the sorghum pathologyprogram is to minimize the possibility thatimproved high-yielding sorghums will be vulner-able to disease epidemics. To achieve this, stablebroad-spectrum resistance must be located andincorporated into high-yielding genotypes.

Screening for Grain-mold Resistance

Short-cycle sorghums need resistance to grainmold as they often mature under wet conditionsand produce grain with low market acceptabilityand poor viability, and may contain mammalio-toxic fungal-produced mycotoxins.

Seventeen fungal species in 11 genera wereisolated from moldy sorghum grain. The mostfrequently isolated genera were Fusarium, Cur-vularia, Tricothecium, and Olpitrichum, in thatorder.

More than 5000 germplasm lines were in-oculated with pathogenic isolates of Fusarium semitectum and Curvularia lunata, either singlyor together, and then compared with nonin-oculated heads in the same row. Conditions inrainy season 1975 were favorable for molddevelopment and only 90 lines with little or nomold were selected for further tests. Unfor-tunately in rainy season 1976, September andearly October had but little rain and this was notconducive to profuse mold growth. However,mold development was sufficient for differen-tiation of low and high susceptibility; of 1421lines tested, 33 were selected. The best materialfrom 1975 was tested at several locations in theInternational Sorghum Grain Mold Nursery,discussed in a subsequent paragraph.

Screening for Downy M i ldew, Lea fDiseases, and Stalk Rot

Work on diseases other than grain molds was

41

focused on learning how to manage or promotethem for screening of germplasm and breedingmaterials. The testing of stability of identifiedresistance was begun through the cooperativeInternational Sorghum Disease Resistance Test-ing Program discussed below.

For sorghum downy mildew, use of early sownalternate rows to serve as an inoculum sourceshows promise for screening test material.

Sorghum grains infected with the anthracnoseand leaf blight causal agents were placed in leafwhorls to promote these two diseases. Theusefulness of this method will be further in-vestigated for other nonobligate leaf pathogens.

Several methods and timings of inoculationswith the charcoal-rot organism are under studyand the technique selected will be used forscreening breeders lines for susceptibility to thisdisease in subsequent seasons.

The International Sorghum DiseaseResistance Testing Program ( ISDRTP)

In collaboration with scientists from African andAsian countries, three nurseries-grain molds(ISGMN), leaf diseases ( ISLDN), and downymildew ( ISDMN)-were widely tested. Theseinternational nurseries serve a vital function inthat sorghum lines are exposed to many pop-ulations of the pathogens under a wide range ofenvironmental conditions, which is essential forthe identification of stable broad-spectrumresistance.

ISGMN. This nursery was sent to cooperatorsat 12 locations in six countries in Asia and WestAfrica. Very wet conditions during maturation inWest Africa gave high levels of mold pressure,but in India the dryness in September and earlyOctober inhibited mold growth. Seventeen of theSO entries had mold ratings of 3 or less (in a 1-5rating scale). The best entry across locations wasE 35—1, a large white-seeded Zera Zera type fromEthiopia.

ISLDN. This nursery was sent to cooperators atnine locations in three countries in Asia andAfrica. Two lines appeared to be resistant across

locations to anthracnose, the most prevalentdisease. Only one lowland site had sufficient leafblight to adequately test for the disease, and here2 lines were free and 26 showed only slightsymptoms. There is a need to identify sites wheremore of the leaf diseases occur consistently, sothat we can adequately test for stability of theresistance.

ISDMN. High levels of sorghum downy mil-dew occurred at three locations in southern Indiaand two entries - Q L - 3 and SC 120-14- were freefrom systemic infection. Eleven entries had meanincidence values of less than S percent andmaximum incidence values of less than 10 per-cent. We gratefully acknowledge the valuableassistance of our cooperators at Dharwar,Mysore, and Coimbatore in providing a goodstart to the ISDMN program. We need to locatesites in other parts of Asia, Africa, and southernUSA where downy mildew occurs regularly.

Nut r i t i ona l Qual i ty

The newly acquired T A A (Technicon AutoAnalyzer) was used for determination of proteinafter standardizing against the MKJ (micro-Kjeldahl) method. Close correlations (r = + 0.997) were obtained between the methods; therange of error was between —1.3 and +1.7percent (Table 11).

Further refinements were made in the pro-cedure for lysine estimation, using the DBC (dye-binding capacity) method. Use of a standardweight of one gram of sample and computing onthe basis of 80 mg protein gave results in closeagreement with those obtained using samples ona constant protein basis of 80 mg. This simplifiedmethod will be used in the future. A correlationof r = 0.89 was also obtained between the U I R(Udy Instrument Reading percentage trans-mission) on a one-gram sample and the M K Jmethod.

Based on these results the UIR/P values werecorrelated with the actual lysine values of sam-ples, using 100 sorghum samples with a proteinrange of 7.3 to 19.1 percent and UIR/P values of

42

Table 11. Comparison of MKJ (micro-Kjeldahl)and TAA (Technicon Auto Analyzer)values of protein in sorghum grain.

Average protein

Protein class Samples M K J T A A Error

(%) (no) (%) (%) (%)

7.0-7.9 6 7.65 7.55 - 1 . 38.0-8.9 10 8.36 8.36 0.09.0-9.9 9 9.53 9.49 - 0 . 4

10.0-10.9 10 10.44 10.52 +0.8

11.0-11.9 7 11.56 11.59 +0.312.0-12.9 10 12.35 12.41 + 0.513.0-13.9 11 13.31 13.34 + 0.214.0-14.9 13 14.44 14.49 +0.4

15.0-15.9 7 15.36 15.59 + 1.516.0-16.9 10 16.44 16.38 - 0 . 417.0-17.9 4 17.28 17.20 - 0 . 518.0-18.9 2 18.05 18.35 + 1.719.0-19.5 2 19.05 19.15 +0.5

between 2.07 and 4.57. A close correlation(r = +0.93) was obtained between UIR/P andlysine concentration (g/100 g protein) and a regression equation calculated. Further work isunder way to determine whether this can be usedwith confidence as a reliable method for pre-diction of lysine values.

A range of 100 germplasm samples of eighttypes (with luster, with subcoat, completelycorneous, almost corneous, intermediate, almostfloury, completely floury, and with waxy en-dosperm) were analyzed in subgroups, takinginto account co lor-whi te through yellow tobrown, red to purple. The 100-seed weight(corrected for moisture) ranged from 1.16 to 4.64g, protein from 9.3 to 16.9 percent, starch from72.8 to 83.6 percent, soluble sugars from 0.8 to1.8 percent, fat from 2.2 to 5.5 percent, fiber from1.0 to 2.3 percent, and ash from 1.2 to 2.2percent.

Routine screening of 6 758 sorghum samplesfor protein and basic amino acids using the T A A

and the DBC methods respectively gave a rangeof6.0 to 21.1 percent (mean 12.1 %) protein. TheUIR/P ratio was 1.84 to 5.57 (mean 3.01) (seeFig 12).

M i c r o b i o l o g y

Efforts are concentrated on developing an assaymethod for screening large numbers of field-grown plants for nitrogenase activity of bacteriaassociated with their roots. The assay is based onthe reduction of acetylene to ethylene by thenitrogenase enzyme. With as little soil distur-bance as possible, soil cores containing plantroots are taken in metal cylinders (15-cm diam-eter, 22-cm length) and incubated in sealed 6-literplastic vessels under an atmosphere of ca. 15percent acetylene in air. Gas samples are taken 17and 24 hours later for gas chromatographicanalysis of ethylene production.

Cores in an atmosphere without added acet-ylene produce little ethylene. It is assumed that

43

Figure 12. Range of estimated lysine content in 6 758 sorghum samples analyzed at ICRISAT.

1050

900

750

600

450

300

150

Total number ofsamples: 6 758

Range 1.84 - 5.57Mean: 3.01

2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8

U I R / P Rat io

acetylene-induced ethylene production is due to

reduction of acetylene by nitrogen-fixing bac­

teria and not due to inhibition by acetylene of the

oxidation of endogenous soil-produced ethylene.

For estimating the amount of nitrogen fixed, a

conversion ratio of 3 C2H

4 is equivalent to 1 N

2 is

used.

Using this assay, sorghum lines grown under

low-fertility (20 kg N, 9 kg P fertilizer/ha)

conditions on Alfisol at ICRISAT Center were

screened for nitrogenase activity in the 1976

rainy season and in the 1976-1977 postrainy

season, as were lines grown at Bhavanisagar in

southern India during the hot dry season of 1977.

These lines included landrace cultivars from

different locations and environments, crosses

and selections which performed well under both

high- and low-fertility conditions at ICRISAT

Center, hybrids and their parents, grain-grass

types, fodder types, and related Sorghum species.

Of more than 347 entries, 110-including some

members of all of these groups - had nitrogenase

activity greater than 50 µ g N/core per day.

There were large differences between cultivars,

even though considerable variation in activity

occurred from plant to plant. For example, the

active line IS 2333 from Sudan had a mean

activity of 325 µ g N/core per day (range 127 to

534 µ g N) when assayed during the grain-filling

stage. This plant-to-plant variability does not

seem to be related to differences in dry weight of

plant top or root. Some lines supported no more

activity than did soil cores without roots, but 26

lines had a mean activity greater than 100 p g

N/core per day.

Activity continued well into the grain-filling

stage and was usually greatest after flowering.

There was no activity in the postrainy season

until the beginning of March. This was probably

due to the cool (night?) temperatures earlier in

the season. The nitrogen-fixing bacteria are

closely associated with the roots, with little

activity in cores with soil only. Activity decreases

as soils dry, although plants do not show wilt

symptoms.

Ninety-four entries were also grown at Bhava­

nisagar during the hot dry summer season. There

were again significant differences between lines

when assayed during the grain-filling stage. The

line IS 2980 from Ethiopia had the largest mean

activity of 127 µ g N/core per day (range 12 to

351 µ g N). A further 16 lines had mean activities

greater than 50 p g N/core per day.

Some lines, including several grain-grass

types, were consistently active in all three grow­

ing seasons. Sudan grass and Sorghum halepense

were also active.

The activity of sorghum roots after flowering

extends laterally at least 30 cm beyond the single

core over the crown of the plant. Figure 13 shows

one experiment, assayed 14 Apr 1977, where the

core with the crown had 114 µ g N/core per day

while activity in surrounding cores ranged from

15 to 180 p g N/core per day.

S o i l a l o n e : 2 2 µ g N / c o r e p e r d a y

S o i l p l u s l o o s e r o o t s : 72 µ g N / c o r e p e r d a y

Figure 13. Nitrogenase activity (pgN/core per

day) of soil cores taken over the crown

of sorghum CSH-1 and in the sur­

rounding unplanted area. Activity

measured in the core is represented by

the number within the circle.

Looking Ahead in Sorghum Improvement

Germplasm. The collection will be maintained

and lines purified and grown out as necessary to

ensure conservation of genetic stock. Lines will

44

be characterized using descriptors recommendedby IBPGR, and ICS numbers will be assigned tonew collections. Introgression work will con-tinue, as will screening in collaboration withother disciplines: pathology, microbiology, andentomology.

Breeding. We aim to bring together andstrengthen traits for improved yield and yieldstability, and to provide source material on anexpanding basis to sorghum breeders in nationalprograms. This will involve incorporatingresistances - by conventional and populationbreeding methods-to diseases, pests, anddrought in our advanced breeding lines. It wil l , ofcourse, be a cooperative effort with ICRISATcolleagues in various disciplines including path-ology, entomology, and physiology.

Increasing emphasis will be placed on ex-change and testing of breeding material withICRISAT cooperating centers and cooperatorsthroughout the SAT. A particular aim will be toextend work on screening for the African Striga species and strains.

Physiology. Studies on the dynamics and de-terminants of growth and yields in variousgenotypes will be continued. We will expandgreatly the work on drought resistance - initiallyby developing suitable screening techniques andsubsequently by investigating the nature of re-sistance and susceptibility - so that breeders canselect positively for favorable traits. We intend tocontinue studies on seedling vigor and cropestablishment, and will conduct collaborativeresearch with the breeders and entomologists forselection criteria and the physiological and mor-phological aspects of pest and Striga resistance.

Entomology. The emphasis of our screeningprogram will move from identification of sourcesof borer and shoot-fly resistance in the germ-plasm to the screening of lines produced in thebreeding program. In the work with midge,development and use of a screening method willhave priority.

Observations on the biology of the major pestspecies will continue and comparisons will bemade between the pest complex at ICRISATCenter and elsewhere in the SAT. Informationcoming from our cooperating programs is essen-tial for this. Overseas testing will be intensified.

The work on Chilo pheromone will enter a phase of attempting to assess whether phero-mones can be used for control of this pest as wellas for sampling of populations.

Pathology. Work will continue on the develop-ment of improved screening procedures for themain sorghum diseases - grain molds, leaf dis-eases, downy mildew, and stalk ro ts -and theutilization of the techniques in the location andincorporation of stable resistance in high-yielding lines.

The multilocational disease nursery program,conducted in cooperation with colleagues innational and regional programs throughout theSAT, will enable the identification of stableresistance and will provide a means of givinglocal programs material with much broaderresistance than could have been identified fromwithin the local program.

Nutritional quality. On completion of the stan-dardization of methods for protein estimationand modification of the screening method forlysine, we will be able to handle a larger volumeof samples for the breeding program. More workwill be focussed on other important areas such ascooking and keeping quality, and on digestibilityof sorghum.

Microbiology. The aim of this program is todetermine whether nitrogen-fixing bacteriaclosely associated with sorghum roots have thepotential for reducing the amount of nitrogenfertilizer required by sorghum crop. We willimprove the techniques for measuring theamount of nitrogen fixation associated withsorghum cultivars and embark on an expandedprogram to identify sorghum lines with enhancednitrogen-fixing activity.

45

P e a r l M i l l e t

Germplasm

We have now received 5048 accessions whichinclude all the available IP (Indian Pennisetum)entries and 351 added this year from parts ofIndia and West Africa. The great majority ofthese accessions are landraces collected fromfarmers' fields, but unfortunately most of theolder collections have been excessively selfed orcontaminated. The recently appointed milletbotanist has therefore been determining whichmembers of the IP collection (including dup-licates) still agree with the catalog, and devisingpractical methods of maintaining variabilitywithin the recent accessions.

A total of 4 317 entries were sown, standardobservations recorded, and selfed seed retainedfrom some. Of 2 250 IP lines studied, 1075agreed with the catalog description. A further800 exhibited useful traits and were also kept.Twenty-six lines were selected for use in thebreeding program. High levels of rust infectionoccurred in the germplasm plots and 128 rust-free plants were identified. Allowing plants sownin hills to intercross under large bags, coupledwith long-term storage facilities, appears to be a feasible way of maintaining collected variability.The postrainy season, where days are short, willbe the most "neutral" season in which to do this;seed quality will also be best at this time.

During the year seeds of 1 540 accessions weresupplied in response to requests (Table 12).

Breeding

Breeding continued with two main approaches(recurrent selection in composites, and varietycrosses) to produce varieties, hybrids, and in-breds for ICRISAT and international yield trials,as well as an array of breeding material fordistribution to breeders (Fig 16). The develop-ment by ICRISAT cereals pathologists of aneffective field technique using infector rows nowallows thousands of entries to be screened agai-nst downy mildew, in both the rainy and post-rainy seasons.

Source Material

The working collection, consisting of 340 linesfrom diverse geographical regions having char-acteristics of agronomic interest, when grownover several seasons and locations indicatedexistence of an enormous variability for day-length reaction, disease reaction, grain size,drought tolerance, protein content, and resto-ration ability on the three cytoplasmic male-sterile systems. Data gathered on this collection,seed of which is available, has been sent forprocessing to the Information Sciences/GeneticResources Program located at the University ofColorado, USA. As further germplasm ac-cessions, including the intermediate forms (shib-ras) from Africa, are acquired at ICRISATCenter, the working collection will be expanded.

In addition to the existing five source pop-ulations in this project (Casady, Nigerian Dwarf,Mokwa-maiwa, Maiwa, and Ex-Bornu) we haveadded four dwarf (D2) populations from IRAT,West Africa.

These populations (Table 13) are being im-proved by progeny evaluation and crosses havebeen made with good progenies to male steriles;to move desirable characteristics into more ag-ronomically advanced backgrounds, crosseshave been made to elite material.

Downy mildew-resistant lines have been iden-tified for recombination in the Casady andNigerian dwarf populations, each crossed ontomale steriles.

Figure 14. Pearl millet germplasm accessions at ICRISAT Center.

Mi l l e t8765432101973 1974 1975 1976 1977

Table 12. Millet germplasm lines supplied to researc h agencies in India and other nations during 1976-1977.

Institution Location Entries

I N D I A :Agricultural Research Station Anantapur, A.P. 60Millet Research Station of Lam Guntur, A.P. 62M.S. University Baroda, Gujarat 4Haryana Agricultural University Hissar, Haryana 2National Dairy Research Institute Karnal, Haryana 94

Punjabrao Krishi Vidyapeeth Akola, Maharashtra 7All-India Coordinated Millet Improvement

Project Rahuri, Maharashtra 330Mahatma Phule Krishi Vidyapeeth Rahuri, Maharashtra 12Indian Agricultural Research Institute New Delhi 2

Punjab Agricultural University Ludhiana, Punjab 96Agricultural Research Station of Durgapura Jaipur, Rajasthan 80Central Ar id Zone Research Institute Jodhpur, Rajasthan 32

University of Jodhpur Jodhpur, Rajasthan 12Tamil Nadu Agricultural University Coimbatore, Tamil Nadu 41Chandra Shekhar Azad University of

Agriculture and Technology Kanpur, U.P. 8Banaras Hindu University Varanasi, U.P. 9

OTHER NATIONS:National Research Council Saskatoon, Canada 22Mr. Guitierrez Palmira, Colombia 10Chinese Academy of Agricultural and Forestry

Sciences Peking, People's Republic of China 12Afdeltng for Dyrefysiologi Biokem. Denmark 15

UNDP Addis Ababa, Ethiopia 1College of Agriculture,

University of Ethiopia Debre Zeit, Ethiopia 8Port Dauphin. S.A. Haiti 4Agriculture Research Centre Tripol i , Libya 14Kasinthula Research Station Chikwawa, Malawi 65

C I M M Y T Mexico 30USAID, Area Development Office Niamey, Niger 114University of Ibadan Ibadan, Nigeria 62Institute of Agricultural Research Samaru, Nigeria 969USAID/USDA Islamabad, Pakistan 449

CNRA Bambey, Senegal 376Plant Breeding Institute Cambridge, United Kingdom 21University College of North Wales Bangor, United Kingdom 5Fundacion Servicio Para el Agricultor Venezuela 18Mennonite Central Committee Kinshasa, Zaire 10

50

Figure 15. Cluster bagging and selfing to maintain pearl millet germplasm.

The landrace population Ex-Bornu from Nig-eria has shown wide adaptability, good yields,and disease resistance in many trials. Seventeengood S3's have been identified and crossed indiallel to choose entries for a synthetic.

Mokwa-maiwa and Maiwa are highly photo-sensitive but appear to carry good downy mildewresistance and seed-weathering resistance. S1's ofthese populations have been crossed to Liguifrom Chad, which is very early and photoin-sensitive. In the F2's, grown in the summerseason of lengthening days, photoinsensitiveselections with maiwa vegetative characters weremade.

The four IRAT synthetics are dwarf, veryleafy, and relatively late maturing, but withexcellent head size. Numerous progenies havebeen test-crossed both to male steriles and to

established inbreds from our variety cross pro-ject. A total 106 lines from 3/4 Souna and 3/4 Ex-Bornu remained disease-free in an initial screen-ing for DM (downy mildew) resistance. Usingartificial inoculation, a further 86 remained freeof smut, and 12 free of ergot.

Finally in this source material project, whichrepresents an intermediate stage between germ-plasm and leading breeding material, we arereconstituting our "dwarf" nursery of inbredlines. Of these, 280 originated from the Jamnagarbreeding program and exhibit extreme advancesin plant habit.

Composite Breeding

Fourteen composites are being improved byintrapopulation methods to generate experimen-

51

NATIONAL TRIALS IPMAT NATIONAL BREEDERS

Figure 16. The pearl millet breeding program at ICRISAT.

tal varieties and best progenies for variety yieldtests (Fig 17). In 1976 two cycles of selection werecompleted on four composites, the first cycle onnine more, and the last is undergoing randommating prior to the commencement of recurrentselection. Selection is effected through testingbetween 225 and 315 progenies at three or morelocations with one replication in the DM nur-sery. Seed for recombination is taken only fromplants remaining disease-free in this nursery. Theselection differential obtained for the mean grainyield of the progeny selected for recombinationhas been about 32 percent, and 25 to 30 progenyare kept per composite.

Yield tests, conducted on several sites in 1976in India, included 22 experimental varieties, 81best progenies, and 7 C0 vs. C1 composite bulksproduced as a result of 1975 progeny tests. Fourexperimental varieties-WC-C75, IVS-A75,SSC-C75, and MC-C75 - gave yields equal to thehybrid check PHB 14 (2070 kg/ha) with betterdowny mildew resistance. WC-C75 was alsosuccessful in A I C M I P trials and has been selec-ted for advanced testing.

Table 13. New source populations from IRAT in West Africa.

Population Country of origin Description

3/4 Souna Senegal Obtained by crossing the local selection "Souna"( IRAT P. 10) 2 to a D2 dwarf source. Selection for resistance

to downy mildew and smut and for goodagronomic characters.

3/4 Ex-Bornu Niger Developed by crossing Nain 1/2 Souna x Ex-( IRAT P. 16) Bornu. Improved Indian germplasm and D2

gene were also included. Resistant to downymildew and smut.

3/4 Hainei-Khirei Niger Obtained by crossing Nain 1/2 Souna x ( IRAT P. 15) Selections of variety P3 Kolo (Zongo x H K )

and Hainei-Khirei. To this were added germ-plasm from Senegal, India, and the D2 gene.Resistant to downy mildew and smut.

Saria Synthetic Upper Volta Resulted from six best lines selected in the( IRAT P. 8) segregating generations of crosses involving

local lines in 1971. Resistant to downy mildew.

52

Figure 17. Generalized scheme for intrapopu-lation improvement in pearl millet at ICRISAT Center.

Individual best progenies from good com-posites increased by sibbing showed yield in-creases of up to 25 percent more than the parentcomposite and equalled the yield of the newAICMIP hybrid BK 560-230.

Though the difference was not statisticallysignificant, an average of 5.3 percent yield gainwas recorded in the comparison of C0 to C1 bulksin seven composites - the best being 13 percent.Reductions in height and increases in downymildew resistance were obtained. Evidence fromthe next cycle (C2) of progeny testing in thesecomposites showed no loss in variability foryield, indicating that continued gains may beexpected. This implies that future improvementswill be reflected in subsequent experimentalvarieties.

One further composite is to be added to thisproject - an Elite Composite to be made of notmore than 20 good tested diverse entries.

Reciprocal recurrent selection (inerpopu-lation improvement) has commenced, primarily

Figure 18. Multiplication plot of an experimental pearl millet variety at ICRISAT Center.

for the production of hybrids, on one com-plementary p a i r - I B and IR. Parents of 23reciprocal full sibs were selected for recom-bination from the 1976 composite progeny testswhere the selection differential was 24.5 percent.The individual hybrid pairs will be developedand evaluated as potential parents.

A diallel of 16 composites (parents and F1's,excluding reciprocals) was tested in rainy season1976 in two environments (high fertility and lowfertility) at ICRISAT Center. The data wereanalyzed according to Gardner and Eberhart, aswell as by Griffing's model. Combining-abilityanalysis indicate that the additive type of geneaction played a predominant role in the de-termination of grain yield. Based on parentalmeans, composite cross performance, heterosis,and the specific heterosis effect, one pair ofcomplementary populations, Serere Composite2(M) x Nigerian Composite, were identified forinclusion in this project. These two populationswill be improved simultaneously by reciprocalrecurrent selection using an inbred tester drawnfrom each population.

Comparison of PopulationImprovement Methods

This project was started in 1976 with the objec-

53

tive of comparing the selection efficiency ofdifferent recurrent selection methods. Thesemethods are S2 progeny selection (S2PS), full-sibprogeny selection (FSPS), recurrent restrictedphenotypic selection (RRPS), and gridded massselection (GMS). One good composite havingwide genetic variability for important characters,the World Composite, was chosen for this study.The first cycle of FSPS, RRPS, and GMS wasplanted in the 1976 rainy season (July-Oct) whilefor the S2PS system half-sibs and S1's weregrown during rainy season 1976 and the hot dryseason (Feb-May) of 1977, respectively. Firstcycle of S2PS and second cycles of FSPS, RRPS,and GMS will be grown in rainy season 1977.These procedures will be compared initiallyduring 1978, but the final comparison will bemade only after completing three cycles of S2PS

and the six cycles of the other recurrent selectionmethods.

Variety Crosses and Synthetics

This project aims to produce improved inbred orpartially inbred lines derived by crossing com-plementary inbreds or varieties and makingpedigree selections for two or three generationstill a degree of uniformity is attained. The lines soproduced are tested for use in synthetics, aspotential hybrid parents, and for performanceper se. They fill an important function in readilysupplying a range of useful clear-cut variabilityto cooperating breeders.

The pedigree lines produced are also reviewedfor further intercrossing. A computer program,using data from a standard set of descriptors

Figure 19. Pearl millet progenies derived from recurrent selection.

54

taken on potential parents, is being developed toensure that the best crosses will be made.

In rainy season 1976, 525 new F1's and 1000F2's were grown. The best selections from thesewere advanced in the 1977 hot dry season nurserywhere 266 superior F3's and 126 F4's wereidentified for the 1977 Uniform Nurseries. Theremaining 1976 selections wil l be grown in rainyseason 1977.

From more advanced lines, 48 progenies-mostly of African x Indian parentage - werechosen after intense selection for tillering (eitherone or many synchronous tillers), disease re-sistance, seed setting, and yield potential forentry into 1977 replicated yield trials; these lines,in nursery form, have also been supplied to 30breeders in the SAT.

Ninety inbreds were tested in two yield trials atseveral locations in 1976. In both trials, a fewinbreds gave yields equal to the hybrid checkPHB 14. In Trial 1, grown at Bambey, Senegal,and at three Indian locations, three inbredsshowed low levels of downy mildew incidence,besides yields equal to the check (Table 14). If thegeneral relationship between inbred and hybridyield levels holds true, then good inbred lines per

se can be expected to produce better hybrids orsynthetics. High inbred yield levels are alsodesirable in hybrid seed parents.

Three synthetics were generated from goodinbreds and tested in IPMAT-2 (see Inter-national Cooperation) in 1976. One of these(SYN 7601, composed of six lines) showed somepotential, but wil l require some refinementbefore release because it exhibited more pheno-typic variability than is acceptable.

Hybrids

Three systems of cytoplasmic male sterility arebeing used to produce high-yielding disease-resistant hybrids.

Potential hybrid parents are developed in boththe Variety Cross and Population Improvementprojects and pass through three stages. Initiallytest-crosses are evaluated in single plots atICRISAT Center, where most have to be discard-ed because of incomplete pollen shedding.Agronomically satisfactory fertile crosses moveto the second stage —a replicated trial at severallocations, including the DM nursery. Thosesurviving reach the third stage of wider nationaland international testing (Fig 20).

Table 14. Grain yields and downy mildew incidence of sele cted inbreds in P M I T 1,1976.

Grain yieldIncrease

DM incidence

ICRISAT Bambey,Centera Senegal Meanb

abovePHB-14

ICRISATCenter

Bambey,Senegal

(kg/ha)

21902507185316571619

±262735

(%) (%) (%)

Souna D2 x Ex-Bornu-161T 166-2 x 700594700638 xSC3(M)-17-1-2PHB 14c

Mean (46 entries)

L.S.D. (0.05)

36753184364734752424

±3961110

(kg/ha)

21902507185316571619

±262735

26492344226422201671

±330752

1962-

- 2 5

1.61.41.42.5

NAd

0.011.3

1.313.7N A

aThe plots at ICRISAT Center had received 80 kg P2O5 and 150 kg N per ha.bMean yield over four locations.c Highest-yielding check.dNot averaged.

55

Figure 20. Flow of genetic material through the ICRISAT pearl millet hybrid program.

Parents found to be good maintainers aredeveloped into A/B pairs. Crosses have beenmade between existing B-lines and between B-lines and sources of DM resistance to generatenew seed parents. F3's from these crosses havebeen selected for testing in India and Africa in1977. Similarly, good R lines have been in-tercrossed to generate new pollen parents.

In 1976, 5054A, 5141A, 111A, MS75A,239D2A, and 67A were used as seed parents inproducing hybrids tested in three third-stagehybrid trials (PMHT 1, 2, and 3, divided on a height basis) and one second-stage trial (PMHT4) conducted in rainy season 1976. Restorer linesoriginated variously from India, Africa,ICRISAT composite progenies, and newly de-veloped progenies from Indian x Africancrosses.

Figure 21. Developing new pearl millet hybrids at ICRISAT Center.

56

Six hybrids in PMHT 1 yielded on averagemore than the best check, BK 560-230 (2420kg/ha) (Table 15). Acceptable levels of resistanceto downy mi ldew-bo th in India and at Kano,Nigeria - were associated with high-yield poten-tial over environments in two hybrids - ICH 118(2847 kg/ha) and ICH 117 (2540 kg/ha). Inhybrid trial 2 (PMHT 2), four hybrids ( ICH 128,ICH 129, ICH 126 and ICH 143) produced morethan the yield of BK 560-230 (2702 kg/ha). Therestorers of ICH 128 and ICH 129 were exoticand, in the case of ICH 129, the male parent wasa population received from Serere, Uganda.PHB 12, from Ludhiana, India, gave the highestyield (2880 kg/ha) in PMHT 3; this was 15percent more than the trial mean yield, and wasclosely followed by BK 560-230 (2 857 kg/ha) and

ICH 162 (2735 kg/ha). In PMHT 4, two pro-genies (EC 298-2 and EC 298-3) from the Pop-ulation Improvement project produced desirablehybrids in combination with 5141A and 111A.

Seed parents 67A, 111 A, and 5054A all pro-duced high-yielding hybrids and those with 111Awere less susceptible to downy mildew. In termsof yield alone, equally good hybrids can be madeusing as pollinators inbred lines, varieties, com-posite progenies, or populations. Tests in Indiaclassed most hybrids in trial PMHT 1 as DMresistant, but the trial at Kano revealed that -except for ICH 118, ICH 117, PHB 14, and ICH119 - this resistance was location specific (Indiansites only). A l l hybrids resistant at Kano had111A as the female parent while all the pollenparents were of African origin.

Table 15. Grain yields and downy mildew incidence of sel ected hybrids in P M H T 1 (1976).

Grain yield DM incidence

IncreaseICRISAT

CenterICRISAT Kano, above Disease Kano,

Hybird Pedigree Centera Nigeria Meanb meanc Nursery Nigeria

(kg/ha) (%) <%) (%)

ICH 110 5054A x SC3(M)-1 3 803 1305 2895 32 2.2 87.8ICH 118 111A x Souna B 3480 2416 2847 30 9.2 2.3ICH 117 111A x 1 / 2 H K 3271 1875 2540 16 5.3 23.8ICH 116 111A x WC 3-7 3073 1903 2531 15 9.7 60.5ICH 115 111lA x 700787 3002 1514 2528 15 5.4 34.9

ICH 114 5054A x 13092 3002 1153 2480 13 2.3 75.7BK 560-230 5141A x K560-D230 3 294 1055 2420 10 2.6 73.0PHB 14 111A x PIB 228 2701 1305 2387 9 4.7 31.3ICH 119 111A x R 2591-30 2783 1264 2335 6 4.7 15.4Mean

(25 hybrids) 2789 984 2193 - 7.5 71.1±274 ±217 ±312

L.S.D. (0.05) 771 610 704

a Received 80 kg P2O5 and 150 kg N per ha.b Mean yield over all test environments.c Increase above mean yield over all locations.

57

Table 16. Performance of the five highest-yielding entr ies over 30 locations in IPMAT-2,1976.

Time to 50% GrainEntry Pedigree flowering Heighta yield Increaseb

(days) (cm) (kg/ha) (%)

ICH 105 5054A x B 282 51 164 2389 27ICH 13 5071A x J 1188 54 189 2 345 25BK 560-230 5141A x K560-D230 50 173 2281 22ICH 108 5054A x Serere 2A-9 47 168 2178 16I C H 107 111A x SC2(M) 50 186 2113 13Local check (average) 51 178 1934 3

Mean over locations 52 183 1876 -

aMean of 26 locations.b Above mean yield over locations.

Table 17. Grain yields of the three ICRISATentries and the highest-yielding check,over 32 locations, in the 30-entryA ICMIP Advanced Yield Trial I I I ,1976.

Entry Grain yield

BK 560~230a

ICH 105ICH 106ICH 107

Mean of 30 entries

(kg/ha)

22702240206017801990

aCheck, ranked first.

and at Maradi, Niger. This type of reactionindicates a location-specific resistance of poten-tially limited usefulness.

The All-India Coordinated Millet Improve-ment Project (AICMIP) tested three ICRISAThybrids during 1976 in its Advanced Yield TrialI I I for limited-moisture conditions. ICH 105(5054A x B 282) ranked second in yield (2 240kg/ha) averaged over trials conducted at 32locations in ten states of India (Table 17). Thishybrid is being reevaluated in 1977 by theA I C M I P in advanced trials, while a further threehybrids ( ICH 110-5054A x SC3(M)-1, ICH 118-111A x Souna B, and ICH 190-111A x EC 298)were accepted into initial trials.

A further 2659 test crosses were initiallyevaluated at ICRISAT Center in rainy season1976. Of these, 109 involving 111A, 5054A, and5141A were retained for second-stage testing.

In ternat iona l Coopera t ion

Network of Breeding Centers

Five millet breeders are now on station inICRISAT's African p rog ram- in Senegal, Up-per Volta, Niger, Nigeria, and Sudan. In India,ICRISAT breeders also operate cooperativebreeding nurseries in the north (at Hissar) and

The four ICRISAT hybrids tested in 1976 inIPMAT-2 all performed well (in comparisonwith commercial hybrid BK 560) at most loca-tions (Table 16). ICH 105 and ICH 13 produced27 and 25 percent more than the mean yield over30 locations. A l l entries in Table 16 were rela-tively resistant to downy mildew in India, butvery susceptible in certain West African loca-tions. I C H 105, for example, had the lowest meanDM severity score over all Indian locations butwas extremely susceptible at Samaru, Nigeria

58

the south (at Bhavanisagar) and work withscientists in the A ICMIP. Between India andWest Africa, the main flow of early generationbreeding material is between ICRISAT andKamboinse in Upper Volta. Other researchprograms in Central and West Africa will drawmainly on Kamboinse for breeding material, butwill participate in cooperative multilocationalyield trials. ICRISAT and Indian breeders arenow cooperating in coordinated composite pro-geny trials, enabling the breeders to select goodprogeny in situ, while contributing to the overallimprovement of the composite.

International Trials

The Second International Pearl Millet Adap-tation Trial ( IPMAT-2) of 21 entries comprisingexperimental and commercial hybrids, varieties,and populations from several origins was sentout to 50 locations in 19 countries in 1976.Results were received from 32 locations (ofwhich 30 could be used) in 11 countries. Theaverage grain yield for the trial over theselocations was 1 876 kg/ha, highest (4221 kg/ha)at Serere, Uganda and lowest (804 kg/ha) atAnand, India (Table 18). At only 10 locations didthe local check or a commercial hybrid give thehighest yield. Among the hybrids, ICH 105, ICH13, and BK 560-230 gave 27, 25 and 21 percentmore yield over the grand mean, while WC-C75was the best of the seven experimental varietiesand populations. The mean yield of the localcheck was near to the grand mean yield overlocations.

Under the high DM pressure in West Africa,Ex-Bornu, WC-C75, NC-S75, and PHB 14 werethe least susceptible. Genotype x environment in-teraction was strong for many entries but ICH13, ICH 107, PHB 14, and WC-C75 combinedhigh yield averages with good stability of yield.

Three hybrids and one experimental varietywere entered in the 1976 A I C M I P National Yieldtrials conducted over 32 and 16 locations inIndia, respectively. ICH 105 came second overall(2240 against 2270 kg/ha) in the hybrid trialwhile WC-C75 came fifth (1 490 against 1 530kg/ha) in the varieties/composites trial, but gave

Table 18. Mean yield performance of IPMAT-2entries over 30 locations in the SAT,1976.

Increase/decreaseover mean yield

Entry Grain yield over locations

(kg/ha) (%)

ICH 105 2389 + 27.34ICH 13 2345 + 25.00BK 560 2281 + 21.58ICH 108 2178 +16.09ICH 107 2113 + 12.63

PHB 14 2084 + 11.08WC-C75 2074 + 10.55Ex-Bornu 2055 + 9.54Syn. 7601 1980 + 5.54MC-S75 1945 + 3.67

Local check 1934 + 3.09LC-S75 1874 - 0.10NC-S75 1866 - 0.53SC13-H75 1861 - 0.80PSB3 1787 - 4.74

Syn. 7602 1759 - 6.23Mc-C0

ICI 2661758 - 6.28Mc-C0

ICI 266 1505 -19.77N H B 3 1390 -25.90ICI 7540 1111 -40.70Syn. 7603 1087 -42.05

Mean (30locations) 1876

the highest fodder yield. Both were selected foradvanced testing.

Distribution of Seed Material

During 1976-1977,5 153 seed lots of germplasm-breeding and source material consisting ofinbreds, restorers, hybrids, synthetics, pop-ulations, population progenies, and experimen-tal varieties-were distributed to breeders in 18countries.

59

Information Dissemination

ICRISAT's Cereal Improvement program pub-lishes and distributes Semi-Arid Cereals, a news-letter devoted to topical research results withcontributions from ICRISAT and cooperators.Semi-Arid Cereals is published in French and inEnglish, and is intended to be a worldwide forumfor all cereals scientists. Apart from this the PearlMillet Improvement program publishes bothformal and informal research reports. The l i -brary is preparing to publish a bibliography ofPennisetum americanum literature.

Physiology

Development and Anatomy

Projects on the growth and development of thepearl millet plant, on the anatomy of pearl millet,and on the morphological and anatomical de-scription of the development of the millet paniclehave been completed. Illustrated reports of thefirst two projects are being prepared and the

results of the third wil l be published in theproceedings of the First Annual Symposium onthe Physiology of Sexual Reproduction inFlowering Plants, held at Ludhiana in 1976.

The report on millet growth and developmentincludes sections on seed morphology, growthstages, and growth patterns of the leaf, stem,root, panicle, and grain. The life cycle of themillet plant consists of nine easily identifiedstages of development, as listed in Table 19. Timerequired for the plant to complete the stages mayvary with cultivar; in Table 19 the Indian hybridHB 3 is compared with the West African land-race cultivar Mel Zengo.

The report on pearl millet anatomy includesillustrations and detailed descriptions of the leaf,stem, root, panicle and grain anatomy in a number of genotypes. Leaf cross-sections (Fig22) show the characteristic Krantz tissue as-sociated with the C4 photosynthetic system.Vascular bundles are surrounded by large sheathcells which contain a high concentration ofchloroplasts (dark staining material arranged insemilunar pattern towards the periphery of the

Table 19. Growth stages in the pearl millet plant.

Growthstage

Majorgrowthperiod

Approximate timeafter emergence

Growthstage

Majorgrowthperiod Identifying characteristic HB 3 Mel Zengo

(days) (days)

01

GS1

''Emergence. Coleoptile visible at soil surfaceCollar of third leaf visible

06

06

2 " Collar of fifth leaf visible 14 1534

GS2 Growing point differentiationFinal leaf visible

22 2843

5 " Boot stage. Head extended into flag leafsheath

35 47

67

GS3

"Half bloomSoft dough

4054

5363

89

Hard doughPhysiological maturity

5865

6975

60

Figure 22. Transverse section of the leaf lamina of pearl millet hybrid HB 3 showing the characteristic bundle sheath cells of the C4 plant.

cells). Surrounding these are the smaller me-sophyll cells, also containing chloroplasts. Mar-ked cultivar differences in the concentration ofchloroplasts in the bundle sheath cells werenoted.

Effects of Drought-screening Treatmentson Crop Growth and Yield

In order to learn more about the effects ofartificial stress treatments used in screening fordrought resistance, detailed measurements ofcrop growth and yield were carried out onirrigated and on stress treatments during the hotdry season. The stress treatments used were a mid-season stress (from 30 to 64 days afteremergence) and a terminal stress beginning atflowering (48 days after emergence).

Crop duration was shortened by 10 days andthe total crop growth reduced by about 30percent in the terminal stress treatment (Fig 23).The mid-season stress caused a marked re-duction in crop growth during the treatmentperiod, as well as an apparent depressive effecton growth rate following termination of stress.Total dry weight and maximum leaf area indexachieved in this treatment were only approxi-mately 60 percent of those in the control, andcrop maturity was delayed more than 20 days.

Grain yields, similar in both stress treatments,were approximately 45 percent of those in thecontrol (Table 20). The way in which yieldreduction occurred, however, differed betweenthe two stress treatments as different yield com-ponents were affected by the different timing ofthe two treatments.

Yield reduction in the terminal stress treat-ment occurred because of the failure of the latetillers to develop (compare the differences inpanicle numbers between this treatment and thecontrol at 64 days and at maturity, Table 20) andbecause of a reduction in size of the grains filledunder stress. Yield reduction in the mid-seasonstress treatment was due entirely to a reduction inthe productivity of the average panicle. Tillerswhose development was interrupted during thestress resumed growth after the stress was termi-nated (again compare panicle numbers at 64 days

Figure 23. Effects of stress treatment on crop dry-matter accumulation. Arrows in-dicate time of application of drought stress. The mid-season treatment was relieved on Day 65; the terminal stress treatment was not relieved, as the crop had senesced by Day 80.

61

Days after seedl ing emergence

40 80 120

Drought period

L.S.D.(0.05)

600

400

200

Drought-period

seasonstress

Mid-

Terminal stress

Control

and at maturity), but the grains produced perpanicle were significantly smaller and fewer innumber.

The data suggest that resistance to a mid-season stress might be related to ability topromptly resume growth and produce near-normal panicles after termination of stress. Re-sistance to terminal stress, on the other hand,might be improved by synchronization of tillerdevelopment and by maximum use of resourcesfor grain-filling under stress conditions.

Screening Breeding Material forDrought Resistance

A two-stage screen is used to measure, in terms ofyield reduction, the response to drought stress inmaterial drawn from germplasm and the breed-ing program.

The first stage consists of a nonstress controlcompared to a single 30-day stress applied atpanicle initiation (about 21 days after planting).These are applied in different parts of the fieldwithout replications, though there are regularplots of check varieties. The second stage is a replicated experiment with four treatments - twoas above, one with a terminal stress from flower-ing onwards, and one with a short stress atpanicle initiation followed by terminal stressfrom flowering.

The 1976 drought work involved the initial

stage only. Avoidant types, which producedgrain yields of 1 500 to 2 000 kg/ha in stress (fromone-half to three-fourths that of their counter-parts in control), were identified. Tolerant en-tries, mainly by means of poststress recovery(tiller production), gave yields equivalent tononstress yields (2000 to 2500 kg/ha).

In 1977, the period allowed for poststressrecovery was reduced, which reduced yields, sothat both the best tolerant and avoidant typesproduced 1 000 to 1 500 kg/ha under stress, whichwas 50 to 60 percent of the yields when grownwithout stress.

The advanced screen introduced in 1977showed that the terminal stress applied at flower-ing (47 days after planting) was critical in termsof grain production. Most entries produced only600 to 900 kg/ha in the two treatments involvingthis stress, 20 to 40 percent of that of nonstressedcontrols.

Initially, within the drought-resistance pro-ject, crosses are being made between resistantlines and also between breeders' advanced prog-enies from other projects. The first set of crosseswere screened as F2 populations in the 1977nursery and single-plant selections were made inthe stress treatment. A group of lines, some ofwhich were selected in 1976 and again in 1977,wil l be evaluated in rainy season 1978 for poss-ible incorporation into either avoidant or tol-erant composite populations.

62

Table 20. Effects of drought-screening stress on yield and yield components.

Panicle number

Seeds perpanicle

MaximumGrain leaf area

Treatment yield index

At 64 daysafter

emergenceAt

harvestSeeds per

panicle100-seedweight

(kg/ha)

Control 2115 3.20Mid-season drought 1 196 2.00Terminal drought 1097 2.60

L.S.D. (0.05) 176 0.7

(no/m2)

2514242

(no/m2)

414424

5

(no)

865589989106

(g)

0.640.480.490.04

Relationship Among Plant Type,Population, and Planting Pattern

Farmers' cultivars and landraces of millet in-clude a wide variety of plant types and are grownin a wide range of planting geometries and plantpopulations. We have initiated studies to de-termine if there are interactions among planttype and planting system; if so, they might beimportant in variety selection and testing. In onesuch study three cultivars, of contrasting heightand tillering habit, were grown at a range of plantpopulations from 3 to 25 plants per square meterin wide rows, narrow rows, and with all plantsequally spaced.

Population seemed to have no effect on grainyields, as all genotypes demonstrated a remark-able ability to adjust to changing crop com-petition. Individual plant yields increased sixfoldas population declined from 25 plants/m2 to 3 plants/m2. The yield per plant was unaffected byplanting pattern at high populations, but was lessin the equidistant planting at the low pop-ulations. These results suggest that populationper se, over a broad range, is not a determinant ofmillet yields, at least where fertility is adequate.

Planting pattern, in contrast, had a highlysignificant effect on yield, which was independentof both genotype and of plant population. Grainyields were significantly different in the order ofwide rows > narrow rows > equidistant plant-ing (Fig 24). The yield differences were due toincreases in dry weight of individual tillers, andto leaf area, grain number, and grain yield (tillernumbers per unit area were constant), emphasiz-ing the role of tiller productivity in maximizingyield.

These studies are now being expanded toinclude hill-planting systems, various fertilitylevels, and a wider range of cultivars.

Millet Genotype Evaluation

Analysis was completed on the large-scale ex-periment of 1976 in which relationships betweenyield, yield components, and length of the majorgrowth stages were evaluated. Results were ana-lyzed separately for the entire set of 50 gen-otypes, as well as for smaller subsets representing

the high-tillering Indian hybrid type and thelower-tillering West African type. Differences ingrain yield between the two subgroups were notstatistically significant, but there were significantdifferences in tiller number per plant, seed num-ber per plant, and harvest index (Indian typeswere superior); and in grain number per panicleand length of the GS2 stage (West African typeswere superior).

The relative contributions of different yieldcomponents in each group were assessed in a multiple linear regression model of yield onpanicle number per plant, grain number perpanicle, and grain size (100-seed weight). Over theentire set, seed number and grain size account-ed fdr 45 and 42 percent, respectively, of thevariation in yield, with panicle number account-

Figure 24. Grain yield of three pearl millet cul-tivars as affected by geometry of planting. Mean data from a series of five planting populations ranging from 3 to 25 plants/m2.

63

Seeding configuration

37.5-cmrow

75-cmrow

Equidistant

L.S.D.(0.05)

cv K 599

cv HB 3

cv 700250

3 500

3 000

2 500

2000

1500

ing for only 9 percent (Table 21). In the high-tillering group seed number per panicle was thepredominant variable determining yield, ac-counting for nearly 70 percent of the observedyield differences. The magnitude of this termindicates that selection for panicle size within thehigh-tillering types should be effective in increas-ing yields. In the low-tillering group, the numberof panicles per plant was the predominant vari-able, but there were also significant contributionsfrom seed number per panicle and seed weight.Thus yield increases in these types should bepossible by improvements in both panicle num-ber and panicle productivity.

The analysis of the relationship of yield to thelengths of the various growth stages confirmedthe hypothesis that an increase in the length ofthe GS2 stage (panicle initiation to flowering) upto an optimum of 25 to 30 days resulted inmaximum yields. The length of the grain-fillingphase, in contrast, varied only slightly among thegenotypes and was not strongly related to yielddifferences. •

Entomology

At ICRISAT Center in 1976-1977, pest damageto pearl millet was minimal. Detailed obser-vations were carried out on blocks of the cropsown in nonsprayed areas; although 63 pestspecies were recorded, none were damaging. The

potentially important pests, Atherigona approxi-mata, Chilo partellus, and Calocoris angustatus were at low levels.

The root aphid, Rhopalosiphum rufiabdo-minalis Fitch, caused some wilting in breeders'plots sown on Alfisols. The ant, Monomorium indicum Forel, removed seed from plot rows andpiled it near nests, causing low plant populationson breeders' and agronomy plots as well.

A list of insects found on pearl millet wasprepared for publication.

Pathology

The fungal diseases-downy mildew (caused bySclerospora graminicola), ergot (caused by Clavi-ceps fusiformis), and smut (caused by Tolypos-porium penicillariae) -cause severe damage andyield reduction in many of the new high-yield-potential pearl millet cultivars and hybrids.Activities in ICRISAT's pearl millet pathologyprogram are directed to the control of thesediseases through the identification and util-ization of stable host-plant resistance. During1976, screening techniques were evaluated andsignificant progress has been made in finding andusing resistance to downy mildew.

Downy Mildew

The traditional screening method of planting test

Table 21. Relative contribution of yield components (all variables in logarithmic form) from the modelY=b1x1 + b2x2 + b3x3.

Fraction of the total sum of squares (R2)

VariableA l l

genotypesaLow-tillering

subsetbHigh-tillering

subsetb

Panicle/plantSeeds/panicle100-seed weight

0.0890.4480.417

0.3950.2470.202

0.1180.6880.176

aTotal of 50 in the complete set.bTotal of 17 in sublet.

64

materials in a sick-plot (i.e. a field in which thesoil contains resting spores of the DM causalfungus) is unreliable and restrictive. A newtechnique was required which (i) would alloweven exposure to inoculation of many thousandsof segregating lines, (ii) would allow flexibility insize and location of the testing site, and(iii) could be used efficiently in the postrainyseason as well as in the rainy season. Evidencefrom two field experiments during the rainyseason clearly indicated that the sporangia(ephemeral asexual spores produced in millionson infected plants each night) played a major rolein the early infection of plants and buildup of thedisease (development of the epidemic) in the fieldwhen high humidities prevailed. This infor-mation was used to develop a screening tech-nique based on sporangia produced by "infectorrows" (a known high-susceptible hybrid) plantedin every third row 15 days before seeding of thetest rows. Inoculum is provided for initiatinginfection in the infector rows by strategicallyplaced pots of infected plants, and the necessaryhumidity is provided by fine-mist irrigation in theevening hours. The test materials are plantedwhen the infector rows show rapid disease de-velopment (15 to 21 days after seeding of theinfector rows) and the mist irrigation is con-tinued twice weekly until the test materials reachheading stage. Throughout the screening area,after every 10 test lines, a known high-susceptible

hybrid (indicator) is planted at the same time asthe test lines in order to measure the success ofthe inoculation process. This technique washighly effective during both the rainy and post-rainy seasons in 1976, and details of the screeningresults are presented below and in the breedingsubprogram report.

Screening Activities at ICRISAT Center

Using the infector-row system, 4 hectares ofbreeding material were screened at ICRISATCenter for resistance to DM in both the rainyand postrainy seasons. This was a joint activitybetween the breeding and pathology staffs. In thepopulation-improvement materials, selfing andsibbing for the next cycle of selection was carriedout in the disease nursery utilizing only D M -resistant plants. A measure of the advanceachieved in this process is presented in Table 22.

In addition to the ICRISAT breeding mat-erials, 230 millet lines were tested in the diseasenursery for scientists working in the All-IndiaCoordinated Millet Improvement Project.

Multilocational Screening

The International Pearl Millet Downy MildewNursery ( IPMDMN). One of the features ofcertain types of resistance is that it "breaksdown" through the ability of the pathogen toovercome the resistance. In an attempt to find

Table 22. Comparison of downy mildew susceptibility in four populations after one generation of selectionunder heavy disease pressure during rainy and postrainy seasons at ICRISAT Center, 1976.

Lines with < 5 % incidence Maximum incidence

Population Rainy' Postrainy Rainy Postrainy

G A M 73 S2'sG A M 73 FSG A M 75 S1'sLate Composite

(%)

3.1. 9.0

0.033.3

(%)

50.016.79.5

68.4

(%)

76.5969.3587.7559.57

(%)

50.041.469.731.0

aResistant plants were selfed in rainy season 1976 and the progeny retested during the postrainy season.

65

stable resistance, a multilocational testing pro-gram was initiated in 1976 in which millet linesresistant at ICRISAT Center were exposed tomany populations of the downy mildew path-ogen. The 1976 I P M D M N , the first diseasenursery in the International Pearl Millet DiseaseResistance Testing Program (IPMDRTP), con-tained 46 entries and was tested by cooperatorsat 24 locations in five countries in Asia and WestAfrica. Ten entries, all of which originated inWest Africa, showed an excellent level of across-location resistance. This is encouraging, forwhile many other cultivars were resistant atIndian locations they were highly susceptible atsome of the West African locations (Table 23). Itis not clear whether these differences in reactionare due to different pathotypes, different en-vironments, or b o t h - b u t it is clear that somemillet lines have more stable resistance thanothers, and that the West African locations areessential to our program for the identification ofstable downy mildew resistance.

Seed-transmission Investigations

The activities of the ICRISAT Millet Improve-ment program require movement of millet seedacross national boundaries and between con-tinents. This movement must not facilitate move-

ment of pathogens. In view of recent reports ofpossible seed transmission of the downy mildeworganism, we continued an intensive investi-gation of possible seed transmission of pearlmillet DM during 1976. Oospores of the causalfungus were found to adhere to external surfacesof pearl millet seed, but in this position they arevulnerable to simple seed-treatment procedures.In order to demonstrate internal seed trans-mission of the disease, infected plants must beproduced from suspect seed in an environmentfree from all external inoculum sources andconducive for symptom development. During1976, we g rew-on sterilized agar medium inlarge sterile boiling tubes-985 plants fromsurface-sterilized seed harvested from infectedheads, and no downy mildew symptoms weredetected. Young seedlings growing in this wayand inoculated with sporangia developed symp-toms in 6 days, indicating that the environmentwas conducive for symptom development. Inseveral such studies to date we have obtained noevidence of internal seed transmission of pearlmillet downy mildew from thoroughly driedmature seed that was surface-sterilized prior toplanting in inoculum-free environments. We arecontinuing to look at the effects of drying andvarious other seed treatments on possible seedtransmission.

Table 23. Comparison of downy mildew incidence in five peal millet cultivarAfrican locations.

India West Africa

Entry HissarICRISAT

Centera Samara Ouahigouaya

Syn 7601I C H 105Ex-BornuWC-C75PHB 14

(%)

0.70.20.00.60.0

(%)

1.21.52.21.7

14.5

(%)

688418174

(%)

181521

14

aIn the DM scneening nursery.

66

Ergot

Less progress has been made with ergot thanwith downy mildew. During 1976, intensivestudies were made on factors influencing in-fection, and the information from these studieswas used to develop an effective screening tech-nique. A limited amount of screening for re-sistance was carried out and the validity of theresults will be rechecked during 1977.

Factors Influencing Infection

Inoculation of heads when the female parts of theflowers are most receptive (protogyny stage)produces maximum infection. Freshly producedspores from the "honey-dew" of infected plantsare more infective than spores of dried "honey-dew" and both produce more infection thanspores contained in dried resting bodies (sclero-tia). Experiments were conducted with dif-ferent inoculum concentrations, different meth-ods of inoculum application, and use of selfingbags to enclose the head after treatment. Thegreatest infection takes place when heads areinoculated by dipping during the protogynystage with spores from fresh honey-dew at a concentration of 100 x 103, and when heads arebagged immediately after inoculation. Inocul-ation in the late evening hours is also beneficialfor increased infection.

Effects of Pollination and the Path of Infection

Heads inoculated after pollen shedding are lesssusceptible than heads inoculated before thisevent. This led to the hypothesis that pollinationand fertilization inhibited infection. We inocu-lated with male-sterile lines (i.e. lines whichproduce no viable pollen) at different stages inthe flowering process and found that the rapidfall in susceptibility occurred in the same way aswith male-fertile heads, with the withering of thestigmas (female parts of the flower). This in-dicates that the infection path is likely to bethrough the stigmas. Some evidence to confirmthis has been obtained from microscopic exam-ination of florets at varying times followinginoculation, and further experiments are plannedwith controlled pollination and inoculation ofmale-sterile heads.

Resistance Screening

About 4000 breeding lines were dip-inoculated(without bagging) during August 1976. Eighty-two lines developed no ergot; these will need tobe thoroughly rechecked during 1977, using theprocedures for optimum inoculation describedabove. A total of 488 millet lines selected inpreliminary screening in 1975 were retested dur-ing 1976 by the dipping and bagging technique.Twenty-two lines developed only slight infection,and these will be tested at several locations in1977.

Smut

Examination of Screening Techniques

It has been considerably more difficult to find aneffective technique for screening pearl millet linesfor smut resistance at ICRISAT Center than itwas for downy mildew and ergot.

During the hot dry summer season of 1976, a reasonably high level of infection was obtainedby injecting a suspension of smut spore-balls intothe top of the boot (immediately prior to heademergence). However this method did not pro-duce good infection at ICRISAT Center duringthe rainy season. Intensive study is continuing onfactors possibly affecting infection. Unti l a reli-able inoculation technique is developed, littleprogress can be made in resistance identificationsat ICRISAT Center.

Rust

Rust, caused by the fungus Puccinia penniseti, isa leaf disease with great destructive potential. A l limproved millets should carry a sufficient level ofstable resistance to this disease if they are to beutilized successfully over a long period. Duringrainy season 1976, 193 lines selected in earlierscreening at ICRISAT were tested at Bhavani-sagar in Southern India where rust normallyoccurs in severe form on susceptible millet.Under these conditions, 10 of the entries de-veloped no rust and 64 entries showed less than 5 percent disease based on a standard (Cobb'sModified) rating scale.

67

During postrainy season 1976-1977, the dis­

ease appeared in severe form in germplasm

material at ICRISAT Center. A total of 128

single-plant selections were made. These selec­

tions and those found free of rust at Bhavani-

sagar will be tested in 1977, in a Preliminary Pearl

Millet Rust Nursery (PPMRN), at several loc­

ations where the disease is known to occur.

N u t r i t i o n a l Q u a l i t y

Standardization of Laboratory Methods

The Technicon Auto Analyzer was found to be

the easiest of several methods of protein esti­

mation tried and gave the best relationship

(r = 0.99) with micro-Kjeldahl values over a

wide range of protein levels. During the year

4 600 samples were analyzed for protein content;

they ranged from 5.8 to 20.9 percent, with a mean

value of 10.6 percent.

For lysine estimation, samples were weighed

to contain constant amount of protein (80 mg)

and analyzed by DBC (dye-binding capacity)

method. A good correlation was obtained

(r = 0.92) between DBC values and actual lysine

concentration. Investigations are in progress to

relate the DBC values obtained on constant

weight of sample and protein values with lysine

concentration in the sample. Lysine content was

estimated in 1400 samples by the DBC method

during the year.

Analysis of Breeding M a t e r i a l

Initial analyses for protein content indicated that

there was a problem of variability within each

genotype, with coefficients of variation ranging

from 12 to 23 percent, and that environment

greatly affected protein level. We therefore se­

lected a number of high- and low-protein lines

and planted a replicated trial in three different

environments in 1976, in order to determine the

stability of protein content. Results indicate

protein content was higher under conditions of

lower productivity. Environment seems to play a

major role in influencing protein content.

Among locations, protein content was highest at

Hissar and lowest at Bhavanisagar. This may

relate to differences in grain-filling periods.

Progenies in several composites showed a wide

range in protein content from 8 to 15 percent (in

one case 19%), indicating possibilities for im­

provement by selection. From analyses on the

same cultivar at different yield levels, the re­

lationship between yield and protein content was

not as strongly negative as expected. Content of

11 percent protein at yields of 3000 kg/ha would

seem feasible. Using the working collection, a

slight negative correlation (r = 0.23**) was

found between oil and protein content, but no

relationship was indicated between 100-seed

weight (0.24 to 1.4 g) and protein content (6.0 to

14.5%).

M i c r o b i o l o g y

One hundred and sixteen lines and landrace

cultivars of pearl millet were grown on Alfisol

at ICRISAT Center under low-fertility con­

ditions (only 9 kg P/ha fertilizer added) over

three seasons - rainy, postrainy, and hot dry

summer-to see if differences between lines in

ability to stimulate nitrogen fixation by bacteria

in the soil could be detected. Nitrogen-fixing

activity was estimated by an acetylene-reduction

assay using cored soil samples containing plant

roots, as was the procedure for sorghum (page

43). Preliminary findings indicate that the single

core probably represents about one-eighth to

one-half of the total activity for plants, de­

pendent on the growth stage.

About 50 lines had some nitrogenase activity

greater than 50 µ g N/core per day, and 12 had

more than 100 µ g N. There were significant

differences between lines despite the large plant-

to-plant variability. The reason for this vari­

ability is yet not clear. Most activity was found

after flowering, and continued well into the

grain-filling stage. Activities during the cooler

postrainy season were lower than in the rainy

season until the temperatures increased in

March. Four lines had consistently high activity

in more than one season. Little activity occurred

in the summer planting.

68

Using a 3:1 ratio to convert ethylene pro­

duction into amount of nitrogen fixed, the high­

est activity was obtained with the line IP 2787

with a mean fixation for five plants of 560 µ g

N/core per day, with a standard deviation of 749,

and a range of activity from 17 to 1 586 µ g

N/core per day. Other lines assayed at the same

time had activities similar to that of soil cores

without plant r o o t s - 6 µ g N/core per day. The

activity on a dry-root-weight basis for IP 2787

was also high (2.29 µ moles C2

H2

g per hour),

though the large support roots near the crown

dominate the root weight of our cores.

Fixation associated with the GAM-73 pop­

ulation assayed throughout the grain-filling

stage varied from 16 to 80 µ g N/core per day

(mean of five plants). For the growth stage at

which nitrogen fixation is most active, individual

plants ranged from 23 to 183 µ g N/core per day.

Nitrogenase activity of GAM-73 throughout the

summer season was positively correlated

(r = 0.7) with soil moisture content in the top 20

cm of an Alfisol (Fig 25).

The minor millets - Eleusine coracana, Pan-

icum spp, P. miliare, P. miliaceum, Paspalum

scrobiculatum, Eragrostis tef, Echinocloa col-

onum and Setaria italica-also had nitrogenase

activity, as did some grasses (Table 24).

Intercropping

In four previous experiments with millet and

pigeonpea at ICRISAT Center, the intercrop­

ping advantage averaged 22 percent. In this

combination, the large difference in crop ma­

turities played a major part in the positive

interaction between the two crops.

Where there is little or no temporal difference

in growth patterns, intercropping may still be

advantageous due to better "spatial" use of

resources. This aspect was probably important in

a 1976 experiment, in which four genotypes of

pearl millet in all combinations with four geno­

types of sorghum were examined. Nine of the 16

combinations gave advantages greater than 10

percent. The two highest advantages-in excess

Figure 25. Relationship between soil moisture

content in the root zone and nitro­

genase activity of pearl millet GA M 73

grown in Alfisol at ICRISAT Center,

hot dry season 1977.

69

of 30 percent -came from the combinations of

latest sorghum with earliest millet (sorghum 21

days later at harvest) and earliest sorghum with

latest millet (millet 14 days later). The com­

bination with the largest difference in height

(sorghum 104 cm taller, and also 9 days later)

gave an advantage of 24 percent.

Next year, pearl millet wil l be grown in

combination with groundnut in order to study

growth patterns, plant population/spacing,

genotype, N- and P-response curves, N fixation,

and weed competition.

Nitrogenase activity 55

50

40

30

20

10

50 70 90 110 130 150

Days after planting

Soil moisture 12

10

8

6

4

Table 24. Tropical grasses and minor millets

with high nitrogenase activity.

Nitrogenase

activity

(µg N/plant

per day)

Napier bajra 1788a

(Pennisetum purpureum x

Pennisetum americanum)

Setaria sphacelata 1353a

Pusa giant napier (P. purpureum) 780'

Pennisetum orientate 369a

Panicum antidotale 261a

Panicum maximum 196b

Panicum miliare 5 1b

Eleusine coracana Sharda 176b

Setaria italica ISE186A 376b

S. italica Arjuna 343b

Sorghum halepense 322b

Chloris gayana 85c

Cymbopogon citratus 117c

Echinocioa colonum 15c

Pennisetum oleopecuroides 267c

aTotal for three cores containing roots of one plant.

bMean of single cores from five plants.

c Mean of single cores from three plants.

L o o k i n g A h e a d i n Pearl M i l l e t

I m p r o v e m e n t

Germplasm. In collaboration with IBPGR, fur­

ther collections of germplasm are planned in

priority areas of West Africa and in Rajasthan in

India.

Breeding. The breeding program will continue

to produce new varieties, synthetics, and hybrid

parents for national and international testing,

together with nurseries and resistance sources for

breeders in national and regional programs.

The breeding program provides for the in-

trogression of desirable material-such as new

sources of disease resistance - into existing com­

posites; these are routinely used in variety

crosses.

New seed parents, particularly if they involve

the A2 and A

3 types, are now an urgent require­

ment. An increased number of crosses have

already been made and additionally the IB and

IR composite pair, designed to produce both

seed and pollen parents, will come on stream in

1977.

To further ICRISAT's role of fostering

cooperation and interchange between millet

scientists around the world, an International

Workshop on Pearl Millet is scheduled to be held

at ICRISAT in September of 1977.

Physiology. We plan to evaluate seasonal ef­

fects on the development, growth, and yield of a

standard set of millet cultivars in order to assess

how such cultivar performance is influenced by

season, and how much importance should be

placed on off-season yield testing and physiologi­

cal studies.

We are planning experiments to help us de­

termine how far we can extrapolate the results of

our off-season drought-resistance screening.

This will include testing the effects of a range of

different stress treatments within the dry season

and the establishment of a small nursery to be

grown in drought-prone areas during the rainy

season.

Work on seedling vigor in millet will be

expanded. Experiments will include comparison

of methods for evaluating seedling-growth rates

under favorable conditions, and development of

methods for screening seedlings for drought

tolerance and ability to emerge from crusted

soils.

Entomology. Observation on pest levels in

standard cultivars will continue. Material com­

ing forward from the breeding program will be

assessed for relative susceptibility to pests, using

standard cultivars as checks. Most of the en­

tomological work on pearl millet will be done in

West Africa where pests are more of a problem.

Any millet pest species that becomes a problem

anywhere in the SAT wil l be investigated.

70

Pathology. The large-scale field screening fordowny mildew resistance at ICRISAT Centerand the multilocational screening in Africa wil lcontinue with an expansion of the latter. In thisway, sources of location-nonspecific resistancewill continue to be identified and incorporatedinto the elite materials in the program. Theefficacy of seed treatment with Ridomil will beevaluated on a multilocational basis.

By appropriate breeding procedures, lines lowin susceptibility to ergot will be used to develophigher levels of resistance. The possibility ofcontrol through pollen management wil l beinvestigated.

Initial screening for smut resistance will con-tinue at Hissar and advanced screening will bemade at several locations in West Africa and atHissar. Smut resistance wil l be incorporated intothe breeding materials by appropriate breedingmethods.

With rust, investigations on variability incausal species are required. The resistance-screening work will continue with initial screen-ing at ICRISAT Center and Bhavanisagar and

advanced screening at several locations in India.The levels of susceptibility to blast in our elite

materials wil l be investigated at known hot-spots. If necessary a screening program will beorganized to identify resistance and to incor-porate the resistance into breeding materials.

Nutritional quality. Standardization of meth-ods for protein estimation will make it possible tohandle a larger volume of materials for thebreeding program. More work wil l be focused onother important areas, such as cooking qualityand digestibility.

Microbiology. In this program, the aim is todetermine if nitrogen-fixing bacteria closely as-sociated with roots of pearl millet have thepotential for reducing the amount of nitrogenfertilizer required by the crop. Techniques formeasuring the amount of nitrogen fixation as-sociated with pearl millet cultivars will be re-fined, and lines with enhanced nitrogen-fixingcapability wil l be sought.

71

T H E P U L S E S

Pigeonpea ( Cajanus cajan)

Chickpea ( Cicer arietinum)

Grain legumes are expected to play increasing roles in providing adequate protein in the diets of theunderfed of the SAT. Chickpea, third-ranking of the world's pulses, is now planted on some 10.5million hectares throughout 10 nations in the SAT. Pigeonpea plantings approach 3 million hectares.Both crops are utilized as human food, and are of vital importance to millions of persons in areaswhere crop production is erratic or otherwise limited. Both have high contents of protein - 1 8 to 24percent and more-and contain some amino acids not found in cereals. When combined with rice,sorghum, millet, or wheat, chickpea and pigeonpea provide an adequate balanced protein-calorie diet.

When compared with many of the world's major crops, the work of the scientist with chickpea andpigeonpea is just beginning. Hopefully breeders of these pulses will benefit from the experiences withother crop species and will be able to make rapid headway. The concentrated effort as initiated byICRISAT is long overdue.

ICR ISAT Goals

ICRISAT pursues two broad goals regarding pigeonpea and chickpea:Assembly, maintenance, and screening of the world's gennplasm resources, including worldwide

73

search for new strains and related species and the provision of seed from the ICRISAT collection forevaluation and use in programs of colleagues everywhere.

To increase, through breeding, the ability of genotypes to produce higher yields, optimum proteincontent and quality, and characteristics favored by the consumer.

In its management of the germplasm collections and genotypes resulting from its work, ICRISATintends to continue its cooperation with plant breeders in all national and regional pulse-improvementprograms, hoping to provide superior genetic materials for the use of others who are attempting todevelop tailor-made lines needed for conditions in a specific location. The segregating of genotypeswith a wide diversity of genetic characters, plant architecture, physiological attributes, and insect anddisease resistance is a major goal.

I C R I S A T Center

ICRISAT Center is the major research facility of the Institute. Located near the village of Patan-cheru on the highway between Hyderabad and Bombay, the facility includes two major soil typesfound in the semi-arid tropics. The red soil (Alfisol) is light and droughty; the black soils (Vertisols)have a great water-holding capacity. The availability of these two soil types provides an opportunityto conduct selection work under conditions representative of many areas of the SAT. Three distinctagricultural seasons - rainy, postrainy, and hot dry - characterize the area. The rainy season, alsoknown as monsoon or kharif, usually begins in June and runs into September; more than 80 percentof the annual rainfall occurs during these months. The postrainy season of October through January,also known as postmonsoon or rabi, is dry and cool; days are short. From February until the rainsbegin again is known as the hot dry season, with daily temperatures of between 36 and 43 C.

Chickpeas are planted in October or November and grow on residual soil moisture; only one gen-eration per year is grown. June and July are the months in which pigeonpeas are planted; they growthroughout the season and on into the postrainy season without irrigation. An additional generation ofearly maturing types is planted at ICRISAT Center in December and grown with irrigation so as toprovide an additional generation for the breeding program.

74

P i g e o n p e a

Previous reports have highlighted the. broadobjectives of the pigeonpea improvement pro-gram, and have listed our observations on theeffects of photoperiod response, rainfall distri-bution, and climatic variations, all of whichcontribute to strong local adaptation of cul-tivars. At ICRISAT Center, we are continuingour emphasis on studies of physiology of theplant, the biology of insect pests and diseases andthe plant's reactions to these, and developmentof pools of genetic diversity, all of which have notonly local but broad application.

Our selection for high-yielding locally adaptedtypes at ICRISAT Center and the croppingsystems under study may have limited appli-cation elsewhere, but are important to the regionrepresented by this location. In 1976 we added a location where early maturity types are adaptedand another location where very late types arebest. We are emphasizing the development ofbreeding populations for use by breeders in thoseregions of adaptation.

In the following paragraphs we have sum-marized what we consider to be the importantdevelopments during the year; where approp-riate we have presented the results accumulatedto date.

Germplasm Col lect ion andEvaluat ion

Germplasm resources were increased by 263accessions collected from five states in India. Weare extremely short of germplasm from Africa(Table 25), and 120 samples collected in easternKenya are still to be cleared by Plant Quarantineofficials.

Efforts to further describe the germplasmduring the year were disappointing. HeavyAugust rains on a poorly drained field contri-buted to unsatisfactory survival and growth of2981 accessions planted for recording theirmorpho-agronomic traits. A set of 1000 lines

were planted at Kanpur, U.P.; Kathalagere,Karnataka; and ICRISAT Center for determin-ing environmental effects on growth and proteincontent. With generally unfavorable weather,only 280 accessions produced seed at all threelocations. These samples have been sent to theCanadian Grain Commission, Manitoba, forrapid protein determination.

A 4-year experiment is in progress to compareperennial growth habit, yield, and regrowthpotential of 45 accessions (5 early, 20 medium,and 20 late). Three of the six replicates wereratooned. Observations of the first year's plant-ing indicate that early types behave as annuals.

Table 25. Countries of origin of pigeonpea germ-plasm at ICRISAT Center.

Country Entries

Australia 14BangladeshBrazil

186

Burma 27Colombia 5

Dominican RepublicGhana

61

GuyanaIndia

44795

Indonesia 1

Jamaica 13MadagascarMexico

12

NepalNigeria

319

Pakistan 4Peru 5Puerto Rico 40SenegalSri Lanka

1055

Thailand 2Trinidad 38USA 2Unknown 360

Total 5431

77

Available data on germplasm accessions arebeing computerized and a catalog is expected tobe issued in 1978. Even without a catalog, wehave strong demand for material (Table 26).

Elite Lines f r om theGermplasm

Lines from plants selected from the germplasmfor agronomic.value were tested in intercroppingand sole cropping at ICRISAT Center and atthree locations in sole cropping by APAU. Insole cropping, the check cultivars were highest inyield, while in intercropping the checks weregenerally low and some of the new selectionswere high. The highest-yielding selections will beretested, and seed is being increased for in-ternational testing. Pigeonpeas of medium ma-turity are usually grown in inter- or mixed-cropping, while breeding material is normallyselected while growing as a sole crop and evalu-ated as a sole crop. In view of this year's results,which are supported by some earlier studies, weare expanding not only testing but also selectionin segregating material grown as an intercrop.

Evaluat ion o f H y b r i dPopulat ions

Evaluation of germplasm lines as parents forhybrids has been a major effort in the breedingprogram. Once a number of crosses have been

made, it is important to determine which hybridsare worth advancing for selection. We experi-mented with testing of F1, F2, and F3 gen-erations from three crosses. The results indicatedthat a yield test of F2's would discriminateamong crosses for additive genetic factors foryield. We also compared parents and F2's of 21crosses. In this test, the correlation between F2

and mid-parent yields was only 0.47, indicatingthat actual F2 yields could not be accuratelypredicted on the basis of performance of parents.

Characters with high heritability can safely bepredicted by parental performance. The corre-lations between F2 and mid-parent values forseeds per pod was + 0.90 and for seed size was+ 0.80. We have concluded that routine testingof F2 populations in replicated tests will increasethe efficiency of breeding for yield by allowingthe discarding of all but the best populations.

Progress w i th Ear ly M a t u r i n gSelections

Our most advanced lines from the breedingprogram are in F6 generation, and these are earlylines of which two generations per year could begrown. Replicated yield tests in the coming yearwill provide our first valid assessment of yieldgains. In early cultivars, generally small seed sizecauses a market disadvantage. Table 27 showsthe increase in seed size achieved in selectionscomparable in maturity with three early cul-tivars. We are currently selecting for an inter-mediate seed size of 10 to 12 g/100 seeds and arecontinuing to investigate the relationship of seedsize with grain yield.

Genetic M a l e Steril i ty

Observations during the year verified that thems1 gene is a simple recessive which conferscomplete pollen abortion in the homozygousrecessive. Pollen-sterile plants are easily identi-fied by the whitish translucent appearance of theanthers. We were interested in the usefulness of

Figure 26. Pigeonpea germplasm accessions at ICRISAT Center.

78

1973 1974 1975 1976 1977

876543210

Pigeonpea

Table 26. Pigeonpea germplasm lines supplied to rese arch agencies in India and other nations dining1976-1977.

Institution Location Entries

I N D I A :Millet Research Station of Lam Guntur, A.P. 36Millet Research Station Madhira, A.P. 36Kakatiya University Warangal, A.P. 26Indian Lac Research Institute Namkum, Ranchi, Bihar 21Agricultural Research Station of Maktampur Broach, Gujarat 100

University of Agricultural Sciences Bangalore, Karnataka 1000College of Agriculture Dharwar, Karnataka 60Medium Research Station Gulbarga, Karnataka 78J.N. Krishi Vishwa Vidyalaya Jabalpur, M.P. 57Mahatma Phule Krishi Vidyapeeth Rahuri, Maharashtra 1

Punjab Agricultural University Ludhiana, Punjab 11Rajasthan College of Agriculture Udaipur, Rajasthan 300K.K. Agricultural Experimental Station Kudumaimalai, Tamil Nadu 1College of Agriculture Madurai, Tamil Nadu 2Crop Introduction & Diversification Station Ootacamund, Tamil Nadu 9

Central Soil & Water Conservation Research & Dehra Dun, U P . 15Training Institute

Indian Grassland & Fodder Research Institute Jhansi, U.P. 1100Chandra Shekhar Azad University of Kanpur, U.P. 1028

Agriculture & TechnologyB.C. Krishi Viswa Vidyalaya Kalyani, W. Bengal 1

OTHER NATIONS:Department of Agriculture Perth, Australia 1000Institute of Nuclear Agriculture Mymensingh, Bangladesh 15Secretaria de Estado de Agricultura San Cristobal,

Dominican Republic310

Crops Research Institute Bunso, Ghana 50Hebrew University of Jerusalem Rehovot, Israel 103

Tropical Agricultural Research Centre Ibaraki-ken, Japan 6Minor Irrigation Projects, FAO Nakuru, Kenya 25Institute of Agricultural Sciences Suweon, Korea 20Department of Agriculture Kuala Lumpur, Malaysia 25Section Cultures Maraicheres, Centre National Kaedi, Mauritania 51

de Recherches Agronomiques

continued

79

Table 26 continued

Institution Location Entries

Department of Agriculture Lalitpur, Nepal 8University of Ibadan Ibadan, Nigeria 84Ministry of Agriculture & Natural Resources Jos, Nigeria 13University of Philippines Laguna, Philippines 408University of Puerto Rico Mayaguez, Puerto Rico 100

Agricultural Research Station Maha Illuppallama,Sri Lanka

24

Hudeiba Research Station El-Damer, Sudan 8Masalato Christian Education Centre Dodoma, Tanzania 50Agricultural Research Station, Bangkhen Bangkok, Thailand 2

Tabk 27. Number of seeds per pod and 100-seed weight of early maturing and check cultivars ofpigeonpea.

Seeds per pod 100-seed weight

Checks & Selections Range Mean Range Mean

(no) (no) (g) (g)

PANT, A-3 (1)Selections 3.5-4.2

3.44.0 9.0-15.6

8.710.1

PRABHAT (2)Selections 3.3-5.1

3.44.1 -

6.310.7

UPAS-120 (3)Selections 3.0-5.3

3.64.1 7.8-15.4

7.711.1

the male-sterility factor in population breeding,and grew a population involving crosses of thesterile with 14 parent lines. We compared thepod-set on sterile plants (resulting from insectpollination) and found 49 percent of 3 735 taggedbuds set pods on male-sterile plants comparedwith 58 percent of 2120 buds tagged on normalfertile plants in the same population. (These budswere all tagged at the peak of flowering, and thepercent pod-set is considerably higher than theobserved pod-set over the entire floweringperiod.)

Under conditions at ICR1SAT Center it ap-pears that pollination of the steriles is sufficient

for them to be useful in population breeding. Weare expanding our studies of natural crossing onthe steriles to determine their possible usefulnessin large-scale crossing bf specific genotypes.

Rhizobium P o p u l a t i o n s i n S o i l

We are using the soil dilution-plant infectionmethod of counting cowpea-type rhizobia usingSiratro (Macroptilium atropurpureum) as the testplant. When 160 strains isolated from pigeonpeanodules were tested, most nodulated Siratro -although 14 percent were slow to do so. A small-

seeded pigeonpea, cv ICP-7332, also nodulates

well in agar deeps and we are now comparing

counts using Siratro and pigeonpea as the test

plant.

Using Siratro for counting, we found cowpea-

type rhizobia varying from less than 100 to more

than 105/g soil within 0.1 ha during the dry

summer. Numbers also varied a hundredfold

within the top 30 cm of soil with little appar­

ent relation to soil moisture-one surface-soil

sample contained 105/g but only 1.3 percent

moisture. Numbers increased tenfold after rain

in early May. The presence or absence of. a

pigeonpea crop in the previous season had little

effect on numbers. Such high populations wil l

make it difficult to introduce new strains by seed

inoculation.

Rhizobium Strains

From 160 strains of Rhizobium isolated from

soils in Andhra Pradesh and Maharashtra, 10

highly effective strains in nitrogen fixation were

selected for field-inoculation trials in Alfisols and

Vertisols. In pot trials with cv ICP-1, these

strains produced plants up to 120 percent heavier

at 6 weeks than the uninoculated control, which

received 240 ppm N in the nutrient solution. The

pattern of nodulation showed Rhizobium strain

effects as well as strain x host genotype interac­

tion effects. Some of the selected strains form

nodules containing a black pigment, and we hope

that this will be a useful marker in our in­

oculation trials. At 60 days after planting only 16

percent of the nodules in Vertisol and 12 percent

in Alfisol contained this pigment in uninoculated

soils.

Nodulation and Nitrogen

Fixation

More than twice as many and much larger

nodules were formed during the rainy season in

Alfisol than in Vertisol, but after the rains nodule

formation in Vertisol was greater, so that by 120

days after planting there were 136 nodules per

plant in Alfisol and 200 in Vertisol. However, 96

percent of the nodules in Vertisol and 74 percent

of the nodules in Alfisol were hollow, apparently

the result of insect attack. The grub damage was

apparent from 30 days after planting, and was

more severe in the Vertisol plots. Such attack

could be a major factor limiting nitrogen fixation

by pigeonpea.

During the rainy season, seed inoculation with

Rhizobium in Vertisol significantly increased the

number and dry weight of nodules per plant and

plant weight at 45 days after planting, but no

effect was found at 80 days or on grain yield.

Where rice had been grown for several years,

the soil contained less than 103 rhizobia/g soil

and in this field pigeonpea responded to

inoculation.

Nitrogenase activity per plant increased from

16 µ moles C2H

4/plant per hour at 30 days after

planting in Alfisol to 55 µ moles at 60 days and

decreased to 2 µ moles at 125 days. Nodules

recovered from Alfisol were more active in

nitrogen fixation than those from Vertisol, re­

flecting differences in the amount of nodule tissue

per plant rather than differences in activity per g

nodule tissue.

Nodulation of Germplasm Lines

A preliminary survey showed a wide range in

nodulation of 731 pigeonpea germplasm lines

grown in the field. At 25 days after planting,

nodule number per plant varied among cultivars

from 5 to 48. By 55 days, the number varied from

10 to 64 per plant. The proportion of apparently

active pink nodules and senescent or damaged

nodules also varied a great deal and was not

apparently related to total number of nodules

per plant. For some lines, 40 percent of the

nodules were hollow at 55 days.

Screening for Disease and Insect Reaction

Phytophthora Stem Blight

In the unusually wet August of 1976, a stem

disease developed which selectively killed cul-

81

tivars in experimental plots. Identified as Phy-tophthora blight, the disease appears to be causedby a species, not definitively identified as yet,different from P. drechsleri var. cajani describedfrom New Delhi in 1970. A laboratory-screeningtechnique has been developed; use of a 15- to 20-day-old culture in a 1:200 dilution of inoculum tospray-inoculate 15-day-old seedlings gave thebest results. Field observations confirmed bylaboratory tests indicate that cv ICP-7065 andNP 69 are good sources of resistance. Screeningof parent lines and segregating populations wil lbe a routine part of the resistance-breedingprogram.

Fusarium Wil t

The development of the two wilt-sick plots, 'A 'and 'B', was continued by incorporating stubbleof diseased plants and growing susceptible cul-tivars. The percentage level of "sickness" in plot'A, ' as determined by the wilt incidence insusceptible cv Sharda, was 52.3 and in the newerplot 'B ' it was 26.5.

Breeding material was screened in the wilt-sickplot 'A. ' Generally high wilt incidence occurredin crosses involving the susceptible parent ICP-6997. Crosses involving NP(WR)-15 performedwell. Of the 189 materials tested, 11 F2 ,4 F3, andone germplasm selection showed less than 25percent wilt. One F2 population, ICP-1 x NP(WR)-15, showed only 5.8 percent wilt.Only 7 single-plant progenies out of 318 screenedin the sick plot 'B ' remained wilt-free. In lab-oratory screening only 2 single-plant progeniesout of 541 screened were unaffected by wilt.None of the 48 materials from intergenericcrosses (Cajanus x Atylosia) was found promis-ing against wilt in laboratory testing. When"water-culture" wilt screening was done on 1000individual plants of each of several lines, thosewith the highest frequency of survival were 15-3-3, BDN-1 , No. 148, and ICP-7035.

Sterility Mosaic

We have reported earlier the leaf-stapling tech-nique for infesting test plants with mite vectors of

sterility mosaic. We have now completed screen-ing of the existing germplasm, using this tech-nique on young plants. The five resistant linesreported last year - ICP-3783, -6986, -6997,-7035, and -7119-have been confirmed to beresistant, as have four previously unreportedlines-ICP-3782, -7687, -7942, and -8113. Oneline, ICP-2376, shows symptoms which we de-scribe as ring spots, and we are investigating itfurther. We have identified 29 more lines whoseresistance is still to be confirmed in the nextgeneration.

Last year we screened breeding material usingthe leaf-stapling technique. This year we haveestablished a large screening area using theinfector-row technique. NP(WR)-15, a wilt-tolerant but sterility mosaic-susceptible cultivar,was planted early and inoculated with sterilitymosaic. Paired rows of the infector were spacedto permit planting 12 rows of test material inbetween; and with this spacing, susceptible mat-erials show close to 100-percent infection. Plantsnot showing symptoms are inoculated by leafstapling to detect possible escapes.

Screening for wilt and mosaic can be donesimultaneously by using infector rows forspreading sterility mosaic in a wilt-sick plot. Weconsider both of these diseases sufficiently im-portant to give priority to combining resistanceto them in improved cultivars. We have de-termined that the sterility mosaic-resistant linesICP-3783 and -7035 have field tolerance to wiltand Phytophthora blight. ICP-7119 has fieldtolerance to wilt as well as resistance to sterilitymosaic.

Insect Reaction

Levels of insect damage were determined onsmall plots of 4558 germplasm accessions andlarger plots of breeding material under pesticide-free conditions. Differences in cultivars wereobserved, and those breeding lines with lowlevels of damage wil l be retested in large rep-licated plots. Derivatives of intergeneric hy-brids (F6), previously unselected for insect re-action, gave very interesting results. Groups oflines from crosses involving three species of

82

Atylosia were compared and podfly and borerdamage was least in crosses with Atylosia scara-baeoides (Table 28). Some antibiosis in A. scara-baeoides to Heliothis armigera was also found.When compared with those fed on pigeonpea,the period of larval development was extendedand weights were lower in larvae and pupae whenthe insects were fed on A. scarabaeoides (Table29). In this regard, A. sericea was intermediatebetween A. scarabaeoides and the pigeonpeacheck.

Intergeneric Hybr id iza t ion

We reported last year that derivatives fromcrosses of pigeonpea with Atylosia species werehigh in protein percentage. These results were

confirmed during the 1976-1977 year. We planmore detailed analyses of the Atylosia speciesthemselves and are attempting to make crossesamong the species, as well as additional crosseswith Cajanus. At present we have 10 Atylosia species available (Table 30). The most recentlycollected, A. cajanifolia, appears to be the closestwild relative of Cajanus. It was collected in theBailadilla Hills in Bastar district of MadhyaPradesh. A. rugosa, also added during the year,was collected near Ootacamund in Tamil Nadu.

Screening for Soil Factors

Screening for waterlogging tolerance under fieldconditions was unsuccessful. Chambers havebeen constructed for this purpose, and tech-

Table 29. Development of Heliothis armigera Hub. immature stages feeding on Atylosia scarabaeoides, A. sericea, and a pigeonpea check.

Larvaea Pupaea

Food plant

3rd instar Larvallength weight

Larval-development

period

Weight Pupalperiod

A. scarabaeoides A. sericea Pigeonpea check

(cm) (mg)

2.3 133.42.7 201.33.2 321.8

(days)

31.829.526.8

(mg)

186.1209.3276.8

(days)

12.913.813.9

aAverage of 12 insects.

83

Table 28. Pest damage in derivatives of intergeneric hy brids (F 6) grown without insecticide use.

Borer-damaged

pods

Podfly-damaged

pods

Meanplantyield

C. cajan x A. lineata C. cajan x A. scarabaeoides C. cajan x A. sericea Pigeonpea check

(%)

59.850.456.357.5

(°/o)

6.46.18.1

17.4

(g)

5.26.98.0

12.7

niques will be tested for reliability of results.Preliminary work is in progress on developmentof precise screening for salinity tolerance.

Ratoon ing

Last year we reported increased production byratooning. Pigeonpeas are normally harvestedwhen the pods mature. However, if the plants areleft in the field they go on to produce a secondflush of flowers and a second crop of pods. Thissecond crop is produced during the hot dryseason when the fields cannot be used for othercrops; the well-established deep root system ofthe pigeonpeas is able to exploit water which isstill available in the soil.

The first crop of pods can be harvested bycutting off the pod-bearing branches; theratooned plants then go on to produce a secondcrop. The ability of cultivars to produce a goodsecond crop after ratooning differs considerably,but some are able to produce almost as much inthe second crop as they did in the first.

We have investigated the effects of differentratooning treatments, at the time of the firstharvest, on second-harvest yields. When theplants were ratooned progressively lower down,closer to the ground, there was more regenerativevegetative growth before they again beganflowering. Plants which were not ratooned at all,from which the first crop of pods had beenplucked by hand, were first to begin floweringagain and they produced a second crop of podssooner than any of the ratooned plants. Thesenonratooned plants produced the highestsecond-harvest yields. The lowest yields camefrom plants which were ratooned closest to theground.

These results clearly showed that the besttreatment of all was nonratooning, pfobablybecause the second crop of pods was producedwith the minimum delay at a time when waterstress was becoming progressively greater. Theyields obtained were quite considerable. Forexample, in an experiment on a deep Alfisol cv148 and AS-71-37 produced a mean second-harvest yield of 1 000 kg/ha without ratooning,

Table 30. Atylosia species available at ICRISATCenter.

Species Entries

A. scarabaeoides 12A. sericea 3A. lineata 3A. platycarpa 1A. volubilis 2A. trinervia 1A. grandifolia 1A. albicans 3A. cajanifolia 1A. rugosa 1

Figure 27. Plants of Atylosia cajanifolia, ap-parently the closest wild relative of pigeonpea, collected in Bailadila Hills of Madhya Pradesh, India.

84

compared with 500 kg/ha when ratooned at a height of 60 cm. The mean yield during firstharvest was 1 200 kg/ha. Hence the second-harvest yield without ratooning was more than80 percent of the yield of the first harvest, whichemphasizes the importance of attempting toexploit the second-harvest potential of pi-geon peas.

Pigeonpea as a W in te r C r o p

In India, pigeonpeas are normally planted at thebeginning of the rainy season (Jun/Jul).Medium-duration cultivars take about 5 to 7 months to mature and late cultivars more than 7 months. The flowering of medium- and long-duration cultivars of pigeonpeas is inhibited bythe longer day lengths of the summer and theearly part of the rainy season; they are photo-period sensitive "short-day" plants. If they areplanted at the beginning of India's winter season(Oct) they flower and mature sooner because ofthe shorter day length during the winter season;

there is less growth because of the coolerweather.

Last year we found that pigeonpeas could begrown successfully as a nonirrigated winter crop.Because the plants are much smaller than usualthey can be grown at much higher populationdensities than in the normal season. This year wecarried out a more-detailed trial to investigatethe potential of pigeonpeas as a winter crop.Early, medium-, and late-maturing cultivarswere compared. Although the late types were thelast to mature. they matured in just under 5 months compared with more than 7 months inthe normal season; the medium-duration cul-tivars matured in 41/2 months and the earlycultivars in 4 months. The early cultivars gavethe lowest yields and the medium-duration cul-tivars the highest. Of these, cv C -11 was the bestwith yield of more than 1 700 kg/ha; cv ICP-1gave 1 500 kg/ha compared with only 1 300 kg/hain the normal season. These results indicate theconsiderable potential of pigeonpea as a wintercrop.

Figure 28. Cultivar differences in response to soils waterlogged for 6 days in chambers (on right) contrasted with normal growth on well drained soils (on left).

85

Figure 29. Screening for salt tolerance. In upper photo, four cultivars grow in normal field soil; the cultivars below show a differential response to added salts (40 meq/kg soil).

86

Figure 30. Pigeonpeas planted in the postrainy

season are competitive in yield with

full-season pigeonpeas planted in June

or July. Cultivars of different maturity

are shown in an experiment in which

effects of planting date and spacing are

also studied.

Observations showed that nodulation and

nitrogenase activity of these winter-sown pigeon-

peas was much poorer than on those sown in the

rainy season. At 40 days after planting, nodules

per plant ranged from 6 for cv 7065 to 17 for cv

C - l l , which also had the greatest nitrogenase

activity of 4.68 µ moles C2H

4, plant per hour.

Nodule efficiency - nitrogenase activity per unit

nodule weight - was also much lower than for

the rainy season planting, with most activity at

40 days (133 µ moles C2H

4/g dry nodule per

hour for C-11). By 70 days, nitrogenase activity

was virtually nil. Insect grubs also attacked

nodules on this winter crop.

Residual Effects of

Pigeonpeas on Subsequent

Crops of Pigeonpea

Last year we found that when pigeonpeas were

87

grown in soil where pigeonpeas were grown the

previous year, the plants were stunted and un­

healthy and gave very poor yields. This year we

observed the same effect in some, but not all,

parts of ICRISAT Center. Other crops such as

sorghum, millet, and castor were unaffected.

This deleterious effect on pigeonpea growth

could not be brought about by adding pigeonpea

residues to normal soil, nor was the effect

reduced by removing pigeonpea residues from

soil when pigeonpea had been grown previously.

Therefore a toxic effect of pigeonpea residues can

be ruled out. We now have some preliminary

evidence that the harmful effects might be ex­

plained by the build-up of parasitic nematodes.

This possibility is being investigated further.

Effects of N i t r o g e n Fertil izer

on Pigeonpea G r o w t h and

Y i e l d

Improved cereal pigeonpea intercropping sys­

tems will almost certainly involve applying nitro­

genous fertilizer to the cereal. We investigated

the effects of a starter dose of 20 kg N ha as well

as a liberal dose of 200 kg N ha and 20 tons/ha of

farmyard manure on the growth. nodulation,

and yield of cv ICP-1 grown on Vertisol and on

Alfisol. Vegetative growth was increased sig­

nificantly by 200 kg N ha but nodulation and

nitrogenase activity, measured by the acetylene-

reduction technique, was surprisingly little af­

fected. In Vertisol at 30 days after planting there

were 15 nodules per plant with 21 mg dry weight

for the 200 N treatment compared with 29 per

plant (52 mg) for no N fertilizer. At 60 days these

differences had disappeared. Nitrogenase ac­

tivity per plant followed the same pattern. Nod­

ules that formed, regardless of the N treatment,

were equally efficient in nitrogen fixation per g of

nodule.

However by the time of harvest, although the

nitrogen-fertilized plants had more stem material

and had produced more leaves, there were no

significant differences in grain yield. These re­

sults indicate that yield was not being limited by

nitrogen supply.

so far we have found none in which the weight ofthe later-formed pods shows a distinct andsignificant decline. This type of measurementseems to be sufficiently simple and reliable for usein screening large numbers of plants from thegermplasm collection in an attempt to findcultivars in which pod-set is not limiting theyield, if such cultivars exist.

Source-sink Relat ionships

Flower Removal

Experiments were carried out in the field toextend last year's observations of the effects ofexperimentally delayed pod-set on yield. Usingthe early and medium-duration cultivars, wefound that when all flowers were removed fromthe plants for 1 to 7 weeks, there was nosignificant reduction in yield in any of thecultivars. This ability to compensate so well forthe loss of early formed flowers and pods isprobably of considerable importance in enabling

Nitrogen fixation by the nodules is apparentlyadequate for the yield potential of this cultivar atICRISAT Center. Farmyard manure treatmentenhanced nodule activity in the early stages, buthad no significant effect on growth or yield.

Seed Size a n d Seed l i ng S ize

The influence of cultivar differences in seed sizeon seedling growth were investigated with twentycultivars from seeds with 100-seed weights rang-ing from 4.5 to 22.2 g. At first there was a closerelationship between seed size and seedling sizewith 100-seed weights of up to about 16 g, but astime went on size of the young plants bore lessrelation to the size of the seeds from which theywere derived (Fig 31). The greater "seedlingvigor" of the larger-seeded cultivars could be ofimportance when seedlings face adverse con-di t ions-such as, for example, surface crusts onAlfisol.

Ovule and Seed A b o r t i o n

The extent of ovule and seed abortion during poddevelopment in a range of genotypes was esti-mated. In some small-seeded cultivars, up to 90percent of the ovules developed into matureseeds, but in other cultivars more than half of theovules or young seeds aborted. There was a negative correlation (r = - 0.81**) between thepercentage realization of the potential seed num-ber and 100-seed weight.

Pod-set

In previous years we found that pods formed inthe later reproductive phase were the same size asthose formed earlier. This suggested that pod-setwas suboptimal. Cultivars in which pod-set is notlimiting yield might be expected to show a declinein the average weight of late-formed pods, whichwould indicate a limitation of nutrient or assimi-late supply. We have compared weights of earlyand late-formed pods in more than 30 cultivars;

Figure 31. Relationship between 100-seed weight and total dry weight of seedling shoots of 20 pigeonpea cultivars.

88

Seed weight ( g / 100 seeds)

4 6 8 10 12 14 16 18 20 22 24

B. 56 days after sowing

35

30

26

22

18

14

10

A. 14 days after sowing

0.6

0.5

0.4

0.3

0.2

0.1

the plants to produce some yield under unfavor-able environmental conditions and also to with-stand insect attacks; it has important impli-cations for the development of minimum-inputpest-management systems.

Effects of DefoliationLast year, which was an unusually wet year withlate rains, we found that removing different pro-portions of the leaves from the plants through-out the reproductive phase led to proportionatedecreases in yield. This year the rainy seasonended unusually early and as a result growth andyields were reduced. Under these conditions,with four different cultivars, we found that halfor more of the leaves could be removed through-out the reproductive phase without affecting theyield significantly. Data for cv ICP-1 are plottedin Figure 32. These results indicate that leaf areawas not limiting pod-set and yield. The differencefrom last year's observations suggests that underconditions of moisture stress the photosyntheticefficiency of the leaves may be limited by watersupply; perhaps the removal of half the leavesresults in an increased water supply to theremaining leaves and an increase in their photo-synthetic efficiency.

Diseases and Pests

Fusarium Wilt

In our continued studies of the wilt disease, wefound three pathogenic Fusarium species as-sociated with wilted plants: F. udum, F. oxys-porum, and F. solani. Twelve distinct culturalisolates of F. udum, and two each of F. oxys-porum and F. solani are being maintained forfurther studies.

Preliminary studies indicated that F. udumcolonized roots of pigeonpea plants in the sickplot between 16 and 30 days after planting. Thepathogen could be detected in the plants prior tothe appearance of wilt symptoms. In infectedplants, F. udum was isolated from the root, stem,branches, and pod hulls, but not from seed. The

89

Figure 32. Effect of different degrees of defoli-ation throughout the reproductivephase on yield and yield components ofcv ICP-1.

Defo l i a t i on (%)

0 10 20 30 40 50 60 70 80 90 100

32

28

24

20

16

12

8

4

0

Y i e l d per p lant

L.S.D.(0.05)

160

140

120

100

80

60

40

20

0

Pods per p lant

L.S.D.(0.05)

Seeds per pod

L.S.D.(0.05)

3

2

1

0

10

8

6

4

2

0

100-seed weight

L.S.D.(0.05)

fungus remained restricted to the undergroundparts of plants not showing wilt symptoms. Twodistinct types of symptoms-wi l t ing with andwilting without a necrotic band on the stem -were observed.

Our observations of the past 3 years at1CRISAT Center indicate that the appearance ofwilt symptoms increases rapidly from Novemberonwards. We have started studies of the yield lossin relation to time of wilt ing, and preliminaryresults show a progressive decline in losses withsymptoms appearing in successively later stagesof pod formation and development. Surveys ofdisease incidence may tend to overestimate yieldlosses if seed production on diseased plants isignored.

Sterility Mosaic

The causal agent of this disease has not yet beenidentified. We have been unable to transmit thedisease either by mechanical transmission or bygrafting. In the case of diseases caused by a mycoplasma-like organism, treatment with tet-racycline has been found to give remission ofsymptoms. Tetracycline applications, made inseveral ways, had no effect on symptoms ofplants infected with sterility mosaic. Electronmicroscope observations to date have not re-vealed virus or mycoplasma-like drganisms.

Although the organism has not been iden-tified, the vector has been known for severalyears. Our studies have shown that a minimum30-minute inoculation-access period was re-quired for transmission by the mite (Aceriacajani). A single mite can transmit the causalagent.

Although infected plants are usually sterile,seed is sometimes produced and we have grownseveral thousand seeds from diseased plants. Inno case has sterility mosaic been transmittedthrough the seed; we consider that the causalagent is not seed-borne.

Disease Surveys

Surveys are being made each year to assess lossesin farmers' fields. This year the central and

western districts of Maharashtra, not included inour previous survey, and the states of Karnatakaand Tamil Nadu were surveyed. Wil t and sterilitymosaic were found to be the most importantdiseases, with a very low incidence of root rot andleaf diseases observed. The relative importanceof sterility mosaic increased in areas furthersouth (Table 31).

Table 31. Observations of wilt and sterilitymosaic in farmers' pigeonpea fields inthree states of India, 1976.

SterilityWilt mosaic

State Range Mean Range Mean

Maharashtra (western & central districts)KarnatakaTamil Nadu

— ( % ) — — ( % ) —

0-38 7.6 - 3.90-17 1.1 0-95 9.70-65 1.4 0-93 12.8

Pest Surveys

The relative importance of various insect pestson pigeonpea was recorded by surveying andsample collection in Karnataka, Tamil Nadu,Maharashtra (Akola and Sholapur districts),Bihar, Orissa, Andhra Pradesh, and Uttar Prad-esh. The borer damage ranged from 3 to 84percent. Severe pod damage was recorded insouthern states of India (Table 32).

H. armigera was found to be the major causeof loss in pigeonpea but Exelastis atomosa Wals.and Maruca testulalis Geyr. were also commonas pod borers.

Nut r i t i ona l Qua l i t y

One of our objectives is to improve the nu-tritional quality of pigeonpea. We recognize thatthe first priority is to increase production ofpulses, so have not included nutritional factors asselection criteria in the breeding program. Em-phasis to date has been on screening the germ-

90

Table 32. Range of pod damage to pigeonpeas caused by different pests in samples collected from variousstates of India (1975-1977).

Species

Lcpidopterousborers

(H. armigera,Bruchids(Calloso-

E. atomosa; Podfly bruchus Hymenoptera Actual podState Maruca (Melanagro- chinensis; (Taraostig- damage by

testulalis) myza obtusa) C. maculatus) modes sp.) insects

(%) (%)(%)

Uttar Pradesh 2.6-12.3 9.9-36.0 0 - 0 . 2 0 - 0.2 15.5-48.7Bihar 3.1-31.3 8.5-56.3 0 - 0.7 Below 1.0 15.5-70.6Orissa 11.8-28.2 23.5-28.1 1.0- 4.4 0 .2- 1.9 37.5-52.4Andhra Pradesh 15.3-67.9 1.1-14.6 0 -22.2 0 -34.3 18.2-71.4Maharashtra 10.9-83.5 1.8-44.3 0 - 1 . 4 15.5-84.6Karnataka 3.1-81.1 2.4-19.7 0.9-43.8 0.5-54.1 27.3-88.3Tamil Nadu 12.4-73.7 0.6-32.6 1.1-46.6 0 .7- 4.7 26.1-81.5

plasm for total protein content, and conductingstudies on techniques for estimating protein andsome of the limiting amino acids. These studiesare essential for identifying rapid and economi-cal procedures that will be needed for implement-ing a breeding program for higher quality.

Comparison of Methods of ProteinEstimation

The Kjeldahl method of estimating total nitro-gen is considered to be the standard method. TheMKJ (micro-Kjeldahl) method gives very similarresults, and we have used it as a standard methodin our studies. We also use the T A A (TechniconAuto Analyzer) and DBC (dye-binding capacity)methods. We found a very high correlation(r = +0.96) between M K J and T A A when test-ing whole seeds. However, the correlation ofM K J and DBC determinations was lower(r = + .87). This lower relationship was attri-buted to interference of seed coat pigments in theDBC method. Whole seed having the sameprotein content as dhal samples gave a higherreading on the U D Y instrument used in the DBCmethod (Fig 33).

Figure 33. Relationship between DBC method and MKJ procedure in determining protein content of pigeonpea.

When dhal samples were analyzed, both T A A( r = +.96) and DBC (r = +.94) were highlycorrelated with M K J . It was concluded that forrapid screening of dhal samples, either DBC orT A A could be used, while for screening whole-seed samples T A A should be used.

91

Mic ro -K je l dah l p ro te in percent ( N x 6 . 2 5 )

20 21 22 23 24 25 26 27 28

Dha l

Whole seed

60

50

40

30

Whole-seed Protein vs. Dhal Protein

For rapid screening it would be desirable to useground samples of whole seed; preparation ofdhal of many small samples is tedious and time-consuming. We determined the protein contentof dhal and whole-seed samples of 43 cultivarsusing the M K J technique, and found a cor-relation of r = + .94 (Fig 34). The same samplesby T A A gave a correlation of r = + .97, andwith the DBC method r = +.88. It was con-cluded that for rapid screening, whole seed ratherthan dhal samples could be analyzed.

Errors of Determination of Protein

In order to establish reliable procedures forcomparing protein content of different pigeon-pea lines, we conducted an experiment to com-pare sources of error. Seed of 10 cultivars grownin three replicates were sampled twice, and thesesubsamples were each analyzed twice by TAA.The experiment was repeated on 2 different daysin the laboratory.

The experimental error (random variationsamong cultivars in different replicates in thefield) was the most important source of error. Itwas three to four times as large as error due to

sampling within seed lots or determination errorin the laboratory. The results indicate thatreliable estimates of genetic differences in proteincannot be made on individual plants or plotsbecause of interaction with the environment.However, with the three replicates used, it waspossible to identify small differences amongcultivars.

Sulphur Amino Acids and Tryptophan

Protein quality of pigeonpea is determined inpart by the limiting amino acids. Pigeonpea seedprotein is normally low in methionine, cystine,and tryptophan. Microbiological assays for theestimation of sulphur amino acids and the use ofthe Leco sulphur determinator for the estimationof total sulphur in pigeonpea samples are beingtried, in order to understand whether the totalsulphur determination would be an indirectuseful way to screen for sulphur amino acids.Two colorimetric methods for the estimation oftryptophan are being evaluated for rapid-screening purposes.

Composition of Some PigeonpeaCultivars

Ten released cultivars of pigeonpea were ana-lyzed for protein, starch, soluble sugars, fat,crude fiber, and ash contents (Table 33). Proteincontent ranged between 23 to 27.3 percent,whereas starch percentage ranged between 56.3to 64.1. Fat, crude fiber, and ash contents rangedbetween 1.2 to 2.2, 1.0 to 1.2, and 3.3 to 4.3percent respectively.

In ternat iona l Coopera t ion

We are maintaining a close liaison with pigeon-pea breeding programs in India and in othercountries. We furnish detailed annual reports ofour various disciplines to cooperating scientiststo keep them abreast of developments prior toformal publication of our results. We also fur-nish breeding material; 21 breeders from 8 countries were supplied segregating populations

92

figure 34. Relationship between whole-seed pro-tein and dhal protein in pigeonpea by MKJ method.

Whole-seed p ro te in percent ( N x 6.25)19 20 21 22 23 24

r = + 0 . 9 4 * *28

27

26

25

24

23

4.428 + 0.939 x

and selected lines. Many requests for cultivars tobe tested for local adaptation have been filled;requests have ranged from types for grain pro-duction to specialized types for forage pro-duction or for use as windbreaks.

Looking Ahead in PigeonpeaImprovement

We now have quantitative evidence showing thatpod-set is limiting pigeonpea yields; this con-clusion was generally accepted before such datawere available. However, we now have a simplescreening technique for identifying genotypes inwhich pod-set is not limiting. A l l genotypesexamined so far are limited by pod-set; we wil lscreen the germplasm and segregating materialfor the exceptional types.

Disease-screening techniques now availablemake possible a shift in emphasis in the breedingprogram. Where we have been selecting for yieldin early generations, we can now screen forresistance to wilt, sterility mosaic, and phy-tophthora blight resistance sequentially, anddefer selection for yield until more advancedgenerations when apparent heritability is higher.

Sick plots for disease screening are being de-veloped to permit simultaneous selection forresistance to more than one disease.

The possibility of utilizing F1 hybrids incommercial plantings will be studied intensivelyin the immediate future. In addition to yieldadvantages, hybrid utilization can be expected toresult in wider adaptation and would facilitatethe combining of desired characters directly ingenotypes for planting by farmers.

Additional emphasis will be placed on de-veloping methodology for screening for insectresistance. Evidence of antibiosis in wild speciesof Atylosid indicate the need for identifyingspecific factors that can be transferred bybreeding.

It is clear that the total supply of pulses canbest be increased through improved yields. How-ever, since they are a source of protein in the diet,we wil l attempt, through intercrossing, to de-velop unusually high protein lines. These wil l beexamined for protein quality before high proteinis included as a primary selection criterion.

We wil l encourage wide-scale testing of pi-geonpea as a postrainy season crop, utilizingmedium-late rather than early cultivars; earlyplanting has given postrainy season yields higherthan those of the rainy season for some cultivars.

Table 33. Chemical composition of dhal samples of some pigeonpea cultivars, expressed on a moisture-free basis.

Soluble CrudeCultivar Protein Starch sugars Fat fiber Ash Total

(%) (%)

HY-3C 23.1 61.6 5.3 2.2 1.1 3.3 96.6ICP-1 24.7 61.2 4.8 1.2 1.2 4.0 97.1ST-1 23.0 61.6 5.2 1.4 1.1 3.7 96.0No. 148 24.5 63.2 5.7 1.4 1.2 3.9 99.9T-7 25.9 64.1 5.5 1.4 1.2 3.7 101.8

T-17 27.3 57.7 4.7 1.7 1.2 4.1 96.7T-21 24.6 59.0 5.3 1.5 1.1 4.0 95.5BDN-1 24.2 60.8 5.4 1.8 1.0 3.9 97.1C-11 24.3 61.8 5.1 1.5 1.2 4.2 98.1Gwalior-3 26.6 56.3 5.0 1.5 1.2 4.3 94.9

93

The scope and limitations of taking a secondgrain crop from short- and medium-durationcultivars planted in the rainy season wil l bestudied intensively; this appears to be a goodpossibility for increasing production withoutadditional inputs.

The possibility of cropping Vertisols which arenormally fallowed in the rainy season by includ-ing pigeonpea as a component in the precedingpostrainy season cropping combination and per-

mitting it to continue into the rainy season wil l befurther investigated. The payoff in increasedrainy season cropping could be tremendous.

International testing wil l continue to expandas advanced lines are developed. The first ICPselection nursery was distributed for planting inthe coming season. Consisting of 38 advancedlines selected for large seed for possible use asvegetable types, it was sent to 16 locations in 11countries.

94

C h i c k p e a

The objectives of the chickpea improvementprogram are: (i) to develop high-yieldingdisease- and pest-resistant cultivars with goodgrain quality, (ii) to furnish advanced breedinglines and segregating populations to nationaland local breeding programs, and (iii) to sup-port regional and national programs throughexchange of information, germplasm, and train-ing of personnel.

Chickpea is a relatively short-season crop,maturing in approximately 4 months atICRISAT Center and in a slightly longer time atHissar in northern India. Thus there is ampletime for growing two generations per year, butthe climate at ICRISAT Center and Hissar doesnot permit growing a second generation. For 2 years we grew the second generation in Lebanon,but unsettled conditions there and problems inprocessing so many samples through quarantineforced us to discontinue. We also utilized theLahaul Valley in northern India for off-seasonadvance, but the growth conditions were lessthan optimum and there was also a problem ofacquiring land; we abandoned that location afterthe 1976 crop. We are presently growing off-season test plantings at three sites in Kashmir,hoping to find a location to supplement the workat ICRISAT Center and at Hissar. Research inbreeding, physiology, and germplasm evaluationhas been carried out at both of the abovelocations; research in pathology, entomology,and microbiology has been concentrated atICRISAT Center.

Germp lasm Resources

We now have 11140 accessions of Cicer arie-tinum L. Figure 2 shows the geographic originof the accessions. During the year, collectiontrips were made in Afghanistan and in India - i nthe states of Gujarat and Maharashtra, and theeastern area of Uttar Pradesh. A number ofaccessions were contributed by colleagues. A summary of new accessions is listed in Table 34.

Figure 35. Chickpea germplasm accessions at ICRISAT Center.

Table 34. Chickpea germplasm accessions ob-tained in 1976-1977.

Source Entries

Agricultural Research Station, Bad-napur, Aurangabad, India 29

Survey collection from Afghanistan 30Agricultural Research Station, Wagga,

Australia 19

Agricultural Research Station,Arnej, Ahmedabad, India 12

Survey collection from Gujarat andMaharashtra, India 133

Survey collection from EasternU.P., India 49

All-India Pulse Workshop 24

Total 296

Evaluation and Maintenance

Germplasm accessions planted during the yearwere evaluated primarily for disease resistance,although this was not our original purpose. Thefield planted at ICRISAT Center turned out tobe a hot spot for Fusarium wilt, and approxi-mately half of 3784 lines planted were dis-eased. Observations on 24 morphological andyield characters could be recorded properly foronly 850 accessions. At Hissar, only 67 of 2 363

97

1973 1974 1975 1976 1977

Chickpea1110

9876543210

accessions planted remained free of stunt. How-ever, it was possible to record full data on 615.

A total of 181 promising lines were grown inreplicated tests for more precise evaluation;harvest index ranged from 14.5 to 70.8 percentand flowering duration from 24 to 66 days atICRISAT Center.

Seed Distribution

Requests were filled as indicated in Table 35. Inaddition, 8 591 samples were utilized by otherdisciplines within ICRISAT.

Breeding Program

Hybridization

Our crossing block consists of lines selected fromthe germplasm and is a working collection for thebreeding program. This year there were morethan 300 strains selected for various traits,including high yield, plant type, maturity period,seed size, podding habit, disease resistance, pro-tein content, wide adaptation, and other specialcharacters. Based on performance, a number oflines are dropped each year, and others are addedfrom the germplasm pool.

Table 35. Chickpea germplasm supplied to research a gencies in India and other nations during 1976-1977.

Institution Location Entries

I N D I A :Osmania UniversityKakatiya UniversityHaryana Agricultural UniversityRegional Research Station

Hyderabad, A.P.Warangal, A.P.Hissar, HaryanaRaichur, Karnataka

1020

10637

I A R IAgricultural Research StationPunjab Agricultural UniversityPunjab Agricultural UniversityTamil Nadu Agricultural University

New DelhiNayagarh, OrissaGurdaspur, PunjabLudhiana, PunjabCoimbatore, Tamil Nadu

557

226363

50

Chandra Shekar Azad University ofAgriculture & Technology

Banaras Hindu UniversityKanpur, U.P.Varanasi, U.P.

531000

OTHER NATIONS:University of OttawaMr. GuitierrezUniversity of ReadingCentre National de Recherches Agronomiques

CanadaPalmira, ColombiaEnglandVersailles, France

9102412

Minor Irrigation Project, FAOAgricultural Research Centre, C/o U N D PDepartment of Scientific and Industrial

ResearchI C A R D AChief Agricultural Research Officer

Nakuru, KenyaTripoli , Libya

Christ Church, New ZealandSyriaLusaka, Zambia

1020

151050

98

During the past year 2 000 crosses among theseselected parents and among hybrids were made.In the nurseries at ICRISAT Center and atHissar we grew a total of 14244 segregatingpopulations and progeny rows for selection.

Selection for High Yield

Emphasis in selection has been on high yield.Pedigree selection at the two breeding sites hasprovided an opportunity to find segregants adap-ted to two diverse environments. Inferior pro-genies can be rejected on the basis of visualobservation, but because of differential podfilling, variation in seed size, and the quantity ofmaterial involved, it has been necessary to makefinal selections among the best lines on the basisof yield measurements. Frequent checks are usedfor comparison with the progeny rows, which arenot replicated.

Breeding for Plant Type

Chickpea has a relatively high harvest index, and

the desi types are considered to be efficientproducers. These are relatively short cultivars,and we are investigating the possibility of in-creasing plant height as a means of providingmore sites for pod production. A comparison ofyields of some check cultivars and selected F3

progenies is given in Table 36. Relative yields perplot show that the tall parents were not adaptedto Hyderabad conditions. We are pleased withthe performance of the F3 lines, since it has beenpossible to select some that are superior in yieldto the adapted check (Annigeri). We have madesome backcrosses to the high-yielding parent,and expect the introduction of more genes fromthe locally adapted parent to produce higher-yielding segregates.

Diseases of Chickpea

Before breeding for disease resistance could beinitiated, knowledge of the importance and dis-tribution of chickpea diseases was necessary.During the last three seasons extensive surveys

Table 36. Comparative performance of normal and tall ch ickpea cultivars and their F 3 progenies during1976-1977 at ICRISAT Center (yield based on single met er-length rows, not replicated).

Cultivar/progeny Height Maturitya Yield

(cm) (kg/ha)Dwarf parent

Annigeri 30-35 E 2654G-130 35-40 L 2371

Tall parent

K-1184 65-70 V L 499K-1481 65-70 V L 747

F3 progeny

H-208 x K-1258( -37) 54 M 3448K-4 x K-56567(-18) 60 M 3144F-378 x K-1184( -28) 60 M 2995H-208 x K-1258 ( - 3 4 ) 55 M 2931Annigeri x K-1480 ( - 3 5 ) 51 M 2816

aE = early, M = medium, L = late, VL = very late

99

were carried out, particularly in India. Thelocations included several experiment stationsand farmers' fields in India, Ethiopia, Iran,Lebanon, Syria, and Turkey.

In India three diseases seem to predominate.These are stunt (cause considered to be a virus),wilt caused by Fusarium oxysporum, and root rotcaused by Rhizoctonia bataticola. Other causes ofdrying/wilting include fungi such as F. solani, R.solani, Sclerotium rolfsii, Sclerotinia scle-rotiorum, Operculella padwickii; a mosaic virus;termites; root-knot nematode; lack of moisture;salinity; and frost damage. In Iran (at Karaj),chickpea stunt was common and bean yellowmosaic and alfalfa mosaic viruses were alsopresent. In Turkey, traces of root rot by R.bataticola and stunt were observed, but themajor problem was Ascochyta blight. In Ethi-opia, stunt was the major problem around DebreZeit. In Syria and Lebanon, R. bataticola rootrot, stunt, and Ascochyta blight were observed.

On the basis of our intensive studies it is nowpossible to state that there are distinct diseaseswhich can be considered as components of theso-called "wi l t complex." It is now possible todiagnose, without much difficulty, these differentdiseases; the ultimate effect of the complex ispremature drying of plants. A pamphlet describ-ing and illustrating these diseases wil l be pub-lished by ICRISAT; it should be a useful diag-nostic aid.

Breeding for Disease Resistance

A wilt (Fusarium oxysporum f. sp. ciceri) sick plotand a multiple-disease sick plot were developedfor screening. By growing crossing block entries(more than 450) in the multiple-disease sick plot,it was possible to identify about 60 promisinglines or cultivars. The laboratory technique(water-culture) for wilt screening was improvedfurther and a good correlation noted betweenfield and laboratory results. Preliminary in-dications of the existence of races of F. oxys-porum have been obtained.

More than 1200 germplasm accessions, cross-ing Mock entries, and wild Cicer species werescreened for Ascochyta blight resistance by in-

oculating seedlings in an Isolation Plant Prop-agator. Of these, more than 40 cultivars orgermplasm accessions were found promising(rating 3 to 5 on a 9-point scale). Two wildspecies which showed resistance were Cicer re-ticulatum and C. anatolicum.

With the successful screening in the wilt-sickplots this past year, plans are being made toscreen breeding lines in the coming season. Thereis a backlog, on which screening wil l be initiated,of material in F1 to F5 generations from crossesinvolving resistant parents.

Seed-borne F u n g i

The genera of fungi isolated from nonsterilizedcv JG-62 seed were Alternaria, Aspergillus, Cur-vularia, Fusarium, Penicillium, Phoma, Rhizoc-tonia bataticola, and Rhizopus. Only Fusarium and Alternaria were isolated from surface-sterilized seed. The germination percentages ofnonsterilized and surface-sterilized seed were 82and 94, respectively.

Fusarium oxysporum was found internallyseed-borne. Anatomical studies revealed that atleast one of the locations of Fusarium in seed isthe hilum region. Cultivars varied in the extent ofcarrying seed-borne inoculum. Of JG-62, P-436,and Chafa, Chafa carried much less inoculumthrough the seed.

Seed treatment with Benlate-T (0.15 %) (Ben-omyl 30% and Thiram 30%) eradicated Fus-arium completely, and eliminated almost allother fungi.

Rhizobium Nodu la t i ngChickpea

Chickpea rhizobia are known to nodulate onlyCicer spp. At present there is no reliable methodof counting their numbers in soil. We can nownodulate chickpeas growing in axenic culture,using a sand:vermiculite root medium enclosedin 22- by 200-mm test tubes, and hope that such a system can be used for a soil dilution-plantinfection method of counting chickpea rhizobia.

We have 160 isolates from chickpea nodulescollected in India and western Asia, with 60 of

100

Figure 36. A promising row of chickpea in the breeding nursery where disease losses are heavy, as evidenced by blank areas in the field.

these verified as Rhizobium. There are largedifferences among these strains in nodulationand nitrogen fixation, and field-inoculation trialsindicate an interaction among strain, soil type,and host cultivar. When chickpea Rhizobium isabsent from the soil, seed inoculation produces a marked zonation of nodulation, with nodulesforming on, or close to, the primary root in a band from the seed to 12 to 15 cm below it. Insoils containing chickpea Rhizobium pop-ulations, nodules form initially on the primaryroot, but then also form on the secondary roots.

Nodulation and Nitrogen Fixation

Large areas of chickpea are sown following rice.

At ICRISAT Center, chickpea after rice re-sponded to inoculation with an increase innodulation, nitrogen fixation, and a 64-percentincrease in grain yield (Table 37). Uninoculatedplants formed virtually no nodules. Inoculatedplots yielded 1 800 kg/ha, as did plots given a large application of N fertilizer, indicating thatN2 fixation by nodulated plants was adequate.

Five cultivars of different duration were com-pared at ICRISAT Center. At only 17 days aftersowing, there were differences in nodule numberper plant. Nodule number and nodule weight perplant, as well as nitrogenase activity measured byacetylene reduction assay, increased up to the61st day but declined by the 81st day due tonodule senescence. Nodule weight per plant

101

Table 37. Yield of chickpea following rice at

ICRISAT Center, 1976.

Treatment Dry matter Grain

(kg/ha) (kg/ha)

Control 1483 1090

Inoculated+150 kg

N/ha 2391 1762

Inoculated 2679 1801

L.S.D. (0.05) 364 278

SE± 161 123

C.V. (%) 7.4 7.9

increased sixfold between the 27th and 61st days,

mainly through growth of already-formed nodu­

les. Nitrogenase activity over this same period

increased fourfold. Nodules were most active per

unit of nodule tissue at the time of the first assay

(at 17 days) and there was little difference among

fanes. The medium-duration desi cv 850-3/27

formed the largest number of nodules and the

most nodule tissues. From the 17th day on, it had

more than double the number of nodules of the

other cultivars. These nodules also had the

greatest nitrogen-fixing activity (43 µ moles

C2H

4/plant per hour at 61 days) and 850-3/27

was the only cultivar retaining any nitrogenase

activity at 81 days (5 µ moles/plant per hour).

This cutivar also yielded significantly more than

did the other four cultivars.

There was a marked diurnal periodicity in

nitrogenase activity, with activity increasing

rapidly from 0600 hours until about 0900, and

then declining rapidly, with a smaller secondary

peak at 1800 hours. Soil temperature apparently

had little effect on this pattern.

We surveyed the nodulation of 258 entries in

the working collection of chickpea germplasm

used in the breeding program at ICRISAT

Center. Nodule number and weight per plant

varied significantly among lines. A few cultivars

had high nodule numbers per plant at both the

18th and 50th days after planting. Some cultivars

could be identified as having consistently high or

low nodulation. Among cultivars, numbers ran­

ged from 8 to 97 per plant at 50 days after sowing,

with nodule dry weight ranging from 2 to 105

mg/plant. A survey of 200 germplasm lines

indicated that nodule number per plant was

greater at Hissar; nodules there also remained

active longer.

Nodules at ICRISAT Center were attacked,

apparently by insect grubs. The pest concerned

and the consequences of the damage will be

investigated.

C o m p a r a t i v e G r o w t h a n d

D e v e l o p m e n t

We compared the growth and development of a

number of chickpea cultivars at ICRISAT Cen­

ter and at Hissar. As usual, the yields at Hissar

were considerably higher than those at ICRISAT

Center. For example, the cultivar which perfor­

med best of all in our trials at ICRISAT Center,

cv 850-3/27 (which is of medium-duration),

yielded 2029 kg/ha; its yield at Hissar was 3413

kg/ha. Patterns of accumulation of dry matter

during the reproductive phase, in the entire shoot

system as well as in the pods of this cultivar, are

plotted in Figure 37 for the two locations, along

with corresponding climatic data. At ICRISAT

Center, the plants matured 48 days after flower­

ing began, but at Hissar during the same period

(although there was more accumulation of dry

matter in the shoot system than at ICRISAT

Center), almost no pod formation took place.

During this period the maximum and minimum

temperatures at Hissar were considerably lower

than at ICRISAT Center. At Hissar, the cooler

temperature apparently suppressed pod-set and

consequently favored vegetative growth; as the

temperatures at Hissar rose, both pod develop­

ment and the rate of growth of the entire plant

increased. The maturation of the plants at Hissar

took place as the temperature and evaporation

rose to levels higher than those encountered

during the chickpea-growing season at

ICRISAT Center. These and other observations

described below indicate that senescence and

maturation of chickpeas are accelerated by hig­

her temperatures and/or higher rates of

evaporation.

102

Figure 37. Dry-matter production and pod weight per plant from onset of flowering to maturity in chickpea cv 850-3/27 grown at ICRISAT Center and Hissar. Flowering began on 10 Dec 76 at ICRISAT Center and on 11 Jan 77 at Hissar - 59 and 74 days after sow-ing, respectively.

Effects of Shading

The effects of shading chickpeas with clothshades placed horizontally above the cropcanopy throughout the reproductive period wereinvestigated. At ICRISAT Center and at Hissarthe shading resulted in a striking delay in senes-cence and maturity. At Hyderabad, shades which

cut out 50 percent of the light gave no reductionsin yield; on the contrary they generally broughtabout small increases. Thicker shades, which cutout 80 percent of the light, resulted in significantdecreases in the yield of early or medium-duration cultivars planted early in the season,but led to no significant reduction in the yield ofcultivars which were maturing late in the season.The senescence-delaying effects of the shades andthe actual stimulatory effects of 50-percent shad-ing at ICRISAT Center suggest that the shadeswere protecting the plants from stresses as-sociated with the heating of the leaves by thesunlight. The stress involved could either bemoisture stress (increased by the radiation loadand consequent transpiration from the leaves)and/or a direct heat stress. We found that theeffects of shading were similar with and withoutirrigation of the plants, which makes it less likelythat the shading effect can be explained in termsof a reduced moisture stress. Also we found thatshading had much less senescence-delaying effectand no yield-stimulating effect on winter-grownpigeonpea. (Pigeonpeas are known to be moretolerant to high temperatures than are chick-peas.) Taken together, these results suggest thatat least at ICRISAT Center the senescence andmaturation of chickpea are accelerated by heatstresses resulting from the radiation load on thecanopy.

Source-sink Relationships

Effects of defoliation. Field experiments werecarried out at ICRISAT Center this year toinvestigate the effects of different degrees ofdefoliation throughout the reproductive phaseon yield. Results (mean values shown) for cvs850-3/27 and Annigeri with and without i rr i -gation are plotted in Figure 38. The yield wasreduced roughly in proportion to the degree ofdefoliation in both irrigated and nonirrigatedplants. This yield reduction was largely owing toa reduction in pod number per plant, althoughthere was also a reduction in 100-seed weight,especially after the higher degrees of defoliation.Seed number per pod was affected only very

103

Figure 38. Effect of different degrees of defoli-ation throughout the reproductive phase on yield and yield components of chickpea cv 850-3/27 and Annigeri (mean values shown) with and without irrigation.

slightly by the treatments. Similar results wi th cvJG-62 were obtained in a separate experiment.

These results suggest that leaf area was l imit-

ing yield. This conclusion contrasts with theresults obtained this year in similar experimentswith pigeonpea (see p 89 of this report), and alsowith results obtained last year (1975-1976) withchickpea, when we found that up to 50 percent ofleaves could be removed without affecting theyield significantly. Last year the experimentswere conducted with plants which were plantedlate. It seems possible that the discrepancybetween the two sets of results may be explainedin terms of the effects of moisture and/or heatstresses to which late-planted plants would bemore exposed.

Effects of flower removal. This year in experi-ments with early sown plants at ICRISAT Cen-ter we found that all the flowers could beremoved from the plants for up to 4 weeks afteronset of flowering without causing a decline inyield, if the plants were irrigated. Nonirrigatedplants had a significantly reduced ability tocompensate for flower removal. Both irrigatedand nonirrigated plants were able to compensatemore or less completely for the removal of up to50 percent of the flowers throughout the repro-ductive period. These compensations involvedboth an increase in the number of pod-bearingnodes per plant and an increase in the number ofseeds per pod.

Last year (1975-1976) we found that flowerremoval from late-sown plants resulted in a decline in yield and also in total dry matterproduction. The differences between the 2 years'results might be because of the greater moistureand heat stresses to which the later-sown plantswere exposed; these stresses would have limitedtheir ability to continue growing and settingpods.

The "double-podded" Character

Although most chickpea cultivars produce a single flower per node, some produce two flowersand are capable of setting two pods per node.Two pods are formed at only some of the nodes,usually at the more-basal earlier-formed nodes.We confirmed last year's observations that ex-pression of this "double-podded" character is

104

100-secd weight

i r r i ga ted L.S.D.(0.05),

non i r r iga ted

30

20

10

0

40

30

20

10

0

8

6

4

2

0

non i r r iga ted

Y i e l d per p lan t

D e f o l i a t i o n (%)

0 20 40 60 80 100

i r r iga tedL .S .D.(0.05)

Pods per p lan t

i r r i ga ted

L .S .D.(0.05)

non i r r iga ted

strongly influenced by environment. At Hissaronly 3 percent of the pod-bearing nodes of cv JG-62 were double-podded; on early sown plants ofthe same cultivar at ICRISAT Center, 32 percentof the pod-bearing nodes were double-podded,whereas on later-sown plants this proportionincreased to 45 percent. However, at ICRISATCenter the absolute number of double-poddednodes per plant was the same on the early andlater-sown plants; the later-sown plants simplyhad fewer single-podded nodes. The percentageof double-podded nodes at ICRISAT Center wasreduced in plants which were defoliated duringthe reproductive phase. This indicates that theexpression of the double-podded character isinfluenced by the source-sink balance within theplant, but we do not know through whichmechanism expression of this character is in-fluenced by environment.

Last year we found - by cutting off the secondflower at each flowering node of double-poddedcult ivars-that the double-podded characterconferred a small but significant advantage inyield over plants with only a single pod at eachnode. We obtained the same result again this yearin experiments carried out at ICRISAT Center.

Position of Pods

Last year in an experiment at ICRISAT Centerwe found that moving the pods into an exposedposition above the leaves had no effect on yield.This year we carried out the same experimentwith four cultivars at Hissar and again found nosignificant effect of pod-position on the yield. Wecan conclude that the "exposed-pod" characteris unlikely to be useful in breeding for higheryields.

Cultivaral Differences

Response to spacing. In farmers' fields, standsof chickpea are often poor and patchy. Underconditions where plant stands are variable, cul-tivars which are more plastic in their response tospacing will yield better than those which are lessplastic. We have studied the plasticity of a rangeof cultivars, using three different methods: by

Figure 39. Population density studies identify chickpea cultivars that will compen-sate for the poor stands frequently encountered by SA T farmers.

comparing yields as the row spacing is systemati-cally increased (in a " f an " design); by comparingyields produced by the border rows of the plotswith yields produced by central rows; and instandard plant-population trials with populationdensities ranging from 8 to 100 plants/m2. Withall methods cultivaral differences were found atboth ICRISAT Center and Hissar, but theresults from the border-row method did notagree well with the other methods. The differ-ences in plasticity were sufficiently large tosuggest that cultivaral performances in standardyield trials may not give reliable indications ofyield at lower populations. In a plant populationtrial at Hissar, for example, cvs G-130 and L-144at the standard population of 33 plants/m2

yielded 3 087 and 3 427 kg/ha respectively, whileat 8 plants/m2 their respective yields were 3 043and 2 200 kg/ha; cv L-144 was much less plastic inresponse to population density than was cv G-130, and the relative performance of the twocultivars was reversed.

Seed size and seedling growth. We comparedthe growth of the seedlings of 23 cultivars fromseeds ranging in 100-seed weight from 4 to 30 g.There was a close and highly significant re-lationship between 100-seed weight and the dry

105

weight and leaf area of the seedlings at 16 daysafter sowing (r = 0.82**), but by 30 days aftersowing the relationship was less close(r = 0.57**) (Fig 40). The larger-seeded cul-tivars also had larger leaves (r = 0.86** for therelationship between maximum area per leaf on80-day-old plants and 100-seed weight). Therewas no significant relationship between 100-seedweight and total dry weight at harvest (r = 0.07)or yield (r = 0.13).

Figure 40. Relationship between 100-seed weight and total dry weight of seedlings of 23 chickpea cultivars.

Susceptibility to iron chlorosis. A number ofcultivars grown at ICRISAT Center showed a marked yellowing of the younger leaves duringthe vegetative phase. The plants recovered spon-taneously later in the growing season. The yel-

lowing symptoms could be relieved by sprays offerrous sulphate, indicating that the symptomswere those of iron chlorosis. In a trial with twoiron-chlorosis-susceptible cultivars we foundthat the sprayed plants gave 35 percent moreyield than the controls, showing a distinct yield-reducing effect of this undesirable cultivaralcharacteristic.

Response to moisture stress. We made a pre-liminary attempt to identify drought-tolerantcultivars of chickpea by growing 71 diversecultivars on soil of limited moisture-holdingcapacity (an Alfisol). Three sets of replicatedplots were irrigated during the growing seasonand three were not. Plants in the irrigated plotsmatured later and, on average, produced aboutthree times as much dry matter and yield as didthe nonirrigated plants, indicating that the latterwere subjected to quite severe moisture stress.

Duration of the cultivars (measured by thenumber of days to flowering) was significantlynegatively related to the yield under nonirrigatedconditions (r = - 0 . 3 9 * * ) , but was not signi-ficantly related to the yield under irrigation(r = 0.04). The ratio of the yield without irr i-gation to the yield with irrigation (which can beregarded as an index of drought tolerance) wassignificantly negatively related to the number ofdays to flowering (r = - 0 . 3 3 * * ) . In otherwords, the earlier cultivars tended to be moredrought-tolerant and to yield better under non-irrigated conditions than did the later cultivars,but this tendency explained only a small part ofthe cultivaral differences.

Although the correlations between yield perplant with and without irrigation (r = 0.60**)and between yield per plant without irrigationand the "drought-tolerance index" (r = 0.48**)were significant and positive, they are not strongenough to effectively iden t i f y -on the basis ofyield under conditions of adequate moisture -cultivars which would yield well under droughtconditions. Nor can plants which are most"drought tolerant" (as judged by ratio of yieldwithout to yield with irrigation) be reliablyidentified simply by the yield produced undermoisture stress.

106

300

250

200

150

100

50

0

600

500

400

300

200

100

00 4 8 12 16 20 24 28 32

Seed weight ( g / 100 seeds)

B. 30 days after sowing

A. 16 days after sowing

Growth Studies in ControlledEnvironments

Our field studies on growth of chickpea are beingsupplemented by work in the Plant EnvironmentLaboratory at Reading University in England.Under a collaborative project financed by theMinistry of Overseas Development, R. J. Sum-merfield and his colleagues have completed in-itial experiments in which several cultivars weregrown to maturity in controlled environments.With appropriate light sources, the plants closelyresembled those grown in the field in India. Atpresent an experiment involving different dayand night temperatures and daylengths is inprogress; this has already revealed very markedand distinct effects of both day and nighttemperatures.

Insect Pest D a m a g e

Most of the current chickpea entomology re-search is concerned with Heliothis armigera (Hub.) which is known to be the most damagingpest of this crop throughout Asia and Africa. Inthe 1976-1977 season at ICRISAT Center thisinsect was again the only really important pest onchickpea; during early growth of the plants thelarvae were found feeding on the leaves and laterthey destroyed flowers and bored into the greenpods.

In an attempt to quantify the yield loss causedby these insects on the kabuli and desi types,plots in which the plants were sprayed withendosulfan and from which the larger H. ar-migera larvae were removed by hand were com-pared with nontreated plots. The percentage ofpods damaged, number of pods harvested, andseed yields are shown in Table 38.

Pest control gave a 55-percent increase in yieldfrom the kabuli type; most of this clearly resultedfrom the increase in number of pods harvested.In the desi type there was no final reduction inpod-set in the nonsprayed plots and the smallincrease in yield from pest control was associatedwith the reduction in pod boring. These resultsconfirm earlier observations that kabuli types are

Table 38. Comparison of pod damage and yieldin chickpea cultivars under sprayedand nonsprayed conditions atICRISAT Center, 1976-1977.

Borer-damaged Pods

pods harvested Yield

(%) (no) (kg/ha)

JG-62 (desi):Sprayed 1.9 19096 1408Not sprayed 8.2 20804 1324

L-550 (kabuli):Sprayed 2.5 12019 972Not sprayed 10.6 8153 626

more heavily attacked by H. armigera, parti-cularly during the flowering stage.

Pesticide Residue Analysis

D D T and endrin are the most commonly usedinsecticides for H. armigera control on chick-peas, but both of these chemicals have well-known disadvantages. We used endosulfan,which has fewer disadvantages, but its relativelyhigh cost will discourage most farmers. In-secticide use during the flowering and green-podstage introduces the possibility of toxic residuesin the harvested seed. Samples of seed fromendosulfan-treated plots were sent to the Tropi-cal Products Institute in England for toxic-residue analysis. Endosulfan sulphate levels ofup to 0.34 mg/kg of seed were detected; such a level may cause acceptance problems.

Screening for Susceptibility

It has been established that kabuli-type chick-peas are generally more susceptible to H. ar-migera attack than are the desi types and there issome evidence to suggest consistent differences insusceptibility within those types. This has en-couraged intensification of the work in develop-ing a screening program for "resistance" to H.armigera.

107

In an experiment to determine the effect of plotsize in detecting susceptibility differences bet-ween cultivars, trials with plots of 4.8 and 20.0m2 were compared. Results of these trials, inwhich 13 cultivars were tested show the smallerplot size to be at least equal in efficiency indetecting differences, both in percentage of borerdamage and in yield (Table 39). This was a welcome result, for it wil l permit additionalreplication without increased demand on seed orland. The susceptibility rankings of the cultivarswere in good general agreement in the two trials;C-235 and L-345, both desi types, were the leastsusceptible while the kabuli types suffered mostdamage.

In a preliminary attempt to screen the germ-plasm collection for susceptibility, 8894 lineswere sown in unreplicated small plots, with twocheck cultivars alternating after every 20 lines. H.armigera attacks were scored at the green-podstage, and the percentage of damage was re-

corded when the pods were harvested. There wasa fertility gradient across the field and H. ar-migera was more common in the better growingareas. Pod damage ranged from 0 to more than50 percent in the check plots, so the individualrecords from the unreplicated germplasm linescan have little value, except perhaps in indicatingthe more obviously susceptible. Of the checkplots, 28 percent of the C-235 (n = 219) and 19.5percent of the BEG-482 (n = 221) were free fromH. armigera damage, but only 11 percent of the8 629 germplasm lines harvested had no damage.This is interpreted as an indication that C-235,and to a lesser extent BEG-482, are less suscep-tible than the majority of the germplasm cul-tivars. This was not unexpected, as both checksare well-adapted desi cultivars that have beengrown generally without pesticide protection.

Although the natural infestation of H. ar-migera in the screening trials was augmented bythe release of laboratory-bred moths, infes-

Table 39. Comparison of pod-borer damage and yields of 13 cultivars tested in small (4.8 m 2) and large(20.0 m2) plots in randomized blocks with four replicates.

% Borer damagea Yield

Cultivars Small plot Large plot Small plot Large plot

(arcsin scale) (g/m2) (g/m2)(arcsin scale) (g/m2) (g/m2)

C-235 4.9 3.4 82.5 118.0BR-70 4.4 10.6 47.7 68.8L-345 3.0 2.6 60.4 74.6L-2937 7.0 6.9 79.8 91.4850-3/27 18.1 12.6 59.8 139.3JGC-1 7.5 8.3 76.9 113.0IC-6037 4.9 6.6 73.5 95.0RS-11 6.1 7.1 81.0 84.2NP-34 12.0 8.1 69.4 71.4P-3090 18.2 16.6 68.5 55.2NEC-143 13.3 11.0 44.4 79.1Rabat 13.6 14.5 61.3 72.2IC-682 9.5 9.3 77.3 86.0

S.E.± 9.33 15.35c . v . ( % ) 21.5 22.0 27.4 34.8

apercent borer damage, bated on arcsin transformation.

10S

tations were low and uneven. Most of the plantsthat were free from attack were probably escapesrather than "resistant." Improvements in screen-ing techniques are of obvious priority. Attemptsto screen plants in net houses were abortive, forthe plants grew poorly and were etiolated. Tech-niques involving replication of small plots infields of uniform fertility and the utility ofinoculating individual plants with eggs or younglarvae wil l be evaluated.

Another approach is to identify charactersinvolved in reducing the susceptibility of plants.Chickpea exudes an acidic liquid throughout itsgrowing period and it has been generally as-sumed that this exudate deters many insect pests,although one report suggests that it attracts H.armigera! Collaboration with the Centre forOverseas Pest Research in the United Kingdomhas been established; chemists are analyzing theexudates. A preliminary analysis has shown thatthe exudate contains aconitic acid, which isknown to be a feeding deterrent for some insects.It may be possible that these analyses will beuseful in explaining, or even detecting, suscepti-bility differences conferred by the exudatecomposition.

Nut r i t i ona l Qua l i ty

Estimation of Protein Content

Two rapid methods of estimating nitrogen con-tent were compared with the micro-Kjeldahl(MKJ) method as a standard. With 98 pairs, theTechnicon Auto Analyzer and dye-binding ca-pacity methods gave results (r = 0.99 and 0.98,respectively) highly correlated with M K J . Thebiuret method, developed for rapid determi-nation of protein in cereals, gave slightly lowercorrelations and considerably higher standarderrors than the other methods. We have con-cluded that either the Technicon Auto Analyzeror dye-binding capacity method is satisfactory.

The distribution of protein content (N x 6.25)in 6014 germplasm samples is shown in Figure41. Content ranged from 14.3 to 30.9 percent.

Total Nitrogen and Nonprotein Nitrogen

Nitrogen in seed occurs in protein molecules andalso in free amino-acids, other organic com-pounds, and in inorganic compounds. Esti-mation of protein content by multiplying nitro-gen content by a particular factor is subject to

error to the extent that nonprotein nitrogen ispresent in the sample. We analyzed 98 samples ofchickpea and found the percentage of nonpro-tein nitrogen to range from 0.16 to 0.73. Ex-pressed as percentage of total nitrogen, thenonprotein nitrogen ranged from 5.8 to 16.5. Thepercentage of nonprotein nitrogen in the sampletended to increase as protein increased (Fig 42).These preliminary results indicate the need foradditional studies of seed nitrogen.

109

Figure 41. Range of protein content of 6014 chickpea germplasm samples.

Number of samples: 6014Protein %:Range 14.3- 30.9Mean 20.1

1350

1200

1050

900

750

600

450

300

150

15 17 19 21 23 25 27 29 30Protein content (%)

Figure 42. Relationship between total nitrogen and nonprotein nitrogen in chickpea.

lower than in kabuli cultivars, which rangedfrom 56.1 to 59.7 percent. Other consistentdifferences included higher crude fiber content indesi cultivars than in kabuli cultivars, and lowerfat content in desi cultivars.

In ternat iona l Coopera t ion

Furnishing of superior breeding material to localbreeding programs is the most hopeful way ofmaking progress in increasing chickpea yields. Inthe past year we distributed screening nurseriesof 100 early and 100 late-maturing advancedlines from our breeding program. Results havebeen received on the early group from 11 lo-cations and on the late group from 13 locations,all in India. Results from nine other countrieswere not received in time for analysis.

l imit ing Amino Acids

As in most legumes, the sulfur-containing aminoacids-methionine and cystine-are usually de-ficient in chickpea. We analyzed 30 samples witha range of nitrogen content from 2.40 to 4.85 percent for total sulfur. Although differences insulfur percentage were small, there appeared tobe an inverse relationship between percentage N and percentage S (Table 40). The sulfur/nitrogenratio expressed as percent emphasizes the inversetrend.

Similar results were obtained when methio-nine content was determined by microbiologicalassay for 78 samples. There was a negative andsignificant correlation between percent proteinand methionine as percent of protein (Fig 43).These results indicate that the additional proteinformed in high-protein lines may consist offractions which are low in the sulfur amino acids.

Composition of Some Chickpea Cultivars

Five kabuli and five desi.cultivars were analyzedfor protein, starch, soluble sugars, fat, crudefiber, and ash content. Protein (N x 6.25) rangedfrom 19,4 to 24.4 percent, with no differencebetween desi and kabuli (Table 41). Starchcontent in desi cultivars (47.2 to 52.5%) was

Figure 43. Relationship between protein and methionine contents in chickpea.

110

Protein (%)17 19 21 23 25 27 29 31

1.6

1.5

1.4

1.3

1.2

1.1

r = -0 .451* *+ 1.621 -0 .015 x

Total nitrogen (%)2.5 3.0 3.5 4.0 4.5 5.0

0.75

0.65

0.55

0.45

0.35

0.25

r =+ 0.802**

= -0 .232 + 0.165 x

Table 40. Total sulfur and nitrogen in 30 chick-pea germplasm lines.

N S S/N

(%)

7-07-08-0

High N lines (12)Medium N lines (8)Low N lines (10)

3.85-4.85 0.13.05-3.84 0.12.40-3.04 0.1

(%)

7-07-08-0

.21 3.7-5.0

.23 4.5-6.9

.24 6.5-8.8

Table 41. Chemical composition of some chickpea cultiv ars, expressed on a moisture-free basis.

Soluble CrudeCultivar and Source Protein Starch sugars Fat fiber Ash Total

(%)(%)Desi types

JG-62 (India) 22.0 47.2 4.1 3.8 10.1 3.1 90.3Pyroz (Iran) 19.7 52.5 5.6 3.5 7.1 3.2 91.4NEC-240 (USSR) 22.3 47.5 4.0 3.8 7.9 3.1 88.6NEC-1639

(Pakistan) 24.0 47.5 4.2 4.0 10.0 3.4 93.1USA-613 (USA) 20.9 48.0 4.8 3.8 7.4 3.2 88.1

Kabuli typesL-550 (India) 19.4 59.7 4.3 5.7 3.2 3.1 95.4NEC-34 (Iraq) 21.0 57.9 5.5 5.3 6.0 3.1 98.8Lebanese Local 21.5 57.6 5.2 5.0 5.6 3.1 98.0

(Lebanon)NEC-10 (Jordan) 21.4 58.6 4.6 5.3 5.1 3.3 98.3NEC-143 (Sudan) 24.4 56.1 5.3 4.8 5.6 3.1 99.3

We considered each location as a replicate,and analyzed the results received. One line in theearly group and five lines in the late group yieldedsignificantly higher than the check when aver-aged over all locations. Lines with high averageyields from each set will be offered for testing inthe all-India coordinated trials. Lines perform-ing best at some of the locations are being furthertested by the local breeders.

Other sets of material distributed include theobservation nursery containing cultivars for useas parents in crossing, yield trials of desi andkabuli cultivars, and disease observation nurs-eries. Results of the trials and disease obser-vations are summarized and distributed tocooperators.

Breeding populations supplied on request areusually sent in the F3 generation for localselection. A summary of all materials supplied tocooperators over the past two years is listed inTable 42.

Communication is supplemented by exchangeof visits with cooperators. We have visited 15 ofthe 28 countries with whom we are cooperating,and have had scientists from 13 of the countriesvisit ICRISAT.

Two trainees from Ethiopia and one fromSudan completed a course in chickpea breeding.

Look ing Ahead in ChickpeaImprovement

A catalog of chickpea germplasm will be pub-lished soon. Fill ing of gaps in the collection wil lcontinue, with high priority areas being theBundelkhand region, Rajasthan, and MadhyaPradesh in India, as well as Ethiopia, Afghanis-tan, Pakistan, Bangladesh, Iran, and Turkey.Work will be intensified on evaluation andutilization of wild Cicer species.

Additional emphasis will be given to breedingmore productive plant types, and responsivenessto fertilizer plus irrigation wil l be investigated inlate-planted material at Hissar.

Field screening for Ascochyta blight wil l beinitiated to supplement laboratory tests, andfield-screening techniques for resistance to stuntwill be investigated.

Selected germplasm lines wil l be resurveyed atICRISAT Center and Hissar to identify thosewhich form large numbers of active nodules. A

111

Table 42. Test and breeding material furnishedto cooperators for the 1975-1976 and1976-1977 seasons.

Material Countries Loca-tions

Sets

Yield trialsScreening nurseriesF3 bulk populationsDisease nurseries

(no)

2824187

(no)

54632816

(no)

1761528716

breeding program to increase nitrogen fixationwill involve the best lines as parents. We wil lattempt to elucidate the factors causing the rapiddecline in nitrogenase activity during pod-fill atICRISAT Center.

Highly effective Rhizobium strains will betested as inoculants under field conditions. Weplan to measure the amount of nitrogen fixed bythe chickpea crop and to determine the residualeffect on a following cereal crop.

Techniques for screening breeding material forpod-borer susceptibility will be investigated.

112

More work will be carried out to identifydrought-tolerant cultivars and to develop reli-able field-screening techniques for this character.Methods for identifying salt-tolerant cultivars,using chambers containing artificially salinizedsoil, and for screening cultivars for the ability togerminate under conditions of limited soil mois-ture wil l be investigated. Work will be carried outat Hissar to compare growth and developmentwith and without irrigation before sowing.

At ICRISAT Center, the influence of heatstress and moisture stress on source-sink re-lationships and yield will be investigated inexperiments involving early and late plantingswith and without irrigation. Further efforts willbe made to develop a simple field method formeasuring plasticity differences in cultivars, asreflected by their response to differentpopulation-densities.

Characterization of the protein fractions inchickpea will be done, and rapid testing tech-niques for the desirable fractions will beinvestigated.

International cooperative activity will be in-tensified, both in the area of chickpea improve-ment and in expanding the training componentof the program.

Groundnut ( Arachis hypogaea)

Groundnuts - a s sources of human and animal feeds and as an economic crop - are one of the mostimportant legumes of the semi-arid tropics. Of the world's total production, two-thirds is produced inthe SAT. Yields are low, however, averaging only 500 to 800 kg/ha. Yields in the United States andsome other areas can average around 2500 kg/ha and in some areas will often be 5000 kg/ha.

ICR ISAT Goals

ICRISAT's groundnut program can be summarized into three broad-based objectives:To assemble, maintain, and screen a world collection of cultivated and wild Arachis material.To seek, through breeding programs, to increase yields and incorporate resistance to importantpests and diseases. Improvement of quality is also an important goal.To provide introductions and segregating populations to all groundnut breeders in the SAT.

I C R I S A T Center

The soil and climatic conditions at ICRISAT Center are ideal for groundnut research applicable toSAT agriculture. In addition to the major effort conducted under the rainfed situation of the normalseason, irrigated crops will be produced during the postrainy and hot dry seasons in order to facilitatethe breeding program.

113

O I L S E E D

G r o u n d n u t

The groundnut improvement program commen-ced in Apr i l 1976 with research in breeding,microbiology, pathology and virology, andgermplasm. In November a mycologist wasrecruited and work started on screening tech-niques for leafspot and rust fungi. In 1977 anentomologist was recruited, and we anticipatethat by early 1978 physiological and cytogeneti-cal work wil l commence. Although much of theyear has been necessarily spent in organizing andobtaining equipment, considerable progress hasbeen made - particularly in virology and obtain-ing germplasm, in spite of strict quarantinerestrictions. Much of our disease-resistant ma-terial has now been released from quarantine,and we expect an increase in the tempo of thebreeding program.

ICRISAT's groundnut program aims to in-crease the low yields (around 800 kg/ha or less) ofgroundnuts obtained by small-scale farmers inthe SAT, mainly by incorporating resistance tothe prevalent diseases, such as leafspots and rust.Physiology research will concentrate on identify-ing sources of drought tolerance, because unreli-able rainfall patterns also contribute to lowyields.

Germplasm

The International Board for Plant Genetic Re-sources (IBPGR) has nominated ICRISAT as a major germplasm center for Arachis. Priority hasbeen given to assembling the extensive col-lections of Arachis hypogaea already existing atresearch stations in India. Thanks to the cooper-ation of Indian institutions, we feel this task hasbeen largely accomplished during the past year.Details of these collections are given in Table 43.

There will be many duplicates in this collectionand in rainy season 1977 the total collection willbe planted out according to botanical type tohelp identify and eliminate duplicates. Syste-matic controls to facilitate yield evaluation willalso be planted. Many of the accessions from

some centers do not have sufficient backgrounddata on their pedigree or origin. The accessionregisters at the National Bureau of Plant GeneticResources, New Delhi, were therefore searchedand all the information on groundnut importsfrom 1947 to 1976 were extracted and published.This will assist in relating original import num-bers (EC numbers) to local germplasm col-lections. Collection of local material has also

Figure 44. Groundnut germplasm accessions at ICRISAT Center.

Table 43. Accessions of A. hypogaea from re-search centers in India.

CollectionNo. Source Accessions

1 Kadiri 10972 Karimnagar 2733 Ranchi 2234 Jalgaon 2455 Pantnagar 11

6 Rajasthan 587 Tindivanam

(AICORPO) 4638 Junagadh 11599 Tindivanam (OES) 29

10 Coimbatore 29

11 Ludhiana 49612 Akola 11013 Amravati 16014 Pollachi 29715 Bombay (BARC) 916 Mahabaleshwar 4

117

Groundnut8765432101975 1976 1977

been undertaken, particularly in those areaswhere improved cultivars have not been released.

Import of exotic cultivars from major germ-plasm centers abroad has been necessarily slow,due to the strict plant quarantine measures toprevent the spread of new seed-borne pathogensinto India. At present, imported seeds are plan-ted in a screenhouse at the Rajendranagarquarantine station and inspected regularly fordisease symptoms. After approximately 6 weeks,young healthy plants are released to us fortransplanting in the postquarantine area atICRISAT Center. Fortnightly quarantine in-spections are carried out until the plants mature,and their produce is then released to us. Even so,during the year approximately 750 cultivars havebeen released to us and a further 90 accessionsare in the postentry quarantine block. A further900 exotic cultivars are awaiting primaryquarantine clearance. Among the importantmaterial now released are two cultivars with rustresistance (a further 14 rust-resistant lines are inthe postentry quarantine area), two breedinglines with resistance to Aspergillus flavus, mate-rial resistant to leafspots, and an accession withextreme earliness. We have managed to have thisbreeding material released by asking for priorityprocessing in relation to normal germplasmresources. Several wild Arachis species, includingthose with disease resistance, have also beenreceived and established.

During rainy season 1976 some 2 000 cultivarsfrom Indian research centers and 331 exotic lineswere planted and evaluated. Due to the erraticrainy season and abundance of fungal and viral

diseases, the collection was evaluated for ad-ditional sources of resistance to rust andleafspot-but without success. A dry periodduring August and September allowed us toscore for drought tolerance. Thrips counts werealso taken on all cultivars; 21 lines with littledamage caused by this pest (and virus vector)were identified for further study. In the 1976-1977 postrainy season 178 new exotic cultivarswere planted in addition to the 331 harvestedfrom the rainy season crop. The yield from thiscrop was higher than had been obtained pre-viously and the collection was also screened forreaction to bud necrosis virus.

Groundnut germplasm was supplied to re-search workers in India and several other nations(Table 44).

Breeding

The main breeding programs are aimed at theincorporation of disease resistance into high-yielding commercially accepted cultivars. As wewere totally dependent on the importation ofdisease-resistant parents, we had to awaitquarantine clearance of such material before wecould proceed with our main breeding objectives.In the interim a team of operators was trained inemasculation and pollination techniques, usinggermplasm from local sources. Plants were raisedinitially in pots kept in the open; despite plantprotection measures, bud necrosis virus seriouslyaffected the rate of successful pollinations. A newset of parents was raised in the new screenhouse

Table 44. Groundnut germplasm lines supplied to rese arch agencies in India and other nations during1976-1577.

Institution Location Entries

Agricultural Research InstituteDepartment of Primary IndustriesAgricultural Research CentreAgricultural Research StationUniversity of West Indies

Karimnagar, A.P. IndiaBrisbane, AustraliaSemongok, MalaysiaMaha Illuppallama, Sri LankaSt. Augustine, Trinidad

236364625

118

Figure 45. Field hybridization techniques with groundnut.

completed in December 1976, and with humid-ifiers to provide high humidity at the time ofpollination, we achieved a success rate of 23percent - which is still less than expected forgroundnuts. Although the constraints have notbeen fully identified, recent figures indicate a further rise in our percentage of successful

pollination. As we can achieve large numbers ofpollinations, we have planted over 300 seedsfrom these initial crosses together with theirparents to eliminate selfs from the F1 plants.Many Indian centers do their crosses on field-grown plants so we have planted cultivars onwide-spaced ridges to compare field and screen-

119

house methods. These cultivars represent dif-ferent botanical varieties having high yield andother desirable characters.

Breeding for resistance to the leafspot fungicaused by Cercospora arachidicola and Cercospo-ridium personation is under way in collaborationwith scientists at Reading University, UK , andthe British Ministry of Overseas Development.At Reading triploid hybrids were producedbetween the tetraploid groundnut, Arachis hy-pogaea, and two wild diploids, A. chacoense andA. cardenasii, which are highly resistant to C.arachidicola and C. personatum respectively.Another diploid, A. sp. H L K 10, reputedlyresistant to both fungi, was also crossed with A.hypogaea. These sterile triploids were treatedwith colchicine at Reading to produce fertilehexaploids. Cuttings of these hexaploids, plussome of the original triploids, were sent toICRISAT and after rooting in a quarantinescreenhouse they were held planted. A highnatural infection of C. personatum was producedby interplanting the hybrids with highly suscep-tible cultivars and spreading infected leavesaround the plants. A l l the plants were scored forreaction to leafspots and other diseases, as well asfor flower and pod production. Plants appearingto be highly resistant were selected for furtherscreening and backcrossing to the cultivatedgroundnut. The pod production on the hexa-ploids varied from zero to more than a hundredpods per plant.

We have so far harvested approximately 100seeds from crosses involving the two rust-resistant cultivars, PI 259747 and PI 298115, andhigh-yielding but rust-susceptible parents. Four-teen additional resistant F3 lines (FESR 1-14)derived from a natural cross between PI 298115and an unknown pollen donor are presentlybeing assessed in the postentry quarantine area.As these lines are still segregating, each plant willbe tested for rust resistance, yield, and qualitycharacters.

Introductions released from quarantine andcurrently being used in other hybridization pro-grams include breeding lines with resistance toAspergillus flavus and others with earliness.

As conventional groundnut breeding pro-

grams are slow, we are investigating methods ofreducing the time needed to advance segregatingpopulations. Ethephon, an ethylene-producingcompound formulated as a dust, is currentlybeing used to break seed dormancy in freshlyharvested normally dormant cultivars. In thelaboratory more than 90 percent of the freshlyharvested dormant seed germinated within 24hours of harvest. A field trial in postrainy season1976 with freshly harvested seed from the pre-vious rainy season gave similar results (Fig 46).

Figure 46. Seedling emergence in Ethephon-treated dormant groundnut (postrainy season 1976).

We have developed a close working relation-ship with AICORPO, the All-India CoordinatedResearch Project on Oilseeds. We planted fournational trials under nonirrigated rainy seasonconditions at ICRISAT Center in July 1976 andreplanted two of them on our own initiative inDecember 1976 under full irrigation for com-parison purposes. The results (Table 45) showthat rainy season yields, shelling percentages,and kernel weight this year were much lower thanthose of the postrainy season due to leafspots,rust, and erratic rainfall in August and Septem-ber 1976.

120

Trea tmentsEthephon Ta lc

(%) (%)100 : 075 : 2550 : 5025 : 750 : 100

14 18 22 26 30 34 81

Days after p lan t i ng

140

120

100

80

60

40

20

0

Table 45. Performance of cultivars in cooperative n ational groundnut trials [AICORPO trial CVT(Bunch)] at ICRISAT Center during rainy and postrai ny seasons, 1976.

Mean time to 50%flowering Mean yieldb

Shellingpercentage 100-kernel weight

Cultivars Rainya Postrainya Rainy Postrainy Rainy Postrainy Rainy Postrainy

(days) (days) (kg/ha) (kg/ha) (°/o) (%) (g) (g)

Robut 33-1 29.6 40.8 1711 3715 55.0 79.2 32.18 54.58X14-4-B-19-13 26.3 38.8 1017 3173 60.0 72.3 32.62 42.44Ah-8189 27.0 38.8 1004 2448 62.5 77.7 22.02 31.82Argentinec 27.5 38.3 962 3025 64.5 77.3 22.31 40.54Pol-2 27.3 39.5 933 3063 64.0 74.2 27.30 34.42

X9-2-B-25-B 26.5 38.0 904 3440 64.0 74.2 25.48 43.86X14-4-3-8-B 26.8 38.0 895 2336 59.0 70.8 27.65 43.54FSB-7-2 27.0 38.0 895 3 333 62.0 78.6 25.75 36.72OG-69-6-1 26.5 38.3 863 3490 66.0 74.2 26.52 32.22MGS-7 26.8 38.5 855 2992 64.0 76.8 24.24 32.86

OG-71-3 26.5 38.0 838 2817 58.0 74.7 26.17 36.33OG-1-13-3 26.8 39.3 835 3 344 64.0 77.6 27.83 44.28NG-268 27.3 39.0 804 3317 52.0 78.7 23.54 29.82MGS-8 27.0 38.5 802 3525 68.5 76.3 25.76 42.75Ah-8254 27.5 38.3 754 3 242 60.0 78.3 22.45 40.35

99-5 26.5 39.0 738 3512 58.0 80.0 27.34 41.66JH-89 27.0 39.0 736 2988 68.0 74.2 24.95 45.86OG-3-24 26.8 37.8 723 3173 70.0 76.8 24.24 42.57148-7-4-3-12-B 26.0 38.5 715 3312 65.0 75.8 23.90 36.68MGS-9 26.8 39.0 698 3 709 62.0 77.8 22.32 30.26

FSB-7-5 27.3 39.0 684 2454 58.0 75.3 22.30 46.29Ah-8253 27.0 38.5 669 2769 66.0 75.8 20.10 34.22Tifspanc 27.0 38.0 479 2781 60.0 77.6 21.16 35.61

L.S.D. (0.05)C.V. (%)

19116.77

81418.55

aIn rainy season trials, cultivars reached maturity in 102 days; in postrainy season trials, 160 days were required.bMean yield over four replications.cArgentine and Tifspan were controls.

Pathology

Bud necrosis virus (BNV) was again observed infield plantings at ICRISAT Center in August1976, some 40 to 50 days after planting, and thedisease reached serious proportions. Surveyscarried out in Andhra Pradesh showed thatbetween 30 and 70 percent of the plants in some

fields were infected with the virus. This diseasehas been reported from all groundnut-growingareas of India. The first symptoms on newlyopened leaflets vary from small indistinct chlo-rotic rings to larger concentric rings interspersedwith green islands. Later necrotic streaks appearon the petioles and spread down the stems. Oftenthe terminal bud becomes necrotic and dies,

121

leading to proliferation of axillary shoots. Theplants become stunted and later-formed leavesare smaller in size and show varied symptomssuch as mottling, vein clearing, cupping, andcurling; they sometimes become filiform. Earlyinfected plants may die; even in less severe casesyield is greatly reduced. If kernels are formedthey are small and wrinkled with dark lesions onthe testa.

The virus is easily transmitted mechanically aslong as an antioxidant such as 2-mercap-toethanol is included in the phosphate buffer atpH 7.0. Several other antioxidants were testedbut 2-mercaptoethanol was superior in that morelocal lesions were produced on the cowpea assay

tissues were fixed in paraformaldehyde and 50 % gluteraldehyde and sent for electronmicroscopyto the United States, Australia, and Japan. Virusparticles were not observed. Various insects weretested as possible vectors; of these only thrips(tentatively identified as Scirtothrips dorsalis) successfully transmitted the virus.

Based on all the evidence available, it becameapparent that the bud necrosis virus closelyresembled tomato spotted wilt virus (TSWV)which has been reported only occasionally ongroundnuts in USA, South America, SouthAfrica, and Australia. Therefore, TSWV anti-serum was obtained from sources in three dif-ferent countries; peanut mottle virus (PMV) and

Figure 47. Effect of virus on groundnut seed quality in petri dishes.

host (Vigna unguiculata cv C152). Using infectedsap with the buffer and 2-mercaptoethanol, some16 genera of plants showed symptoms whenmechanically inoculated.

Physical properties of the virus were investi-gated using classical techniques and cowpea asan assay host. The thermal inactivation point(TIP) was calculated to be between 45 to 50 C and the dilution end point (DEP) was between1 0 - 2 and 1 0 - 3 . Healthy and freshly infected leaf

peanut stunt virus (PSV) antisera were alsoobtained from USA. Sheep red blood cells werecoated with antisera of TSWV, PMV, and PSV.Various dilutions of healthy and virus-infectedgroundnut leaves were reacted against tannedblood cells coated with different sera. The cellscoated with TSWV antibody gave titers up to1/400 for healthy and 1/3200 for infected leafextracts. PMV- and PSV-coated cells did notreact with either healthy or bud necrosis-infected

122

leaf extract. The titers obtained with infected leaf

extracts were eightfold those of healthy leaves,

indicating that TSWV antigens were present in

bud necrosis-infected leaves. Confirmatory evid­

ence was obtained from latex agglutination tests.

Latex particles (0.46 µ in diameter) coated with

TSWV antiserum were reacted with different

dilutions of healthy and infected leaf extracts.

Healthy leaf extracts gave a titer of 1/100 com­

pared with 1/400 in extracts from infected plants,

again suggesting the presence of TSWV antigens

in bud necrosis-infected plants.

During rainy season 1976, 331 cultivars from

USA were scored in the field for the presence of

BNV. No accessions were found to be virus-free,

and the incidence varied from 30 to 80 percent.

Further field screening was carried out on the

1976-1977 postrainy season crop but only four

germplasm lines (Acc. Nos. 2188, 2575, 2372,

and 2932) showed late infection, which may

indicate some field tolerance to the virus. One

row of Acc. No. 1107, however, remained virus-

free. A l l the field-scored lines and the unaffected

row of Acc. No. 1107 were then inoculated in the

laboratory, using the mechanical method per­

fected earlier. Apart from one plant of Acc. No.

2372, all accessions developed bud necrosis when

mechanically inoculated. Even when grafted

with infected scions, the single plant of Acc. No.

2372 apparently remained virus-free; the control

plants developed typical symptoms. This plant

has now been multiplied by vegetative cuttings

and will be retested as a possible source of

resistance to the virus.

Several other viruses collected from surveys in

India are also under investigation. One virus

caused stunting and produced oval-shaped chlo-

rotic spots on the leaves; the spots tended to fade

as the plant aged, leaving mild mottle symptoms.

There was no necrosis or proliferation of axillary

shoots. When infected scions were grafted onto

healthy plants, typical symptoms were produced

in 2 to 4 weeks. This disease has been tentatively

named chlorotic spot virus (CSV). Mechanical

transmission of the virus could be achieved by

using 0.05 M potassium phosphate buffer at pH

7.0 with added 0.02 M 2-mercaptoethanol. Good

assay hosts were French bean (Phaseolus vul­

garis) and cowpeas. Nicotiana rustica was also

used extensively for laboratory tests, as large

amounts of the virus could easily be recovered

from it. Some physical properties of the virus

have been determined. The virus was infective

after 72 hours' exposure to room temperature

(25-30°C) and also when held at 4°C. The

thermal inactivation point (TIP) of the virus was

determined to be around 80°C. Concentration of

the virus was achieved by centrifugation at

30 000 rpm after extraction in phosphate buffer

and clarification. Further purification was achi­

eved by two cycles of sucrose gradient centrifu­

gation at high speeds. Several different zones

were detected and the zone containing virus

particles was detected by performing infectivity

assays on French beans and cowpeas. Using the

purified virus, we are now in the process of

producing CSV antiserum. Fixed leaves were

sent to Australia for electronmicroscopy and

thin sections showed the presence of large in­

clusion bodies normally found in "caulimo

virus" infections. Copper mesh grids coated with

Formvar membranes were supplied to us from

Japan and our purified virus was placed on the

grids before being fixed and stained. Leaf dip

samples were also prepared. Electronmicroscopy

was carried out in Japan and icosahedral virus

particles of 40 to 50 nm in diameter were clearly

discernible. Further work is in progress to de­

termine the precise diameter of the particles.

Results indicate that this virus is extremely stable

and has a low sedimentation coefficient between

100 and 200 Svedberg units. It is thought that

CSV may belong to the group of viruses contain­

ing D N A .

Another disease has been tentatively named

"vein-banding disease" (VB). During surveys in

Andhra Pradesh, the incidence varied from 1 to 4

percent in most fields, but on a few farms up to 10

percent of the plants were infected. Diseased

plants have also been seen at ICRISAT Center.

Early symptoms show vein-banding on newly

emerged leaflets giving an. oak-leaf pattern. Leaf

size is considerably smaller and the tips flex

downwards. Internode length is reduced and this

is followed by shoot proliferation, profuse

flowering, and peg production. Although pegs

123

are numerous, only a few develop into maturepods and the kernel size is considerably smaller.Typical symptoms were observed in 3 to 4 weekson healthy plants grafted with diseased scions.Using the buffer and antioxidant described ear-lier, mechanical transmission was made to sev-eral host plants. Chenopodium quinoa showedchlorotic local lesions and the infection becamesystemic. On soya (Glycine max cv. Bragg),systemic mosaic mottling occurred. Whenmechanically transmitted to groundnuts, thevein-banding symptom was produced but noaxillary shoot proliferation, stunting, profuseflowering, or peg formation occurred and thesubsequently formed leaflets were chlorotic andmalformed. Mechanical transmission from C.quinoa and G. max back to groundnuts alsoproduced these atypical symptoms. Small quan-tities of seed from infected seed remained healthywhen germinated. The results at present indicate

Figure 48. A rust-resistant groundnut cultivar growing at ICRISA T Center.

that more than one agent may be involved in thisdisease. One component is a sap-transmissiblevirus. The profuse flowering, pegging, and axil-lary shoot formation are similar to symptoms

produced by mycoplasma or Rickettsia-likeorganisms.

Priority has been given to the two mostimportant worldwide foliage fungi ofgroundnuts - rust (Puccinia arachidis) and leaf-spots (Cercospora arachidicola and Cercospo-ridiumpersonatum). Rust has spread at an alarm-ing rate over the last few years to all the majorgroundnut-producing areas. However, little isknown about its biology, the presence or absenceof physiological races, or its methods of survival.Infected leaf debris collected in rainy season 1976and exposed to natural weather conditions wasassessed at weekly intervals for viability of theuredospores. At the time of collection, germi-nation was in the region of 70 percent, after oneweek this had fallen to 30 percent and in a furtherweek to 1 percent; after that, uredospores did notgerminate. In the 1976-1977 postrainy season,the results were similar. Only the uredial stage ofthe fungus was found, despite intensive searchesfor the telial stage which has been recorded inSouth America. None of the 37 weed plants and11 cultivated plants inoculated with rust pro-duced symptoms.

To assist the breeding program, a reliablescreening technique for detecting rust resistanceis urgently required. Plants 30 days old wereinoculated with uredospore suspensions andkept at 100 percent relative humidity at 25 to30°C for 48 hours before being transferred toambient conditions. Abundant uredosori wereproduced in 8 to 10 days. This technique will bestandardized to provide a high inoculum sourcefor field testing populations produced by thebreeders in their rust-resistance breedingprogram.

A detached-leaf technique for detecting sour-ces of resistance is also being investigated. Thistechnique will have the advantage of testing, witha known intensity of inoculum under controlledconditions, large amounts of material in a smallarea. During some initial tests, leaves which weredetached 5 days after unfolding were found to bethe most suitable. After detachment, the petioleswere placed in various culture solutions with orwithout additives. The most suitable media forkeeping the leaves in good condition were found

124

to be either Hoagland or Shrive and Robbins

solutions with added kinetin at 20 ppm or

benzimidazole at 20 to 60 ppm. The detached

leaves were sprayed with uredospore suspensions

and incubated under 100 percent relative hu­

midity and temperatures of 25 to 30°C for 2 days.

The leaves were then maintained at 70 to 80

percent relative humidity and cool temperatures.

Sori began to develop 8 days after inoculation;

they soon darkened and enlarged before ruptur­

ing and releasing masses of uredospores. Com­

parisons of the detached-leaf technique with

whole-plant inoculation show that disease de­

velopment is similar in either case.

The leafspot fungi cause worldwide reductions

in groundnut yields each year. In India and parts

of Africa, C. personatum is the dominant species;

in USA and other parts of Africa, C. arachidicola

is more important. We need to develop reliable

screening techniques to assist breeders in their

selection of resistant segregates. Although it has

been reported that both fungi can be cultured on

artificial media to produce conidiospores, spore

production on the recommended media was not

sufficient for our purposes. Other media, which

had been recommended for different species of

Cercospora, were tried but success was limited.

Recently we have gained much more satisfactory

results with malt agar supplemented with ino-

citol and thiamine hydrochloride; further in­

vestigations are being carried out with this

medium. Conidial suspensions from infected

leaves have given lesions on young plants in 13

days after an initial 72 hours of incubation at

high humidities followed by transfer of the plants

to ambient conditions.

Cultures of Aspergillus flavus are being isol­

ated and maintained for use when the breeding

program for resistance to this important fungus

gets under way in the near future.

Two disease surveys were carried out in

Andhra Pradesh. At most sites leafspots and

rusts were the major pathogens. Other important

fungi included Aspergillus flavus, A. niger, Lep-

tosphaerulina arachidicola, Sclerotium rolfsii,

Rhizoctonia bataticola, and Fusarium spp. At

ICRISAT Center Rhizopus spp. was causing seed

rots in stored kernels as well as in the field, and a

Pythium sp. caused seedling damage in the

screenhouse. Cultures of many of these fungi are

being maintained for further study.

M i c r o b i o l o g y

Unlike many grain legumes, groundnuts con­

tinue to form numerous active nodules well into

the grain maturity stage. In rainy season 1976

cultivar Kadiri 71-1 had formed an average of 70

nodules/plant 28 days from sowing, with appro­

ximately half of them on the primary root. By 90

days there were 125 nodules per plant and this

increased to 190 per plant by 111 days. Although

the first-formed nodules senesced by 70 days,

nitrogenase activity as measured by acetylene

reduction assay continued until just before

harvest.

Nodulation and nitrogen fixation was ap­

parently little affected by seed treatment with

ethephon, or by application of herbicides and

fungicides.

A significant diurnal periodicity in nitrogenase

activity was noted in cv Kadiri 71-1 when

examined 81 days after sowing in postrainy

season 1976-1977. Activity increased rapidly

until 1400 hrs and then declined until dawn (Fig

49). After-dawn activity again increased

rapidly, suggesting a close link between nitro­

genase activity and photosynthesis. The rapid

decline after 1400 hrs may be related to the cloud

cover which developed by 1430 or to reduced

photosynthetic activity because of previous sto-

matal closure. This periodicity means that for

valid comparisons of nitrogenase activities under

different treatments, the assays would need to be

done on the same day and at the same time of

day.

In the postrainy season the effect of drought

stress on nodulation and growth of cv Kadir i

71-1 was determined by withholding every alter­

nate irrigation to some plots. Nitrogenase ac­

tivity per plant reached a maximum at 54 days

after sowing, just before the first stress was

applied. By 66 days, activity had declined to 54

µ moles C2H

4/plant per hour for control plants

irrigated 8 days previously and to 16 µ

moles/plant per hour for plants stressed for 20

125

days. Upon irrigation at 68 days, activity was

increased to 87 µ moles/plant per hour in both

treatments, and this pattern was repeated during

subsequent stress and irrigation periods. Stressed

plants produced less top and root growth and the

yield of 1 481 kg/ha was only half that of control

plants receiving regular irrigations.

Figure 49. Nitrogenase activity (µM C2H

4/g

dry weight of nodule per hour) in

groundnut cultivar Kadiri 71-1, 81

days after planting, ICRISAT Center,

postrainy season 1976.

There were significant differences in nod-

ulation between the more than 400 germplasm

lines screened during the rainy and the postrainy

seasons. A few cultivars had much better than

average nodulation and a few were poorly nod­

ulated. During field surveys, nodulation was

noted to vary a great deal between locations and

soil types.

L o o k i n g A h e a d i n G r o u n d n u t

I m p r o v e m e n t

In the germplasm program we intend to develop

Jinks with the IBPGR to enable us to prepare a

set of descriptors and an acceptable international

groundnut classification system based on tax-

onomic and agronomic characters. We expect to

have available shortly large quantities of germ­

plasm for cooperators; and we will be increasing

our own collection by additions from abroad.

Germplasm catalogs will be produced as soon as

possible.

As far as breeding goes, we hope to produce

through a large-scale crossing program high-

yielding material with stable resistance to the

major pathogens. Other breeding objectives are

to incorporate earliness and dormancy with high

yield potential.

The virologists will continue their work on

characterizing and purifying the viruses under

current investigation. They will develop inter­

national and regional arrangements for the sur­

vey of all groundnut-growing areas to precisely

identify important viruses. Rapid germplasm-

screening techniques will be perfected to assist

breeding programs.

Efficient fungal-screening techniques for large

scale field testing of breeding populations with

resistance to leafspots and rust will receive

priority in the immediate future. For the major

foliage fungi, we hope to set up testing centers in

a number of countries to monitor stability of

resistance and to detect the development of

physiological races. We hope to develop rapid

methods of laboratory screening for the de­

tection of lines with resistance to invasion by

toxin-producing fungal strains, such as yellow

mold (Aspergillus flavus).

Planning for the entomology program is now

complete; it will get under way in late 1977.

Objectives of the program include survey of the

harmful and beneficial arthropods of ground­

nuts, study of the vectors of diseases, and

assessment of the role of insects as pollinators in

groundnuts.

In microbiology, differences between lines in

nodulation and the interaction with Rhizobium

strain will be characterized to see if nitrogen

fixation increases with nodule number. Selected

lines with enhanced nitrogen-fixing activity will

be incorporated in a breeding program for

increased yield of total plant dry matter and

126

kernels. We also plan to examine the relationshipbetween carbohydrate storage and movementwithin the plant and nitrogen fixation, again inorder to select plants for use in a breedingprogram for increased nitrogen fixation andhopefully yield. The effect of drought stress onnitrogen fixation will be examined in more detailto characterize differences between lines at vari-

ous growth stages. We plan to survey the majorgroundnut-growing areas in order to determinepatterns of nodulation in groundnuts growing inthe field.

By 1978, we hope to initiate cytogenetics andphysiology programs. In the physiology project,the major study will be on locating sources ofdrought resistance in groundnuts.

127

The goals of ICRISAT from the beginning have embraced development of improved systems of farm-ing. In genera], the goals of the Farming Systems section parallel those long-range goals ofICRISAT - t o increase food production in the SAT, and to make it more reliable from season toseason and year to year.

Specifically, the goals of the Farming Systems research programs can be presented in threestatements-

- t o aid in generating economically viable labor-intensive production technology winch makes a better use of the productive potential of resources while at the same time conserving and improvingresources.

- t o assist in development of technology for improving land and water management and resourceconservation systems which can be implemented and maintained during the extended dry seasons, thusproviding additional employment to people and better utilization of available manpower.

- t o assist in raising the economic status and quality of life for the people of the SAT by aiding inthe development of systems of farming which will increase and stabilize agricultural output.

At any location, these objectives must be accomplished by providing optimum conditions for rainyand postrainy season cropping through proper management of the soil and of the total precipitationthat falls on the land and by better utilization of the improved environment through more-productivecropping systems. In some areas tins will also require collection and storage of runoff and the efficientutilization of stored water and groundwater.

129

F A R M I N G S Y S T E M S

F a r m i n g S y s t e m s

Rainfed agriculture has failed to provide even theminimum food requirements for rapidly increas-ing populations in many developing countries ofthe semi-arid tropics (SAT). Although thereasons are many, the primary constraint to agri-cultural development in the seasonally dry trop-ics is the lack of suitable technology for soil andwater management and crop production underundependable rainfall conditions. The severity isamplified by generally high evaporative demandsand, in many areas, by shallow soils with limitedwater-holding capacity.

The situation is intensified because farmershave increased the cropping areas and livestocknumbers in an attempt to provide the foodrequired for expanding populations. Vast areashave been over-cropped and over-grazed, result-ing in serious soil degradation. The decreasingproduction potential has intensified the quest foradditional land. To break this vicious circle,more stable forms of land use-forms whichpreserve, maintain, and better utilize the pro-ductive capacity of available resources-are ur-gently needed.

The Farming Systems research program atICRISAT Center is striving to do research whichwil l make a contribution to an improved eco-nomic status and quality of life for small farmerswith limited means who comprise the majority ofthe farming population of the SAT. The largerfarmers, who occupy most of the land andemploy much of the landless labor are alsoexpected to benefit. By investigating all facets ofresource inventory, development, management,and utilization and all factors involving cropproduction in a systems approach, scientists atICRISAT expect to develop basic principles,approaches, and methodologies which can bereadily adapted as alternative, economicallyviable farming systems in the SAT.

To meet its objectives, the Farming Systemsresearch program is involved in the followingactivities:

1) To assemble and interpret existing base-

line data in several areas of science relevantto agriculture in the SAT.

2) To assemble and communicate to cooper-ators basic and applied research resultsrelated to farming in the SAT.

3) To conduct basic and supportive researchin agroclimatology, hydrology, soil physics,soil fertility and chemistry, farm imple-ments, land and water management, agron-omy, and cropping systems.

4) To perform simulation or systems-analyticstudies based on climate, soil, and socio-economic information, so as to predictpotentials of new crops, cropping systems,and soil-, water-, and crop-managementpractices.

5) To organize international cooperative t r i -als to rapidly gain information about theperformance of a practice, technique, orapproach over time at the same locationand/or across locations.

6) To provide support and expertise forICRISAT training programs which areconcerned with farming systems.

7) To perform research on resource manage-ment, crop production, and resource con-servation at ICRISAT Center and selectedbench-mark locations.

Results of the past year's work are reported inthe following major categories:

Research in subprogram areasWatershed-based resource utilization researchCooperative research with national programs

Research i nS u b p r o g r a m s

Much of the research in the various subprogramsis carried out on a wide spectrum of activities or

133

experiments within the particular disciplines in-volved. These range from basic studies to appliedstudies which may find immediate applicationfor increased production. However, each sub-program has a problem-oriented focus towardsthe stated ICRISAT objective: " T o developfarming systems which wil l help to increase andstabilize agricultural production through betteruse of natural and human resources in theseasonally dry semi-arid tropics."

"Watershed-based resource utilization re-search" is conducted on watershed units onAlfisols and Vertisols. It is the operational testingground for principles and leads developed in thevarious subprograms where rainfall-use ef-ficiency and economics of alternative farmingsystems are investigated. Thus scientists in allsubprograms in the Farming Systems researchprogram, as well as economists, plant breeders,entomologists, physiologists, pathologists, andmicrobiologists, are involved in facets of theoperational-scale systems research in the water-shed units.

Agroc l imato logy

Crop production and crop suitability, for a particular cultivar at a particular location, aredetermined primarily by interaction of moisture,temperature, and radiation. In planning foragricultural development and transfer of agricul-tural technology, climatic data is essential. TheAgroclimatology subprogram is therefore pri-marily concerned with the collection, assembly,and interpretation of climatic data for ICRISATCenter and other locations of concern in theSAT. The quantification of the moisture en-vironment for crop growth is also being at-tempted. A sample analysis for the Hyderabadarea of India was reported in ICRISAT's 1975—1976 Annual Report.

The distinctive characteristics of the semi-aridtropics have major influence on the distributionof natural endowments-soils, rainfall, and cli-mate. These areas are well supplied with radiantcnergy, but because of variations in the weathersystem and orographic influences, a variety of

rainfall patterns are produced. Because of thehigh evaporative demand during most of thegrowing season, variations in the timing andamounts of precipitation are generally the keyfactors influencing the agricultural-productionpotentiality of a given SAT region.

Weather at I C R I S A T Center

This year both the onset and withdrawal of therainy season were earlier than normal. Thisfacilitated the timely planting of rainy seasoncrops, but adversely affected the seeding ofpostrainy season crops, which was delayed. Thetotal amount of rainfall received during the yearwas near normal at 752 mm (normal is 800 mm).Monthly totals of precipitation recorded fromJune 1976 to May 1977 are compared with long-term averages in Table 46. July and Augustreceived respectively 30 and 90 percent morerainfall than average, while September receivedonly about 40 percent of the average. These dataonce again show that rainfall in the SAT is quite

Table 46. Average monthly rainfall and rainfallreceived during 1976-1977 atICRISAT Center.

Month Average rainfall Rainfall

(mm) (mm)

JunJulAugSep

115.5171.5156.0181.0

86.0219.3298.7

74.0

OctNovDecJan

67.023.56.05.5

0.629.7

FebMarApr

11.012.524.0 7.5

May

Total

26.5

800.0

36.0

751.8

134

erratic and undependable. It also confirms thefindings reported on rainfall dependability forHyderabad in the 1975-1976 Annual Report. Itwas shown that, though September is the rainiestmonth, the dependability of rainfall is lower inSeptember than in July.

During the postrainy season (Oct-Feb) of1976-1977, very little rainfall was received (30.3mm; normal 113 mm). The growth and develop-ment of postrainy season crops were adverselyaffected due to the resultant atmospheric drou-ght and relatively higher temperatures. Thedependability of rainfall during this period isquite low; the amounts received from year to yearvary widely.

Daily and weekly values of precipitation re-ceived during the 1976 rainy season (Jun to Oct)are shown in Figures 50 and 51. The maximumamount of rainfall received during one single daywas 112 mm in August; there was a 93-mm rain inJuly. There were ten occasions when rainfall wasin excess of 20 mm/day and 19 occasions when itexceeded 10 mm/day. Set on standard weeks (Fig51), the highest weekly total was 132 mm forweek No. 29 (16-22 Jul). This amount of rainfallis about three times the average value based on

Figure 51. Weekly rainfall at ICRISA T Center, 1976.

70 years' data and it is six times the dependablerainfall expected at 75 percent probability.

The monthly amounts of rainfall recorded forthe past 5 years are shown in Figure 52 for therainy season period extending from June toOctober. The dispersion of curves clearly showsthat the early period of the rainy season, i.e. Juneand July, is somewhat more stable compared tothe months of August, September, and October.

The distribution and the amounts associatedwith high-intensity rainfall recorded during 1976are shown in Figure 53. During rainy season1976, there were 58 rainy days recording more

135

Figure 50. Daily rainfall at ICRISAT Center, 1976.

93

112120

6 0

50

40

30

20

10

0J u n J u l A u g S e p O c t

140

120

100

80

60

40

20

022

Jun26

Jul30

Aug34

Sep38 42

OctStandard

weeks

Figure 52. Monthly rainfall at ICRISAT Center, 1972-1976.

Figure 53. Rainfall intensities and total rainfall of selected storms at ICRISA T Center in 1976.

than 1 mm rainfall. There were two Auguststorms with an intensity exceeding 50 mm/hr.The first was a 60-mm rainfall in one hour nearmidnight on 18 Aug, followed by a second stormof 32 mm in 30 minutes just after midnight. Thetwo high-intensity rains were separated by abouthalf an hour.

The large number of rain-gauge stations distri-buted over 1394-ha ICRISAT Center permitsthe close examination of the uniformity of rain-fal l . The observations showed that the Septem-

ber rainfall was fairly well distributed while theearly rain in June was highly localized andnonuniform in distribution.

In semi-arid areas, extreme variation in rain-fall can be encountered even over a small area. A comparison of the spatial distribution of rainfall(total seasonal Jul-Sep) at ICRISAT Center forthe years 1974, 1975, and 1976 revealed widevariations. In 1975 the total rainfall varied from1000 to 1 300 mm. The variation was much lessin 1974 and 1976. The data for 22 August 1976showed a variation from 1.4 to 65.6 mm in N-Sdirection, demonstrating that rain gauges shouldbe located close to each experiment if quanti-tative evaluation of the moisture regime isdesired.

The trends of weekly mean maximum-minimum temperatures for 1976-1977 areshown in Figure 54. A maximum of 41°C (106° F)was recorded on 21 May and the minimum of9°C (48°F) recorded on 30 January 1977. Thedaily maximum temperatures generally rangedbetween 25 to 30° C during both the rainy seasonand the postrainy season. The daily minimumtemperatures ranged between 20 and 25° C dur-ing June through early October, and between 10and 20°C from mid-October to February.

Soil thermometers were placed at depths of 1 to 150 cm in a grassed Vertisol and readingsrecorded twice daily (thermometers embedded at100- and 150-cm depths were read only at 0717hours). Observations for various depths areplotted in Figures 55a and b. In the surface 30cm, there was little difference between the morn-

Figure 54. Weekly average maximum and min-imum air temperatures at ICRISAT Center, 1976-1977.

136

12 16 20 24 28 32 36 40 44 48 52 4 8 12Standardweeks

40

35

30

25

20

150

Minimum

Maximum

Normal1976-1977

22Jun 18Jul 3Aug 18,19Aug 19Aug

R a i n f a l l i n tens i t y

> 3 0 mm/hr

T o t a l r a i n f a l l

> 25 mm

120

100

80

60

40

20

0

1974

1973

1975

Normal19721976

Jun Ju l Aug Sep Oct

450

400

350

300

250

200

150

100

50

0

ing values. However, in the afternoon, variationsat different depths are more pronounced. Sea-sonal variations in temperature at shallowdepths are clearly shown.

The average weekly hours of sunshine at theICRISAT agri-meteorological observatory in1976-1977 are plotted in Figure 56. During Julyand August, relatively fewer hours of brightsunshine (on an average < 6 hours) were re-corded, while > 9 hours were observed from themiddle of September to March except on somedays in November, when there was cloud cover.

Average weekly wind velocities, recorded atthe 3-meter height, are plotted in Figure 57. Theperiod between Apr i l and July was quite windy,with average daily wind speeds in excess of 10km/hr frequently observed. The prevailing windwas from the west during this period. A max-imum daily average wind speed of 33 km/hr wasrecorded on 7 June 1976. During September-

Figure 57. Weekly average wind velocities at ICRISAT Center, 1976-1977.

March, the prevailing wind was from the east inthe afternoon and from the south in the morning.

Weekly means of the relative humidity, re-corded at 0717 and 1417 hours, are plotted inFigure 58. The relative humidity values in themorning were above 70 percent for the periodbetween June to January. However, the valuesexceeded 70 percent at 1417 hours during oneweek in July and in August.

The class-'A' pan evaporation data read dailyat 0830 hours are shown in Figure 59. Rates of 4 to 6 mm per day were observed during the periodmid-July to late January. Evaporation ratesabove 10 mm/day were frequently observedduring Apr i l to June, with the maximum dailyevaporation of 14.1 mm recorded on 28 Apr i land again on 30 Apr i l .

137

a . 0 7 1 7 h o u r s

150-cm depth

15-cm depth

30-cmdepth

30

25

20

15

10

60

50

40

30

20

Figure 55. Weekly average soil temperatures, re-corded at 0717 and 1417 hours, at ICRISAT Center, 1976-1977.

12 16 20 24 28 32 36 40 44 48 52 4 8 12Standardweeks

Soil surface

5-cm depth

b . 1417 h o u r s

5-cmdepth

1976-197712

10

8

6

4

2

0

12 16 20 24 28 32 36 40 44 48 52 4 8 12Standard

weeks

12 16 20 24 28 32 36 40 44 48 52 4 8 12Standard

weeks

Figure 56. Weekly average hours of bright sun-shine at ICRISAT Center, 1976-1977.

1976-197726

24

22

20

18

16

14

12

10

8

6

4

Normal

Figure 58. Weekly average relative humidity at ICRISAT Center, 1976-1977.

Dew was recorded in November, December,and January. Total for the period was 3.5 mm.

Characterization of Rainfall Patterns inthe SAT

Average monthly, seasonal, and annual rainfalldata based on long-term weather records, aregenerally used for assessing the precipitationcharacteristics and length of the humid season.Several workers (Thornthwaite, 1948; Tro l l ,1965; and Cocheme and Franquin, 1967)1 haveused comparisons between average rainfall andevapotranspiration to arrive at conclusions re-garding moisture availability of crops. Recently,

Hargreaves (1975) suggested that due to greatvariability of rainfall in the tropics, the climaticmoisture availability to crops should be based ona comparison of dependable rainfall (amount ofrainfall expected with 75% probability) andpotential evapotranspiration. Robertson (1976)used the Markov-chain method to analyze dryand wet spells for defining rainfall distribution inmonsoonal Malaysian tropical areas.

To compare the relative usefulness of theabove methodologies for characterizing rainfallin SAT conditions, we selected two adjacentareas of India-Hyderabad (17°27'N) andSholapur (17°40'N) situated nearly 500 kmapart. Both are in the Vertisol soils region. Othercharacteristics are listed in Table 47.

Study of rainfall, moisture index, and length ofthe growing season for the two locations fromgeneralized annual, seasonal, or monthly datashows that the areas are quite similar agroec-ologically. On similar soils (e.g. deep/mediumdeep Vertisols), therefore, one would expectsomewhat similar agricultural potentialities.However, results of. farming/cropping systemsresearch carried out at Sholapur Research Sta-tion of the All-India Coordinated ResearchProject for Dryland Agriculture and atICRISAT Center over the past 5 years or soreveal the following:

At Hyderabad it is possible to obtain yields inexcess of 5 metric tons per hectare by adoptingpigeonpea/maize intercrop or maize/chickpeasequential crop combinations under good agro-nomic management (rainfall-use efficiency is ofthe order of 6 to 10 kg/mm), whereas at Sholapurrainy season cropping is fairly undependable. A short-duration crop of pearl millet followed by a sorghum grown on conserved moisture is suc-cessful at Sholapur. But yields from year to yearare highly variable and rainfall-use efficiency atSholapur is quite low.

The aim, therefore, is to characterize therainfall climatology that is agronomically re-levant. We selected short-term climatic (weekly)data instead of month/season or annual data.The rainfall amounts have been characterized in

138

1References for this section are listed on page 197.

Figure 59. Weekly average evaporation at 1CRISAT Center, 1976-1977.

Table 47. Climatic characteristics of the Hyderabad and Sboiapur areas of India.

Characteristic Hyderabad Sholapur

Mean annual rainfall (1931-1960) 764 mm 742 mmProportion in Jun-Sep 76 % 75 % Annual mean PE 1757 mm 1802 mmMoisture indexa -56 .4 -58 .7Growing seasonb 12 Jun-8 Nov 8 Jun-22 Nov

(130 days) (148 days)H + Mc 130 days 140 daysH + M + M D 170 days 165 days

Climatic classification:Troll's Dry semi-arid Dry semi-aridHargreaves' Semi-arid Semi-arid

aThornthwaite's. See Rao et al (1971 a).bK.rishnan(1974).c According to Cocheme and Franquin's method. See Raman and Srinivasamurthy (1971).

relation to their relevance for crop water avail-ability. Once the crop is planted, the waterrequirement is fairly continuous, and hence theconditional probabilities of rainfall occurrenceare important.

Analysis by Markov-chain model for initialand conditional probabilities of R/PE>0.33meets most of the requirements as shown by theplots (Fig 60) for Hyderabad and Sholapur.2,3 Itis evident that rainfall distribution at Sholapur ishighly erratic, as only a couple of dispersedpoints of initial probability exceed the 70-percentthreshold. The conditional probability of wetperiod followed by wet period [P(W/W)] alsofollow a fairly similar pattern. In comparison,Hyderabad rainfall analysis shows that it has a dependable rainfall distributed between 18 Juneand the end of July and from about mid-Augustto mid-September. This clearly. brings out thatrainfall during the rainy season cropping periodat Sholapur is highly erratic and therefore unde-

pendable and is probably one of the majorenvironmental factors that has led to lowagricultural-production efficiency in that area.Hyderabad seems to have much more favorableclimate for crop production during the rainyseason.

Additional agronomically relevant infor-mation that such an analysis reveals, (Fig 60) isas follows:

(i) In Hyderabad, onset of the cropping seasonis fairly consistent around mid-June. So dryseeding in Vertisols is a feasible practice. InShoiapur, such a practice cannot be recom-mended because of uncertainty as to whenthe rains will start.

(ii) At Hyderabad, it is evident from therainfall-probability analysis that midseasonbreaks in the continuity of rainfall are likelyto occur on an average of 4 to 6 years of a 10-year period. Obviously one would not selecta crop cultivar that would be in an activephase of growth during this period. Either a sole short-duration crop (which can beexpected to complete most of its life cycleprior to the break in rainfall) or a long-

139

2R/PE (The rainfall/potential evaporation ratio) is calledthe Moisture Availability Index (MAI); dependable rainfallis considered to be the criterion.

3 India Meteorological Department has published PE valuesfor most of the districts in the subcontinent. See Rao, et al.(1971b).

Figwe 60. Initial and conditional rainfall probabilities of RIPE > 0.33 at two selected Indian SAT locations.

duration base crop with a short-durationintercrop would be best suited for theHyderabad environment.

(ii i) The rainfall analysis for Sholapur showsthat crops with indeterminate nature anddrought-hardy crops would be more suit-able, whereas at Hyderabad one could pro-bably select determinate or even drought-

sensitive crops, depending upon soilconditions.

(iv) The cost/benefit ratio for recycling of runoffwater would be much more favorable in theSholapur region when compared (for sim-ilar soil types) to Hyderabad.

(v) The wet/wet probabilities of rainfall ateither location show that in about 4 of 10

140

S H O L A P U R 1 7 °4 0 ' N 75° 54 'E E l e v a t i o n 479 m.

100

90

80

70

60

5°

40

30

20

10

0

C o n d i t i o n a l(W/W)

70%

I n i t i a lCond i t i ona l

( W / D )

StandardWeek 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov ' DecMonth

Date 8 22 5 19 5 19 2 16 30 14 28 11 25 9 23 6 20 3 17 1 15 29 12 26 10 24-31

100

90

80

70

60

50

40

30

20

10

0

C o n d i t i o n a l(W/W)

I n i t i a l

70%

Cond i t i ona l( W / D )

H Y D E R A B A D 1 7 °2 7 ' N 78° 28 'E E l e v a t i o n 545 m.

years, rainfall has a tendency to continueafter the normal date of recession. Cropssensitive to wet weather at maturity shouldnot be selected, particularly in the Vertisols.

The above comparison between Hyderabadand Sholapur is just one example of the employ-ment of climatic data for selecting crops orvarieties to suit the weather. Depending upon thenature of the investigator's interest, such a procedure could be used for any station (forwhich data is available) for planning severalcultural practices-including selection of meth-ods of land layout, seedbed preparation, selec-tion of sowing dates and methods of sowing,weeding, degree of mechanization, and type ofequipment to be used.

Hydro logy

In most farming systems of the SAT, only a relatively small portion (about 20 to 50 %) of theannual rainfall is actually used for crop pro-duction. To determine the potentials for im-provement, the Hydrology subprogram is con-cerned with the fate of all water, not only thatwhich is immediately available in the crop rootzone for evapotranspiration, but also that whichruns off or drains to deeper layers beyond theroot profile. Although considerable informationis available on the hydrology of large catchmentswith perennial vegetation, our knowledge of theactual water balances of small agricultural wa-tersheds is fragmentary. Most hydrologic studiesat ICRISAT are conducted as part of watershed-based resource utilization research or of cooper-ative research. The objectives of the Hydrologysubprogram are:

(i) To contribute to the quantification of runoffprobabilities, groundwater hydrology, anderosion behavior under alternative manage-ment treatments in agricultural watershedsfor various agroclimatic zones.

(ii) To assist in the development of hydrologicmodels and simulation programs for theinterpretation and extrapolation of hydro-

logic research findings to major agrocli-matic zones.

(iii) To develop methodology and equipment forhydrologic research.

Sediment Distribution and RunoffSampling

Measurements of sediment density variationduring runoff events show that considerableerror in erosion estimates might result whenthese are based on few samples (Fig 61). Sedi-ment concentrations at relatively brief peaktimes may be twice as large as those observedduring most of the remaining part of a hy-drograph. The vertical sediment distribution atthe converging and diverging sections of Parshallflumes was determined for Vertisols and Alfisols.The distribution of suspended particles at dif-ferent flow depths was relatively stable on Ver-tisols. However, a greater variation was observedfor Alfisols; this is presumably caused by a largerfraction of coarse sediment in eroded materialfrom such soils. Thus, accurate erosion estimatesrequire sampling at frequent intervals and sev-eral flow depths during a runoff event. Twodifferent types of sediment samplers, based onthese requirements, are presently being tested onresearch watersheds (Fig 62).

Observations on the Runoff Process inSmall Watersheds

Preliminary studies on modelling and simulationof the runoff process characterizing small agri-cultural watersheds have shown the overridingimportance of rainfall intensity and canopydevelopment. The effect of the weighted meanintensity (WMI ) 4 of storms on the rainfall-runoffconversion system during two distinct cropcanopy periods in the rainy season on BW1 (150-

4 To determine W M I , a s torm of varying rainfal l intensity issubdivided in to sections of similar intensity I 1 , I 2 , . I n ;if the precipitation in these sections is R1, R2, Rn,then

WMI - [ ( I1 /R I) + (I1/R2)+....(In/Rn) /(R1+R2+....

141

0.025

0.020

0.015

0.010

0.005

0

0 30 60 90 120 150 180 210 240 270

Sediment concentration

Runoff rate

Minutes

Figure 61. Sediment concentration and runoff rate during two storms on watershed BW1.

0 30 60 90 120 150 180 210 240 270

Minutes

0.05

0.04

0.03

0.02

0.01

0

Figure 62. Two types of sediment samplers being tested for research on the watersheds.

142

cm bed at 0.6 % slope) and BW4 C (rainy seasonfallow, field bunded) is plotted in Figure 63. Inthe preemergence and canopy-developmentstage (from emergence to about 20 July), crop-ped watersheds show a greater sensitivity tohigh-intensity storms than when a full canopy ispresent. Comparing the runoff behavior on BW1with that observed on BW4C, it is evident thatalthough bedded cultivation systems appear ef-fective in reducing runoff up to a W M I of about25 mm/hr, at greater intensities both watershedsreact similarly.

Figure 63. Effect of rainfall intensity and crop canopy on runoff.

Soil Physics

The purpose of the Soil Physics subprogram is toprovide a quantitative description of the physicalstate and dynamics of water during its passagethrough the soil-plant-atmosphere continuum.The research is directed at the soil-plant-atmosphere system which, because of its highlydynamic nature, must be treated as a functionalunit rather than as separate components.

Physical Characterization of AlfisolProfiles

Alfisols have a large number of stones distributedthroughout the profile; hence, sampling for bulkdensity using a core sampler 2 inches in diameter

is not feasible. The bulk density and stoniness ofan Alfisol profile was studied by taking 15-cm-thick samples to a depth of 180 cm with a 30- x 30-cm metal sampler. Three sets of such sampleswere taken from a pit dug in the RW3 area. Theresults are summarized in Figure 64.

Infiltration in Alfisols and Vertisols

Four sets of square concentric iron frames(1.5 x 1.5 meter inner, 2.4 x 2.4 outer) were usedto run infiltration experiments on the Alfisols ofRW1 and RW3 and the deep Vertisol of BW5.The 45-cm buffer zone between the framesprevented horizontal flow of water from theinner ring. A water head of 3 cm was maintained

143

Figure 64. Bulk density, particle density, and percent stones in Alfisols.

BW 4C 15 Aug - 30 SepBW 4C 1 Jun - 2 0 JulBW 1 15 Aug - 30 SepBW 1 1 Jun - 20 Ju l

0

30

60

90

120

150

180

0.6 1.0 1.4 1.8 0.6 1.0 1.4 1.8Bulk density

(g /cm 3 )Stone-free bulk density

(g /cm 3 )

0

30

60

90

120

150

1802.1 2.3 2.5 2.7 0 20 60 100

Stones > 2 mm (%)Part ic le density( g / c m )

Stone-freeWith stones1.8

1.6

1.4

1.2

1.0

Stones > 2 mm (%) Stones > 2 mm (%) 0 20 40 60 0 20 40 60

Weighted mean in tens i t y ( m m / h r )0 10 20 30 40

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

Figure 65. Initial and final moisture profiles of differentially charged soil profiles.

postrainy season crops of pigeonpea, chickpea,and safflower. The end-of-season profiles (Fig66) reveal that the surface 30 cm was depletedbelow the 15-bar percentage. In the 30- to 180-cmsection, slopes of the residual moisture profilewere roughly the same for the different crops.About 40 percent of the "available" water (fieldcapacity minus 15-bar percentage) in the 30- to180-cm section was not used by the crops. It issuggested that this end-of-season moisture pro-file may be a better indicator of the "cropextractable" water on deep Vertisols than thetraditionally used "available" water.

Assuming that the deep Vertisol profile wasfully charged at the beginning of the season, 115mm was used by the postrainy season crops, 40mm was lost by evaporation, and 50 mm was left

Figure 66. Moisture profiles at crop harvest in deep Vertisols.

as residual moisture in the 30- to 150-cm section.The 0- to 30-cm layer lost about 50 mm, much ofit by evaporation. An unmeasured but perhapsagronomically significant amount of moisture isbelieved to have moved upwards into the rootingzone from depths below 180 cm. This assumptionis supported by tensiometer data that showedupward hydraulic gradients in excess of five timesgravity in the 150- to 180-cm section duringextended periods of time beginning inNovember.

ConclusionsFrom the experiments conducted and data ob-tained during the past year, it was concluded thatthe physical properties and processes whichgovern the retention and movement of water inthe soil-plant-atmosphere continuum can bemeasured with reasonable precision. Thesemeasurements can be used for field studies of thewater balance, time and depth pattern of profile-water depletion, and crop water-use efficiencieson the deep Vertisols; however, further work wil lbe required to develop means of coping with the

145

FallowPigeonpeaChickpeaSafflower

0.20 0.24 0.28 0.32 0.36 0.40 0.44Moisture content (volumetric basis)

0

30

60

90

120

150

180

210

30

45

60

75

90

105

120

135

150

165

180

Full capacityInitial 9 NovFinal 23 Feb

32 .34 .36 .38 .40 .42Volumetric water content

A : Initially recharged soil profileB : Partially recharged soil profile

Table 48. Effect of row spacing and nitrogen applicati on on sorghum upon grain yields of sole andintercropped sorghum and pigeonpea on a Vertisol* at ICRISAT Center, 1976-1977.

Rowspacing

Zero N on sorghum 60 N on sorghum

CropRow

spacing Yield LER Yield LER

(cm) (kg/ha) (kg/ha)

Sole sorghumSole sorghumSole pigeonpeaSole pigeonpea

45904590

1530176015401400

-3520332013701370

-

Intercropping (alternaterows 45 cm)

Sorghum 1660 1.01 2980 .87Pigeonpea 1050

Total Land Equivalent Ratio (LER)L.S.D. (0.05) sorghum-1060; pigeonpea-310.

0.71

1.72

920 .67

1.54

aAvailable, N, P. and K prior to fertilization:0-to 15-cm depth, 78, 6.0, and 356 ppm, respectively

15-to 30-cm depth, 63, 1.1, and 237 ppm, respectively

This is a 4-year experiment, and some of thetreatments have had only their first annualapplication. Some preliminary information wasobtained, however. In the sorghum series, 10 and20 kg/ha of P as super phosphate and 30 ton/ha ofF Y M (farmyard manure) gave significant yieldincreases. In all other treatments there was a positive trend, but it was not significant. Thepearl millet yield was extremely low due toextreme drought in September and October andthere was no significant difference in pigeonpeayields (Table 49).

An experiment was initiated to evaluate theresponse of chickpea genotypes to phosphoruson a Vertisol. Ten promising genotypes fromchickpea breeders were tested at four levels of P (0, 10, 20, and 40 kg/ha).

Since this was the first crop grown afterclearing the land, the average yield levels werevery low (320 kg/ha). There was a small butsignificant response to phosphorus application.The genotype mean yields were significantly

different, but the interactions between P levelsand genotypes were not significant.

Seasonal Changes in Nutrient Status

A better understanding of seasonal changes innutrient status throughout the year under vari-ous soil and crop management systems is vital tothe development of management systems forcrop residues and organic wastes, as well as forarriving at soil- and water-management practiceswhich optimize nutrient availability to crops andminimize losses. There is some evidence of a buildup of nitrate in the soil profile during thenoncropped season prior to the rainy season(Krantz, et al. 1944 and Wetselaar, 1961a). Inorder to study this phenomenon on a year-around basis, a soil sampling program to meas-ure the seasonal changes in nitrate nitrogen and"available" N in the upper 90 cm of a Vertisoland an Alfisol was initiated.

To obtain a statistical measure of the vari-

147

Table 49. Sorghum and pearl millet/pigeonpea intercr op grain yields as affected by different treatmentson an Alfisol at ICRISAT Center, 1976-1977.

Intercropping

Treatment P applied Sorghum Pigeonpea Pearl millet

(kg/ha) (kg/ha) (kg/ha) (kg/ha)

1a 0P 680 620 3402 20 P as Phosphate Rock 930 660 3203 40P 850 790 4004 80P 770 710 4805 160P 1270 580 2206 80 P Phosphate Rock + Sulphur in 1100 690 350

Pellets

7 80 P " + 3 0 tons/ha 1600 740 490F Y M mixed

8 80P " +5P(SP) 1160 700 480banded

9 5 P as Super Phosphate (SP) banded 1280 760 43010 10P 1790 750 42011 20P 2040 610 58012 30 Tons/ha F Y M 1710 850 540

L.S.D. (0.05) 660 - -

aAvailable N, P, and K prior to fertilization:0- to 15-cm depth, 80, 2.8, and 126 ppm, respectively

15- to 30-cm depth, 94, 3.6, and 108 ppm, respectively

ability, soil samplings were taken in each of thefour replicates of the "Steps in Improved Tech-nology" experiment. Four treatment combi-nations involving local vs. improved fertilizationand management, plus a fallow plot, were sam-pled. Soil samples were taken at 0 to 15,15 to 30,30 to 60, and 60 to 90 cm at monthly intervalsfrom an area of 9 x 12 m in size in each treat-ment. Each sample consisted of a composite of 10subsamples taken at random. The samples wereimmediately treated with toluene, dried at 60° C,and thoroughly mixed. A 5-g sample was thenextracted with CuSO4 and the nitrate nitrogenConcentration determined on a 5-ml aliquot ofextracting solution by the phenol disulphonicacid method.

This investigation was started in July 1976 andwill be continued for several years in an attemptto establish a trend in seasonal status of nitrateand available nitrogen in these soils. Earlysamplings indicate large variability in nitrateconcentration with depth and time. Thus, largenumbers of replications and repeated samplingswill be necessary to establish reliable trends. Thepresent data will be combined and summarizedwith next year's data and reported at that time.

Boron Toxicity on Ratooned Pigeonpea

Severe necrosis of the leaf margins in ratoonedpigeonpeas was observed in March 1977. Thesymptom was the most severe on the oldest leaves

148

and graded to no symptom on the youngestleaves. The affected areas, which occurred inspots in fields ST1 and BW8 C, also showeddepressed plant growth. Pigeonpea leaf samplestaken at random within the affected areasshowed high levels of boron in all cultivars. Theaverage was 393 ppm B vs. only 151 ppm B in theleaves of "normal" plants from adjacent areas.

The average water-soluble boron content ofthe 15- to 90-cm soil depth showed that the boronlevel in the affected areas (1.50 ppm) was twicethat in normal soil (0.72 ppm). There were alsomore sodium and other soluble salts in the"affected" soil compared to the normal soil, butthese were not reflected in the analysis of plantsamples.

Toxicity symptoms did not occur during therainy season or early postrainy season. Theirappearance on the nonirrigated ratoon cropduring the February-March period is believed tobe due to increase in boron concentration in thesoil solution due to the reduction in soil moistureand also due to greater moisture extraction fromthe lower depths (where boron levels are higher)as a result of the drying of the surface soils.

ICRISAT's pulse physiologists induced thesame symptom in potted pigeonpeas by addingincrements of boron to the soil. This is the firstknown case of boron toxicity in pigeonpea.

F a r m P o w e r a n d E q u i p m e n t

During the past year the staff of this subprogramhas been expanded with the addition of a prin-cipal scientist and two engineers. The FarmPower and Equipment subprogram is attemptingto define and develop power-machinery systemsfor different levels of improved management.These systems have to be technically feasible andeconomically viable. Specific objectives andorientation of this subprogram are still evolvingand considerable time has been spent in projectdevelopment. The following five projects havebeen conceived and wil l be undertaken as fa-cilities, equipment, and staff are developed:

1. Determination of available power and

machinery resources and the requirementsfor improved farming systems in the SAT.

2. Development of improved tillage andplanting techniques and equipment andmachinery systems.

3. Development of improved harvesting tech-niques and equipment and machinerysystems.

4. Effective utilization of available power andgeneration of new sources of power andenergy.

5. Determination of design criteria, stan-dards, and cost estimates for the manufac-ture of selected farm equipment and smallmechanical power units and design of test-ing equipment.

Equ ipment Development

A major emphasis of the subprogram is tomodify, adapt, and evaluate existing implementswhich can play an important role in improvedfarming systems. In order to provide the properfocus to the subprogram and to avoid dupli-cation of work being conducted elsewhere, a survey is being organized to ascertain availableequipment as well as current and past researchand development on machinery and implementssuitable for small-scale farmers in the SAT. Aneffort wil l also be made to determine future needsof farm power and equipment for this group offarmers. The survey is being organized jointly byICRISAT and the International LivestockCentre for Africa ( ILCA). Major emphasis wi l lbe placed on Africa and Asia.

The machinery-development phase was grea-tly aided by the assistance of Mr. Jean Nolle, a designer of considerable experience in animal-drawn machinery. Most of his efforts were spenton modifying and improving the "Tropiculteur"for use in the farming systems being developed atICRISAT for Vertisols and Alfisols. This wheel-ed tool carrier was described in ICRISAT'sAnnual Report for 1975-1976.

149

150

Figure 67. Broadbeds being re-formed for the third year on an Alfisol at ICRISAT Center.

Since the 150-cm bed-and-furrow system hasbeen adopted in the watershed-based resourceutilization research, it was necessary to adaptand modify equipment for making these. Wherethe soil had been previously tilled, a ridger bodywas set at each end (150 cm apart) of the toolbar.An oblique deflector was attached behind eachridger inclined towards the center to form theridge. This provides a slight crown, which isdesirable to provide proper drainage of the bed.Figure 67 shows broad beds being re-formed forthe third year. For interrow cultivation, a steer-able toolbar is used (Fig 68). The animals andimplement wheels in all operations move in thefurrows.

It is relatively common in the SAT to use two-wheel carts with various animals. In some cases a four-wheel trailer replaces the two-wheel cart.

This has the advantage of being able to carry a much heavier load and to remove the verticalforce from the necks of the animals. Such a trailerwas designed to fit on the tropiculteur, using thetropiculteur as the two front wheels and framefor the trailer. A turntable is mounted on top toprovide steerage. A frame with rear wheelscompletes the vehicle. This trailer can be fittedwith various attachments for handling produceor to carry pipes used for supplemental irrigation(Fig 69).

Tillage Requirements

Tillage, including mechanical weed control andseedbed preparation, usually requires muchpower in the production of most annual crops.Thus it is desirable to minimize tillage as much as

possible. Excess tillage tends to accentuate theformation of crusts, particularly on Alfisols.Efforts will be concentrated on determining thetillage requirements and defining the physicalcharacteristics desirable in a seedbed. On Ver-tisols it has been observed that beds are suf-ficiently friable to be plowed during the early dryseason; at these times flat-planted areas becomevery hard and very difficult and costly to t i l l .Tillage studies to quantify these observations areplanned.

Planting Requirements

"Ebra" seeders, made in France, were used forplanting all crops (except pearl millet) atICRISAT Center. These machines use an in-clined rotating plate as a metering mechanism,

and prior to the 1977 planting season the plateswere modified for the seed size and seeding ratesof the cultivars being planted at ICRISAT. Thisinvolves having a plate correct in thickness, holesize, and spacing in relation to the drive wheel.These unit planters, each with its independentdrive wheel, provide a convenient method ofplanting two or more intercrops simultaneously.Compared with available planters, the French-manufactured machines provide a considerabledegree of precision planting for the small-scalefarmer. The major requirement of a good seederis to meter the required amount of seed atuniform intervals and to place the seed in propercontact with moisture for quick germination.This involves having the required meteringmechanism, furrow openers, covering devices,and press wheels. Experience in the past year

Figure 68. Interrow cultivation of intercrops on a Vertisol.

151

Figure 69. Four-wheel semitrailer, using the tropiculteur as the front wheels.

indicates a need for further development andstudies in these areas.

Energy Sources in the SAT

Power or energy is often a major constraint toagricultural production in the SAT. In someareas, particularly parts of Africa and northeas-tern Brazil, the size of the area cropped each yearis determined by the available human power forcertain critical operations. Mechanical, animal,and human power are all being used in variousparts of the SAT, usually in various com-binations. It is important that equipment shouldbe available to properly perform all operationsso that each power source can be utilized to itsfullest potential. In the watershed-based resource

utilization experiments, all operations from landdevelopment through interrow weeding are nowbeing conducted with animal-drawn machinery;traditional and improved machines are beingused. The improved machinery is being used forboth the bed and the flat cultivation. Thus it ispossible to compare machinery use in threefarming systems. Human power is used for someof the intrarow weeding, fertilizer topdressing,and harvesting. Threshing and supplementalirrigation are done with mechanical power.These operations have provided considerableexperience in the use of a variety of implementsand have shown deficiencies, indicating a needfor modifications. Since crops are being grown inthe rainy as well as the postrainy seasons,timeliness is important to permit crops to beplanted as early as possible in both seasons.

152

A study is being initiated to analyze all oper-ations in the watersheds in terms of theirmechanical, animal, and human power needs.This will help determine which operations havehigh energy requirements and merit particularattention in order to reduce power requirementsand operating costs, and to decide which oper-ations can best be done by mechanical, animal,or manual power.

A cropping calendar is being developed andtime-series studies initiated to determine wherepower, equipment, and labor bottlenecks occur.This analysis provides the information needed toquantitatively compare power, equipment, andlabor requirements for different crops and crop-ping combinations in a particular farmingsystem.

Village-level Studies of Power andEquipment Requirements

To understand farmers' requirements, a cooper-ative project is being undertaken with the Econ-omics program to evaluate farm power andmachinery systems for selected traditional farm-ing systems. The data collected by the Economicsprogram in its village-level studies are beinganalyzed to assess the extent of utilization ofavailable power and equipment sources underpresent farming systems. This will be followed bya study of the constraints to adoption of im-proved power-machinery systems.

Land and Water ManagementThe goal of the Land and Water Managementsubprogram is to assist national and regionalresearch organizations in the development andimplementation of improved resource manage-ment technology. Such technology must moreeffectively conserve and utilize the rainfall andthe soil to facilitate the introduction of hew crop-production systems capable of maintaining grea-ter productivity and assuring more dependableharvest. Most farms in the SAT are quite smalland the farmers have very limited financialmeans. The attitude of farmers towards new

techniques is conditioned by experience of un-certainty and risk. Methods for soil and watermanagement, conservation, and use should besuitable for implementation in this situation andbe associated with immediate and clearly visibleimpacts on the levels and stability of production.Major areas of land- and water-managementresearch are:

1. Land-management technology providingan improved moisture environment forcrops, runoff and erosion control, and(where necessary) increased infiltration ofrainfall without causing drainage problems.

2. Surface-drainage techniques facilitating a better growth environment for plants andimproving the workability of the soil with-out increases in erosion.

3. Design-criteria for waterway systems thatwill safely convey excess water from theland with minimum interference to agricul-tural operations.

4. Alternative technologies for use of avail-able surface- and ground-water on rainfed crops, providing increased benefits throughstabilization of crop production during therainy season and by lengthening the grow-ing season.

5. Improved systems for runoff collection andre-utilization along with utilization ofgroundwater to increase available waterresources on a watershed basis.

Solutions of resource-management problemsmust be tailored to specific site conditions due tothe diversity of environments in the SAT (clim-ate, soils, crops, topography, etc.). Focus of theLand and Water Management subprogram istherefore on the development of principles, ap-proaches, and methodologies, and on cooper-ative research projects and the training of re-search and extension personnel. Much of theland- and water-management research atICRISAT is executed as part of watershed-based

153

resource utilization research; small-scale studiesof present and improved soil- and water-management technologies wil l be discussed in thefollowing sections.

Evaluation of PresentResource-conservation Practices

Contour and field bunding are the most commontechniques to conserve soil and water on rainfed

lands in the Indian SAT. Although well-designedand maintained bunds undoubtedly conserve thesoil (see Watershed-based Resource UtilizationResearch), their usefulness for moisture con-servation which enhances crop production, par-ticularly on heavy soils, is less certain. At eachbund, one can distinguish a seepage area, a borrow pit, a zone where runoff is impoundedfrequently, a transition area intermittently underwater, and a zone which apparently is notaffected by the bund (Fig 70).

154

Figure 70. Effect of contour bunding on yields in a maize/pigeonpea intercrop on a Vertisol at ICRISAT Center, 1976.

D o w n s l o p e U p s l o p e D i s t a n c e f r o m c o n t o u r b u n d ( m )

- 1 0 0 10 2 0 30 4 0 50 6 0

B u n d 5

B u n d 6

B u n d 7

P I G E O N P E A2 0 0 0

1 0 0 0

0

D o w n s l o p e U p s l o p e

- 1 0 0 10 20 30 4 0 50 6 0

D i s t a n c e f r o m c o n t o u r b u n d ( m )

M A I Z E

B u n d 5

B u n d 6

B u n d 7 4 0 0 0

3 0 0 0

2 0 0 0

1 0 0 0

0

Seepageaffected

area BundBorrow

pi tSubmergedafter runoff Transit ion zone Land not submerged in the monsoon

The effects of three contour bunds on theyields of rainy season maize grown as an in-tercrop in pigeonpea are shown in Figure 70.These data were collected on Vertisol watershedunits BW6 C and D; the relative land areaoccupied by bunds in this area amounts to about3 percent, while the additional area affected bythe bunds varies from 2.7 to 3.6 percent. At bund5, crop yields in the submerged and transitionareas exceeded those in nonaffected areas byabout 10 percent; at bund 6, maize yields in theaffected area were approximately 10 percent lessthan those farther away from the bund. At bund7, yields near the bund were around 10 percentless than in nonaffected areas. Pigeonpea yieldsfrom the same areas are summarized in Figure70. In all cases, yields in the submerged zone wereless than in areas at greater distance from thebund. At bunds 5, 6, and 7, yields were reducedby 50, 30, and 60 percent, respectively.

The results obtained during the past seasonshow a rather mixed effect of bunds on maizeyields in the affected areas. During the rainyseason of 1976, only three runoff-producingstorms occurred and the rains terminated beforethe wet-season crops became mature in Septem-ber. The waterlogging effects, frequently as-sociated with contour bunds (Fig 71), were lessserious than, for example, in 1975. However, the

Figure 71. Waterlogged sorghum/pigeonpea in-tercrop above contour bunds.

relative area influenced by bunds is often so smallthat the overall yield effect, even if positive,becomes marginal. An increase of 10 percent inmaize yields of the intermittently submerged areaat bund 5 contributed less than 1 percent to totalyields because only 2.7 percent of the entire areais actually affected by the bund. Given theseobservations on contour bunds in a relativelylow rainfall year, it appears that no substantialmoisture-conservation effect on crop yields canbe expected from this practice under most con-ditions in the SAT.

Bed and Flat Planting

Runoff and soil erosion are processes whichoriginate on cultivated land. Although mechani-cal protection structures such as bunds may serveas a second line of defense, the treatment of theland and/or cultivation methods are most criticalin controlling the effects of high-intensity rain-fall, particularly under conditions of limited cropcanopy development, even when land slopes aremoderate. Land levelling is costly and oftenrequires large-scale equipment; also many of thesoils in the SAT do not have the depth requiredfor substantial earth movement. Therefore, a search was initiated to identify rather simple landtreatments which could be implemented by thefarmers.

Soil- and water-management and productivityaspects of several systems of land cultivation andplanting were evaluated on Alfisols and on deepand medium deep to shallow Vertisols. Re-latively large plots (0.3 to 0.4 ha) were used forthese comparisons because the relevant factors(runoff, erosion, drainage, operational require-ments, etc.), do not express themselves ade-quately on small plots. Also, the recommendeddistance between contour bunds is taken as a guideline for appropriate plot size. Runoff anderosion were measured on only two of the fourreplicates. Visual observations of factors such asweed conditions, drainage, workability of thesoil, plant-stand establishment, etc., wererecorded.

On deep Vertisols (BW5 A) three cultivationtreatments were compared - flat planting, nar-

155

Table 50. Effect of land management on runoff, erosion , and yield on deep Vertisob at ICRISAT Center,1976-1977.

Yields

LandTreatment

Maize Chickpea Valuea

LandTreatment Runoff Erosion Nonirrigated Irrigated Nonirrigated Irrigated

(mm) (kg/ha)

141 24077 110

110 170

(kg/ha) (Rs/ha)

Flat plantingNarrow ridgesBroad ridges

(mm) (kg/ha)

141 24077 110

110 170

274032403170

490 610450 650740 940

3170 33603540 38603940 4260

aTotal value of maize and chickpea.

156

row (75-cm) ridges, and a bed system withfurrows at 150 cm. Slopes averaged 0.6 percent.Maize was planted during the rainy season,followed by a chickpea crop in the postrainyseason. Because the rains terminated early, theseedbed was very dry when chickpea was plantedand one irrigation of about 75 mm was appliedon two replicates of each treatment. Data aresummarized in Table 50.

Although there was more runoff from thebroad ridges than from the narrow ridges, actualwater conservation for crops was not increasedbecause most runoff took place after the rootprofile was filled. Erosion quantities in all plant-ing systems were acceptable; the high-intensitylong-duration storms came after substantialplant cover had developed. Average maize yieldson both types of ridged plots were superior tothose from flat planting, although rather largevariations between plot yields were observed -presumably caused by minor differences in soiltype. Chickpea yields on broad ridges weresuperior to those obtained on narrow ridges orunder flat planting, either with or without i rr i -gation at planting time. These differences may beexplained by the greater friability of the soil inbroad ridges (which makes it easier to achieveadequate planting depth under dry conditions),operational problems in planting a second cropon narrow ridges, and difficulties in adequately

irrigating flat-planted areas. The total grossvalue of the yields of maize and chickpea onbroad ridges exceeded that obtained with flatplanting by 770 and 900 Rs/ha, respectively, forthe nonirrigated and irrigated systems. The in-crease on narrow ridges over flat planting wasabout half of that with broad ridges.

On medium deep and shallow Vertisols (BW8B), identical land treatments were imposed; a maize/pigeonpea intercrop was grown. The re-sults (Table 51) show that runoff and erosionwere very small. Yields on broad ridges weresuperior to those under flat and narrow-ridgedplanting during both the rainy and postrainyseason. Relatively low yields on narrow ridgesare explained by the unsuitability of this treat-ment for intercropping systems, resulting inproblems at planting time as well as in maintain-ing adequate weed control.

In conclusion, it appears-on the basis ofyields obtained - that broad (150-cm) beds arean appropriate land-management technique onVertisols. In addition it has been observed thatpreformed beds provide for faster and moreuniform planting as well as a wider range of rowspacings required by various crops (Fig 72). Ifearly showers occur at planting time, the top ofthe bed dries more quickly, thus facilitatingearlier planting. Germination of weeds has beenfound to be less in bed systems than in flat

Table 51. Effect of land management on runoff, erosio n, and yield on medium deep Vertisols at ICRISATCenter, 1976-1977.

Yields

Land Treatment Runoff Erosion Maize Pigeonpea Valuesa

Flat plantingNarrow ridgesBroad ridges

(mm)

141917

(kg/ha)

102020

(kg/ha)

270024703030

(kg/ha)

530480620

(Rs/ha)

3 78034454310

aTotal value of maize and pigeonpea.

systems. Beds and furrows make application ofsupplemental water possible without furtherland treatment. Soil on the beds remains friable,facilitating postrainy season planting and landpreparation after harvest of the second crop.Finally, with broad beds, crop land is not takenout of production for conservation purposes, asis the case with mechanical soil-protection struc-tures such as bunds. The implementation andmaintenance of beds from year to year is feasiblewith animal-drawn equipment. Al l fertilizers andplant residues are concentrated in the beds, thusincreasing the productive potential of the soil inthe plant zone.

Optimal Use of Supplemental Water

One drawback of "normal" irrigation in the SATis that frequently only small areas of rainfedagriculture are replaced by irrigated agricultureoriented towards high-value crops which requirelarge quantities of water. Total water resourcesin most semi-arid regions are too limited toenvisage large-scale conventional irrigation.Therefore, ICRISAT's research on the ut i l -ization of available groundwater or collectedsurface runoff must aim a,t the development ofviable technology providing for the supplemen-tal use of limited water resources to stabilize andbackstop rainfed agriculture.

In 1976-1977, the treatments designed forAlfisols and shallow to medium deep Vertisols

157

Figure 72. Cropping patterns and row spacings on 75- and 150-cm beds.

Narrow beds and furrows are adapted to75-cm rows only

Maize

Broad beds and furrows are adapted to manyrow spacings

Maize

75150

Sorghum, pearl mil let or soybeans

45 60 45

Groundnut or mungbean

•30 60 30

Cereal/Pigeonpea intercropping

Cereal PP Cereal Cereal PP Cereal

45150

45

75

during the rainy season-which were based ongrowth-stage sensitivity to drought stress and onrainfall probabilit ies-were not applied becausethe rainfall was adequate in quantity and distri-bution. This is the second time in succession thatthis situation has occurred. It appears necessaryto explore the possibility of conductingsupplemental-irrigation research on a multi-location basis during the rainy season. Throughemployment of cooperative programs, one cancover a wider range of rainfall conditions andobtain the necessary data more rapidly.

Cropp ing Systems

Intercropping has again received major attentionthis season. Another major effort involved more-detailed investigations in areas of particularpromise, notably plant population and genotypeevaluation. In the intercropping work, the aimsare to identify the various aspects of populationand spacing, and to assess their independenteffects as well as their interrelationships. In thework with genotypes, the aims are to identifydesirable genotype characteristics and to exam-ine the problems of evaluating relatively largenumbers of genotypes.

Intercropping performance is assessed by a "Land Equivalent Rat io" (LER) which is therelative area of sole crops required to produce theyield or yields obtained in intercropping. Thispermits direct comparison of crops with differentyield levels and it can be used either for anindividual crop or for the total of all crops. In thelatter case, a total greater than unity indicates a yield advantage for intercropping, e.g. a LER of1.2 indicates a yield advantage of 20 percent.

There has also been a further examination ofthe possibility of growing two successive cropson deep Vertisols where traditionally only a single postrainy season crop is grown. Again,both "sequential" sowing (seeding followingharvest of the first crop), and "relay" sowing(seeding shortly before harvest of the first crop)have been examined.

The influence of legume/cereal intercroppingon the development of pest populations has been

examined with particular emphasis on the impor-tant pest, Heliothis armigera (Hubner). Infor-mation on the pest/parasite relationship in in-tercrops has also been obtained.

Plant Population and SpatialArrangement in Intercropping

In maize/pigeonpea intercropping, response tototal plant population (i.e., both crops com-bined) at various row-arrangement patterns wasexamined. It was found that to achieve max-imum advantage from intercropping, the totalpopulation had to be higher than that for eithersole crop alone. For either sole crop, the op-timum population was six plants per squaremeter. At this level in intercropping, the yieldadvantage was 30 percent (Table 52). Althoughthis is quite a substantial advantage, it wasincreased to 45 percent at a total population ofnine plants/m2. The data indicate that this mightbe further increased at even higher populationsand this will be investigated further next season.

A particularly interesting aspect of this experi-ment was the effect of maize competition on theefficiency of the pigeonpea plant (Table 53). Theabsence of substantial reproductive growth ofthe pigeonpea until after maize harvest is con-sidered to be due to maize competition. Aftermaize harvest there was considerable compen-sation in reproductive growth, especially in num-ber of pods per plant. Thus compared to the solecrop, the harvest index (HI) of the pigeonpea wassubstantially increased by intercropping. In thetreatment with the highest proportion of maizethe HI value was increased from 17.2 to 30.4percent.

In the postrainy season, a new intercroppingdesign permitted examination of populationchanges in either crop. Four chickpea pop-ulations were established in main plots and thepopulation of a safflower intercrop was sys-tematically changed within these. A 10-percentchange between each safflower population gave a fourfold change in 15 steps. These populationeffects were distinguished from spatial arrange-ment effects by examining all population treat-ments at each of two row arrangements.

158

Safflower was the more competitive crop and ittended to reduce chickpea yield. But, except atvery low populations, safflower was little affectedby changes either in its own population or in thatof chickpea. Thus the best intercropping com-binations were those which maximized chickpea

yield. These were moderately low safflower pop-ulations (to minimize safflower competition)combined with high chickpea populations (tomaximize chickpea population response) (Fig73). These combinations gave substantial yieldadvantages in alternate-row intercropping but

159

Table 52. Effects of plant population on grain yields and Land Equivalent Ratios (LER) inmaize pigeonpea intercropping on Vertisol at ICRISAT C enter, 1976-1977.

Total plant population

3/m2 6/m2 9/m2

Yield LERa Yield LERa Yield LERa

(kg/ha) (kg/ha) (kg/ha)

Sole maize 2 769 - 3 398 (1.0) 3 331 -Sole pigeonpea 988 - 1035 (1.0) 954 -

Intercroppingb

Maize 1637 .48 2205 .65 2400 .71Pigeonpea 780

Total Land Equivalent Ratio (LER)

.75

1.23

672 .65

1.30

768 .74

1.45L.S.D. (0.05) between Total LER = 0.16

aLER calculated from optimum sole crop yields (i.e., 6 plants/m2).bMean of three row arrangements (See Table 53).

Table 53. Effect of different row arrangements on pigeo npea when intercropped with maize on Vertisols atICRISAT Center, 1976-1977.

Pigeonpea Pods Pigeonpeatotal dry per grain Harvestmatter plant yield Index

(kg/ha) (no/plant) (kg/ha) (%)

Sole pigeonpea 5 750 73.8 991 17.21 row pigeonpea/ 3 560 128.0 889 25.01 row maize1 row pigeonpea/ 2820 131.3 785 27.92 rows maize1 row pigeonpea/ 1800 162.4 545 30.43 rows maize

L.S.D. (0.05) 600 19.8 141 3.5

Figure 73. Population effects on saffower/ chickpea intercropping at two row arrangements on Vertisols at ICRISAT Center, 1976 (3-point mov-ing average).

little or no advantage when the planting patternwas one row safflower to two rows chickpea. Theadvantages in alternate rows may have beenpartly due to a beneficial effect of shading on thechickpea, an effect which has been observed inthe chickpea physiology research.

Genotype Evaluation

In cooperation with the pearl millet breeders, 40pearl millet genotypes were subjected to initialintercrop screening with pigeonpea, Setaria, andsorghum. A large number of plant charactersand yield determinants were measured on ail

genotypes and these were correlated with in-tercropping performance. Unfortunately, yieldsof pigeonpea and Setaria were poor and nocorrelations with these crops were significant.With sorghum, the best correlations were withdifference in height and difference in maturity.These effects were substantiated by a yield trialwhich examined all intercropping combinationsof four pearl millet genotypes x four sorghumgenotypes (Table 54). For two crops so similar,the advantages of intercropping were remark-ably high. Nine of the 16 combinations gaveadvantages ranging from 10 to 32 percent. Thetwo highest advantages were from the com-binations of latest sorghum with earliest milletand earliest sorghum with latest millet. The thirdhighest advantage was from the tallest sorghumwith shortest millet. But other advantages couldnot be explained in terms of height and maturitydifferences. It seems likely that there are manyother factors involved, so these crop combi-nations will be studied further next season.

Three-crop Intercropping

The effects of adding intercrops of Setaria (90days), groundnut (134 days), or both to pigeon-pea (184 days) were examined on both soil types(Table 55). The Setaria/pigeonpea combinationgave a large advantage on both soils; groundnutgave a large advantage on Alfisols, but less onVertisols. The three-crop combination gave lar-ger advantages than the two-crop combinationon Vertisols, suggesting that the different ma-turity periods of the three crops allowed somecomplementarity in terms of the use of growthresources over time. However, on Alfisols, thethree-crop combinations gave smaller advan-tages than did the two-crop combinations. Inview of the dry period during September, thesedata suggest that three-crop intercropping maybe beneficial only if there is sufficient moisture tosupport all three crops.

Relay and Sequential Cropping

In 1975-1976, there was good evidence that a rainy season as well as a postrainy season crop

160

Saf f lower (p lan ts / m3)0 5 10 15

One row sa f f lower : T w o rows ch ickpea

1.2

1.1

1.0

0.9

0.8

0.7

0

1.1

1.0

0.9

0.8

0.7

0

One row safflower : One row chickpea

13.326.740.053.3

Ch ickpea ( p l a n t s / m 2 )

Table 54. Land Equivalent Ratio and height and matur ity differences of sorghum and pearl milletgenotypes in intercropping on Alflsol at ICRISAT Ce nter, 1976-1977.

Land Equivalent Ratio Height Maturity

Sorghum 4- Pearl millet Sorghum Millet Total ( S > M cm)a (S > M days)a

Y 75 + PHB 14 0.80 0.52 1.32 + 59 + 21GE 196 + G A M 75 0.25 1.06 1.31 - 2 - 1 4Y 75 + G A M 75 0.69 0.55 1.24 + 104 + 9 CSH-6 + PHB 14 0.64 0.54 1.18 - 2 + 8 GE 196 + Ex-Bornu 0.14 1.03 1.17 - 8 7 - 9

Y 75 + Ex-Bornu 0.42 0.71 1.13 + 19 + 14Y 75 + G A M 73 0.57 0.55 1.12 + 102 + 18CSH-6 + Ex-Bornu 0.38 0.72 1.10 - 4 2 + 1 IS 9237 + Ex-Bornu 0.39 0.71 1.10 - 2 5 + 11CSH-6 + G A M 75 0.53 0.53 1.06 - 4 3 - 4

GE 196 + PHB 14 0.20 0.85 1.05 - 4 7 - 2IS 9237 + PHB 14 0.45 0.59 1.04 + 15 + 18GE 196 + G A M 73 0.22 0.82 1.04 - 4 - 5CSH-6 + G A M 73 0.48 0.51 0.99 + 41 + 5 IS 9237 + G A M 73 0.42 0.56 0.98 + 58 + 15IS 9237 + G A M 75 0.35 0.61 0.96 + 60 + 6

a S > M refers to the amount the sorghum exceeds pearl millet in height (cm) or in maturity (days).

Table 55. Yields and Land Equivalent Ratios of pigeonp ea intercropped with Setaria and groundnut atICRISAT Center, 1976-1977.

Pigeonpea Setaria Groundnut Total

Yield LER Yield LER Yield LER LER

AlfisolP S S S a

P G G GP S G SP S G G G S

VertisolP S S SP G G GP S G SP S G G G S

(kg/ha)

400400350270

7001070

810630

0.850.850.740.58

0.570.880.660.52

(kg/ha)

1860

1420970

1970

18401480

0.75

0.570.39

1.11

1.040.84

(kg/ha)

970240380

22070

250

0.690.170.27

0.410.130.46

1.601.541.481.24

1.681.291.831.82

aP = pigeonpea, S - Setaria, and G = groundnut.Distance between all intercropping rows was 37.5 cm.

161

could be grown on deep Vertisols, though tradi-tional practice in India is to grow a postrainyseason crop only. These investigations werecontinued by growing six postrainy season crops(i) after fallow and (ii) after a crop of maizegrown on 75-cm ridges. There were three dates ofsowing equivalent to 24, 12, and 0 days beforemaize grain harvest. At the two earlier dates, thetreatments following maize were (i) relay sowninto the full maize crop and (ii) relay sown afterremoval of alternate maize rows for greencobs;the last sowing date was essentially a sequentialsowing after maize. (Due to abnormal droughtconditions in September, a "come-up" irrigationwas applied after the first planting.) These treat-ments also provided three "shade" comparisons:(i) no shade (after fallow), (ii) half shade (afterharvest of greencobs), and (iii) full shade (underful l maize crop).

Germination at the second time of sowing was

Single postrainy Sequential planting Relay planting in Relay planting inseason crop greencob-grain grain treatment

treatment

Figure 74. Monetary values of sequential and relay planting after maize.

poor until rain fell just after the third sowing;therefore there was little difference in terms ofperformance between these dates. It can be seen(Table 56) that the growing of a rainy seasonmaize crop had little or no effect on the yield ofthe postrainy season crops. The half-shade treat-ment obtained by the greencob harvest appearedto have a small beneficial effect on the yield ofsafflower, sunflower, and cowpea. Pigeonpeaand sorghum benefitted from earlier sowing,whereas chickpea and safflower benefitted fromlate sowing.

Monetary values of the different systems aresummarized in Figure 74. For single croppingafter fallow, the highest yield of the three sowingdates is taken. For sequential cropping aftergrain maize, the mean of the two latest sowingdates is taken because in practice they germi-nated at the same time. The "greencob" systemgave the greatest returns because of the high

162

Maize grainMaize greencobsMaize grain

10.5

9.0

7.5

6.0

4.5

3.0

1.5

Chickpea

Saff lower

Cowpea

Sunflower

Sorghum

Pigeonpea

Maize grain

163

value of greencobs; in practice, the feasibility ofthis system depends on the marketability ofgreencobs. After this system, two of the bestsystems were chickpea planted after maize andpigeonpea relay planted 24 days before maizeharvest.

Cropping Systems Entomology

The influence of legume/cereal intercropping onthe development of pest populations has beenexamined with particular emphasis on the impor-tant pest, Heliothis armigera (Hubner). This yearinformation on the pest/parasite relationship inintercrops has also been obtained.

Heliothis armigera (Hubner). H. armigera wasrecorded feeding on SO cultivated plant speciesand 51 weed species. The most significant "carry-over" hosts in the hot summer season are theweeds Datura metel, Acanthospermum hispidum, and Gynandropsis gynandra. However, out-of-season crops of tomato, cowpea, and maizegrown under irrigation carried a large number oflarvae-up to 156000 larvae/ha on cowpea inMarch. In January 103 000 larvae/ha were foundon pigeonpea, while on chickpea in the samemonth up to 248000 larvae/ha were present.Generally, populations on all host plants werelow from Apr i l to June.

Peak oviposition in H. armigera is generallyassociated with the flowering of its host plants.Detailed observations were carried out on mothbehavior in cowpea and irrigation was noted tocause a considerable increase in the number ofadults present in the crop.

Nineteen parasitoids of H. armigera have beenrecorded. Parasitism levels varied, depending oncrop and season. Parasites were recorded in allmonths except June. Mermithids were importantearly in the season (Jul-Sep), Diptera in Octoberand January, and Hymenoptera in most monthsbut particularly in September-October andDecember-February. Overall parasite levels werefar higher on crops which had not been sprayed(28 %) than on sprayed crops (10%) (Table 57).This was also true in other areas of AndhraPradesh where parasitism rates of 1 to 3 percent

were recorded in DDT-sprayed intercroppedpigeonpea compared with 22 percent in onenonsprayed field. Egg parasites recovered wereTrichogramma confusum Viggiani, and Microche-tonus curvimaculatus Cameron. No virus par-ticles were recovered from a number of diseasedlarvae submitted to the Boyce Thompson In-stitute. Studies on chickpea grown in 25 farmers'fields revealed high parasitism levels on H.armigera and this had a considerable influence onpod damage, which was generally low (up to 8 % pods damaged). Three farmers applied D D T atearly podding stage but with no obvious benefit.

Insect relationships in intercropping. Tech-niques for assessing pest and pest-parasite ratiosand yield losses in intercrops were devised usingtwo cultivars of pigeonpea, HY-2 (semierect) andHY-3A (erect) in Alfisol and in deep Vertisol.The treatments were sole-crop pigeonpea (PP),pigeonpea intercropped in alternate rows withCSH-5 sorghum (PP/S), with HB 3 pearl millet(PP/PM), with H-1 Setaria (PP/St), and pigeon-pea mixed with sorghum within rows (PP + S).

Several pod borers were recorded damagingpigeonpea, but the pod borer H. armigera was byfar the most common species present and anobvious cause of yield loss. Damage by pod fly,Melanagromyza obtusa (Mall.), increased duringthe later stage, particularly on cv HY-3A.

Eggs laid and number of larvae of H. armigera were far lower on pigeonpea grown on the Alfisolthan on the deep Vertisol, and a similar trendappeared in moth numbers trapped (Fig 75).Larval parasitism by Diptera was much higherthan by Hymenoptera on pigeonpea, and levelsdeclined as both cultivars approached maturityin both soil types.

No differences in levels of M. obtusa damagewere observed between trials grown on the twosoil types, but time played an important role.Cultivar HY-3A was more susceptible and heav-ily attacked. An increasing trend of pod sheddingin cv HY-2 in Vertisol and cv HY-3A on Alfisolwas observed. Up to 62 percent of shed podswere damaged, 56 percent by lepidopteranborers, 2 percent by pod fly, and 4 percent by a hymenopteran pest, Taraostigmodes sp.

164

Table 57. Larval parasitism on Heliothis armigera (Hubner) in sprayed versus pesticide-free areas,ICRISAT Center (Sep-Mar 1976-1977).

Larval parasitism

Parasites Pesticide-freea Sprayed area

(%) (%)

Diptera

Eucarcelia (Carcelia) illota CurranGoniopthalmus halli Mes.

8.39.9

3.11.4

Others (three species)

Total

1.0

19.2

0.6

5.1

Hymenoptera

Diadegma sp.Eriborus argenteopilosus CameronCampoletis chlorideae Uchida

5.33.10.6

4.10.30.8

Others (five species)

Total

0.2

9.2

0.1

5.3

Overall parasitism (%)Larvae collected (total)

28.43828

10.47877

aArea BA-25 at ICRISAT Center has been set aside as a perpetual pesticide-free area.

The final yield-loss data were obtained usingthe actual weights of damaged and undamagedpods and seeds and calculating the potentialyields if all pods had been undamaged. Loss inseed yield caused by insect pests was lowest insole-drop blocks of cv HY-2 in Vertisol and of cvHY-3A in Alfisol (Table 58). Increased damageto pigeonpea in Vertisol resulted in reducedshelling percentages in both the cultivars.

The yield of both cultivars was higher onAlfisol in intercropped blocks but cv HY-2 in thesole-crop block yielded more on Vertisol (Table58).

Crop proportion (pigeonpea/sorghum) studies.On sorghum (CSH-5) and pigeonpea (ICP-1)sown in various crop proportions in late June inVertisol, the percentage larval parasitism on H.

armigera on sorghum declined with the increasein plant population, but the levels were onlysignificantly different (P< 0.05) during the peakparasite activity on pigeonpea (Table 59). Datafrom similar trials during 1975-1976 on both soiltypes revealed that fewer egg and larval numberswere present on blocks with fewer plants per ha.Pest numbers, pest-parasite ratios, and p lant -density relationships will affect the efficiency ofselection procedures for resistance to H. ar-migera in the pigeonpea screening program.

Pest monitoring on sorghum in intercrop/cropproportion trials. It was confirmed that shoot-fly, Atherigona soccata Rond., attack was moresevere on later sowings. Damage levels werehigher on sorghum grown on Alfisols with up to43 percent compared with a maximum of 20

165

Figure 75. Heliothis armigera (Hubner) eggs and larvae per 100 terminals on intercropped pigeonpea (cv HY-2 and HY-3A) and moths trapped at two locations, ICRISAT Center, 1976-1977.

percent in Vertisols. This was associated withslow initial growth of seedlings in Alfisols. Heav-ier attacks were also obtained on sorghum due toslow growth in some Vertisol areas sown afterunweeded fallow. Orius sp., a predatory bug, andearhead bugs, Calocoris angustatus Leth., werepresent in higher numbers on Alfisols than onVertisols (S3 compared with 25 per 40 earheadsand 16.5 compared with 3 per 40 earheads,respectively). Predatory spiders were found ingreater numbers on sorghum earheads on A l -fisols, while thrips were more numerous onVertisols (21 compared with 15 per 40 heads).

Light-trap studies and insect fauna at ICRISATCenter. A third light trap at a Vertisol water-shed was commissioned in Apr i l 1977. Three

years of regular light-trap monitoring of morethan 50 important pests of the SAT on legumesand cereals is revealing basic information onseasonal variations in pest species.

Four-week moving means during 1974-1977for H. armigera moths recorded from Trap 1 areshown in Figure 76. The first and the third peakswere further displaced during this year, but thesecond appeared in the same week as observedduring 1974-1975. A fourth peak at the end ofApr i l this year was from a summer cowpea cropat ICRISAT Center.

The most significant pests and beneficial faunato date on the range of crops were authenticated.A list of parasites, predators, and hyperparasites,along with a detailed list of pests and insects, hasbeen compiled.

166

The terminal is d ista l 25-cm portion of the pigeonpea branch.

Oct Nov Dec Jan Oct Nov Dec ' Jan

cv H y - 2

Deep V e r t i s o l

c v H y - 3 A

Deep Vert isol

Oct Nov Dec JanOct Nov Dec Jan

EggsLarvae

c v H y - 3 A

A l f i s o l20

Oct Nov Dec Jan Oct Nov Dec Jan

30

20

10

Deep V e r t i s o l4 week moving means

Weekly to ta l s

A l f i s o l

600

400

200

60

40

20

Eggs

La rvaecv H y - 2A l f i s o l

Table 58. Pod numbers produced per 25 plants, yield , and percentage of insect-caused yield loss onintercropped pigeonpea (cv HY-2 and HY-3A) on Alfiso l and Vertisol, ICRISAT Center,1976-1977.

HY-2a HY-3A

Treatment Alfisol Vertisol Alfisol Vertisol

PP/Sb Pods (no/25 plants)Yield (kg/ha)Yield loss (%)

930.2320.2

19.7

710.8169.635.1

582.7175.743.9

525.884.443.0

PP/Sc Pods (no/25 plants)Yield (kg/ha)Yield loss (%)

790.7336.523.2

426.6120.339.6

707.2279.2

41.3

455.069.951.1

PP/PM Pods (no/25 plants)Yield (kg/ha)Yield loss (%)

1 045.0302.2

18.0

765.6219.3

23.9

837.5375.038.9

424.494.654.6

PP/St Pods (no/25 plants)Yield (kg/ha)Yield loss (%)

1 175.5405.7

24.1

841.2229.527.3

888.0426.7

35.6

651.6200.049.0

PP Pods (no/25 plants)Yield (kg/ha)Yield loss (%)

1101.7317.220.7

1257.8489.5

22.6

1 277.0634.5

31.0

677.2193.855.0

L.S.D. (0.()5)Pods (no/25 plants)Yield (kg/ha)Yield loss (%)

NSNS

3.8

NSNS

9.0

227.1169.2NS

86.1NSNS

a Inclusive of two pickings.bPP = Pigeonpea, S = Sorghum, PM = Pearl Millet, St = Setaria.cSown mixed within rows.

Agronomy and W e ed Science

To capitalize on an improved soil- and water-resource base, it is essential to develop croppingsystems which wil l provide a continuum ofproductive crop growth from the onset of therainy season as far as possible into the postrainyseason. Because of the great potential for in-creased production by intercropping and otherdouble-cropping systems, additional emphasishas been placed upon basic studies (reported in a separate section under the Cropping Systemssubprogram) in this field. Agronomic researchreported here involved preliminary studies on

row direction and lodging, weed-managementresearch, and a continuation of the "Steps inImproved Technology" experiments.

In the SAT, many weed-control problemsremain unsolved due to limited weed research oncrops and cropping systems. At ICRISAT Cen-ter, the major approach to weed management isthrough habitat management, and research ef-forts are being oriented to obtain more basicinformation on the crop-weed balance. Weedresearch on individual crops was designedmainly to determine the competitiveness withweeds of distinctly different cultivars of themajor crops and to study the tolerance of thesecultivars to some selected herbicides.

167

Effect of Row Direction and ArtificialLodging Upon Pigeonpea Yields

A preliminary trial was established to study theeffect of row direction (north-south vs. east-west)

of two cereals and pigeonpea alone and inintercropping systems.

Pigeonpea yields in the north-south directionwere significantly higher than those in the east-west direction on both Alfisols and Vertisols.

Table 59. Parasitism levels on Hetiothis armigera (Hubner) from sorghum (CSH-5) and pigeonpea(ICP-1) grown in various crop proportions in deep V ertisol, ICRISAT Center, 1976-1977.

P a r a s i t i s m

Plant population(Pigeonpea/Sorghum)

Sorghum PigeonpeaPlant population

(Pigeonpea/Sorghum) Mid-Sep Late Nov Mid-Dec Early Jana Early Feb

(1000/ha)

32.0:85.515.0:128.511.5:145.08.0:150.0L.S.D. (0.05)

32.037.039.053.0NS

(1000/ha)

32.0:85.515.0:128.511.5:145.08.0:150.0L.S.D. (0.05)

40.537.931.526.9NS

9.59.58.0

15.0NS

32.037.039.053.0NS

34.550.551.059.07.3

39.434.417.930.2NS

aPeak activity of parasites observed.

a B e g i n n i n g 1 J u l .

Figure 76. Catch of Heliothis armigera (Hubner) from August 1974 through May 1977 in Trap No. 1 (at Crap Improvement Building), ICRISAT Center. Plotted as 4-week moving means.

168

J u l A u g S e p O c t N o v D e c J a n F e b M a r A p r M a y J u n5244403632282420161284

48

40

32

24

16

8

0

1 9 7 4 - 1 9 7 5

1 9 7 5 - 1 9 7 6

1 9 7 6 - 1 9 7 7

Weeka

(%)

Plants artificially lodged appeared to increase theflush of new branches at the nodes above the bentarea; however, there was no significant effectupon pigeonpea grain yield in either soil.

Setaria planted in the north-south row direc-tion in the pigeonpea intercropping system gavesignificantly higher yields than plantings in theeast-west row direction. In the maize/pigeonpeaintercropping system, the maize yield trend wassimilar, but the differences were not significant.These results are in agreement with results onwheat (Santhisegaram 1962) and on maize (Dun-gan et al. 1955).

Evaluation of Genotypes of Low-growingLegumes in a Pigeonpea/Cereal IntercropSystem

Several genotypes of low-growing legumes weretested as a third crop in a pigeonpea/cerealintercrop system. These low-growing legumeswere planted on broad beds in between thepigeonpea and cereal rows, thus giving five rowsof crops for 150 cm of space.

In the Alfisols the introduction of mung beans,cowpeas, black gram, or groundnut had nosignificant effect upon the grain yield of therapid-growing upright Setaria. There was, how-ever, a small but significant depression in pigeon-pea grain yield when low-growing legumes wereintroduced.

In spite of the slight pigeonpea yield reduction,the total monetary value was greater in the three-crop intercrop systems than in the two-crop(pigeonpea/Setaria) system. There were, how-ever, large differences in the genotypes used,especially in cowpeas and groundnuts. The bestcowpea (1152) and groundnut (Robut 33-1)cultivars in the three-crop combinations pro-duced total monetary values of Rs 5230/ha,about 50 percent above that produced in two-crop systems.

The large differences in yield between gen-otypes of low-growing legumes and the differen-tial effect upon the yield of the other twointercrops needs further research to evaluategenotypes and identify crop systems which aremutually compatible.

Steps Towards Improved Technology

In the development and implementation of im-proved technology, there are many facets or"steps" involved. If one attempted to researcheach of the individual phases in combination, thetotal number of combinations would becomeunmanageably large; also, the effects of manyindividual facets have been investigated pre-viously. For convenience the many steps weregrouped into the following four phases: Variety,Fertilization, Soil and Crop Management, andSupplemental Water (where needed).

Sorghum on Alfisols. In 1975, a sorghumexperiment involving comparisons of traditionalvs. improved technology on an Alfisol wasinitiated. This experiment was repeated in 1976with a slight modification in treatments. Duringthe main crop (rainy) season the rainfall distri-bution was adequate and fairly uniform; sup-plemental water was not required. Thus, Treat-ments 9 and 10 were considered as additionalreplications of Treatments 4 and 8, respectively,and only the first three phases can be consideredin interpretation of the main crop yields; how-ever, Treatments 9 and 10 were consideredseparately in the ratoon crop (Table 60). Therewas a significant response to improved fertili-zation. Improved variety and improved manage-ment as single factors showed an upward trend,but this was not significant. Sorghum yield in the1975 experiment showed the same trends. Incomparing yields of improved vs. traditionallevels for these three factors applied individually,the sum of the increase was 1489 kg/ha; however,when all three factors were combined (Treatment8), the increase was 3092 kg/ha-(Table 60).

Figure 77 shows the yield increases of im-proved over traditional technology from theapplication of the three steps (Variety, Fertili-zation, and Soil and Crop Management) singlyand combined. Yield increases f rom the threesteps combined was double that of the sum of theincrease due to the same three steps appliedsingly, thus illustrating the large synergistic effectof the three steps when applied together in a system. There was a slight synergistic effect when

169

17©

Figure 77. Sorghum yield increases from im-proved variety (V ) , fertilization (F), and soil and crop management (M) singly and in combination over tradi-tional technology, ICRISAT Center, 1976.

any two of the factors were combined; however,the magnitude of this effect was far less. Similarsynergistic effects were observed in maize yieldsin North Carolina (Krantz and Chandler 1975),and in wheat yields in India (Krantz et al. 197S)when all major "steps" were combined in a production system.

Yields of the ratoon crop were very lowcompared to that of the main crop, due to poorratoonability and heavy attack of shoot fly. Totalyields and net benefits are given for ratoon cropas Well as the main crop (Table 60). However, noattempt can be made to make either agronomicor economic interpretation of treatment effects upon the ratoon crop because of the poor

ratoonability. In cooperation with the breeders itis planned to screen a large number of genotypesfor ratoonability. The factors affectingratoonability are also being investigated so as tocapitalize on this method of double-cropping. Assoon as good ratoonability can be achieved,treatment effects can be reliably interpreted andproper conclusions drawn.

The sorghum fodder yield of PJ8K was nearlydouble that of CSH-6. In contrast, the harvestindex of the PJ8K variety was very low (7 to 24)compared to CSH-6 which ranged from 26 to 49.

Maize/pigeonpea intercrop in Verttsols. Thetreatments and yields of maize and pigeonpea ina Vertisol are shown in Table 61. No water wasapplied to the maize crop or the main pigeonpeacrop and Treatments 9 and 10 were considered asadditional replications of 4 and 8, respectively.As in the case of sorghum, maize yields weresignificantly increased due to improved fertili-zation, but improved variety or improved soiland crop management as a single factor did notincrease yields significantly. However, with im-proved fertilization, the improved maize varietygave a highly significant response to manage-ment. These results again demonstrate the pos-itive synergistic effect of one improved practiceupon another.

In the case of pigeonpea, however, the story isquite different. Relatively little synergistic effectcan be observed. This is believed to be due to thefact that the improved variety was not responsiveto fertilizer and was not appreciably better thanthe traditional variety. Also the general yieldlevel of this and all other pigeonpea experimentsat ICRISAT Center was much lower in 1976 thanin 1975, probably because of the September andOctober drought. The drought occurred duringthe critical flowering period after the removal ofthe competing cereal intercrop. However, im-proved soil and crop management, which in-cluded closer row spacing on pigeonpea, gaveconsistent and significant increases. In bothvarieties, there was also a significant yield in-crease in the pigeonpea ratoon crop due to a 5-cm water application immediately after harvestof the main crop. Since this is the first season in

171

Three Three T w o s tepssteps steps in insingly combination c o m b i n a t i o n

Synergistic effect ofsteps in combination

Sum of yield in-creases from singlesteps

V,F,M

3000

2000

1000

0V V V

M F

MV,M

F

V,F

F

M

F,M

172

which ratooning of pigeonpea has been tried,more research is needed before attemptingconclusions.

Economic analysis of sorghum and maize/pigeonpea intercrops. In the economic analysis,we considered total costs of input and the totalvalue, including sorghum fodder. In the sorghumtrial on Alfisols, the net benefit was by far thehighest with Treatment 8. The next highestbenefit was recorded for Treatment 4, showingthat the traditional variety also responded re-asonably well to improved management andfertilization. In all treatments with traditionalF Y M fertilization, the net benefits were negative(Table 60).

The highest rate of return in themaize/pigeonpea intercrop was for Treatment 8 (Fig 78).8 The net benefits were highest forTreatments 8 and 10. Traditional methods(Treatment 1) showed negative net benefits and

Figure 78. Returns from various steps in im-proved technology with a maize/ pigeonpea intercrop on a Vertisol at ICRISAT Center, 1976-1977. Treat-ment numbers (top of bar) are ex-plained in Table 61.

8Costs recorded for an application of 10 cm/ha is Rs 590,which equate Rs 5900/ha-m or about $70/acre-foot. Tanktechnology and water application research is under way tofind lower-cost methods of providing supplementalirrigation.

rates of return, while all the other treatmentsshowed a positive rate of return (Table 61;Fig 78).

Weed management in "Steps in Improved Tech-nology" trial. Weed growth early in the rainyseason is very rapid and if weeds are not removedon time, serious yield reductions can result. InAlfisols, which dry rapidly after a rain, handweeding can be easily achieved. However, inVertisols one may face serious problems whenfrequent rains while crops are in the seedlingstage prevent the control of weeds by cultural orhand-weeding methods. During the past season,the improved technology treatments received a minimal amount of herbicides which was veryhelpful in reducing weed competition in the earlystage. This was particularly true on the Vertisols,where with traditional management the averagenumber of woman days/ha for hand weeding was48. With improved management, where alachlorwas used at the rate of 0.75 kg/ha and whereeffective cultivation was possible in the broadbed and furrow system, only 10 woman days perhectare were needed. Thus, the hand weedingcost/ha was Rs 171/ha more in the traditionalthan in the improved management. In the A l -fisols, the total of two hand weedings required 30and 56 woman days/ha for improved and tradi-tional management, respectively; hand weedingcosts in traditional and improved treatmentswere Rs 252 and 135 per ha, respectively.

Further investigations on Vertisols are underway to try to develop an effective low-cost meansof controlling weeds in the early stages under thewide range of rainfall conditions which occursfrom year to year.

Yield comparisons: random samples vs. entireplot. In the maize/pigeonpea intercrop experi-ment on Vertisols, grain yields were determinedby two methods-random selection of four subsamples (1.5 by 8 m) vs. harvest of the entire plot.Although there was some variation betweenreplicates, within given treatments the two pro-cedures correlated very closely. The average yieldof all treatments determined by the random-sample procedure was 44 kg/ha higher than the

173

Variable costs (Rs/ha)

1200 1400 1600 18001000

56

23

180160140120100806040200

-20

974

810

1

yield obtained from harvesting the entire plot.This represents a difference of 2.4 percent. This,along with the information obtained by compar-ing the two harvest systems in several water-sheds, gives confidence in the random-samplingprocedure.

Weed-management Research

Intercropping. To investigate the influence ofintercropping different crops on the trends inweed infestation, and to determine the natureand extent of weed problems in various

pigeonpea-based intercropping systems, twofield trials were conducted. Intercrops of cowpeaand maize suppressed weeds in the early stages tothe greatest extent, followed by mung, sorghum,and groundnuts. Weed infestation on both soiltypes was about the same in the early part of theseason, but late-season weeds yielded two to fourtimes higher weed weights in Vertisols than inAlfisols (Fig 79). The effect of intercrops on weedgrowth was perceptible even by the time of firsthand weeding at 25 days. Though cowpea ef-ficiently suppressed weeds in the early stages,weeds reappeared after the harvest. Systems with

Figure79. Weed growth in pigeonpea-based intercrop systems on Alfisols and Vertisols, ICRISAT Center, 1976-1977.

174

Pigeonpea (alone)

P igeonpea/pear l m i l l e t

P igeonpea/ mungbean

Pigeonpea/raaize

Pigeonpea/groundnut

Pigeonpea/sorghum

Pigeonpea/cowpea

Ver t i so l

0 30 60 90 120 15Q 180

Days after plantingDays after planting0 30 60 90 120 150 180

2 400

2 200

2000

1800

1600

1400

1200

1000

800

600

400

200

0

A l f i s o l

maize and sorghum as intercrops recorded lessweed growth at the final harvest of pigeonpea.

Preliminary weed observations recorded fromother agronomic intercropping trials indicatedthat cultural and biological factors like plant-population pressure, crop species, and genotypesaffect the crop-weed balance in intercropping.Increase in plant-population pressure resulted ina significant reduction in weed infestation, andspreading types of pigeonpea suppressed weedgrowth to a greater extent than did the compacttype.

Even though intercropping can be a potentialbiological tool for management of weeds, thesystem by itself does not completely avoid weeds.Therefore, direct weed-control studies were alsoconducted to evaluate different herbicides onsorghum/pigeonpea intercrops. In general, tr i-azine herbicides like ametryne, prometryne, andterbutryne performed well and were safe for bothpigeonpea and sorghum (Table 62). The perfor-mance of fluchloralin was excellent on pigeonpea

but was phytotoxic to sorghum, while atrazinewas quite safe on sorghum but toxic to pigeon-pea. Further weed-management studies on in-tercropping wil l be conducted to investigatemeans of manipulating various physical, biologi-cal, and cultural factors.

Double-cropping. In a trial to determine thecomparative efficacy of alachlor and atrazineapplied to dry and wet soils (after rains), therewas no significant difference in yields of sorghumand maize. The dry applications were as efficientas wet applications, even though weed-controlefficiency usually tends to be greater in case ofwet applications. It was concluded that eitherherbicide may be applied either dry or after rains,depending upon convenience, without alteringits efficacy. This information has practical impli-cations in that the herbicides may be appliedimmediately after dry planting before the rainsfor rainy season cropping in deep Vertisols of theSAT.

Table 62. Efficacy of different preemergence herbici des in CSH-5 sorghum intercropped with HY-3Apigeonpea on Alflsols, at ICRISAT Center, 1976-1977 .

Weed dry matterSorghum yield Pigeonpea yield at harvest

(% of weed- (% of weed-Treatments Rate (grain) free check) (grain) free check) Sorghum Pigeonpea

(kg/ha) (kg/ha) (%) (kg/ha) (%) (kg/ha)(kg/ha) (kg/ha) (%) (kg/ha) (%) (kg/ha)

Dinitramine 0.5 1400 39 290 85 1450 840Devrinol 1.0 1870 52 270 79 1350 1310Proraetryne 1.5 2 790 78 320 94 510 810Terbutryne 1.5 2910 81 270 79 570 1140Ametryne 1.5 3090 86 310 91 690 720

Destun 1.5 810 23 190 56 2320 1790Fluchloralin 1.5 1000 28 360 106 1710 1140Atrazine 1.5 2500 70 - - 2720 1160Alachlor 2.0 1860 52 180 53 1120 1270Weed free - 3 590 100 340 100 - -Weedy check - 1190 33 160 47 3650 2200

L.S.D. (0.05) - 660 - 140 - -

175

A trial was conducted to explore the possi-bilities of employing the minimum- and zero-tillage concept to reduce the time lag betweencrops and to conserve soil moisture, both ofwhich are very critical for the optimum growth ofa postrainy season crop of chickpea. The max-imum yields of chickpea were obtained in thefollowing three treatments: optimum tillage andhand weeding, paraquat and prometryne with nocultivation, and paraquat and three cultivations(Table 63).

Individual crops. Major emphasis is being di-rected to determine the competitiveness withweeds and herbicide tolerance of different cul-tivars of sorghum, pigeonpea, pearl millet, chick-pea, and groundnut. Field trials involving manycultivars of these crops conducted during the1976-1977 seasons indicated that crop cultivarsdiffer in their tolerance to a given herbicide andalso vary in their weed competitive ability. Asthis " intra specific differential response" may berelated to environmental conditions, continuingresearch on this problem is planned before

Table 63. Crop stubble and postrainy seasonweed management for chickpea onVertisols at ICRISAT Center,1976-1977.

Weed drymatter at

Treatments Yield harvest

(kg/ha) (kg/ha)

1 Hand weeding + three 1060 300cultivations

2 Three cultivations 790 10703 Paraquat (1.0 kg/ha) 1030 790

+ three cultivations4 One minimum cultivation 530 980

+ paraquat (1.0 kg/ha)5 Paraquat (1.0 kg/ha) 1030 650

+ prometryne(0.75 kg/ha)and no cultivation

L.S.D.(0.05) 240 420

176

drawing definite conclusions. Similarly, experi-ments are also being continued to determine theability of different crop cultivars to suppressweeds and their ability to grow well in spite ofthem. In groundnut, spreading and semispread-ing cultivars were significantly more competitivewith weeds than was the bunch type. In pearlmillet, tall and profusely tillering cultivars com-pete with weeds more efficiently than dwarf andpoor tillering types. In sorghum, plant height andinitial seedling vigor were the characters iden-tified as responsible for weed competitive ability.

Herbicide screening. To determine the croptolerance and weed-control activity of someselected herbicides, two herbicide-screening t r i -als were conducted. Visual evaluation of cropinjury and weed control from preemergenceapplications of herbicides recorded on sorghum,pearl millet, pigeonpea, chickpea, and ground-nut are as follows: On sorghum, atrazine (1 kg/ha), prometryne (1.5 kg/ha), ametryne (1.5kg/ha), terbutryne (1.5 kg/ha), and Tribunil (2.0kg/ha) showed promise. Among the herbicidestested on pearl millet, only atrazine (0.5 kg/ha)and Tribunil (1 kg/ha) were found safe on thecrop. Nitrofen (2 kg/ha), dinitramine (0.5 kg/ha),prometryne (1.0 kg/ha), ametryne (1.0 kg/ha),terbutryne (1.0 kg/ha), Modown (2.0 kg/ha), andTribunil (2.0 kg/ha) were all found to be effectiveon pigeonpea. Among the herbicides evaluatedon groundnut, dinitramine (0.5 kg/ha), alachlor(2.0 kg/ha), nitrofen (2.0 kg/ha), Prefar (4.01/ha),and devrinol (4 1/ha) proved promising. Onchickpea, nitrofen (2.0 kg/ha), alachlor (2.0kg/ha), dinitramine (0.5 kg/ha), Prefar (4 1/ha),prometryne (1.0 kg/ha), ametryne (1.0 kg/ha),and terbutryne (1.0 kg/ha), were found to beeffective.

Weeds and weeding systems survey. In a fewselected villages representative of the IndianSAT, surveys were made in collaboration withICRISAT economists to observe weed problems,weeding systems, and the extent to which thefarmers understand their weed problems. Theexisting methods of weed control are handweeding with a small hoe and intercultivating

with the traditional blade harrow or "Guntaka."The weed control activities of farmers of Mah-bubnagar and Akola villages is clearly related tothe quality of the resource base. The better thegrowth environment for the crops and weeds, thebetter the weed control. On the farms where weedmanagement was not satisfactory, the majorweeds found were Cynodon, Cyperus, Des-modium, Digitaria, Echinochloa, Amaranthus, Celosia, Corchorus, Crotalaria, and Ipomoea.

Look ing Aheadin the Fa rm ing SystemsSubprograms

Agroclimatology. Correlation of crop yields toweather conditions at a network of bench-markSAT locations and determination of transfera-bility of agroproduction technology among dif-ferent areas will be attempted. Characterizationof crop moisture availability in order to delineatehomogeneous areas in the SAT is also beingundertaken.

Activities during the initial phases of this workwill include (i) establishing contacts for develop-ment of a meteorological data bank for locationsof interest in the SAT; (ii) conducting agricul-tural climatic characterization studies for se-lected locations of Africa, Latin America, andIndia; and (iii) determining relationships be-tween crop production and environmentalfactors.

Hydrology. Collection of hydrological data onalternative resource-management systems willcontinue at ICRISAT Center as well as in thecooperative programs. Work on refinement ofsediment samplers-to improve accuracy andrel iabi l i ty-wi l l continue. Data analyses wil l becomputerized, and simulation of the runoff pro-cess will be attempted in cooperation with otherresearch organizations.

Soil physics. Field studies of profile-water re-charge and use wil l be conducted and detailedwater balances for different periods of time wil l

be computed for crops grown on Vertisols andAlfisols during the rainy and postrainy seasons.In cooperation with ICRISAT's physiologistsand climatologists, a more-complete quantifi-cation of crop development, moisture stress,energy balances, and phasic root developmentwill be used together with data on profile-waterdynamics to evaluate a water-use and -yieldmodel suitable for the soil-crop-atmosphere con-tinuum at ICRISAT Center.

Soil fertility and chemistry. The long-termphosphate rock experiment will be continued.More experiments for the study of response ofsorghum to N in an intercrop situation and pearlmillet/groundnut intercrop responses to N and P are planned. Monthly sampling for determi-nation of nutrient status will continue.

Farm power and equipment. Ongoing projectswil l be continued, with emphasis on adaptationof available farm machinery and developmentand testing of additional equipment needed toimplement improved farming systems. Specialemphasis will be given to adaptation of simplelow-cost planting units which can sow success-fully on Vertisols and Alfisols under a wide rangeof moisture conditions.

In multicropping situations in the SAT, thefirst crop is frequently harvested during the endof the rainy season. Thus, considerable dryingmay be required to bring the grain moisturecontent to a safe level for storage. Two ap-proaches to this problem will be studied —

(i) Maize wil l be stored on the ear and thenshelled after it is dry. Open-sided cribs,through which air may pass freely, wil l bebuilt for storing the maize ears.

(ii) A solar grain dryer, using a flat-plate col-lector to heat air by solar energy, is beingdesigned. The prototype wil l be used to dry,on an experimental basis, sorghum andmillet grain during the late rainy season.

Land and water management. Present projectswill be continued in order to improve metho-dology and to provide additional data on

177

land-management and supplemental-water-utilization techniques. Research on improvedpractices wil l be initiated at several locations inIndia in addition to that at ICRISAT Center,and similar activities may be initiated at SATlocations in West Africa. Investigations (incooperation with ICRISAT economists) of thewater balances and water-use efficiencies in exist-ing small- and medium-scale runoff collectionfacilities in Indian villages have begun; theobjectives of these studies is to develop pro-cedures that wil l provide a more effective use ofavailable water. The possibility of involvingother research institutions in this phase of theprogram is being explored.

Cropping systems. In the intercropping work, itis intended to put considerable emphasis ondetailed study of growth and resource use incombinations of proven advantages, with thehope that ways in which yield advantages areachieved may be identified and improved. De-tailed work on plant-population effects wil lcontinue, and genotype studies wil l be increased,hopefully to include all the ICRISAT crops. Inthe sequential and relay cropping studies, com-parisons of maize and sorghum as rainy seasoncrops will continue; postrainy season crops willbe limited to chickpea, pigeonpea, and sorghum.Sorghum genotypes wil l be screened forratoonability.

In cropping systems entomology, larger plotswil l be used to provide a more-realistic " f ie ld"situation that wil l permit regular sampling with-out seriously affecting population levels. Pest-parasite surveys in farmers' fields will be expand-ed. With the indications that some insecticidesprays may still be needed in intercroppingsituations, a prime objective of this phase of ourwork will be development of integrated pest-management systems. Importance of plant type,plant populations, and spacing wil l be furtherstudied.

Collaboration with organizations-such asCOPR, Boyce Thompson Institute, and CIBC -on the possibility of biological and viral controlof Heliothis is being discussed and a survey ofbiological and microbiological agents on this

insect in India is planned for the coming season.Plans for establishing a trap grid throughout thesubcontinent so as to study migratory behaviorof H. armigera in India will hopefully mature.

Agronomy and weed science. Further studies todetermine the effects of different crop com-binations, sequences, geometries, and genotypes,along with other physical, biological, and cul-tural factors on weed growth are planned. In thecrop-oriented weed research, the major emphasiswill be to determine growth characteristics re-sponsible for weed competitiveness and to studythe differential tolerance of different cultivars ofICRISAT crops to commonly used herbicides. Inaddition to the field trials at ICRISAT Center,observations in a few selected villages are plan-ned to measure the success of farmers' own weed-control methods and to assess the payoffs ofadditional weed control over the existing system.This effort will help to develop an understandingof the crop, soil, climatic, and social situations inwhich improved weed management could havethe greatest impact.

A minimal forage-crop program is being in-itiated to evaluate forage grass and legumemixtures as to their yield of palatable forage, soil-erosion control, longevity through hot and dryseasons under heavy grazing pressure, and ra-pidity of regrowth at the onset of the rainyseason. These features are important formanagement of grassed waterways and tankbunds. The "Steps in Improved Technology"experiment wil l be continued with sorghum onVertisols and pigeonpea/pearl millet intercrop onAlfisols.

W a t e r s h e d - b a s e dResource U t i l i z a t i o nResea rch

The watersheds described in Table 64 representICRISAT's field laboratory for "systems re-search." Here the concept of a watershed-based

178

farming system is examined on an operational orfield scale, using animal draft power. This in-volves the monitoring of data on a Wide range ofactivities - which include inventories of naturalresources, complete water-balance studies(measurements of rainfall, rainfall intensity,runoff, soil erosion, nutrient loss, consumptiveuse by plants, and deep percolation into shallowgroundwater), investigations of rainfall-use ef-ficiency, total annual production, human laborand bullock power utilization, and fertilizer,pesticide, and other material inputs used in thesystems. The data are interpreted for each of thesoil-, water-, and crop-management treatmentson Alfisol and Vertisol watershed units.

Since water is the first limiting natural factorfor crop production in the SAT, improving themanagement of the water and the soil for in-creased crop production becomes the primaryaim of the Watershed-based Resource Ut i l -ization research. In rainfed agriculture, only therain which falls in a given area is used, thus the

watershed or catchment is the natural focus forstudies of water management in relation to cropproduction, resource conservation, andutilization.

In the operational-scale research phase, econ-omic evaluations are made in cooperation withthe staff of the Economics program. A l l laborand draft-animal time spent in each watershedunit is recorded. At the end of each crop year, theinput-output data from each of the watershed-management treatments being compared is as-sembled and reviewed. Measurements of relativeprofitability, rates of return on capital, labor,and animal power use, etc., are made and thosetreatments which appear consistently promisingcan be singled out for further study. Data fromICRISAT and other research institutions arealso being employed in studies involving modell-ing and simulation of rainfall-runoff relation-ships and crop yield-moisture responses. Theobjective here is to assess the potential for waterharvesting and supplementary irrigation under a

Table 64. Land-, water-, and crop-management treatments on watersheds at ICRISAT Center.

Watersheda

Plantingmethod

Technologyusedc

Bundtyped

RunoffStorage (m3)eNo. Areab

Plantingmethod

Technologyusedc

Bundtyped

RunoffStorage (m3)e

Deep Vertisols

BW1 3.41 Bed (0.6%) Impr Graded -BW2 3.96 Bed (0.6%) Impr Field -BW3A 5.21 Bed (0.4%) Impr Graded 4000BW3B 2.15 Flat Impr Field 4000BW4A 3.07 Flat Impr Field -BW4B 2.48 Flat Impr Graded 4000BW4CN f 1.31 Flat Impr Field -BW4CS f 2.14 Flat Trad Field -B W 5 B N f 0.45 Bed (0.4%) Impr Graded -BW5BS f 0.67 Bed (0.4%) Trad Graded -BW6A 1.67 Flat Trad Field -B W 6 B N f 2.90 Flat Impr Contour -BW6BS f 1.79 Flat Trad Contour -BW7A 3.74 Bed (0.6%) Impr Graded 2000

continued

179

Table 64 continued

Watershed8

Planting Technologymethod usedc

Bundtyped

RunoffStorage (m3)eNo. Areab

Planting Technologymethod usedc

Bundtyped

RunoffStorage (m3)e

Medium Deep and Shallow Vertisols

BW6C 2.66 Flat Impr Contour -BW6D 1.80 Flat Impr Contour 6500BW7B 2.70 Bed (0.6%) Impr Graded 2250BW7C 2.54 Bed (0.6%) Impr Graded 1750BW7D 4.09 Bed (0.1%) Impr Graded 4100BW7Eg 0,89 Bed (0.6%) Impr Graded -BW7F g 0.73 Bed (1.0%) Impr Graded 500BW8A 2.37 Flat Trad Field -

Alfisols

R W I C 0.86 Flat Impr Graded 2900RW1D 1.15 Bed. (0.6%) Impr Graded 2900RW1E 1.58 Flat Impr Graded 2900RW2B 2.87 Bed (0.6%) Impr Graded 12000

Grassed Watersheds

BW12 2.25 - Ungrazed - -RW3A 2.87 - Ungrazed - -RW2A 28.90a- Grazed - -

aThe two cropping systems applied on deep Vertisol watersheds BW1, BW2, BW3 A, BW3 B, BW4 B, BW7 A, and on mediumdeep Vertisol watersheds BW6 C/D, BW7 B, BW7 C, BW7 D, BW7 E, and BW7 F consisted of a maize/pigeonpea intercrop andmaize followed by a sequential chickpea crop.Deep Vertisol watersheds B W 4 A and B W 6 A were sown to a sorghum/pigeonpea intercrop and sole sorghum.Deep Vertisol watersheds BW4C, BW5 B, and BW6 B were fallowed in the rainy season; sorghum and chickpea were grown inthe postrainy season. On medium deep Vertisol watershed BW8 A, a sorghum/pigeonpea intercrop and a sole sorghum crop were

grown.On Alfisol watersheds, sorghum, groundnut and a Setaria/pigeonpea intercrop were grown in the rainy season and a ratoonsorghum crop was grown in the postrainy season. (Varieties and yields are given in Tables 71, 72, and 73.)

bAll areas in hectares; large portions of BWS B, BW8 B and RW2 are not included, these areas are used for field-scale replicatedexperiments.

c Improved technology Indicates the following: Improved cultivars. Improved implements with animal power. Init ial landpreparation (plowing) on Vertisols immediately after harvest of the last crop to»kill weeds and stubble; land preparation iscompleted during the extended dry season. On Alfisols, initial plowing is executed whenever an early shower occurs. Grassedwaterways are established. Plant protection is executed with minimum use of pesticides. Fertilization: 75 kg/ha of 18-46-0 atplanting and 67 kg N/ha side-dressed.Traditional technology simulates local practices: Local varieties. Traditional implements with animal power. Tillage is startedafter rains soften the soil. Farmers' field bunds remain the field boundaries. Plant protection with minimum use of pesticides.Fertil ization: Farmyard manure at 50 cartloads/ha every second year.

dGraded bunds in bed plantings are low terrace-cum-channels which serve as drains for the bed-and-furrow system and as turn

space. Graded bunds in flat plantings are broad bunds with a channel of 0.4 % grade. Field bunds are the original villagers' fieldboundary bunds. Contour bunds are of standard size and design according to locally adopted specifications.

eSeveral watersheds are sometimes served by the same runoff storage facility (e.g. BW3 A, BW3 B, and BW4 B); those watersheds,where supplemental irrigation is available, are frequently divided into a North and South (or East and West) section, water beingapplied to only half of the watershed.

T h e rainy season-fallow watersheds (4 C, 5 B and 6 B) are in the postrainy season divided into subunits where improved and localtechnology is applied.

(The watershed units BW7 E and F have been eliminated from comparisons because of the presence of eroded soils, salinity, and a disproportionate area under field bunds (see Table 70).

180

wide range of agroclimatic and economicenvironments.

By investigating all facets of resource in-ventory, development, management, and util-ization and all factors involved in crop pro-duction in a systems approach, ICRISAT scien-tists expect to develop alternative viable systemsfor farmers in the SAT. Although the research isfocused on the small farmers of limited meanswho comprise the majority of the farm pop-ulation of the SAT, the large farmers who occupymuch of the land and employ most of the laborare also expected to benefit.

Watershed Operat ions andDevelopment

Research Watersheds on Vertisols andAlfisols

The locations and layout of present and pro-posed research watersheds were presented inICRISAT's Annual Report for 1975-1976.Some important characteristics for each water-shed unit studied during the 1976—1977 seasonshave been summarized in Table 64. On the basisof previous results, a broad-bed system of 150 cmwidth was applied to several watersheds (Fig 80).

On the deep Vertisols, crops were again plan-ted in dry soil during the first half of June.Experience over the past 4 years shows that thispractice appears quite suitable for establishingcrops early, thereby gaining in effective length ofgrowing season and in attaining greater rainfall-use efficiency. H owever, the success of thistechnique depends upon: (i) the feasibility ofplanting relatively deep ( > 6 cm) to protect theseed from germination after early, small showersand also upon seedling tolerance to potentialdrought stress (e.g. sorghum, pigeonpea, andmaize), (ii) sufficient reliability that after theselected sowing date, rains-once they beg in -will continue [the probability of a rainy week( > 20 mm) followed by another rainy week mustexceed 45%], and (iii) absence or control ofrodents or other factors which might damage theseed in dry soil.

Figure 80. One row of pigeonpea and two rows of a cereal crop on a 150-cm bed on a Vertisol at 1CR1SA T Center.

Watershed Development and FeasibilityStudies

Lined, elevated inlets for runoff storage units(tanks) were used previously to create above-ground storage and thereby increase storage-to-excavation ratios in BW7. Satisfactory ex-perience with unlined inlet channels justified theremoval of the cement and stone slab lining fromthe BW7 B tank which decreased costs from Rs2.22/m3 of storage to about Rs 1.75/m3. Previoustank construction in BW7 was executed usingexclusively human labor and draft animals. A new tank was constructed below BW8 using 40hr of bulldozer time to build the elevated inletchannel and the bund up to the elevation wherethe bund width was adequate for efficient dozeruse; labor was then used to finish further bundconstruction and shaping. This circular tank(capacity 4800 m3, storage/excavation ratio = 2.6) was completed at a cost of Rs 1.10/m3 ofstorage (Fig 81); these results compare favorablywith earlier tank development costs in BW7which ranged from Rs 2.00 to 2.50/m3 of storage.

181

Figure 81. The BW8 tank partially filled follow-ing intense August storms.

Water-balance Studies

Runoff

In watersheds on deep Vertisols, which werecropped in the rainy season and cultivated tograded beds, 7 to 10 percent of the seasonalrainfall ran off; while under flat-planted con-ditions runoff varied from 10 to 17 percent(Table 65). Runoff from the BW3 B watershedexceeded that observed from BW1 and BW3 A.These observations again confirm the hypothesisthat a bed-and-furrow system can be effectivelyused to uniformly conserve moisture on the land.This is also evident from Table 66; in 1973 and1974, when the BW3 B watershed was cultivatedto a 0.8- and 0.4-percent graded ridge-and-furrow system, respectively, runoff from thisarea was very similar to that on BW1 which hasalways been maintained in a 0.6-percent graded-ridge treatment. In 1975 and 1976 when BW3 B was changed to a 1-percent graded-ridge systemand flat field-bunded conditions, respectively,runoff was (in relation to BW1) considerablyincreased.

Runoff under cropped contour-bunded con-ditions (BW6 C/D) was of a magnitude similar to

the bed-and-furrow watersheds; however, in thissituation additional infiltration occurred on onlya small portion of the land and drainage pro-blems frequently existed due to stagnant waterabove the bund. (See Land and Water Manage-ment subprogram.) Runoff on cropped, bedded,medium deep Vertisols (BW7 B,C,D, and F)amounted to about 10 percent with somewhathigher values on steeper (1.0%) slopes.

The largest quantities of runoff were observedunder rainy season fallow, field-bunded con-ditions; on BW4 C about one-third of theseasonal rainfall (and 74% of the 21 Jul and 18,19 Aug precipitation) ran off. A comparison ofthe "fal lowed" watersheds BW4 C, BW5 BS andBW6 B indicates that broad beds in a rainyseason fallow situation affect total runoff to a similar degree as contour bunds. However, thedisadvantage of inadequate drainage on part ofthe land (Fig 82), associated with contour bunds,does not exist under ridged conditions; plantingcan be executed within a few days after rain.Although this management system is not suitablefor Vertisols at ICR1SAT Center, this findingmay be relevant in those areas where the earlyrainy season showers are less dependable andthus "dry planting" is more risky. Peak runoffrates observed this year (and also earlier) indicatethat under cropped bedded conditions and underrainy season fallow, values of 0.05 to 0.1 and 0.1to 0.15 m3/sec per ha, respectively, are obtained.

Except under contour-bunded situations,runoff on the Alfisols amounted to about 20percent of the seasonal rainfall (Table 67). Onshallow soils runoff was considerably increasedby beds and furrows (RW1 D); in case runoff iscollected for later re-use, this may be desirable onlight soils as long as erosion is not excessive.During high-intensity long-duration storms, al-most 50 percent of the rainfall was lost under flatas well as bedded planting. Peak discharge rateson these soils generally range between 0.15 and0.20 m3/sec per ha.

Erosion

Soil losses measured at watershed outlets arepresented in Table 68. Although the observed

182

183

Figure 82. Ponded water above a contour bund on rainy season fallowed land.

erosion from cropped catchments on Alfisolsgenerally exceeded that which occurred in similardeep Vertisol situations, all quantities are wellwithin acceptable limits.9 The presence of gradedbeds and furrows under rainy season fallowconditions reduced erosion considerably com-pared to a flat, field-bunded situation (Table 68,BW5 B vs. 4 C); the observed difference would beeven greater if erosion within watersheds wereconsidered (Fig 83). The highest intensity long-duration storm (on 19 Aug) caused 80 percent ofthe seasonal soil loss under rainy season fallow ina flat field-bunded watershed (BW4 C). Soilerosion apparently increases much more withgreater rainfall intensities under rainy seasonfallow than under cropped beds and furrows(Fig 84).

Figure 83. Channel deposition in an uncropped Vertisol following a heavy late-season rain at ICRISAT Center.

9The runoff collection tank on BW5 A (about 6.5 ha) wascleaned from sediment deposited in 1974, 1975, and 1976;the total quantity of eroded material removed was 18metric tons, an average of less than 1 ton per hectareannually.

184

Table 66. Rainfall and runoff on BW1 and BW3B at ICRIS AT Center, 1973 through 1976.

BW1 BW3 B

Year Treatment Rainfall Runoff Treatment Rainfall Runoff

1973197419751976

Ridged 0.6%Ridged 0.6%Ridged 0.6%Beds 0.6 %

(mm)

700810

1040690

(mm)

45 Ridged 0.8 % 116 Ridged 0.4%161 Ridged 1.0%73 Flat

(mm)

660810

1050700

(mm)

44105193105

An experiment comparing broad beds andfurrows with flat cultivation at 0.4- and 1.0-percent slopes was conducted on an Alfisol byK. A. Shams, a graduate student from Sudan.10

The first runoff-producing storm occurred 18days after planting a sorghum/pigeonpea in-tercrop when the crop was small (Leaf AreaIndex of 0.3) and thus did not provide muchplant cover. Even though treatments did notinfluence the amount of runoff, the amount oferosion under flat cultivation was 2.7 times thatof the bed-and-furrow system (Table 69). Therewas no significant difference between slopes, butthe erosion tended to be higher at the 1.0-percentslope. The pattern of subsequent storms wassimilar; however, the amount of erosion per rainwas greatly reduced and the magnitude of differ-ences much less due to the overriding effect of theincreased plant cover later in the season.

10Shams, K. A., 1977. "The effect of bed vs. flat cultivation attwo slopes upon runoff, erosion, crop growth, and yield."Thesis submitted to Andhra Pradesh Agricultural Un-iversity in partial fulfillment of the requirements for thedegree of Master of Science in Agriculture.

185

Figure 84. Effect of rainfall intensity on soil loss in a fallowed (BW4 C) and a cropped (BW1 ) watershed on a deep Vertisol.

Weighted mean intensity (mm/hr)

0 5 10 15 20 25 30 35 40

BW 4C

BW 1

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.01

0

Table 69. Soil erosion from first erosive storm(90 mm on 21 Jul 1976) on an Alfisol atICRISAT Center.

Cultivationmethod

Percent slopeCultivation

method 0.4 1.0 Mean

FlatBroad ridge

L.S.D. (0.05)

(kg/ha) (kg/ha)

852 1309327 470

(kg/ha)

1081399365

Groundwater

Most groundwater levels, measured at piezom-eters, indicate no substantial change betweenMay 1976 and May 1977 (Fig 85). An exceptionis found where observations were taken near tonewly constructed Lake ICRISAT (BW8). Ondeep Vertisols, the groundwater levels increasedby about 250 cm during the rainy season; this is

equivalent to approximately 160 mm of water. 11

Thus, about 25 percent of the seasonal rainfallpercolated through the root profile. On shallowand medium deep Vertisols, the storage capacityof the root profile is filled more rapidly andtherefore groundwater levels increase earlierthan in the deep Vertisols (Fig 85); the totalquantity percolated through the profile duringthe rainy season amounted to about 265 mm, or40 percent of the precipitation. On the shallowAlfisol watersheds, the quantity of water lost togroundwater can, from lysimeter observations,be estimated at about 230 mm, or about 35percent of the rainfall. Thus, even during a rainyseason of moderate total precipitation, very largequantities of moisture may be lost from the rootzone particularly if, as in 1976, the duration ofthe rainy season is short while the total seasonalquantities are near normal.

Crop Season Evapotranspiration

Detailed data on evapotranspiration of severalcrops during both seasons were collected by theSoil Physics subprogram making use of ly-simeters, neutron moisture probes, and ten-siometers installed on watershed units (see page144). The precipitation from 1 to 21 June (about35 mm) did not contribute to plant growth,although final presowing cultivation was facilit-ated. The evapotranspiration of a sole maize (cvDeccan Hybrid 101) crop of 100 days duration,grown on deep Vertisols, was estimated at 350mm; rainfall was well distributed and crops werenot subject to drought stress. Maize (cv SB-23)grown as an intercrop in pigeonpea was of 93days duration, and the evapotranspiration of theintercrop combination was estimated at about310 mm up to maize harvest time. No detaileddata were collected on medium deep and shallowVertisols; however, no serious wilting of cropswas observed and therefore similar figures maybe used. On the basis of lysimeter data frommedium deep Alfisols, it can be estimated that

11 The groundwater component is computed on the basis ofan estimated S % available storage capacity of the deepsubsoils.

Table 68. Erosion on Vertisol and Alfisol water-sheds at ICRISAT Center, 1976.

Watershed Treatment Soil loss

(kg/ha)

BW1 R/F ,0 .67% 800BW2 R/F,0.6% 300BW3A R/F, 0.4% 300BW4A Flat, bunds 700BW4B Flat, Gr. bunds 900

BW4C Fallow, bunds 9200BW5B Fallow, R/F, 0.4% 3800BW6A Flat, bunds 2200BW6C/D Flat, C. bunds 200BW8A Flat, bunds 500

RW1C Flat, Gr. bunds 2000R W 1 D R/F, 0.6% 1800RW1E Flat, C. bunds 400RW2B R/F, 0.6% 2000

186

the evapotranspiration for sorghum (CSH-6)was approximately 340 mm; although sorghumis of longer duration than maize, once dryweather had set in during September, wilting wasobserved in both sorghum and groundnut onRW1 and RW2.

Beginning after the first week of September,soil moisture levels started to decline and post-rainy season crops were sown under extremelydry conditions; emergence was poor and standsinadequate in nonirrigated areas. Newly plantedsorghum on deep Vertisols can be estimated tohave used about 200 mm of water (1.70 mm ofprofile moisture and 30 mm of rainfall). Wherechickpea was sown on previously fallow deepsoils, or where it was irrigated (with about 60mm) after planting, the evapotranspiration dur-ing the postrainy season can be estimated atapproximately 180 mm. Chickpea sown as a second crop without supplemental irrigationmay not have used more than 75 mm on the deepVertisols and even less on medium deep soils.The evapotranspiration of a ratoon sorghumcrop on deep Vertisols amounted to less than 130mm; on Alfisols, where the ratoon sorghum wasirrigated twice, only about 125 mm was used.Pigeonpea grown on deep Vertisols used about175 mm after the maize harvest.

The estimates of postrainy season evapo-transpiration on Vertisols do not include contri-butions through upward water movement fromsoil layers below 180 cm depth. There is some

Figure 85. Comparison of seasonal changes in groundwater levels in a medium to shallow Vertisol and a deep Vertisol. Surface elevation of the medium to deep Vertisol is 533.02 meters; that of the deep Vertisol is 537.82 meters.

evidence that this quantity, particularly in a drypostrainy season like 1976, can be quite sub-stantial. From moisture data collected in June1976 and May 1977, it appears that total profilemoisture up to 180 cm depth decreased by about30 mm during this period.

In conclusion, total evapotranspiration of theintercropping systems of maize and pigeonpeacan on deep Vertisols be estimated at about 485mm (310 and 175). When sequential crops ofmaize and chickpea were grown without i rr i -gation, the total evapotranspiration was 425 mm(350 and 75); when the chickpea was sup-plementally irrigated up, approximately 530 mm(350 and 180) was used. In situations where onlya rainy season crop was grown on Vertisols,seasonal evapotranspiration probably did notexceed 400 mm even where somewhat longer-duration crops were used. Only approximately200 mm was used if a single postrainy seasoncrop was grown. On Alfisols, the total evapo-transpiration of sorghum followed by an irr i -gated ratoon crop was approximately 465 mm(340 and 125). The evapotranspiration of a singlerainy season crop on Alfisols was about 340 mm.

Runoff Collection and Use

Substantial quantities of runoff occurred onlytwice during the rainy season. In tanks with highseepage rates - such as most of those on BW7 -much of the water collected during the firstrunoff event in July had disappeared by the timehigh-intensity rains of sufficient duration occur-red again in August. A sequential crop ofchickpea was sown around 5 October on Ver-tisols planted to sole maize in the rainy season.Soil moisture had by this time receded to suchdepths that planting into the moist zone byanimal-drawn equipment was successful in onlypart of the watershed areas, thus very unevenstands resulted. Because rainfall probabilities arehigh in September and early October [the pro-bability of a wet week (20 mm) following a dryweek generally exceeds 50 % ] , supplemental i rr i -gation was delayed until mid-October.

When it was decided to apply water to chick-pea on Vertisols for germination and stand

187

M A M J J A S O N D J F M A M J

Deep Vertisol

Month

Medium toshallow Vertisol

538537536535534533532

531. 0

establishment and to sorghum on Alfisols forratooning, water remained in only the BW3,BW5, BW6, BW7 A and D, RW1, and RW2taiaks. For deep tanks, less than 10 percent of thecollected water evaporated. From 10 to 90percent of the stored water was lost due toseepage in different tanks; the greatest seepagelosses occurred in medium deep Vertisols. Therelative quantity of water actually used fromtanks therefore varied considerably-on BW3about 85 percent of the collected runoff wasapplied to the land. This is equivalent to appro-ximately 40 mm on the contributing watershed.On the RW1 watershed, the quantity used forsupplemental irrigation amounted to only 45percent of that collected. However, because ofgreater runoff from these soils, this was equiva-lent to about 85 mm.

Yield response under supplemental irrigationcompared to crops grown on residual moisturealone varied considerably, due to differences instand establishment at the beginning of the dryseason for both chickpea and ratooned sorghum.A 20-mm rain occurring 6 November causedmost of the nonirrigated chickpea to germinate,which further complicated comparisons. Forexample, on BW3 A (broad-ridged) supplemen-tal irrigation (80 mm) at planting increased yieldson the average from 780 to 1 310 kg/ha; underflat-sown conditions the respective yields were480 and 850 kg/ha on BW3 B and 310 and 560kg/ha on BW4 B. General yield levels on themedium deep Vertisols were lower; supplementalirrigation at planting increased yields on BW7 B from 500 to 1000 kg/ha and on BW7 C from 320to 1080 kg/ha. The effect of supplemental i r r i -gation on ratoon sorghum on Alfisols was disap-pointing, yields on RW2 B increased from 540 toonly 1060 kg/ha after two irrigations of 50 mmeach, one at time of harvest of the first crop and a second during heading of the ratoon crop.

Water Balances and Effective Rainfall

In deep Vertisols the total evapotranspirationduring the crop season for the intercrop situ-ations was about 485 mm and for sequentialcropping approximately 425 mm. This "effective

rainfal l"1 2 constituted about 70 and 60 percentrespectively of the total rainfall in the growingseason. Of the remaining portion of the pre-cipitation, around 10 percent was lost as runoffand 25 percent disappeared from the profilecontributing to groundwater. In watershed unitswhere runoff was collected and re-used forchickpea, the actual runoff loss decreased byabout half; however, effective rainfall was in-creased to almost 70 percent because of greaterprofile-water use under supplemental irrigation.Minor discrepancies in the water balance may beexplained by small changes in the profile-moisture status or by inaccuracy of the ground-water estimate. On double-cropped flat-plantedwatershed units, the effective rainfall may havebeen of similar magnitude. Runoff averagedaround 15 percent, thus the groundwater re-charge was slightly less at about 20 percent.

The application of improved technology onthe medium deep Vertisols resulted (on some-what steeper slopes under ridged conditions) in a runoff of about 10 percent; however, the portionof the precipitation contributing to groundwaterwas much larger (around 40 %) due to the limitedmoisture-holding capacity of these soils. There-fore, effective rainfall contributing to cropgrowth on these medium deep soils amount-ed to only about 50 percent of the rainfall. Therelatively low effective rainfall on the mediumdeep Vertisols is an important factor explainingthe low yields obtained from the postrainyseason crops.

On the Alfisols, runoff amounted to 20 to 25percent of the precipitation in ridged systems; thegroundwater component can be estimated at 35percent and the actual evapotranspiration of a rainy season crop at only about 45 percent.

The traditional practice of growing only onecrop, either in the rainy season or in the postrainyseason, resulted in a relatively low effectiverainfall. On deep Vertisols where a rainy seasoncrop was grown, the actual evapotranspiration

12 "Effective rainfal l" is defined as the percentage of theprecipitation during the crop season (for a two-cropsystem, generally taken as 1 Jun to 1 Feb) actually used asevapotranspiration by a soil-crop complex.

188

was no more than 50 percent of the seasonalrainfall, about 20 percent ran off and 25 to 30percent percolated down to groundwater. Onmedium deep soils there was less runoff andtherefore a proportionately greater groundwatercomponent. The rainy season-fallow system, asin previous years, resulted in the smallest effec-tive rainfall (25-30 %), large quantities of runoff(35%), and a substantial groundwater contri-bution (much of the rainfall in this system is lostas evaporation from bare soil and transpirationfrom weeds during the rainy season).

From the discussion of water balances, it isevident that even in a relatively low rainfall year,very substantial quantities of water are notavailable for crop production due to runoff anddeep percolation from the profile. Appro-ximately the same rainfall occurring on deep andmedium deep Vertisols and Alfisols resulted ingreatly different moisture environments for plantgrowth, length of growing season, and totalwater use. In 1976, management methods result-ing in decreased runoff had as their primaryeffect an increased groundwater component. AtICRISAT Center and in many similar situationsacross the SAT, groundwater is not easily ac-cessible due to poor aquifer conditions. Thus,two areas of research need to be furtheremphasized - (i) exploration for a better use ofthe root profile storage capacity and (ii) testingof management methods which increase runoffinto small watershed tanks (particularly in thelater part of the rainy season) without causingsignificant erosion.

Crop Production andEconomic Data

The main objective of resource development andmanagement is to increase agricultural pro-duction on a sustained basis. To accomplish thisgoal, systems must be developed which caneffectively utilize the current season's rainfalldirectly, as well as supplemental water fromavailable surface-stored water and/or ground-water. This system must also minimize soil

erosion and conserve resources for future gen-erations. By researching alternative soils- andcrop-management systems in field-scale oper-ational units, it is possible to study means ofincreasing and stabilizing production as well aspotentials for reducing peak labor and draft-power demands. Emphasis is being given to thedevelopment of "non-monetary" inputs such astimeliness of operations and improvement ofsoil- and crop-management practices in order tooptimize the returns from improved seed, fertili-zation, plant protection, and improved animal-drawn equipment.

Yield Comparisons of Harvest of RandomSamples vs. Entire Plots

In most watersheds, entire replicates (0.2 to 0.8ha) were harvested and yields determined afterrandomly selected yield samples had been taken.In sole maize, the average yield computed fromthe sample-based system was 120 kg/ha less thanthat of the whole-plot-based system. In the caseof intercropped maize, the average of the sample-based yields were 75 kg/ha less than the plot-based yields. Thus, it appears that the random-sample harvest procedure gives a sufficientlyaccurate measure of whole-plot yield.

In order to obtain information on the netproduction in each watershed unit, yields wereadjusted according to the actually cropped areain each watershed. In other words, the areaoccupied by field bunds, contour bunds, andgrassed waterways was subtracted in order toobtain the percentage of net cropped area in eachwatershed unit. As shown in Table 70, there isconsiderable variation in net cropped area; this isparticularly true in BW2 where wide field-boundary bunds were retained and in the smallunits in the lower part of BW7.

Crop Production on Vertisols

In 1976, two cropping systems (a maize/pigeonpea intercrop and sole maize followed by a chickpea sequential crop) were adopted on mostof the watershed units on the Vertisols. In otherunits, sorghum/pigeonpea intercropping and

189

Table 70. Total land area and percentage netcropped area in the Vertisol watershedunits at ICRISAT Center.

Total Net croppedWatershed area area

(ha) (%)

BW1 3.41 95BW2 3.96 87BW3A 5.21 89BW3B 2.15 92BW4A 3.07 91BW4B 2.48 87

BW4C 3.45 97BW6A 1.67 86BW6B 4.69 90BW6C 2.66 79B W 6 D 1.80 87BW7A 3.74 84

BW7B 2.70 81BW7C 2.54 81B W 7 D 4.09 80BW7E 0.89 78BW7F 0.73 62BW8A 2.37 93

sole sorghum plus ratoon-cropping systems weregrown. These double-cropping systems werecompared to traditional postrainy season singlecropping under different soil- and water-management practices (Table 71). Each croppingsystem was duplicated in each watershed unitand the crop yields and rupee values given aremeans of the two replicates. Yield and economicdata presented in Tables 71, 72, and 73 are fromnonirrigated areas.

On deep Vertisols, yields of the broad bed- andfurrow-system (BW1, 2, 3 A) were higher thanthose in the adjacent flat-planted watersheds(BW3 B, 4 B). The average gross returns of thethree bedded watersheds for the intercrop andsequential crop systems were Rs 4920 and Rs3680 per ha, respectively (Table 71A). These

values were Rs 710 and Rs 810 per ha higher thanthe comparable flat-planted systems. BW7 A wasexcluded from the comparisons because of poorgrowth due to inadequate drainage in past years(see footnote b., Table 71 A). In the sorghum-based system13 gross benefits were about thesame as in the flat-planted maize-based systems(Table 71). As in past years, the most strikingcontrast in production was that of the improveddouble-cropping on beds versus traditional flat-planted single cropping in the postrainy season.The gross benefit of traditional postrainy seasonsorghum was only Rs 950, or 19 percent of theimproved intercrop system (Table 71 A, C).Another disadvantage of the postrainy seasonsingle-cropping system is that clean cultivation ispracticed four or five times during the rainyseason, which leaves the soil unprotected andsubject to raindrop erosion during high-intensitystorms (see page 182).

The yields and gross monetary values of themaize/pigeonpea intercrop on both the deep andmedium deep Vertisols were consistently muchgreater than that of the maize-chickpea sequence(Tables 71,72). This large difference in the 1976-1977 season was due to (i) poor-quality DeccanHybrid 101 seed used in the sequential system,(ii) poor germination of chickpea due to severedrought in late September and October, and(iii) unusually high prices of pigeonpea at har-vest time. In 1975-1976, the maize plus sequen-tial chickpea was a good combination with maizeand chickpea yields being about 50 and 100percent higher, respectively, than in 1976-1977.

On the medium deep Vertisols, average yieldsand gross benefits in the bed-and-furrow water-sheds at the 0.6-percent slope (BW7 B, C) wereconsistently better than those obtained under flatplanting. However, the magnitude of the differ-ences is less than on the deep Vertisols. Yieldsand gross benefits in BW7 D (beds at a 1.0%slope) were lower than those obtained with bedsat 0.6-percent slope or flat planting. Yield valuesand gross benefits on the medium deep Vertisols

13 Although the sorghum-based systems compare relativelyfavorably in 1976, this system was unsuccessful in the 3 previous years because of moist conditions at harvest time.

190

191

Table 71. Grain yields and rupee values of several farm ing systems on deep Vertisok at ICRISAT Center,1976-1977.

Intercroppinga Sequential croppinga

WatershedNo.

Plantingmethod

Tech.used Slope Maize

Pigeon-pea

Grossvalue Maize

Chick-pea

Grossvalue

A. Double-crop systems(%) (kg/ha) (kg/ha) (Rs/ha) (kg/ha) (kg/ha) (Rs/ha)

BVV1 Bed Impr 0.6 3360 720 4870 3 310 600 3840BW2 Bed Impr 0.6 2910 900 4960 3090 450 3410BW3A

Mean

Bed Impr 0.4 3 590

3290

670

760

4930

4920

2970

3120

760

600

3800

3680

BW3BBVV4B

Mean

FlatFlat

ImprImpr

30302790

2910

680560

620

44703940

4210

23502930

2640

440280

360

27503000

2870

BW7A b Bed Impr 0.6 2460 590 3730 2270 660 3030

B. Sorghum/pigeonpea intercropping and sole sorghum + ratoon cropa

Pigeon- Gross GrossSorghum pea value Sorghum Ratoon value

(kg/ha) (kg/ha) (Rs/ha) (kg/ha) (kg/ha) (Rs/ha)

BVV4A Flat Impr 2580 670 4130 2720 700 3080BW6A Flat Trad 540 390 1920 410 - 660

C. Single-crop systems (postrainy season only)a

Sorghum Chickpea Gross value

(kg/ha) (kg/ha) (Rs/ha)

BW4CN Flat Impr 500 490 1230BW6BN Flat Impr

Mean

1480 - 1330

990 1280

BW4CS Flat Trad 600 - 660BW6BS Flat Trad 1130 - 1240

Mean 870 950

aVarieties used: Improved -maize intercrop, SB-23; pigeonpea, ICP-1; sole maize, Deccan Hybrid 101; chickpea, local; sorghumCSH-6; Traditional -local.

b The maize yields in BW7 A were only 74 percent of the average of the other deep Vertisol watersheds (BW1,2,3 A). Gross returnsfor the two cropping systems on BW7 A were about Rs 1190 and Rs 650/ha less than the average of BW1,2, and 3 A. This is dueto the presence of some areas of medium deep soils and a poor growth area in the center caused by years of inadequate drainage.However, crop growth has substantially improved since 197S and it is expected that BW7 A will behave similarly within a fewyears. Nearby areas in STI showed poor growth in 1976 due to boron toxicity. This and other causes are being investigated.

Table 72. Grain yields and rupee values of several fa rming systems on medium deep Vertisols at ICRISATCenter, 1976-1977.

Intercroppinga Sequential croppinga

Ws Planting Tech.No. method used Slope Maize

Pigeon-pea

Grossvalue Maize

Chick-pea

Grossvalue

(%) (kg/ha) (kg/ha) (Rs/ha) (kg/ha) (kg/ha) (Rs/ha)

A. Double-crop systems

BW7D Bed ImprBW7B Bed Impr

1.00.6

21702060

390620

29403470

22102460

130410

22703250

BW7C Bed Impr

Mean

0.6 2740

2400

730

675

4350

3910

2150

2305

500

455

3220

3235

BVV6C Flat Impr Contour 2470 410 3260 1850 390 2660BW6D Flat Impr

Mean

Contour 2040

2255

690

550

3640

3450

1760

1810

310

350

2 370

2520

B. Sorghum/pigeonpea intercropping and sole sorghuma

SorghumPigeon-

peaGrossvalue

BW8A Flat TradBW8A Flat Trad

(kg/ha)

640630

(kg/ha)

440

(Rs/ha)

22101010

aVarieties used: Improved - Maize intercrop, SB-23; Pigeonpea, ICP-1; Sole maize, Deccan Hybrid; Chickpea, local; Sorghum,CSH-6. Traditional -local.

were substantially less than on the deep Vertisols.This is believed to be due to greater moisturestress during the severe September drought andinadequate moisture for the postrainy seasoncrops. Yields and benefits from bedded culti-vation on both deep and medium deep Vertisolsare likely to improve since the beds are left inplace and soil structure in the broad beds isexpected to improve due to reduced soilcompaction.

The sorghum yield from rainy season croppingon BW8 A (traditional technology) was 630 kgha (Table 72 B). It is interesting to note thatsorghum and the addition of the pigeonpeaintercrop more than doubled the gross monetary

value. This was true even in a season when therainy season rains stopped one month earlierthan normal, which caused some moisture stressfor the late-maturing pigeonpea.

Crop Production on Alfisols

Grain yields and monetary values for the threecropping systems used on the Alfisols (RW1 and2) are given in Table 73. RW1 E, which wascontour bunded and flat planted along thecontour, consistently produced the lowest yieldof the three cropping systems. The reduced yieldsappeared to be due to erosion and sedimentationbetween the bunds, which caused reduced plant

192

-

stands and growth. The average gross value forthe three cropping systems in the broad bed-and-furrow system treatment (RW1 D) was Rs 3 400,which is 39 percent greater than that of RW1 E (contour bunded) and 9 percent greater thanR W l C which was flat planted along gradedbunds at 0.6-percent slope.

The pigeonpea yield on the Alfisols was gen-erally greater than that on the Vertisols in spite ofthe low September/October rainfall and therelatively lower moisture-holding capacity of theAlfisols compared to Vertisols. A possible reasonfor this is that the Setaria intercrop on theAlfisols matured earlier and therefore compe-tition was removed at an earlier date than withmaize (cv SB-23) and sorghum grown on theVertisols.

Rainfall-use Efficiency

In a year of less-than-average rainfall, it becomescritically important to utilize all available waterwith utmost efficiency to maintain optimumlevels and stability of crop production. Develop-ing a system for the best possible use of the totalseasonal rainfall by means appropriate for thefarmer of the SAT is the basic objective of theFarming Systems research program. Thus, effec-tive rainfall as well as "water-use efficiency" needto be optimized in order to arrive at a high andprofitable "rainfall-use efficiency" (RUE). 14 TheRUE's were calculated on the basis of cropproduction under alternative treatments in re-search watersheds (Tables 71, 72, 73). The esti-mated RUE values obtained from several farm-ing systems on selected watershed units are listedin Table 74.

Gross RUE's varied from Rs 69/cm for a maize/pigeonpea intercrop in BW2 to only Rs9/cm for sorghum grown with local technologyafter rainy season fallow on BW4 C. On deep

14"Water-use efficiency" is defined as the agricultural pro-duction (in yield or value) in relation to the actual crop-related evapotranspiration (in cm). "Rainfall-use ef-ficiency" is the product of effective rainfall- and water-useefficiency.

193

Table 74. Rainfall-use efficiencies obtained inalternative fanning systems on Ver-tisols and Alfisols at ICRISAT Center,1976-1977.

Water-shed Cropping Systems Value of GrossNo. and water usea,b produce RUEC

(Rs/ha) (Rs/cm)

Deep Vertisols:

BW2 Maize/Pigeonpea 4960 69Maize/Chickpea 3410 47

BW3 A Maize/Pigeonpea 4930 68Maize/Chickpea NS 3800 53Maize/Chickpea S 4511 63

BW4 A Sorghum/Pigeonpea 4130 57Sorghum/Ratoon 3080 43

BW4C Sorghum Local 660 9Sorghum Improved 1330 18

BW6B Sorghum Local 1240 17Sorghum Improved 1330 18

Medium Deep Vertisols:

BW7B Maize/Pigeonpea 3470 48Maize/Chickpea NS 3250 45Maize/Chickpea S 3850 53

BW8 A Sorghum/Pigeonpea 2210 30Sorghum 1010 14

Alfisols:

RW2 B Setaria/Pigeonpea 2640 40Sorghum and RatoonNS 3160 49Sorghum and RatoonS 3520 54Groundnut/Safflower 3180 49

a Total rainfall during the growing season (Jun to Feb) was720 and 650 mm on BW1-8 and RW2, respectively.

bThe symbol NS when added to a cropping system indicatesthat crops were grown rainfed and on residual soil moisture;the symbol S stands for the application of supplementalwater.

cGross RUE is the rainfall-use efficiency calculated on thebasis of the total vaiue of the products grown in a croppingsystem.

Vertisols, the RUE values obtained on water-sheds in broad beds were two to three times thoseobtained with traditional rainy season fallowwith local and improved varieties. RUE's onmedium deep Vertisols and Alfisols were sub-stantially lower than those obtained in deepVertisols. On both deep and medium deep Ver-tisols (at the 0.6 % slope) the RUE's obtained forintercrops and for the combined sequential cropswere higher on broad beds than on flat-plantedsoils. There are clear indications that in years ofearly rainy season termination, the use of sup-plemental water to irrigate a second crop atplanting time may be critical to substantiallyincrease RUE on at least part of the land and tomake it feasible to grow a sequential crop. OnAlfisols, the RUE's on watersheds in broad bedsgenerally exceeded those on flat-planted areas.

The range of RUE values clearly shows thepotentials of improved resource use in a year ofrelatively low rainfall in the late rainy season. It isof course realized that RUE values derived fromgross returns, although reflecting the actualquantity of production, only partially reflect thedifferences between alternative systems of farm-ing because of differentials between the costsincurred in different systems. However, sometentative calculations show that when input costsare taken into account, the differences in RUEvalues between improved and traditional systemsbecome even greater.

Impl icat ions and Future Plans

Experience in watershed-based research over thepast few years and agroclimatic analyses showthat in many years sufficient moisture is availableon deep Vertisols for intercropping or sequentialcropping consisting of a short-duration rainyseason crop followed by a dry season crop; theearly termination of the rains the past year hasbeen an exception. However, harvesting, drying,and threshing of crops during the latter part ofthe rainy season and the susceptibility of mostpresent sorghum and millet genotypes to grainmold and weathering pose a serious problem (asa temporary solution, maize is grown on many

194

Vertisol watersheds). Thus, to make intercrop-ping and sequential cropping viable with speciesnow grown widely in the SAT three options mustbe simultaneously investigated: (i) generation ofshort-duration genotypes which better withstandhumid weather conditions at maturity and whicheither do not ratoon (to facilitate planting a sequential crop with minimum tillage) or willproduce a viable ratoon crop; (ii) the develop-ment of harvesting, drying, and threshing tech-niques suitable for the rainy season, and/or(iii) the generation of determinate and high-yielding genotypes of 130 to 150 days durationand concurrently the development of supplemen-tal irrigation availability for establishment of a sequential crop (or ratoon crop).

A large portion of the season's rainfall percol-ates through the root profile after the capacity ofthe soil to store moisture is exceeded; this processoften begins before half of the rainy season isover and is of the greatest magnitude on mediumdeep Vertisols and Alfisols. Even on deep Ver-tisols, however, frequently as much as 20 percentof the rainfall contributes to groundwater. Inmany areas in the SAT this component cannot beeasily recovered because of unsuitable aquiferconditions. Thus, it becomes important to in-vestigate techniques which permit a more ef-ficient use of available profile storage throughgreater extraction and/or which permit a largerportion of the rainfall to be collected in reservoirsduring the later part of the rainy season forsubsequent utilization.

During the past 4 years, it has been observedthat on Alfisols substantial quantities of waterescape early in the season when the profile is notyet fully charged, partly due to surface sealing.Means to increase infiltration of larger quantitiesof rainfall into the soil during the preplantingand establishment period must be sought.

When the Vertisol profile is fully charged at thebeginning of the postrainy season, the oppor-tunity for substantial returns from supplementalwater is reduced. Saving the available water untilthe summer season usually means large evap-oration and seepage losses. Therefore, longer-duration rainy season crops of high yield poten-tial must be explored on deep Vertisols so that

collected runoff can be used for profile rechargeand second crop establishment. Ratooning ofpigeonpea or postrainy sorghum are other possi-bilities which are being explored for effective useof stored water.

High seepage losses, particularly on mediumdeep Vertisols with gravelly subsoils, seriouslylimit the potential for supplemental irrigationfrom collected runoff. Although in some situ-ations eroded sediment reduces losses while inothers compacted salt/soil mixtures are effective,additional research on cheap but highly im-permeable sealing materials is urgently needed.

Precision planting of seed at the exact finalspacing required presents problems in manysituations common in the SAT. The nonavailab-ility in some areas of seed graded according tosize, along with problems with germination,frequently results in plant populations of lessthan optimum. Gap filling, even if done early,does not result in satisfactory growth of late-planted material. It will therefore, often benecessary to overseed crops and later thin to therequired plant density. Thus, there is a need foradjustable continuous-flow planters operatingindependently of seed size. There is also anurgent need to develop planting equipmentwhich can be satisfactorily used over a widerange of soil moisture conditions.

Resource utilization research on natural wa-tershed units of an operational size wil l becontinued and intensified during the next fewyears. Five additional catchment units are beingdeveloped on Alfisols; efforts to develop ap-proaches which substantially increase rainfall-use efficiency on these soils as well as on themedium deep Vertisols have so far had onlylimited success. Increased attention will be givento a better understanding of the movement ofwater in the soil profile-plant-atmosphere systemoperating under alternative resource-man-agement and cropping systems. Ratooning ofpigeonpea with or without the use of sup-plemental water, which appears promising (see"Steps in Improved Technology", page 171) willbe applied on an operational scale. The use andconservation of collected runoff water, parti-cularly where seepage rates are high, wil l be

195

further investigated, Management systems re-sulting in greater runoff without significant ero-sion wil l be explored on soils with limited profilestorage capacity. Research wil l be oriented to-wards operational problems related to harvest-ing, drying, and threshing of rainy season crops.

Research on alternative resource-managementand cropping systems on an operational scale toincrease and stabilize production in rainfed ag-riculture has been initiated at two RegionalCenters by the All-India Coordinated ResearchProject for Dryland Agriculture. Possibilities forsuch cooperative research are being consideredat other Centers. Data on improved resourcemanagement collected at several locations char-acterized by different agroclimatic conditionswill provide a better base for extrapolation ofresearch results and for early implementation ofimproved farming systems. Possibilities arebeing explored to test improved farming systemsunder real-world conditions in pilot projects todetermine their social implications.

C o o p e r a t i v e Research

ICRISAT is one of the research centers wherenew concepts, approaches, and methodologies toarrive at improved farming systems will begenerated. However, before application in thediverse regions of the SAT, these principles haveto be integrated into viable packages or systemsand adapted info applicable site-specific tech-nology through cooperative research.

Simulation techniques are used to quantifyand predict the hydrologic behavior of agrocli-matic environments under alternative resource-management technologies. Also, basic infor-mation on crops and cropping systems, de-veloped in various national programs, is used tomatch improved systems to the resources of a region. These studies wil l greatly reduce thenumber of research alternatives to evolve viablefarming systems. For effective simulation stud-ies, a limited number of bench-mark locationsmust be identified. The diversity encountered in

the SAT (climate, soil, people, livestock, capital,etc.), dictates the need for associated researchefforts at such locations. Preliminary investi-gations are under way to identify bench-marklocations in Africa.

During the past year, discussions with Indianresearch programs15 have resulted in twocooperative research proposals which have beeninitiated at several locations across the IndianSAT.

Hydrologic studies to improve land and waterutilization in small agricultural watersheds

The objective of this project is to derive region-specific design criteria for improved resourcemanagement which will more effectively con-serve and utilize the rainfall and the soil andwhich, when integrated with new crop pro-duction systems, increase productivity and as-sure dependable harvests. The project has beenstarted at Bangalore and Hyderabad.

Research on resource development, conservation,and utilization in rainfed areas

The objectives of this project are to develop a research program for testing the bed-and-furrowsystem of cultivation and its modifications underseveral agroclimatic conditions and also to quan-tify the production effects of presently acceptedsoil and water conservation practices. The pro-posed research locations are Akola, Bangalore,Bellary, Hyderabad, ICRISAT Center, Indore,Ranchi, and Sholapur.

15The All-India Coordinated Research Project for DrylandAgriculture and the Central Soil and Water ConservationResearch and Training Institute.

196

References

Cocheme, J., and P. Franquin 1967. A study ofthe semi-arid area South of the Sahara in WestAfrica. FAO/UNESCO/WMO/Inter AgencyProject, pp 117-129.

Dungan, G. H., U. S. Sisodia, and G. D. Singh.1955. The benefit of sowing maize for fodder inNorth and South lines. Allahabad farmer (In-dia) 29:8-13.

Hargreaves, G. H. 1971. Precipitation, dependa-bility and potential for agricultural produc-tion in North East Brazil. EMBRAPA andUtah State Univ Publ No 74-D159.

Hargreaves, G. H. 1975. Water requirementsmanual for irrigated crops and rainfed agricul-ture. EMBRAPA and Utah State Univ PublNo 75-D158.

India Meteorological Department. Climatologi-cal tables of observatories in India (1931 — I960). Poona.

Krantz, B. A., A. J. Ohrogge, and G. D. Scar-seth. 1944. Movement of nitrogen in soils. SoilSci Soc Amer Proc 8:189-195.

Krantz, B. A., and W. V. Chandler. 1954.Fertilize Corn for Higher Yield. Bull 376, N Car Agr Exp Sta (foreword by R. W.Cummings).

Krantz, B. A., et al. The Agronomy of DwarfWheats. Indian Council Agr Res, New Delhi(foreword by M. S. Swaminathan).

Krishnan, A. 1974. Some climatological featuresof the semi-arid tropical regions of the world.(Proceedings) International Workshop onFarming Systems, ICRISAT, pp 53-91.

Raman, C. R. V., and B. Srinivasamurthy. 1971.Water Availability Periods for Crop Planning.Sci Rep No 173, India Meteorol Dept, Poona.

Rao, K. N., C. J. George, and K. S. Ramasastri.1971a. Climatic Classification of India. SciRep No 158, India Meteorol Dept, Poona.1971b. Potential Evaporation Over India. SciRep No 136, India Meteorol Dept, Poona.

Robertson, G. W. 1976. Dry and wet spells.UNDP/FAO, Tun Razak Agr Res Cen.Sungh: Tekam, Malaysia, Project Field Rep,Agrometeorology; A-6.

Santhisegaram, K. 1962. Thesis. Waite Inst Unitof Adelaide, Adelaide, Australia.

Thornthwaite, C. W. 1948. An approach to-wards rational classification of climate. Geogr Rev 38:55-7 38:55-64.

Trol l , C. 1965. Seasonal climate of the earth. InRodenwaldt, E. and Jusatz, H. (Eds), WorldMaps of Climatology, Berlin, Springerverlag,p28.

Wetselaar, R. 1961. Plant and Soil 15:110-120.

197

E c o n o m i c s

P r o g r a m

The Economics Program was able to essentiallycomplete eight major research studies during the1976-1977 year at ICRISAT. A l l were aimed athelping to identify the major socioeconomic andother constraints to agricultural development inthe semi-arid tropics and to evaluate alternativetechnological and/or institutional measures re-quired to alleviate these constraints. The mainfindings from the studies are discussed under thetwo subprogram headings of Production Eco-nomics and Marketing Economics.

Product ion Economics

Resource Base and Cropping Patterns

Results from the Indian village-level studies -which have been under way in six villages inthree agroclimatic zones since May 1975-indicate large regional differences in croppingpatterns. In the Sholapur District of Mah-arashtra, which generally has deep Vertisols,farmers plant nearly 60 percent of their crops inthe postrainy season following a rainy seasonfallow (Fig 87). In the medium deep Vertisol areaof Akola District, also in Maharashtra, almostall crops are sown in the rainy season. In theAlfisol villages of Mahbubnagar District,Andhra Pradesh, around 85 percent of thecropped area is sown in the rainy season. InShirapur village of Sholapur District, which hasvery deep Vertisols, small farmers were found tokeep a much higher proportion (78 %) of theirlands fallowed during the rainy season comparedto large fanners (55 %). If this is representative ofthe estimated 18 million hectares of land notcropped during the rainy season in SAT India,then technologies which enable crops to begrown in the rainy as well as the post-rainy season may contribute a proportionately

larger impact on the small farmers, as well assubstantially increasing food production.

Except in villages with little irrigation orwithout the very deep Vertisols which storemoisture well for relatively assured postrainyseason cropping, small farmers sow a higherproportion of their land to intercrop mixturesthan do large farmers (Fig 88). The latter have a higher proportion of sole crops, but interestinglyenough they grow many more different species ofcrops than do the smaller farmers. One might betempted to conclude that intercropping researchwould hence tend to offer proportionately morebenefits for smaller farmers. However, it wasfound that virtually all of the high-yieldingvarieties (HYVs) in the six villages were sown assole crops. To the extent then that new in-tercropping technology will embrace HYVs, it isnot clear that smaller farmers will benefit more.The question remains as to why the HYVs arenot intercropped, whereas the local cultivarsgenerally are. Perhaps the HYVs are less riskythan the locals and/or they were never evolved orrecommended under an intercropping situation.More research is under way to test thesehypotheses.

In the villages with more irrigation (Dokur)and/or very deep Vertisols (Shirapur), 60 to 100percent of the land is planted to sole crops. InAurepalle, where drought-resistant castor is ex-tensively grown, the proportion of sole croppingis also relatively high -more than 40 percent. Itseems from this evidence that the less-assured (interms of rainfall) areas could benefit more fromintercropping research. It also suggests that ifimprovements in soil- and water-managementtechnologies enhance the resource base andmake the environment more stable, one may seemore sole cropping practiced. This is becauseintercropping seems to be primarily practiced asa risk-reducing procedure.

In Kalman Village, as many as 60 differentintercrop combinations were grown and 10 to 20were not uncommon in the other villages (Fig89). This increased complexity in cropping sys-tems seemed to be associated with a moreheterogeneous natural resource base. These haveevolved largely through farmers' informal trial-

201

Figure 86. Location of the six villages in SAT peninsular India selected for Village Level Studies.

and-error experimentation and are aimed atsatisfying their various needs. Under these cir-cumstances, intercropping research would seemto have its greatest potential contribution if itsubstantially embraces new technological ele-ments such as HYVs and land- and water-management techniques new to the farmers.

Weed-control Economics

In a study of the economics of weed controlpractices in six SAT areas of India, it was foundthe most-frequent and -timely operations occur-red in the relatively assured rainfall zone ofAkola where cotton dominates the cropping

202

MAHBUBNAGAR DISTRICTANDHRA PRADESH

AKOLA DISTRICTMAHARASHTRA

SHOLAPUR DISTRICTMAHARASHTRA

PENINSULARINDIA

Aurepalle Dokur

Mahbubnagar D i s t r i c t

[ A l f i s o l s , uncertain

ra in fa l l (71 cm)]

Shirapur Kalman

Sholapur D i s t r i c t

[Deep and medium-depth

Ver t i so l s , uncertain

rainfall (69 cm)]

Kinkheda Kanzara

Akola D i s t r i c t

[Medium-depth Vert isols,

stable rainfal l (82 cm)]

Figure 87. Proportion of postrainy season cropping to gross cropped area in six SAT villages of India, 1975-1976.

system on Vertisols (Fig. 90). Here more than 90percent of intercrop sorghum fields are interculti-vated two or more times. More than 60 percentof the intercultivation operations are initiatedwithin 15 days after sowing. Other crops inAkola follow a similar pattern.

In the Alfisol area of Mahbubnagar, where therainfall pattern is relatively less assured, weedgrowth is somewhat less vigorous and rapid.Intercultivation operations are less frequent andare generally initiated 26 or more days aftersowing. Little hand weeding is carried out onrainfed crops, and most weed control is given toirrigated or cash crops.

In the deep Vertisols of Sholapur Districtwhere postrainy season sorghum is a prevalentcrop, little intercultivation and no hand weedingoccurs. The single intercultivation performed bymost farmers is very much delayed (more than 6

weeks after sowing) and its purpose is to closecracks in the soil for moisture conservationrather than for control of weeds, which appearmuch less of a problem in postrainy season crops.Several harrowings are performed on fallowlands during the rainy season; these substitutefor intercultivation in the postrainy season crop.

The conclusion from the examination of pre-sent weed-control practices in the six villages isthat, in general, farmers allocate weed-controleffort rationally across regions and crops. Thehigher the crop value and the more vigorous theearly weed growth, the more intensive and timelyweed control becomes. This indicates that underimproved farming systems which embraceHYVs, fertilizers, improved implements, andland and water management, farmers would tendto increase weed-control efforts.

An evaluation was made of the estimated

203

Smal l farm

Medium farm

Large farm

80

70

60

50

40

30

20

10

0

Aurepalle Dokur

Mahbubnagar D i s t r i c t

Shirapur Kalman

Sholapur D i s t r i c t

Kinkheda Kanzara

Ako la D i s t r i c t

Figure 88. Proportion of gross cropped area devoted to intercropping in six SAT villages of India, 1975-1976.

Figure 89. Crop combinations used in intercrop-ping and crops used in sole cropping in six SAT villages of India, 1975-1976.

relative costs of three alternative weed-controlpractices under conditions existing in the Akolaregion of Maharashtra. The three plans weredesigned to achieve about the same level of weedcontrol using (i) human and animal power only,(ii) partial herbicide use plus human labor, and(iii) herbicides only. The cost of the pure her-bicide treatment was so high that it could neverbe feasible at existing wage rates. As Figure 91shows, for food-grain crops under SAT Indianconditions, even partial herbicide use combinedwith animal and human methods was two to fourtimes more costly than the traditional animal andhuman methods. The basic reasons for this arethe low wage rates for human and bullock powerin India. For small farmers using their ownfamily labor in which the opportunity cost isgenerally lower than ruling wage rates, there iseven less incentive to use herbicides. In SATAfrica, where real wage rates may be two to fourtimes those in SAT India, herbicides may be a more viable proposition.

204

Mahbubnagar Sholapur AkolaDistr ict District Distr ict

Crop combinationsin intercropping

Sole crops60

50

40

30

20

10

0

Small farms

Medium farms

Large farms

100

90

80

70

60

50

40

30

20

10

0

Figure 90. Frequency and timing of weed-control practices on selected rainfed crops in three SAT districts of India, 1975-1976.

205

Dis t r ibu t ion

Number of operat ions

0 1 2 3 4

Postra iny season sorghum on Ve r t i so l s , Sholapur D i s t r i c t

70

60

50

40

30

20

10

0

Sorghum mixture on A l f i s o l s , Mahbubnagar D i s t r i c t

70

60

50

40

30

20

10

0

Sorghum mixture on V e r t i s o l s , A kola D i s t r i c t

60

50

40

30

20

10

0

0-15 16-25 26-35 36-45 46+

Delay from sowing (days)

D is t r ibu t ion

Figure 91. Comparison of costs of alternative weed-control methods in Akola District, Maharashtra, India, 1975-1976.

Unless herbicides make possible substantialyield increments not achievable by traditionalmethods (such as a preemergence herbicide),then widespread use of herbicides will be out ofthe question for a long time to come in SATIndia. Wage rates would have to rise by morethan SO percent to make the partial herbicideplan attractive in pearl millet. For other crops,the wage rise would have to be even more. Realwage rates in SAT India have been stagnant forthe past decade or so. Hence such increases areunlikely in the near future.

Herbicide research and technology in SATIndia could adversely affect income-earning op-portunities of one of the most disadvantaged ofall rural socioeconomic groups, namely femalelaborers, and in particular hired females fromlandless and small farm households. From 70 to90 percent of the hand weeding is performed byhired female labor (Table 75) whose wage ratesare generally about half those of hired males. Thelabor displacement that herbicides could cause inSAT India could be substantial, particularly incrops like cotton and sorghum in the assured

rainfall areas. The African SAT situation, withits higher wage rates and opportunity costs, maybe quite different. Landless laborers are rare inmany African areas, and animal power forintercultivation is often not available. Weed-control problems there seem substantial, andwould justify a greater allocation of ICRISAT'sweed-control research resources than is the casefor India.

Group Action and Organization

A review of the extensive economic and anthro-pological literature on group action and organi-zation, together with an examination of variouscase studies, revealed that several requirementsmust be met if group action is to be successful.These requirements are especially relevant in thecontext of developing and implementing im-proved land- and water-resource-utilizatioatechnologies, where environmental or planningunits are often much larger than individualholdings. These conditions also seem helpful indistinguishing the types of technology designs

206

Intercultivation with bullocks and hired labor

Intercultivation with bullocks and family labor

Partial herbicide and hired labor

150

100

50

0

Groundnut Sorghum Pearl millet Chickpea Mixed crop

Table 75. Sex distribution of hand-weeding activity on sample farms in three districts in SAT India, cropyear 1975-1976.

Akola District,Maharashtra

MahbubnagarDistrict,

Andhra Pradesh

SholapurDistrict,

Maharashtra

Cotton Sorghum Groundnut Sorghum

Handweeding by females (%)Handweeding by hired

females (%)Days of weeding wage

employment per samplefarma

98 9691 77

78 24

10077

5

10070

11

a Based on 8-hour days.

which require a minimum of outside adminis-trative input and control. Such attributes seemdesirable if viable adoption of watershed-basedsystems is to be ensured.

The conditions necessary for successful groupaction appear to be:

(i) the resource or activity around which thegroup forms must be both collective andorganizational in nature - that is, action ofall members of the group is required forits creation and the activity is not availableunless potential beneficiaries organize them-selves to procure or create it;

(ii) the basic profit from the group activity mustat least equal the rate of return required tomotivate private investment, plus a pre-mium or compensatory profit to cover thetransactions cost of organizing people andof having them forego some of their in-dividual decision-making powers;

(iii) there must be a functional identity in thecollective good which participants in thegroup can recognize; that is, the benefitsshould be distributed so that no one mem-ber or group can coerce others to ensure thathe or they gain a disproportionate share -

or a benefit qualitatively different from thatderived by the others; and

(iv) group size must be appropriate to ensure itslongevity for the tasks at hand. Smallgroups (5 to 15 farmers) generally appearbest for tasks of limited duration, whilelarger groups (up to 90 to 100 farmers) aresuited to longer-term purposes. The smallerthe group, the greater are the chances offactionalism and eventual breakdown of thegroup; the larger the group, the greater arethe transactions costs of organizing them toagree. Optimal size is dependent on thetechnological, economic, and political en-vironment in each situation.

From detailed analyses of case studies ofgroup action - successful and unsuccessful - inIndia, it seems small groups of farmers could beviable for one-time tasks like land shaping andredistribution of plots such as may be required toimplement ICRISAT's watershed-based tech-nologies. Supplementary irrigation out of earth-ern runoff-collection tanks may require largergroups. However, more research is required totest these hypotheses and to determine whetherthe above conditions are sufficient, in addition tobeing necessary, for successful group action.

207

Credit and Risk

Data from ICRISAT's village-level studies andstudies by others conducted during droughtperiods in arid and SAT India were analyzed todetermine the degree of success farmers have inmanaging their assets and consumption overdrought cycles. The studies include data fromtwo areas of Rajasthan, one in Gujarat, and twoin Maharashtra for the drought years 1963-1964, 1969-1970, and 1972-1973.

It was found that the largest reduction inconsumption during droughts occurred withceremonial expenses (40 to 60 %), followed byexpenditure on "protective" foods like milk,sugar, fruits, vegetables, and fleshy foods, etc.(30 to 40%). In all except one area, expendi-

ture on education, medicines, etc. fell duringdroughts by about 40 percent. Despite a rela-tively unchanged expenditure on food grains, theconsumption of food grains fell by 10 to 25percent during the drought years due to higherfood-grain prices.

The value of "productive" assets, consistingprimarily of livestock and implements, fell by 15to 40 percent in drought compared to predroughtyears (Fig 92). The decline in "nonproductive"assets like jewelry and consumer durables isgenerally less marked. The "productive." assetsare not recouped fully in postdrought years.With the frequency of droughts which occur inthese areas, assets may never be recovered todesirable levels before another cycle of assetdepletion-replenishment is initiated.

Jodhpur Barmer Banas Kantha Sholapur Aurangabada

(Rajasthan State) (Gujarat State) (Maharashtra State)

a Postdrought data not available.

Figure 92. Changes in household productive and nonproductive assets, because of drought, in five districts

of India.

208

Predrought year Drought year Postdrought year

100

80

60

40

20

0

100

80

60

40

20

0

Jodhpur Barmer Banas Kantha Sholapur Aurangabad

Compared to predrought years, indebtednessincreased by 160 to 320 percent in drought years.In three of the four areas (data for Aurangabadwas not available), indebtedness increased fur-ther after droughts. However, borrowing con-tributed only between 5 and 15 percent of thesustenance requirements during drought periods(Fig 93). The importance of government reliefprojects in the sustenance of farmers in drought-prone areas is clearly demonstrated; governmentrelief work provided 25 to 55 percent of suste-nance in the various areas. This was followed bymigration and remittances from outside (12 to27 %) and sale of assets (12 to 26 %).

Of the small role played by credit in droughtsustenance, it was found that less than 15 percentcomes from institutional sources in Rajasthan,whereas the bulk (65 to 80 %) comes from tradersand moneylenders. In Gujarat, institutionalsources provided about 40 percent of droughtcredit, traders 30 percent, and relatives most ofthe remainder. In the Maharashtra villages, most(55%) of the drought credit came from in-

stitutions, then from large farmers (25 %), andfrom relatives (20%). In postdrought years inRajasthan, the share of credit from moneylenders, large farmers, and relatives increased,while that from traders decreased. In Gujaratand Maharashtra, institutional credit substan-tially increased in importance after the droughtsand that of traders and large farmers decreasedduring this period. In all areas, the proportion ofcredit in kind rather than in cash increasedsubstantially (from 1:1 to 4:1) during droughts.

This evidence lends support to the thesis thatduring drought periods credit is primarily usedfor consumption, while in postdrought periods itis used for crop production and asset replenish-ment. Cooperatives and banks traditionally donot lend for consumption purposes, and thisexplains their relatively minor role during theactual drought. Although interest rates can riseby more than 50 percent in drought years (partlydue to added risks of consumption loans), pri-vate lending agencies seem to offer the requiredflexibility in terms of repayment which more

Figure 93. Sources of sustenance income during drought years in selected households in five districts of India.

209

Government re l ie f work, e tc .

Sale of assets

Borrowing

Household product ion/storage

Other: migrat ion, g i f t s , e tc .60

50

40

30

20

10

0

(Rajasthan State) (Gujarat State) (Maharashtra State)

Jodhpur Barmer Banas Kantha Aurangabad Sholapur

nearly suit the requirements of borrowers duringsuch times. Links are established between factorand product markets to enable repayments to bemade using a variety of vehicles-such as ex-change labor, grain, etc. Formal credit agenciescould profit by providing more flexibility inrepayment terms through integration of creditservices with factor markets more explicitly.

Effect of "Green Revolut ion" in Whea t onPulse and Nutr ient Production in India

High-yielding varieties of wheat were introducedinto India in the mid-1960's, and their effect onthe production of pulses and major nutrients hasbeen evaluated by ICR1SAT. Linear trend lineswere fitted to data from the six major wheat-growing states of India for the 10-year periodpreceding 1964-1965, and separately for thesubsequent 10-year period. The states were Pun-jab, Haryana, Uttar Pradesh, Bihar, Rajasthan,and Madhya Pradesh. It is apparent that 22percent of the expansion in wheat hectaragewhich occurred in the latter period could beaccounted for by reductions in the area sown topulses, particularly chickpeas. Eight percent ofthe expansion in wheat came at the expense ofwinter rice and barley. The vast majority of thegrowth in area sown to wheat in the six states wasdue to increases in cropping intensities resultingfrom the HYVs and the expansion in irrigation,and an increase in net sown areas.

Total annual trend food-grain production in1974-1975 in the six states would have been 13.4percent less, had HYVs of wheat not beenintroduced. This takes into account the re-duction in production of pulses, winter rice, andbarley, as well as the increased wheat production.The following percentage reductions would haveoccurred in total annual 1974-1975 productionof various nutrients: protein, 10.0; energy, 13.5;methionine and cystine, 15.5; tryptophan, 11.3;leucine, 5.9; and isoieucine, 2.0.

Per-caput production of food grains and theabove nutrients would also have been less hadHYVs of wheat not been introduced. Generally,the per-caput reductions would amount to be-tween 1 to 2 percentage points less than the above

figures. Total lysine production would have beenabout 6.4 percent more had HYVs not beenintroduced. Calories, rather than protein oramino acids, seem to be the first limiting nutrientin the diets of most Indians (including those inthe least affluent socioeconomic categories). Vita-min A and vitamin B complex, calcium, copper,iron, and zinc follow. The small reduction inlysine is hence an acceptable price to pay for thesubstantial increases in production of foodgrains, energy, and protein resulting from use ofthe HYVs. When the record production year of1975-1976 is compared with projected pre-greenrevolution trends, the production of all nutrients,including lysine, shows a substantial increaseover what would have occurred had pre-greenrevolution conditions continued. The percentageincreases in 1975-1976 nutrient production as a result of the green revolution are estimated to beas follows: protein, 20.1; energy, 21.9; lysine, 6.9;methionine and cystine, 20.9; tryptophan, 32.5;leucine, 16.2; and isoieucine, 12.1.

Even though the prices of chickpea relative towheat have risen since 1964-1965, the real pricesof protein and energy from both sources hassubstantially fallen. For chickpea, the real pricesof protein and energy have fallen more than 20percent between 1964-1965 and 1974-1975. Forwheat the reduction has been more than 33percent. Every hectare of 1974-1975 wheatwhich substitutes for a hectare of chickpea alsoadds a further 55 kg of protein and 2527000kilocalories of energy to what the chickpeawould have produced.

Hence, it is clear that the net nutritionalimpact of the new HYVs of wheat introduced inthe mid-1960's in India was both positive andsubstantial. The success of these wheats clearlyillustrates how a plant-breeding strategy whichemphasizes increased yield potential can result insignificant improvements in aggregate nutri-tional well-being.

Breeding strategies which emphasize yield andyield stability offer the best prospects for improv-ing nutritional well-being of the least nu-tritionally and economically affluent groups inthe SAT. Increased yields and production offood grains has a direct impact on prices and real

210

incomes of the least-affluent groups, who spend a large amount of their incomes on food grains.Increased real income will enable them to pur-chase additional food grains and hence improvetheir nutrition. If yield potentials of cereals,pulses, oilseeds, and other crops could all besubstantially increased, nutritional improve-ments will follow as these all have complemen-tary nutritional compositions. It is not necessarythat any one of the grains have a "balancedcontent" of all nutrients — which is the underly-ing premise of cryptic quality breeding pro-grams. People eat more than one food item eachday. Making available additional quantitiesof all foods is the appropriate nu-tritional strategy. Nutrit ion education can helpensure that the correct mix is consumed. It isunreasonable to expect any one grain to providea balanced diet.

Marke t i ng Economics

Market Surpluses of Food Grains

Access to markets provides a primary incentivefor farmers to adjust farm organization andcropping patterns to capitalize on the compara-tive regional advantages in production of variousenterprises. A study of the research literature andpublished data on market participation by In-dian food-grain producers showed that the bulkof production does not reach formal marketchannels. The proportion of production whichreaches the market varies across regions, crops,and years. Two factors seem to consistentlyinfluence household marketed surplus (produc-tion minus quantities retained for consumption).Generally the larger the farm production of foodgrains, the larger the amount of marketed sur-plus. On the other hand, as family size increases,the amount of farm food grain marketed tends tofall.

Table 76 shows that the official estimates ofsorghum arrivals in Indian assembly markets asa proportion of 1972-1973 production arearound 14 percent. For chickpea they are around26 percent. Increases in production in sub-

sequent years, together with relaxation of in-terstate and intrastate trade restrictions, leads usto believe market arrivals may have increased.ICRISAT's "educated guesstimate" of currentmarket arrivals for sorghum and millets isaround 15 to 20 percent, while for chickpea andpigeonpea it is between 35 and 40 percent.

Chickpea has shown a strong upward trend inarrivals at regulated markets since 1969-1970(Fig 94). Arrivals of rice and groundnuts havebeen relatively steady, while wheat, millets, andsorghum showed increases between 1969-1970and 1972-1973, followed by downward trends.

World Sorghum and Millet Trends

World demand for coarse grains in the year 2000is estimated by the World Food Council to bedouble or more the demand experienced in 1970.About one-fifth will be for direct human con-sumption and four-fifths will be fed to animals.In the LDCs (less-developed countries), demandfor coarse grains is expected to triple by the year

Table 76. Market arrivals (as a proportion ofproduction) of wheat, rice, sorghum,and chickpea from villages to assemblymarkets.

Year Wheat Rice Sorghum Chickpea

(%)(%)

1960-61 20 NAa NA NA1966-671967-68

2024

NANA

NANA

NANA

1968-69 29 22 13 27

1970-71 31 26 13 271972-73 31 24 14 261974-75 25b 24b

1b 35b

1976-77 30b 25b 17b 37b

Source: Indian Agriculture in Brief, 9th, 10th, 12th, 13th,and 14th editions. Directorate of Economics andStatistics, Ministry of Agriculture, Government ofIndia.

a Data not available.bICRISAT estimates.

211

Figure 94. Index of market arrivals of major food grains in selected regulated markets.

2000. Some 60 percent will be for humanconsumption.

Comparing the projected annual compoundgrowth rates of demand for coarse grains, usingWorld Food Council data, with past annualtrend growth rates of sorghum and millet pro-duction from 1964 to 1974, a picture of excess-demand imbalance emerges for the LDCs (Table77). Demand for coarse grains in the LDCs isestimated to grow 3.23 percent per year while,historically, production of sorghum and millethas grown by only 2.13 percent. The past pro-duction growth rate in LDCs of the SAT hasbeen even less than this -2 .11 percent per year.The picture for the developed countries shows anexcess-supply position emerging, with an annualdemand growth of 2.14 percent and past pro-

duction increasing at 3.48 percent. Hence theprojected future overall world supply-demandbalance for coarse grains seems favorable. Iftrade flows permit the excess production in thedeveloped countries to move to the grain-deficient LDCs, then their position will beimproved.

About 80 percent of the world's sorghum area(Sorghum vulgare, S. Sudanese, and S. almum) islocated in LDCs of the SAT, but the productionfrom this area represents only 60 percent of theworld's total output. Since 1964 the area devotedto sorghum by these countries has increased by235 000 hectares per year, but the increase hasmainly been in Africa (south of the Sahara),Mexico, and South and Central America. In SATAsia, the sorghum area has declined and yield

212

1 9 6 9 - 70 7 0 - 7 1 71 -72 72 -73 7 3 - 7 4 7 4 - 7 5

Y e a r

Rice

M i l l e t s

S o r g h u m

Whea t

C h i c k p e a s

G r o u n d n u t s

150

140

130

120

110

100

90

80

70

6 0

0

Table 77. Projected demand and past trends in coarse- grain production.

Annual projected compoundincrease in demand for all

Annual trend compoundincrease in sorghum and

Region coarse grains millet production, 1964-1974b

(%) (%)

LDC (in SAT) market NAC 2.11economies

LDC (non-SAT) market NA 2.37economies

Total - LDC market 3.23 2.13economies

Total - Developedcountries market

2.14 3.48

economiesWorld 2.41 2.50

aSource: Aziz, Sartaj. "The World Food Situation Today and in the Year 2000". Paper presented at World Food Conference,1976, Iowa State University. (Dr. Aziz is Deputy Executive Director, World Food Council).

bCalculated by ICRISAT from data contained in various editions of the FAO Production Yearbook. cData not available.

trends have also been relatively stagnant, in-creasing (on a hectare basis) only 14 kg/year.This increase is much less than that in otherareas, where sorghum is used primarily foranimal feed.

World sorghum production increased from a trend value of 37.6 million metric tons in 1964 to51.3 million in 1974 (Fig 95), representing anannual trend increase of 1.37 million tons. Pro-duction has risen fastest (360%/year) in SATLDCs of South America, followed by Australia(a developed SAT country) at 223 percent peryear.

There has been little change in the world areaof millets (Pennisetum americanum, Eleusine coracana, Panicum miliaceum, Setaria italica, and Echinocloa crusgalli) since 1964. The 68million hectares of millets grown in the world isshared about equally between the People's Re-public of China (and other centrally plannedeconomies) and LDCs in the SAT. Approx-imately half of the world's millet area is devotedto the Pennisetums. Setaria occupies about 23percent, Panicum 14 percent, and Eleusines 8 percent.

Unfortunately, the two major millet-growingregions in the world have the lowest yields.Yields have been rising by 17 kg/ha annually inthe centrally planned economies, where Setaria isthe principal species. Yields in the SAT LDCs(where Pennisetum predominates) have beenrising only about 6 kg/ha each year.

In the countries with centrally planned econ-omies, total millet production has been increas-ing by about 600 000 tons per year since 1964(Fig 96). In the SAT LDC areas, millet pro-duction has been rising by around 200000 tonsper year. World millet production amounts toabout 47 million tons annually. It has beenincreasing by 800000 tons per year. Apparentlysome 85 percent of the world's millet is used forhuman consumption.

Sorghum and Millet Trends in India

On the basis of secondary data it was shown that,in India, up to a monthly per-caput income levelof around Rs 15, the per-caput consumption ofcoarse grains increases, reaching a peak of just

213

less than 7 kg/month. At higher income levels,per-caput consumption declines. About SO per-cent of India's population has a monthly incomeof less than Rs 15. Hence, population and incomegrowth in India wil l continue to place heavystrains on demand for coarse grains.

Growth in coarse-grain production in Indiahas not kept pace with demand during the last 20years despite introduction of HYVs. In 1971,

production was estimated to have fallen behinddemand by 1.7 million tons. If these trendscontinue, the shortfall could be around 4 milliontons by 198S. To meet this from local productionwill require an annual production growth of 3.2percent per annum. It seems if this increase is tobe achieved, yields per hectare must rise as landfor expanding the area under coarse grains isvery limited.

214

Figure 95. World sorghum production trends.

a Equat ions represent l inear trend l ines f i t t ed by ord inary leastsquares regress ion a n a l y s i s .

bFigures in parentheses are t -va lues .

( t )

Y e a r

1964 65 6 6 6 7 6 8 6 9 70 71 72 73 74

Count r ies w i th centra l lyp lanned economies

P= 1 5 6 . 7 5 - . 0 6 4 t ( - . 2 5 )

28000

26000

24000

22000

20000

18000

16000

14000

12000

1600

1500

1300

1100

900

700

500

300

100

200

180

160

140

120

Count r ies w i t h l ess -developed market econo-mies ( i n SAT)

P = 1 9 4 1 5 . 7 0 + 640.59 t ( 4 . 6 5 ) b

P = 15952 .30+565 .1 t ( 2 . 2 2 )

Count r ies w i t h developedmarket economies (otherthan SAT)

Count r ies w i t h l ess -developed market econo-mies (other than SAT)

P = 761.51 + 5 0 . 3 5 t ( 4 .55 )

P = - 6 4 . 1 8 + 1 1 5 . 8 2 t ( 5 . 3 6 )

Count r ies w i t h developedmarket economies ( in SAT)

Equat ions represent l inear trend l ines f i t t ed by ordinary leastsquares regress ion a n a l y s i s .

F igures in parentheses are tavalues.

Figure 96. World millet production trends.

215

28500

27500

26500

25 500

2 4 5 0 0

23500

22 500

21500

20 500

0

22000

2 1 0 0 0

20000

19000

18000

17000

16000

15000

0

2 0 5 0

2 0 0 0

1950

1900

1850

1800

1750

0

70

60

50

40

30

20

10

0

1964 65 66 61 6 8 6 9 70 71 72 73 74

Y e a r

( t )

Countr ies w i th cent ra l lyp lanned economies

P = 20492.70 + 583.59 t ( 7 .82 ) b

P= 16013.80 + 207.60 t (1.25)

Countr ies w i th l ess -developed market econo-mies ( in SAT)

P=1807 .55 + 20.32 t (3.70)

Countr ies w i t h l ess -developed market econo-mies (other than SAT)

Countr ies w i t h developedmarket economies ( i nSAT)

P = 2 1 . 0 2 + 1.66 t (1.63)

Count r ies w i t h developed

market economies (other

than SAT)

P = 6 3 . 5 5 - 4 . 9 1 t ( - 13 .15 )

Interregional Trade and Welfare

Spatial equilibrium models were used to explainthe potential contribution of improved marketchannels and trade on cropping patterns, ag-gregate productivity, and the welfare of con-sumers and producers. Examples of rice, sor-ghum, and chickpeas grown in Andhra Pradesh,Madhya Pradesh, and Maharashtra were selec-ted for study.

Results suggested that if trade in food grainswere restricted to 10 percent of "free-trade"levels, food-grain production would be about 2 percent less in these three states. Most of thereduction would occur in chickpea, and to a lesser extent in sorghum (Table 78), becauseregions would be prevented from concentratingon crops for which they have a comparativeadvantage.

Relative prices of food grains would be alsomarkedly affected by trade restrictions. Forexample, prices of the three grains would fall inMadhya Pradesh (Table 78). This would meansubstantial reductions in the production of thethree crops in that state. In Andhra Pradesh,chickpea prices would rise by about 140 percent.Prices and production of rice in Maharashtrawould rise and for chickpeas they would fall.This illustrates how restrictive trade policies canencourage production of crops in areas where the

crops do not have a comparative advantage, andaway from those where they do grow compara-tively well.

Pulse production in India has declined sub-stantially in recent years. The analysis aboveillustrates how food policies could be used toenhance pulse production and help arrest thistrend. In the process, welfare of consumers andproducers as a whole could be improved. How-ever, the distribution of these welfare gainswould differ amongst the states and betweenconsumers and producers.

Look ing Ahead in theEconomics Program

In the Economics Program, we expect -

— to continue our village-level studies insouthern India for a further year or two, andto extend them to include areas in northernIndia and Africa;

— to develop simulation models which can helpevaluate the potential for water harvestingand supplementary irrigation on differentcrops in different agroclimates;

Table 78. Estimated price and production changes re sulting from food-grain trade restrictions in threestates of India.

States

Andhra Pradesh Madhya Pradesh MaharashtraOverall

Crops Price Production Price Production Price Production Production

(Change in %)

- 5- 7

- 1 7

RiceSorghumChickpea

- 8 - 1- 3 8 - 4

+ 143

- 3 4- 3 7- 6 1

(Change in %)

- 5- 7

- 1 7

+ 68 +29+ 38 - 4- 6 2 - 2 2

+ 1 - 5

- 1 3

aFrom zero production in the unrestricted case to only 44000 metric tons.

216

— to identify necessary conditions required forsuccessful group organization and action forimproved land and water management basedon the watershed concept and to initiatesocioeconomic experiments to test if theserequirements are sufficient for sustainedadoption;

— to develop techniques to assess importance ofinclusion of risk-reduction characteristics innew technologies;

— to have quick screening methods for selectinggrain-types preferred by consumers;

— to identify market constraints to increasedfood-grain production in the SAT;

— to have identified more precisely humannutritional needs in the Indian SAT andICRISAT's role in improving nutritionalstatus.

217

change of germplasm of the ICRISAT crops andto short-term consultancies from senior staff.Building on this, a second phase of assistance hasevolved which provides various forms of trainingto candidates from SAT countries and includesposting of ICRISAT scientists to selected sites inSAT countries to undertake research programsof a regional nature as well as to assist in nationalprograms.

West African Cooperative Project

In January 1975, ICRISAT and UNDP enteredinto a 3-year contract which had as its primeobjectives cooperation with and strengthening ofexisting West African agricultural research pro-grams to develop improved higher-yielding vari-

Figure 97. An important aspect of cooperation with crop scientists all over the world is exchange of germplasm samples and seed. During 1976-1977, more than 38 000 samples were despatched to scientists in more than 70 countries.

219

C o o p e r a t i v e

P r o g r a m s

Since its founding, ICRISAT has been com-mitted to the important and fundamental task ofassisting the SAT countries of the world increasetheir food-production capabilities. As rapidly asresearch programs were developed and plantmaterial and technological information beca'meavailable at ICRISAT Center, steps were takento initiate programs of cooperation and assis-tance with the national research programs inSAT countries internationally. Initially, the pro-grams were limited to the provision and ex-

eties of sorghum and millet that provided con-sistent and reliable yields, and the transfer of newtechnology on crop-production systems. Train-ing of candidates from the region was alsorecognized as vitally important. The drought-stricken Sahel area below the Sahara was thefocus of concern; sorghum and millet are thestaple foods of the majority of the people in thisarea and the food-population balance has beenand continues to be seriously strained. Thecountries making up this region are Senegal,Gambia, Mauritania, Mal i , Upper Volta,Ghana, Togo, Nigeria, Niger, Benin, Chad, andCameroon-a total of 12 in all.

The strategy for implementing an effectiveprogram of assistance, which would have anearly and measurable impact, was outlined.Basically, the concept was to post ICRISATscientists to existing research institutions atBambey, Senegal; Samaru, Nigeria; Kamboinse,Upper Volta; and Maradi, Niger. These scien-tists would be linked to and would work inconjunction with ICRISAT headquarters in Hy-derabad, other ICRISAT cooperating centersinternationally, and other national programs inthe region as well as program of the host country.Subsequently, in 1977, the contract was expand-ed to include Sudan with ICRISAT scientistsheadquartered at Wad Medani. Additional sup-port from other agencies such as Ford Foun-dation, USAID, and the Dutch Government hasprovided for additional scientists at Bamako inMal i and Kambomse in Upper Volta.

Implementation of the program has continuedto gain momentum and has become fully oper-ational for the 1977 crop year. The total team of14 scientists is now in place. The major start-upcapital items (buildings, plot, equipment, andvehicles) have been constructed or purchasedand each component within the network is readyto contribute to the overall objective of assistingthe small farmer to increase his food production.

Project manager. The project manager wasdeputed to ICRISAT by I R A T in July 1975. Hehas established an excellent regional office inDakar, Senegal, from which he provides themanagement and supervision of the overall pro-

ject and maintains close liaison with all agenciesthat work on programs of assistance in theregion. Recently a French-English translationservice has been added to the office and much ofthe excellent relevant research information pre-viously available only in French is being dissemi-nated to English-speaking scientists as well.

Senegal. During the year, a millet breeder andsorghum/millet entomologist were posted to theBambey Research Station and areas of responsi-bility were outlined in collaboration with theStation Director and scientists. It was agreedthat the millet breeder would be involved in threetypes of activities - (i) assisting the national pro-gram through introduction and evaluation of thebest ICRISAT Indian and East African ma-terials for testing, crossing, and further selectionof superior types; (ii) setting up coordinatedregional trials in Senegal, Mal i , Mauritania, andGambia, involving the most promising materialfrom all the West African countries; and(iii) providing to the core program at ICRISATCenter the improved germplasm selected fromlocal and regional trials in Senegal and elsewherein West Africa. This material is then evaluated infurther breeding programs and internationaltrials. During the off season several nurserieswere grown under irrigation and a substantialbeginning was made toward these objectives.

The entomologist will undertake national,regional, and international responsibilities withmajor emphasis on sorghum pests. Initially, themost serious insect problems in the region mustbe identified in order to determine the ap-proaches whereby short- and medium-term re-search is likely to be most fruitful. The economicimportance of the wide variety of pests is to beestablished through observations in farmers'fields as well as through controlled plot experi-ments at Bambey. Screening of ICRISAT andlocal material for various insect resistances is anintegral part of the program.

Mali. An agronomist-selector joined duty atthe Sotuba Research Station in Apr i l 1976 andcarried out the first set of ICRISAT screening,yield, and agronomic trials during the year,

220

utilizing six widely separated locations within thecountry. The primary focus was on screeningsorghum and millet nurseries, supplied by thecore programs at ICRISAT, for local adapta-bility as well as to determine potential in-tercropping practices and fertilizer responses.Close collaboration was established with thenational scientists and IRAT personnel as well aswith the USAID-sponsored "Operation M i l "project at Mopti.

Upper Volta. In 1975 and 1976 a sorghumbreeder and plant pathologist respectively wereappointed to assist the national sorghum im-provement program and to initiate ICRISATsregional effort in that area. In 1977 a milletbreeder and two agronomists were added to theteam. Major emphasis is placed on developingphoto-insensitive, shorter-cycle, medium-heightvarieties of sorghum and millet capable of with-standing drought conditions at higher plantpopulations. Initial screening of large numbersof nurseries from ICRISAT Center for diseaseand insect resistance and general adaptabilitywas conducted and promising materials areavailable for multi-locational testing throughoutthe region in 1977. Several very promising linesof sorghum having good levels of disease andinsect resistance and potential high yield understress conditions were identified. A limited test-

Figure 98. ICR1SA Ts laboratory and office building constructed at the research station in Kamboinse, Upper Volta.

ing program on hybrid sorghum also revealed a good potential for the immediate transfer ofIndian hybrids to West Africa, although therewas evidence that hybrid seed from Indianparents might be more difficult to produce in theWest African regions. A number of on-farmtrials and demonstrations, featuring improvedvarieties and cultural techniques, were carriedout in collaboration with farmers in UpperVolta. Expressions of interest by increasingnumbers of small farmers reflect their appreci-ation of this effort.

With the addition of a full-time millet breeder,the limited program (previously carried out as anadjunct of the sorghum-improvement research)wil l be expanded significantly and will follow thegeneral broad research objectives as for sor-ghum. Similarly, in 1977 agronomy experimentswill be introduced. The initial effort will bedirected towards developing techniques that willbe within reach of the average small WestAfrican farmer. Such techniques will aim atutilizing the natural environment as much aspossible without relying heavily on chemicalsand other inputs. Studies on the interactionbetween cereals and legumes as intercrops and inrotation, crop-residue management and its effecton percolation and retention of rain water,population densities in sole and intercroppedstands, etc., will form a part of this program.

Niger. In 1977 a pearl millet breeder was postedto the Tarna Research Station to work jointlywith the national program on millet improve-ment and to expand ICRISAT's regional re-search network on this major food crop. Closecollaboration with scientists in the region, par-ticularly in Upper Volta and Nigeria, will bemaintained. Emphasis will be placed on breedinghigher-yielding drought-tolerant shorter types ofplant that will be resistant to diseases and pestsand will grow well in higher plant densities.

Nigeria. A plant pathologist to work on thediseases of millet and sorghum was appointed in1977 and headquartered at the Institute forAgricultural Research at Samaru. He will assistthe pearl millet breeder appointed there a year

221

earlier. A season of crossing and testing has beencompleted. Materials supplied by ICRISATCenter as well as those available locally andelsewhere in the West African region were stud-ied. The trials contained synthetic progenies,composites, experimental varieties, adaptationentries, hybrids, and inbreds. Selections weremade on the basis of adaptability, uniformity,yield, and resistance to disease and insects.

East Afr ican Cooperative Programs

Sudan. Following discussions held in late 1976by the Government of Sudan, UNDP, andICRISAT it was agreed that an expansion of theUNDP/ICRISAT West African program to in-clude cooperation with Sudan on sorghum andmillet improvement would be in the overall bestinterests of all concerned. Subsequently, a sor-ghum breeder and a millet breeder were recruitedand stationed at the Agricultural Research Cor-poration headquarters at Wad Medani. Thesescientists wil l work closely with the nationalprograms on these crops and at the same timeaccept a regional responsibility as an East Af-rican extension of the total ICRISAT researchnetwork. A great deal of plant material fromICRISAT Center as well as that obtained locallyand elsewhere in Africa wil l be tested in the firstyear. A total of eight millet nurseries and sevensorghum nurseries are planned-these wil l in-clude exotics, segregating populations, and ad-vanced yield trials. An ambitious crossing pro-gram is also envisaged.

Tanzania. A proposal for a sorghum and pearlmillet improvement program for Tanzania- tobe operated in collaboration with the Govern-ment of Tanzania, I ITA , and USAID -has beenunder active discussion for some time and isexpected to get under way by posting a plantbreeder and agronomist at the Ilonga ResearchStation before the next crop year. Tanzania hasexpressed renewed interest in increasing its pro-duction of sorghum and millet and the majorobjectives would be to develop varieties andmethods of culture which wil l give improved andmore-stable yields, better grain quality, and

greater dependability of harvest. Particular at-tention will be paid to varieties of short tomedium duration, short to medium height, goodgrain quality, drought tolerance, and resistanceto important pests and diseases of the area -which include Striga, birds, shoot fly, and stemborer.

Ethiopia, Kenya, Uganda, Somalia. ICRISATcollaborated closely with these countries bysupplying and exchanging germplasm and vari-ous nurseries, including advanced lines of ap-propriate ICRISAT crops, for further testingand selection within national programs. A num-ber of scientists from these countries visitedICRISAT Center and scientists from ICRISATvisited many of these trials. Close workingrelationships have been established. ICRISAT iscontinuing to explore possibilities for strength-ening of cooperative work on breeding high-altitude, cold-tolerant sorghum for the region.

Southern Afr ica

The network of ICRISAT cooperative centerscontinued to expand during the year to includeBotswana and Lesotho. Scientists from thesecountries visited ICRISAT and have been sup-plied with seeds of appropriate crops as re-quested for incorporation into their crop-improvement programs.

South American Cooperative Program

Brazil. There was a marked increase in col-laboration between Brazil and ICRISAT, includ-ing exchange of sorghum and millet materials,exchange visits by scientists, and training andworkshops. ICRISAT scientific teams visited theSAT region of Brazil twice during the year - t h efirst team consisting of four scientists and thesecond team of five. Attention focused in thedirection of sorghum and millet crop improve-ment, land and water management, intercropp-ing, and selection and development of a re-search station representative of the major SATareas of the Brazilian northeast. Three Brazilianscientists visited ICRISAT Center to become

222

familiar with the Institute's ongoing programs.One attended the International Sorghum Work-shop. A research fellow was brought toICRISAT Center for a 2-month training sessionin Agroclimatology.

Cooperation with C I M M Y T

Responsibility for the high-altitude cold-tolerantsorghum-breeding program introduced byC I M M Y T in Mexico was turned over toICRISAT, with C I M M Y T still providing thefacilities and logistic support for its continu-ation. Reciprocally, ICRISAT is providing a similar service to the C I M M Y T Asian Maizeprogram now based at ICRISAT Center.

Asian Cooperative Programs

ICRISAT continued to strengthen its ties with a number of countries in Asia during the year.A large number of trials of the various crops weresupplied to Pakistan, Bangladesh, Sri Lanka,Burma, Thailand, Afghanistan, and Nepal, andsenior scientists from headquarters visited sev-eral of these countries for follow-up discussionsand observations of the growing nurseries. A formal working agreement between the PakistanAgricultural Research Corporation andICRISAT was finalized to permit a more directflow of scientists and materials betweeninstitutions.

India. A formal Memorandum of Understand-ing has been signed with the Indian Council ofAgricultural Research for all cooperative pro-grams with ICAR Institutions programs andAgricultural Universities in India. Under thisagreement, in the meetings of the Joint AdvisoryCommittee, the cooperative programs of mutualinterest are discussed.

Based on the suggestions of the Policy Com-mittee, the Government of India has been re-quested to authorize ICRISAT's setting up re-search substations at Bhavanisagar (Tamil NaduAgricultural University, Coimbatore), Dharwar(University of Agricultural Sciences, Bangalore),Gwalior (Jawaharlal Nehru Krishi Vishwa Vidy-

alaya, Jabalpur), Hissar (Haryana AgriculturalUniversity, Hissar), and in Jammu and Kashmir.The Bhavanisagar center is primarily for re-search on sorghum and millets; Dharwar centerfor research on downy mildew and other diseasesof sorghum; Gwalior for long-duration pigeon-pea, and Hissar for pearl millet, chickpea, andshort-duration pigeonpea. These centers are tobe established with the cooperation and supportof local agricultural universities. For off-seasonlimited work on chickpea, a subcenter in theKashmir Valley is considered desirable in prin-ciple. Preliminary trials on summer crops ofchickpea have been started for the purpose.

Excellent cooperative work on Striga of sor-ghum has been started at Punjab Rao KrishiVidyapeeth, Akola, and in postgraduate trainingat Andhra Pradesh Agricultural University,Rajendranagar.

Cooperative programs in farming systems re-search, involving ICRISAT, the Dryland Pro-grams, and selected agricultural universities aregetting under way. Locations of the work andcooperating universities are: Sholapur and Ma-hatma Phule Agricultural University, Rahuri;Bangalore, Bijapur (University of AgriculturalSciences, Bangalore); Indore (Jawaharlal NehruKrishi Vishwa Vidyalaya, Jabalpur) and Ranchi(Rajendra Agricultural University, Bihar); andHimayat Nagar, (coordinated cell of ICARDryland Agricultural Program). The projectsinclude study of resource utilization under pre-sent management practices, watershed-based re-source development and agricultural productionin village areas, hydrologic studies to improvewater utilization in rainfed agriculture of theIndian SAT, and studies on water utilization intank-irrigated areas.

In village level studies, conducted in six vil-lages of Sholapur, Akola, and Mahbubnagardistricts, the cooperation of the agriculturaluniversities in Maharashtra and Andhra Pradeshin certain phases of the work is mutuallybeneficial.

Testing of the promising material of sorghum,pearl millet, pigeonpea, and chickpea is beingdone through the All-India Coordinated Projects

223

for these crops. ICRISAT has developed goodfacilities for screening pearl millet for downymildew disease, and promising lines from theIndian program are being screened, thus provid-ing valuable support to the national program.Likewise, sorghum from the Indian program isalso being screened for shoot fly-resistance underICRISAT's program. Similar cooperation isbeing extended for screening for other importantdiseases in pulses.

Other Organizations

ICRISAT recognizes a great backup resource inthe scientific laboratories and institutions of itshost country - India - and of the various nationsrepresented in the Consultative Group. Thesewil l be cultivated and developed as they areidentified and mutual interests and opportunitiescan be matched. Among those of particularinterest are the sorghum protein-quality pro-gram at Purdue University; the conversion anddisease-resistance programs with sorghum atPuerto Rico and at Texas A and M University,the program on physiology of stress at theUniversity of Nebraska; the drought-stress pro-gram at Saskatoon; the projects on stimulants togermination of Striga and orobanche at Sussexand the Weed Research Organization of the U K ;physiological studies on tropical-plant responsesto closely controlled environmental conditions inthe phytotrons at Reading, England, and theCSIRO, Australia; programs on water relationsand on legume improvement at Cambridge; theroot-development studies at Letcombe Labora-tories, UK ; nitrogen-fixation and rhizobium cul-tures for tropical legumes by the CSIRO, Aus-tralia, the Taximetrics Laboratory of the Uni-versity of Colorado, and many others. Gen-erally, the major financing of these projects wil lcome from sources outside ICRISAT, but theirresults wil l be of considerable importance toICRISAT's program. Through joint collabo-ration, the relevance of such projects to the semi-arid tropics can be sharpened; facilities andscientific talent of such institutions can comple-ment those of ICRISAT and relieve it of certainsegments of work; ICRISAT can in turn provide

needed facilities and environments for study ofthe field aspects of the problems under in-vestigation. Other international institutes havealready made considerable progress in develop-ing this type of cooperation, and ICRISATconsiders it to be an important avenue indevelopment.

Fellowships and T ra in ing

Educational opportunities and experiences wereprovided at ICRISAT Center for scientists,research workers, scholars, and agricultural ad-ministrators. A l l were employed or preparing foremployment in rainfed semi-arid agriculturalregions of the tropics. Practical field and lab-oratory experiences were provided to developskills for more effective participation in research,production, extension, training, or adminis-trative activities in their national or localprograms.

Research fellows holding advanced degreescame from India, Sri Lanka, Brazil, and Koreafor practical and theoretical experience in theirfields of expertise. They worked with ICRISATscientists on current field and laboratory re-search projects.

Research scholars from the Sudan and uni-versities in India, Canada, Kenya, the Nether-lands, and the United States initiated or con-tinued their thesis research projects as theyworked towards their M.Sc. or Ph.D. degrees.

In-service trainees from 19 countries partici-pated in programs of 2 weeks to 6 months inlength. The programs, tailored as nearly as possi-ble to meet the individual needs of each student,include practical experience in field and lab-oratory research, production and resource-management, and crop improvement. French-speaking trainees (Chad, Mal i , Mauritania,Niger, Senegal, and Upper Volta) complete a 2-month intensive course in the English languageat the Central Institute of English and ForeignLanguages on the Osmania University campus inHyderabad prior to entering their respectivetraining programs.

224

Table 79. Persons completing long-term training prog rams during the 1976-1977 report year.

Name Country Program

In-service:Anne Sara Senegal Pearl milletDiallo Mamadou Senegal Pearl milletOuattara Sou Upper Volta Pearl milletLassana Tigana Mal i Pearl milletKern Tolna Chad Pearl millet

Hamissan Mijiaba Niger Pearl milletThiombiano Nadia Luis Upper Volta SorghumAdama Coulibaly Mali SorghumAbbo Nassengar Chad SorghumSani Idi Niger Sorghum

Ankaurao Souleymane Niger SorghumAbdoa Iro Niger SorghumSani Mamman Nigeria Crop ProductionA . M . Fagge Nigeria Crop ProductionA.A. Adeniji Nigeria Crop Production

A.A. Okolo Nigeria Crop ProductionIbrahim I. Kurba Nigeria Crop ProductionSamuel D.S. Kwabe Nigeria Crop ProductionS.K. Thakre India StrigaS.T. Khade India Striga

Adama Moussa Niger StrigaTaye Wolde Mariam Ethiopia ChickpeaGetahun Terefe Ethiopia ChickpeaAbdalla Fageer Ahmed Sudan ChickpeaY.C. Panchal India StrigaNagaraj India StrigaSubbaiah India Striga

Research Fellows:Assanee Pachiburavan Thailand Cereal PathologyU.P. des S. Waidyanatha Sri Lanka MicrobiologyU.D. Han Korea Farming SystemsJ.S. Jung Korea Farming Systems

Research Scholars:B.K. Hadi India APAU/ IMCChandrashekhar Reddy India APAU/ IMCAnjali Yadav USA EconomicsO.I.J. Ogombe Otieno Kenya Physiology

Junior Principal Scientists:Peter Lawrence Australia Sorghum ImprovementJasper Skinner USA Sorghum ImprovementS.O. Okiror Uganda Pearl millet Improvement

225

Table 80. Participants in the ICRISAT trainingprograms, by country.

1974-75 1975-76 1976-77

Research Fellows:India - - 1Korea - - 2Brazil - - 1Sri Lanka - - 1Thailand - - 1Upper Volta - 1 -

Research Scholars:Sudan - 1 1India - - 5United States - - 2Canada - - 1Netherlands - - 1Kenya - - 1

In-service Trainees:Nigeria 4 8 6India 3 11 71Sweden - 1 -West Germany - 2 -Bangladesh - 1 -Upper Volta - - 2Senegal - - 2

Mali - - 2Chad - - 2Niger - - 6Thailand - - 1Ethiopia - - 2Sudan - - 1

Apprentices(university

students):India 1 3 11USA - - 1

Student agricultural engineers and agricul-tural economics students from India and UnitedStates completed 1- or 2-month apprenticeshipsworking with engineers and scientists in landdevelopment; land and water management; farmmachinery research, operation, and mainte-nance; or data processing and analysis.

Workshops , Conferences,and Seminars

Exchange of information for the transfer oftechnology and cooperative research is essentialand workshops and conferences provide aneffective channel for this purpose. The 1976—1977 report year saw an international workshopand several smaller conferences and workshops.ICRISAT also cooperated with research agen-cies in India and elsewhere in organizing con-ferences and workshops, and ICRISAT scientistsparticipated in practically every major scientificmeeting involving one or more of its five cropsgrown in the SAT environment.

International Sorghum Workshop

Fifty-two sorghum specialists from 26 nationsparticipated in the 1977 International SorghumWorkshop at Hyderabad 6-12 March. The pur-pose of the Workshop was to exchange infor-mation about sorghum research in the SAT,determine how best to develop cooperative re-search among sorghum-improvement workers,review the arrangements for exchanging seedmaterial and information, and to assess trainingneeds and the mechanism for identification oftrainees.

One of the highlights of the Workshop was thepresentation of country papers. These, alongwith information gleaned from questionnairessubmitted by workers from the 26 nations,provided the most up-to-date appraisal of theSAT sorghum situation available anywhere.Such data is essential for international planning.

Face-to-face interchange of ideas and sug-gestions was perhaps the most valuable featureof the Workshop. Information flow (from na-tional program to ICRISAT, from ICRISAT tonational program, and between national pro-grams) is essential if the sorghum workers of theworld are to provide the most effective assistanceto the small SAT farmer of limited means.

The All-India Coordinated Sorghum Im-provement Project served as co-sponsor for theWorkshop.

226

Figure 99. ICRISAT's program for inservice trainees emphasizes practical agriculture along with scientific principles. Fertilizer solutions is the topic here.

DPAP Training Conference

A 6-day training conference was held in Apri l for35 agricultural officers from five Drought ProneArea Program districts in three Indian states.The purpose was to acquaint the officers with thelatest developments in crop production tech-nology and resource management, and to estab-lish communication exchange with developmentworkers and researchers. The program was joint-ly organized by the All-India Coordinated Re-

search Project for Dryland Agriculture, theCentral Soil and Water Conservation Researchand Training Institute, agricultural universitiesin the three states, DPAP, and ICRISAT.

Chickpea Breeders' Meeting

The annual chickpea breeders' meeting was heldon 24-25 February at ICRISAT Center. Thirtybreeders from India and a breeder from Israel

227

met to review the progress of breeding, pa-thology, germplasm, entomology, microbiology,physiology, and nutritional quality and to tourICRISAT Center. Interspecific hybridization inchickpea occupied one half-day session, and thefinal afternoon of the meet was spent in the field,where the visiting breeders selected lines thoughtto be useful in their own programs.

Seminars

Institute and program seminars are a regularfeature of life at ICRISAT. Institute seminars arescheduled monthly, and many of the programsschedule weekly seminars. Visiting scientistsusually present one or more seminars duringtheir stay at ICRISAT. These activities provide a

valuable forum for exchange of ideas andinformation.

Field Trips and Visits

ICRISAT frequently schedules field trips andother short-time activities for visiting groups.The Institute considers this to be a valuableavenue of information exchange. Occasionally,such activities serve a special purpose. Forexample, during September, all of the milletbreeders in India visited ICRISAT Center tobecome more familiar with the research programand to make their own selections fromICRISAT's breeding nurseries. Discussions withthese workers regarding the future direction ofICRISAT's efforts in breeding and pathologywere quite useful.

228

P l a n t Q u a r a n t i n e

For the proper execution of crop improvementprograms at ICRISAT and elsewhere in theSAT, the exchange of valuable seed material withour scientists in the outreach programs and othercolleagues in cooperative programs throughoutthe world was carried out through the PlantQuarantine Services of Government of India.Sorghum, pearl millet, pigeonpea, chickpea, andgroundnut were inspected and cleared at theCentral Plant Protection Training Institute, loc-ated in Hyderabad. Seed material of other cropsessential in the Farming Systems research pro-gram was cleared at the National Bureau of PlantGenetic Resources in New Delhi. Unrootedgroundnut cuttings were released by the Direc-torate of Plant Protection, Quarantine, andStorage, Faridabad.

Expor t o f Seed Ma te r i a l

During the year, 38 886 samples of sorghum,pearl millet, chickpea, pigeonpea, and ground-nut were exported to 72 countries of the world(Table 81).

A total of 20 676 sorghum samples were sent to38 countries. International cooperative trialsconsisting of population progeny lines, grainmold resistant, pests and disease resistance,Striga resistant, and grain grass cultivars num-bering 12632 were sent to our scientists andcooperators/collaborators in Botswana, Cam-eroun, Ethiopia, Ghana, Kenya, Malawi, Mal i ,Mexico, Niger, Nigeria, Pakistan, Saudi Arabia,Senegal, Somalia, Sudan, Tanzania, Thailand,and Upper Volta. Segregating material andcultivars totalling 6316 were sent to 33 countries,the bulk of it going to Bangladesh, Botswana,Brazil, Cameroun, Ecuador, El Salvador,Kenya, Malawi, Malaysia, Nicaragua, Nigeria,Tanzania, Yemen, and Zambia. Germplasm cul-tivars totalling 1 728 were exported to Bangla-

desh, Ethiopia, Malawi, Malaysia, Mexico,Somalia, Venezuela, and West Germany. SaudiArabia and Venezuela received 36 drought-resistant, salt- and heat-tolerant lines.

Of the 8081 pearl millet samples sent to 29countries, 4702 went out as Internationalcooperative trials to 19 countries. Downy Mi l -dew, Ergot, and Smut Nurseries were sent toNigeria, Senegal, and Upper Volta; 2731 cul-tivars consisting of germplasm, varieties, com-posites, synthetics, and segregating materialswere sent to Ethiopia, Libya, Mexico, Niger,People's Republic of China, Senegal, Venezuela,West Germany, and Zaire. Forty-eight sampleswere sent to Canada, England, and Denmark forcarrying out biochemical studies.

International Chickpea Cooperative Trialsmaterial formed the bulk of chickpea export. Ofthe total of 7 218 trial samples, 6 661 were sent tocooperators/collaborators in 27 SAT countries.Germplasm and segregating cultivars numberingS6S were sent to Botswana, Canada, Italy,France, Kenya, New Zealand, Syria, USA, Ven-ezuela, and Zambia. Thirty-four samples weresent to various agencies in England and USA forphysiological and biochemical studies.

Pigeonpea samples were sent to 34 countries.Seed samples totalling 494 went as vegetable-type observation nurseries to Belize, El Salvador,Dominican Republic, Panama, Philippines,Puerto Rico, Venezuela, and Zambia. Germ-plasm cultivars numbering 1277 were exportedto Australia, Ghana, Malaysia, Mauritania,Puerto Rico, and Tanzania. Varietal trial entriesand varietal purification cultivars numbering 966were sent to 22 countries. Twenty-eight sampleswent to Canada and Denmark for biochemicalstudies.

A total of 146 groundnut cultivars were ex-ported to Australia, Malaysia, Pakistan, SriLanka, and West Indies. One-hundred sampleswent as germplasm cultivars; 46 observationaltrial entries were exported to Sri Lanka.

Twenty samples of safflower, paddy, wheat,cowpea, and dried neem leaves were examinedand cleared at the National Bureau of PlantGenetic Resources, New Delhi, for despatch toNigeria, Taiwan, Thailand, and Upper Volta.

229

Table 81. Seed samples exported or imported by ICRISAT daring 1976-1977.

Sorghum Millet Chickpea Pigeonpea Groundnut Other

Country Exp Imp Exp Imp Exp Imp Exp Imp Exp Imp Exp Imp

Botswana 929 22Cameroun 388Cape Verde Islands 52Central African

Republic 52Egypt 490

Ethiopia 4313 413 69 414Ghana 325 51Kenya 1335 10 25 1Libya 14 250Malawi 389 65 4

Mal i 309 120Mauritania 51Niger 735 277Nigeria 949 2 1966 222 15 1 1 Senegal 615 1378 60 74 1

Sierra Leone 20Somalia 892 60Sudan 2154 595 195 2Tanzania 529 120 150 98Tunisia 135

Uganda 60 100Upper Volta 1905 1319 64 10 2

Zaire 20 14Zambia 33 50 103 1Afghanistan 150 20 13

Bangladesh 730 7 60 681 1Burma 358 12Iran 150Iraq 185Japan 6 43

Jordan 135Korea 5Malaysia 142 25 36Nepal 184 12New Guinea 14

Pakistan 791 1406 1073 20 3People's Republic

of China 35 12Philippines 274 38

continued

230

Table 81 continued

Sorghum Millet Chickpea Pigeonpea Groundnut Other

Country Exp Imp Exp Imp Exp Imp Exp Imp Exp Imp Exp Imp

Saudi Arabia 92 4 25Sri Lanka 52 49 19 46

Syria 60 588Taiwan 25 1Thailand 1708 2 60 249 33 8USSR 6Yemen Arab

Republic 238 60 74 11

Argentina 25Brazil 100 11 120 8Chile 60 189Colombia 10 60 10Dominican Rep. 424

Ecuador 107El Salvador 51 63Mexico 442 46 414 8Netherlands 3 15Nicaragua 50

Panama 38Peru 164Puerto Rico 184 17 14Venezuela 44 18 20 104West Indies 26 67 25

Belize 38Denmark 15 7 24England 21 30 202 82France 12 7Italy 10Spain 35 15

Turkey 280West Germany 2 4Australia 34 25 1000 36 47Canada 12 9 4New Zealand 40USA 79 275 18 338 4

20676 716 8081 124 7218 48 2765 40 146 598 20 259

Total 21392 8205 7266 2805 744 279

231

Figure 100. ICRISA T plant quarantine scientists and quarantine officials of the Government of India work together to assure minimum delay in import and export of seed materials.

Impo r t o f Seed Ma te r i a l

ICRISAT received 1 785 samples of sorghum,pearl millet, chickpea, pigeonpea, and ground-nut along with seed material of some other cropsfor use in our Farming Systems research pro-gram from Afghanistan, Australia, Bangladesh,Brazil, Colombia, England, Ethiopia, France,Japan, Kenya, Mexico, Philippines, Nether-lands, Nigeria, Puerto Rico, Senegal, Thailand,Upper Volta, USSR, USA, and Zambia (Table81).

Bulk of the sorghum introductions came fromEthiopia (413) and the USA (275) - Washington12, Texas 16, Minnesota 17, Mississippi 18,Illinois 22 {Shoot-fly resistant sorghum x sugarcane crosses), Nebraska 81, and Purdue

109. These materials were imported to utilizetheir yield, physiological, microbiological,disease- and pest-resistance attributes and to fillgaps in our germplasm collections. A total of 124lines of pearl millet germplasm and diseaseresistance material were introduced from Sen-egal and Upper Volta.

Forty-eight chickpea cultivars were intro-duced from Afghanistan, Australia, and Nether-lands under the germplasm exchange program.Agricultural research agencies of France, Nig-eria, and Puerto Rico formed the main sources ofpigeonpea germplasm and breeding lines toICRISAT.

For groundnut improvement, 192 unrootedcuttings of interspecific hybrids of Arachis spp.were received from Dr. Phil Moss's collection at

232

the University of Reading, U K , for testing theirreactions to Cercospora leafspot. In addition,406 cultivars -comprising germplasm lines, dis-ease resistant material, released cultivars, andgenetic marker lines -were received from NorthCarolina and Georgia in the USA, Japan, PuertoRico, and Zambia.

In other crops, 259 samples of Acacia, beans,Bermuda grass, corn, cowpea, Cicer herbarium, jojoba, Leucaena latisiliqua, Phaseolus atropur-pureus, Setaria, and Sesamum were cleared at theNational Bureau of Plant Genetic Resources,New Delhi for the Farming Systems and CropImprovement programs.

Quarantine Regulationsfor Release of Pearl Milletand Groundnut Imports

Pearl Millet: The ICAR Quarantine Commit-tee on pearl millet seed imports recommended on21 Dec 1976, certain procedures that, if followed,would permit pearl millet seed to be introducedwithout the risk of the introduction of pearlmillet downy mildew. The procedures were:(i) seed should be collected from downy mildew-free plants; (ii) seed should be physiologicallymature, dry, and treated with fungicides beforedespatch; and (iii) on arrival the seed should besoaked in a 1:1000 solution of HgCl2 for 10minutes, then immediately washed in fourwashes of sterile water. After washing, the seedbe transferred immediately into a water bath setat 55°C for 12 minutes. Following the waterbath, the seed should be placed in an incubatorset at 35° C for 12 hours and then at 40°C for anadditional 12 hours. This procedure was fol-lowed by the quarantine officers at Central PlantProtection Training Institute and 124 samples ofpearl millet seed imported from West Africa werereleased to us for planting in the Post-EntryQuarantine Isolation Area at ICRISAT Center.

However, in view of a letter received fromDr. Neergaard, Director General of the DanishInstitute of Seed Pathology for DevelopingCountries in Copenhagen, in Feb 1977, on his

Institute's inability to guarantee the hot-watertreatment of pearl millet seed for purposes ofquarantine clearance, Dr. M.S. Swaminathan,Director General, Indian Council of Agricul-tural Research, organized a meeting of ICRISATand Indian government scientists in Apri l 1977to review the quarantine arrangements forICRISAT seed. In this meeting it was decidedthat a committee should reconsider the method-ology of treating future pearl millet seed con-signments for quarantine clearance.

The appointed committee met and concludedthat treatment of seeds with CIBA fungicideCGA-1-82/50W at a rate of 3 g/kg seed in 1000ml water for 6 hours, in addition to the pro-cedures recommended in the December 1976meeting, should afford adequate safeguardagainst the risk of the pathogen being introducedthrough pearl millet seed imported byICRISAT.

Groundnut: The earlier procedure of growingimported seed material in the Post-EntryQuarantine Isolation Area and releasing the seedof the healthy plants was discontinued. Afterprolonged discussion with the Directorate ofPlant Protection, Quarantine, and Storage,Government of India at New Delhi, the pro-cedure for releasing groundnut seeds was fin-alized. Ten seeds from each sample would begrown in the nethouse at Central Plant Pro-tection Training Institute and kept under strictobservations for a period of 6 to 8 weeks.Seedlings showing the slightest symptoms of anexotic disease were discarded; healthy seedlingswere released for planting in the Post-EntryQuarantine Isolation Area at the ICRISATCenter. The material was again inspected untilharvest by Government of India and ICRISATscientists. This procedure, though slow, is work-ing out satisfactorily, and 406 cultivars have beencleared during the year.

Postentry Quarant ine

It was decided in ICAR Quarantine Committeemeetings in December 1976 and reconfirmed in

233

the meeting of Apr i l 1977 that all the importedseed material released after quarantine inspec-t ion would be grown for the first generation inthe Post-Entry Quarantine Isolation Area atICRISAT Center. The crops would be inspectedby Government of India quarantine officers andICRISAT scientists at weekly intervals untilharvest, and any plant showing symptoms ofexotic disease would be destroyed.

Consequently, 124 cultivars of pearl milletimported from West Africa were grown in thePost-Entry Quarantine Isolation Area andexamined periodically. Some plants showingdoubtful disease symptoms were destroyed, butall cultivars are still growing, and it is hoped thatwe wil l be able to harvest seed material f rom allreleased cultivars for our experimental work.

Six-week-old seedlings of 406 cultivars ofgroundnut released after careful inspection for 6 to 8 weeks in Central Plant Protection TrainingInstitute nethouse were planted in the Post-Entry

Quarantine Isolation Area during the year. Simi-larly 192 unrooted cuttings of Arachis spp., aftergrowing for 6 to 8 weeks in our nethouse, werealso transplanted to this area. These plants wereexamined each fortnight and all doubtful plantswere removed and destroyed. However, not oneof the released entries was totally lost due todisease incidence.

Facil it ies Prov ided byI C R I S A T

To assist in the speedy clearance of its seedmaterial, ICRISAT has provided an inspectionroom and fumigation chamber at Central PlantProtection Training Institute. This facility wasformally presented to CPPTI by Director L.D.Swindale in Apr i l 1977.

234

C o m p u t e r S e r v i c e s

ICRISAT's Computer Services unit is equippedwith a "DECdata" system 550, which it operatesas a dedicated timesharing system utilizing theRSTS/E (Resource Sharing Time Sharing Exten-ded) operating procedure. Resources of thesystem are accessible to ICRISAT scientiststhrough terminals located in the computercenter.

Objectives

The goal of the Computer Services unit is tointegrate the use of the ICRISAT computersystem into the daily routine of the research,administrative, and service activities of the In-stitute. In order to achieve this goal, the Com-puter Services unit is

(i) developing interactive systems which areeasy to use,

(ii) providing data-entry services, and

(iii) conducting seminars on computer usageand programming.

Current Stage ofDeve lopment

The major emphasis in Computer Services dur-ing the past year has been on the improvement ofthe statistical-analysis capability of the system.The file structure used to store research data inon-line disk files was improved to permit thestoring of larger collections of data more ef-ficiently. Eleven new data-editing options wereintroduced and the entire set of editing routineswas grouped into a file-editing subsystem. Thissubsystem permits the different editing optionsto be invoked through the specification of a two-letter code. The statistical-analysis routines wereintegrated into a single system called the CropResearch Integrated Statistical Package(CRISP), which permits the use of a set ofcommands for selecting the desired analysis. A single computer-system command is required to

gain access to CRISP. Under CRISP, the usercan proceed through a sequence of analyses bychoosing the appropriate options and does nothave to revert to computer-system commands.At present there are 42 statistical-analysis anddata-manipulation routines available underCRISP. Access to the file-editing subsystem isgained through two CRISP commands.

The system developed to store and maintainVillage Level Studies data collected by the Eco-nomics Program was improved by the additionof a subsetting capability. This feature permitsthe selection of a subset of data from a data file,based on a logical combination of conditionsimposed on the data. A conversion routine wasdeveloped which permits restructuring of theEconomics Program file structure into the struc-ture required under CRISP, in order that existingstatistical analysis routines can be used to ana-lyze this data also. A modified form of the systemdeveloped for the Economics Program is beingused to store and maintain ICRISAT's germ-plasm data. During the next year, an improvedsystem for storage and retrieval of large databases will be developed to accommodate both theEconomics and germplasm data.

An improved system for performing randomi-zations, printing seed-packet labels, and generat-ing field books was introduced to the users inearly May, 1977. Under this system the name anddescription of, factor identification for, and theperson responsible for the experiment are storedin a file on disk. The resultant randomization isalso stored in this file. This information is used toinitialize the data file required to store the datafor analysis, using routines under CRISP. Datacan be entered directly from field books withoutreordering, thus reducing the occurrence oftranscription errors. The field book system isaccessed through commands under CRISP.

Work has progressed on the computerizedfiscal accounting system and various aspects of ithave been under test since Apr i l . Under thissystem, all operational, capital, and special ac-counts wil l be stored in a disk file. Each day a transaction file is created from input at a terminaland the accounts in the budget file are updatedfrom these transactions. Once the system is fully

235

operational, all accounts wil l be current to theprevious day and the status of any account wil lbe readily accessible.

Dr. Jerry A. Warren, statistical consultantfrom the University of New Hampshire in theUnited States, visited ICRISAT again duringJanuary and February, 1977. He conductedclasses in regression analysis and experimentaldesign, consulted with scientists concerningplanning of experiments and required analyses,and instructed staff members how to best utilizethe analysis routines under CRISP.

Look ing A h e a d inCompute r Services

Projects scheduled to receive attention in thecoming year include:

1. Completion of the fiscal accounting systemand the addition of a payroll system.

2. Development of on-line systems for in-ventory control and purchase-orderstatistics.

3. A compact system for the maintenance ofInstitute mailing lists.

4. Improvement of the germplasm-retrievalsystem and development of inventory-control systems for the management ofavailable seed.

5. A personnel data base for the PersonnelDivision.

6. Investigation of word-processing andcomputer-generated offset masters forprinting.

236

L i b r a r y

The Library was shifted from its crowded quar-ters in the Banjara Hills district of Hyderabad toICRISAT Center on 28 February 1977. The newquarters, although temporary, are spacious andaccommodative for the time being. PermanentLibrary facilities are expected to be ready some-time near the end of 1978.

Acquisition

The Library continues to acquire only highlyspecialized research documents pertaining to theareas of interest to ICRISAT scientists. Theprocurement of each document, therefore, re-quires special efforts. Growth in the Librarycollection is presented in Table 82.

Catalog

The card catalog of back-volumes of periodicalholdings has been maintained in addition to theusual Official Catalog, Public Catalog, ShelfList, and Accession List. Back volumes and thesubscribed completed volumes of periodicalshave been bound. In all 1 133 books and periodi-cals were bound during the period of report,making a total of 4000 bound volumes ofperiodicals available to Library users.

Table 82. Accessions of the ICRISAT Library.

TotalMar 1974

TotalMar 1975

TotalMay 1976

Added1976-1977

TotalMay 1977

Books 580Bound volumes of periodicals 231Microforms -Periodicals subscribed 232

23Reprints for master file of biblio-

graphies -

26881045

311344

52252551

471394

23851469

389193

76104020

860587

Books 580Bound volumes of periodicals 231Microforms -Periodicals subscribed 232

23Reprints for master file of biblio-

graphies - - - 1400 1400

Total (less periodicals) 13890

During the period under report 109 docu-ments were supplied to libraries and scientists inICRISATs areas of interest over the world.These figures reveal the excellent relations thatthe ICRISAT Library is maintaining with otherlibraries.

The ICRISAT Library has up-dated its col-lection of the annual reports of various in-ternational and other institutions with areas ofresearch similar to ICRISAT. It now has a unique collection of annual reports from 130institutions of 28 countries. This helps ourscientists to avoid research-in-parallel and inplanning their research-in-series.

Another unique collection consists of copies ofall the papers and tour reports of ICRISATscientists.

Documentation

In the field of documentation, the Library haspublished the following:

ICRISAT Library: Monthly List ofAdditions.

Catalog of Periodicals: ICRISAT Library.1976.

Bibliography of Theses on ICRISAT Spe-cialities. 1977.

Cooperation in Agricultural Libraries in In-dia. Paper presented at ICAR-PAU Seminar

237

on Agricultural Librarianship and Documen-tation, Ludhiana, 2-5 Feb 1977.

Role of Information Science in AgriculturalResearch; Sorghums and Millets InformationCenter. Paper presented in International Sorg-hum Workshop, Hyderabad, 6-12 Mar 1977.

Sorghums and Mi l le tsIn fo rma t ion Center

The Sorghums and Millets Information Center(SMIC) originated from an idea presented at theWorkshop for the Development of an Inter-national Sorghum and Millet Information Net-work held 12-13 May 1975 in the United States.With funding assistance from the InternationalDevelopment Research Centre ( IDRC), SMICwas established in December 1976.

The concept of an information/documen-tation center recognizes that research workers donot have ample time to scan the ever-increasingnumbers of periodicals, even those in theirparticular areas of interests. Unlike the tradi-tional library, an information/documentationcenter reviews the contents of periodic literature,culls out relevant literature, publishes that infor-mation in the form of bibliographies, and collectsmaster files from which copies can be preparedfor ready dissemination.

Specific Objectives

SMIC's specific objectives are:

(a) to collect documents on sorghums andmillets dating from 1969 onwards;

(b) to set up an appropriate document storageand retrieval system;

(c) to bring up to date the existing bibliog-raphies on sorghums and millets;

(d) to provide bibliographic search servicesfor scientists in developing countries;

(e) to provide upon request, particularly fordeveloping countries, copies of material inthe document collection;

(F) to establish a question-and-answer service,using the advice of the scientists inICRISAT's research network; and

(g) to issue a newsletter to reinforce the re-search network and to improve com-munications with outside institutions.

Work Accomplished

Master copies of references cited in the existingstandard bibliographies of millets are beingprocured. So far, about one-fourth of the ap-propriate documents have been secured. Theyhave been obtained from the collection inICRISAT Library and through the collectionsearch of local specialized libraries. Cooperationextended by local libraries has been most helpful.

Efforts are being made to collect master files ofall the references pertinent to sorghum andmillet, so that any scientist may get a copy of thematerial at any time. The master copies so farcollected have been filed according to the entrynumbers of the main bibliography.

Work in Progress

Lists of reference publications, abstracting, re-viewing, and other periodicals in English,French, and Spanish are being prepared forpurchase on priority basis.

Future Plans

The Sorghums and Millets Information Centerwould be responsible for the collection, docu-mentation, and retrieval of all information onsorghums and millets. The work wil l be under-taken primarily in English and French lan-guages; other languages may be included as perrequirement and importance. The ICRISATLibrary has already a good collection of sorg-hums and millets bibliographies in its possession.

238

Figure 101. The ICRISAT Library and the Sorghums and Millets Information Center are tooling up toserve the information needs of the scientists working with sorghum, pearl millet, pigeonpea, chickpea, and groundnut throughout the SAT.

After documentation and collection work isover, SMIC intends to finalize its retrieval systemand produce at regular intervals supplements tothe existing bibliographies on both crops. It

hopes to develop the capability of performingspecific subject searches on demand, and tomaintain profiles of sorghum scientists' specificsubject areas of research.

239

Published by:

International Crops Research Institutefor the Semi-Arid Tropics

1-11-256, BegumpetHyderabad-500016 (A.P.) India

Editor: G. D. Bengtson. Visualizer: S. M. Sinha. Color pages design: G. K. Guglani. Illustrations: A. A. Majid. Illustrationscomposing: T. R. Kapoor. Photography: H. S. Duggal (with technical assistance of A. B. Chitnis). Coordination: Administrativestaff, Information Services. Photo Typesetting: The Macmillan Co. of India Ltd., Bangalore. Printing: Vakil & Sons Ltd.,Bombay.