Citation
Dryland wheat in India

Material Information

Title:
Dryland wheat in India the impact of technical change and future research challenges
Series Title:
Economics working paper
Creator:
Byerlee, Derek
Indian Council of Agricultural Research
Place of Publication:
Mexico D.F
Publisher:
CIMMYT
Publication Date:
Language:
English
Physical Description:
vii, 54 p. : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Wheat -- India ( lcsh )
Wheat -- Economic aspects -- India ( lcsh )
Wheat -- Climatic factors -- India ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )
Spatial Coverage:
India

Notes

Bibliography:
Includes bibliographical references (p. 51-53).
Statement of Responsibility:
Derek Byerlee in collaboration with the Indian Council of Agricultural Research.

Record Information

Source Institution:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services (UFDC@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
33414114 ( OCLC )

Full Text
Dryland Wheat in India:
The Impact oTechnical Change and
Future Research Challenges
Derek13yerlee
in collaboration with the
Indian Council of Agricultural Research
C. I *M M*Y* T
INTERNATIONAL MAIZE AND WilFAT IMPROVEMENT CENTER
Anartano postal 6-641, C.P 06600, Mixico. 13 F., M6xico




E C O N O M I C S
Working Paper 92-05
Dryland Wheat in India:
The Impact of Technical Change and
Future Research Challenges
Derek Byerlee
in collaboration with the
Indian Council of Agricultural Research*
Derek Byerlee is an economist with the International Maize and Wheat Improvement Center (CIMMYT), Mexico. This publication was prepared as a background paper for the 1990-91 CIMMYT World Wheat Facts and Trends: Wheat and Barley Production in Rain fed Marginal Environments of the Developing World (1991). The views expressed in this paper do not necessarily reflect policies of CIMMYT or ICAR.




CIMMYT is an internationally funded, nonprofit scientific research and training organization. Headquartered in Mexico, the Center is engaged in a research program for maize, wheat, and triticale, with emphasis on improving the productivity of agricultural resources in developing countries. It is one of 17 nonprofit international agricultural research and training centers supported by the Consultative Group on International Agricultural Research (CGIAR), which is sponsored by the Food and Agriculture Organization (FAO) of the United Nations, the International Bank for Reconstruction and Development (World Bank), and the United Nations Development Programme (UNDP). The CGIAR consists of some 40 donor countries, international and regional organizations, and private foundations.
CIMMYT receives core support through the CGIAR from a number of sources, including the international aid agencies of Australia, Austria, Belgium, Brazil, Canada, China, Denmark, Finland, France, India, Germany, Italy, Japan, Mexico, the Netherlands, Norway, the Philippines, Spain, Switzerland, the United Kingdom, and the USA, and from the European Economic Commission, Ford Foundation, Inter-American Development Bank, OPEC Fund for International Development, UNDP, and World Bank. CIMMYT also receives non-CGIAR extracore support from the International Development Research Centre (DRC) of Canada, the Rockefeller Foundation, and many of the core donors listed above.
Responsibility for this publication rests solely with CIMMYT.
Printed in Mexico.
Correct citation: Byerlee, D. 1992. Dryland Wheat in India: The Impact of Technical Change and Future Research Challenges. CIMMYT Economics Working Paper 92-05. Produced in collaboration with the Indian Council of Agricultural Research. Mexico, D.F.: CIMMYT.
Abstract: This study focuses on the dryland wheat areas of central and southern India (over half of India's total rainfed wheat area), where wheat is produced on deep vertisols in a fallowwheat system depending heavily on the conservation of monsoon rainfall. Although wheat yields have increased more slowly in dryland areas than irrigated areas, perhaps half of the dryland wheat area is planted to improved varieties (mostly tall, many of them durum wheats) possessing superior drought tolerance, rust resistance, and high quality grain which fetches a price premium. Fertilizer use remains low. Research resources appear to have been allocated to dryland wheat in accordance with its relative importance in total production, but the declining importance of dryland wheat may suggest that less priority be given to wheat research in this environment in the future. However, the close association between the incidence of poverty and dependence on dryland agriculture may justify further research in dryland areas. Consideration should also be given to interactions between favored (irrigated) and marginal (dryland) areas occurring through food and labor markets. Finally, although progress in breeding research may be slow because of the difficulty of the target environment, progress in crop and resource management research may be more rapid. Developing better moisture conservation practices is clearly a research priority, given the limited groundwater supplies for irrigation.
ISSN: 0258-8587
AGROVOC descriptors: Wheats, dry farming, fallow systems, food production, production factors, agricultural development, plant breeding AGRIS category codes: E14, F08, F30
Dewey decimal classification: 338.14
ii




Contents
Page
iv Tables
v Illustrations vii Acknowledgements
1 Introduction
3 Trends in Rainfed Wheat in India 16 Agroclimatic Situation for Rainfed Wheat in India 25 Wheat in the Farming Systems of Central and Southern India 29 Varietal Development and Diffusion 32 Crop and Resource Management 37 Wheat Marketing and Consumption 44 Research Resource Allocation to Dryland Wheat 48 Conclusions 50 Appendix A. Estimated Quantity of Fertilizer Applied to Wheat in Major
Wheat Producing States of India, 1969-88 51 References
..III




Tables
Page
4 Table 1. Average wheat yields, 1988-90, and growth rates in total wheat area by state, 1961-87
8 Table 2. Average rainfed and irrigated wheat area, 1984-86, and growth rates in rainfed and irrigated wheat areas by state, 1961-88
9 Table 3. Compound annual growth rates in components of wheat area in India (%/yr), 1961-88
11 Table 4. Trend growth rate (%/yr) in irrigated wheat area by state, India,
1965-76 and 1976-88
13 Table 5. Annual growth rate (%/yr) of irrigated and rainfed wheat yields by
state, India, 1972-85
24 Table 6. Correlation matrix for simulated wheat production, Sagar District,
Madhya Pradesh, India
25 Table 7. Cropping pattern in two villages of Raisen District, Madhya
Pradesh, India
27 'Table 8. Comparison of sole cropping of wheat and chickpeas with
intercropping, Begumgunj, Madhya Pradesh, India
28 Table 9. Approximate cost of cultivation for partially mechanized dryland
wheat, Khuria, Sagar District, Madhya Pradesh, India, 1990
1
28 Table 10. Approximate cost of cultivation for non-mechanized dryland durum
wheat, Arnej, Ahmedabad District, Gujarat, India, 1990
29 Table 11. Percentage of dryland area sown to various types of wheat 30 Table 12. Classification of wheat varieties released in India, 1966-91 31 Fable 13. Yield and associated characteristics of local and improved wheat
varieties, Indore, India, 1971-76
39 Table 14. Total wheat production, estimated consumption, and percentage
self-sufficiency, central and southern India, 1984-86
40 Table 15. Rural and urban per capita wheat consumption in central and
southern India compared to Punjab and all India, 1977-78
45 Table 16. Summary of congruency analysis of allocation of wheat research
resources by zone, India
50 Table Al. Estimated quantity of fertilizer applied on wheat, major wheat
producing states, India, 1969-88
iv




Illustrations
Page
2 Map 1. States of India and key rainfed wheat production sites
7 Map 2. Approximate distribution of irrigated and rainfed wheat in India
17 Map 3. Area of black soils in relation to rainfed wheat area, India
5 Figure 1. Trends in rainfed and irrigated wheat area in India, 1951-88
6 Figure 2. Rainfed wheat area by state, India, 1986-88
6 Figure 3. Rainfed wheat area as a percentage of total wheat area within states, India, 1986-88
9 Figure 4. Wheat area as a percentage of gross cropped area, India, 1961-88 10 Figure 5. Cropping intensity and trends for districts grouped by
percentage of wheat area that is rainfed, India
12 Figure 6. Trends in yields of irrigated and rainfed wheat in India and
annual growth rates (g), 1968-86
13 Figure 7. Average yields of rainfed and irrigated wheat in 10 states, India,
1981-85
14 Figure 8. Yields of rainfed and irrigated wheat, Madhya Pradesh and
Punjab, India, 1969-88
15 Figure 9. Share of rainfed wheat in total wheat area and production, India 18 Figure 10. Cumulative probability of rainfall before the growing season
(stored moisture) in four rainfed sites, India.
18 Figure 11. Cumulative probability of rainfall during the growing season in four rainfed sites, India
19 Figure 12. Mean monthly rainfall and evapotranspiration, Sagar, Madhya
Pradesh, India
20 Figure 13. Relationship between rainfall and rainfed wheat area and yields,
Ahmedabad District, Gujarat, India, 1978-88
21 Figure 14. Relationship between rainfall and rainfed wheat area and yields,
Sagar District, Madhya Pradesh, India, 1955-86
v




Page
22 Figure 15. Average temperatures during the wheat growing season at
four sites, India
23 Figure 16. Effect of soil temperature and moisture on wheat germination,
Indore, India, 1986-87 (mean of 10 varieties) 24 Figure 17. Simulated wheat yields at Sagar, India 31 Figure 18. Percentage of total wheat area under modern varieties in
districts grouped by size of rainfed area, India, 1966-83 33 Figure 19. Estimated fertilizer use on wheat in India, 1969-90 34 Figure 20. Relationship between fertilizer use in the rabi (winter) cycle
and irrigation in districts of Madhya Pradesh, India,
1983-84
36 Figure 21. Number of irrigations given to wheat in Uttar Pradesh, India 37 Figure 22. Real farm gate wheat prices in districts grouped by irrigation
status
38 Figure 23. Ratio of farm harvest prices for wheat in Madhya Pradesh and
Karnataka to prices in Punjab, India
38 Figure 24. Ratio of wholesale prices of bread wheat in Sagar and durumn
wheat in central and southern India to the price of bread
wheat in Ludhiana, Punjab
41 Figure 25. Relationship between wheat consumption and income, rural
areas, central and southern India, 1977-78
41 Figure 26. Share of different cereals in total cereal consumption, rural
Maharashtra, India, 1977-78
42 Figure 27. Relationship between state agricultural income, level of
poverty, and earnings of rural labor households
43 Figure 28. Real agricultural wages in districts grouped by the percentage
of wheat area irrigated
43 Figure 29. Real wages and their trends in districts grouped by the
percentage of wheat area that is rainfed, India
vi




Acknowledgements
This report draws heavily on data and information provided by many individuals in India. We particularly thank Dr. R.S. Paroda, formerly Deputy Director General (Crops), Indian Council of Agricultural Research (ICAR), and Dr. J.P.Tandon, Director, All India Coordinated Wheat Project, Indian Agricultural Research Institute (IARI), for arranging visits to research stations in Central and Southern India. In addition the following wheat researchers and their teams were extremely generous with information, time, and hospitality:
Dr. A.K. Singh, Head, Regional Wheat Research Center, IARI, Indore,
Madhya Pradesh.
The late Dr. R.P. Sheopuria, Head, Wheat Research Program, Regional
Research Station, Powerkheda, Madhya Pradesh.
Dr. V.S. Tomar, Head, Regional Research Station, Sagar, Madhya Pradesh.
Dr. R.R. Hanchinal, Head, Wheat Research Program, University of
Agricultural Sciences, Dharwad, Karnataka.
Dr. B.S. Jadon, Senior Scientist, GAV Wheat Research Station, Vijapur,
Gujarat.
Numerous other individuals at these research centers also provided valuable information and assistance.
Several colleagues at CIMMYT contributed to the development of this paper as well. I would especially like to thank Daphne Taylor for research assistance in analyzing agroclimatic and production statistics. I also appreciate the help of Mark Bell for valuable discussions on soils and moisture conservation in dryland wheat, as well as for generating the computer simulations for Sagar. The assistance of Hector Tovar and Laura Saad in drawing the maps and diagrams, Maria Luisa Rodriguez for patiently typing the report, and Jos6 Luis Delgado in doing the layout is gratefully acknowledged. Finally, careful reviews of the paper by Gunvant Desai, Tony Fischer, John Foster, R.R. Hanchinal, and Miguel L6pez-Pereira were most helpful in completing this report. Any remaining errors are of course the author's.
vii







Dryland Wheat in India: The Impact of Technical Change and Future Research Challenges
Derek Byerlee
in collaboration with the Indian Council of Agricultural Research
Introduction
Both CIMMYT and national agricultural research systems (NARSs) have been called upon to give more attention to marginal areas in allocating research resources. This emphasis appears to originate from four concerns:
" that marginal areas have largely been bypassed by the technological
breakthroughs of the past three decades, especially the Green Revolution;
" that a large number of people depend on marginal areas for their survival
and that increasing population pressure is forcing more people into marginal
areas;
" that the people who depend on agriculture in marginal areas are among the
poorest groups of the population; and
" that resource degradation, especially soil erosion, is most serious in marginal
areas.
This paper is part of a larger effort by CIMMYT in collaboration with NARSs to 1) define the extent of marginal maize and wheat areas with greater precision, 2) characterize the agroecological and socioeconomic conditions under which maize and wheat are grown in these areas, and 3) identify research opportunities for these areas. The paper focuses on the dryland wheat areas of central and southern India in the states of Madhya Pradesh, Gujarat, Maharashtra, and Karnataka (see Map 1), which by almost any standards must be classified as marginal.1 Average farm-level wheat yields are only about 0.7 t/ha, and even on research stations yields average less than 2 t/ ha against a yield potential (under irrigation) of over 5 t/ha. Furthermore, India, along with Turkey, Iran, and China, has the largest concentration of dryland wheat area in the developing world. Finally, dryland wheat in India is largely produced using residual soil moisture; thus these dryland wheat areas represent a unique drought environment within CIMMYT's general definition of mega-environments (Fischer and Varughese 1990).
This study focuses on the major part (over half) of the total rainfed wheat area in India. Extensive areas are sown to rainfed wheat in northern India as well, although moisture conditions for wheat production in these areas are relatively
1 CIMMYr has classified as "marginal" those areas where experimental yields are normally less than
40% of potential yields defined by available solar radiation and temperature (CIMMYT 1989a).
1




more favorable. Hence this paper uses the terms "rainfed" or "unirrigated" to refer to all rainfed areas in India, whereas the term "dryland" refers specifically to wheat production in Central and Southern India (i.e., a subset of rainfed wheat areas).
To provide a perspective on dryland wheat's significance within the wheat economy of India, the paper reviews statistics on wheat area and yield, disaggregated by irrigation status and state. This review is followed by an
AFGHANISTAN
JAMMU
AND
KASHMIR CHINA
". ..... TIBET
HIMACHAL
PRADESH
PAKISTAN PUNJAB .
--.,.. -- ,'BHUTAN
HARYANA
.0: SKIM
DELHI ARUNACHAL PRADESH
..NEPAL UTTA R ....
RAIASTHAN ... PRADESH ASSAM
...... .... 'NAG .ALAND
".**..51.'*..*Patna MEGHALAYA iMANIPUR
GUJARAT *~ BIHAR TRIPURA
Are ^" .' Sagar IN IA--::N MIZORAM
Arnej INDIA BENGAL
'... MADHYA PRADESH :...
... .--- . .. el BURMA
BANGLADESH
MAHARASHTRA ORISSA
Arabian ..,
Sea '
.-..... ANDHRA
Segaum,, PRADESH GOA Annigen Bay of Bengal
KARNATAKA
KERALA
Indian Ocean
Map 1. States of India and key rainfed wheat production sites.
2




examination of the agroclimatic conditions under which wheat is produced in the dryland areas of central and southern India. The paper then goes on to examine dryland wheat in some depth to answer three major questions:
1) Has the Green Revolution in wheat in India bypassed the dryland areas
or, more specifically, what has been the impact of technological change in
dryland areas over the past three decades?
2) Should more research resources be allocated to addressing the special
challenges of dryland wheat?
3) If so, what are appropriate research priorities for achieving more rapid gains
in productivity in the future?
This report draws on three major sources of information. The first source consists of official statistics on rainfed and irrigated wheat area and yields by state, published by the Directorate of Economics and Statistics, Ministry of Agriculture, New Delhi. These official statistics were supplemented by districtlevel statistics on rainfed wheat area, yields, wage rates, and other parameters, assembled by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) and others, for eight states in India.2 Second, with the collaboration of the Indian Council of Agricultural Research (ICAR), the major research stations in the dryland wheat belt of central and southern India were visited to observe wheat research and production firsthand and to collate the extensive research that has been conducted by Indian wheat scientists at these stations (see Acknowledgements). Finally, published research reports and articles on dryland wheat in India were extensively reviewed. Together, these various sources of information provide a reasonably complete picture of the current status and future prospects for dryland wheat in India.
Trends in Rainfed Wheat in India
This section draws on official wheat production statistics to examine overall trends in all rained wheat in India. Although complete statistics for total wheat area and production were available for 1961 through 1990, production statistics disaggregated by irrigation status were incomplete. Hence the period of analysis varies depending on the data available, presenting some difficulties in making comparisons across tables.
Trends in Wheat Area
Trends in wheat area by state-Total wheat area in India has expanded by
2.5-3.0% annually since the early 1960s, increasing from about 13 million hectares to over 23 million hectares in the 1980s. The area sown to wheat grew
2 District-level statistics were available for the state- of Haryana, Karnataka, Madhya Pradesh,
Maharashtra, Punjab, Gujarat, Rajasthan, and Uttar Pradesh.
3




much faster than the area sown to any other food grain. 3 Wheat area expanded most rapidly in the northeastern states of Assam, Bihar, Orissa, and West Bengal, which currently make up 10% of India's total wheat area (Table 1)." The states in the traditional wheat belt of northwestern India (Uttar Pradesh, Punjab, and Haryana) also recorded above average growth in wheat area. The slowest expansion occurred in central and southern India (the region that is the subject of this paper), where wheat area increased at about only 1.0% annually.
The past three decades have seen a dramatic shift in the amount of wheat grown on irrigated land compared to rainfed land. The area of irrigated wheat expanded at 5.9% per year from 1961 to 1988, from just over 4 million hectares to over 17 million hectares (Figure 1). During the same period, rainfed wheat area declined steadily from 8.8 to 6 million hectares as irrigation systems steadily expanded. The share of rainfed area fell from 67% of all wheat area in 1961 to
Table 1. Average wheat area, 1988-90, and growth rates in total wheat area by state, 1961-87
Average Annual growth
wheat in wheat
area, 1988-90 area, 1961-87
Area State (000 ha) M%
All India 23,385 2.4
Northwest Himachal Pradesh 377 2.5
Jammu and Kashmir 241 1.0
Punjab/Haryana 4,981 4.4
Rajasthan 1,651 2.1
Uttar Pradesh 8,585 3.1
Northeast Assam 99 14.0
Bihar 2,005 4.2
Orissa 42 7.2
West Bengal 334 7.9
Central and South Gujarat 48 7 0.7 (ns)
Karnataka 249 -0.3 (ns)
Mad hya Pradesh 3,454 0.7
Maharashtra 818 0.2 (ns)
Note: All trends are significant at the 1%c percent level except where noted as not significant (ns).
3 The annual growth rates in area sown to individual food grains in India for 1960-89 are: rice, 0.63%o;
maize, 0.84%; wheat, 2.52%o; and sorghum, -0.61 %.
4 The high rates of growth in wheat area in AMsam, Orissa, and West Bengal are the result of the very
low base area of wheat. Wheat was not a traditional crop in these areas until the availability of early
maturing varieties, which allowed wheat to be grown after the main crop, rice.
4




just 25% in 1986. The common statement that crop area is expanding into marginal areas is certainly not valid for wheat in India, where favored (irrigated) areas have steadily become more important relative to marginal areas in the wheat economy.
Four states dominate rainfed wheat production: Madhya Pradesh, Uttar Pradesh, Maharashtra, and Karnataka. The largest area of rainfed wheat is in Madhya Pradesh (2.2 million hectares), which accounted for 42% of the rainfed wheat area in 1986-88 (Figure 2). The next largest concentration of rainfed wheat area is in Uttar Pradesh (21% of the rainfed wheat area), followed by Karnataka and Maharashtra in southern India (together these states account for 10% of the total rainfed wheat area). In 6 of the 13 wheat growing states, more than half of the total wheat area was rainfed in 1986-88 (Figure 3).
The distribution of rainfed and irrigated wheat based on district-level data clearly indicates the concentration of rainfed wheat in northern Madhya Pradesh, where Sagar and Vidisha Districts each report sowing over 200,000 ha of wheat, virtually all of it under dryland conditions (Map 2). Districts adjoining Madhya Pradesh to the north, such as Kota and Tonk in Rajasthan and Hamirpur and Banda in Uttar Pradesh, are also important producers of rainfed wheat.
Million ha
20
Irrigated area
15
10
Rainfed area
5
0
1952 1956 1960 1964 1968 1972 1976 1980 1984 1988
Figure 1. Trends in rainfed and irrigated wheat area in India,1951-88.
5 Statistics on rainfed and irrigated area are taken from the Government of India, Area and Production of
Principal Crops in India (various years). The release of statistics on rainfed and irrigated area is delayed
by two to three years, and hence the series used here extends only to 1988.
5




Rainfed wheat cultivation extends south and west from northern Madhya Pradesh but is much less dominant in the cropping systems. Throughout central and southern India, rainfed wheat is grown on deep black vertisols (see the section on "Soils," later in this paper, and Map 3). Rainfed wheat in northern India is grown on alluvial soils but rainfall is generally much higher and temperatures are cooler.
Maharashtra
6.8% Karnataka Orissa 0.1%
3.1% 1Assam 1.6% Uttar Pradesh West Bengal 1.4%
22.6% PdGujarat 1.9%
Jammu and Kashmir 2.8% Punjab 2.7%
Haryana 1.3%
Rajasthan 4.5%
Himachal Pradesh 5.2%
Madhya Pradesh Bihar 6.8%
39.2%
Figure 2. Rainfed wheat area by state, India, 1986-88.
Haryana 3
Punjab 5
Orissa 7
Rajasthan 10
Uttar Pradesh 13
Bihar 17
Gujarat 27
West Bengal 37
Maharashtra 42
Madhya Pradesh 77777777= 61
Karnataka 74
Jammu and Kashmir 76
Himachal Pradesh 83
Assam_________All India 22
0% 20% 40% 60% 80% 100%
Rainfed wheat area as a percentage of total wheat area Figure 3. Rainfed wheat area as a percentage of total wheat area within states, India, 1986-88.
6




Approximately 2.7 million hectares of rainfed wheat are subject to dryland conditions in central and southern India (including neighboring districts of Uttar Pradesh and Rajasthan). About the same area is sown to wheat in rainfed areas of the northwestern and northeastern states, but moisture conditions are more favorable. Several districts in northeastern Uttar Pradesh produce over 75,000 ha of rainfed wheat under these more favorable conditions.
Although rainfed wheat area still accounts for nearly one-quarter of India's total wheat growing area, much rainfed area has been converted into irrigated area. Rainfed area has decreased in all states except Assam.6 Irrigated area has expanded most rapidly in central and southern India and in the state of Bihar (Table 2). In Madhya Pradesh (the major producer of dryland wheat), only 6% of the wheat area was irrigated in 1960 compared to over 40% in 1990.
A = 40,000 ha rainfed wheat
.... A = 40,000 ha irrigated wheat
** : .. ..*
A A
" ~s so ..... ......'.. .,A A
Map 2. Approximate distribution of irrigated and rainfed wheat in India.
6 In particular, the share of the northern hill areas in rainfed wheat production has jumped from 3% in
1961-65 to 8% in 1986-88
7




Components of area increases-The growth rate of wheat area, ra, can be decomposed into three components:
r =r +r.+r, (1)
where:
rc= growth rate of cultivated area;
r= growth rate of cropping intensity; and
r= growth rate of wheat's share of gross cropped area (i.e., the sum of
area of all crops grown in a year).
The results of this decomposition for the years 1961-88 are given in Table 3 for rainfed, irrigated, and total wheat area. In irrigated areas, the two major components of the rapid increase in wheat area were the expansion of irrigated area (2.28% per annum) and the increase in wheat's share of irrigated cropped area (3.10% per antum). This share more than doubled from 1961-65 (about 15%) to 1986-88 (33%) (Figure 4). Meanwhile, rainfed wheat area declined at a rate of -1.69% per annum for two reasons: irrigation facilities reduced the amount of
Table 2. Average rainfed and irrigated wheat area, 1984-86, and growth rates in rainfed and irrigated wheat area by state, 1961-88
Rainfed wheat Irrigated wheat
Growth rate Growth rate
Area in in area, Area in in area,
1986-88 1961-88 1986-88 1961-88
States (000 ha) (%) (000 ha) (%)
All India 5,218 -1.8*** 17,693 5.9 ***
Himachal Pradesh 311 -2.2 64 3.8 ***
Jammu and Kashmir 177 0.9 (ns) 57 4.9 ***
Punjab/Haryana 221 -3.0*** 4,659 5.9
Rajasthan 177 -3.5*** 1,507 4:2
Uttar Pradesh 1,107 -2.8** 7,235 5.9
Assam 104 15.3*** 0 0.0
Bihar 330 -1.7*** 1,540 9.6
Orissa 3 na 46 na
West Bengal 138 na 221 na
Gujarat 83 -2.3 230 1.8 **
Karnataka 200 -1.2*** 69 8.9 ***
Madhva Pradesh 2,188 -0.9** 1,398 8.4
Maharashtra 335 2.7** 448 5.6 ***
Note: *, **, *** denote significance at the 10%, 5%, and 1 levels, respectively; na = not available; ns = not significant.
8




rainfed area (-0.56%), and wheat's share of the remaining rainfed cropped area became smaller (-1.52%). Wheat accounts for only 5% of the rainfed cropped area (Figure 4). Increasingly, rainfed wheat area is concentrated in central and southern India, where wheat is a less important crop.
Cropping intensity in both rainfed and irrigated areas has steadily increased (at 0.39% per year in rainfed areas and 0.51% per year in irrigated areas) (Figure 5). However, increased cropping intensity is not an important component of the expansion in wheat area, compared to the growth in irrigated area and the greater importance of wheat in irrigated cropping patterns.
Table 3. Compound annual growth rates in components of wheat area in India (%/yr), 1961-88
Rainfed Irrigated Total
wheat area wheat area wheat area
Growth in cultivated area -0.56 2.28 0.14
Growth in cropping intensity 0.39 0.51 0.43
Growth in wheat's share of gross cropped area" -1.52 3.10 2.06
Growth in wheat area -1.69 5.89 2.63
Note: All trends are significant at the 1% level. a Gross cropped area is the sum of area of all crops grown in a given year.
Wheat as percentage of gross cropped area
35%,
30%
25%
20%
15%
10%
Rainfed land
5%
0%
1960 1964 1968 1972 1976 1980 1984 1988
Figure 4. Wheat area as percentage of gross cropped area, India, 1961-88.
9




Trends in wheat area: 1976-86 vs. the Green Revolution decade-It may be hypothesized that irrigated wheat area grew most rapidly in the Green Revolution period and has slowed in recent years. To test this hypothesis, a spline function of the following form was used:
In (Y) = b. + blw, + b2w2' (2)
where:
In (Yt) = logarithm of wheat area in year t;
wi = year t, and 0 if t < 1976; and
w 2 = t-1976 if t > 1976
The coefficient b2 tests for significant differences in growth in irrigated wheat area between 1976 and 1988 (defined here as the post-Green Revolution period) compared to the period from 1965 to 1976 (the Green Revolution period). The coefficient is negative if growth has slowed.
The results (Table 4) support the hypothesis that in the post-Green Revolution period irrigated wheat area has increased at a significantly slower rate across most states. In the eastern states of Bihar and West Bengal, growth in irrigated
Cropping intensity, 1983 Growth rate of cropping intensity, 1961-83 (%/yr)
132 .73
124
112 112 115
.47 .4
.34
.25
....... I I"1I1/1/
80-100% 60-80% 40-60% 200% 0-20% 80-100% 60-80% 40-60% 20-40% 0-20%
Percentage of total wheat area that is rainfed Percentage of total wheat area that is rainfed
Figure 5. Cropping intensity and trends for districts grouped by percentage wheat area that is rainfed, India.
10




wheat area for 1976-88 appears to have slowed sharply to less than half the rate for 1965-76. Among the components of increased wheat area, cropping intensity is the only one whose growth did not decline in 1976-88.
Given these trends, and the fact that much wheat area is already irrigated and construction of new irrigation facilities has slowed, irrigated wheat area in India is likely to grow much more slowly in the future than in the past. This means that, in rainfed as well as irrigated areas, higher yields will be the main source of production increases.
Trends in Wheat Yields
Although data exist on yields of rainfed and irrigated wheat by state (Government of India, various issues), the data are incomplete. In some cases there are significant differences between the official average yield statistic for the state and the derived yield (i.e., weighted average yield based on rainfed and irrigated area and yields).7 For all India, however, there is a very high correlation (r=.98) between the derived yields and the official statistics.
Table 4. Trend growth rate (%/yr) in irrigated wheat area by state, India, 1965-76 and 1976-88
Period
1965-76 1976-88
All India 8.4 2.8 *
Himachal Pradesh 4.4 0.4 *
Jammu and Kashmir 10.7 1.3 ***
Punjab/Haryana 8.6 2.5
Rajasthan 7.1 2.0
Uttar Pradesh 8.0 3.3
Assam na na
Bihar 14.7 1.7 ***
Orissa na na
West Bengal na na
Gujarat 5.7 -4.4 *
Karnataka 17.7 0.7 ***
Madhva Pradesh 13.2 4.4
Maharashtra 10.9 0.3 ***
Note: *, **, *** indicate that the trend growth rate in the second period is significantly different to that in the first period at the 10%, 5%, and 1% levels, respectively; na = not available.
7 Disaggregated data on irrigated and rainfed wheat yields are released several years after they are
gathered, so the latest data available for many states are from 1985 or 1986.
11




Figure 6 illustrates yield trends in irrigated and rainfed areas in India from 1968 to 1985, the last year for which data are complete for all the major wheat producing states. For all of India, the estimated trend yield in 1985 was 2.2 t/ha in irrigated areas compared to 0.9 t/ ha in rainfed areas. From 1972 to 1985, wheat yields grew at an annual rate of 2.8% in irrigated areas compared to 1.4% in rainfed areas (Table 5). Over the same period, the growth in wheat yields for the entire country 3.1% was higher than growth in yields of either irrigated or rainfed wheat, because of the increasing share of wheat area under irrigation.
Across states, average yields of rainfed wheat are highest in Punjab and Haryana (1.5 t/ha) and lowest in the drier, hotter central and southern states of Madhya Pradesh, Maharashtra, Gujarat, and Karnataka (0.4-0.7 t/ha) (Figure 7). The relative difference between yields of rainfed and irrigated wheat is lowest in the eastern states of Uttar Pradesh, Bihar, and West Bengal. In these states, the ratio of rainfed to irrigated wheat yields is 0.6-0.7, compared to 0.5 in Maharashtra and Madhya Pradesh and only 0.2 in Gujarat. Over time the ratio of yields of rainfed wheat to irrigated wheat for all of India has widened from an estimated
0.5 in 1960 to 0.38 at present.'
In most states, yields of irrigated wheat grew more rapidly than yields of rainfed wheat. The exceptions are Punjab and Jammu and Kashmir, where the remaining rainfed wheat area is in the submountain higher rainfall areas
Yield (t/ha)
2.5
2 Irrigated
g=2.8% ....
1.5
Rainfed
1965 1970 1975 1980 1985 1990
Figure 6. Trends in yields of irrigated and rainfed wheat in India and annual growth rates (g), 1968-86.
8 These figures are based on projections of yields and production of rainfed wheat at the historic growth
rate.




Table 5. Annual growth rate (%/yr) of irrigated and rainfed wheat yields by state, India, 1972-85
Growth in irrigated Growth in rainfed
wheat yield wheat yield
All India 2.8 1.4
Himachal Pradesh 1.3 (ns) 0.8 (ns)
Jammu and Kashmir 0.8 (ns) 3.3
Punjab 2.2 3.1
Haryana 2.6 1.4 (ns)
Rajasthan 2.8 2.1
Uttar Pradesh 3.1 2.4
Assam na 1.2 (ns)
Bihar 3.0 2.0
West Bengal 4.4 3.0
Gujarat 3.0 -0.1 (ns)
Karnataka 4.8 3.0
Madhya Pradesh 2.5 1.6
Maharashtra 2.2 1.6 (ns)
Note: The actual years for which data are available vary by state. For major wheat producing
states, complete data are available for 1972-85. All trends are significant at the 10% level or
less, except where denoted by "ns" (not significant); na = not available.
Punjab "
Haryana
Rajasthan Irrigated
Uttar Pradesh '- / 7 Rairrfed
Bihar
West Bengal ./
Madhya Pradesh ./
Maharashtra /
Gujarat
Kamataka //X-"/
0 1 2 3 4
Yield (t/ha)
Figure 7. Average yields of rainfed and irrigated wheat in 10 states, India, 1981-85.
13




(Table 5). Growth in rainfed wheat yields was nevertheless positive and significant in almost all states except Haryana, Gujarat, and Himachal Pradesh.9 The performance of rainfed and irrigated wheat in contrasting environments is illustrated in Figure 8 for two states that have complete yield data. The Punjab registered the most rapid gains in yields of irrigated wheat. Yields of rainfed wheat also increased steadily under the relatively favorable environmental conditions for rainfed wheat in the Punjab. In fact, both absolute yield levels as
Yield (kg/ha) 4,000
3,500 Madhya Pradesh
3,000 2,500
Irrigated 2,000 1,500 1,000
500 -Rainfed
0 -I
1969 1973 1977 1981 1985 1989
Yield (kg/ha)
4,000
3,500 '
3,000
2,500
2,000 1,500 1,000
500
0
1969 1973 19*77 1981 1985 1989
Figure 8. Yields of rainfed and irrigated wheat, Madhya Pradesh and Punjab, India, 1969-88.
9 Growth rates of yields of rainfed wheat at the state level are generally higher than at the national level
because in recent years rainfed wheat production has become more concentrated in the drier lower
yielding areas of central and southern India, which depresses the overall average growth rate.
14




well as relative changes in rainfed wheat yields in the Punjab are remarkably similar to those for yields of irrigated wheat in Madhya Pradesh (Figure 8), a dry environment where the water supply for most irrigated wheat is fairly limited.
An unexpected finding is that variability in rainfed wheat yields is quite low (coefficient of variation around trend of 6.8% for all India) and is in fact less than yield variability in irrigated areas (coefficient of variation around trend of 8.1% for all India). This difference in yield variability occurs in part because a high share of rainfed wheat is grown under residual moisture; if farmers see that moisture conditions are inadequate, they reduce their wheat area. Hence rainfed area is often more variable than yields (see the section on "Rainfall," below). In central and southern India, limited supplies of irrigation water increase yield variability in socalled irrigated areas.
The Declining Importance of Rainfed Wheat Production Reductions in the area sown to wheat under rainfed conditions, and the fact that yields in irrigated areas have grown twice as fast as yields in rainfed areas, imply that the share of rainfed wheat production has fallen drastically. Rainfed areas accounted for 23% of the wheat area in 1988 but produced only about 11% of the wheat crop. This is a sharp decline from 46% of the wheat area and 27% of production in 1970, and an estimated 67% of the area and 50% of production in 1960 (Figure 9). These trends are significant because the allocation of research resources to rainfed wheat must adjust to the relative importance of rainfed wheat in the overall wheat economy (see "Research Resource Allocation to Dryland Wheat," below).
Share of rainfed wheat to total wheat
80
% Rainfed wheat area
= % Rainfed wheat production
60
40
20
0
1960 1970 1986
Figure 9. Share of rainfed wheat in total wheat area and production, India.
15




Agroclimatic Situation for Rainfed Wheat in India
Three sets of agroclimatic variables soils, rainfall, and temperature interact to influence the distribution and productivity of rainfed wheat in India. Rainfall distribution and soil type determine the extent to which rainfed wheat production depends on residual moisture. Deep black vertisols and good monsoon rainfall enable wheat to survive almost exclusively on residual moisture from the monsoon. In areas with lighter soils, some rainfall during the growing season is required to obtain even a minimum wheat yield. Where wheat depends on residual moisture, there are clear advantages in planting early to best exploit the end of the monsoon rains. However, high temperatures in September and October limit the extent that the planting date can be moved forward. To clarify these interactions, the following sections summarize key agroclimatic factors, with emphasis on the implications for dryland wheat in central and southern India.
Soils
The dryland wheat belt of central and southern India extends through the zone of black vertisols or cotton soils, which covers 546,000 km2 (Map 3). The depth of these soils varies by as much as 2 m; their depth, together with their high clay content of 40-60%, gives them a high moisture holding capacity. These soils are generally low in organic matter and on sloping land are subject to serious erosion and degradation (Put and van Dijk 1989).
The black vertisols have two key properties relevant to crop production. First, because of their high clay content, they are exceedingly sticky when wet and on drying develop large, deep cracks. When very dry or very wet, these soils are difficult or impossible to till, and at the onset of the monsoon there is a very short "window of opportunity" when moisture conditions permit land preparation and planting (Virmani, Rao, and Srivastava 1985).
Second, the vertisols can retain a great deal of moisture. In the deep vertisols in the higher rainfall areas of northern Madhya Pradesh, 400 mm of moisture or more may be stored in the soil profile at the end of the monsoon, at least half of it available for winter crops such as wheat. However, since vertisols are impermeable when their water content is high, they are subject to waterlogging in high rainfall areas. This adds to the difficulty of managing kharif (summer) crops and favors a fallow-wheat rotation.
To the north, in the alluvial soils of the Gangetic Plain and the Terai (lowlands), rainfed wheat must depend less on stored moisture and requires significant growing season rainfall. To the south, in Maharashtra and Karnataka, high winter temperatures favor sorghum as the winter cereal crop in rainfed areas.
Rainfall
Any analysis of the quantity and variability of rainfall must distinguish between moisture stored in the soil and growing season rainfall. The level of stored
16




moisture is determined by many factors, including rainfall, evaporation, previous crop or fallow, and soil type. An analysis of those factors is not attempted here; rather cumulative probabilities of rainfall in September and October are plotted for three key sites where dryland wheat is grown (Sagar in Madhya Pradesh, Arnej in Gujarat, and Annigeri in Karnataka) and compared to a rainfed site in northeastern India (Patna) (Figure 10). At Sagar, in the largest dryland wheat growing district in India, it is clear that rainfall is highly variable prior to sowing, even though average annual rainfall is 1,237 mm, over 90% of which falls in the monsoon. Even at Sagar there is just slightly more than a 50% probability of receiving 200 mm of rainfall in the pre-sowing period. South and west of Sagar, monsoon rainfall decreases and becomes more unreliable. At Arnej in Gujarat, the chance of receiving 200 mm of rainfall in SeptemberOctober is only 20%. In contrast, at Patna in the northeast there is a 66% chance of receiving 200 mm of rainfall prior to sowing.
A __ = 40,000 ha rainfed wheat
A.'%A A,.."A
A A ?. .... ...
A A..
Map 3. Area of black soils in relation to rainfed wheat area, India.
17




Growing season rainfall is even more erratic but shows similar patterns (Figure 11). At Sagar there is less than a 20% chance of receiving 100 mm or more of rainfall during the crop cycle and a 50% chance of receiving 50 mm or less. In Gujarat, there is practically no chance of receiving any significant rainfall in the growing season.
Cumulative probability (%)
100- - -
A.s.'nnigen --80
800 Arne's g" SsS. [oI "" 'r ....
A60j9{,.e Sagar
tX
40 Patna
40 /
20
0 200 400 600 800
Moisture (mm)
Figure 10. Cumulative probability of rainfall before the growing season (stored moisture) in four rainfed sites, India. Note: Stored moisture was calculated for Amej and Sagar (100% rainfall in September plus 75%
rainfall in October); Annigeri (100% rainfall in September); and Patna (100% rainfall for
September and October).
Cumulative probability (%)
100 Arnej -" ,J 0eri
/'Patna A e
80 f .
i ,;
60- .1 Sagar
40
20
0'
0 200 400 600 800
Moisture (mm)
Figure 11. Cumulative probability of rainfall during the growing season in four rainfed sites, India. Note: Growing season moisture was calculated for Arnej and Sagar (25% rainfall in September
plus rainfall through to February); Annigeri (October through to January); and Patna
(November through to March).
18




At all sites, stored moisture is critical for the survival of the wheat crop. Figure 12 shows mean monthly rainfall in relation to potential evapotranspiration at Sagar. Wheat is planted just at the time that precipitation falls below evapotranspiration, so that during the entire growing season, evapotranspiration usually exceeds the amount of rainfall.
These different rainfall patterns influence cropping patterns. In Annigeri (a wetter site), double cropping is common in most years. At Sagar the preferred rotation is fallow-wheat, because of the problems of managing the heavy black soils in the monsoon season and because the moisture for double cropping is usually insufficient. In Gujarat, in many years there is not enough moisture for even a single crop.
In addition, the relationship between 1) the amount of rainfall and 2) rainfed area and yield is quite different at a dry site, such as the district of Ahmedabad in Gujarat, and a wetter one, such as the district of Sagar in Madhya Pradesh (Figures 13 and 14). The monsoon is so uncertain in Gujarat that farmers first decide whether to plant wheat or leave the land fallow, depending on how much moisture is available at planting time. Between 1978 and 1988, there was almost a linear relationship between area sown to wheat and the amount of monsoon rainfall (Figure 13). The exception is the very wet year of 1981, when 1,000 mm were recorded. After the unusually dry monsoon of 1987, practically no wheat was sown. Over the same period, rainfed yields also tended to be quite variable (Figure 13).
mm
450
Rainfall
400
350
300 -Wheat
250 -harvest
200 -patn
150
100
50
0~
May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr
Figure 12. Mean monthly rainfall and evapotranspiration, Sagar, Madhya Pradesh, India.
19




In Sagar where rainfall is higher, both wheat area and yield are relatively more stable (Figure 14).1" Only a weak relationship exists between September-October rainfall and wheat area sown (r = 0.30); monsoon rainfall is almost always sufficient to plant a crop of wheat and the coefficient of variation (CV) of rainfed
Yield (t/ha)
1,000
900 r = 0.71
cv = 33%
800 M
700
600 M
500 400 300
200 M
100
0
0 200 400 600 800 1,000 1,200
Monsoon rainfall plus growing season rainfall (mm) Area (000 ha)
1.0
0.9. r = 0.95
cv = 51%
0.8.
0.7.
U U
0.6. 1981
0.5
0.4
0.3
U
0.2
0.1 1987
0.0 .
0 200 400 600 800 1,000 1,200
Monsoon rainfall (mm) Figure 13. Relationship between rainfall and rainfed wheat area and yields, Ahmedabad District, Gujarat, India, 1978-88.
10 Wheat area and yields in Sagar show significant upward trends. Hence data in Figure 14 are expressed
in deviations from the long-term linear trend.
20




wheat area is only 8%, compared to 51% in Ahmedabad District. Although yields are significantly and positively correlated with total rainfall from September to February (r = 0.66), yields are also quite stable with a CV of 15.5%.
Temperatures The final important variable in wheat production is temperature. Practically the whole dryland wheat area of central and southern India lies in the warmer tropical zone (defined by CIMMYT as including areas where mean temperatures
Deviation of wheat yield from long-term trend (t/ha)
0.3
r = 0.66
cv = 15.5%
0.2
o.1 .............ra...............
U U p .......
..................
-0.1........
-0.2
-0.3
0 200 400 600 800
Rainfall, -eptember-February (mm)
Deviation of wheat area from long-term trend (t/ha)
40
r = 0.30
30 A A A A cv = 7.7%
20
10 A A A ................... .
. A ....... ...... .............
0- A A .. AA.........
0 ................. ...............
-10 A
-20 A
-30 AA
-40
0 200 400 600
Rainfall, September-October (mm)
Figure 14. Relationship between rainfall and rainfed wheat area and yields, Sagar District, Madhya Pradesh, India, 1955-86.
21




for January exceed 17.5'C). In the extreme north of the dryland wheat belt (in northern Madhya Pradesh), winter temperatures average a little below 17.5'C and there is some risk of frost; this favors the cultivation of wheat rather than sorghum. To the south, temperatures increase to a mean January temperature of 23-24")C in Karnataka (Figure 15), which along with Sudan is the warmest area in the world where wheat is grown commercially.
The most critical relationship between factors influencing dryland wheat production in central and southern India is depicted in Figure 16. At the end of the monsoon in September, a high level of moisture is stored in the soil (up to 357o), and wheat planted early has the potential to use this moisture. However, soil temperature is also high and can impede germination and seedling emergence. Over time, both the soil moisture level and temperature decline. At later planting dates, germination/ emergence rates may increase because of lower soil temperatures (although this may be counteracted by inadequate moisture). However, the low probability of growing season rainfall means that late planted wheat is likely to develop in a soil moisture profile that falls steadily from an average of 25% at planting to 13% in February. Because the dryland wheat crop is nearly always forced to maturity by moisture stress in late February, it avoids extreme heat stress at flowering and grain filling (except in the south, where the entire growing season is relatively hot) and the more critical period of heat stress occurs at the seedling stage.
Temperature ('C)
30
28
26 -,
Own. .. An ij
22-maa
20- aa
18Pan
16
Oct Nov Dec Jan Feb Mar
Figure 15. Average temperatures during the wheat growing season at four sites, India.
22




Simulation of Wheat Yields, Sagar District The EPIC model (Williams et al. 1990, Sharpley and Williams 1990) was used to simulate crop growth and yields over 50 years based on soil and climatic conditions at Sagar, Madhya Pradesh (M. Bell, pers. com.). The simulation confirms most of the findings reported above. With only moisture as the limiting factor (i.e., soil fertility, diseases, and other factors are held at non-limiting levels), average yields were 2.2 t/ha with a CV of 25%. In other words, yield variability is considerably lower than in other marginal rainfed areas (Belaid and Morris 1990), reflecting the extreme importance of residual moisture relative to growing season rainfall at Sagar. Simulated yields ranged from an absolute minimum of 1.2 t/ha to a maximum of 3.4 t/ ha (Figure 17).
In contrast, when moisture is included as a non-limiting factor (i.e., under full irrigation), potential yields at Sagar are close to 6 t/ha with a CV of 6% (Figure 17).11 Thus, drought stress permits average yields at Sagar to reach only about one-third of their potential a finding consistent with CIMvMT's definition of a marginal environment. Correlation coefficients derived from the dryland simulations are given in Table 6. As expected, yield is highly correlated with cumulative available moisture (CAW). These correlations also indicate that
47 27. 170
46 26- Emergence ..
" 45 -,.x 25. A.. ," 160
"0 44 24. 1 -'
E E 150
m 43- 23" 42. 22S
- .- moisture 140
0 41 2* 2.
40 20 ..1
3 .
C" 39 "19 *. SoiluA
. E
. 38 E 18 temperature 0 120
-12
37 J17.
36 16 110
35 15
8 Oct 15 Oct 20 Oct 25 Oct 30 Oct 5 Nov Planting date
Figure 16. Effect of soil temperature and moisture on wheat germination, Indore, India, 1986-87 (mean of 10 varieties).
Source: IARI, Indore (1990).
11 The variability in yields under irrigation is caused by temperature fluctuations.
23




growing season rainfall can be important, as shown by the high correlation between CAW and growing season rainfall. The low correlation between CAW and September-October rainfall is probably due to variation in soil moisture conditions at the beginning of this period as well as the loss of much monsoon rainfall through runoff.
Cumulative probability
1
0.8
Rainfed
0.6
0.4
Irrigated
0.2
0.1
2 3 4 5 6 7 8
Yield (t/ha)
Figure 17. Simulated wheat yields at Sagar, India.
Note: Yields simulated using EPIC model (Williams et al. 1990, Sharpley and Williams 1990).
Table 6. Correlation matrix for simulated wheat production, Sagar District, Madhya Pradesh, India
Correlation coefficients' Cumulative
available September- Novembermoisture October March
Yield (CAW) rainfall rainfall
Yield 1.00 0.81 0.28 0.63 ***
CAWb .. 1.00 0.26 0.62 ***
September-October rainfall .... 1.00 0.05
November-March rainfall .... 1.00
Note: *, **, ***, indicate significance at the 10%, 5%, and 1% levels, respectively. a Based on 50-year simulation. b Cumulative available moisture.
24




Wheat in the Farming Systems of Central and Southern India
In northern India where rainfall is higher, the farming systems in which rainfed wheat is grown are similar to adjacent irrigated farming systems. In the northeast, wheat may be grown on residual moisture after rice; in the northwestern foothills, wheat is often grown after maize or kharif pulses. However, in the drier areas of central and southern India, dryland farming systems are quite distinct from those in the adjacent irrigated areas. In general, dryland wheat in central and southern India is grown after fallow (to conserve moisture) and single cropping is the norm. The rest of this section briefly describes the main characteristics of these dryland systems.
Wheat in the Cropping Pattern
Farms tend to be larger in the dryland regions of central and southern India than in irrigated areas, although the poorer land quality means that farm incomes are low and many farmers cannot meet their subsistence needs. For example, farm surveys consistently show average farm sizes of 5-7 ha in rainfed Madhya Pradesh (Singh, Jain, and Rao 1984; Singh and Raje 1984; Ryan and Sarin 1981), which is much higher than in irrigated areas. Although farm size is larger, cropping intensity in rainfed areas
is lower, averaging just over 100%, Table 7. Cropping pattern in two indicating that a very small area is villages of Raisen District, Madhya double cropped. Pradesh, India
In the main dryland wheat belt of Percent cropped area
Madhya Pradesh, 70-90% of land is Rampura
left fallow in the kharif season Kalan Papda
(Ryan and Sarin 1981; Singh, Jain, Rabi and Rao 1984; Michaels 1982). Most Wheat 27 51
of this land is sown to wheat in the Wheat intercrop 28 5
rabi (winter) season (Table 7). The Chickpeas 7 a 2
major reasons for kharif fallowing Lentils 11 b 25
in this area are (Michaels 1982): Linseed 1 5
Kharif
" The need to conserve moisture Rice 5 3
for wheat in the rabi season, Sorghumc 3 3
since wheat is the main Pigeon peace 12 2
subsistence crop preferred over Other 15 4
other cereals, such as sorghum. Total 100 100
" The difficulty of land Percent land
preparation, given that soil fallowed in kharif 76 90
moisture conditions may
change very quickly at the Source: Singh, Jain, and Rao (1984).
ch ng of the monsoon from a Includes chickpeas intercropped with linseed.
beginning ob Includes lentils intercropped with linseed.
too dry to too wet. c Includes various intercrop combinations.
25




" The risk of waterlogging and flooding during the monsoon.
* The difficulty of controlling weeds in the kharif season.
Given that most of the rain falls during the kharif season, the system is relatively inefficient in using available moisture; only about one-quarter of the annual rainfall may be used for crop evapotranspiration in the rabi season.
The main rotation is kharif fallow-rabi wheat. Pulses, chickpeas, and lentils are also important in the rabi cycle. In areas where soils are shallow and less fertile, pulses tend to dominate. In medium-depth better drained soils, soybeans have been found to be suitable for cultivation in the kharif season and have spread quite rapidly in some areas in recent years. In the district of Sagar, for example, rainfed soybean area grew from practically zero to over 70,000 ha in the late 1980s. Where soybeans are produced in the kharif season, land is typically left fallow in the rabi season. Pandey (1986) shows that even in this relatively favorable dryland wheat area sufficient moisture is available to plant wheat after soybeans in less than one out of every three years. Much of the expansion in soybean area has occurred on larger farms where the opportunities for substitution between wheat and soybeans are greater. In contrast, farmers who have less land place considerable priority on meeting subsistence needs for grain and straw, so the area on which soybeans substitute for wheat is smaller (Foster et al. 1987).
Further south, where temperatures are warmer, rainfall is lower, and there is no risk of frost, the importance of wheat in the rabi cycle declines. Wheat is replaced by rabi sorghum, although where it can be grown wheat continues to be the preferred subsistence crop. Cotton rather than soybeans is the dominant cash crop. Both rabi sorghum (planted in September) and cotton (planted in August) compete directly with wheat. Kharif fallow is commonly practiced because of unreliable monsoon rains and the risk of kharif crop failure. However, double cropping is practiced in the higher rainfall areas of northern Karnataka (usually wheat after mung beans).
Intercropping
Throughout the dryland wheat areas, wheat is often intercropped. In Madhya Pradesh the common intercrops are chickpeas, lentils, and linseed (Table 7). In Maharashtra and Karnataka wheat is intercropped in rows of safflower with 5-6 rows of wheat (25-cm spacing) for every row of safflower. These intercrop combinations appear more profitable than sole cropping (Table 8). Intercropping is also likely to reduce risks (Jodha 1981). For example, in on-farm experiments in dryland Karnataka over four years, the wheat-safflower intercrop was more profitable than either safflower alone or wheat alone in three years out of four, although on average safflower alone was most profitable (RR. Hanchinal, pers. com.).
26




Roles of Livestock and Wheat Straw Livestock are also important in the farming systems of the dryland wheat areas, and wheat straw is a valuable byproduct of wheat production. According to cost of production surveys, wheat straw accounts for 26% of the total value of wheat production in Madhya Pradesh (rainfed and irrigated areas), compared to only 12% in the Punjab. In rainfed districts, the value of wheat straw tends to be much higher. The value of straw is 35% of the value of the wheat crop in Madhya Pradesh (at post-harvest prices), and in dryland Karnataka the value of straw almost equals the value of grain.12 In drought years the value of straw may exceed the value of grain. As the use of tractors and irrigation increases in dryland Madhya Pradesh, the value of straw may be decreasing. Nonetheless, on smaller holdings where bullocks continue to be used, producing wheat straw for fodder is an important subsistence objective in wheat production.
Cost of Cultivation and Risk
In assessing the role of wheat in dryland farming systems, it is important to recognize that wheat is a relatively low-cost, secure crop largely produced for home consumption (both of grain and straw). Tables 9 and 10 summarize the costs of cultivation for partially mechanized wheat production in Madhya Pradesh and non-mechanized wheat production in Gujarat.
In both cases input use is very low, since land preparation is minimal (though sufficient to control kharif weed growth), little or no fertilizer is used, and there is no weed control in the crop cycle. Labor is only about 160 person hours/ha,
Table 8. Comparison of sole cropping of wheat and chickpeas with intercropping, Begumgunj, Madhya Pradesh, India
Wheat Chickpeas Wheat/chickpeas
sole cropped sole cropped intercropped
Yield (kg/ha)
Wheat 667 .. 582
Chickpeas .. 572 125
Value (Rs/ha)
Wheat (grain) 1,200 .. 1,048
Chickpeas .. 1,200 263
Fodder 132 64 140
Variable cost (Rs / ha) 962 920 954
Net benefits (Rs/ha)a 370 344 498
Source: Foster at al. (1987).
a Return to land and capital.
12 The price of wheat straw in dryland Karnataka is reported to be Rs 1.0-1.5/kg compared to a farm
gate price of wheat grain of Rs 3.0-3.2/kg (R. Hanchinal, pers. com.). The grain-to-straw yield ratio is
generally assumed to be 1:2.
27




about one-third of the labor used in irrigated areas. In Sagar, Madhya Pradesh, a yield of about 300 kg/ha is needed to pay the variable costs of wheat production. The average yield of rainfed wheat in the district is about 900 kg/ha with a low variability (see "Simulation of Wheat Yields, Sagar District," above). In addition, since harvesting costs are over 25% of total costs and are proportional to yields, variability in economic returns from dryland wheat is lower than variability in yields. Farmers in Sagar regard wheat as a secure crop. Small farmers in this area produce wheat largely for home consumption; the modest marketable surplus of wheat generated by larger farmers is intended for specialized markets that pay a premium for quality wheats (see "Prices," below).
Table 9. Approximate cost of Table 10. Approximate cost of culcultivation for partially mechanized tivation for non-mechanized dryland
dryland wheat, Khuria, Sagar durum wheat, Arnej, Ahmedabad
District, Madhya Pradesh, India, 1990 District, Gujarat, India, 1990
Cost Cost
(Rs/ha) (Rs/ha)
Land preparation Land preparation
Two cultivations by tractor 200 Cultivate twice with
bullocks plus planking 200
Planting
Seed (100 kg/ha) 400 Planting
Planting (by drill) 100 Seed (60 kg/ha) 200
Sowing (with bullock) 100
Fertilizer
(40-20-0 kg/ha of N-P-K) 350 Harvesting
Cutting (8 days/ha) 160
Harvesting Threshing (1/20 yield) 120
Cutting (1/20 share) 220
Thresher (Rs 10/100 kg) 125 Total costs 780
Total costs 1,395 Revenues
Grain
Revenues (600 kg / ha at Rs 3.75 / kg) 2,250
Grain Straw
(1,250 kg/ha at Rs 3.50/kg) 4,375 (600 kg/ha at 0.30/kg) 180
Straw
(1,880 kg/ha at Rs 0.70/kg) 1,316 Net return to
land and management 1,650
Net return to
land and management 4,300 Minimum yield
to cover variable costs 200 kg/ha
Minimum yield
to cover variable costs 300 kg/ha Source: Farmer interviews.
Source: Agricultural Extension Office, Khuria.
28




In the harshest environment for dryland wheat, the Bhal tract of Gujarat, farmers use even lower levels of inputs (Table 10) and can recover variable costs with a yield of only 200 kg/ha. Yields in this area are somewhat more variable, but recall that farmers can adjust their wheat area to the amount of moisture available at sowing (Figure 12). Hence incomes and food supply may vary quite substantially from year to year, but farmers face little risk once they have decided whether to plant wheat or not.
Varietal Development and Diffusion
Over 50% of India's dryland wheat area is now planted to improved or recommended varieties. Most of these varieties were adopted after 1976, when it was estimated that only 15% of all rainfed wheat area was sown to improved varieties (Desai 1982). Traditionally a large part of the dryland wheat area of central and southern India has been sown to durum wheat, which has dual uses in bread (chapati) or in many specialized foods. Durum wheat still predominates in Gujarat, Maharashtra, and Karnataka, but most of the wheat sown in Madhya Pradesh is now bread wheat (Table 11). Triticum dicoccum is also grown in the south, but mostly under irrigation.
A considerable number of wheat varieties have been released for rainfed areas in India over the past three decades, especially for the dryland areas of central and southern India (Table 12). Varieties released for dryland central and southern India tend to be tall compared to varieties for higher rainfall areas elsewhere in India. Nearly all of the latter varieties are semidwarf wheats, many of which are recommended for irrigated areas as well. Some of the varieties released for rainfed areas have been widely adopted, especially C306 (and Sujata, a selection from C306) in Madhya Pradesh; MACS-1967 (durum) in Maharashtra; Bijaga Yellow (durum) in Karnataka; and A-206 and GW-1 (also durum wheats) in Gujarat.
Over time, the proportion of varieties released specifically for rainfed conditions has declined, although the absolute number of rainfed varieties has not changed
Table 11. Percentage of dryland area sown to various types of wheat
Bread Durum Triticum
State wheat wheat dicoccum All
Madhya Pradesh 85 15 0 100
Gujarat 0 100 0 100
Maharashtra 0 98 2 100
Karnataka 0 98 2 100
Total 63 37 0 100
Sources: Interviews with wheat breeders in each state.
29




much because the total number of varieties released has rapidly increased in recent years. The declining share of rainfed varieties is, of course, consistent with a decline in the share of wheat produced under rainfed conditions a theme that will be discussed later. The proportion of durum varieties has averaged about 14%, which is well above their share in production. All but two of the durum wheats were released for central and southern India.
It is estimated that over half of the rainfed wheat area is now sown to recommended varieties. It is also evident that adoption of improved varieties is closely correlated with the availability of irrigation (Figure 18). In predominantly rainfed areas, improved varieties came to be adopted in the early 1970s, when a large part of the irrigated districts had already been sown to improved varieties.
Improved varieties have been adopted in dryland areas because of two main reasons. First, local land races are highly susceptible to leaf and stem rust. When weather favors the buildup of rust inoculum, disease losses are very high (Sheopuria 1990). The release of varieties resistant to rust has been a major factor in stabilizing yields. Second, farmers in dryland areas grow varieties whose high grain quality fetches a premium price (see "Prices," below); to be successful, new varieties must have high quality grain.
Table 12. Classification of wheat varieties released in India, 1966-91
Five-year period
1966-70 1971-75 1976-80 1981-85 1986-91 All
No. of varieties released per year' 2.8 6.6 6.2 10.8 6.3 7.0
Percent varieties for rainfed areasb 29 46 23 28 40 38
By zone'
N and NW Plains 43 27 36 45 37 34
NE Plains and Far East 7 12 23 19 18 17
Central 7 24 29 17 8 19
Peninsular 0 21 10 6 16 13
Northern Hills 0 12 3 8 16 9
Other 43 3 0 6 6 8
Percent durum wheat 7 15 16 9 5 12
Percent rainfed varieties
that are semidwarfs 0 40 57 67 93 47
Source: Jain and Byerlee (forthcoming 1992).
a Includes state and central releases.
b Includes varieties released for both rainfed and irrigated areas.
30




Although improved varieties have spread to dryland areas, it appears that genetic yield gains in varieties released for these areas have been modest (Hanchinal 1988). In Gujarat, for example, the successful durum variety GW-1 (released in 1986) had a yield advantage of only 9% over the farmers' variety A-206 (released in 1954) in testing over six years at three sites (Wheat Research Institute, Vijapur, pers. com.). Even though the yield advantage of GW-1 is modest, this variety has replaced A-206 on a significant area because of its higher grain quality (and hence higher price) and its improved rust resistance. Similar experiments conducted over six years at Indore found local varieties yielding better than the improved varieties recommended for the area (Table 13). The slightly better performance of the local varieties appeared to result from
Percentage of total wheat area
100
0-20% Rainfed
80
20-40% Rainfe
60
406%Rainfed
40
20
0'
1966 1969 1972 1975 1978 1981
Figure 18. Percentage of total wheat area under modem varieties in districts grouped by size of rainfed area, India, 1966-83. Source: District data.
Table 13. Yield and associated characteristics of local and improved wheat varieties, Indore, India, 1971-76
Varieties Yield Seedlings/m2 Ears/rn2
Locals (Malvi, Meghdoot, and Kathia) 1.66 213 282
Improved (KH65 and C306) 1.47 157 231
Source: Upadhvaya et al. (1986).
31




their 30-50% higher germination rate. Nonetheless, because of better disease resistance the improved varieties have been widely adopted. For example, the variety C306, grown on a wide area in Madhya Pradesh, combines reasonable disease resistance with high grain quality and excellent drought tolerance. Experiments have consistently shown the superior performance of C306 when it is grown in residual moisture or with limited irrigation (Sheopuria, pers. corn.).
Overall, breeding improved wheat varieties for dryland conditions in central and southern India is a formidable challenge. Varieties require an ability to germinate in hot soils, long coleoptiles to emerge from deep planting, and a deep root system to exploit receding residual moisture as the season progresses (Hanchinal 1988). Grain quality is also a major criterion for farmer adoption, given the high market premiums for quality. Input responsiveness seems to be a secondary criterion, since there is almost never enough moisture to produce more than 2.0-2.5 t/ha. Nonetheless, slow progress has been made; successful (widely adopted) varieties have been released about once every 20 years in each state. As a result, yields are stable in those few years when rusts (leaf rust and, for durum wheat, stem rust) are a problem.
Crop and Resource Management
Although considerable emphasis has been given to developing varieties for dryland areas, it is generally recognized that improvements in crop and soil management to conserve moisture and to utilize available moisture more efficiently play a major role in increasing productivity for these areas (Bolton 1979, Byerlee and Winkelmann 1981, CIMIMT 1989a). This section reviews the potential for increasing the productivity of wheat in dryland areas of central and southern India through increased fertilizer use, fallowing, moisture conservation, and investment in irrigation.
Fertilizer
Statistics on fertilizer use in wheat are available only for 1976. Nonetheless, data on rabi fertilizer off take in states where wheat is the major crop, combined with estimates of the proportion of rabi fertilizer used on wheat, can be used to approximate trends in fertilizer use on wheat (see Appendix A). Overall, the average amount of fertilizer used on wheat for all India is around 110 kg nutrient/ ha, compared to only 25 kg nutrient/ ha in 1969 (Figure 19). However, as expected, fertilizer use on wheat at the state level closely reflects the area under dryland wheat. Average fertilizer use on wheat is about 45 kg nutrient/ ha in Madhya Pradesh compared to 200 kg nutrient! ha in the Punjab (Figure 19) and 95 kg nutrient/ ha in Uttar Pradesh. Nonetheless, fertilizer use in Madhya Pradesh was virtually nil in 1969. The increase in fertilizer use in large part reflects the expansion of irrigated area, discussed below.
32




As with the use of improved varieties, use of fertilizer in rainfed and dryland wheat areas expanded rapidly in the past decade. In the 1970s, it was estimated that only 10% of the rainfed wheat area was fertilized, at an average rate of 35 kg nutrient/ha13 (Desai 1982). In contrast, it is estimated that 70% of the irrigated area was fertilized at nearly 80 kg nutrient/ha. Although comparable data are not available for the 1980s, a compilation from various sources indicates that perhaps half of the dryland wheat area in central and southern India is now fertilized. For example, in 1989 it was estimated that 41% of the dryland wheat area in Karnataka and 65% of the dryland wheat area in Gujarat was fertilized.4 In Madhya Pradesh, it is estimated that 65% of the total wheat area was fertilized. Assuming that 90% of the irrigated area was fertilized, this implies that 50% of the dryland wheat area in Madhya Pradesh was also fertilized. However, fertilizer use varies a great deal between districts in Madhya Pradesh. Less than 10% of the wheat area is fertilized in some districts where levels of irrigation are low (Figure 20). Also, fertilizer use in rainfed wheat varies from year to year depending on moisture conditions. Even moderate doses of nitrogen applied early in the season may depress yields in very dry years due to excessive vegetative growth. However, in those few years when rainfall in the growing season is timely, farmers may top-dress to boost yields.
Fertilizer (kg nutrient/ha)
200
180
160 Punjab
140 120 100 80
60 Mdy rds
40 20
1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 Figure 19. Estimated fertilizer use on wheat in India, 1969-90. Source: See Appendix A.
13 These data, which are an average of all India, are likely to overestimate the use of fertilizer in drier
areas of central and southern India.
14 Monitoring and evaluation reports of the state extension system.
33




The general recommendation is to apply 40-20-0 kg N-P205-K20 / ha, but farmers approximate these doses only in the wheat belt of northern Madhya Pradesh. In Karnataka, farmers who apply fertilizer use only about 20 kg nutrient/ ha (equally divided between N and P205), and in Gujarat the average is closer to 10 kg/ha.
These variations in fertilizer use reflect expected fertilizer response. For low doses of fertilizer, the estimated grain-to-nutrient ratio for rainfed wheat in Madhya Pradesh is about 10 kg wheat to 1 kg nutrient (Fertilizer Association of India 1985, Rao 1976). In drier areas, this response is lower. In Maharashtra, the estimated ratio is only 5.5:1 (Kohli 1976) and no response was reported in over six years of fertilizer response trials in Gujarat's dryland wheat areas (Maliwal 1989).
Although the area and dosages of fertilized dryland wheat are low, substantial progress has been made over the past decade in increasing fertilizer use. In fact, increased fertilizer consumption may account for much of the increase in dryland wheat yields since 1970. If 50% of rainfed wheat is now fertilized at an average rate of 30 kg nutrient/ha, this could account for an increase in average dryland yields of 150 kg/ha. Under the extreme limitations of moisture in dryland wheat production in central and southern India, only very slow growth in fertilizer use can be expected in the future.
Rabi N-P-K (kg nutrient/ha)
200
150
100
Jr =0.73
50 m U U
0U
0 20 40 60 80 100
Percent of total wheat area under irrigation
Figure 20. Relationship between fertilizer use in the rabi (winter) cycle and irrigation in districts of Madhya Pradesh, India, 1983-84.
34




Fallow Management and Moisture Conservation Clearly the success of dryland wheat production in central and southern India depends on conserving moisture during the monsoon season (see "Rainfall," above). During this season, farmers attempt to maintain a clean fallow through two to four tillage operations using a shallow scraper-harrow, which kills weeds and creates a shallow soil mulch in the top 5 cm of the soil surface that helps conserve moisture.
Another widely used technique is to level and bund fields, so that monsoon rainfall can stand in the fields for up to two months. This system, locally called the haveli system, achieves several purposes: 1) it prevents runoff and hence soil erosion; 2) it conserves moisture, since 50% of monsoon rainfall may be lost to runoff (Singh and Raje 1984); and 3) it helps prevent the germination of kharif weeds, especially Karsgrass (Saccharum pontaneum), a serious perennial weed prevalent in the wheat growing areas of Madhya Pradesh.
The main disadvantages of the haveli system are: 1) the capital cost to level the field and construct bunds and 2) the need to dedicate a bunded field to rabi cropping and hence the loss in flexibility to substitute between rabi and kharif crops. The practice has, however, become widespread in the main dryland wheat belts of Madhya Pradesh.
Finally, the most efficient use of moisture may be to plant kharif crop land that is now commonly left fallow. The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) has devoted two decades to research on technologies to allow kharif cropping in deep vertisols in higher rainfall areas in a double cropping sequence. This technology depends on broad beds and furrows across the slope to facilitate drainage and moisture conservation. The technology, extensively demonstrated over several years in Madhya Pradesh, has had only partial success (Foster 1987, Walker and Ryan 1990). Farmers tend to adopt elements of the technological package, but the key element, a bullock-drawn wheeled tool-carrier, has not been adopted because of doubled operating costs, high initial investment outlays, farmers' increasing use of tractors, and the uncertainty associated with double cropping. Low adoption is also explained by farmers' strong desire to maintain a considerable proportion of area under wheat to produce both subsistence grain and fodder. Double cropping by planting wheat after soybeans greatly decreases the probability that moisture will be sufficient for planting wheat (hence the risk of crop failures in the rabi season) (Pandey 1986).
Improvements in the physical structure of soils may also help conserve moisture. Organic manures, such as farm yard manure (FYM) and green manures, are recognized as helping improve moisture conservation. However, farmers' use of FYM tends to be quite low in the black vertisols and is often confined to cash crops (e.g., cotton). Promising experimental results have been obtained through the rotation of wheat with deep-rooted crops such as castor, which help to improve the physical structure of soils (IARI 1990).
35




Irrigation
The most widely used strategy to deal with the serious problem of drought stress and low yields in dryland wheat is investment in irrigation facilitates to provide limited or full irrigation to the crop. The number of tubewells in central and southern India has increased rapidly over the past two decades, facilitated by village electrification, so that now over 40% of the wheat area is irrigated.
In most areas, wheat receives priority in the allocation of scarce irritation water supplies, although the water supply is often limited and the wheat crop receives a restricted number of irrigations. However, most studies show that even one irrigation at planting will increase yields by 0.5-1.0 t/ha. Overall it is estimated that half of the irrigated wheat area in Madhya Pradesh receives less than 50% of the recommended number of irrigations. (More precise data on irrigated wheat are available for Uttar Pradesh; they indicate that only 30% of the irrigated wheat area receives three or more irrigations; see Sinha, Aggarwal, and Chopra 1985 and Figure 21). Hence an important priority for research is to develop appropriate recommendations for the large area of wheat grown under limited irrigation.
A major challenge to the future Percentage area expansion of irrigated or partly
34% irrigated wheat area in central and
southern India is the over-exploitation 29% of groundwater in many areas, leading
26% to a declining water table, higher
irrigation costs, and in some cases serious water shortages in dry years. Given these trends, the rate of
expansion of irrigated wheat area in central and southern India is likely to slow in the future.
11%
In areas where underground water is not available, it may be possible to "harvest" the monsoon runoff
(estimated at 50% or more of total rainfall) in storage tanks in order to 0 1 2 3 ormore provide one or two irrigations to the
3 ormore wheat crop. While there has been
Figure 21. Number of irrigations considerable experimentation with this
given to wheat in Uttar Pradesh, technique, its adoption is limited
India. because of the high investment costs
(Pandey 1986) and the need for Source: Sinha et at. (1985). neighboring farmers in a watershed
area to develop close cooperation.
36




Wheat Marketing and Consumption
Prices
For all wheat growing areas of India, real prices received by farmers for wheat have shown a steady downward tendency except for a short period during the world food crisis of 1973-75 (Figure 22). In general, farmers in rainfed areas receive a somewhat higher price than farmers in irrigated areas and this difference has widened over time. Before the Green Revolution in the mid-1960s, there was no difference in prices received by farmers in rainfed and irrigated areas, but by the 1980s this difference was consistently 20% or more. The data from Sagar District, where fine-grained wheats (C306) are produced, suggest that the price premium received by farmers was first noticeable in the late 1960s and was as high as 30% in the 1970s and 1980s (Figures 23 and 24).
For durum wheat grown in South India the price premium is even higher. Even before the Green Revolution, there was a considerable price premium for durum wheats (Figure 24). By the 1980s this difference had increased to about 50%.
The increase in wheat prices in dryland areas relative to irrigated areas reflects a tendency for farmers in dryland areas to specialize in premium quality bread wheat and durum wheat. As dryland wheat area has steadily declined as a proportion of total wheat area, and in the absence of significant technological change and with a very small marketed surplus, the supply of premium quality wheats has tended to contract relative to demand and has led to upward pressure on prices. Hence although dryland wheat farmers have also seen their
Real farm gate wheat price (Rs/100kg)
350
300
Districts
60-100% rainfed
250
200
Districts
150 80-100% irrigated
100....
1966 1969 1972 1975 1978 1981
Figure 22. Real farm gate wheat prices in districts grouped by irrigation status. Source: District data files. Wheat prices deflated by the Wholesale Price Index.
37




real wheat prices decline as a result of the Green Revolution, this decline has been less than it would have been in the absence of their comparative advantage in producing premium quality wheats.
Ratio
1.9
1.8 Madhya Pradesh
1.7 T1TT Karnataka
1.6
1.5
1.4.
1.3
1.2.
1.1
1.0
0.9 0.8 0.7
0.6 0.51
0.4 0.3
0.2 0.1
0.0 d 1
1958-62 1962-67 1968-72 1973-75 1982-86
Figure 23. Ratio of farm harvest prices for wheat in Madhya Pradesh and Karnataka to prices in Punjab, India.
Ratio
1.5
1.4
1.3
1.2
1.1I 1
1.0 0.9
0.8
0.7
0.6 0.5 0.4
0.3
0.2
0.1
0.0
1954-60 1963-67 1968-74 1981-85
Sagar fine white Durum C & S India
Figure 24. Ratio of wholesale prices of bread wheat in Sagar and durum wheat in central and southern India to price of bread wheat in Ludhiana, Punjab.
Note: Durum price is average of Dhanduka (Gujarat), Amroati (Maharashtra), and Hyderabad
(Andhra Pradesh).
38




Wheat Consumption
The estimated wheat balance sheet for central and southern India is given in
15
Table 14. Over 3 million tons of wheat are imported annually from northern India, accounting for one-third of all wheat consumed. Almost no wheat is locally procured and only a small proportion of local wheat production (10% in Madhya Pradesh and Karnataka and 20% in Maharashtra) enters the market. As expected, wheat self-sufficiency declines from north to south. Madhya Pradesh almost achieves self-sufficiency in wheat whereas southern states have a very low level of self-sufficiency.
Per capita wheat consumption is also higher among the predominantly wheat consuming population in the northern part of this region (Table 15), but it is still well below consumption levels in northern states such as the Punjab. Note also that wheat consumption is generally much higher in urban areas, where the public food distribution system concentrates its efforts. For example, in Maharashtra and Karnataka, urban per capita consumption is almost three times higher than in rural areas. In the southern states, annual wheat consumption is below 10 kg per capita, lower than most tropical countries (Byerlee 1985).
Another characteristic of wheat consumption in India is the strong tendency for consumption to rise with increasing income (Figure 25). This is especially
Table 14. Total wheat production, estimated consumption, and percentage self-sufficiency, central and southern India, 1984-86
Public
Wheat distribution Total Selfproduction of wheat' consumption b sufficiency State (000 t) (000 t) (000 t) (%)
Madhya Pradesh 4,170 184 4,354 96
Gujarat 1,246 350 1,596 78
Maharashtra 881 919 1,800 49
Karnataka 165 458 623 26
Andhra Pradesh 9 304 313 3
Kerala 211 211 0
Tamil Nadu 654 654 0
Total, central and
southern India 6,471 3,084 9,556 68
Source: Government of India, Directorate of Economics and Statistics (1988). a Net of small quantity procured in Madhya Pradesh. b Sum of production and public distribution, assuming private marketing of wheat from
surplus to deficit states is negligible.
15 This balance sheet reasonably assumes that private marketing of wheat from surplus northern states to
deficit states of central and southern India is unimportant.
39




marked in the southern states, where wheat consumption is concentrated in the highest income groups. In general, there is a strong trend to substitute wheat and rice for the traditional food staples, especially sorghum (Figure 26).
These data suggest that wheat consumption in the southern states could expand rapidly in the future as incomes rise and as the population becomes more urbanized.
Dryland Wheat and the Poor
When assessing research priorities, it is important to know if there is an association between dryland wheat production and the incidence of poverty. In general India's unirrigated areas tend to be poorer than its irrigated areas, and this gap has widened over time. States having a low percentage of irrigated cropped area, such as Madhya Pradesh, also tend to have a low rural per capita income and a high percentage of the population living in poverty (Figure 27).
This picture becomes clearer after an examination of district-level agricultural wages (Figure 28). In 1966 before the effects of the Green Revolution were felt, the gap in agricultural wages between rainfed and irrigated areas was relatively small. From 1967 to 1971, the gap widened sharply and has been maintained ever since (Figure 28). The real daily wage rate (measured in kilograms of wheat) in largely irrigated districts is double that in largely rainfed districts (Figure 29). These data suggest that dryland wheat is produced (and consumed) mostly by the poor. However, except for northern Madhya Pradesh, dryland wheat does not constitute as large a share of poor producers' incomes or poor consumers' diets as wheat does in northern India. It should be noted also that the largest
Table 15. Rural and urban per capita wheat consumption in central and southern India compared to Punjab and all India, 1977-78
Area Rural areas Urban areas All
Central and southern India
Madhya Pradesh 57 93 64
Gujarat 43 74 53
Maharashtra 20 53 32
Karnataka 6 17 9
Andhra Pradesh 1 10 3
Kerala 3 5 4
Tamil Nadu 2 7 4
Punjab 138 113 131
All India 49 58 51
Source: National Sample Survey.
40




density of poor peope occurs in northeastern India; many are concentrated in districts where rainfed wheat is grown although under more favorable conditions compared to central and southern India.
Per capita wheat consumption (kg/yr)
180'
Madhya Pradesh 160 140 120 100 80
60 40 20 0
0 40 80 120 160 200 240
Per capita income (Rs/mo) Figure 25. Relationship between wheat consumption and income, rural areas, central and southern India, 1977-78. Source: National Sample Survey. Percentage of total cereal consumption
100
90 06,o, Coarse grains
80 70
60
50 a..
40 30
20 10
0 40 80 120 160 200 240
Per capita income (Rs/mo) Figure 26. Share of different cereals in total cereal consumption, rural Maharashtra, India 1977-78. Source: National Sample Survey.
41




Earnings of rural labor households (Rs/yr) 1,000
r = 0.83
900 Haravana
Punjab
800
700 U
Bihara
600
500
400 Madhya Pradesh
4 Orissa
300
200 ..
1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000
State agricultural value added per worker (Rs/yr)
Percentage of rural population in poverty
80
Orissa r = 0.88
70 '
Bihar West Bengal
60
Madhya-,, Maharashtra
50 Pradesh M'.Karnataka
Uttar
Pradesh
40 Gujarat
Rajasthan 30 '
30 Haravana
20
10 Punjab
0 1 1aI S SSI
1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000
State agricultural value added per worker (Rs/yr)
Figure 27. Relationship between state agricultural income, level of poverty, and earnings of rural labor households.
42




Real wage (Rs 1980)
12
80-100% irrigated
10 **
8 4 60-80% irrigated
4 0-40% irrigated
0 1 1 1 1* I I I I I I I II* *
1965 1970 1975 1980 1985
Figure 28. Real agricultural wages in districts grouped by the percentage of wheat area irrigated.
Agricultural real wage Growth of real agricultural
(kg wheat/day), 1980-83 wage (%/yr), 1966-83
7.8 1.7
6.8
1.4
4.3 1.0 1.0
3.4 3.6
0.4
m,,K,,, m I .-: I///III
80-100% 60-80% 40-60% 2040% 0-20% 80-100% 60-80% 40-60% 2040% 0-20%
Percentage of total wheat area that is rainfed Percentage of total wheat area that is rainfed Figure 29. Real wages and their trends in districts grouped by the percentage of wheat area that is rainfed, India.
43




Research Resource Allocation to Dryland Wheat
Given the low and lagging productivity of dryland wheat, an important question for India and for developing countries in general and CIMMYT is whether more research resources should be allocated to try to increase productivity in this sector. While this paper does not attempt to answer this question specifically for the case of India, it does apply a simple method, called congruency analysis, to illustrate the kinds of analysis that could be undertaken to address this question.
Congruency Analysis
A number of methods of varying complexity have been applied by economists to analyze the optimum allocation of resources between research programs, whether they are organized by commodities, regions, or problems. The simplest approach, congruency analysis, examines the allocation of research resources in relation to the share in value of production. In other words, if commodities/ regions A and B account for 70% and 30% of the value of production, respectively, then research resources should be allocated to A and B in a 70:30 ratio (Barker 1988, Scobie 1984).
In the present study, the issue is whether the value of production in dryland wheat is consistent with the allocation of research resources to dryland wheat. To examine this question, Indian wheat production is divided into three major zones (following approximately the agroclimatic zones used by the All India Coordinated Wheat Program) and two moisture regimes:16
1. Northwest Plains (irrigated);
2. Northeast Plains (irrigated);
3. Central and Southern India (irrigated);
4. Northern Plains (rainfed); and
5. Central and Southern India (rainfed).'7
The distinction between zones 4 and 5 is important, since farmers in the more favorable rainfed areas in northern India can often use varieties developed for nearby irrigated areas, while in central and southern India quite distinct breeding programs are needed for dryland wheat.
Table 16 presents the area and production of wheat in each zone. While Zone 5 (dryland central and southern India) accounted for 16% of the wheat area in
16 For purposes of exposition, th- conventional wheat production zones have been slightly modified;
central and peninsular India have been aggregated and the Northern Hills have been ignored.
17 Includes all rainfed wheat in Madhva Pradesh, Maharashtra, Karnataka, and Gujarat, as well as some
districts of Uttar Pradesh and Rajasthan that border Madhya Pradesh.
44




1986, it provided only 7% of the wheat produced. Moreover, given the projected trends in growth of irrigated compared to rainfed area, by 2000 this zone may provide only 5% of India's wheat production. However, since ramnfed wheat receives a price premium (the section on "Prices," above), the share of dryland wheat in value of production is presently slightly higher.
Some idea of the historical allocation of wheat breeding resources can be obtained by examining the number of varieties released for each of the zones listed above (Table 16). In general, there is a good correspondence between the share of varieties released and the share of value of production. The main exception is the high share of varieties released for irrigated central and southern India relative to the Northwestern Plains (Table 16). For rainfed areas, the share of varieties released is somewhat higher than their share in the value of production, except for the last 10-year period, when only two rainfed varieties were released in Central and Southern India.
Table 16. Summary of congruency analysis of allocation of wheat research resources by zone, India
Irrigated Rainfed
Central Central
Northwest Northeast and and
plains plains southern North southern
(1) (2) (3) (4) (5) All
Percent area' 38.6 25.4 9.9 10.1 16.1 100
Percent production" 51.4 25.5 10.0 6.5 6.6 100
Percent value of production b 49.4 24.5 11.5 6.3 8.2 100
Percent value of production
weighted by povertyc 29.7 36.9 17.3 5.6 10.5 100
Percent value of production
weighted by poverty and
expected research progress d 31.8 39.5 18.6 4.5 5..6 100
Percent allocation of
research resources d
1973-90 32.6 15.6 26.0 13.5 12.3 100
1985-90 33.1 16.8 25.8 13.9 10.3 100
Percent of varieties released d,,
1966-91 37.1 16.4 23.6 12.9 10.0 100
1980-91 43.0 13.9 21.5 15.2 6.3 100
a Based on current area and production.
b Based on projected production and prices. c Weighted by reciprocal of agricultural per capita income. d Excludes northern and southern hills.
e Excludes varieties recommended for more than one zone or for both irrigated and rainfed
areas.
45




Modification of Congruency Analysis The analysis described above fails to take into account two other important factors in allocating research resources: the need to direct resources toward alleviating poverty, and the expected rate of research progress.
Those living in rainfed areas have already been shown to be generally poorer than those in irrigated areas. Real wages in rainfed districts are less than half the wages in irrigated districts. Income differences may be even greater, since laborers in irrigated areas work more days per year. In general, states with a significant area of dryland wheat, such as Madhya Pradesh, have a high proportion of the rural population living in poverty (Figure 27), although other largely irrigated states, such as Bihar, also have a high level of poverty. Whichever income measure is used, the Northwest Irrigated Plains is clearly the wealthiest zone.
Turning to the second issue, there is considerable evidence that, at least for wheat breeding programs, the expected rate of progress from research is lower in dryland areas (CIMMYT 1989b). Although wheat breeders have made yield gains of close to 1 % annually in irrigated or well-watered areas, this rate of progress may be halved in areas where moisture stress is common. Even in countries such as Australia that have made concerted breeding efforts for dryland areas, the rate of yield gain from plant breeding has rarely exceeded
0.5% annually for a significant period.
The evidence examined in this paper also suggests that progress has been slow in developing varieties of wheat for dryland India. Indeed, central and southern India are among the most difficult wheat growing environments in the world. The evidence would thus suggest that research progress in dryland wheat is likely to be slow.
To include the effects of the level of poverty and rate of research progress into the analysis, two indices can be constructed. One index is based on the value of production weighted by the reciprocal of rural incomes or wages, and the second index is based on both the poverty level and the expected rate of research progress. The weighted index based on poverty level clearly favors increased investment in dryland wheat, because it assigns a higher weight to poor people (Table 16). It also significantly increases the allocation to the Northeast Plains a poor zone as well and decreases the allocation to the Northwest Plains. However, if research progress is expected to be slower in the dryland areas, the weighted index (bottom line) suggests less difference from the share based on value of production, although once again the Northeast Irrigated Plains maintains its importance (Table 16).
Overall the results suggest that, to justify an increased allocation of research resources to dryland wheat, planners would need to attach a very high weight to alleviating poverty or they would need to believe that research progress will be
46




much more rapid than in the past in India (or in other countries). If the latter is the case, an increased allocation to the Northeast Plains appears also to be a high priority.
Other Considerations
In a highly simplistic approach such as the congruency analysis described above, many issues are not considered. These include:
1. Breeding vs. crop management research: The analysis has considered only
gains from breeding research. In dryland areas, research to conserve
moisture, to use moisture more efficiently, and to conserve soil resources may
be more important than breeding research. Evidence from other countries (e.g., Australia and Turkey) suggests that investment in crop and resource
management research may provide high payoffs in dryland areas.
2. Wheat vs. other crops: In central and southern India, producing wheat after
fallow in a receding monsoon moisture profile is an inefficient way of using
available annual rainfall. Better utilization of annual rainfall implies the
development of methods for planting and managing kharif crops that is, if
the objective is to improve productivity in dryland areas, then research on
wheat may not be the priority. The rapid diffusion of soybeans in central
India is an example of how productivity can be increased by changing the
cropping pattern. In addition, productivity gains in sorghum and millet may
provide greater benefits to the poor who (outside of northern Madhya
Pradesh) depend more on these crops for income generation or consumption.
However, the priority that farmers give to wheat grain and straw suggests
that wheat will continue to be important.
3. Maintaining market premiums for dryland wheat: Farmers in dryland
areas now enjoy a considerable price premium for quality in the case of bread
wheat or because they have a monopoly on durum production. This
premium has widened as dryland wheat area has declined. This implies that wheat breeding for dryland areas must give a high weight to quality. At the same time, efforts to develop durum wheats for irrigated areas are likely to
reduce the price premium that farmers in dryland areas now receive.
4. Poor producers vs. consumers: The analysis presented above assumes that
the benefits of research accrue to producers. In fact there is substantial
evidence that poor consumers gain most from research through lower food
prices. Since many of the poor in dryland areas (including small farmers) are net purchasers of wheat, they benefit from increased production in irrigated
areas, which lowered food grain prices (Renkow 1991).
47




Conclusions
This review of dryland wheat production in India has been motivated by the need for a better indication of the rate of technical progress in dryland areas as well as the need to assess the prospects for future research payoffs in these areas. The paper has distinguished between the rainfed wheat areas of northern India and the dryland areas of central and southern India. In the northern areas, moisture and temperature conditions are generally more favorable for wheat production; the technologies developed for nearby irrigated areas are often applicable. In contrast, wheat in central and southern India is produced under marginal moisture conditions requiring technologies quite different from those used to produce wheat in adjacent irrigated areas.
Although rainfed wheat has traditionally been important in national wheat production in India, the extraordinary success in expanding irrigation facilities has resulted in a continuing decline in the area sown to rainfed wheat. In addition, yields of rainfed wheat have grown at only half the rate of yields of irrigated wheat over the past 25 years, further reducing the share of dryland wheat in national wheat production to an estimated 12% today.
Most dryland wheat is produced in deep vertisols in a fallow-wheat system depending heavily on the conservation of monsoon rainfall. While this practice represents a relatively inefficient use of total available moisture, for farmers it substantially reduces the risks inherent in kharif cultivation (the difficulty of working heavy soils in the monsoon season, waterlogging problems, etc.). At the same time wheat provides the main source of subsistence food and fodder (wheat straw). Hence, surprisingly for a marginal rainfall area, wheat is regarded by farmers as a secure crop. An examination of historical yields and a crop simulation confirm the relatively low risk in wheat production under these conditions. Moreover, in the driest areas, such as Gujarat, farmers adjust the area sown to wheat according to the amount of moisture available at planting time. This strategy may lead to variability in incomes, but it helps farmers avoid risk. Finally, wheat grown under these conditions is a low-input crop that requires a yield of only 200-300 kg/ha to cover variable costs of production. Even in the driest years, farmers are likely to cover their costs.
Although yields in dryland areas have expanded more slowly than yields in irrigated areas, slow but steady progress has been achieved. Perhaps half of the dryland wheat area is now planted to improved varieties. These varieties, many of them durum wheats, are distinguished by their superior drought tolerance, resistance to rusts, and high quality grain (which fetches a price premium in the market). Semidwarf wheats, which spearheaded the Green Revolution elsewhere in India, have made virtually no impact on dryland wheat production in central and southern India.
Likewise, fertilizer use on dryland wheat has expanded steadily since 1976, although the level of application is low and over half of dryland wheat may still
48




be unfertilized. Experience with fertilizer use has been most positive in the main dryland wheat belt in northern Madhya Pradesh. In the driest area (e.g., Gujarat), no economic response is obtained from fertilizer use.
Clearly a priority in increasing the productivity of dryland wheat is to find ways to conserve more moisture. Farmers have developed systems to trap monsoon rainfall for the subsequent wheat crop. Considerable research has also been undertaken to find ways to produce kharif crops. However, in general there is insufficient moisture to allow double cropping, and given the substantial investment in water control/ drainage required for monsoon cropping, and the importance of wheat for food and fodder, these technologies have not been widely adopted.
Farmers' major strategy for increasing productivity has been to invest in irrigation, especially tubewells, with the result that the production of wheat under irrigation has expanded rapidly throughout the dryland wheat belt. However, irrigation is only possible where there is sufficient groundwater at a reasonable depth; it seems that much of the area that is currently irrigated is unsustainable because of limited groundwater supplies.
Given that dryland areas of India have generally lagged behind irrigated areas in the rate of technical progress, there is pressure to increase productivity in dryland systems. An important question is the tradeoff in allocating wheat research resources between favored (i.e., irrigated) areas and dryland areas. While it is not the place of this paper to recommend the level of priority that should be assigned to dryland wheat research in India, a simple analysis urges some caution in increasing investment in the marginal areas relative to favored areas. Some evidence suggests that breeding resources have been allocated to dryland wheat in accordance with its relative importance in total production. In the future, the declining importance of dryland wheat may suggest a declining priority to wheat research in this environment.
There are, however, a number of qualifications to this tentative conclusion. First, we have assumed that research progress will be slower in dryland areas. In the case of crop and soil management research, this may not be the case. Second, the close association in India between the incidence of poverty and dependence on dryland agriculture may justify further research to increase agricultural productivity in these areas. Finally, some consideration should be given to the interactions between favored and marginal areas occurring through food and labor markets. Given that many households in marginal areas are net food purchasers, any decline in food prices (e.g., through increased productivity in irrigated areas) may benefit the poor in marginal areas (Renkow 1991). On the other hand, given the specialization of marginal areas in premium wheats, success in efforts to produce premium wheats in irrigated areas (e.g., through high quality durum wheats) may have adverse effects on prices received by wheat producers in marginal areas. The resolution of these questions requires more in-depth analysis than is afforded by this paper.
49




Appendix A
Estimated Quantity of Fertilizer Applied to Wheat in Major Wheat Producing States of India, 1969-88
Table A.1. Estimated quantity of fertilizer applied on wheat, major wheat producing states, India, 1969-88'
Share of
rabi Fertilizer applied to wheat (kg/ha)
fertilizer
State to wheat 1969 1970 1971 1972 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1986 1987 1988 1989 1990
Bihar 0.63 20 36 28 29 27 37 40 45 52 52 57 60 50 62 1(00 10)8 112 102 98
Haryana 0.90 32 34 39 45 40 53 65 88 89 89 94 101 90 105 120 137 136 163 167
HimnachaliPradesh 0.43 3 3 4 4 6 6 7 7 8 7 8 9 10 11 14 16 15 19 2(1
Jammu and Kashmir 0.80 3 2 6 6 8 17 17 13 22 29 28 18 69 75 55 63 53 56 70 C
Madhya Pradesh 0.67 2 5 7 10 16 16 21 21 24 20 26 26 25 32 39 49 47 63 63 L
Punjab 090 54 54 61 81 76 85 96 114 137 146 168 178 171 179 195 194 198 198 195
Rajasthan 0.90 13 16 22 27 24 28 35 38 42 45 49 46 49 58 81 82 84 104 103
Uttar Pradesh 0.75 33 42 35 41 31 44 59 69 77 69 76 86 93 103 93 101 89 108 112
All states 25 31 30 37 33 42 52 61 68 68 77 83 83 92 99 107 102 115 118
Source: Fertilizer Statistics of India (various issues).
a Calculated by multiplying rabi fertilizer offtake by the estimated share applied to wheat in each state. The estimated share to wheat was computed from Desai (1982) and Cost of Cultivation Surveys.




References
Barker, R. 1988. Methods for Setting Agricultural Research Priorities: Report of a Bellagio
Conference. Working Paper in Agricultural Economics No. 88-3. Ithaca, NY: Cornell
University.
Belaid, A., and M.L. Morris. 1990. Wheat and Barley Production in Rainfed Marginal
Environments of WANA Region: Problems and Prospects. CIMMYT Economics Working
Paper 91/02. Mexico, D.F. CIMMYT.
Bolton, F.E.V. 1979. Agronomic yield constraints in rainfed cereal production systems. In
Vol. 2 of The Gap Between: Present Farm Yield and Potential Major Constraints and
Possible Solutions. Mexico, D.F.: Ministry of Agriculture and Agrarian Revolution
(Algeria), International Center for Agricultural Research in the Dry Areas, and
CIMMYT. Pp. 2-16.
Byerlee, D. 1985. "Wheat in the Tropics: Economic and Policy Issues." In Wheats for More
Tropical Environments: A Proceedings of the International Symposium. Mexico, D.F.:
CIMMYT. Pp. 303-315.
Byerlee, D. and D.Winkelmann. 1981. Accelerated Wheat Production in Semi-Arid Developing
Areas: Economic and Policy Issues. CIMMYT Economics Paper 81/2. Mexico, D.F.:
CIMMYT.
CIMMYT. 1989a. Toward the 21st Century: CIMMYT's Strategy. Mexico, D.F.: CIMMYT.
CIMMYT. 1989b. 1987-1988 CIMMYT World Wheat Facts and Trends. The Wheat Revolution
Revisited: Recent Trends and Future Challenges. Mexico, D.F.: CIMMYT.
Desai, G.M. 1982. Sustaining Rapid Growth in India's Fertilizer Consumption: A Perspective Based
on Composition of Use. IFPRI Research Report 41. Washington, D.C.: International
Food Policy Research Institute.
Fertilizer Association of India. Fertilizer Statistics (various issues). New Delhi: Fertilizer
Association of India.
Fischer, R.A., and Varughese, G. 1990. Developments in CIMMYT Wheat Program in 1989.
Annual Wheat Newsletter 36:94-95.
Foster, J.H., K.G. Kshirsagar, M.J. Bhende, V. Bhaskar Rao, and T.S. Walker. 1987. Early
Adoption of Improved Vertisol Technology Options and Double Cropping in Begumgunj,
Madhya Pradesh. Economics Group of Resource Management Programme, Progress
Report 77. Patancheru, India: International Crops Research Institute for the SemiArid Tropics.
Government of India. Area and Production of Principal Crops in India (various issues). New
Delhi: Directorate of Economics and Statistics, Ministry of Agriculture.
Government of India. 1988. Bulletin of Food Statistics, 1986. New Delhi: Directorate of
Economics and Statistics, Ministry of Agriculture.
Hanchinal, R.R. 1988. "Development of Wheat Varieties for Rainfed Conditions." Draft
Paper, Dharward, India: University of Agricultural Sciences.
51




IARI. 1990. Indian Agricultural Research Institute Regional Station (Wheat Breeding). Indore:
Indian Agricultural Research Institute.
Jain, KB.L., and D. Byerlee. 1992. "Wheat Breeding in India: Trends and Impacts." Draft
paper. Mexico, D.F.: ICAR/CIMMYT (forthcoming).
Jodha, N.S. 1981. Yield stability and economics of intercropping in traditional farming
systems. In Proceedings of the International Workshop on Intercropping. Patancheru,
India: International Crops Research Institute for the Semi-Arid Tropics. Pp. 282-292.
Kohli, A.K. 1976. Yield per unit of land and inputs, quantity and quality-wise in relation to
other cereals with reference to Maharashtra. In Wheat in the Changing Social Context.
New Delhi: Government of India.
Maliwal, G.L. 1989. Bhal and Coastal agroclimatic zone: Important research
accomplishments. Arnej Station, Gujarat Agricultural University, India. Mimeo.
Michaels, G. 1982. The Determinants of KharifFallowing on the Vertisols in Semi-Arid
Tropical India. Ph.D. thesis. St. Paul: University of Minnesota.
Pandey, S. 1986. Economics of Water Harvesting and Supplementary Irrigation in the SemiArid Tropics of India: A Systems Approach. Ph.D. dissertation. Armidale, Australia:
University of New England.
Put, M.., and M.P. van Dijk. 1989. Semiarid Indian Agriculture: Land, Water and Crop
Management in Southern India. Bulletin No. 316. Amsterdam: The Royal Tropical
Institute.
Rao, M.V. 1976. Wheat production prospects in India. In Wheat in the Changing Social Context.
New Delhi: Government of India.
Renkow, M. 1991. Modeling the Aggregate Effects of Technological Change on Income Distribution
in Pakistan's Favored and Marginal Production Environments. CIMMYT Economics
Paper No.4. Mexico, D.F.: CIMMYT.
Ryan, J.G., and R. Sarin. 1981. Economics of technology options for deep vertisols in the
relatively assured rainfall regions of the Indian semi-arid tropics. Paper presented at
a Seminar on Management of Deep Black Soils, New Delhi.
Scobie, G.M. 1984. Investment in Agricultural Research: Some Economic Principles. CIMMYT
Economics Working Paper. Mexico, D.F. CIMMYT.
Sharpley, A.N., and J.R. Williams. 1990. EPIC-Erosion/Productivity Impact Calculator: 1. Model
Documentation. USDA Technical Bulletin No.1768. Washington, D.C.: United States
Department of Agriculture.
Sheopuria, R.R. 1990. A write up on wheat facts and trends in marginal environments of
Madhya Pradesh, India. Powerkheda, India: Krishi Vishwa Vidyalaya Zonal
Agricultural Research Station. Mimeo.
Singh, R. P., and S.R. Raje. 1984. Optimistic production in rainfed vertisols: A case study of
an operational research project. Fertilizer News 29(4): 59-71.
52




Singh, R.P., B.L. Jain, and V. Bhaskar Rao. 1984. Features of Traditional Farming Systems in
Selected Villages of Madhya Pradesh. Economics Group of Resource Management
Programme, Progress Report 61. Patancheru, India: International Crops Research
Institute for the Semi-Arid Tropics.
Sinha, S.K., P.K. Aggarwal, and R.K. Chopra. 1985. Irrigation in India: A physiological and
phenological approach to water management in grain crops. Advances in Irrigation 3:
129-212.
Upadhyaya, Y.M., K.N. Ruwali, and K.S. Murty. 1986. Evaluation of genetic improvement in
wheat cultivars under stress conditions. In New Genetical Approaches to Crop
Improvement, K.A. Siddiqui and A.M. Faruqi (eds.). Karachi, Pakistan: PIDC Printing
Press.
Virmani, S.M., M.R. Rao, and K.L. Srivastava. 1985. Approaches to the management of
vertisols in the semi-arid tropics: The ICRISAT experience. In Management of Vertisols
for Improved Agricultural Production. Proceedings of an IBSRAM Inaugural
Workshop. Patancheru, India: International Crops Research Institute for the SemiArid Tropics.
Walker, T.S., and J.G. Ryan. 1990. Village and Household Economics in India's Semi-Arid Tropics.
Baltimore: Johns Hopkins.
Williams, J.R., P.T. Dyke, W.W. Fuchs, V.W. Benson, O.W. Rice, and E.D. Taylor. 1990. EPICErosion/Productivity Impact Calculator: 2. User Manual. USDA Technical Bulletin No.
1768. Washington, D.C.: United States Department of Agriculture.
Recent Publications from CIMMYT Economics
Working Papers
90/01 The Design and Management of Call System Training in On-Farm Research (R.
Tripp, P. Anandajayasekeram, and G. Sain) (Also available in Spanish)
90/02 Economic Losses from Karnal Bunt of Wheat in Mexico (J.P. Brennan and E.J.
Warham, with J. Herndndez, D. Byerlee, and F. Coronel) (Also available in Spanish)
90/03 Public and Private Investments in Maize Research in Mexico and Guatemala (R.G.
Echeverrfa) (Also available in Spanish)
90/04 Sources of Growth in Wheat Yields in Pakistan's Punjab: Is There a Sustainability
Issue? (D. Byerlee and A. Siddiq)
90/05 The Maize Subsector in Paraguay: A Diagnostic Overview (M.L. Morris with M.
Alvarez and M.A. Espinoza) (Also available in Spanish)
90/06 Technical Change, Productivity, and Sustainability in Irrigated Cropping Systems of
South Asia: Emerging Issues in the Post-Green Revolution Era (D. Byerlee)
91/01 Land Prices, Land Rents, and Technological Change: Evidence from Pakistan (M.
Renkow)
53




91/02 Wheat and Barley Production in Rainfed Marginal Environments of West Asia and
North Africa: Problems and Prospects (A. Belaid and M.L. Morris)
91/03 Extension of On-Farm Research Findings: Issues from Experience in Southern Africa
(A. Low, C. Seubert, and J. Waterworth)
91/04 Chimanga Cha Makolo, Hybrids, and Composites: An Analysis of Farmers' Adoption
of Maize Technology in Malawi, 1989-91 (M. Smale, with Z.H.W. Kaunda, H.L.
Makina, M.M.M.K. Mkandawire, M.N.S. Msowoya, D.J.E.K. Mwale, and P.W.
Heisey)
92/01 Economic Criteria for Establishing Plant Breeding Programs (J.P. Brennan)
92/02 Technical Change and Wheat Productivity in the Indian Punjab in the Post-Green
Revolution Period (D.S. Sidhu and D. Byerlee)
92/03 Rainfed Maize Production in Mexico: Trends, Constraints, and Technological and
Institutional Challenges for Researchers (A. Hibon, with B. Triomphe, M.A. L6pezPereira, and L. Saad)
92/04 Returns to Wheat Research in Nepal (M.L. Morris, H.J. Dubin, and T. Pokhrel)
92/05 Dryland Wheat in India: The Impact of Technical Change and Future Research
Challenges (D. Byerlee in collaboration with the Indian Council of Agricultural
Research)
Economics Papers
1 Determining Comparative Advantage Through DRC Analysis: Guidelines Emerging
from CIMMYT's Experience (M.L. Morris) (Also available in Spanish)
2 Triticale Production in the Central Mexican Highlands: Smallholders' Experiences
and Lessons for Research (J. Carney) (Also available in Spanish)
3 Continuous Economic Analysis of Crop Response to Fertilizer in On-Farm Research
(M.A. Jauregui and G.E. Sain)
4 Modeling the Aggregate Effects of Technological Change on Income Distribution in
Pakistan's Favored and Marginal Production Environments (M. Renkow)
54




f { ~ A
A
A
A
4
A
A A A S V
A
A
A
A
A
A




COPYRIGHT NOTICE
@ CIMMYT International Maize and Wheat Improvement Center The CIMMYT International Maize and Wheat Improvement Center holds all rights to the source text and shall be considered the copyright holder for the text and images of these publications.
The CIMMYT International Maize and Wheat Improvement Center has made this publication available to the University of Florida, for purposes of digitization and Internet distribution. The CIMMYT International Maize and Wheat Improvement Center reserves all rights to this publication. All uses, excluding those made under "fair use"~ provisions of U.S. Code, Title 17, Section 107 are restricted. Contact the CIMMYT International Maize and Wheat Improvement Center for additional information and permissions.