wre@oo %o( R300 WORLD BANK TECHNICAL PAPER NUMBER 300 Technologies for Rainfed Agriculture in Mediterranean Climates A Review of World Bank Experiences Peter A. Oram and Cornelis de Haan EED: 2MHDDITIE' iUEVE _ BON AND TI ICO *NV2RONMI ME MEl VEog Eli 'xi 'VE RES f < ' - < VE RI. Wi GIL. Ag XRwlP U.... . .'LANUATURE LAND TEUEM RECENT WORLD BANK TECHNICAL PAPERS No. 226 Bindlish, Evenson, and Gbetibouo, Evaluation of T&V-Based Extension in Burkina Faso No. 227 Cook, editor, Involuntary Resettlement in Africa: Selected Papersfrom a Conference on Environment and Settlement Issues in Africa No. 228 Webster and Charap, The Emergence of Private Sector Manufacturing in St. Petersburg: A Survey of Firms No. 229 Webster, The Emergence of Private Sector Manufacturing in Hungary: A Survey of Firms No. 230 Webster and Swanson, The Emergence of Private Sector Manufacturing in the Former Czech and Slovak Federal Republic: A Survey of Firms No. 231 Eisa, Barghouti, Gillham, and Al-Saffy, Cotton Production Prospectsfor the Decade to 2005: A Global Overview No. 232 Creightney, Transport and Economic Performance: A Survey of Developing Countries No. 233 Frederiksen, Berkoff, and Barber, Principles and Practicesfor Dealing with Water Resources Issues No. 234 Archondo-Callao and Faiz, Estimating Vehicle Operating Costs No. 235 Claessens, Risk Management in Developing Countries No. 236 Bennett and Goldberg, Providing Enterprise Development and Financial Services to Women: A Decade of Bank Experience in Asia No. 237 Webster, The Emergence of Private Sector Manufacturing in Poland: A Survey of Firms No. 238 Heath, Land Rights in Cote d'lvoire: Survey and Prospectsfor Project Intervention No. 239 Kirmani and Rangeley, International Inland Waters: Concepts for a More Active World Bank Role No. 240 Ahmed, Renewable Energy Teclhnologies: A Review of the Status and Costs of Selected Technologies No. 241 Webster, Newly Privatized Russian Enterprises No. 242 Barnes, Openshaw, Smith, and van der Plas, What Makes People Cook with Improved Biomass Stoves? A Comparative International Review of Stove Programs No. 243 Menke and Fazzari, Improving Electric Power Utility Efficiency: Issues and Recommendations No. 244 Liebenthal, Mathur, and Wade, Solar Energy: Lessons from the Pacific Island Experience No. 245 Klein, External Debt Management: An Introduction No. 246 Plusquellec, Burt, and Wolter, Modern Water Control in Irrigationi: Concepts, Issues, and Applications No. 247 Ameur, Agricultural Extension: A Step beyond the Next Step No. 248 Malhotra, Koenig, and Sinsukprasert, A Survey of Asia's Energy Prices No. 249 Le Moigne, Easter, Ochs, and Giltner, Water Policy and Water Markets: Selected Papers and Proceedingsfrom the World Bank's Annual Irrigation and Drainage Seminar, Annapolis, Maryland, December 8-10, 1992 No. 250 Rangeley, Thiam, Andersen, and Lyle, International River Basin Organizations in Sub-Saharan Africa No. 251 Sharma, Rietbergen, Heimo, and Patel, A Strategyfor the Forest Sector in Sub-Saharan Africa No. 252 The World Bank/FAO/UNIDO/lndustry Fertilizer Working Group, World and Regional Supply and Demand Balancesfor Nitrogen, Phosphate, and Potash, 1992/93-1998/99 No. 253 Jensen and Malter, Protected Agriculture: A Global Review No. 254 Frischtak, Governance Capacity and Economic Reform in Developing Countries No. 255 Mohan, editor, Bibliography of Publications: Technical Department, Africa Region, July 1987 to April 1994 No. 256 Campbell, Design and Operation of Smallholder Irrigation in South Asia No. 258 De Geyndt, Managing the Quality of Health Care in Developing Countries No. 259 Chaudry, Reid, and Malik, editors, Civil Service Reform in Latin America and the Caribbean: Proceedings of a Conference No. 260 Humphrey, Payment Systems: Principles, Practice, and Improvements No. 261 Lynch, Provision for Children with Special Educational Needs in the Asia Region No. 262 Lee and Bobadilla, Health Statisticsfor the Americas No. 263 Le Moigne, Subramanian, Xie, and Giltner, editors, A Guide to the Formulation of Water Resources Strategy No. 264 Miller and Jones, Organic and Compost-Based Growing Mediafor Tree Seedling Nurseries (List continues on the inside back cover) WORLD BANK TECHNICAL PAPER NUMBER 300 Technologies for Rainfed Agriculture in Mediterranean Climates A Review of World Bank Experiences Peter A. Oram and Cornelis de Haan The World Bank Washington, D.C. Copyright C) 1995 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing November 1995 Technical Papers are published to communicate the results of the Bank's work to the development com- munity with the least possible delay. The typescript of this paper therefore has not been prepared in accor- dance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibili- ty for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatso- ever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy por- tions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of sub- jects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Ina, 75116 Paris, France. ISSN: 0253-7494 Peter A. Oram is research fellow emeritus at the International Food Policy Research Institute, Washington, D.C. Cornelis de Haan is livestock adviser in the Agriculture and Forestry Systems Division of the World Bank's Agriculture and Natural Resources Department. Library of Congress Cataloging-in-Publication Data Oram, Peter. Technologies for rainfed agriculture in Mediterranean climates: a review of World Bank experiences / Peter A. Oram and Comelis de Haan. p. cm. - (World Bank technical paper, ISSN 0253-7494; no. 300) Includes bibliographical references (p. ). ISBN 0-8213-3433-6 1. Dry farming-Developing countries. 2. Agricultural innovations-Developing countries. 3. Agricultural development projects-Developing countries. 4. Mediterranean climate. I. Haan, C. de. II. Title. III. Series. SB110.0724 1995 338.1'6'091724-dc2O 95-37034 CIP Contents Foreword ....... ...................................................................... vii Abstract ............................................................................. ix Abbreviations And Terms ............................................................................. xi Acknowlegements ............................................................................. xiii Executive Summary ............................................................................. xv CHAPTER 1 INTRODUCTION ..............................................................................I CHAPTER 2 LOW-RAINFALL ZONES: CLIMATE, TOPOGRAPHY, SOILS AND LAND USE ..............5 MEDITERRANEAN ARID AND SEMI-ARID WARM SUB-TROPICAL WINTER RAINFALL ZONES ............5 Climate and Topography ..............................................................................5 Subclimates in RainfedAgricultural Areas ..............................................................................6 Climate Areas in WANA ..............................................................................7 Climate Areas in China .............................................................................8 SOILS AND LAND SUITABLILTY ......................................................................................................................8 Soils in WANA and China .....................................................8 Land Agricultural Suitability .....................................................9 Land Use Across Agricultural Zones ................................................................................ 10 CHAPTER 3 FARMING SYSTEMS IN WEST ASIA AND NORTH AFRICA .13 FARMING SYSTEMS IN MARGINAL RAINFED ENVIRONMENTS .13 Cereal-Livestock Systems .13 Technical Options for Cereal-Livestock Systems .15 FARMING SYSTEMS IN SEMI-ARID ECOLOGICAL ZONES .23 Increasing the Contribution of Pulses.24 Introduction of Other Crops .27 Replacement of Food Legumes with Fodder Crops .28 Mountain Watersheds.28 CHAPTER 4 LAND RECLAMATION AND SOIL AND WATER CONSERVATION .31 LAND RECLAMATION ... 31 Destoning.31 SoilAmendment .................................................................................................................................I ............. 33 SOIL AND MOISTURE MANAGEMENT AND CONSERVATION .34 Conservation Tillage.35 Contour farmingg.38 Terracing.39 Veegetative Cover.39 Conclusionsfrom Project Experience .40 WATER HARVESTING AND USE .................................................................................................................... 43 Water Harvesting ..................................................... 44 Supplementary Irrigation .................................................... 48 iii ASSESSING THE COSTS AND BENEFITS OF SOIL AND WATER CONSERVATION MEASURES ............ S0 CHAPTER 5 PROSPECTS FOR TECHNOLOGICAL CHANGE IN RAINFED CROP PRODUCTION.. 55 FACTORS INHIBITING TECHNOLOGICAL CHANGE ......................................... ................................ 55 GENETIC IMPROVEMENT OF KEY FOOD CROPS .............................. ............................................ 57 Incorporating Tolerance ofAbiotic Stress ........................................................................... 58 Combating Biotic Stress Factors ...................... 59 Meeting Farmer and Consumer Preferences Through Breeding ............................................... 61 CROP PROTECTION ............................................. 63 Changes in Farming Practices ............................................... 63 Use Of Resistant Varieties ................. 63 Chemicals .......,...63 Biological Pest and Disease Control ................................................ 67 Integrated Pest Management ................................................. 67 FERTILIZER USE ............................................... 69 Factors Influencing Use of Fertilizer ............................................. 70 Fertilizer Response: Experimental Data ............................................ 71 Increasing Fertilizer Use Efficiency in World Bank Projects ............................................ 72 Co.onclusions ............................................. 74 MECHANIZATION ........................................... 75 Machinery Specification in Project Documents ............................................ 77 Major Gaps in Bank Provision for Farm Mechanization ............................................. 77 Mechanization and WANA ............................................ 79 Bank Support for Investment in Machinery Contracting ............................................. 80 Environmental Effects of Mechanization ..............................,,.. 81 Optimum Balance Between Mechanization and Animal Draft ................................................ 81 Future Demand For Farm Machinery ................................................. 82 Considerationsfor WANA ................................................ 83 CHAPTER 6 ANIMAL PRODUCTION ............................................... 85 IMPROVING NUTRITION ............................................... 88 Increasing Domestic Feed Production ................................................. 89 Improving Natural Pasture ................................................ 90 Bank Experiences ................................................ 91 Lowland Steppe and Marginal Rainfall Rangelands ................................................ 92 Bank Experience ................................................ 96 OPTIONS FOR TECHNOLOGY IN MARGINAL STEPPE AREAS ............................................... 97 Ammonium or Urea Treatment of Straw ..................................................100 Policy and Socio-Economic Measures ................................................ 101 Bank Experiences ................................................. 102 MAINTENANCE OF ANIMAL HEALTH ............................................... 102 Animal Health Provisions in Bank-Funded Projects ................................................ 103 Availability of Vaccines and Drugs ................................................. 104 Reaching Farmers and Herders ................................................ 104 Payingfor Animal Health Services ................................................. 105 New Technologies in Veterinary Medicine ................................................ 106 ANIMAL BREEDING, GENETICS, AND BIOTECHNOLOGY ............................................... 106 Genetic Potential of Local Breeds in Semi-arid Regions ................................................. 109 Breeding Strategy .................................................111 Artificial Insemination ................................................ 113 Potential of Biotechnology in Animal Breeding ................................................. 114 Embryo Transfer ................................................ 115 iv Hormone and Immunology Technology .............................. 116 Future Perspectives .............................. 116 Conservation ofAnimal Genetic Resources .............................. 117 Conclusions .............................. 118 REFERENCES ..................................................................................................................................................F121 ANNEX: DATA ................................ 133 LIST OF FIGURES 1.1 Types of Projects Reviewed ..............................................................................................................................................2 1.2 Countries Represented ......................................2 3.1 Changes in Producer Price Ratios in Algeria, 1970 and 1987 ................................................................................ 14 6.1 North Africa: Changes in Cattle, Sheep and Goat Population 1961-65, 1971-81 and 1980-90 .................................. 86 6.2 West Asia: Changes in Cattle, Sheep and Goat Population 1961-65, 1971-81 and 1980-90 ...................................... 86 LIST OF BOXES 3.1 Is Fallowing Necessary? ................................................................................ 17 3.2 Options for Forage Production in WANA ................................................................................ 18 4.1 Main Water and Conservation Techrniques ................................................................................ 35 4.2 Conservation Tillage in the United States ................................................................................ 36 4.3 Features of Water Harvesting ................................................................................. 44 4A.4 Benefits of Supplemental Irrigation in Syria ................................................................................ 49 4.5 The Mexican Case Study ................................................................................ 53 4.6 The Role of the Bank in Financing Soil and Water Conservation ................................................................................ 54 5.1 Experience w ith Triticale in WANA ................................................................................ 62 5.2 Adoption Factors in Herbicide Use ................................................................................. 65 5.3 The Limitations of IPM ................................................................................................................................................. 68 5.4 Packages Recommnended in Bank Projects ................................................................................. 76 5.5 Mechanization Policy in WANA ................................................................................. 79 6.1 Community Range Development in the Middle Atlas ................................................................................ 93 6.2 Fodder Species ................................................................................. 98 LIST OF TEXT TABLES 2. 1 Classification of Mediterranean Bioclimatic Zones in West Asia by Precipitation ............. ...............................6 2.2 Land Use in West Asia and North Africa, 1989 ................................................................................. 10 3. LAverage Yields of Barley Grain in the Different Agricultural Stability Zones of Syria 1979-89 ...................... 15 3.2 Use and Environmental Adaptation of Different Species of Lathyrus and Vicia in the West Asia and North Africa Region and Priority Research Objectives for these Crops at ICARDA ............ ............................ 22 3.3 Comparative Annual Yieldsfor Cool-Season Food Legumes and Cereals, WANA and Worldwide 1989-91 ................................................................................ 24 5. 1 Relationship Between Area, Percent of Crosses Made, and Percent of Scientists by Wheat Type, 1990 ..................... 56 5.2 Average Ranking of Breeding Priorities for Spring Bread Wheat by 100 National Research Programs ................................................................................ 59 5.3 Farmers Opinion of Triticale Compared with Durum Wheat and Barley .................................................................... 62 5.4. Average Fertilizer Consumption in WANA, 1961-63 and 1988-90 .............................................................................. 69 5.5 Nitrogen-to-wheat Price Ratios in Selected Countries 1991-1992 ............................................................................... 70 5.6 Number of Tractors and Combine Harvesters in Relation to CultivatedArea in Selected Regions and Countries, 1989 ................................................................................ 76 v 6.1 Productivity ofRegional Sheep Populahons, 1981 ...................................................................... 87 62 Productivity ofRegional Goat Populations, 1981 ....................................................................... 87 6.3 Important traitsfor Sheep and Goat Production in Developing Countries ................................................................ 112 6.4 Possible Biotechnical Solutions to Current Problems in Developing Country Livestock Production ........................ 119 vi Foreword The introduction of improved agricultural technology to develop sustainable production systems will be a key factor in determining whether the world's future population will be adequately nourished and clothed. This is especially the case in the more resource-poor areas such as the rainfed cool subtropics found in the Mediterranean basin and Northern China. In these less favorable environments, the growing population pressure will assure that the downward spiral of poverty and resource degradation will continue unless special attention is paid to developing sustainable farming systems. In low-potential areas, the development of technology is hindered by the low and variable rainfall and extreme diversity of soil resources--which greatly reduces wide application of any technology-- and by the later start of international agricultural research efforts in this domain. Furthermore, agricultural research has only recently started to focus on risk reduction in spite of the priority that smallholders in such resource-poor environments attach to the reduction of risk. Finally, crop and livestock research has mainly been carried out in separate institutes on a commodity basis, neglecting the strong linkage between crop and livestock in these areas. Tlhis technical paper is part of a general assessment of available technologies, and how they can be transferred into World Bank lending operations, that the Agriculture and Natural Resources Department has been undertaking for over the last few years. The paper reviews the adequacy of present crop and livestock technologies, identifies technical and socio-economic constraints to their adoption, and seeks to identify prospective technologies for pilot testing and full-scale introduction in future World Bank- funded projects. As such it is directed at World Bank task managers, although it should also be of interest to policymakers, researchers and others involved with the promotion of sustainable agricultural development. Alex F. McCalla Director Agriculture and Natural Resources Department vii I Abstract This document provides a review of past expericnces with the introduction of agricultural technologies in World-Bank funded projects in Mediterranean climates, with the emphasis on North Africa and the Middle East. It is intendcd for planners involved in agricultural development of the region. First, it describes the recent trends in the farming systems of the region and the--only partially successful- past attempts to change the prevailing mono-culture cereal systems with legume/cereal farming systems. "Pipeline" technologies are then described, which might offer new opportunities to diversify currcnt mono- culture systems and thus make the overall farrming system more sustainable. Second, the document reviews the experiences with specific technologics in the area of land, crop and livestock improvement and provides an assessment of their impact and the reasons for the successes and failures in their introduction in World Bank-funded and other projects. Thirdly, it provides an overview of the use of technology by World Bank staff, assessing the specificity, scope, and physical as well as socio-economic and environmental appropriateness of the technologies used. It concludes, that while a good array of "best bet" technologies is available, more attention is warranted by World Bank staff in making the correct choices. ix Abbreviations and Terms ACSAD Arab Center for the Studies of Arid Zones and Dry Lands Al artificial insemination AOAD Arab Organization for Agricultural Development billion 1.000 million BNF biological nitrogen fixation CGIAR Consultative Group for International Agricultural Research CIMMYT Centro Internacional de Mejoramiento de Maiz y Trigo CSC cottonseed cake ERR economic rate of return ha hectare(s) HP horsepower IARC international agricultural research center ICARDA International Center for Agricultural Research in the Dry Areas ICRISAT International Crops Research Institute for the Semi-Arid Tropics kg kilogram(s) ILRI International Livestock Research Institute K potassium LU livestock unit(s) m meter(s) mm millimeter(s) mt metric ton(s) N nitrogen P phosphorus pH a measure of the acidity or alalinity of a substance on a scale of 0-14 (7 equals neutral) PAR project appraisal report PCR project completion report PPAR Project Performance Audit Report SAR staff appraisal report SI supplementary irrigation t ton TLU tropical livestock unit(s) WANA West Asia and North Africa Please note that unless indicated otherwise, all dollar amounts are current U.S. dollars. xi Acknowledgements The authors are grateful for the inputs of many individuals in the preparation of this paper. Jitendra Srivastava and Hamdy Eisa of the Bank and Jacques Brossier of INRA, France greatly contributed to the initial discussions on the scope and outline of the pper. Staff from INRA Morocco, especially R.K Bansal and 0. El Gharras on agricultural mechanization and on adoption of crop and livestock technology in Morocco were especially helpful. Peer reviewers for this paper were John Cole and Johannes ter Vrugt. xiii Executive Summary This report assesses the effectiveness of crop and livestock technologies in Bank projects, for agricultural areas with a Mediterranean climate.' These zones comprise about 25 percent of the developing world's land area and are inhabited by some 750 million people. In some countries, especially in west Asia and North Africa ( WANA), these climatic zones cover 75-100 percent of the land area. They contain about 25 percent of the ruminant livestock population of the developing world. Much of the zone is characterized by a fragile natural resource base and by risk-prone and resource-poor farners. Soils are often degraded, low in fertility, organic matter and water-holding capacity, and are easily susceptible to wind and water erosion. Runoff losses during rainfall commonly exceed 50 percent. Because of the growing population pressure on these less-favorable environments, the downward spiral of poverty and resource degradation will continue unless technologies are developed to make the farming systems of these areas technically, socially, and economically sustainable. FARMING SYSTEMS In China and WANA, land and water resources are currently under heavy pressure from expanding human and animal populations, and little land remains that can be brought into cultivation. The future emphasis in agricultural development should thus be on raising the productivity of the existing resource base by improving water use efficiency and soil management and by raising yields and cropping intensities. Since 1970 there have been important changes in land use with three main trends. These are: * Greater diversification of annual crop rotations in higher rainfall (greater than 400 mm annual precipitation) and irrigated land, with less dependence on cereals and reduced fallow. * An expansion of tree crop farming in all zones with less than 250 mm annual precipitation especially on sloping land that is unsuited to annual cropping * Increasing monoculture of cereals (mainly barley) in areas with annual rainfall of less than 400 mm with reduced fallow and cultivation encroaching on marginal grazings ( barley is the key crop but it is intended for consumption by runinant animals rather than by humans; warning signs of falling barley yields in marginal ecozones suggest that this system may not be sustainable). Theoretically, the most attractive alternative to barley mono-culture at the system level in zones with less than 350 mm annual precipitation is a food or forage legume, since legumes fix atmospheric I Low-rainfall areas (less than 600 mm of annual precipitation) with winter rainfall either in a sub-tropical ( West Asia and North Africa (WANA) and Latin America) or a cold winter (Eastern Turkey, northern Iran, Afghanistan) and in northern China. xv nitrogen to help maintain cereal yields, and oilseeds, the other potential altemative crop, are not sufficiently drought-tolerant. Thus, in the past, considerable emphasis in WANA countries was on fallow replacement, with barley, pulses, and ley farming based on medics. These changes have generally not been successful, however, becaise in spite of considerable efforts by scientists, local government and international institutions, food legumes occupy only 4.5 percent of the total crop land, and forages occupy even less. The main reason for this disappointing result is that food legumes are perceived by the farmers as more risk-prone than cereals, with their greater susceptibility to day-length and temperature variability and to drought, salinity, diseases, and weeds. Legumes also have harvesting problems. Furthermore, in many countries, food legume prices are not subsidized, whereas wheat usually is. Wheat and sometimes barley markets are more or less guaranteed by governments, whereas those of pulses usually are not, and prices and markets for fodder are purely opportunistic. The most positive World Bank-supported experience so far has been in Turkey, where a massive and well-integrated research, extension, seed production, credit, and marketing effort to encourage the replacement of fallow or continuous cereals by a cereal food legume system has resulted in an overall reduction of the large fallow area by 33 percent in about seven years. This was on relatively favorable terrain, however, in an area with 350 mm-450 mm annual rainfall. Farmers in less favorable areas have been more reluctant to give up their cereal area or even to substitute legumes for fallow. Similarly, long and well-financed efforts (including by the Bank) to introduce the Australian ley farmiing system with reseeding annual forage legumes (medics) have met again with limited success. Biological (lack of adapted cultivars and complex agronomic practices requirements) and social (uncertainties of land ownership, grazing rights, and grazing management of the fallow crop) constraints were the main reasons for the lack of adoption of this system. The future (if any) of this system lies on larger farms. On the basis of recent technology developments, two altemative farming systems might now develop. * Improved harvesting techniques for lentils and leguminous crops, high yielding winter chickpeas cultivars with better resistance to the diseases Ascochyta blight might change the farmers' reluctance to adopt legumes and the barley-pulse system might develop as an acceptable alternative for the 250 mm-350mm annual rainfall area. For the zones down to 200 mm annual precipitation, the recently developed more robust annual forage legume such as Narbonensis or Lathurus sativa might develop as an alternative system. * On the other hand, the new barley ecotypes with higher biomass and good drought and cold tolerance could reinforce their preference for barley, and might lead to a barley mono-culture system. This would still require limited fertilizer use to prevent a decline in yields from mining the soils. The impact of continuous barley cultivation on disease built-up has not yet been adequately researched and should thus be carefully monitored. It should be noted, however, that the social and other constraints that impede the adoption of ley farming in WANA appear less likely to hamper its use in South America, where private farns are larger, or with cooperatives in Northwest China, which are more mechanized. LAND RECLAMATION AND SOIL AND WATER CONSERVATION The scope for land reclamation is limited in WANA. The main technology for land improvement promoted in Bank projects have been destoning, which has proved very popular with farmers. xvi Technologically, experience from the various projects, especially in North Africa (Morocco), suggests that the critical factors determining success of destoning are good soil quality as determined by prior soil surveys, the combined use of mechanical techniques for heavy clearance and ripping, manual collection fDr stone removal in carts or tractor trailers, and the terracing or other conservation techniques once stones are removed from sloping land ( in order to prevent erosion). Experience shows also that destoning is most successful where owner-farmers are the main beneficiaries; cooperative and collective efforts at destoning generally performed poorly. Financially, 1995 costs in Morocco using contractors were of the order of $1,000/ha for machinery work and $500/ha for labor. Some costs are recovered by sale of the stones. An estimated investment cost below $2,0501ha or a stone density under 450 m3/ha is required to permit a satisfactory economic rate of return (ERR) of 10 percent. Water harvesting and use have received little attention in Bank-funded projects. This review discusses different measures to improve water use and efficiency, stressing the need for an integrated approach. Supplementary irrigation has the potential to greatly improve yields, however; it involves a learning process, which farmers find difficult. Excessive water use is common. Results in land conservation projects in the United States and Mexico and in Bank-funded projects in WANA show that investments simply to control erosion are often unprofitable, even at low discount rates or soil erosion tolerance levels. There are significant differences between sites, however, and pilot operations are recommended to test techniques and estimate and demonstrate benefits. A majority of the Bank projects reviewed had mechanization components; in several cases these were the largest single item. Equipment was provided for destoning, land clearance, conservation works, and on-farm uses. Generally, the emphasis was on tractors and tillage implements, combine harvesters, and small irrigation equipment. Equipment to improve efficiency of input use (such as seed drills, ferdlizer applicators, and sprayers) as well as forage harvesters and animal-drawn equipment were generally neglected. The important issue of the complementarity and balance of animal draft and mechanized equipment and the potential role of contractors in bridging the gaps was largely ignored. CROP TECHNOLOGIES This paper examines available technology for crop improvement, crop protection including weed control and fertilizer use. The development of improved cultivars has until recently focused on bread wheat for more favorable environments. Barley has received less attention. Furthermore, the adoption rates of modem cultivars of barley in WANA and China is lower than in other regions of the world. Recent surveys show that farmers are generally well aware of the existence of the modem varieties; however, seed availability is sometimes a problem. Even more importantly, the farmers feel that the local varieties outyield the modem ones, especially under drought conditions, and have preferred consumer qualities (including straw and grazing qualities). Better cold-resistant varieties of barley are now being developed. Careful scrutiny of eventual new varieties on these characteristics is thus recommended, when designing a project on new varieties. A good potential exists to improve crop management techniques at the farm level in WANA. Deep plowing and the use of a rototiller, which can have detrimental effects on soil conditions, is often practiced. Conservation tillage, which would be more beneficial for the soil, is impeded by the lack of special equipment, and more importantly by farmer needs to save weeds and straw for livestock feeding. Other techniques can also be improved. * Seed rates for cereals are often excessively high. Research shows that lower rates would economize on high seed costs without decreasing yields. xvii * Relatively few farmers treat their seeds against diseases and pests before sowing them. This is a cheap and effective means of reducing losses which does not require expensive machinery, and can be done on-farm. * Weeds cause heavy losses with cereals and especially with food legumes. Yet weeds are often poorly controlled, and in low rainfall areas up to 70 percent of farmers do not weed at all, because weeds provide feed for livestock or salad material for home consumption. The use of herbicides is very limited, due both to ignorance of their value and to lack of cheap, efficient, modem sprayers. In view of the large returns attributable to efficient weed control, Bank-funded projects need to pay more attention to extension and spray equipment. * Fertilizer use is another important potential source of productivity growth. Methods of application are inefficient, application levels are low in upland areas, and in most semi-arid/arid zones many farmers use none. Yet regional trials on barley show average increments of 50-100 percent in grain yields, even where annual rainfall is 250 mm or below. Farmers perceive fertilizers as too costly, however, even though nitrogen-to-wheat price ratios are generally more favorable in WANA and South America than in other regions. Insufficient rainfall and risk aversion are major reasons for non-adoption. Considerable progress has been made in WANA at low costs to farmers in crop protection through genetic resistance to biotic stresses-for example against stem rust of wheat-and significant advances are being reported from biological pest control. In addition to the limited weed control and seed treatment of cereals, other chemical controls are also not widely used in low rainfall zones. Two serious problems for which satisfactory controls have yet to be found are the parasitic plant Orobanche crenata, and the Ascochyta blight, both of which seriously damage legumes. An integrated pest management (1PM) approach with a combination of genetic, biological, chemical, and cultural controls is being tried with Orobanche. ANIMAL PRODUCTION In all regions with Mediterranean climates, livestock are of primary importance. Recent surveys in the principal WANA countries show that more than 70 percent of the farmers in areas with annual rainfall below 350 nmm rely on small rnminants as their principal source of income, especially on small farms, and animal production represents 30-40 percent of agricultural GDP in WANA. The combination of continuing strong demand for meat and milk with an inelastic land resource and rising imports of feed grains poses particularly severe problems for the WANA region. If escalating meat imports are to be avoided, both productivity per head and domestic feed supplies must be increased. The number of small runinants has also risen by about 70 percent since 1961-65 in China, with a large share of the increase in small ruminants is probably in the colder and drier areas of northern China, where pastures are a major component of land use. In South America, cattle is the dominant species, although their number have declined over the last decade. The production depends heavily on the export markets, which have been greatly depressed by dumping from the European Union (EU). The recent reforms under the General Agreemtne On Trade And Tarrifs (GATT) might offer new opportunities for the South American beef exporters. Shortage of feed of the right quantity and quality is generally accepted as the most formidable technical constraint on livestock development in this region. Hence, the Bank has given priority to improving animal xviii nutrition in many projects in WANA as well as in northem China, and (in combination with veterinary measures), in South America. Improving Feed Options There are several options in this area. Improving the management of natural grazing. Stocking rates on range lands in WANA are reported to have fallen from 2.8 to 1.5 tropical livestock units per hectare since 1960, although opinions are divided on the extent to which the resource is irreversibility degraded. The situation has often been aggravated by govermment policies which have disturbed traditional management structures for the communal grazings without putting effective altematives in their place, and which have unwittingly provided incentives to overstocking, barley cultivation, and other detrimental practices. Best bet' Technology does in fact exist for the steppe range areas, but it is limited. The constraints imposed by the extremely low and unreliable rainfall preclude capital intensive investments in fencing and reseeding, and simple, relatively low-cost techniques are required, the results of which can be readily demonstrated to range users in order to gain their confidence and cooperation in applying them on open rangelands through communal grazing agreements. These techniques include: Passive techniques. This involves small block seeding techniques using wind and water forces to distribute seed. Deferred grazing would allow degraded areas to rest and regenerate. Phosphate application (25-60 kg ha). This is recommended for pastures, accompanied by deferred grazing to allow seed biomass to increase before returning to continuous grazing. It is promising and relatively low cost, especially for legume-rich swards. The effects are cunulative over the years. Micronutrients (e.g., molybdenum) can have dramatic effects, but have yet to be identified in WANA. Planting or preferably seeding perennial shrubs into steppe or upland rangelands. Numerous (mostly indigenous) species are being tested in WANA, including several of Atriplex, Acacia, ATtemesia, Opuntia, and Salsola. Seeds can be sown into shallow pits or furrows, and a combined cultivar/seeder has been developed in Jordan. However established, they need protection for 3-5 years before grazing is permitted, and careful matching of species to environments is needed. Reseeding (usually with legumes). Sometimes this is supported by phosphoric fertilizer, and chemical weed control to eliminate noxious or unpalatable species. Improving conservation and utilization offorage crops. This includes conserving and utilizing forage crop by-products, including measures such as urea or ammonium treatment of straw to improve its digestibility and palatability. This has become standard farm practice in China, but it is not yet widely used in WANA. It involves treatment of straw with 5 percent urea in solution, or 3 percent anhydrous ammonia in order to increase its digestibility, but it is most effective when fed in conjunction with 2.5 kg/day cotton-seed cake or other protein supplement. In cattle the combined treatment raised daily liveweight gain by 41 percent, but there is little data on its use for sheep. This may be a 'best-bet' technique for WANA, but promotion through Bank-funded projects should be preceded by analysis of the availability and costs of materials and labor for recommended levels of application, and expected benefits. Extension training is also essential; there have been instances of poisoning in sheep in Egypt from improper use of the technology. xix Supplementary feeding of grain and other 'concentrates'. These can be usedeither to prevent catastrophic losses of aninals as a result of drought or to provide extra nutrients at critical periods of the production cycle such as breeding, pregnancy, and fattening. Drought relief programs with substantial feed inputs, while justified from a social and human viewpoint for range producers, might have a negative impact cn the rangelands, and may impede the normal range recuperation after the drought. "Stratification" of ruminant production across ecozones. This technique can take advantage of ecological complementarities and can especially reduce reliance on natural grazings during certain periods of scarcity. Stratification includes the establishment of feed lots raising slaughter weights of sheep to 40 kg/head from the current regional average of around 20 kg/head. So far, large-scale fattening operations have not done well in WANA due to a combination of technical, institutional, marketing, health, and policy factors, the latter include inadequate price differentials for fat sheep, high feed and pharrnaceutical price problems, inadequate research, and inexperienced extension staff. Fattening, of beef cattle by smallholders, on the other hand, has done very well in China. Experience shows that technical measures are unlikely to succeed where government policies give incorrect or contradictory signals, or where property rights are not clearly defined and understood by range users. Property rights vary both anong and within countries. It is essential for appraisal missions to undertake a carefil review of current government policies and institutions governing use and management of the range so as to identify the appropriate institutional framework to allow government to contribute most effectively without sfifling local initiatives. Improving Animal Health Nhile a range of ruminant diseases and metabolic disorders is identified, it is noted that the main factors affecting flock production of sheep in WANA are low lambing percentages and high young mortality. In China serious losses in animals and in productivity resulted from mineral deficiencies (sodium, selenium, and copper) and from untreated worms. In some Bank-funded projects, it seemed likely that improved nutrition or improved breeding (or both) would eliminate loss of productivity from disease. Experience shows that this is not the case-the combined action of better feeding and improved veterinary care are synergistic, while the introduction of exotic genes through breeding may actually increase susceptibility to disease, for example in WANA and China. Apart from better nutrition and direct measures to prevent diseases a number of Bank-funded projects included support for improving housing, shelter, and water supplies. In areas of severe heat, drought, or extreme cold, these measures may reduce vulnerability to disease. They may also raise efficiency by enabling animals to use their available feed for production rather than to cope with climatic stress. While it does not appear that lack of vaccines against key diseases is a major problem, non-technical factors, particularly high inflation rates, shortages of foreign exchange, or government cutbacks in funding of agriculture have led to shortages of medicines, cutbacks in veternary services, and reduced support to mobile clinics and extension staff mobility. Unifying crop and livestock services into one national extnsion system has not worked well for livestock, especially for support of disease prevention and artificial insemination (Al). The village-level workers are rarely well trained in animal management and are too preoccupied with other tasks to be the link between livestock technicians and the flock owners. The use of lay staff (paravets, herders) in the provision of veterinary services may be an alternative. To provide adequate services to smallholder producers can be a heavy burden on government budgets, consequently there is pressure to charge users for the services, and to privatize some of the veterinary tasks xx (clinical treatments). Generally these changes have been more successful than anticipated by the project authorities for those interventions felt critical by the herders. "Best bet" technologies specifically for animal health in low rainfall ecozones, will concentrate on vaccination against major endemic diseases and parasites (which can have a high pay-off particularly when combined with improved nutrition), and improvements in management techniques to reduce losses and unthriftiness in pregnant ewes, lambs and kids. "Pipeline" technologies include increased attention genetic resistance to internal parasites (especially in small ruminants) the increased use of multivalent and thermos-stabile vaccines. Animal Breeding And Genetics Progress in raising the productivity of ruminant livestock in developed countries has depended considerably on genetic improvement; in some cases, by creating new genetically stable breeds of cattle and sheep. Animal breeding has also played an important role in South America, especially Argentina and Uruguay. Consequently, Bank-funded projects have tended to allocate high priority to genetic improvement. Genetic improvement of ruminants may be a best-practice technology if it enables animals to perform better within constraints imposed by their environment, for example, greater stress tolerance, disease resistance, and better feed conversion. Importation of purebreds in an environment requiring a higher plane of nutrition, management, and health care than can realistically be foreseen within a decade is not advisable. Insufficient attention has been paid to characterizing and evaluating the potential of local genotypes, especially with respect to sheep and goats. There is a tendency in Bank-funded projects to rely almost exclusively on importation of Merino sheep and to ignore the attributes of local breeds and the potential for selective breeding and/or crossing among them. Artificial insemination (AI) is a valuable tool, but one to be used with discretion. It can be important in accelerating breed improvement in dairy cattle, where nutrition and health are not severely limiting their performance. It is not appropriate fpr large-scale genetic improvement of sheep or goats because of the high cost and practical difficulties. Itnportation of purebred animals and their distribution to small numbers of farmers within large animal populations, and without workable means of monitoring their future performance, is an inefficient use of resources and should be rigidly controlled. Instead priority should be given to building the institutional capacity, which could make eventual large-scale genetic improvement programs (self- financing AI service for cattle, milk recording and producer based selection system (open nucleus breeding system)). Biotechnology holds promise for the future with respect to livestock improvement, particularly in the use of embryo technology, and in inmmunization/vaccine development. At present, however, it is mainly a research tool. The use of embryo technology and growth hormone (BST) is not appropriate where inadequate management, nutrition, and health services are major impediments. CONCLUSIONS This study concludes that there is a considerable amount of proven technology available for further dissemination in Bank projects in WANA, especially for the higher-rainfall areas. This includes destoning, the use of improved wheat , barley and food legume varieties, better agronomic practices, pest and weed control and several animal health and breeding technologies. For similar regions in China, the urea-cottonseed-straw feeding technology has also proven to be quite promising. As can be expected the availability of technologies decreases as the aridity increases. xxi In addition, there are a number of "pipeline" technologies including improved varieties (especially those for less favorable environments), Integrated Pest Management (IPM), and livestock vaccines and breeding technologies based on molecular biology techniques for testing in future projects. Few technologies exists for the range lands of the low-rainfall areas. For those areas, the emphasis can only be on conservation and arresting the degradation now observed in many areas. Socio- institutional interventions, through the creation of producers organizations and allocation of grazing rights to such groups, and the establishment of appropriate policies are the main instruments for these areas. Some technical interventions, such as the introduction of fodder shrubs can be considered once these institutions and policies are in place. The record that emerges from this study regarding the appropriateness and correctness of the technological choices in Bank-funded projects is mixed. For example, the description of the technology to be used often lacks specificity, and technologies are often poorly justified. Plant varieties, fertilizer levels, and pest control methods are frequently stated in the most general terms ("improved varieties", use "appropriate levels of fertilizer" and "minimum or no pesticides"). Rates of fertilizer application vary greatly and their scientific basis is doubtful, due to inflexible recommendations based on research institute trials, lack of good soil testing services, and a narrow choice of fertilizer formulations with fairly high nutrient concentrations often containing potash as well as nitrogen and phosphate in a region with little potash deficiency. Most Bank-funded projects in WANA provide for fertilizer, often at rates of NPK which seem quite high for low rainfall zones, although their validity is sometimes difficult to assess for lack of location specificity. However, little is said about measures to ensure its efficient and timely application, to prevent avoidable losses, or to enhance the generally low organic matter status through nutrient cycling and biological nitrogen fixation. Among Bank projects there is limited use of available technologies; that is, the entire range of available technologies is not used or suggested. For example, few projects have given attention to supplementary irrigation, water harvesting ,and IPM . On the other hand, once a technology is proven in the first project, such as destoning, it is generally widely introduced in that region. Also, not enough attention is given to the social and economic framework in which the technology has to operate. On the social side, labor and land tenure constraints are often not given the appropriate weight, as demonstrated by the continued emphasis on the introduction of the ley system. Furthermore. sometimes the analysis and the proposed technologies are contradictory. For example, it was stated for Turkey that the main current need was equipment rather than more tractors; still the project supplied more tractors. There is a lack of coherent policies, especially in crop protection, as shown by the lack of a holistic approach in past control efforts, the vagueness in the prescription of pesticides and the unenforceable statements, such as "no pesticides will be used in this project". Environmental aspects of project interventions are often not sufficiently analyzed, for example in the effect of mechanization on soil condition and on the effect of pesticide use. Little attention was paid to adverse environmental or social effects of mechanization, and in no case was survey, monitoring, or evaluation of its impact provided for. xxii Chapter I INTRODUCTION Virtually all World Bank rural development projects include some components of agricultural technology. The majority depend heavily on the effectiveness of that technology for the achievement of the project's goals. Agricultural technology is important not only for Bank projects, however. In many developing countries the amount of land available for cultivation is shrinking fast. Many of the earlier means of increasing agricultural production and productivity, such as irrigation, are unattractive either because natural resources (such as water itself or the availability of irrigable sites) have been exhausted, or because the costs of these means are too high. Moreover, over the last decade, concerns about sustaining economic development by protecting the environment have emerged. Because of constraints on land development, increasing the productivity of the existing natural, and human resources is crucial to maintaining and expanding agricultural production. It is rarely possible to expand agricultural production at low cost without significant technological change. The two main instruments available to increase the productivity of existing resources are technology and human energy and skills. Although the performance of agricultural projects has improved recently, it has been disappointing in the past and the Bank has been concerned about understanding the reasons for poor results. One important aspect of this review is whether the best technological choices were made in Bank projects and whether technology was used to best advantage. This study, a first stage in agricultural technology assessment, seeks to survey available agricultural technologies, to examine their historical and current use in Bank projects, and to identify how new or existing technologies might be incorporated most effectively into future lending operations. It focuses on - Low-rainfall areas (less than 600 mm of annual precipitation) in the Mediterranean sub- tropical winter rainfall climate zone of West Asia and North Africa (WANA) and Latin America * The Continental cold temperate semi-arid and arid winter rainfall climate zone found in parts of WANA (eastem Turkey, northern Iran, Afghanistan) and in northern China. These zones were selected because they comprise about 25 percent of the developing world's land area and are inhabited by some 750 million people. In some countries, especially in WANA, arid and semi- arid lands constitute the vast majority (up to virtually all) of the land area. Arid and semi-arid zones contain about 25 percent of the ruminant livestock population of the developing world. Many of these dry areas are characterized by a fragile natural resource base and by poor farmers. Soils are often low in fertility, organic matter and water-holding capacity, and are very susceptible erosion. Runoff losses during rainfall commonly exceed 50 percent. The downward spiral of poverty and resource degradation will continue unless special attention is paid to developing sustainable farming systems for and and semi-arid lands. 2 The basic approach in this assessment consists of an in-depth review of the relevant World Bank projects in the geoclimatic regions referred to above, focusing on rainfed crops and livestock technologies. Seventy World Bank reports covering 52 projects in nine countries were reviewed. These reports are listed in annex 1 (table Al) and represented below in figures 1.1 and 1.2. Where possible, appraisal, performance audit and project completion reports were examined. A range of types of projects was covered, including credit (17 loans, including two sector adjustment loans) research and extension (8), rural development (6), agricultural development (8), livestock (6), fruit and vegetable production (three), watershed management (3), and land reclamation (1). The nine countries represented are Turkey (12 projects), Morocco (13), Argentina (4), Syria (4), Tunisia (7), China (4), Algeria (3), Chile (3), and Jordan (2). Most projects reviewed include FiReael. TYva,ofProieRetWI components involving crop management, input use, mechanization, and livestock production. WaterShed Mngmt Land Recl. The study is confined to Livestock 60% 2% Credit relatively recent projects, because they _1l2% 11 il ll l | | _ 32% give the best indication concerning the - l l _ | i | | | _Bank's use of up-to-date technology A&Dev. Thus, nine of projects reviewed were 15% started in the present decade, 21 in the Fnuit deg RwirDev. Research& 1980s, two in the 1970s, and two in the 6% 12% Extenion late 1960s. While there are advantages in including a substantial proportion of up- to-date project appraisal reports in the mix so as to be able to assess the Banks use of the latest technology in its Figure1.2 CounlriRepreser ted projects, there are also important benefits to be reaped from being able to review what actually happened during the life of Algeria Chile Turkey some of the projects by examining the China 6% 6 4%23% 8% ! project completion reports and the conclusions of the independent auditors. These can also be a valuable guide to the Bank's ability to learn from projects, Tu3nsia occo especially those with sinilar goals. Syria Argerntina 24% 8% 2% It is only possible to understand why a particular technological paradigm succeeded or failed by also understanding the setting in which it was expected to perform. A technology that may succeed in one situation may fail in another, as already shown by a similar analysis of technology for small-scale farmers in Sub-Saharan Africa (Carr 1989). Failure to adopt or use a technology effectively in one situation may be caused by lack of access to the necessary inputs or credit, by lack of markets for the product for which the technology was designed. Family labor constraints are also a problem. To understand these complex reasons for success or failure of technology, the review paid particular attention to the environmental, social, institutional and economic setting in the country, and the specific project locality. Failure to appreciate these external factors seems to have been the nemesis of a number of the projects reviewed. 3 Chapter 2 of this study reviews the main environmental parameters that determine land use and the rainfed agricultural potential in the semi-arid and arid regions selected. Chapter 3 highlights the key issues in crop production strategy, describes farming systems in the arid and semi-arid regions and upland watersheds, and examines the technical options for improved crop management in those marginal areas based on research and experiences in Bank projects. Chapter 4 looks at the problems of managing soil and water resources in the low-rainfall areas. Chapter 5 reviews the current use of technology for crop improvement in those areas and farming systems under four principal topics. Experience in the management of ruminant animals in the difficult agricultural environments of the low-rainfall and upland zones is the subject of Chapter 6. Chapter 2 LOW-RAINFALL ZONES: CLIMATE, TOPOGRAPHY, SOILS, AND LAND USE This chapter describes the features of the two low-rainfall zones that cover the 51 World Bank agricultural projects reviewed in this technology assessment. These zones cover three geographical regions that include the countries in which the World Bank projects reviewed are located (the project countries appear in italics below). * WANA, defined by the United Nations Food and Agricultural Organization (FAO) to include twenty countries stretching from Morocco to Afghanistan, In alphabetical order these countries are: Afghanistan, Algeria, Bahrain, Cyprus, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Qatar, Saudi Arabia, Sudan, Syria, Tunisia, Turkey, United Arab Emirates, and Yemen * Northwestern China * Latin America (Argentina and Chile). These geographical areas and countries will be mentioned according to their climatic zone, with some reference to analogous areas in the world. MIEDITERRANEAN ARID AND SEMI-ARID WARM SUB-TROPICAL WINTER RAINFALL ZONES The largest single contiguous area of Mediterranean climate, including WANA and also most of southem Europe, lies around and to the east of the Mediterranean Sea. The climate area stretches over 5,000 kIn, from the Atlantic in the west to Afghanistan in the east. Outside of WANA, other important areas of Mediterranean climate are in southern and western Australia, parts of South Africa, the western seaboard of North America (Oregon, California), northwestern Mexico, and in South America (parts of Chile and Argentina. Climate and Topography The characteristic main climatic pattern of the sub-tropical Mediterranean zone is hot, mainly or totally dry summers and cooler winters, with limited precipitation largely concentrated in the winter season. Annual precipitation in the Mediterranean climate ranges from 0 millimeters (mm) to 1,200 mm, but this study is concerned with areas that could be classified as arid or semi-arid, with annual rainfall 6 between roughly 200 mm and 600 mm. Some 88 percent of the arable land in West Asia falls into this category (table 2.1). The annual rainfall decreases sharply from the coastal Mediterranean littoral in North Africa south towards the Sahara, and in West Asia rainfall decreases both south and east of the Mediterranean Sea. Much of the Arabian Peninsula is atypical and lies in the semi-arid tropics with sparse rainfall mainly in summer.2 The variablility of rainfall is high, both within between seasons, with a covariance (c.v.) of between 20 and 50 percent. hence rainfed agriculture is more than usually a risky occupation. The zone is quite heterogeneous, since slope and altitude moderate temperature. In West Asia, this heterogeneity is particularly strong. For example, there are marked differences in winter temperatures in Turkey between the mild Mediterranean littoral, the inland mountain valleys, and the high plateaus. These differences greatly affect the yields of crops and the farmers' choice of crops, cultivars, and farming systems, and allow some diversification of land use in rainfed land (Oram 1979). Table 2.1. Classification of Mediterranean Bioclimatic Zones in West Asia by Precipitation ZONE TYPE ANNUALRADNFALL (rm) 1 PERCENT OF ARABLE AREA IN WEST ASIA Super-arid <.200. 0 Semi-arid 1 720044a00 Semi-arid 2 A400-600 14 Sub-humid 600-800 10 Humid 800-1200 i Super-hurid i> 1200 1 Note: Figures represent the share of total arable area in west Asia in each of the respective ecological zones. Source: Adapted from Barbero and Queezel 1981, pp. 69-74. Subclimates in Rainfed Agricultural Areas Rainfall and its distribution are not the only important determinants of crop distribution. Altitude, topography, and latitude all modify temperature. Although altitude moderates high summer temperatures, it also increases the risk of damage to crops from cold (in upland areas) and from unpredictable frosts (in adjoining lowlands). This limits the choice of crops and of cultivars in areas of high risk from low temperature. Such areas may be characterized as marginal in terms of their overall climatic constraints even though their average rainfall is often higher than that in many lowland areas with no serious frost hazard. 2The WANA region corresponds to sub-zones IVI, IV2, 1V3 and IV5 in Troll and Paffen's (1965) map of seasonal climates of the earth, and to the Csa, BSk, BShs, and BWh regions and subtypes defined by Koppen and Geiger (1936). Koppen and Geiger characterize the Csa zone (the principle agricultural region of WANA) as having a temperature exceeding 22 °C in the warmest month. Temperatures in summer are much higher in the BShs steppe and the BWh desert zones. 7 Climate Areas in WANA For the purpose of looking at faming practices and the scope for diversification of agriculture in the Southem Mediterranean (chapter 5) it is convenient to disaggregate areas in WANA with rainfed agricultural potential into five components that are the focus of this paper. These are all climatic zones with predominantly winter rainfall. Arid and super-arid steppe zones with mild winters and average annual rainfall below 200 mm. These comprise the bulk of the grazing lands in the region. Grazing lands are, however, being invaded sporadically by mechanized barley production in years of good rainfall where annual precipitation exceeds 150 mm. Management of pastoral areas is discussed in more detail in Chapter 6. * Arid to semi-arid "marginal" zones with mild winters and annual precipitation between 200 mm and 400 mm (a 75-day to 180-day growing season). These are mainly situated at lower elevations inland from the Mediterranean littoral or are foothill areas of mountain ranges in Morocco, Algeria, Tunisia, Libya, Israel, Jordan, Syria, Turkey and Iraq. Their main vocation is for barley cultivation (especially for sheep feed) towards the lower rainfall margins, and for wheat cultivation where annual precipitation is between 350 nmn and 400 mm. There is an extensive area where wheat and barley overlap in areas where annual precipitation is between 250 mm and 350 mm. Pulses and forage legumes can be grown in rotation with cereals where annual precipitation is above 250 mm. Cereal-fallow or continuous barley prevails in areas where annual precipitation is below 250 mm (where the growing season is less than 75 days). * Semi-arid zones where annual rainfall is between 400 mm and 800 mmn (180-day to 220- day growing seasons) with mild winters, at relatively low altitudes along the Mediterranean littoral from Morocco (including the Atlantic coast) through North Africa to West Asia (Israel, Jordan, Lebanon, Syria, Turkey). There is more scope for crop diversification in these zones. Cereals occupy the largest share of arable land, but fallow diminishes, and pulses, sunflowers, sugar beets, vegetables, and melons enter the rotation with cereals. Fruit trees, (notably olives, pistachio, and almond) and land pine forests (Pinus halipensis, P. pinea, P. reticotata) are relatively important in these areas. * Upland plateau zones with adequate or limited precipitation (300 mm to 600 mm annually), but with cool or cold winters (Anatolian Turkey, northeastem Syria, north- central Iran, Iraq, Algeria). They have a growing season of between 120 days and 180 days. * Upland mountainous zones with adequate annual precipitation (400 mm to just over 800 mm). Winters in such areas near the sea may be cool but have a high frost hazard (parts of Morocco and Algeria, northwestern Tunisia, northern Jordan, northwestern Syria, and Cyprus). Winters may also have prolonged cold (as in Morocco in the High Atlas, Lebanon, high elevation mountainous areas of northern and eastern Turkey, northern Iraq, northern Iran. and Afghanistan). These West Asian high mountain and plateau areas are 8 analogous to the climate zones of northem China to the east of the Hindukush and Pamir ranges. ClimateAreas in China In China, the zone with a somewhat similar climate to these upland areas is the large area of arid and semi-arid land in the northwest with a Continental climate of cold winters and relatively short but hot summers. This area lies north of Tibet on the westem border of China and south of the Mongolian Peoples Republic and the former Soviet Union. The continental semi-arid zone comprises the Provinces of Xinjiang and the Ningxia Autonomous Republic together with most of Gansu, westem Inner Mongolia, and smaller portions of northern Shaanxi and northeastern Qinghai. Stone and others (1985) distinguish two major agro-ecological sub-zones for wheat production in northwestem China: the Xinjiang winter and spring wheat zone, and the northwestern spring zone. Aridity limits crop production to irrigated perimeters in much of the zone. The Xinjiang winter and spring wheat zone is analogous to irrigated areas of north central Anatolia in Turkey, and the northwestern spring wheat zone can be compared with irrigated areas of the highlands in eastern Turkey. The Hexi corridor in Gansu is a flat arid plain with an elevation between 1,100 meters (m) and 2,600 m. It is subject to strong winds, sandstorms, and salinity, with a frost-free period of 150 days to 170 days, but an annual rainfall below 200 mm. In the World Bank's project area in northern Xinjiang, the average daily temperatures range from a low of -17°C in January to 27°C in July and August. The winter is long with five months below freezing and around five months when no frosts occur. T here are significant differences in microclimate within this area as a function of precipitation, altitude, the number of frost-free days, and the hours of sunshine. Annual precipitation ranges from under 200 mm to over 500 mm. In China's northwestern wheat region, precipitation is low and the climate is marked by long cold winters and short summers that are very hot at lower elevations. Water is the principal limiting factor for agriculture. In two of the sub-zones irrigation is necessary for any cultivation whatsoever, while in the other two subzones dry farming is possible. Average annual rainfall in the project area is 280 mm to 380 mm. There is little snow in winter, but the ground nornally freezes to about a meter. There is an average of 135 frost-free days, and the growing season for rainfed crops is from June through September. SOILS AND LAND SUITABLILTY Apart from a harsh climate, the quality of soils in WANA and China make the land suitable for a limited variety of crop production and forage. Soils in WANA and China On the whole, soils in WANA are not a major constraint to rainfed production. They tend to be of neutral to high pH, and calcic with an adequate base status. Except where gypsum is present (as in westem Egypt, Syria, and Iraq), the soil is in general not difficult to cultivate. In a number of countries, especially in West Asia and the Maghreb countries, fine-textured soils derived from limestone, including fertile "terra rossa" soils, represent a major portion of the agricultural area. In North Africa and parts of the Arabian Peninsula, however, there are also large areas of deep sandy soils that are low in nitrogen and organic matter and are extremely permeable. These soils are easily erodible, especially when cultivated 9 mechanically or when subject to severe grazing. They demand careful management for sustained agricultural use. In China, the soils in the World Bank Xinjiang project areas are mainly gray desert, gray calcareous, and meadow soils which occur, respectively, in the plains areas, in the foothills where they are associated with alluvial deposits, and at higher elevations. The organic matter content of the soils in all project areas is reasonably high. The soils are generally well structured with deep profiles, and present no problem for crop production when irrigated and drained. In the northwestem spring wheat area, the Loess Plateau, one of China's main geographic features, extends over large parts of Gansu and its neighboring provinces of Ningxia and Shaanxi. The loess soil forming the plateau is a wind-deposited silty loam of considerable depth (large tracts 50 m to 100 m deep) overlying sandstone bedrock. Although an excellent agricultural soil, loess is highly susceptible to erosion when deprived of its vegetative cover. Natural erosion of the friable soil has been exacerbated since settlers moved into the province several hundred years ago, but the problem worsened considerably in the 1950s when the rural population grew beyond the carrying capacity of the land. This is one of the most difficult issues that the Bank project aims to address. Land Agricultural Suitability Excluding Egypt, Libya, and the Sudan, seventy-one percent of the total land area of 678 million hectares (ha) in West Asia has no growing period because of cold or lack water or both (FAO 1978). Of this total, only about 102 million ha (15 percent) are classified as climatically suited to rainfed agriculture (minimum length of growing period is between 90 days and 119 days). Of the total arable land in West Asia, 87 percent is in the winter rainfall subtropics, 11 percent in the cool semi-arid tropics (summer rainfall) zone, and 2 percent is in the temperate zone. However, almost 70 percent of the useable total is in the two most marginal rainfall zones with respective growing periods of between 90 days and 119 days and between 120 days and 149 days (FAO 1978). Another 92 million ha with a growing period of less than 90 days is considered unsuitable for rainfed fanning, but suitable for irrigated cultivation. The actual area irrigated in 1990 in WANA was 22.2 million ha (15.6 million ha in West Asia, and 6.6 million ha in North Africa) but the unutilized irrigation potential based on water availability is only about 6.8 million ha (5.2 million ha in West Asia and 1.6 million ha in North Africa). Recent projections suggest that a continuation of the growth of irrigated area at the average 1980-90 growth rate of about 1.6 percent would exhaust the estimated irrigation potential for the region of about 7 million ha in about 20 years (Oram 1993). Both surface and groundwater availability are severely limited, however. The cost of water development is high, and agriculture faces competition for water from urban and industrial users. Considerable scope does remain for improving water use efficiency, land management, yields, and cropping intensities in existing irrigated land. Measures to combat salinity, to reclaim saline and waterlogged land, and to use lower-quality water require urgent attention. Such measures would contribute to raising yields and cropping intensities, and to introducing higher-value crops. Irrigated cropping intensities are currently only 135 percent overall, a very low level compared with what is potentially attainable with good management. This highlights the importance of raising the productivity of both rainfed and existing irrigated land in WANA. The area of higher-potential rainfed land is limited and altitude is also a constraint. For examnple in Tunisia only 10 percent of the country receives over 600 mm of annual rainfall, while 50 percent has less than 200 mm. In Syria 65 percent of wheat and 70 percent of the barley area is located in "marginal" ecological zones where rainfall is below 350 mm (in Tunisia the figures are much the same). In Jordan around 85 percent of both wheat and barley is grown in "marginal" ecological zones (Tadros 1992). In addition to low rainfall, high altitude is a major limiting factor. According to Kassam (1988), 21 percent of 10 (Tadros 1992). In addition to low rainfall, high altitude is a major limiting factor. According to Kassam (1988), 21 percent of the land in WANA is at altitudes exceeding 1,500 m and a much larger area (about 40 percent) exceeds 1,000 m, suggesting that almost half of the area may suffer such constraints. Tables A2-A6 (in the annex) present further data on land use and on high-altitude area and production of barley and wheat. Land Use Across Agricultural Zones Annex tables 2.3 and 2.4 show the aggregate distribution of land use for the 22 countries in WANA, and changes in the pattern of use between 1973 and 1989. For WANA as a whole, cultivated land occupies only 7 percent of the total land area. Wheat and barley dominate cultivated land use almost everywhere in WANA, even in humid zones . There was a significant shift in land use over the period, with a decrease of almost 2 percent in the annual arable cropped area and an increase of 19 percent in the area under perennial crops, especially fruit trees, nuts, olives, dates, and vines. The latter increase is partly because of attempts to exploit local comparative advantage for exports, but also because of increased demand and firmer Table 2.2. Land Use in West Asia and North Africa, 1989 prices in the region and a perception of (thousand acres) producers that tree crops in certain ...... ... ~~~~~~~ecological niches are a more appropriate or .. .. .. .... .. ... ..less risky form of land use than annual L ..and area;::; 1,543,351 crops. Arable land 96,892 The area under permanent pastures Permanent crops 8,717 has decreased by about 2 percent, Permanientt. pasture 320,225 apparently confirming reports of progressive erosion of grazings and :yForest and woodland .. .; ;;}: };;; ;;: ; y:; :110,491; deforestation due to pressures of increasing Other land: 947, i . . . .. . i i . . ....026 human and animal populations on the Irrigated IAgiculture 21,725. natural resources of the region. The , ~ ~ ~ ~ ~ ~ ~~~~~ . ... . .. .. ; - . . ... .. ........... .. ~decline in the forest area is approximately Source: Agrostat Land Use Data Tape, FAO. of the same magnitude, but represents a considerably higher proportion of the relatively small total forest resource. In China's Xinjiang wheat area, both fall- and spring-sown wheat are dominant, and often the only crop. It is generally grown at elevations between 300 m and 600 m, but spring wheat may be planted up to 1,900 m. Despite the limited potential for rainfed farming, northern Xinjiang has extensive natural pastures, but of variable quality. On the state farms in the 200 mm rainfall area around Bole, pastures are sparse semi-arid steppeland, but in the higher elevation areas excellent quality natural pastures exist. In the lower and warmer areas where farming is possible, individual farming households maintain small flocks on wastes and crop residues, although some animals which are not breeding may be moved to upland grazings in spring. Further north, semi-nomadic systems of livestock management exist, reminiscent of those in Eastern Turkey and Iran, with flocks being kept on grazings in the valleys and plains from mid-March to mid-July and again from mid-September to early December. In the winter (early December to mid-March), livestock graze in the foothills for protection from winds and storms. In the high summer (mid-July to mid- September), livestock visit lush high altitude pastures before snow closes them off. In China's northwestern spring wheat zone, Bank projects cover three agricultural regions. 11 * In the Hexi corridor, with an annual rainfall of only 150 mm, irrigation is essential for crop production and about 80 percent of the total cultivated area of 530,000 ha is irrigated. * In the Dingxi region, the annual precipitation ranges from about 200 mm to 400 mm. Rainfed crops occupy 1.27 million ha, and 190,000 ha are irrigated. * In southem Gansu, where annual rainfall averages between 400 mm and 600 mm, 1.37 million ha are rainfed and 130,000 ha are irrigated. Gansu's irrigated area has more than doubled since the early 1950s with expansion of irrigated area in the Hexi corridor accounting for more than 90 percent of this increase. Crops, mostly grains, comprise over 70 percent of agricultural production from Gansu. Another 20 percent of Gansu's total agricultural production is accounted for by animal husbandry and forestry products. Grassland areas in Dingxi, and southern Gansu, totaling about 14.3 million ha, supply fodder for sheep, goats, and draft animals. Timber and other wood products come from a forested area of about 3.6 million ha, most of it in southern Gansu. Chapter 3 FARMING SYSTEMS IN WESTASIA AND NORTHAFRICA This chapter describes the main crops and cropping systems in WANA, and suggests variations in land use and overall cropping systems to improve crop production in the marginal zones. FARMING SYSTEMS IN MARGINAL RAINFED ENVIRONMENTS With the limited scope in WANA for increasing the area of irrigated agriculture or high potential arable land more attention is warranted for the marginal environments. These are environments characterized by abiotic stresses that severely inhibit plant growth, and which are difficult to ameliorate by technology or other means (Morris, Belaid, and Byerlee 1991, part 1). This marginality also depends on the crop, the setting (there are different problems at different altitudes, for example), the temporal distribution of constraints, and the type of cultivation technology. Table A.8 documents the importance of these marginal environments with respect to wheat. Cereal-Livestock Systems A cereal-livestock production system dominates farming in marginal areas of WANA, with barley and sheep as the main commodities. A high proportion of this production is located at the lower end of the marginal rainfall spectrum (annual precipitation of from 350 mm to 200 mm or below) (Cooper and Bailey 1991, Mazid and Hallajian 1984). The range lands provide a high proportion ( some 55 percent) of the feed supply of national flocks in most WANA countries. There has been a significant expansion in recent years of the area under integrated crop-livestock systems based on barley as the primary source of feed--often without fallow. In some WANA countries, barley-livestock systems are believed to account for approximately half of the sheep production. In current low rainfall cereal-livestock systems, barley is the key crop but it is intended for consumption by rnminant animals rather than by humans. It is an attractive system conceptually since the barley can be used for grazing in its vegetative stage (especially in late winter when other grazing is extremely scarce). The weeds from the crop can be fed to animals, the grain harvested for feed (or sold), the straw fed to animals or sold, and the stubble can be grazed. In recent years, barley straw prices in several WANA countries have often approached those of the grain (CIMMYT 1991). If barley crops fails to mature, the vegetative growth at least provides grazing and straw. An ICARDA survey of 168 farmers in northern and western Syria found that 49 percent grazed their barley green every year, and almost all the remainder grazed occasionally, the latter based their decision mainly on their expectations of the barley grain harvest (Mazid and Hallajian, 1984). Comparisons of grain yields from grazed and ungrazed barley showed no significant differences. 14 Figure 3.1. Changes in Producer Price Ratios in Algena 1970 and 1987 The principal factors driving the growth of cereal- livestock systems in low rainfall zones of WANA 50 have been the rapid expansion of national 50 consumption of livestock products, the high levels of protection against imported meat, the consequent 40 increase in meat prices relative to those of cereal 35 U Barleylbroad grains . and the increasing gap between domestic 30 wh*SIeepbread supply of feed and feed requirements (Oram 1988). 25 wheat Table A9 (annex) presents the data shown graphically 20 a Beef/bread above. The relative prices of wheat and barley grain 15 I wheat for sale are not the main issue, since the dominant *0 component in terms of farm income is not the cereal * * but the animal, and in any analysis of the economic potential of the system, and of the scope for 0 improvement this should be the focus. For example, 1970 1987 in a survey of 440 farms in Jordan, 42 percent of the farmers owned small ruminants and on all farms with a livestock component, animals were the main source of on-farm income. Surveys in Algeria, Jordan, Morocco, Syria, and Turkey (Taurus mountains), show that livestock contribute respectively up to 77 percent, 82 percent, 100 percent, and 60 percent of farm income in low rainfall areas, it being especially important on smaller farms (Oram and others 1994, Somel 1991, Tully 1986, and Glenn 1988). In the Jordan and Moroccan surveys, the highest on-farm incomes per hectare also occurred on the farm types with livestock (NCARRT 1991, Oram and others 1994). Many of these farmers-even in areas with annual rainfall below 300 mm-grow both wheat and barley. While they use wheat mainly for their family subsistence and barley for their livestock, this is not always the case. The proportion of barley relative to wheat on those farms increases with decreasing rainfall, and there were no wheat-livestock farms in the rainfall zone below 200 mm annual precipitation. Farms in rainfall zones below 350 mm annual precipitation depend more on cereals with livestock than those located in areas with annual precipitation above 350 mm. In general, the lower the annual precipitation, the more farmers depend on combined cereal-livestock production. Low rainfall restricts the potential for crop diversification and intensification. Seventy-one percent of farmers in the highest annual rainfall zone grew some other crop than cereals, compared with 48 percent in the intermediate (200 mm-350 mm) zone and only 15 percent in the zone below 200 mm. However, the contribution of these marginal areas to overall barley supply is limited. For example, in Syria, Cooper and Bailey (1990) demonstrate that productivity of barley is so low in the zone below 250 mm, that 47 percent of the national barley area the zone contributed only 19 percent of national production. This was even more pronounced in the zone with less than 200 mm, which represented 19 percent of the barley area, but produced a mere 6 percent of the national total (see table AIO in the annex.) As barley has expanded into these more marginal zones, Cooper and Bailey (1990) report that national average yields have declined from 780 kg/ha to 681 kg/ha, between the periods 1960-70 and 1975-85, while coefficients of variation of yields have risen sharply nationally from 37 percent to 50 percent over the same period despite a reduction in the variability of rainfall over the same period. Despite the increase of the barley area in Syria by approximately one and a half times its 1979 level in ten years , the official goal of providing 60 percent of the energy requirements of the national sheep flock from barley has only been met in about one year in five in the past decade compared with one year in two in the 1960s. Clearly any further increases in sheep numbers comparable 15 to the threefold expansion between 1960-65 and 1989-90, or the 50 percent growth from 9.3 to 14 million head from 1979-81 to 1988-90 can not be accommodated. The issue is rather whether feed production can be maintained at its current level without permanent damage to the environment in the marginal cereal-producing areas and adjoining rangelands. Similar issues have to be faced in Jordan, Algeria, Libya, Morocco, and Tunisia. Since further large-scale incursions of cultivation into the rangelands can not be sustained, solutions that address policy distortions and seek technological change must be introduced. Table31l. Average Yields of Barley Grain in the Different Agricultural Stability Zones ofSyria ,1979-89 ZONE 1 2 3 4 5 AVERAGE ANNUAL < 350 250-350 <250 200-250 >200 RAINFALL (mm) Mean yield (kg/ha) 1,367 885 576 480 359 C.v. ( percent) 37 51 81 103 114 Percent of yearswith no 0 1 36 46 64 harvest Barley area 1989 (000 ha) 43 646 580 867 400 Mean share of barley production 16 42 17 8 6 1979-89 (percent) . ,_,._._. _. Source: Cooper and Bailey 1990. Technical Optionsfor Cereal-Livestock Systens There are two main technical opportunities for increasing productivity at the systems level in rainfed farmi, g in zones that have precipitation that is both low and highly variable. These are reducing fallow and replacing cereals with other forage or food (fodder) crops. Reducing fallow. Fallow in WANA has traditionally been a fourteen-month break between successive cereal crops. This basically forms a one year cereal, one year fallow rotation. The benefits claimed for fallow under this system have been related to five factors: moisture conservation, weed control, control of cereal diseases, improved soil fertility (especially nitrogen status), and easier workability for tillage and seeding. On the other hand, a main reason for the substitution of continuous barley for barley-fallow in the steppe margins seems to be the capacity of the former to provide a more consistent supply of feed from grazing as well as grain and straw. While there is experimental evidence to support the positive effects of fallow in some agro-ecological situations, its overall value has come under increasing scrutiny as pressures on available arable land and the need for more animal feed have increased in WANA. In areas with 375 mm to 450 mm annual precipitation in Turkey, a massive and well-integrated research, extension, seed production, credit and marketing effort supported by the World Bank in the rainfed marginal areas has resulted in an overall reduction of the large fallow area by 33 percent in about seven years. This campaign highlights the importance of improved crop varieties (cereals and legumes), 16 seed availability, cultivation practices, weed control, selective fertilizer use, and market development, for domestic consumption and/or exports, to encourage the successful replacement of fallow or continuous cereals by a cereal food legume system. In some countries price support policies favoring wheat, but giving no stability to legume production, have acted as a disincentive to farmers adopting a new crop involving additional costs and uncertainties (Oram and Belaid 1990). Despite the successes achieved in increasing the area under food legumes in Turkey and in WANA as a whole between 1980 and 1990--when the regional area and production respectively doubled from approximately 2.25 million ha and 2.19 million tons in 1980 to 4.17 and 4.21 million tons in 1990--the yields of the major pulse species have hardly changed, and in overall terms have actually declined marginally despite enhanced research at ICARDA, ICRISAT, and in some national programs. A bigger success was hindered by farmers perceptions that food legumes are more risk prone than cool-season cereals, with their greater susceptibility to length of day and temperature variability and to drought, salinity, diseases, and weeds, and with harvesting problems. Furthermore, in many countries, food legume prices are not subsidized or supported in any way, whereas wheat usually is. Wheat and sometimes barley markets are more or less guaranteed by governments whereas those of pulses usually are not, and prices and markets for fodder are purely opportunistic. Moreover, wheat is the primary subsistence crop in a region with the highest per capita consumption of wheat in the world. Hence, governments and farmers are very sensitive to any decline in the output of this commodity. Research stations, and development institutions such as the World Bank have placed considerable emphasis on fallow replacement in the WANA countries. The general conclusion is that fallow replacement is not a priori a "best bet" technology, but is beset with numerous pitfalls. First of all, both the technological and the farming systems options are extremely limited. Farming system options include barley-fallow (with weeds contributing forage), barley-barley (or possibly wheat), barley-forage legumes (or grass-legume mixtures), and ley farming based on medics, as in Australia. Alternatives to barley-fallow or barley-barley systems become increasingly restricted as the average precipitation declines . If legumes are to stand a chance of adoption they must fit well technologically with the farmers' equipment, budget and labor force. They must clearly contribute more to total biomass than altemative land use systems without depressing barley yields significantly compared with those from fallow or continuous cropping. Second, technical, economic, and psychological constraints to farmers' adoption of potential new cropping systems are often not adequately assessed. Measures which appear to threaten farnily subsistence, increase costs or labor requirements, or raise risks are likely to be resisted and these perceptions need to be understood through contacts with local research and extension staff, and through group or individual contacts with the farmers themselves. Third, the elimination of fallow in favor of continuous cereal cultivation (whether wheat or barley) should be discouraged in ecological zones below 200 mm and on slopes with shallow soils in upland areas without adequate precautions against erosion. It is therefore especially urgent to develop sustainable cropping systems which avoid cereal monoculture in such areas, while providing economic alternatives to fallow. Such policies would not greatly reduce national wheat and barley supplies. For example, Cooper and Bailey (1990), using data from the results of a joint ICARDA/Syrian National Soils Directorate study of 75 sites over four years show that by applying fertilizer and/or other improved practices (seed dressing and drill planting) to a reduced area under barley with annual fallow, but excluding areas with less than 200 mm annual precipitation, the government's planned targets for barley production could be achieved three years out of four with increased stability of production. 17 _-. . ._, ,.,-.... . .. .. .. .- ....... -..,. .. ......... ........---,.,-. B 31. IsFfowing. Nemesary? ... ss,.'. '... 5, *. , 1. .. ... . .v ' '-- .'" ' '' - .-'''' ' - .. '.' . ...... R' ent,' resear casts doubt'on thX benefit' of fallow. First, results from seera Mdte c es show'tha't'fllow offers little advantage for moisture conservation in areas wifth more thae. 450 nmurainall. Second, even in more marginal areas, benefits are:being questioned. For example, i Northern Syra (280 mm average annual precipitation) onyl 10 percnt of the rain that falls during the fallow season remained in the,soil-profile at planting tim*e (fHarri 1989) Research in Turkey on the Anatolian Plateau p.r.ovdes water use efficiencies .of between2:23 percent and-31 percent, with close: links to soil depth (Guler and Karaca 1988). Fallow efficiencies as low as 8 percent are reported in Southeastern Anatolia. In addi*t an uncultivated fallow may deplete moisture and add to the weed population in the crop year. On the other hand, fallow may' inmrove soil fertility depending on the legume content of the weed flora, and nmay provide vegetable material. for human consumption or feed for livestock, although little evidence is available- -concer'ning'the value of thie weeds for fodder or for human nutrition. Furthermore, continuous cereals- when grovn'without fertilizer, generally give the worst results in economic terns, with declining yields after a few years (CARIDA 1994). Fourth, in zones with an annual precipitation of 200 mm to 300 mm where the coefficients of variance of barley yield are very high, the potential for nitrogen and phosphate fertilizer application to barley in a barley-fallow rotation should be explored and, where technically and economically beneficial, widely demonstrated. An alternative option now being explored by ICARDA is to sow barley with nitrogen and phosphate fertilizer between widely spread hedges of a shrub species such as Atriplex. The shrubs also benefit from the fertilizer, especially in the establishment phase, and once established provide additional forage. They also act as a buffer against wind and water erosion. The Atriplex hedges are planted transverse to the prevailing wind. In ongoing ICARDA trials in Syria in a 215 mm rainfall zone, there was no adverse interaction between the barley and the hedges in terms of height or evenness of the crop. Further comparisons of barley-barley and barley-vetch rotations in the hedgerow system are underway. Several tools and support technologies have been developed to facilitate the introduction of continuous cropping. In Turkey, an index using annual rainfall, average relative humidity, and average annual temperature of the months with temperatures above 0°C was found to be useful in delineating the fallow/annual cropping area boundary (Guler and Karaca 1988). A Syrian model shows that where the only alternative to cereal-fallow in really marginal rainfed areas is continuous cereal cropping, it may better from the point of view of yield stability and long-run sustainability to retain the fallow with a light N-20kg- P205 60 kilograms per hectare (kg/ha) fertilizer application which will provide some grazing. The snag is that if the same fertilizer is applied to continuous barley it may out-produce barley-fallow in all but the driest years, at least in the short run. Forages as an alternative to fallow in barley-livestock systems. National data from six WANA countries (Algeria, Morocco, Tunisia, Jordan, Syria, and Turkey) show that about 2 million ha of forage are cultivated annually, representing 3.9 percent of total arable area in those countries overall, with a maximum of 10.3 percent of total arable area in Algeria and a minimum of 0.6 percent in Tunisia (annex table A13). On the other hand, the potential of reducing fallow and introducing fodder crops was thought to be very high: the area of fallow waiting to be put to directly productive use in the region, was estimated by Carter (1978) at 23 million ha out of a total of 30 million ha of fallow overall in nine countries. Consequently, from 1970 onwards, the replacement of fallow by forage crops has been a major objective in the World Bank livestock projects in Syria and Turkey, credit projects in Jordan, Morocco, 18 Tunisia, and Turkey, watershed management projects in Algeria, northwestem Tunisia, and eastem Anatolia in Turkey, and rural or agricultural development projects in Morocco, Tunisia, and Turkey. In most of these projects, as well as in the projects in China, a parallel goal was to increase the area under leguminous forage crops in arable rotations at the expense of cereals (usually wheat) particularly on sloping land, in addition to any gains made as a result of fallow replacement. Generally efforts were also made to increase the productivity of the forage crops by fertilizer application (mainly phosphate), use of improved varieties where those were available, and timely planting. 6Box. 3. ,2 tions fox' Fobragef Pro.dutio i W NA Options f~~~~~or frag production in.lU...oe........ t-.....0.;. ---. ,;0.0.. 0,}lV -ttt .-00fl,X .....f.X-Et,.t.-.- -f,i.. tH,-.. .,i0 Ann-- -ual species, including vetches,peas, chicklin . (L ..hyrus s ) ..e.s..m -X-- g g-.-:.;(Trifolium alexandrnum) nd other clovers.T. h. ey : .may..be gra.ed, .. .s. tf t;:0C00000:.- :0i- :i0:0S}.0}0:see d, used as;green forage,or mr comonly), cut for hay0$;0!0f0- :f--0 0D-3f0-0: :-00:"0'000;.-.u0.0:.0-Pa00-sture1 legume atbtae :longer-tem i the ; fann ytemsi roato with: cel.: ...... .. ..... These. may.be.resed .n an uas (Medio go.spec es- - medcs-i ubteran iC tX:tE;;iE;:f:00 igCl ;:#i;:ov er) o r perennias(alalfa, asulla, sainfoin, aphicarpic vetch) ::E7- :S::fEiE::E:d fffI;'$:S}::S0;*jC;0f:7:$.'-SPerenn iasS Cfor:.-bay or- tSsilage; with :little tgrazing such 'alfalfa (M' cg sativa saikbi (Qno0.0 s afa)0- ' . ' .....sulla.(Hed.saru' ..... . f:;:-,t-,;-:;}: E$t;-ff.iS::fi. .E. ........ ...... .....f: S . -::;:fX .SS-;ffl .S;:; ..ff$E: .. ...... . .. ... ..... .i:: Annualforage species. The introduction of an annual fodder or forage species is the most common option for the WANA region, partly because its management approximates most closely that of the cereals with which the annual forage (and food) legumes are most commonly grown in rotations, and partly because of its flexibility. The forage legumes are often sown together with a cereal (especially oats), so that the forage yield is higher and the cereal benefits directly from the nitrogen fixed by the legume (see Oram 1956,1ICARDA 1992a, and annex table 3.10). The general impression from World Bank project completion and audit reports is of limited success in getting farmers to replace fallow with legumninous forage crops, and of failure to persuade them to reduce their cereal area (or that of other crops) in favor of food or forage legumes in farming systems. Nevertheless some progress has been made in substituting leguminous crops for fallow, especially in Turkey, where about 1.5 million additional ha have been brought into legumes to replace fallow in part of the Anatolian plateau between 1970 and 1990. However, the emphasis has been heavily on the food legumes (pulses) Durutan (1991), reviewing the Bank experience in this area, comments: 19 Tremendous production increases were recorded in food legumes: 333 percent for lentil and 213 percent for chickpea, between 1980 and 1990. However, similar achievements were not recorded for the feed legumes due to ineffective extension, limited marketing and lack of incentives provided by the government. Efforts are needed for better integration of crops and livestock in the farming systems (p.4). Reseedingpasture legumes. High hopes have been pinned on the reseeding annual legumes which, although native to the Mediterranean Basin, have formed, until recently, the backbone of the wheat/sheep farming systems in the analogous climatic zones of Australia as part of a "ley farming" system (Ruthenberg 1980). In Australia there are with about 12 million ha of land under "regulated leys" based almost entirely on leguminous species originally native to the Mediterranean zones of Southern Europe, North Africa, and West Asia. Those leguminous fodder species have enabled a highly productive and biologically sustainable ley farming system alternating wheat and temporary pastures (principally for sheep) to be developed in that country. Starting in the early 1970s, this concept was widely tested through considerable national efforts in many Mediterranean countries, including Cyprus, Iran, Iraq, Jordan, Syria, Turkey in West Asia, Algeria, Egypt, Libya, Morocco, and Tunisia in North Africa, and France, Italy, Greece, Spain and Portugal in Europe. In several WANA countries this work was supported by expatriate financial and technical assistance, especially involving Australian 'know-how' (150 man-years of expertise in Libya alone). Despite the efforts directed to testing and introducing the system farmers over the last three decades, success in getting it widely adopted by farmers or pastoralists has proved elusive. National statistics on the adoption of annual or perennial forage crops in rotations are too poor to admit of comparisons with ley farming systems, but the area under ley farning is extremely low given the resources devoted to popularizing it, and its rate of expansion has been much slower than that of food legumes. In a number of countries where research and extension programs were initiated with expatriate assistance in the seventies and eighties, the area sown to medics and similar species in ley systems appears to have declined substantially and in very few cases has the system been widely embraced by farmers. The economic viability of this system, especially on small farms, has yet to be convincingly demonstrated. It has proved difficult, even experimentally, to show that cereal yields after medic are better than those after fallow: a recent economic analysis by ICARDA (1992) states "The Medic-wheat rotation did not compete well with other rotations in terms of whole farm income under standard price assumptions. Rotations with labor- intensive crops (lentil and waternelon) did best in the economic analysis. Medic pasture becomes competitive on larger 64 ha farms, but remains the poorest option on farms below 16 ha." Halse (1993) reviewing Australian efforts to introduce the system into the WANA region, reaches somewhat similar conclusions with respect to wheat yields but points to the additional income generated by the livestock. Buddenhagen (1990) questions the much-touted benefit from nitrogen fixation which he believes is 'over- stated and over-thought,' especially when legume grain and stover are removed from the fields, when he thinks there may be a net negative balance. If this is the case or a zero balance situation exists, then reasons other than Nitrogen fixation must be adduced to account for any benefit accruing to the cereal yields after legumes. Therefore he argues that more rigorous research is required to clarify such enigmas. Other reasons underlying this disappointing outcome for ley farming are grouped here for convenience under generic headings. They are elaborated in Christiansen and Boulanour (1993). * Biological constraints, especially lack of adapted cultivars (low cold tolerance) 20 * Management problems, particularly with small farms, including constraints imposed by uncertainties of land ownership and grazing rights, lack of fencing and water on farms for animals, inappropriate cultivation techniques and seeders, lack of hay- and silage- making equipment, and farmer and extension worker lact of familiarity with the sensitive techniques of good pasture establishment * Lack of input availability, including problems of supply of phosphate, appropriate herbicides, seeds, and appropriate equipment (ICARDA has developed a hand-pulled seed harvester to respond in part to the latter problem) * Entrenched farming traditions such as overstocking of seeds (leaving an insufficient amount in the soil for regemeration, deep plowing, and poor weed control * National economic policies that act as disincentives to ley farning, especially subsidies given to alternatives to legume forages such as barley grain and wheat, and lack of support for animal sciences and research and extension staffs for ley farming * Farm structure and tenure systems that may restrict the adoption of ley fanming. This last point bears some comment, since it has quite broad implications. The agrarian structure in many WANA countries is very skewed, with 60 per cent or more farms under 5 ha. These represent, however, under 25 per cent of total agricultural area., and S per Farms over 50 ha have about 40 per cent of the total agricultural area. The smaller rainfed farns are often in ecozones of less favorable precipitation, and/or on more difficult soils or terrain. These small farms, however have most of the animals. Larger farms have larger flocks, but represent a lower proportion of total animals. The smaller farms cannot easily adopt ley fanning because of the costs of fencing and water installations, and if they are tenants have no incentive to make such investments because of insecurity. A further problem exists with respect to communal or customary grazing practices, which in some countries do not respect sown pastures as they tend to do with standing sown crops. If sown pastures interfere with perceived traditional rights, the stock owners may even cut fences (Springborg 1986). Thus, enabling ley farmers to reap the full benefits of this innovative technology may not be easy, and the social and economic obstacles may well be greater than the technical ones. In a realistic review of the future of the ley farming system in the tropics and the Mediterranean McCown (1993) concludes: The strength of the ley system is its flexibility to respond to need for reallocation of resources between crop and animal production and the elegance of the soil restoration linkage. As appealing as ley farning is, it appears to be adaptive economically and socially only in conditions that occur in but a few places and times. In spite of the provision of several well-adapted pasture legumes, the evidence indicates that the prospects for ley farming in most of the tropics and the Mediterranean basin are poor where it is not already well established. Because the reasons for this lie mainly beyond technology and production ecology, the returns for continued R&D can be expected to be low. There may be places within these large regions where conditions may favor a ley system. If so, R&D that can eventually be viewed as effective and efficient will begin with 21 research which elucidates the existence of such a niche, and only then proceed to a plan for filling it (p. 4). Recent surveys suggest that the expansion of barley cultivation for sheep production seems a more attractive option to farmers than ley farming in the low-rainfall rainfall ecozones of WANA. Barley is more tolerant of drought and salinity than most other crops, including wheat. It is a versatile crop which can be grazed green during the rainy season, harvested for grain in early summer after a period of recuperation from winter/spring grazing, and provides reasonably palatable straw and stubble grazing after harvest. In the marginal rainfall zones grazing straw is regarded as more valuable than the grain, since in really dry areas good grain harvests are rare. Little or no fertilizer is applied to the barley in these risky areas, so this is close to a zero cash cost technology. Its cultivation is extending opportunistically into the steppe, but although this appears undesirable it is very difficult to stop when no legal sanctions exist or are unenforceable. World Bank-funded projects in the Mediterranean environment have generally eschewed ley fanning technology. It was proposed only in Tunisia (Northwest Agricultural Production Project) and in Morocco (Oulmes-Romani Project), but no results are given. Summarizing, all experience indicates that, unless the social, economic, and technical problems impeding the adoption of ley farming can be overcome, the incorporation of annual forage legumes into arable rotations, particularly species of Vicia and Lathyrus, for early grazing followed by harvesting for hay and silage, or grain and straw, is a better altemative than ley faming under WANA conditions. T'he comments made with respect to WANA countries conceming the future of ley farming based on the Australian pattem, do not necessarily apply to parts of Latin America, where it is reported that large areas are targeted for development projects based on leguminous pastures in rotations with grain crops. Temperate species of pasture legumes are suited to some areas of the Southern Cone countries. While Mediterranean-style ley farming systems are unlikely to be valid in northern or northwestern China because of major ecological dissimilarities, there may well be good opportunities for developing a cold temperate ley farming system along northem European or Canadian lines, using cold-tolerant species of grasses and legumes in medium term rotations with other crops. This would be appropriate to cattle and/or sheep production on larger well mechanized farms of the type foreseen in the Bank's Hebei, Xinjiang, and Gansu projects, or possibly to cooperative dairy or fattening enterprises, providing pooled machinery to enable smaller farmers to produce forage, either for their own animals or to provide feed supplies to a central unit where animals are managed professionally for meat and/or milk on a profit sharing basis. Perennial fodders. Perennial forage species, especially alfalfa, are widely grown in irrigated land, and also in upland rainfed areas where cold tolerance is required. Sainfoin production is being encouraged in Turkey for slopes greater than 8 percent . These species are not generally so well suited to rainfed systems in low rainfall lowland ecological zones, although Hedysarum (Sulla), shows promise in Tunisia. Future research directions. Key issues facing the replacement of fallow by fodder legumes. which emerge from past experience are: * Based on its on-station and extensive on-farm trials with forage legumes ICARDA makes the following recommendations concerning their ecological adaptability (table 3.2) 22 Table 3.2. Use and Environmental Adaptation of Different Species of Lathyrus and Vicia in West Asia and North Africa Region and Priority Research Objectives for these Crops at ICARDA -.. ECIE .CUSE ATION PRIOUT R ESEARCH OBJ I :-fL. -:sativus0gj i.-70: ': ji::GZ,G,: <300c 3 mm! rainl04: iResistance to- Orobanche and'f:oliar diseaes, thighl: ; (comm:on Xchicking :S: -i f;;l000.moderate:-cold;:: ;iHI, low BOAA con tent:::002i .-;i'i;:00000.0X'-:00X 0: L.scicera -- 3G,S <0 : raini' Resistence to Orobanbcheand foliar deases, high .. ..:. .. .. .0. .. . ..... ..0 t 03t n':S t:::Ipoedtoditlrac-t ; :i:0000 i' t:it00.-040 ;;ti (doazffionhickling) mo:derate vold HI, ~low BOAA contentit Lt, icilaolatus ff Wl}f GZ 300 X 0 mm rain RAdaptation tolow-rainfall barley ease, h :d g (subtcierranencl margnal see edness t h dfiI ( fchicmmn ) vet H:.- moderate cId tannIn cont en BOA c .... - V.t narbonensisp0-0; i: tG,S 3500mm rain;7: Reduneds pod shatteringgpd,tsse;Xo:0: 7 V. ervnsia G,S 30 mm rain ImprovesH, I scochyta le igh re sistanc llV. panoioa'sp-0-00 l GZ,H, :350 mmrain:0000: Rleducediodshaferin cold 00;:7 ::--000 : ..:iE :0 i!: 0 . t -...........t 4X ;,0 < - T 7 -iC - 00 : ..: X ;... .- .:CSi .. .. WCEi : (mHungarianmvetch) S highr:iri : :a t t lo ii::4fe 0 Q :: ; :.: Note: G = Grain GZ = Grazing H = Hay S = Straw Source: ICARDA. The contribution (if any) of legumes to the yields of the succeeding crop through residual soil nitrogen not used by the legume itself. This is one of the most complex and controversial agronomic issues requiring further research (Buddenhagen 1990). Moreover, advantages from nitrogen fixation by forage legumes are visible to the farmer primarily through increased yields of cereal grain or through increased yields of the legumes forage for grazing, hay, or total dry matter. Thus ICARDA correctly emphasizes that improved legume performance without detriment to the subsequent barley yield is essential if barley- forage rotations are to become sufficiently attractive for widespread adoption by barley- zone farmers. 23 The scope for improved management in the direct contribution of the legumes to feed supply, and what (if any) impact this may have on the barley yield. Evidence from on- farm trials in a 250 mm-300 mm annual rainfall zone of Syria indicates that applying 50 kg/ha P205 to the legume can both raise yields of legume hay, grain, and straw compared with unfertilized plots and increase barley yields compared with fallow. As an alternative to fallow or continuous cereal cultivation experimental evidence suggests that the best relatively low cost technology may be to apply a limited P205 dressing (40 kg/ha) to the most promising legume, which should stimulate N fixation as well as having a residual N and P benefit on the barley, and to treat the barley seed with a suitable fungicide. Drilling the barley and legume seed should also be beneficial to yields, as well as reducing seed costs, but many farmers do not have the necessary equipment. ICARDA claims to have convinced cooperating farmers in Syria of the value of the fertilized forage legumes in an integrated system with barley, but the current trends towards continuous low-cost barley for livestock where precipitation is under 250 mm, and the fact that the area of forage crops remains so low regionally suggests that this will not be an easy task. Experience in Bank projects in WANA countries lends support to this conclusion. FARMING SYSTEMS IN SEMI-ARID ECOLOGICAL ZONES The potential for diversification of cropping systems increases at average annual rainfall levels of 400 mm-600 mm or a growing season of 90 days to 150 days. Cereals--both barley and wheat, still dominate land use in this zone. Approximately 60 percent of the land rated as most suitable or suitable for wheat by FAO falls within this rainfall bracket in WANA. Wheat becomes more important than barley above 350 annual precipitation mm and the dependence on livestock in the farming system decreases as rainfall becomes less erratic. Cash crops such as food legumes, oilseeds such as sunflower and sesame, vegetables and melons, tend to replace forage legumes in the crop mix. Drought-tolerant tree crops such as olives, almonds, and pistachios are important and their area is increasing in this zone, especially in deep sandy soils in lowland areas and on hilly lands, where stoniness or steep slopes preclude annual cultivation. Fallow remains important in areas with an annual rainfall of between 300 mm and 400 mm, especially where mechanized cultivation is economically and technically feasible. The scope for technological change increases with precipitation. The adoption of modern varieties becomes more general together with improved tillage and seeding practices, as does the use of fertilizer, and to a more limited extent herbicides. In the upland areas, severe winters and higher frost hazard impose a short effective growing season and limit the options for diversification of crop production. There is often a sharp transition from moderate to high temperatures in late spring. In these areas, considerable areas of fallow remain in wheat-based systems: cereals predominate in land use, followed by food and forage legumes and vegetables, with fruit trees becoming important in mountainous areas. The importance of livestock relative to crops varies considerably, depending on altitude, topography and distance from major population centers. Thus in Turkey the big breakthrough in fallow substitution by food legumes in rotations with wheat was in the cool winter undulating plateau area of Central Anatolia, further to the east, as the terrain becomes more mountainous and winters harsher, the importance of livestock in the farm system is much greater. This may account for larger areas of fallow remaining in those areas because of the value of the weeds for grazing. It also creates opportunities for annual or perennial forage legumes to be cultivated with the cereal, however. Average yields of wheat and barley in upland zones are generally 24 lower than those in the plains, by 38 percent and 18 percent respectively, thus providing scope for raising yields through wider use of modem wheat and barley varieties with cold tolerance plus improved agronomic practices since the proportion of area under Mexican varieties in upland areas of Turkey and other WANA countries is still relatively low, especially with respect to durum wheat. Major efforts have been directed in this ecological zone towards increasing the contribution of pulses, other crops and fodders. They are described below. Increasing the Contribution of Pulses Because of their nutritional value, pulses (food legumes) are currently the most widely grown crop after cereals. There are expanding opportunities for trade in pulses for human food and for animal feed, both within and from the WANA region, and the prevailing high prices of meat may stimulate demand for pulses by low-income urban consumers even further (Oram and Belaid 1990). Pulses are by no means a new crop in WANA. All of the cool-season species (nearly 85 percent of the pulse production in the region) are native to the region, and most have been grown there for millennia. They are also important in the higher altitude regions of WANA.. Over a million ha of pulses are grown in upland areas of Morocco, Iran, Iraq and Turkey at altitudes over 1,000 m. Table 3.3. Comparative (metric tons per hectare) Annual Yields for Cool-Season Food Legumes and Cereals, WANA and Worldwide, 1989-91 FOOD LEGUMES CEREALS 1l8-91 ;1989i-91 Faba Bean Chickpea Lentil Peas Overall Wheat ey Top National Yield 3.62 1.21 1.27 4.95 - 7.89 5.89 Worldwide"2 Top WANA Yield3 21714 0.95 0.67 0.83 4-- i4.96 3.89 World Average 1.34 0.72 0.71 1.82 1.17 2.48 i 2.32 iYield DEf0 WANA Average 1.31 0.85 0.65 0.67 0.76 1.72 1.27 Yield Note.: lAll yields are from countries with a significant area of the crop -- at least 40 ha. Top world yields for legumes: Germany (faba bean), Australia (chickpea), Canada (lentil), France (pea). & for cereals: Ireland (wheat), Switzerland (barley). Top WANA yields for legumes: Egypt (faba bean), Turkey (chickpea), Turkey (lentil), Morocco (pea) & for cereals: Turkey (wheat), Turkey (barley). Irrigated crop, otherwise all yields listed are principally or entirely from rainfed land. Source: Oram 1992, FAO Agrostat Production Tapes 1989-91. The average growth of area for pulses in the region overall between 1961-63 and 1989-91 was 1.65 percent a year and the growth rate of yield only 0.69 percent. Growth accelerated somewhat in the 25 late eighties, as the pulse area rose by 4.6 percent p.a. and yield by 1.1 percent. Although 80 percent of the area increase in West Asia was due to the successful fallow replacement campaign in Turkey mentioned above, the area also expanded in other WANA countries by 42 percent overall, especially in Algeria, Egypt, Morocco, Iran, and Syria. Yield perfornance was highly variable, improving in Egypt, Morocco, Iraq, Jordan, Lebanon, and the Yemen, but declining slightly in Algeria, Syria and Turkey. Among the three main food legumes, faba bean production rose by 200,000 tons during the 1980s through growth in area and yield, mainly in North Africa. Chickpea and lentil production rose primarily due to area expansion. Of the three species, only faba beans respond well to irrigation, hence the bulk of chickpea and lentil production originates in rainfed areas where they have to compete for land with cereals other crops. There are several major impediments to increased yields and wider farmer adoption of food legumes in WANA that are apparent from both the World Bank's experience and from other projects. First, pulses have received much less research effort overall than cereals, and there are many more species involved, since most legume scientists work on both food and forage legumes. Data from the mid- 1980s show that legumes are allocated less than a third of the scientific resources of cereals both nationally and within the CGIAR system. (Oram 1992) Consequently fewer, improved cultivars have been released, and yields are 30 to 60 percent lower than those obtained in Europe or Australia (Annex table 3.12). . Second, because of bottlenecks in seed production and high costs of available seed, the majority of farmers still plant local land races with high susceptibility to biotic stresses, especially diseases. Seed shortage is cited by ICARDA (1992b) as a main constraint to wider use of winter chickpeas. Some of the most serious diseases are seedborne, hence there is a need not just for seed multiplication, but for the application of rigorous standards of hygiene to seed production. Review of national seeds organizations in Morocco, Syria, Tunisia, and Turkey shows that seed availability for pulses is low, often not more than 5 or 10 percent of requirements, that of forage legumes is generally lower, with pasture legumes much further behind. Governments rarely subsidize sales of legume seeds (unlike wheat), and private seed companies prefer to produce hybrids or vegetable seeds where they can have some control over the rights to their products rather than self-pollinating crops such as winter cereals and food legumes where they have no particular comparative advantage over government organizations or entrepreneurial farmers. Third, legumes have greater sensitivity than wheat or barley to variations in length of days, temperature, drought, and salinity. Hence their yields tend to be more variable This is reflected in higher coefficients of variation of yields of legumes than wheat, worldwide. This may be compounded by cereals receiving priority in site, seedbed preparation, weeding, and harvesting; legumes are more likely to be grown under conditions of relative agronomic neglect. This results in sub-optimal sowing dates, which can affect growth due to poor stand establishment, cold damage, and aggravated weed competition. Improved performance under high-altitude and low-rainfall situations through resistance to abiotic stresses is also being sought through breeding. Cold -tolerant lines of chickpeas have been released in six WANA countries, and in the case of lentils promising lines are in the pre-release stage in Turkey and Iran. Tolerance of drought, combined with flexibility to perform well in years of better rainfall is a further important breeding objective. Although this is a difficult objective it is essential to make progress if farmers' perceptions that pulses are more risky than cereals are to be modified. Analysis of recent surveys of 770 farners in Syria and 1,060 farmers in Morocco (Oram and others 994), shows that while a majority are aware of their potential benefits of food legumes a high proportion of those farmers reject them as being too risky, as well as costly. Since increased costs also increase risks these is a strong correlation between those factors. No other reasons for adoption received such high ratings. Fourth, food legumes are more susceptible than cereals to weed competition, and are more likely to be damaged by herbicides. Surveys of farm budgets on 131 farms in the Settat area of Morocco (low altitude vertisol, annual precipitation 377 mm) show how important weed control of legumes is compared 26 with winter cereals. Farmers on small (less than 10 ha) to medium (10-30 ha) farms spent from 11 percent to 18 percent of their cash on weed control, mainly for lentils, compared with only 1 to 2 percent for wheat and barley. Much of this outlay was for hand weeding of the legumes: some farmers used animal drawn weeders, but very few applied herbicides). The parasitic weed Orobanche is life-threatening to all legumes, but especially to faba beans. It has proved very difficult to control by normal farm techniques, so research is focusing on the use of "burn-out" herbicides such as glyphospate, and on integrated pest management (ICARDA 1992b). Fifth, legumes have also proved difficult to mechanize compared with cereals, although less so for planting than for harvesting. In Morocco, harvesting represents 41 percent of total farmer expenditures on this crop. (Dahan and Solh 1988). Much of this expenditure is for labor. Lentils cause particular problems because of their short straw, hence the losses from shattering and pod dehiscence associated with hand harvesting may actually be aggravated by mechanical harvesting, despite its potential savings in labor costs. Harvesting difficulties are accentuated by the uneven seedbed preparation, ridge cultivation, hand sowing, and high weed infestation common in food legume production, as well as by stoniness. No fully satisfactory form of mechanization has yet been devised. Sixth, legumes are susceptible to a wide range of serious diseases and pests, which are an important source of yield instability and farmer resistance to legume cultivation. Consequently a high proportion of the work on varietal improvement is being devoted to pest and disease problems (see chapter 4). In particular the threat of increased losses from Ascochyta blight is retarding the spread of winter- chickpea cultivation. On the other hand, where the Ascochyta incidence is low, yield increases of 155 percent have been found with considerably higher financial returns (table 3.15) compared with traditional spring sowing. The conventional wisdom that a legume contributes more nitrogen to the system than it takes out is being challenged. It is argued that when the legume is harvested for grain, and straw is also removed, there may in fact be a net loss (Buddenhagen 1990) . Recent work at ICARDA also suggests that local Rhizobia in WANA are not very effective sources of biological fixation so that even indigenous species of legume may benefit from inoculation with strains of Rhizobia selected for higher efficiency. Since the nitrogen contribution of legumes to yields of cereals grown in rotation with them is often held up to farmers as an important plus factor in their favor , these issues require further investigation. Where evidence shows increased cereal yields following a legume, even when the legume is harvested for grain leaving little nitrogen in the soil, the reasons cited by McCown and others (1985), for this phenomenon include: (a) breaking the cereal disease/pest cycle, (b) reduction in phytotoxic and allopathic problems, (c) improvement of soil structure and (d) favorable positioning of nitrogen in the soil profile. At ICARDA, research has focused on the effects on soil nitrogen of introducing different legumes into wheat-based systems. Six years of results showed unchanged total soil nitrogen concentration in rotations including lentils, melons, or fallow, but increases of 15-20 percent in rotations including grazed annual medic pasture or forage vetch. The large increases in total soil nitrogen and organic matter in the wheat-medic rotations indicate that there were large inputs of organic nitrogen to the soil (ICARDA 1992c). These results are consistent with those in other WANA countries as well as in Australia and Israel which show that the available residual nitrogen in the soil is lower when legumes are harvested for grain than when they are used for forage or grazing, with differential effects on the following cereal crop yields (table A28). In the upland areas, major improvements are still possible through improved pulse varieties. The fallow replacement program in Turkey was achieved mainly with existing varieties, and was largely the result of the expansion of the legume area in place of fallow together with improved crop management rather than an increase in yields. Problems of cold tolerance and weed competition in fall sown lentils and vetches in Turkey have impeded their wider use, but research on cultivars which can be sown in winter 27 has led to the development of several chickpea and lentil lines, which can take advantage of the longer growing season and have a higher yield potential . While lentils seem relatively free from pests and diseases at higher elevations, chickpeas appear at higher risk from Ascochyta blight and cold. The only WANA country which has a major upland area of faba beans is Morocco, where about 92,000 ha was reported in 1987 (64 percent of the national total) compared with a total of 50,000 ha from peas, chickpea and lentil combined. Yields of faba bean average about 1,000 kg/ha there, over double those of lentils, and five times those of chickpea. In Egypt, however, where improved varieties are grown under irrigation, yields exceed 2.5 tons/ha, some of the highest in the world. Thus there are problems both of fact and of perception constraining the wider adoption of pulses by farmers in rainfed farming systems. Some benefits such as their contribution to soil nitrogen may be demonstrated through higher cereal yields, but other issues such as increased costs may best be resolved by assessing returns in terms of grain and biomass from both cereals and legumes, supported by a budget analysis of gross and net returns from the system as a whole that includes all costs. This is essential if policies to expand legume production are to stand more chance of success. Bank appraisal reports in WANA generally accord high priority to increasing the area under legumes. If this policy is to succeed more resources should be allocated by governments to support of research on food and forage legumes, to training specialist extension staff in techniques of their evaluation and demonstration, and the availability of seed, phosphate, and rhizobia where needed. Introduction of Other Crops The diversity of rainfed farming in WANA's low altitude zones is already greater than is commonly believed above 400 mm, with oilseeds, corn, rainfed summer vegetables, melons and water melons, forages, and possibly cotton being grown without irrigation, or, increasingly, with supplementary irrigation. The area of oilseeds, principally sunflowers, has risen tenfold since 1961 to over 1 million ha, and doubled in the last decade; yields have also increased substantially to 1.3 mt/ha. Cereal-sunflower- fallow is said to be the best rotation. However, Pala (1991) reports that sunflowers are entering the system in the 350 mm to 600 mm rainfall zones of Turkey with more stable rainfall, but with considerably greater depleting effects on soil moisture than legumes. Other annual crops such as sugarbeets, continuous cereal, or cereal-oilseeds also tend to be more competitive for soil moisture and nitrogen than legumes The effects of other crops on depleting wheat yields disappears in Turkey where annual precipitation exceeds 500 mm, although temperature may restrict the scope for diversification. From this evidence and from experience in Australia it appears that annual oilseeds are a risky proposition in areas of WANA with under 400 mm of rainfall and would benefit from supplementary irrigation even further up the rainfall spectrum. International market prospects for sunflower oil and canola are very competitive, and it seems unlikely that large-scale production in WANA would be remunerative even with higher and more stable yields in the absence of domestic oil-extraction industries. 3Leaving legumes weedy directly reduced lentil yields by 21 percent, and those of vetch by 29-48 percent, as well as having cumulative effects through increasing weed populations. 28 Replacement of Food Legumes with Fodder Crops In upland areas, experience indicates that while food legumes are important components of cereal- based rotations in upland areas, forage species are of even greater significance. There has been a good deal of effort in Turkey to develop forage production for hay in plateau areas and mountain valleys, both through annual grass/legume mixtures in relation with cereals and through perennial legumes for grazing and hay (Onobrychis viciafolia and Medicago sativa). Data on the total area of these crops is difficult to access, but it is not inconsiderable and suggests another alternative for diversification in upland zones where livestock (especially cattle) are important, or could become so if cereal prices remain at their current historic low levels and meat prices continue to rise. Mountain Watersheds While much of the upland plateau area has a relatively level terrain suited to mechanized agriculture, there are considerable areas of WANA with mountainous terrain, where farms are small and often difficult to cultivate other than by animal draft or hand labor.4 Although such areas are often major watersheds of considerable importance to the sustainability of agricultural and urban activities in lowland regions--especially through their influence on water supplies--they tend to be overlooked by policymakers and research and extension programs. This is often a function of their relatively low population and difficulty of access. One result of this "benign neglect" has been that the fragile resource base has progressively been eroded by centuries of competition among users (including farmers, herdsmen, and foresters) leading to deforestation and massive erosion. In isolated communities which may be cut off from downstream areas by snow, a considerable degree of basic self-sufficiency is desirable, although this may be complemented by trading vegetables, fruit, and livestock products for grain and pulses at urban markets where the latter are accessible, and by appropriate storage techniques. The vocation of these dissected mountain and valley areas is very different from that of the upland plains--for example, while subsistence production of wheat for family use remains important, large-scale cereal culture is not a sustainable form of agriculture on steeply sloping lands, although, in some countries, governments have unwisely subsidized cereal production in mountain areas on social grounds. On the other hand, forestry, agroforestry, controlled grazing of livestock, fruit and nut production on protected slopes or terraces, and production of irrigated vegetables, fodder, and trees (e.g., poplars) in valleys are well-suited to mountainous topography. Stratification of production is possible at different altitudes within the mountains and between mountain valleys and lowlands--for example, young animals reared on mountain pastures can be fattened on grain or irrigated fodders "downstream". Tourism is a source of revenue in some mountain areas, for example in Turkey, Syria, Lebanon, Jordan, and Morocco, as well as in southern Europe. Government investment in infrastructure and services (roads, electrification, water supply, health, education, research, and extension) and its encouragement of private investment in transport services, marketing, and local agro-based industries is crucial to both social and technical development of such areas. Land reform may also be necessary. Farms in watershed areas are often both small and highly fragmented, with titles to land a bone of contention. Rational resource management may require significant shifts in current patterns of land use (often the historical result of isolation) to permit specialization according to ecological and/or economic comparative advantage, and the 4 The average farm size of 63 farms in eight villages of the Taurus mountains area of Turkey was 4.1 ha ( ICARDA 1990 b). In Morocco, for 61 farms in the high Atlas Zone, it was 6.5 ha (Oram and othersl994). 29 development of a sufficient bulk of produce to satisfy market demand for fresh produce or the needs of processing plants. This may not be feasible without land consolidation and cooperation among groups of producers. Thus in considering new Bank projects in upland areas it is essential both to take a watershed management approach, and to enlist the cooperation and participation of local authorities and local communities from the planning phase through implementation. Chapter 4 LAND RECLAMA TION AND SOIL AND WA TER CONSER VA TION WANA has 1.25 billion ha of uncultivated area, more than ten times the current cultivated area in the region ( FAO, 1990). In addition, there are 476 million ha of pasture and 153 million ha of forest, much of it in semi-arid or arid ecozones of limited production potential. The potential for bringing new unused land into cultivation by normal techniques is virtually exhausted. Attempts to rehabilitate severely degraded steppe rangelands have had little success, principally because of social problems related to property rights, but also because of high relative costs, especially in drier areas and in remote mountain areas. Afforestation has had mixed results. Meanwhile, both cultivated and uncultivated agricultural land continues to be degraded, leading to loss of productivity, and in some cases to irreversible damage as well as to off-site externalities. The region's governments face three major tasks in managing their agricultural resources: (a) bringing unused land into productive use where technically and economically feasible; (b) preventing further deterioration of existing agricultural land, and (c) increasing the efficiency with which soil and water resources are used. This chapter focuses on these three tasks and identifies specific technologies for resource management to complement the discussion of crop and livestock management in Chapters 3, 5 and 6. LAND RECLAMATION Destoning and soil amendment are two principal techniques of land reclamation in WANA. The former has been used or mentioned in Bank-funded projects more frequently than the latter. Destoning Bank-funded land reclamation projects in rainfed zones have focused on destoning, that is, removing rocks and smaller stones from the land for arable cultivation. Depending on the nature of the subsoil, land to be destoned may have to be ripped mechanically before destoning, for example where there is a shallow layer of limestone close to the surface (as in Syria). This is followed by bulldozing or rock raking to remove the larger rocks before mechanical or manual stone clearance. In some cases, blasting may be necessary. Subsoiling may be required as a prelude to agricultural use. The debris may be used for road-building, for terrace foundations, for buildings, or for run-off control and water- spreading. Destoning operations can bring new land into cultivation or can improve the quality of existing land for buildings and raise its value, as well as providing supplemental local employment. In WANA the scope for this form of reclamation may be limited by soil, aridity, or slope, while in South America where land is more abundant, it may not be economically viable. Bank experience with destoning comes from projects in Jordan (1st and 2nd Credit Projects 1963, 1974), Syria (Southern Region Agricultural Development Project 1987), and Morocco (El Hajeb Pilot Project, Meknes Agricultural Development 1975, Middle Atlas Rural Development 1981, and Oulmes- 32 Romani Agricultural Development 1982). More recently, destoning operations have continued in all of those countries, both with government and external funding, and increasingly through farmers initiatives using local contractors. These experiences offer several conclusions. * Manual clearance (collection of surface stones for removal in wheelbarrows or animal- drawn carts) can help to facilitate future farm operations, especially tractor cultivation and mechanized harvesting. It is not a solution for large-scale reclamation, however. * Large-scale reclamation requires a combination of mechanized clearance and breaking up of larger surface stones, ripping and excavation of buried stones, andstone collection either manually (loading into a tractor trailer), or mechanically (by raking). On land with many deep stones, two passes of the ripper may be needed. * The nature and power requirements of the mechanical equipment, and hence the costs of the operations, can vary widely with the size and depth of the stones that have to be removed, and the geology of the substrate. Mixed tractor clearance and manual collection and loading of smaller stones can be done with a 70-80 HP tractor, and a hectare can be destoned in three to five days with the help of three laborers. An 80-120 HP tractor equipped with one rake can clear a hectare an hour on reasonably flat land with no big or deeply buried stones. Deeper ripping (60-80 cm), blasting, and unearthing of buried debris requires more powerful traction equipment. - Destoning is only technically and economically feasible on soils with a potential depth of 50 cm-100 cm after ripping and clearance. Soil and sub-soil survey and analysis should be undertaken to determine the physical and chemical status of the area to be reclaimed as a basis for decision-making both with respect to the type of operation and the equipment required for destoning, and the future agricultural potential. -- Destoning on sleep slopes may be difficult, and if undertaken should be followed as soon as possible by anti-erosion measures and planting cover crops. At intermediate degrees of slope, destoning can be combined with terracing, although this raises the costs. e Destoning and clearance of the land is moderately capital-intensive. A cost per hectare of $1,880 (1990) for a 2,000 ha operation is estimated by the Oulmes-Romani SAR, using D7 or D8 bulldozers for ripping and subsoiling under contract. Breaking of large stones and loading and unloading stones into transport supplied by the project was done by labor supplied by the farmers. According to the SAR the selected areas should have an estimated investment cost below $2,050/ha (1990), or a stone density of under 450 m3/ha to allow for a satisfactory ERR of 10 percent. In the Meknes project the contractual component cost was approximately $580/ha (1990); whereas that for collection and evacuation of stones by the beneficiaries was around $1,345/ha (1990). This 5000 ha destoning operation generated 5,500 person-years of labor--roughly one person-year per hectare. Contractual destoning work for private farmers in 1995 in semi-arid Khouribga Province of Morocco indicates a machinery cost of $1,000/ha, and labor costs of $450/ha, partly recovered by sale of stones. 33 These projects show that destoning is most successful with individual land ownership. Cooperative efforts failed because of lack of motivation, and the impossibility of identifying individual inputs for the amount of work individual members contributed. Thus nearly 80 percent of the destoning was the work of private farmers. The PPAR for the Meknes project rates the operation as successful, engendering high interest on the part of the beneficiaries and having a positive demonstration effect on neighboring areas where some farmers undertook destoning on their own initiative and at their expense. In addition to direct economic and agronomic benefits, it led to higher cropping intensities and the introduction of higher value crops and livestock. Although the interest in destoning in subsequent Bank-funded projects has been modest, despite the praise directed at the Meknes operations, destoning is still apparently seen as a potentially productive option in WANA. Research on destoning methods is provided for in the 1992 Bank Project for Pilot- Watershed Management and Forestry in Algeria. A new project is being developed for consideration by IFAD, involving cooperation between ACSAD and ICARDA for the reclamation and agricultural development of a large area for rainfed agriculture not far from Aleppo in Syria. Destoning is not a high-profile, glamorous technology, and the scope for technical improvement is probably limited to increasing mechanical efficiency so as to reduce costs. Nevertheless, wherever the availability of new land is exceedingly tight, as in WANA, and land can be reclaimed in areas of adequate rainfall and workable soils, it may be a relatively quicker, lower-cost and simpler approach than irrigation. Private initiatives for destoning and related land improvement operations should be encouraged, using rental machinery, perhaps on a pro rata per-hectare cost-sharing basis among groups of private landowners. Soil Amendment A second approach to upgrading or reclaiming land is through soil amendment to correct acid, salt, and nutrient imbalances, especially those due to micronutrient toxicity or deficiency. Acidity. Crops vary considerably in their range of tolerance to acidity (for example among cereals, rye and triticale are very tolerant, wheat moderately tolerant, and barley least tolerant). However, in much of the WANA region soils are of calcareous origins; acidity is therefore not a widespread problem, and lime application has consequently received very little attention in Bank-funded projects in this region. Salinity. Salinity is a more pervasive problem, and is of particular importance in remote mountain areas that lack infrastructure and in semi-arid regions because of the lack of rainfall to leach salts from the surface horizons. Salts may be transported by wind-borne deposits (a problem in some areas of WANA), by drainage water downstream from irrigation projects (off-site effects), by rising groundwater (usually in irrigation projects where water is being used wastefully), by saltwater intrusion (often by overpumping wells in coastal areas), by saline water from deep aquifers; and by seepage of excess soil water below the root zone beyond that needed by plants in dryland areas (for example from uncropped fallow in wet years). This water can move downhill, picking up dissolved salts en route, until it encounters some barrier in the soil when it surfaces in a seepage area. Although these problems are encountered most commonly and are of greatest economic importance in larger irrigation systems, they can arise in semi-arid rainfed areas, especially where rainfall is highly variable. Such problems are especially likely to occur when supplementary irrigation is practiced or with minor irrigation from streams, water harvesting systems, and well-sinking operations. The latter may be because of pre-existing soil salinity, of saline ground water, or of overpumping of coastal aquifers leading to saline intrusion. These hazards are not referred to in the 34 projects reviewed, nor in most cases are provisions made for drainage. Where groundwater is of high quality, and soils are not initially saline or are highly permeable (or both), drainage may not be necessary. The reports of rising water tables from some countries in WANA and South Asia raise a red flag concerning water management and salinity, however Micronutrients. Micronutrient problems are also rarely mentioned in Bank-funded projects although in some semi-arid winter rainfall areas of the world deficiencies, for example of molybdenum and selenium in wheat and pastures, or boron in sugarbeet and alfalfa, and boron in wheat in Syria, are reported (see Chapter 5). Toxicity problems are also reported. Excess selenium in drainage waters may severely limit the agricultural potential and affect human and animal health. While deficiencies are often controllable cheaply and relatively easily by soil amendment or foliar sprays once diagnosed, toxicities such as aluminum present more intontable problems. This is because they tend to be related to other serious soil problems such as acidity which are difficult or costly to control. In such cases breeding for crop tolerance may be the best approach. SOIL AND MOISTURE MANAGEMENT AND CONSERVATION In semi-arid WANA, Mexico, the west coast of South America, and in northern China (despite its colder climate), both land and water are scarce resources, so that technologies that conserve soil and water are of crucial importance to sustainable agriculture. Moreover, many if not all of the techniques that prevent soil erosion and degradation also help to conserve moisture or to improve water use efficiency (or both), while measures to improve water use efficiency may also reduce soil loss and degradation as a result of flooding, waterlogging, and salinization. Thus there is considerable synergy among these techniques, although it may be difficult to measure the full benefits of investment in soil and moisture conservation. This is especially the case in upland watersheds, when a significant proportion of the benefits may occur outside the immediate area. A further problem is that to some extent such damage may arise from natural causes, and assessing the contribution of natural and anthropomorphic factors is troublesome methodologically. A wide range of physical, mechanical, and biological measures can contribute to soil and moisture conservation and thus to sustainable agriculture in semi-arid regions. These include: - -Engineering structures to reduce gully erosion, check spate flows in the rainy season, spread run-off water or divert it to storage areas for future use * Mechanical techniques such as conservation tillage (including no till), contouring, and broad-based terraces that reduce sheet erosion from wind and water, conserve moisture in situ, and channel excess water in a controlled fashion through prepared waterways (either to drainage lines or for water harvesting and storage) * Vegetative techniques (such as strip-cropping, vegetative contour bunds, grassed waterways), combined with physical or mechanical techniques, or in some cases as an alternative to those techniques * The incorporation of these and other conservation-oriented techniques into sustainable drought management farming systems, for example, early seeding, use of drought-tolerant 35 and/or early maturing cultivars, balanced plant nutrition, effective weed control, integrated pest management, and timely harvesting5 Range improvement and controlled grazing that can contribute both to reduced erosion and to recharge of the ground water table by decreasing run-off, as well as increasing the productivity of the resource itself * Tree planting, either through afforestation, windbreak development, woodlot or other plantations, and agroforestry.6 Approaches to soil and water conservation are described in Srivastava and others (1993). These techniques are often complementary rather than mutually exclusive, and may form components of a conservation plan on watersheds or even on farms. This complementarity can cause problems in trying to assess rates of return to individual conservation measures. High- quality monitoring and evaluation procedures would help to determine such benefits more precisely, but the Bank-funded projects reviewed have an almost uniformly poor record in this respect. Box 4.1. Main Water and Conservation Techniques Main water and conservation techniques include: * Conservation tillage, sometimes termed residue management or MART (mininum and reduced tillage), including no-tillzero tillage * Contour farming * Terracing in various forms, with or without conservation or other tillage * Biological conservation (techniques douces in North Africa) * Water harvesting. Source: USDA/SCS (1992). Conservation Tillage Conservation tillage normally forms part of a cropping system but differs from conventional cultivation practices in a reduced frequency of operations and in less frequent or no inversion of the soil. Residue from the crops is left in on the soil surface to prevent erosion and to help conserve moisture. 5A problem with the integrated package approach is that it depends significantly on the availability of energy to perform the essential operations as nearly as possible at the optimum time with respect to the interaction between the physical enviromnent and the biotic components of the system. This is where breakdown often occurs, especially where the field operations depend heavily on human labor, as in many developing countries; 6 Not only can this contribute to reducing wind and water erosion and improving soil organic matter, but in areas where tree removal for cultivation of arable crops has led to salinity, salt-tolerant trees can be used to reclaim this land for productive use. Trees can also provide fodder for livestock, thus helping to reduce pressures on the natural grazings; 36 Weeds are controlled by a combination of cultivation measures and herbicides. In the case of no-till (the most extreme form of conservation tillage) there is no inversion of the soil, and weeds are controlled almost entirely by chemicals. No-till disturbs residue only in a narrow band along the row since the seed is placed precisely through chisel-like tines, accompanied by fertilizer if required. There are three main variants of residue management, namely no-till, ridge till, and mulch till, with the degree of soil disturbance increasing and the dependence on herbicides for weed control decreasing from no-till to mulch-till. No till and mulch till are practiced in well-drained soils for crops such as cereals or soybeans normally grown on a flat surface; ridge till is used for crops such as potatoes that are normally grown on ridges, or for com and soya on fine-textured soils with poor intemal drainage. All of these systems as currently applied in North America eschew the use of the moldboard plow; in Eastem Europe and the former Soviet Union (where residue management practices ought to be more widely practiced but are not), the soil is plowed with discs or moldboards at least once in four years in "reduced" tillage systems, whereas in standard tillage practices it is inverted every year. The value of conservation tillage systems that maintain substantial residues on the soil from year to year has been so convincingly demonstrated in the United States that farm conservation and anti-erosion farn plans involving the maintenance of at least 30 percent of residues on the soil have been made mandatory on most farms with effect from 1993. There are also important cost savings through fuel conservation and the lower power requirements of these techniques, which reduce machinery maintenance and may allow less powerful tractors to be used. Because of the reduced traffic over the land and the avoidance of bare soil, compaction losses are also reduced, facilitating more timely operations in wet weather. Higher cropping intensities, especially through the elimination of fallow, are reported from Canada and the United States. In view of the important technical and economic benefits that could be expected from residue management techniques with respect to erosion control, moisture conservation, crop yields and cropping intensities, this would certainly seem a "best bet technology". It is mentioned as an approach to be applied in more recent Bank-funded projects, for example, in Argentina (Agricultural Credit lI), Tunisia (N.W. Rural Development), and Turkey (East Anatolia Watershed), and hinted at obliquely in terms such as "develop improved and more timely cultivation practices" and "improve moisture conservation", in several others. No results were yet specified in any of the completion or audit results reviewed. Moreover, the references in the appraisals are generally vague and give no indication, which of the various "conservation tillage" practices was actually to be applied in the project. Even in the East Anatolian project, where "conservation tillage" using sweep tines on slopes over 4 percent is specified, there is a subsequent confusing mention of "minimum" tillage, creating doubt on which technique is actually being proposed. Box 4.2. Conservation Tillage in the United States Thirty six million ha were planted in the United States in 1992 to tillage management systems that leave 30 percent more residue on the soil, and another 30 million hectares to systems retaining 15-30 percent residue, representing nearly 60 percent of total cropped area. Best estimates for the year 2000 are 70-90 percent of all cropland retaining 30 percent or more residue (Cameron and Oram 1994). The bulk of the area under conservation tillage is under mulch till systems, but the share of no-till has risen sharply from just under 20 percent of the total in 1989 to nearly 32 percent in 1992 (CTIC 1993). Three possibilities may account for this uncertainty. First, the appraisal teams may be acquainted in general terms with conservation tillage practices but are not familiar with the details of these techniques. Second, there was not enough knowledge or experience of these techniques by extension staff in the 37 techniques. Second, there was not enough knowledge or experience of these techniques by extension staff in the countries concerned to pursue their implementation in the projects, and third, the equipment (especially seeders) and chemicals needed to enable farmers to apply them successfully were not available, even if there was know-how concerning the correct methodology. At the farm level, the following common constraints to the adoption of the techniques outlined above are widely reported. They are: * The techniques represent a learning experience even for relatively sophisticated and experienced farmers and some lose money and/or heart after a short time. * New and more expensive machines may have to be purchased, especially drills and sprayers. * Seeding into large volumes of residue, as is the case in Western Europe (where straw yields may average 10-15 tons/ha) is problematic. * Cultivating and seeding in wet, "no-till" heavy clay soils is difficult. * The longer-term impact of the frequent use of herbicides on soil, water, plant or animal life, on residues in products, and on the build-up of resistant weeds is a reason for concern. * Spread of diseases through the residues on the soil might be a problem. * Entrenched attitudes of some farmers who like to see clean, well-tilled soils and regard deep plowing with moldboard plows as essential to a good crop. * Farmers have other uses for the residues--especially straw,for example animal feed or bedding, thatching. This is of special importance in the WANA region. Although there is little evidence of conservation or no-till techniques being used to a significant extent by farmers in any of the countries reviewed in this assessment except Argentina and China, both of the last two constraints seem likely to affect the adoption of such practices. Certainly many farmers and even crop scientists in WANA remain wedded to deep moldboard plowing, despite considerable experimental evidence from that region to suggest that this is not essential to moisture conservation and may even be detrimental (Oram 1956, ICARDA 1991b), as well as from China (ICDF 1988), Brazil and the Ivory Coast (CIRAD 1993), and North America. ICARDA states: Mechanized tillage has become common in WANA only relatively recently, and the techniques and implements farmers use often appear ill-matched to the agricultural needs. Deep tillage, which has a high fuel cost, may be unnecessary, even detrimental; and research at Tel Hadya has shown that it does not improve water infiltration or crop yield over shallower tillage. Some yield reduction has been experienced using a zero-tillage, direct-drill planter, but this is because of a limitation on the machine's row-spacing adjustment. The technique itself shows great promise for local conditions. 38 These problems should not deter development agencies from examining the prospects for introducing or encouraging existing on-farm use of modern residue management techniques where conditions appear propitious. Contour Farming Land in the study region is frequently tilled and planted up and down, rather than across slopes, either because land ownership or tenure does not give the farmer any choice, or because it is convenient. On steeper slopes, or in erodible soils even on gentle slopes, this leads to serious losses from both sheet and gully erosion, especially if the soil is exposed, with high runoff losses and poor moisture retention for crops being grown on the sloping land. Eventually the land may go out of production. Such problems are very common and are not confined to annual crop production. A high proportion of the new fruit and olive tree plantations being established in west Asia is on slopes with trees running straight-up and down the hillsides in wide rows with no vegetative cover to stabilize the soil. Experience with alternative approaches to cultivating slopes has shown that soil losses from sheet and gully erosion can be reduced by up to 50 percent with simple contouring, and possibly by 75 percent with strip cropping combined with cultivating along the contour. This also helps to conserve moisture in situ. Contour fanning involves cultivating and planting the crop across rather than down the slope, as nearly as possible on the contour. There are variants ranging from growing only one crop at a time on the contoured land to stripcropping with equal bands of two or more crops running across the slope, thus combining crop rotation and contouring. Another alternative, usually confined to steeper slopes, is to incorporate buffer strips of perennial crops, most commonly of grasses such as Vetiver, or grass-legume mixtures at intervals between the other crops being grown on the contoured fields. It is desirable in mechanized farming to plant field borders of grass at each end of the contoured field to allow turning of implements. Contour farming does not necessarily require major changes in cropping systems and may even increase flexibility, since more than one annual crop or annual and perennial crops can be incorporated into the contour system at any one time. Moreover, conservation tillage techniques, including no-till, and residue management can be practiced on contoured fields, although herbicides have to be applied very judiciously where more than one crop is being grown or where annuals are alternated with perennial crops, or with grass strips (or both). Contouring, conservation tillage, and terracing may also be combined, especially on steeper slopes (over 9 percent), where contouring alone may not give sufficient protection. These techniques have the merit of being relatively easy to apply once the basic principles are grasped by farmers and they understand how to lay out the contour lines and strips. They are also inexpensive, only adding a few dollars per hectare to conventional cultivation costs if single crops are planted on the contour and a little more for strip cropping. Establishing buffer strips generates an additional expense for seed and fertilizer, but once established only maintenance costs are needed, for example for weed control, and for drainage of excess water. Consequently, a number of Bank-funded projects have attempted to introduce or to encourage the adoption of contouring, often associated with broad-based terraces, especially on less extreme slopes in hilly land and in watersheds. These projects include Argentina (Agriculture Credit II), China (Gansu Provincial Development), Tunisia (N.W. Rural Development), Jordan (Agricultural Credit), Turkey (Eastern Anatolia Watershed Management), Morocco (Loukkos Rural Development, Oulmes-Romani, Agricultural Development, and Fez-Karia-Tissa Agricultural Development), and Syria (Southern Region Agricultural Development). In a few cases conservation tillage practices on this contoured land are also advocated. Since these and some other projects combined terracing with contouring, the results will be 39 discussed after the review of terracing techniques because the constraints that impeded effective use were usually common to both. Terracing Terraces are embankments around hillsides constructed of earth, stones (or both), which slow down run off and allow soil particles to settle out, thus reducing erosion and sediment load in drainage water. On highly erodible soil, they also help to stabilize slopes and to prevent landslides. Depending on their width, they may be cultivated with annual or perennial crops, mechanically or by animal draft equipment, or planted with perennial grasses, legumes; shrubs, trees or some combination of these. Their spacing is determined mainly by the steepness of the slopes, soil type, and whether other conservation practices are incorporated. Three main types of terrace exist. * Broad-based terraces suited to arable cultivation (particularly when combined with residue management techniques) are laid out running parallel around the contour. Normally they are relatively flat in appearance and are confined to slopes less than 9 percent, with farmable slopes not exceeding 5:1 (a 5 foot horizontal drop to every foot of vertical drop). Both front and backslopes are farmable. * Grassed backslope terraces are those with a 2:1 backslope which can only be farmed on the frontslopes (downhill side of the backslope). * fNarrow-based terraces typically have 2:1 slopes on both the front and the backslope. Terraces may be designed primarily to collect and store water until it can infiltrate into the ground or drain away from a stable outlet (commonly an underground pipe) or as a channel to slow runoff and carry water to a stable outlet such as a grassed waterway. Storage terraces should be able to control run- off from a ten-year frequency storm and also to hold ten years of sediment accumulation. Terraces must be maintained to keep them operating efficiently. This includes removal of sediment buildup in the channel and around the intake, repair of eroded or otherwise damaged sections and intakes, weed and rodent control, seed and fertilizer treatment to maintain good vegetation on front and backslopes, and maintenance of grassed waterways used as drainage/runoff outlets from terraces to prevent gullying. Terraces, except on very steep hillsides or where narrow rock terraces are made by hand, are generally constructed mechanically, and in the United States the earthworks alone may cost from $500- 620 per hectare, excluding maintenance and upkeep. For handmade stone walls a figure of $500 ha has been suggested for Tunisia). However, the fact that terraces are being abandoned or poorly maintained in some mountainous areas with once magnificent hand-built terrace systems may be indicative of unsustainable costs of repair and maintenance, and/or the lack of labor due to migration or off-farm employment. Thus priority consideration should be give to erosion control through a combination of conservation tillage, contouring, and strip-cropping before embarking on a more costly terracing program that may not give a much higher degree of protection. Vegetative Cover Most of the projects reviewed employ vegetative as well as physical/mechanical techniques to help stabilize soils, and to prevent excessive run-off, erosion, downstream flooding, and siltation of dams and irrigation systems. In addition to residue management, these measures include planting cover crops, buffer 40 strips, grassed waterways, grassed markers for contouring, development of natural or sown pastures, and shrub or tree plantations. Conclusionsfrom Project Experience Experiences with these various approaches to soil and water conservation have been quite mixed: They are summarized below by reviewing those individual projects of which they are important components. * Syria, Southern Region Agricultural Development (1987). Destoning and Terracing. Serious erosion occurred because terraces were not constructed soon enough after destoning, leaving bare soil exposed to water run-off on steep slopes. Because of a general suspension of the collaboration between the World Bank and Syria, the project was never completed. * Tunisia, N.W. Rural Development (1981). Contour terraces were to be sown to pernanent pastures on steeper slopes, and some rock terraces planted to trees. These measures succeeded where they did not disturb local farming practices, but conservation tillage, contouring, vegetative strips and grass markers on slopes of 2-8 percent in agricultural zones involving a large area of farm land were resisted. The project failed because it did not take adequate account of the difficulties involved in contouring areas of small scattered plots and did not capture the hearts and minds of local people, but also because it was incredibly complex and in many respects conceptually unrealistic and over- optimistic. Costs per hectare for contour markers on lower slopes appear high by U.S. standards, and those for terraces are low. This project seems to have failed in its most crucial objectives of reducing erosion and increasing on-farm crop productivity, despite subsidies for inputs, energy production, and incomes. Imposed on-farm soil conservation measures including contour farming, grassmarkers, and gully control, and attempts to replace cereals by forage legumes or permanent pasture were resisted by the farmers. Transferring soil conservation and natural pasture improvement under direct management of the project had limited success. * Turkey, Erzerum Rural Development (1982). Terracing for soil conservation and channel control, mainly to reduce flooding, was implemented in eight sub-projects totaling 1,775 ha. This was considered to be successful in achieving its objective but the costs and benefits are not specified in the audit. Physical infrastructure targets were generally well met, but were beginning to deteriorate by 1992 due to lack of maintenance. * Morocco, Loukkos Rural Development (1980) This was a pilot project for land management systems in an eroded mountainous area. It involved contouring on farms in mid-altitude zones, strip planting of pastures or shrubs in abandoned farmlands and on some farms on steeper slopes, stone terraces and gully control dams. Contour strips on farms were not well accepted and had to be replaced by private grass/legume plantings on two-thirds of the farms, while the erosion works proposed for higher altitudes had to be modified as the slopes proved too steep. Forage shrubs (Acacia cyanophilla) did well and 41 increased feed supply in eroded collectively owned land. Costs of establishing forest plantations and shrubs were seriously under-estimated, however, and targets for forest maintenance were not meet. Little data is available on the adoption of other proposed on- farm anti-erosion works. China, Gansu Provincial Development Project (1987). Rainfed terracing to be undertaken on 74,000 ha in the Gansu River Basin included wide benches on 13,500 ha for cropping of existing arable land with 15-25° slopes; narrow terraces planted with shrubs on 21,000 ha exceeding 25 percent slope, and grass and legume planting on 40,000 ha of steepest slopes at higher altitudes in order to shift the focus of fodder production and grazing from the lowlands. This is an area of high erosion hazard but fertile soils. Implementation has been quite successful, using mainly hand labor. * Algeria, Pilot Forestry and Watershed Management (1992). The watershed management component on two microcatchments totaling 10,400 hectares involves research on destoning, anti-erosion contour and contour rock-lines, improved fallow and range management, investments in on-farm fruit and fodder trees, pasture establishment, small livestock, spring capping and water storage. The appraisal document contains a first- class analysis of the reasons for the lack of success of many soil and water conservation measures, but its diagnosis of the nature of erosion is narrow and dogmatic and makes no reference to the crucial issue of the shape of the slope as a deterninant of runoff and soil loss. * Turkey, Eastern Anatolia Watershed Rehabilitation (1993). This is not a pilot project, but probably should have been since it is very complex and involves large areas of uncertainty. To counter this uncertainty, the project concept is flexible, and the project relies heavily on the cooperation of the local population for its success. It approaches watershed management as an integrated system, rather than as an uncoordinated set of piecemeal initiatives. Soil conservation components include (a) support for mechanically prepared dryland bench terraces on sloping fields to be planted to cereals, annual and perennial legumes, grapes, and tree crops, (b) conservation tillage using sweeps (thus not zero tillage) on slopes over 4 percent (equipment and methods for preparing seedbeds on slopes exceeding 4 percent is signalled as an applied research need), (c) improved management and rotational grazing on common rangelands, and (d) soil ripping and conservation with earth terracing; corn planting and forage/pasture seed broadcasting in forest areas. In addition, gully rehabilitation through checkdams and tree planting will be undertaken. The project aims to improve 54 degraded microcatchments covering 400,000 ha in seven years as part of a much larger effort to protect the upper Euphrates watershed. Audit or completion results from other Bank-funded projects were not available, but the above projects at best had mixed success, and tend to have common causes of failure. First, over-ambitious targets were set for crops, especially forage, pasture (range), and for livestock production. Second, there were often unrealistic assumptions concerning the willingness of farmers to change their cropping patterns, and especially to reduce cereal production in favor of forage legumes and permanent pastures. This reluctance may have been due in part to government price supports for cereals, 42 and in part to farmer risk perceptions, both with respect to the potential reduction in their main subsistence and income earning crop (wheat), and to a perception that legumes are a less reliable crop. Third, underestimation of the difficulties of persuading farners to adopt on-farm soil conservation measures such as conservation tillage, contouring, and terracing, partly because these were unfamiliar to them and involved new machinery, planting techniques, and extra costs, but most particularly because land tenure and the layout and size of their farms and fragmented plots did not permnit individual farmers to adopt them routinely even if they had wanted to. The common property nature of natural range and even of some forest grazings was a fourth problem. The degree of success achieved in their improvement seems to be related to the success of the project staff in enlisting the cooperation of the users, and the incentives provided to persuade them to stock and manage the resources conservatively as in the Middle Atlas project in Morocco. A fifth cause of failure was lack of involvement of the local people (communities or individual farmers) in the planning, implementation, and evaluation of the projects. Although almost all of the appraisals contain some references to participatory approaches, the audit and completion reports do not shed much light on how much was achieved or by what methods. At best (Middle Atlas range improvement) it only succeeded in part of the project, in some cases (Northwestern Tunisia Rural Development Project) it seems to have largely involved top down enforcement and even expropriation. Even the 1993 East Anatolia Watershed Project in Turkey, which emphasized participation, presented local committees with a pre-defined set of techniques; the committees were hardly involved with selection of these. The general picture that emerges from most projects is that local people accept works, whether infrastructural or agricultural that offer them incentives or do not disrupt normal resource use and farming practices. Local people resist those techniques that appear to threaten their rights to land or their livelihood. The task of persuading them to forgo some freedom or short-term gains in the name of sustainability, so as to improve their future lot or that of their children has proved difficult. Expecting such sacrifices voluntarily to prevent externalities that principally benefit people they do not even know may be virtually impossible. Sixth, in most of the projects, difficulties were experienced because of a lack of local knowledge and understanding of the real problems of the area involved, because of inadequate or wrongly targeted research, and too few and poorly trained extension staff. Often these problems could not be corrected during the duration of the project. For example, although some projects had built-in research components, neither the time nor the funds were available to achieve useful results before project completion. This contributed to misunderstandings in project conception both with respect to technical and to socio-economic constraints, and hence to failures to enlist local support and to propose realistic goals and ways of achieving them. Seventh, many of the projects reviewed aimed at rural, agricultural, or watershed development, usually with some difficult mountainous terrain. Thus they tended to involve land use on farms, rangelands, and forests, at various altitudes, requiring a complex web of infrastructural, agricultural, pastoral, and forestry investments for their success. Because these cut across both central and local government as well as community rights and responsibilities they often required an equally and sometimes excessively complicated set of management and administrative arrangements to coordinate, finance, and supervise the project components. Finally. several of the technical aspects of the projects were not part of nornal farm or natural resource management practices in the areas concerned (conservation tillage, contour farming, terracing, gully control, various forage and shrub management techniques, rotational grazing, etc.). While it is not always clear from the appraisals how well these were understood even by the planning staff, it is clear that they were rarely well understood by the local technical staff. Nor, apparently, is there sufficient 43 recognition by the Bank of the need for social scientists and for multidisciplinary interaction in project planning and implementation. The various Bank-sponsored research and extension projects in the WANA region make some references to the need to address this issue, but the research projects still appear overwhelmingly commodity and especially plant breeding oriented. The extension staff generally lack sufficient subject-matter support on the technical as well as the socio-economic aspects of natural resource management, and the village-level workers are not properly trained to deal with the complexities of this type of project. WATER HARVESTING AND USE Future conflicts in the WANA region may well be related to water rather than to oil, and such disputes seem likely on an increasing scale not only among but also within countries. Two main reasons are the progressive exhaustion of the irrigation potential in most countries, and the rapid growth of population, especially of urban population. Demand for water from principal cities with significant political influence on governments is already leading to transfers of water from rural areas to meet urban needs in some countries, with shifts of villages to more remote areas and the abandonment of wildlife reserves. At the same time, water rationing is being introduced in some major cities. This paper will not review technology for irrigated farming per se, but will discuss raising the efficiency with which natural precipitation is used by plants, optimizing its availability within the fanning system, and supplementing it where possible through harvesting and storage of runoff or flood water (either above ground or through recharge of the water table). This must be considered a high priority. Several complementary approaches can contribute to this objective, including: * Breeding crops and animals with good tolerance of drought * Drought-avoidance farming techniques such as early or late planting, use of early maturing varieties, weed control, and timely harvesting * Maintenance of healthy plants and animals through breeding for pest and disease resistance, prophylactic or curative treatments to supplement inherent resistance, adequate and well-balanced nutrition, and provision of shelter and water to livestock * Adoption of cultural practices that conserve moisture efficiently in situ and reduce run-off, for example, conservation tillage and residue retention, contour farming, terracing, cover crops or other measures to increase soil organic matter, and in some circumstances, fallowing.' * Water harvesting, either through spreading run-off or flood water directly on to agricultural or range land, or through its collection and storage for subsequent use for irrigation, human use, or to provide drinking water for animals '7 The benefits of fallow with respect to moisture conservation, particularly where weeds are allowed to flourish to provide livestock feed, are extremely controversial, however. So is the value of deep plowing, even though this may help to control weeds. 44 Supplementary irrigation, to complement natural precipitation in meeting the needs of crops and trees. Box 4.3. Features of Water Harvesting * 7 Water harvesting is applied in arid and semi-arid regions where runoff often has an intermittent character. Because of the ephemeral nature of flow, storage is an integral part of water harvesting systems * Water harvesting is based on the utilization of surface runoff, and requires a runoff producing and a runoff receiving area. * Most water harvesting systems use water only near where it falls (Matlock and Dutt, 1986). They therefore do not include the storage of river water in large reservoirs or the mining of groundwater. * Water harvesting systems are relatively small-scale operations in terms of catchment area, volume of storage, and capital investment Source: Boers and Ben-Asher 1982, Bruins 1986. The first three approaches are primarily biological in nature, whereas the next three are essentially physical and mechanical. These physical techniques reduce also the more devastating downstream effects of uncontrolled runoff through physical damage to crops and structures, and siltation of water storage reservoirs, and by helping to recharge the water table. Thus any strategy designed to conserve both soil and water should take an integrated and holistic approach to the planning and implementation of these various complementary measures. This is also mandated by their ecological and site specificity. This section will concentrate primarily on water harvesting and storage techniques, which received surprisingly little direct attention in the projects reviewed, either as a means of augmenting ground or surface water availability, or as a source of supply in their own right. Water Harvesting This is probably the oldest approach to supplementing natural precipitation, ranging in complexity from merely constructing stone barriers or small dams to check the flow of surface runoff water and spread it over cropland (with or without storage) to quite sophisticated underground structures designed to tap the water table and channel it downstream to a point where it can be used above ground. Water harvesting , sometimes described as "runoff farming", is often used as a term for a range of measures for collecting, storing, and utilizing runoff water from different sources, either for crop production, livestock, or domestic purposes. The main source of water for harvesting is usually rainfall, and runoff flow may be overground, through stream channels or cross-country; or sub-surface, either through flow to rivers or to groundwater reservoirs. Reij and others (1988), describe common terminology in depth and note that there is a considerable ambiguity in the various descriptions used by their sources. Nevertheless they agree on the common features of water harvesting (see box 4.3). Some water harvesting techniques may be specifically 45 designed to recharge the underground water table, or to capture sub-surface rather than surface runoff and eventually bring it into use for surface irrigation or domestic water without pumping or other lifting devices (ancient structures known as qanats in Iran, foggara in the Yemen and Saudi Arabia, and karezes in Baluchistan). These are stone shafts to trap water, leading to tunnels up to 5 km long, sloping down the underground water table, to outflows serving surface canals. They are currently expensive to construct and to maintain and are therefore used mainly either to irrigate high value crops (especially fruit trees and vegetables) and/or to provide domestic water. Other methods of classifying water harvesting systems are based on using criteria such as: * Source of water (streams, groundwater, rainwater), and the form of runoff which may involve large, erosive flows from long catchments (floodwater harvesting), or sheet runoff from short 50-150 microcatchments (rainwater harvesting) * Catchment and storage techniques, an elaboration of the above in terms of storage * Form of runoff and water conveyance (water spreading; stream diversion, microcatchment) * Geomorphology, from small to large systems, including short microcatchments (less than 100 m), terraced wadi systems with check dams to retain and infiltrate runoff water, hillside conduit systems, liman systems (checkdams impounding arable land with spillways for excess water), wadi floodwater runoff systems to raise water on to adjacent fields. Most of these items are included in the classification developed by Pacey and Cullis (1986). Environmental factors, including climate, physiography, soils, vegetation and the interaction of these, greatly influence the potential for water harvesting. In terms of climate, a lower limit of 100/150 mmn precipitation is considered critical for water harvesting in the arid zone with a winter rainfall climate and 200 mm with summer rainfall. The chances of success are much better in the semi-arid zones. However, the rainfall characteristics are more crucial than the total rainfall, hence the number of rainfall events producing runoff, their frequency, and the probability of a certain minimum level and intensity are better predictors. The high variability of rainfall in arid regions creates problems of design in water harvesting structures. In order to impound enough runoff water they have to be large. Thus the damage resulting from unusually heavy rains is likely to be greater. The size of the catchment area, the permeability and gradient of the slopes, and the length of the slope are important determinants of runoff and also of erosion. In general, runoff efficiency decreases with larger catchment size, due in many cases to the lower average depth of rainfall and especially to greater water losses from infiltration, evaporation, and surface storage along the extended flowpath. North-facing slopes tend to produce greater runoff than south-facing slopes in the northem hemisphere, and hilly areas with rocky slopes generate more runoff than undulating plains with permeable soils. Incised narrow streambeds invite stream diversion techniques, whereas wider channels are more suited to terraced wadi systems. Relationships between slope gradient, slope length and form, erosion, and runoff are complex; steeper slopes produce more erosion, but not necessarily more runoff. Soils that have a high clay fraction, or are shallow, and have numerous rock outcrops enhance runoff. The influence of stone clearing on runoff is controversial. Soil type and vegetation cover are important factors in defining run-off. There is an apparent conflict between on the one hand policies that encourage water harvesting, the success of which depends on securing adequate quantities of runoff water but which may simultaneously 46 increase erosion, and on the other soil conservation policies aimed to prevent or reduce erosion and conserve more moisture in situ or through infiltration to the ground water table. In the Bank projects reviewed the favored policy was to conserve or infiltrate water in situ, rather than promote ex situ water harvesting. However, water harvesting is now receiving increased emphasis by development assistance agencies and national governments, for example in Syria, Jordan, Tunisia, and Morocco. The aforementioned report of Reij and others (1988) distinguishes four categories of water harvesting techniques. These are (a) water harvesting from short slopes (circa 50 m -150 m microcatchments), (b) water harvesting from long slopes, (c) floodwater harvesting within the streambed, and (d) floodwater diversion. More details are available in that report. The farmer's choice of crops for use of harvested water is determined to a considerable extent by the water harvesting technique, which in turn depends largely on the environmental factors discussed above. Very broadly, cereals and most other annual crops require an even distribution of water while trees require water to be concentrated and ponded at specific intervals, as is possible with microcatchments. In some cases even distribution is not feasible without land leveling. Waterlogging is a common problem, so crops that have low tolerance such as pearl millet or grain legumes may be at hazard. So are many tree species, including citrus. Sorghum, because of good drought and waterlogging tolerance, is a useful summer cereal for monsoon water harvesting systems, and barley for Mediterranean winter rainfall climates. Fodder crops and grasses are often relatively drought tolerant, require a less even distribution of water than annual grain and cash crops, and help to bind the soil. Lower economic returns imply the use of low-cost water harvesting techniques, however. Water harvesting systems based on short slopes are very largely used for trees, including fruit, fodder, forest, and shrub species. Long-slope water harvesting systems tend to be more flexible and, depending on the technique and the evenness of water distribution, may be planted to fruit or multi-purpose trees, annual crops, forage crops, and perennial grasses, or used for range reseeding. Streambed flood water harvesting systems and flood water diversion schemes may be utilized for olive or fruit tree (almond) production, with undercrops of barley (as in Jordan, Tunisia and Libya), or for annual crops or forages in pure stands. Cultivated reservoirs are normally used for cereals such as wheat, barley, millet, or sorghum, but in Israel the liman terraces are used for fuelwood production from Eucalyptus species. A problem with this technique is that rains arriving some time after crops are planted may damage them severely if the reservoir is not equipped to prevent unseasonable entry of runoff water. In examining the costs and benefits of water harvesting systems Reij and others (1988) note that empirical evidence that water harvesting will lead to substantial short-term increases in yields is limited. No economic analysis of a water harvesting project in WANA was available. One Israeli analysis on the effect of water harvesting on almond production found a very marginal profitability, requiring further refinement of the technique to increase soil moisture storage, or a change in the tree species to make the system profitable. The study stresses the need for better data so that the value, performance, and costs of water harvesting projects can be properly assessed. The authors provide a range of estimates of labor requirements for different techniques, which are between 40 person-days/ha and 80 person-days/ha for earth bunded microcatchments, between 140 person-days/ha and to 185 person-days/ha for contour works, and 380 person-days/ha for larger trapezoidal bunds for harvesting water from long slopes. In a number of cases the potential returns to water harvesting can be limited by low soil fertility. Constraints to the success of soil and water conservation measures identified in reports from China and WANA suggest the following technical and socio-economic issues: * Poor maintenance of the conservation structures is more commonly due to the lack of participation of the beneficiaries, because structures are communal or are too large for the local labor force to maintain. Imposed top-down decisions maybe resisted, especially if 47 these affect farmers or community freedom of choice over use of their land and water resource. Areas or projects may lack proper equipment. While some works can very adequately be constructed by hand, a combination of hand and mechanization is often needed-especially for larger structures or where rocks have to be collected and transported to the work area * Once works are completed incentives payments and other benefits such as cash wages or food for work on may be discontinued. * There is often nadequate local understanding that soil and water conservation is a social responsibility, and that environmental degradation affects the whole community. This problem is aggravated by the fact that actions in upper watersheds can cause damaging downstream externalities of which the perpetrators may not be aware. A related issue is lack of knowledge and experience of farmers in the construction and maintenance of the necessary works, which is often compounded by weak extension effort, with village level workers themselves poorly trained to advise farmers and others in the technique. In such cases help form local private sources of expertise may be essential. * There is often a shortage of seeds or plants to follow recommended vegetative techniques of conservation management, and knowledge of how to establish and manage the grasses, shrubs, or trees that they can obtain. This is partly a function of the lack of proper nursery arrangements, but there are also research, educational, and financial problems, since the optimum species and management practices are often not well known either to the extension staff or to local people, who may be reluctant to invest their own money in unfamiliar species. These examples show the difficulty in generalizing the potential for water harvesting. It is likely to have limited success where the main dependence is on spreading water from relatively unpredictable flash floods, especially where their potential magnitude requires large relatively costly works so as to prevent their premature destruction. A pre-requisite for the success of small water harvesting and storage projects seems to be a fairly adequate and predictable flow of water that gives farmers an incentive to plant orchards or other remunerative crops, relatively low costs of construction, a potentially productive land resource in close proximity to the structures, and a long enough life to yield a viable economic rate of return. The latter implies a structural ability to stand up to flood pressures (that may raise construction costs to uneconomic heights), prevention of excessive siltation, and (where not required to replenish the water table), prevention of excessive seepage; and relatively easy maintenance--preferably by the beneficiaries. A further critical factor concerns the ability of the planners to enlist the active support and involvement of the local people in the planning, construction, and subsequent maintenance of the necessary works. It is clear from several project completion and audit reports that this has rarely been attempted in the past, and that because the beneficiaries could not internalize the benefits, they resisted either passively or actively. In some projects compulsion was attempted, even though a clear lesson from attempts to introduce soil and water conservation measures and to promote sound environmental management in industrialized countries is that compulsion should be a last resort rather than a first priority. 48 Supplementary Irrigation Since any use of water to augment natural precipitation could fit the description of supplementary irrigation (SI), the technique is open to many interpretations. In this study it is assumed more specifically to include: * Providing crops with water from small streams, wells, cisterns or other storage basins on an ad hoc basis to avoid unpredictable losses from drought emergencies caused by intermittent dry spells during the rainy season * Using such water more precisely to ensure adequate supply to meet physiological need at critical stages in the growth cycles of key crops * Supplying crops with water either in advance of the main rains in order to facilitate early planting or to extend the season at the tail-end of the rains, usually in order to plant a second crop. * Using "deficit" or "limited" irrigation, in situations where both rainfall and water for irrigation are in short supply, so that crops are under some water stress even with SI All of these situations may apply to semi-arid regions where there is some precipitation but where it is either not sufficient or too erratic--or both, to support a good crop--even one with high drought- tolerance, in most years. A further situation can exist where rainfall is sufficient for rainfed production of relatively drought tolerant crops, such as cereals, but additional water can either boost their yields or allow a higher value but more water demanding crop to be grown. This is the case in some areas of WANA where average winter rainfall exceeds 600 mm, as well as in temperate regions of Western Europe to support intensive grassland production. In most cases, however, especially in semi-arid areas, water rather than irrigable land is the main constraint, hence the goal should normally be to achieve optimum output per m' of applied water, rather than to maximize yield per hectare. Probably the least satisfactory conjunctive use of rainfall and SI from the point of view of water use efficiency is where irrigation is supplied ad hoc to mitigate drought periods during the rainy season. Rainfall occurring unexpectedly soon after irrigation may then be of limited use to the crop, especially if it is not at a critical period of growth and the soil may become saturated leading to waterlogging and/or considerable runoff. When the soil becomes saturated late in the growing season the capacity to benefit from and store rainfall prior to the next rainy season is reduced, whereas this is essential to maximizing conjunctive use of rainwater and irrigation. Stewart (1988) describes several limited irrigation techniques being tested in the southern Great Plains of the United States, where drought resistant crops are grown in both rainfed, irrigated, and conjunctive water use situations. A limited irrigation-dryland (LID) farming system has the goal of maximizing the conjunctive use of the variable growing-season rainfall with a fixed and limited irrigation component. Its unique feature is to provide for flexible adjustment of the amount of land irrigated so that more can be irrigated during the years of higher rainfall. A "surge irrigation" system that allows the irrigation of a larger area with a limited water supply by cyclical interruption of the water to furrows and borders, thus decreasing infiltration rates, and increasing water advance rates over the system, require fairly complex layouts of beds, furrows, and furrow dams. A major goal is to minimize water loss from either excessive percolation or runoff. Water is generally applied only to alternative furrows. Fertilizer can be applied in the irrigation water or to the land at right angles to the rows prior to bedding. 49 Box 4.4. Benefits of Supplemental Irrigation in Syria . .. - - . - : .: ' ; - .' ..: . . . -. .--.:.- . . . . .. ,., : .: : * Increases ofapproximately 100'pcent in 'rain and straw yields of ereal:oand 'Se-.' -'- A 'drairatic reduction of 52 percent in yield variability of wheat ' Maj'o'r shifts cropping patternsaway from low-value barley and.pulses to highervalue rops-' L1Increased crop intensities from 85 percent rainfed up to 150 percent with Sl, accompanied by the elimination of falow.' SSource: ICARDA 1991b, 1992a. Combinations of these techniques, for example of surge irrigation with LID, and the use of conservation tillage to utilize residue produced by one crop to enhance the storage efficiency of a fallow prior to a subsequent crop and to control weeds in the fallow; have led to increased water storage, water use, grain yields, and water use efficiency. No-till proved significantly superior to tillage with sweeps or disks in this experiment. Although some farmers apply SI in WANA countries, data on the type of water application, the volumes used, and under which conditions SI is used, are very poor. Empirical data on the potential in WANA is also limited, but trials show yield increases from SI of 50 percent to 500 percent. Barley and wheat cultivars adapted to higher rainfall conditions responded best, highlighting the need for careful matching of varieties to the changing moisture regime. In addition to cereals, trials were undertaken on rapeseed and sunflowers, which also responded positively to SI, especially with appropriate selection of varieties, plant spacing, and cultivation practice. Experiments showed that deficit irrigation of rainfed wheat to 40 percent of the additional water requirement reduced yields compared with full SI, but was still profitable. On-farm trials indicate that farmers tend to over-irrigate, thus both wasting water and depressing yields below those attainable with more efficient water use (I CARDA 1994). T'hese gains are not costless, however, since they involve higher inputs, more labor, and water charges. Consequently, financial returns per kg produced with SI may be lower than those from rainfed production except in very dry years. The great advantage of SI lies in the higher volume of output per ha and the unit value of product. Results of village-level surveys in Syria also show that successful use of SI involves a considerable learning process which many farmers do not fully master. An emerging problem there is a rapidly falling water table due to an expansion in the number of wells and over-pumping. Whether for full supplementary irrigation this poses a serious threat to the long-term sustainability of the gains described above. ICARDA is moving cautiously with respect to supplementary irrigation. It reports that a survey of practices in Syria will be published shortly, and that it hopes to extend trials and demonstrations to other WANA countries by 1998. It appears that most farmers prefer to irrigate their crops fully where they have a well, either because their land is limited, or because they want to grow higher value crops or crop year- round. The 1991 Tunisian Agricultural Expenditure Review states: Investments [in SI] amounted to TD 9 million ($US 11 million), about one-third of the initial target." Part of the equipment is used for other crops, as farmers often felt that the investments were unprofitable if used only for wheat. Many believe that supplementary 50 irrigation of cereals is a "last resort" when rains fail, instead of using it to assist timely establishment of the crop and to provide additional moisture to boost yields. The regions and conditions under which supplementary irrigation is financially attractive to the farmer should be clearly defined prior to any further large-scale promotional campaign. Supplementary irrigation of cereals, legumes, or other annual field crops is hardly reported by researchers in WANA countries other than Syria, and not operationalized in any of the Bank-funded projects reviewed, many of which have significant minor irrigation components. A more penetrating inquiry into the reasons for this could be worthwhile. ASSESSING THE COSTS AND BENEFITS OF SOIL AND WATER CONSERVATION MEASURES Financial data on the impact of soil and water conservation in WANA are lacking. Benefits cited in the Northwestem Tunisian Rural Development Project (Phase II) include fuelwood and additional fodder production from sylvo-pastoral plantings, rangeland conservation, and ravine plantings; increased cereal yields ranging from 30 percent-35 percent in a wet year to 70 percent in a dry year from stone walled terraces and contour banks, combined with improved farming techniques, increased yields and better establishment from pits or V-ditches for fruit trees; and fruit, fuelwood, timber and honey from various types of tree plantings and windbreaks. Estimated returns to farmers after calculating the cost of their labor contributions (paid in kind) for establishment and maintenance are listed below for the first five years of the project. Although some practices involved a net loss over this period, all measures yield a positive benefit by year 5, which is assumed to remain constant thereafter. No discounted cash-flow analysis is provided. One definite conclusion from the various projects reviewed is that measures intended simply to control erosion are viewed at best with reserve by local people, especially if the main benefits accrue off- site while the main costs are incurred in the project area. This suggests that pilot projects may be required to test and demonstrate the potential benefits to the local community before large-scale measures of the type described above are instituted, especially when they may require sacrifices of land, labor, or capital by the people before they can yield financial benefits. A recent World Bank sector review of soil conservation experience in Mexico takes an in-depth look at the extent of erosion in Mexico, the benefits of soil conservation, the constraints to technical change, the policy options for managing soil erosion, and the implications for Bank operations. The report describes the methodology and model used for economic analysis, including sensitivity analysis; with identification of the net benefits for two levels of erosion control in the two major agro-ecological zones of the tropical, and the highland and semi-arid states. These observations accord with those from other Latin American countries, as well as with those from Sub-Saharan Africa. It appears that erosion control measure are more likely to be profitable in tropical than temperate or semi-arid sub-tropical ecozones, and results from the analysis appear to confirm this. In fact, there are strong parallels between the behavior of farmers towards soil conservation in central and northem highlands areas of Mexico, and those of farmers in upland areas of WANA. The report states: 'Farmers' (in those regions) use few soil and water conservation practices. The only widespread technique is some incorporation of animal manure, although this is not universal. Most land preparation is done with mechanical discs, frequently up and down 51 the slope, not on the contour. Most locations do not promote the utilization of simple on- farm structures. There is the almost ubiquitous removal of maize stover from fields, exacerbating the nutrient loss to erosion and decreasing soil cover. Because of the severe risks of drought, frost and hail, varieties of maize and other crop varieties, with shorter growing seasons have been difficult to develop, and in consequence, little crop rotation is practiced and virtually no cover-crops are grown for soil protection. The paper discusses several possible constraints to reducing the costs of erosion in Mexico, most of which would also apply to other countries. These are: Unprofitable techniques. The highest barrier to the adoption of known techniques is that many would be unprofitable. Of the 128 combinations of production techniques and sites modeled for maize, only three were profitable at a real discount rate of 2.5 percent. None were profitable for sorghum (19 cases) and none for beans (36 cases). * Discount rates. High private rates of discount harm incentives to conserve soil if benefits from conservation accrue over such a long period that producers discount those benefits heavily, while having to bear the costs of conservation investments very soon. a Lower yields on farms than on research stations. The gap between crop yields on fiams and on research stations is between 50 and 100 percent for rainfed maize, sorghum, and beans. The gap exists because farmers do not fully apply available technologies, because it would be economically inefficient, or because they lack information. Eliminating that gap, where it is economically appropriate, would have only a small additional effect on the incentives to adopt soil conservation practices. Most investments would not be economical where maize yields are below 0.5 ton per hectare. Lower cost measures would have to be developed for such areas. * Complex technology. Farmers may not use proposed techniques if they are too complex; require drastic changes from their normnal farming practices, or are only feasible through group action. * Economic disincentives. Low crop prices, high taxes, and high material or equipment costs reduce incentives to adopt soil conservation techniques. Some of the constraints which inhibited earlier conservation efforts in Mexico have been alleviated by recent changes in price or other policies. For example, the three main rainfed crops, maize, beans, and sorghum now receive significant price protection. Should those supports be reduced the possibility of shifting to other, and possibly less-erosive crops on rainfed land, appears to be limited. The question of whether raising farm-level yields closer to those at experimental stations would provide a stronger incentive to conserve soil is treated somewhat ambiguously: under 'constraints'...the conclusion seems to be that it would make little difference; whereas it is noted that reducing the gap would raise the return to soil conservation because the base land yield would be raised. Presumably this implies an incentive to adopt conservation techniques. 52 A main conclusion of the report is that wherever possible the aim should be to encourage conservation techniques that can be adopted readily by individual farmers (especially those which maxinize vegetative cover), with minimum need to modify their normal practices or to invest in special equipment. This keeps costs down or may even reduce them (for example, with minimum tillage) while the benefits are generally achieved with less delay. However, even if it is economically profitable to undertake soil conservation investments, private action alone is likely to be insufficient to promote their adoption, due to intractable extemalities. Two major externalities are highlighted: that the private discount rate of producers is greater than the economic discount rate, and that producers cannot capture the full economic benefits to soil conservation investments because information about soil conservation techniques is a public good. As such, techniques to fight soil erosion would not be optimally supplied by the private sector and public investment in it would be necessary to reach optimal levels. Moreover, even if erosion control investments equalize marginal benefits and costs on-farm there are still additional costs of erosion off-farm such as siltation of waterways. This could justify some government support for preventative action. Overall, the report concludes that a government-supported soil conservation program is justified, through action to supervise works contracted to private enterprise with public financing; training programs for extension staff and farmers; and support to conduct relevant research and extension. Neither the Mexican project nor most others attempt to assess the externalities, that is, the quantification of reductions of 'downstream' effects of erosion, which may be extremely damaging (e.g. landslides, siltation of reservoirs et cetera), or partially beneficial (deposits of silt on farmlands below the source of erosion to increase soil fertility there-as in the Nile Delta. Four out of five Bank-funded projects located in watersheds did not do so (Spears and Rowe, 1991). Instead (as in Mexico) the focus of evaluation is the value or value-added of increased production and income from the land directly protected from erosion in the project area or for export. 53 Box 4.5. The Mexican Case Study Some of its major conclusions regarding the benefits of soil protection in Mexico are:.:.. Where erosion is severe in Mexico, yields can be reduced by 75 percent.. Even where it is moderate, 20 percent of yield is lost. This is due to reduced soil depth, loss of organic matter,: lower water-holding capacity, and loss of soil nutrients-raising the cost of fertilizer by 5 to 30 percent of the crop value. Average soil losses are calculated at 2.8 tons/ha/year, ranging from 4.3 t/ha/year in the dry tropics to only 1.1 .tlha/year in the semi-arid subtropics. * Techniques applied to control erosion in Mexico include contour fanning, terracing, field. drainage, vegetative barriers and windbreaks. Available evidence suggests that these techniques can reduce erosion substantially, but does not specify which techniques are most cost-effective and beneficial in terms of financial returns as well as with respect to erosion control. * Economic analysis suggests that most control techniques would not be socially profitable at the real economic discount rate at a tolerance level of 5 t/ha soil loss per year, and hardly any at a tolerance level of 1 t/ha/year soil loss for the three key crops in Mexico: maize, beans, and sorghum, which occupy more than 10.4 million ha (over 60 percent of the rainfed cropped area). A. prograrn to. support only socially profitable techniques of conservation by subsidizing the difference between the profits at the high private discount rate assumed to be applied by fanners and the lower social rate, covering 10.4 million hectares, would cost about US$307 million in discounted present vailueover 20 .years. The discounted present value of farmers' investments (borne by them because the benefits would exceed their costs) would be $374 million and another $547 million would be required for related agricultural extension at estimated cost of $6 per ha per year and supervision costs at $3 per ha'per year. A sensitivity analysis reveals some interesting information on this problem: (a) The private discount rate has little effect on the number of profitable techniques. Many techniques are so unprofitable at the lower social rate that raising the discount rate simply makes them more unprofitable. (b) The target quantity of soil erosion to be achieved with control measures is.very important. If the target loss to erosion is reduced to the lowest technically feasible level i.e. 1 tonlhalyear the cost of the subsidy component rises from US$307 million to US$2.558 billion. (c) High productivity impact from erosion control can probably only be achieved in severely eroded sites. Hence, the total program cost and benefits will vary dramatically according to the sites chosen. (d) A major conservation program would be justified even if the lowered off-farm costs of erosion control ate not counted, which was the procedure used here. Counting such benefits would not spur producers to invest in erosion control, however, because they would capture none of the off-farm benefits.. Therefore, a producer subsidy would be required to capture such benefits. (e) A major soil conservation program would be justified even if the benefits from fertilizer costs saved by erosion control are not counted. Counting such benefits would spur producers to make further investments in erosion coritrol, but would not justify any subsidies for the investments because the benefits would accrue fully to the producers and would do so immediately. 54 The Mexican Sector Review notes that although better information about the problems is worth having, not all of the information would have significant effects on recommendations for controlling soil erosion. More needs to be understood about the association (positive and negative) between farmers attitudes to risk and their willingness to adopt soil conservation measures. It is clear from the Mexican Sector Review, and from numerous other studies, that the magnitudes of soil losses from erosion can vary enormously, within countries and even within farms, as well as from year to year and within years. More importantly, a much more pro-active attitude is needed to soil conservation; and the Mexican Sector Review is probably correct in advising against further expenditure on studies of soil erosion losses there. Assisting in the establishment of facilities to permit the more systematic assessment of soil quality determinants and reference levels over time. This should involve measurement of runoff and sedimentation to assess externalities and effects of water quality, but whether any attempt to assess annual soil losses and their effects on productivity is worthwhile in view of the controversial nature of existing data seems doubtful. A more positive approach to building soil quality through research, and training extension staff to assist in the introduction of improved farming practices, seems a better use of resources. Box 4.6. The Role of the Bank in Financing Soil and Water Conservation The appropriate roles for the Bank lie in supporting co-financed investments to help small farmers, supporting the relevant investments made with commercial loans, institutional development to better provide specialized technical support and policy advice, and research and extension. There would be no Bank role in financing direct public works for soil conservation. Though money is fungible, some targeting could be achieved by: (i) giving additional funds to states and DDRS with more serious erosion, since the spatial distribution of the problem is well understood, and (ii) giving a higher financing percentage to soil conservation activities generally within all states and DDRs Source: Authors. Chapter 5 PROSPECTS FOR TECHNOLOGICAL CHANGE INRAINFED CROPPRODUCTION There are very large differences in the adoption and use of sound production technology in WANA. The labels "modem" and "high-yielding" are not always appropriate since under some conditions there may be good reasons for farmers to follow traditional practices. Perhaps the most striking example in this respect is a comparison among five regions in Syria in 1990, all in zones with annual rainfall exceeding 350 mm and relatively similar temperature and altitudes above sea level, which shows that one region has 28 percent of its land fallow and uses no modem varieties of cereal while all of the other regions have virtually abandoned fallow and have between 85 percent and 100 percent of their cereal area under modem varieties (ICARDA 199 lb) Moreover, the odd region also had the highest proportion of hand harvesting and a low level of herbicide use. FACTORS INHIBITING TECHNOLOGICAL CHANGE Macro-environmental factors, especially precipitation and topography, are the primary influences on farmers' choice of crops and cultivars, and on the nature and levels of adoption of crop production technology. Nevertheless, there are extremely important second-order factors that constrain the adoption of certain technologies within a given macro-environment and that may require in situ analysis to identify and understand. These include: * Local physical endowments such as soil depth and quality, stoniness, and slope. These are especially important determinants of the choice between wheat and barley, as well as of the number and type of tillage practices. The availability and reliability of supplementary irrigation is becoming increasingly a determinant of both technology applied to individual crops and the choice of crops in farming systems. * The availability and price of labor. With rapidly rising labor costs in WANA where out- mnigration and off-farm work is common, the availability or absence of family labor is an important determinant of the use of mechanization and the efficacy of practices, especially weed control and harvesting. Farm surveys in Morocco and Syria (Oram and others 1994) show that in some rainfed farming systems 60 - 80 percent of family income is derived from off-farm work either locally or overseas. * Infrastructure and accessibility to markets, advisory services, inputs, contractual services, and credit. These are often major problems in mountainous regions, and also in steppe lowlands that have low population densities. * Farm size, degree offragmentation of plots, ownership, tenure, and rental arrangements. Smaller and more fragmented farms prevail in the mountain areas, and sometimnes in higher- rainfall ecological zones. 56 Government policy toward agriculture generally (taxes, input subsidies, investments, credit), specific crops or livestock (support prices, state trading and purchases, acreage controls), and research, extension, and veterinary services. Finally, there are a number of third -order technological constraints to raising food production in the region. There is a limited use of cultivars with superior characteristics, especially in the case of durum wheat, barley, and both food and forage legumes. Historically, priority in genetic improvement of annual crops in WANA has been focused on wheat, especially bread wheat, and on the more favorable environments, where progress seemed likely to be most rewarding and where the "Mexican" wheats with high yield potential could be introduced (table 5.1). Other crops have received considerably less research effort. Consequently, fewer improved varieties of wheat have become available to farmers for upland areas (Mexican wheats lack the cold tolerance required for these areas), and of barley for rainfed ecological zones with less than 350 mm annual rainfall (Oram 1988). Table 5.1. Relationship Between Area, Percent of Crosses Made, and Percent of Scientists by Wheat Type (1990) Bread wheat Durum wheat --- . . al . --: Variable w.heat... . .. Spring inte Spring W : .........i: t ... Percent of area ; i73.5 ::i :15.2; 10.9: .:0 -0.4: :-. 100 Percentoffcrosses : : 73.3 Q ; : 13.7 0: 11.90;0:-0 11 -.;: 100 : ;Percet ofscienist 7,147 :102.5;0: 00;000000:00.50:;:.;. 01:000S- Source: Bohn and Byerlee 1993. In recent years, more attention has been paid to relatively neglected crops and zones, and an increasing number of varieties has been released. Even in the case of wheat, however, with 256 'improved" varieties released in WANA countries between 1966 and 1990, the adoption rate in 1990 was only 56 percent of the total wheat area in North Africa and 38 percent in West Asia (Belaid and Morris 1991). These figures include irrigated wheat, where adoption is often almost 100 percent. In comparison, in South Asia, Central Amnerica, and the main wheat zone in the southern cone of South America, the adoption rates were 88 percent, 95 percent, and 82 percent respectively. Considerable scope thus exists for further use of improved genetic material for all crops and situations. However, adoption studies in the region show that farmers are quite conservative in their choice of varieties, although, when they do shift, the evidence indicates that they rarely regress to their "unimproved" material. They may, however, continue to use their "new" varieties long after they have been outmoded by subsequent products of crop imnprovement. Seed availability delays the adoption of improved varieties or reduces their impact. These problems include (a) shortages and high costs of seed of irnproved varieties of cereals, pulses, and vegetables (especially forage and pasture species, because of inadequate government seed services and a lack of private seed industries), (b) poor seed quality and germination, mainly because farmers save their own seed to reduce costs (since many crops have high weed populations, both from wild species and "groundkeepers" from previous cultivated crops. This may perpetuate seed and soil-born diseases, as well as building up weed populations, diluting gains from adoption of new germplasm), (c) lack of seed 57 treatment against seed bom or soil diseases and pests, (d) highly variable and often excessive seed rates for the same crop, partly because official recommendations often fail to take local factors into account (Oram and others 1994). Tillage and seeding operations are often performed inefficiently, and delayed beyond the optimum time, leading to yield losses of up to 50 percent. This is because farmers cannot obtain or afford the best equipment for the tasks (for example seed drills), their holdings are too small or too fragmnented to permnit efficient operation of machines, or they lack draft power to use animal drawn equipment. Finally, farmers may deliberately delay weeding in spring to provide feed for livestock from grazing weeds. Average levels of fertilizer use are low, and farmers are reluctant to use any fertilizer at all in many upland areas and in areas where annual rainfall is less than 300 mm. This is often because of inherently risk averse-behavior on the part of farmers, but is sometimes because of inappropriate recommendations with respect to materials, quantities, timing, and method of application, or because the necessary formulations or applicators are not available or affordable. A common error of research is to provide extension staff with ideal solutions based on yield trials at research stations unsupported by on-farm trials and economic evaluation, and with no "second best bet" offers where local conditions may force farmers to adopt sub-optimal measures. Many farmers lack suitable application equipment and chemicals for weed control or these items have a high cost. Additional hazards are posed by the relative lack of research, the wide range of possible herbicides, and the considerable risks of phytotoxicity attributable to incorrect use, for example on food and forage legumes, or to imprecise research recommendations. In some countries continued use of post- emergence chemicals such as 2-4D to control broad-leafed weeds is leading to a dominant flora of resistant grassy weeds that compete vigorously with cereals and may vector diseases. A further important cause of farmer resistance to herbicides is the wish of those who keep livestock to use the weeds for grazing or to hand-pull them for feed, sometimes for human consumption. Finally, heavy harvest losses occur as a result of shattering and shedding of grain and lodging of straw, especially with local varieties of barley and with short-statured food legumes such as lentils. While combine harvesting may reduce losses with cereals, this is not necessarily the case with legumes, where high labor costs combined with losses in harvesting and storage are limiting to their wider adoption. Approaches to overcoming these difficult problems will now be discussed in more detail. GENETIC IMPROVEMENT OF KEY FOOD CROPS For the marginal areas of this study, genetic improvement should focus on improving yields and reducing their variability through developing resistance or tolerance (often used interchangeably in the literature) to abiotic stresses, for example drought, cold, salinity, waterlogging, and soil acidity. In seeking increased resistance to some factors such as drought, very complex multi-gene problems have to be overcome. Genetic improvement work requires expertise not only in plant breeding and genetics (increasingly now linked to biotechnology) but also in plant physiology, entomology, pathology, botany, and agronomy. A substantial share of the overall effort has to be devoted to "maintenance research", safeguarding existing gains against the ravages of biotic stresses caused by pests, diseases, and weeds.. Thus, table 5.2 shows that although wheat research programs most commonly give priority to abiotic criteria when delineating environments for wheat improvement, resistance to biotic stresses generally receives priority in their breeding work (CIMMYT 1993). Breeding work will not be discussed here in detail. A vast number of lines is under scrutiny; in 1989-90 alone ICARDA made over 1,000 durum wheat crosses and over 1,400 spring bread wheat crosses. Over 10,000 entries of winter and facultative wheat were evaluated in Turkey and Syria for cold and 58 drought tolerance. In the case of barley, 3,248 lines were evaluated in yield trials. Promising material was evaluated for a wide range of ecological habitats, from very dry or cold, to areas where yields were not severely limited by climatic stress, and for different management and input-systems. CIMMYT, through its global mandate, both collaborates closely with ICARDA, and operates its own massive wheat breeding program that provides further valuable material to WANA countries. Incorporating Tolerance of Abiotic Stress Characteristics that are important in assessing suitability to drought-prone conditions for barley and wheat include number of days to heading (earliness), plant habit, peduncle length, number of fertile spikelets, and tolerance of low winter temperatures. Semi-dwarf varieties with resistance to lodging, (for example the "Mexican" wheats) are more suitable for less rigorous environments with mild winters and higher and less variable precipitation, or for farming with irrigation. In the case of barley, two breeding strategies are feasible: to develop early-maturing varieties to escape terminal drought, which may give optimum yields but risks damage from late frosts, or to use vernalizing material adapted to low temperatures, photoperiod, and irradiation regimes, which has better frost resistance. This may give somewhat lower yields but better yield stability. Cultivars that fit neither of these criteria should not be released in areas subject to severe drought or cold. Crops at high elevations have to contend with early drought, prolonged cold, terminal moisture stress and heat because of the abrupt transition to high temperatures in spring. Therefore desirable characteristics differ from those in low elevation areas and include a long vegetative phase with slow primordia development in early stages to survive the severe winters, and a short reproductive phase with rapid grain fill and early maturity to escape tenninal moisture stress and heat. A high fertile tiller number and medium-size lax heads with bold grain is desirable (Tahir and Hayes 1988). Although wide adaptability (that is, yielding well across different environments) has proved valuable in low elevation areas, the great agroclimatic diversity in upland regions, especially mountainous areas with dissected terrain, makes this a difficult goal. 59 Table 5.2. Average Ranking of Breeding Combating Biotic Stress Factors Priorities for Spring Bread Wheat by 100 National Research Programs. In addition to screening lines for performance under abiotic stresses, breeders also l Yield Average ranking devote a major share of their effort to incorporating Yeld Average ranking resistance to specific pests and diseases into new cultivars. Table A21 in the annex lists the major Grain yield 4.8 pests and diseases of cereals and food legumes in Grain yield 4.8 WANA. Considerable success has been achieved in Yield stability 4.5 developing genetic resistance, especially with respect to important diseases, which are often more difficult Straw yield 1.4 to control chemically than insect pests or weeds. Quality Wheat. A major breakthrough has been Baking 3.1 made recently in lasting control of stem rust Protein 2.9 (Puccinia graminis) of spring wheat, a key cereal in WANA and northern China, by in-built genetic resistance. CIMMYT (1993) has estimated that this Biotic stresses alone is worth US$150 million a year worldwide. Diseases 4.4 Moreover, the knowledge of the wheat-pathogen Diseases 4.4 relationship gained in that effort seems likely to open Insects 1.4 the way to developing resistant cultivars to stripe and leaf rusts, two historic plagues causing large Maturity economic losses of wheat yields. Research is also Maturity underway to develop resistance to leaf blotch Earliness 34 (Septoria tnfiic), common bunt (Tilletia caries), powdery mildew (Erysiphe graminis), head blight Abiotic stresses (Fusarium culmorum) and barley yellow dwarf virus Drought 3.0 (BYDV). Some varieties have resistance to more than Heat 2.5 one disease. For example, the durum cultivar Lahn Cold 1.8 has resistance to Septoria, yellow rust, and bunt, as well as high cold tolerance. It appears well adapted Sprouting 1.6 both to upland areas and to irrigated production, and Acidity l 0 has proved promising in Lebanon, Turkey, and Acidity 1.0 Syria. Three other durum lines have resistance to all Salinity 0.6 three wheat rusts, one of the most promising cultivars, Brachova, is resistant to both Septoria Notes: 5 = highest ranking, 0 = lowest. blotch and barley yellow dwarf virus (ICARDA Interestingly this table indicates no priority to wheat 1992d Several un-named bread wheat lines under straw quality, nor to resistance to lodging. Wheat test at ICARDA also show multiple resistance, one straw is not highly prized for animal feed, and the line being resistant to all three rusts, common bunt, widespread adoption of semi-dwarf wheat cultivars Septoria, and yellow dwarf virus. Even if they have has greatly reduced losses from lodging. However, some negative traits such lines are very valuable as straw attributes are likely to have much greater parental material for breeding against diseases. importance in barley or triticale. Resistance to certain insects, including Hessian fly, Source: Adapted from CIMMYT 1993. 60 has also been incorporated into wheat varieties, these include the new bread wheat variety Saada that has been released in Morocco. Barley. Resistance to yellow dwarf virus is considered very important, since this disease is known to increase the susceptibility of barley to other biotic and abiotic stresses, including serious fungus diseases such as powdery mildew as well as aphid pests, drought, and cold. Some barley varieties have cold resistance or tolerance--including the important cultivar Tadmor, and these generally show higher tolerance of drought as well. A second main source of loss in barley, powdery mildew (Erysiphe graminis), is often treated with chemicals in Europe, and to some extent also in WANA, but resistant varieties exist and ICARDA is incorporating this resistance into new cultivars. Resistance to barley rusts (Puccinia hordeii) has been identified in Tunisia and in the Andean Region of South America, but does not appear to be highly stable. Aphids, especially greenbug (Schizaphis graminum) are a damaging pest of barley in Egypt, Ethiopia, and the Sudan, and resistance has been identified in Indian barley cultivars with other desirable characters, which are now being used for crossing purposes. Triticale. Resistant varieties of triticale have been identified to stem rust (Puccinia graminis), bunt (Tilletia species), and smuts (Ustilago) as well as to the cereal cyst nematode (Heterodera avenae). Legumes .Legumes suffer from a diverse range of insect pests and diseases (especially the latter) that greatly limit yields and raise costs to consumers, thus restricting demand and market opportunities. In addition, the parasitic weed broomrape (Orobanche crenata) is an important cause of economic losses, especially with faba beans, and it can put land out of use for legumes for a period of years. Some faba bean varieties resistant to Orobanche have been developed. Chickpeas .When cold-tolerant cultivars are sown in early winter, their yields can be more than double those of spring sown chickpeas. However, this practice can expose the crop over a longer period to unpredictable outbreaks of blight (Ascochyta rabiei), which has confined adoption of winter planting to about 30,000 ha in WANA. A major effort to overcome this disease (sometimes called Anthtacnose), which affects chickpeas, lentils, faba beans, and forage legumes in many countries. There is accompanying work to control another serious disease of these crops, Fusarium wilt, which is often associated with Ascochyta, and which is especially damaging to lentils. Some resistance has been identified to Fusarium in a few cultivated and several wild species. Resistance also exists to chickpea leaf miners and to podborers (Helicoverpa and Heliothis) that are a source of considerable loss of yield in WANA and in Australia, but best results have been obtained from combined use of resistant cultivars and field treatment with chemical sprays (Thiodan 35, 6 cc per liter in 500 liters/ha). Sunflowers. Sunflowers are becoming an important rainfed crop in parts of WANA. Red rust (Puccinia helianthi), and Sclerotinia stem and head rots are important diseases that seriously reduce yields in other regions. Hybrids with genetic resistance to rust are available, and early spring sowing to avoid warm conditions with free moisture on leaves, for example in irrigated areas, can also reduce incidence. Prolonged periods of cool moist weather during seed development predispose crops to head rots, and cultivation of legumes or leguminous forages in the rotation may maintain the pathogens in the soil. Varieties with field resistance are available commercially, but immunity to infection is unknown. However there seems to be little data on the cultivation conditions or use of sunflowers in WANA. In other important sunflower-growing regions, hybrid varieties are mainly used for oilseed production, which can out-yield the older open-pollinated varieties such as Peredovit of Russian origin by 70 percent. These hybrids also mature more evenly, reducing risks from delayed harvesting and bird or weather-related 61 damage, and improving harvest efficiency. Resistance to important diseases, especially rust (Puccinla helianthis), and to Orobanche is common. While bee pollination is beneficial to yield and oil content, it is not as essential with hybrids as with open-pollinated cultivars which have a high degree of self- incompatibility. However reliance on hybrids requires an efficient and probably a private seed industry, which is still the exception in most WANA countries. Experience in Australia indicates that in winter- rainfall ecological zones yields of sunflowers are low, and this appears to be the situation with rainfed sunflowers in Syria and Morocco, where supplementary irrigation can increase plant height, head diameter, and yield (Cameron and others 1991). Canola. Rapeseed (Brassica napus) has the potential to become an important oilseed in higher rainfall areas of WANA. Resistant varieties exist to two important diseases that cause economic losses: blackleg (Leptosphaeria maculans), and white rust (Albugo candida). Research is underway to breed herbicide and insect resistant cultivars with the help of biotechnology. Trials of rapeseed varieties show that they would be unlikely to succeed without supplementation where rainfall is less than 420 mm. Frost also reduced yields to about a third of those expected from experience in Canada and Australia. Rapeseed is currently of minor importance in WANA. Meeting Farmer and Consumer Preferences Through Breeding Screening genetic material for quality is also important. Some key quality aspects are discussed below. Appearance. Farmers in WANA prefer large white or amber-seeded bread wheats to those with red seeds and tend to reject promising new varieties for this reason. Similarly, triticale-a relatively new cereal in WANA--has adoption problems in Morocco because of the high proportion of small and shriveled grains and their brittle nature during threshing. Surveys in Morocco indicate that only about 15 percent of the triticale was retained on-farm, primarily for animals (especially sheep), consumption for human use being negligible. In the case of winter Kabuli chickpeas a main obstacle to their adoption in Morocco is also their smaller seed size (Herzenni 1988). Straw quality. Short straw and poor palatability is a main reason for farmer resistance to "modern varieties" that often have shorter and stiffer straws. While short straw may be valuable to prevent "lodging" in higher rainfall ecological zones where farmers use nitrogen and sell the grain, it is not valuable in drier areas where farmers rate grazing and straw yield and quality highly. Dual-purpose barley varieties with higher biomass to meet these needs are now in the pipeline. Ongoing work includes an attempt to cross wheat and wild barley--"Tritordeum". This has long and non-shriveled grains, unlike early triticale lines. It also has a protein content averaging about 23 percent, almost double that of wheat (Martin 1988). In Cyprus, a self-regenerating pasture barley is being developed by crossing cultivated barley (Hordeum vulgare) with a wild indigenous species, Hordeum spontaneum (Hadjichristodoulu 1988). Poor straw quality is also a main constraint to adoption of triticale in North Africa. Grain quali.y. A genuine obstacle to the adoption of triticale and the new winter-spring habit wheat crosses from CIMMYT which have yield propensity superior to the "green revolution" semi-dwarfs, is that their dough becomes sticky and hard to handle under high-speed mechanical mixing. In the case of 62 barley, the quality of grain for feed is important: farmers in Morocco dislike an otherwise promising new cultivar because its grain is too hard for sheep (Saade and others 1994). Box 5.1 Experience With Triticale in WANA Triticale is a potentially important new cereal for WANA farmers, either for grain or for forage. However, despite ten years of trials and promotion, its promise as an alternative to wheat in acid or waterlogged soils, or to barley in semi-arid zones, has yet to be fulfilled. Quality and marketability are the main factors limiting its use. Only two countries have a significant area grown on farms: Tunisia, with about 16,500 ha reported in 1990, and Morocco with about 5,000 ha. Other countries with some ecological analogies to WANA that have area devoted to triticale include Australia (16,000 ha), China (25,000 ha), Argentina (19,000 ha), Mexico (8,000 ha), and Chile (5,000 ha). Poland, the world's largest producer, with a large area of acid safety soils, has about 700,000 ha. Experience from on-farm trials in semi-arid areas of Morocco cast doubt triticale's suitablity as a replacement for barley under dry conditions. About half of the recipients of seed abandoned it after the first year (annex table A22). A cross-zonal transect survey revealed a somewhat better yield performance compared with barley, but many farmers continued to express doubts about triticale's grain quality (Primov and others 1988). In Tunisia, where Triticale has been adopted principally by large farmers in the wetter northem zone, it out-yielded barley by 22 percent in grain and straw. However, barley suffers from lodging and mildew in this humid climate: its comparative advantage over other cereals lies in drier ecological zones, and in saline, not acid, soils. In an overview of yield data from various sources in Tunisia, Saade (1993) concludes that triticale has a very modest yield advantage over durum (5 percent) and bread wheat (6 percent) in normal years, which can hardly compensate for its lower price. Its yield advantage is greater in drier years. Source: Survey data. Saade (1993). Table 5.3. Farmers Opinion of Triticale Compared with Durum Wheat and Barley (number of respondents) Durum wheat Barley (no. of respondents) (no. of respondents) Criteria: Better Worse Better Worse Early maturity 8 6 3 23 Tillering 10 7 11 15 Drought resistance 11 5 8 10 Resistance to excess water 5 5 10 2 Resistance to Sirocco 9 6 9 6 Disease resistance 12 1 14 2 Resistance to lodging 8 8 21 6 Source: Saade and others (1994). 63 CROP PROTECTION A wide variety of pests take a heavy toll on plants during growth and on stored products after harvest. Estimates of these losses globally and within individual countries are subject to large errors, because of the difficulties in obtaining data, but the economic loss is undoubtedly very large (Cramer 1967). Although economic returns to control of biotic stresses in rainfed crops are not well documented in WANA, the estimated magnitudes of yield losses in cereals and legumes from diseases, insect pests, and weeds reported from experiments and field surveys are often large and compound the effects of abiotic stresses. Consequently a high proportion of the research effort, both national and international, has been devoted to finding ways of reducing these losses. For exarnple, CIMMYT (1993) notes that about 28 percent of the breeding work on wheat in national programs is devoted to maintenance research-to prevent yields slipping back through weed and pest proliferation--basically running faster to stay in the same place! Various strategies to avoid, minimize, or prevent yield losses, including changes in fanming practices, use of resistant varieties, chemical methods of control, biological control measures, and integrated pest management (IPM) have been tried. Of these approaches the first three are in common use by farmers, the last two are predominantly in an experimental stage. What is actually happening will be described in more detail below. Changes in Farming Practices Fanning practices, which reduce the risks of severe loss because of pests include (a) adjusting cultivation practices, and planting or harvesting dates, (b) variation of plant population and density so as to moderate humidity, and (c) managing fertilizer use to discourage pest build-up. Use Of Resistant Varieties From the farmers' point of view, this is potentially the cheapest and most attractive of all approaches to minimizing losses from pests and diseases, since it only requires them to obtain and plant resistant cultivars, and needs no special training, chemicals, or equipment. Provided the cultivars perform well agronomically and in terms of quality this appears to be a 'win-win"solution. Chemicals Although farmers in some countries are reducing chemical use, pesticides are still the first line of defense against economic losses from pests and diseases, especially in irrigated and high-rainfall areas where intensive cropping is practiced, for example on cotton, sugar beet, vegetables, and fruit in WANA. On rainfed cereals, pulses, and other crops in less intensive areas, chernicals are used mainly to control weeds and certain insect pests, especially those which transmit viruses to growing crops or attack stored crops after harvest. In World Bank projects, descriptions of chemical use are often vague. In three projects reviewed, provisions are made for spraying fruit or other trees and grapes with insecticides or fungicides; six such compounds are listed in one rural development project, presumably for potatoes, sugar beet, vegetables, and fruit. In the same project, two herbicides are mentioned, but costs are calculated only for herbicide treatment of wheat. In a second case, a list of six crops to be sprayed is given, and in another fann models show that all seven crops included in the models are to be sprayed, and an unrelated list of available chemicals is cited. Which are to be used on which crop and for what purpose cannot be determined from 64 the tables, and it seems improbable that every crop on a farm would need to be treated. The 'state of the art"in the region is described in the sections below. Insecticides. Insecticides are used widely and often heavily on irrigated crops but little on rainfed crops other than olives and fruit trees. A recent development is the use of Carbofuran to control larvae of the Hessian fly in rainfed wheat, where over half of the plants may be infested in some years. Application of the insecticide alone can raise yields by over 50 percent, but experience in Morocco shows that the effects of fly control can be further enhanced and yields more than doubled by the combination of a resistant variety, Carbofuran insecticide applied to the soil, and weed control by herbicides. Carbofuran interacts synergistically with the herbicides (2,4-D or chlorsulfuron) to raise yields of wheat dry weight and grain above those from use of the insecticide or the herbicide alone, and greatly above the zero control (annex table A23). These treatments are still in the experimental or on-farm trial stage. One problem is that Carbofuran, although effective, is quite expensive and requires careful handling and safety precautions by farmers. Consequently efforts are being made to find safer alternatives. Chemical treatment with Carbofuran (spray), and Promet (seed treatment) has also proved effective against Sitona weevil, which feeds on the root nodules formed in nitrogen fixation and significantly reduces legume yields. For lentils, Carbofuran (equivalent to 20 kg/ha of 5 percent Furadan) gave best results when combined with weed control (Cyanazine and Pronamide) (0.5 and 50mlIha) pre-emergence, early planting in November rather than January, and with phosphate fertilizer. Fungicides. Fungicides are economically less viable in WANA, where rainfed yields are far lower and rainfall sparse and highly variable, than in other areas. It is also noteworthy that farmers in ecologically analogous regions of Australia and North America do not use fungicides on rainfed cereals. Seed treatment. Treating seeds with fungicides or insecticides (or both) to control seed or soilbome diseases or pests of crops is one of the most cost-effective approaches to crop protection, as well as one of the safest both from the point of view of the crop and the user. This method of crop protection is widely used for protecting cereals against damaging diseases such as smut and bunt. At the recommended levels of use these chemicals are extremely cheap, do not require elaborate equipment. Although seed treatment is probably most efficient when done by large scale seed suppliers, it can be done at the farm level in a rotating drum made by the farmer or a local artisan, or even in more primitive vessels. Modern seed treatment formulations are used in small concentrations and have low soil pollutant or residue effect compared with the chemical sprays applied to fruit trees or growing crops such as cotton. The only serious hazard is if treated grain is accidentally eaten by humans or fed to livestock. This may be the main reason subsistence producers do not use seed treatments. Rates of adoption on cereals, which are very high in many other countries, are extremely variable in WANA especially with barley and in drier zones (Oram and others 1994, Erkan and others 1990). Herbicides. A wide range of chemicals and formulations have been developed for control of broad-leaved and grassy weeds, for use in growing crops and pre-emergence of crops. Certainly these are the most widely used crop protection measures in WANA on cereals (both rainfed and irrigated). Herbicide use is also increasing on grain legumes. This is not surprising as yield losses of 30 to 75 percent have been reported in WANA (herbicidal formulations and treatment levels that have proved effective in field trials or farm practice with the main rainfed crops in WANA are listed in annex table A23. Because of the botanical characteristics of cereals (monocotyledons), the control of most weeds (dicotyledons), is a relatively easy. It is more complex in legumes and most other annual or perennial crops (which are also dicotyledons). Although considerable progress has been made in this respect, herbicides are still not widely 65 used by farmers to protect broad-leaved crops against weed competition. In the case of food legumes, for which weed control in early growth is extremely important, pre-sowing or pre-emergence treatments can give good results, although in some cases post-emergence treatment, hand weeding, or mechanical cultivation may still be needed during the growing season to control resurgent or resistant weeds, especially grasses. Box 5.2. AdoptionFactors in Herbicide Use Adoption varies widely by ecozone and farm size, from zero to alnost 100 percent. It seems to average about 40 percent of those farmers who weed their crops. Virtually all of the remainder weed manually, *' Adoption is highest in the more favorable rainfall zones, and with irrigation.: * - Rates of adoption on barley are generally lower than those on wheati even when (as in Tunisia, yields were increased by over 450 kg ha (90 percent)). * : Rates of herbicide adoption are often lower on small farms, where family resources are more adequate for labor-intensive hand weeding, than on larger holdings. * A considerable proportion of the farmers do not weed their rainfed crops at all, up to 75 - percent in low rainfall zones of Jordan, Morocco, and Tunisia, because the weeds are used for forage Source. Authors. Very little information exists concerning levels of on-farm use of herbicides on legumes in WANA; however, one report from Algeria states that only 30 percent of the chickpea and 20 percent of the lentil area are currently treated there, even though almost all farmers cite weeds as their most severe production constraint (Haddad and others 1989). 'While suitable formulations giving good control of weeds without damage to the crop are now commercially available for cereals, most food legumes and sunflower, constraints to further use are: * Lack of familiarity. In a large farmer survey in Morocco only about 10 percent reported that they did not know about them, but 40 percent to 60 percent said that they were not useful or traditionally not used (Oram and others 1994). These perceptions may stem from inadequate understanding of the correct chemicals, rates, timing, and methods of application, which are critical to successful use, as well as to the avoidance of damage to the crop and/or the user. The proliferation of available herbicides, the emergence of herbicide-resistant weed species, and the expanding use of herbicides on crops other than cereals, has compounded the task of advisory services which lack adequately trained subject matter specialists in weed management. * Lack of modem application equipment. Modem herbicides are generally most economical at low volumes that cannot easily be applied using old-style knapsack sprayers. To apply herbicide on a field scale with such primitive equipment involves transporting a lot of water and is very laborious. Modem power-operated backpacks are more suitable, but also much more costly. Contract spraying is not widely reported, although farmers use contractors for cultivation and harvest operations. * Alternative uses of weeds. The use of weeds as livestock feed is often noted as the primary reason for farmer resistance to the use of herbicides (as well as to adopting early tillage to 66 tillage to control weeds and conserve moisture). This is most important in barley-based systems in low rainfall ecozones where farmers may value forage for livestock more than an uncertain grain yield. * Costs. Additional cash costs are required for herbicides and their application, which may not be required where family labor is adequate for hand weeding. This creates perceptions that herbicide use is expensive, although economic data from Tunisia (Khaldi 1992, Byerlee and Winkehnann 1980), Turkey (Mann 1980, Byerlee and Winkelnann 1980), Syria (Pala and others 1987), and Morocco shows net returns to investment in herbicide use (if correctly applied) on wheat from 55 percent to 400 percent. On small farms with adequate family labor lacking alternative opportunities, however, hand weeding is widely preferred, herbicide use and economic viability increases as labor availability decreases essentially with farm size. (annex table A24). It is important that these constraints be borne in mind when recommending herbicide use in Bank projects. A new and potentially important use of herbicides8 on food and forage legumes in WANA is for control of the extremely damaging parasitic weed Orobanche. So far the main focus has been on the use of a contact herbicide, glyphosate, commonly known as "Roundup" (but apparently also as "Lancer" in North Africa). This is widely used for "burn off" weed control in North America. So far results from its use in WANA have been mixed, with good results in on-station and on-farm research but poor adoption rates. the latter is probably because of lack of familiarity with the correct application method and time and (until recently) the high cost of the herbicide. In 1991 the price to 2 farmers in North America was cut significantly, presumably this will have led subsequently to price reduction for roundup in other regions. Weed control was a major component of only one Bank-funded project, the First Argentina Credit, where infestation of arable land by Johnson and Bermuda grasses was to be dealt with by a combination of fallowing, cultivation, and a mixture of unspecified herbicides at an average cost (in 1990 dollars) of $124/ha or $20,000 per farm. 244D, "Iloxon" and "Grammoxone (MCPA)" for post- emergence weed control are referred to in four other projects without specifying the nature of the problem. The use of "Suffix" for wild oat (Avenafatua) control is referred to in two projects. Apart from these relatively specific references "weed control" is mentioned as part of the package in another four projects, but only in vague terms; in some others sprayers are mentioned but not linked to any identified end use, or provision is made in the loan but nothing is actually specified. Given the range of problems discussed above greater precision seems desirable in identifying the crops on which herbicides are to be used, the nature of the weed flow, the products and formulations being recommended and the method and time application. 8 In addition to herbicide treatments, other methods of protecting crops against orobanche are being sought as part of an integrated pest management approach. These include (a) varietal resistance, for example with the ELC32798 "Ghab 12" chickpea cultivar, (b) nitrogen fertilizer applications to inhibit Orobanche development:, (c) the introduction of " break" crops into legume rotations to reduce the seedbank of the parasite, (d) biological control by the agromizzid fly Phytonya orobanchia, and (e) delayed planting of lentils and faba beans by about three weeks, combined with post-emergence application of Imazaquin8 and the use of a resistant cultivar IL 8 in the case of lentils. This latter combination of treatments gave good Orobanche control, even in heavily infested soils, and raised crop yields. 67 Biological Pest and Disease Control Considerable progress is being made in the area of biological pest and disease control, which has the merit of requiring relatively little expenditure on the part of the farmers. Biological control uses natural enemies to destroy and keep pest populations down below the threshold level of serious damage. Examples of successful use of this technique on major rainfed crops important to WANA are given below. Wheat. "Take-all" fungus (Ophiobolus/graminis) causes serious losses ranging from 25 percent to 75 percent world-wide, if not controlled by means of rotations and herbicides. Continuous wheat production is being practiced more widely in WANA and in a number of other Mediterranean analogs including Australia. Altemative methods of control are being developed through bacteria and fungi hostile to the pathogen which are either coated onto the wheat seed or drilled into the soil with the seed. Barley/Wheat. The Russian wheat aphid is a serious pest of cool-season cereals, which is proving vulnerable to parasitism. ICARDA is testing aphid parasites for release in WANA. Fruit trees and olives. Mediterranean fruit fly and olive fly are two major horticultural pests for which biological controls including sterile male releases, have been valuable in some WANA countries, for example Cyprus. The degree of success so far has depended considerably on the extent to which re- infestation can be prevented from other areas outside the main treated area. Control of storage insects . Control of chickpea-seed insects by traditional methods of seed protection (versus insecticides), shows that treating seed with 5 ml olive oil plus 20 g salt per kg seed gave long-lasting treatment virtually as effective as that from insecticides.(ICARDA 1991c) So far no hazards to human health, domestic animals, wildlife, or other crops have been observed following the use of biological controls. The quality of the crop being treated is not affected and rnay be improved. Most of the methods reported can be applied relatively easily and at small cost by individual farmers or--perhaps even more effectively, by communities. When no direct action by the beneficiaries is required, cost recovery is difficult, although governments may conclude that the overall social benefits and relatively low costs make free treatment worthwhile. Integrated Pest Management Integrated Pest Management (IPM) is a relatively new ecologically-based approach that draws on and integrates a range of possibilities for pest control. There is no set prescription, but the method rests on the concept of an economic threshold, normally determined by the size of the pest population, below which the cost of control measures exceeds the expected value of the losses from pests. At farm level the decision to apply control measures depends on the nature and virulence of the pest attack, and the estimated damage, the control options open to the farmer, his knowledge of how to use them, and their probable effect, and finally on the farmer's objectives. IPM is conceptually attractive, but has its limitations.(World Bank 1994a) No economic data related to IPM could be found for WANA, but economic research in the Philippines on the value of applying economic threshold treatments shows that benefits vary with the pest population, and with the use of resistant varieties. Where farmers are not using resistant varieties, returns 68 to spraying are higher than from natural control. With resistant varieties and low pest populations, natural control becomes more attractive than prophylactic chemical treatment, but its returns compared to economic threshold depend on labor costs. The best approach to successful natural control is for farmers to avoid prophylactic sprays and to use resistant varieties. In some Asian countries IPM and other pesticide control measures are being enforced by government regulations. If this approach is to succeed, subsidies on pesticides should be removed, and measures introduced to regulate the import and sale of pesticides. Threshold levels need to be worked out carefully if the goal of reducing pesticide use is to be achieved. Box 5.3. The Limiations of IPM 1PM ....... *f: f :T:: g: There are no fixed recipes for lPMW the approach best suited to a given ecological zone- neeI ;to tbe worked outon the basis oflocation specific interdisciplinaty research. .. ...... . * 0 ;It is of less value 0in areas;where only a few 0farmers 0practice 0it, ibecause of pest inifestation; * i:; ;It requires; judgment on thle 0part dof the farmeri to decidef when to applyj controls,- andf which5 tol g;~~~~~~~~~~~~~~~~~~ s.; . ...... . .. . . g . ..--... . . . .ij;;-- ; ;0 ;; t;;:: :;Q-Qg 0fg ;- Most exenion workers have lmited training and capacity to advise lfarmers.S fff--4' ;t4 l It is more effecive. m areas. where w eaher- nd radio-lied pest eamouitorinp ndetat :;$:3fi:0:4000warningl$ systems ;:exist, iigso: as $0tot advise extension tworkerAs -aLnd :farmers }w.hen cownditios pri spose to pt acks .hat are likely to exc ed.eshold levels.; . .- * :::: :EE Precise::timing of control measures is:critical. V : E: : ::: :0i:: i:::;E:i ::i:;:E::: :s--:ES- -:: *:s Some ibiological control measures :have a delayed action effect. i0Farmers may fall bac;k :on t:: 0000fchemical sprays if teyfeel the measures have not worked.:: 00$ 0::;ti::20::l:0:40:000i::0:i4;:i00;;0: *:: ; ;It !is morfe successful in controlling insects (orlarger pests) than plant diseases. lft00-4 i :Sour*e: Autho'rs.0i0 E; !0t0tl;;:;0 t g0 ii:i i;'0' .; i tfi: 00 i' : t'?' V;L It is difficult to identify' a logical and coherent World Bank policy on pesticide use from the projects reviewed. In part this is because of vague specifications concerning the proposed nature and use of pesticides in most cases where they are mentioned. Perhaps this is because of a perception that in predominantly rainfed farming systems in WANA or northern China, with a limited range of relatively low- value crops, the need for pesticides may be lower than in intensively cultivated irrigated areas. There has been some improvement over the unsatisfactory situation with earlier projects where no mention was made of environmental hazards to producers or consumers from the proposed use of pesticides, even in the case of horticultural export promotion. Current attitudes expressed in appraisals range from extreme statements such as "no pesticides will be used in this project" to references to IPM, but with very little indication of 69 what this implies operationally. A positive and pro-active attitude towards crop protection is required, one that recognizes that IPM is a complex procedure that requires appraisal missions to work closely with national research and extension staff to identify and provide support for the operationally feasible and environmentally sustainable mix of cultural measures, cultivars, biological controls, and pesticides most appropriate to the principal ecological zones and farming systems within the project area. In the case of rainfed crops in WANA, special attention should be given to disease control through genetic resistance and cultural practices, to weed control through tillage practices and careful use of pre- and post-emergence herbicides, to encouraging the use of clean, treated seed, and to the control of Orobanche by IPM. FERTILIZER USE Fertilizer use in WANA has sky-rocketed over the last decades (table 5.4), although most of the increase is in nitrogen use, considerably less on phosphate, and almost none on potash. These figures reflect the predominance of cereals, the relatively low proportion of crops with high potash requirements (mainly irrigated), a low proportion of potash-deficient soils, and the low nitrogen and phosphate status of most soils in WANA. Table 5.4. Average Fertilizer Consumption in WANA, 1961-63 and 1988-90 (kilograms per hectare) Period N P K NPK 1961-1963 - -3.6 1.5 0.3 - S. 1988-1990 40.7 19.2 -25 -624 percent of total NPK 65.3 30.7 4.0 100.0 1990 Source: FAO Agrostat data tapes 1992. Other comparable regions have without exception a higher share of potash in the fertilizer mix, and-except for China, a lower share of nitrogen. The proportion of phosphate is lower than that in WANA in North America and Western Europe (where high levels of use over time have reduced current levels of response significantly) but also in South Asia, India, and China. It is considerably higher in the southern cone of South America, and strikingly so in Australia--reflecting the importance of legumes and legume- based pastures there. Average 1992 levels of fertilizer use (NPK) in west Asia were 58 kg/ha and 66 kg/a in North Africa (the latter, however being biased upwards by 383 kg/ ha in Egypt). While these averages may appear low compared with China or Western Europe, it is important to note that they are above those in other developed countries with large semi-arid areas of rainfed arable land under wheat and other crops similar to those grown in WANA, including Australia, Canada, the United States, and the southern cone region of South America. 70 Growth rates of fertilizer consumption in WANA were Table 5.5. Nitrogen-to-Wheat Price rapid during the 1960s but fell progressively each decade, as Ratios in Selected Countries, 1991-1992 the total quantitity used in the region increased to a level of 4.4 Reil N to Wheat percent a year during the 1980s, (annex table A25). This was ...... well below growth rates during that period in South and East WANA 1.1 - 3 0 Asia, China, India, and the wheat/barley zone of South ChiIIa: 2.8 America. Over the 1980s growth rates in Australia, Canada, Argentina 6.1 the United States, and Europe were, because of environmental Australia 7.0 concerns, and low grain prices, stagnant or negative. Despite USA 4.7 increased fertilizer prices by the end of the decade, nitrogen- Mexico 2.0 wheat price ratios in WANA were more favorable than in most Sorce CIMMYT comparable countries (annex tables A25 and A26). Reports from individual WANA countries on fertilizer use give conflicting signals. Empirical evidence from national research programs and international/regional research institutes (CIMMYT, ICARDA, ACSAD) show good responses of winter cereals to nitrogenous and phosphate fertilizers under both rainfed and irrigated conditions, and, except in areas below 325 mm average annual rainfall or in mountainous areas, most farmers are aware of the benefits and a majority use fertilizer, especially on wheat. In Morocco an extensive survey of 1,062 farms in 17 provinces in 1990 shows that in irrigated and more favorable rainfed zones, over 90 percent of farmers use fertilizer on bread wheat. Adoption rates on barley are around 68 percent. However, in the arid zone with annual precipitation of less than 300 mm, only 15 percent of farmers use fertilizer on barley, 26 percent on bread wheat, and 29 percent on durum wheat (Oram and others 1994). These results confirm findings (Ouassou and El Baghati 1988, ICARDA 1992a) of low use in semi-arid and mountainous areas. Factors Influencing Use of Fertilizer Low rainfall, high cost, and lack of awareness of the advantages are the main reasons given by farmers in Jordan for not using fertilizer (annex table A27), and as well as in Morocco (Oram and others 1994). Analysis across countries and surveys suggests that six main variables influence farmers' decisions to use fertilizer. Magnitude and variability of precipitation. Under most situations in WANA rainfall is the dominant influence on crop response to fertilizer, especially nitrogen. On average, at prevailing low fertilization levels of 40-50 kg/ha of nutrient in WANA, a 100 mm increase in annual rainfall increases the grain-to-nitrogen response ratio by about five (Somel, Mazid, and Hallajian 1984). Given the strength of this relationship, rainfall variability greatly augments the risk of using fertilizer in dryland production zones, especially if the rate of application and material used increases that variability (Somel 1992). The nature of the crop and the rotation. Wheat, especially bread wheat, normally grown in the higher rainfall regions, receives more fertilizer than barley, which often receives no fertilizer. Pulses, especially lentils, commonly grown lower down the rainfall spectrum than chickpeas, may go unfertilized. Chickpeas and faba beans (a crop of the more reliable rainfed and irrigated lands), usually receive some phosphate fertilizer. 71 Price relationships. Although many farmers surveyed complain about the high costs of fertilizer, and of inadequate government support in terms of fertilizer subsidies or higher support prices for cereals. Annex table A29 suggests that at least for nitrogen and wheat, the price relationships are quite favorable. Accessibility of the area. This may be simply a factor of distance from main urban areas or markets (remoteness), or of topography (inaccessibility) or both. Quite apart from problems of marketing products and purchasing inputs, remoteness and difficulty of access tend to engender conservatism with respect to adoption, which is often compounded by poorer than usual extension service support. Inadequate advice on fertilizer use. This takes several forms, including inflexible recommendations from research institutes, usually based on the results of experiment station trials in more favorable areas, lack of soil testing facilities as a basis for diagnostic recommendations, and numerically inadequate, immobile, and insufficiently trained extension staff (especially in remote areas), unable to provide locally-specific advice to farmers. Fertilizer Response: Experimental Data A wide range of fertilizer trials at experimental stations and on farms, on wheat, barley, food legumes and forage crops has been undertaken in the last decade by ICARDA and by national programs, both on individual crops (often in combination with testing promising cultivars), and on crops in rotations. On the whole, these trials indicate that fertilizer use can be economically viable even in low rainfall areas, and the case for applying it to barley in zones down to 250 mm is strong (anenx table A28). On- farm trials and demonstrations in areas with about 250 mm average annual rainfall, showed an average grain yield increase from fertilizer of 44 percent in Syria, 75 percent in Jordan, and 84 percent in Iraq (ICARDA 1992d). In another low-rainfall on-farm trial series in Jordan and Syria, grain yields were increased between 5 percent and 100 percent, with straw yield increases being almost double those of grain. These trials showed that farmers net financial returns could be increased by 50 percent (ICARDA 1990c). In Morocco, returns to nitrogen applied to wheat were calculated for 0, 40, 80, and 120 kg/ha, in two soil types, following three rotations. The results show that benefit/cost ratios calculated at the prevailing nitrogen/wheat grain price ratio of 1.85 generally decrease at higher dose rates, for wheat after barley, as well as after legumes or fallow. Thus, returns to moderate rates of nitrogen application on wheat up to 80 kg N/ha seem likely to be rewarding where it is grown in continuous rotation with another cereal, an oilseed, or melons, but appear more speculative after fallow or a legume--especially a forage legume (Abdel Monem and others 1990). Phosphate fertilizers often show profitable responses when applied to grain legumes and pastures, since phosphate deficiency is widespread in most WANA soils. Over a four-year period pastures top- dressed with 60 kg/ha superphosphate per year produced 300 percent more herbage in Syria than non- fertilized land, reducing the need to feed barley supplements to sheep by about 50 percent (ICARDA 1990c). The feed saving was greatest in the fourth year, although it was the driest. The seedbank rose from 2,000 legume seeds/m2 with 25 kg/ha phosphate to 8,000 seeds/M2 with 60 kg/ha phosphate, and nitrogen fixation capacity also improved. An adequate soil phosphate status is also important to rhizobium development and thus to enhanced nitrogen fixation. The direct response of wheat to phosphate application is harder to detect than that of nitrogen, barley is found to be more responsive, possibly because it is nornally grown in drier areas, and there may be a link between phosphate and drought-tolerance. In both cereals, however, there seems to be a marked synergism between nitrogen and phosphate application. 72 Responses of cereals and pulses to potassium in rainfed areas of WANA are generally low or negligible, hence the application of a compound fertilizer such as 14-28-14 for cereals is wasteful. No trials dealing with soil acidity problems seem to be reported, probably because most soils in WANA are of calcareous origin, and have reasonably high pH levels. Recently, boron toxicity has been identified in cereal areas of Turkey, Syria, and Egypt, causing impaired root growth, leaf yellowing, and death in severe cases. As the potential for correcting macro- nutrient deficiencies is progressively reduced through fertilizer and plant breeding in WANA countries and other developing regions with analogous ecological situations and cropping patterns, the impact of other adverse soil conditions (salinity, acidity, alkalinity, and micronutrient deficiency and toxicity) may well become more limiting. These problems appear to be poorly documented and inadequately studied at present in WANA and in northern China, and the limits they place on crop growth and yields (and especially on non-cereal crops such as cotton, sugar beets and vegetables) are not well quantified. Increasing Fertilizer Use Efficiency in World Bank Projects Fertilizer is an important component of the investment package in most of the Bank projects reviewed. Although the type and quantity of fertilizer is usually specified for the crops involved where farm models are used, the recommendations are often vague in other projects. There is almost no reference to any plant nutrient needs other than nitrogen and phosphate in the Bank projects. The range of fornulations funded in Bank projects is fairly limited, with a strong emphasis on ammonium nitrate (33.5 percent N) or urea (46 percent N) among Nitrogenous materials, and triple superphosphate (48 percent P205) among phosphate fertilizers. Potash fertilizer (potassium sulfate), is only specified in one project (Meknes). Only two compound fertilizers are widely recommended, Di- Ammonium Phosphate (36-48-0) and 14-28-14, the latter mainly in North Africa. The preference seems to be for a combination of a fairly high analysis nitrogenous formulation (ammonium nitrate or urea), with a similar type of phosphate fertilizer (triple superphosphate). Organic fertilizers (manure, oilseed cake, et cetera) are only used in two projects, both in China, possibly because there is a serious shortage of farmyard manure throughout the WANA region. No mention is made of anhydrous ammonia, sludge, or any other liquid fertilizer. The levels of application proposed in WANA projects appear adequate, and in some cases high, especially for rainfed cereals, pulses, and cereal-forage mixtures, given the high coefficients of variation of yield because of the unpredictability of rainfall there. While in many project descriptions no distinction is made between rainfed and irrigated crops, in those cases where irrigation is not a major factor the rainfed rates proposed seem more appropriate to high rainfall situations than to the semi-arid conditions prevailing in the project area, for example 250 kg/ha 14-28-14 plus 100 kg urea for rainfed cereals in Morocco, and 200 kg DAP on rangelands in Turkey. Despite the significant proportion of funds invested in fertilizer, and its potentially large contribution to increased productivity in WANA (and thus to project rates of return), most project appraisals devote little attention to strategy for its most efficient and cost-effective use. This strategy should consider the entire package of technologies, such as: The use of improved, fertilizer-responsive varieties of the key crops. It is not clear from review of the appraisal reports, and particularly the post-project evaluations, that the critical link between fertilizer response and improved variety is adequately appreciated, certainly the references to use of high- yielding varieties in most project documents are vague and unspecified: Soil factors. Most fertilizer recommendations are provided without explicit reference to the soil analysis, or to its use on previous crops and possible residual effects, although such factors are 73 critical in determining the returns to fertilizer. Thus ICARDA (1991b) stresses that calculations of the economics of fertilizer application should never be restricted to the year and crop of application. Efficiency in the method and timing offertilizer application. While more concentrated fertilizers seem to be preferred in Bank projects, because they are more economical to transport, store, and apply, they also require more precise methods of application to avoid inefficient use, because of their more concentrated nature. This should be kept in mind when recommending formulations for Bank projects. This holds especially for those in projects located in areas with many small farms, where equipment for applying fertilizers or organic manure is relatively primitive or even absent, and the materials may have to be hoed in or broadcast on the surface. Newer technologies such as liquid application of fertilizer (anhydrous ammonia 82 percent N), are not mentioned in any appraisal. although it is now commonly used for supplying nitrogen to field crops in North America and might become opportune in WANA and South America for use by contractors, or for direct delivery and storage to larger farmers. Maintenance or enhancement of soil organic matter status. When animal manure is limited, as in WANA, nutrient cycling can enhance soil organic matter. Nutrient cycling is now receiving increased attention as a means of increasing soil fertility, building organic matter, and improving soil moisture content. In many soils, total plant biomass production, on which organic residue depends, is limited by inadequate plant nutrients in the soil. Chemical fertilizers thus hold the key to the system. In the absence of fertilizer farmers may practice long fallows to mobilize soil nitrogen for cereal crops. They may also develop farming systems involving grain legumes or legume-based pastures in rotation with cereals to increase N through BNF. Research in low-rainfall ecological zones in Australia, Canada, and China has shown that a combination of a rotation (involving grain or forage legumes followed by wheat, barley, and in some cases canola or another annual brassica) with conservation tillage (direct drilling), stubble or residue retention, and phosphate application to the legume leads over time both to higher soil fertility and to the most efficient utilization of applied nitrogen (ICDF 1988). Around 50 percent of inorganic nitrogen is used in the cropping season it is applied, part is lost through volatilization and leaching and part carried over to the next season. It is added to between crops by the decomposition of organic matter in the soil or on the surface, especially in a residue management system. Nitrogen fixed by legumes is partly used to sustain their growth, and partly carried over to the next crop (probably a cereal), but it has been shown that less than 20 percent of this is taken up by the following cereal, the majority being transformed into soil organic matter nitrogen that decomposes slowly and provides nitrogen in small amounts to subsequent crops. Fertilizer nitrogen also contributes to the residual soil organic matter reserve through plant decay. Such findings leading to integrated soil management are rarely considered in Bank projects (World Bank 1994c) A number of studies from northern China also emphasize the importance of organic recycling, directly through crop residue incorporation and indirectly through chemical fertilizer and organic manure as a means of restoring or increasing soil fertility, improving water use efficiency, and raising crop productivity (ICDF 1988). Prevention of Soil Nutrient Loss. Measures to prevent loss of soil nutrients through volatilization or denitrification should form an important element of fertilizer strategy, but do not figure prominently in any of the Bank projects reviewed. Losses through volatilization and denitrification can be prevented by (a) early incorporation of manure and fertilizer into the soil, or fertilizer placement into the soil, (b) use of nitrification inhibitors to reduce oxidation of ammonia into nitrates (two commonly used compounds are nitrapyrin and etridiazol, applied at low rates of 0.5 kg/ha, although the former may be less effective at soil temperatures exceeding 250 Celsius), and use of urease enzyme inhibitors to delay hydrolysis of urea to ammonium nitrogen, (c) coating urea to reduce its solubility and rate of hydrolysis to ammonia (sulfur is 74 commonly used commercially, but experiments in India show paste made from the neern tree to be quite effective). The second major form of nutrient loss occurs through runoff, erosion, and leaching. These are major causes of low recovery of nitrogen fertilizer by crops, both in the year of application, and by a following crop. According to the recent comprehensive study by the U.S. Committee on Long-Range Soil and Water Conservation (NRC 1993a), relatively little soluble nitrogen is lost in true runoff. Some nitrogen in the form of ammonium-N is lost along with organic nitrogen attached to soil particles and contributes to nitrogen in surface water. The majority of soluble nitrogen, nitrate, is lost in leaching into and through the soil, either as shallow sub-surface flow or by deep percolation through to groundwater and eventually into surface flows. This research shows that farmers use fertilizers inefficiently, because they do not take account of all sources of N from manure, BNF and other crop residuals. Such "overkill"can greatly increase nitrogen losses, as well as wasting the farmers' money. Very little work along these lines is in progress in WANA. Nitrogen studies mainly concem the residual effects of legumes, sometimes with conflicting results, and of phosphorous. Given the relatively unsophisticated state of on-farm fertilizer management in WANA, it seems optimistic to foresee widespread use of nitrification or urease inhibitors. Soil testing services are reported to be limited in scope in most countries, and where available are often not used, or their recommendations not followed closely by farmers. Under these circumstances, the main effort should be directed to improving fertilizer application practices on farms, along the lines described above, and to educating farmers to take advantage of the benefits of biological nitrogen fixation and to consider the contributions of all sources of nitrogen in determining their use of fertilizers. In recommending levels of fertilizer use in areas with low and highly variable rainfall it is important to avoid advising farmers to apply rates that are economically but not biologically optimum, yielding incremental returns at the border-line of profitability just before the response curve flattens. Given the unpredictability of yields and the gaps between experimental and ordinary farm practice, such advice is likely to cause economic losses in many years. Conclusions Despite the relatively high rates of adoption and application of fertilizers reported from some well- watered regions of WANA, national NPK/ha averages and farm surveys show clearly that little or no fertilizer is being used by many farmers. Although the fertilizer formulations and recommended rates of application (where specified in Bank projects), appear adequate to conditions in the more favorable zones, they tend to reflect blanket research prescriptions and to provide little flexibility for sub-optimal conditions. Moreover they are not clearly specified in several projects, and even where they are it is not always obvious which crops they are intended for, or whether the fertilizer is intended for a basal, a top dressing, or a split application. The use of concentrated materials (urea, DAP, and TSP) on pastures and rangeland is questionable. Single superphosphate is quite adequate, less costly, and may even be more effective. No work on the use of rock phosphate seems to be reported, possibly because phosphate is mined in several countries and may be cheap, or because rainfall is too low. The possibility of using lower-grade material on pastures should nevertheless be explored. More attention needs to be directed in monitoring future projects to the nature and conditions of the farmers who use fertilizer and those who do not. The ecological, social, economic, or institutional factors constraining farmers from using fertilizer efficiently--or using it at all--must be studied more carefully and alleviated where possible if the current low rates of adoption and slow growth of yields are to be improved. With respect to increasing fertilizer use efficiency and reducing avoidable nutrient losses, the absence, high cost, or inadequacy of machinery for its application, and particularly for precision placement may well be an important constraint. This issue is not dealt with satisfactorily in SARs, which rarely 75 specify seed and fertilizer application equipment in proposals for mechanization. Future appraisals should pay special attention to this component of farm mechanization. The relatively low levels of adoption of high-yielding (and presumably fertilizer-responsive) cereal cultivars in WANA seem likely to be a constraint to profitable use of fertilizer. Attention needs to be directed to the reasons for this, especially in Algeria, Iran, Jordan, and Turkey, where adoption rates for wheat are under 40 percent, and in low-rainfall and mountainous zones in all countries. Fertilizer application should not be treated in isolation, but as part as a system that includes cereals, legumes in crop rotations for BNF, reduced tillage, weed control, and residue management, and soil and water conservation. To the extent that organic manure is available this should form an integral part of the package. Machinery specifications for tillage and residue management, weed control, seed and fertilizer application should be geared to the furtherance of this system. Special attention should be directed to practices that reduce excessive use of fertilizer and minimize avoidable losses through volatilization and leaching. National research priorities should be aligned to testing the concepts of nutrient cycling and the enhancement of soil organic matter status through experiment station and on-farm trials as a means cf building sustainable production suited to prevailing ecological and social situations. Extension specialists should receive special training in the implementation of the findings and how to demonstrate them to farmers most effectively at the farm level. Improved private fertilizer distribution networks should be encouraged. Poor access to fertilizer, especially by small farmers, is a reason for low use of fertilizer in several countries. Extension staff should work closely with distributors to advise farmers in correct use of fertilizer. While research on crop response to fertilizer often shows that yields can be increased even in low rainfall areas and with local varieties, the results of economic analysis are more conflicting. Because of the ambiguity of the data, it has been suggested by (Belaid and Morris 1991) that efforts to promote fertilizer use should concentrate on educating farmers about the complex nature of fertilizer response functions and letting them decide for themselves what practices to adopt. This implies that a new and different approach will be needed to replace traditional extension strategies that focused on making blanket recommendations to all farmers. MECHANIZATION Although there were no Bank-funded mechanization-only projects included in this study, virtually all of the projects reviewed had mechanization components, and in several cases they were the largest single investment item. Six involved destoning in conjunction with hand labor. Twelve projects used machinery to construct soil conservation works (banquettes, various types of terraces), three included mechanized land clearance, fourteen supported mechanization as a means of reducing fallow and increasing crop intensity, and several viewed it as an aid to increasing crop productivity through better soil management, timely sowing, and reduction of harvest losses. While project descriptions often mention inproved cultivation techniques, conservation tillage was only referred to overtly in two cases (the Argentina agricultural credit, and the eastern Anatolia watershed project in Turkey). Even in those cases the defunitions are ambiguous. In all other projects, it was not possible to identify the specific improved soil management practices, or the required equipment. Thus, these objectives sound more like an expression of hope than reality. There are important differences in the emphasis on mechanization between projects that deal primarily with area development (watershed management, rural development, and agricultural development), and those that are credit-oriented. While the majority of the former have important 76 components of soil and water conservation works involving mechanized methods, as well as some on-farm mechanization components. The credit projects concentrate primarily on the latter. However there is often overlapping, for example in Morocco and Turkey, where some on-farm mechanization components of agricultural development projects are financed from the credit projects in those countries. On-farm mechanization is easier to track theoretically in projects where the investments are related to farm models, which specify the items to be financed in each model, for example the Morocco Fifth Agricultural Credit. In practice it is difficult to ascertain whether the actual disposition of farm equipment purchased under this project by farmers or cooperatives conformed with those models, of which only four out of the eleven postulated in the SAR actually received machinery or implements other than small-scale irrigation equipment. This project relied to a considerable extent on contractual mechanization services to provide support (mainly for land preparation) to the many farmers relying on hand labor or animal traction for their other farm operations. It is not clear whether farmers not conforming to such models (which, however well-specified, generally represent only a small cross-section of the farms in a project area), can get credit for their machinery requirements or for machinery hire, or how the machinery and implements proposed in appraisal reports eventually get distributed. Table 5.6. Nurmber of Tractors and Combine Harvesters in Relation to Cultivated Areas in Selected Regions and Countries, 1989 ..;(t .... ..f ... iutVated Tractors Ha cultivated Comb'nes Ha cereals .-'.fg.0-..X ..,--T,.,. .. . . . . ..,, . .* .:''L '0 .. ..0 .f .:- ......L ', .. .. ...:. Cotmtry crops Numbers per~~~~ crultora harvesters p ,erea:sbin t-,Nt-,',k,,,,,J' (,,''.,' 00 0, ha) Total ____ China 6115 ... ~~~~~~~~~~~~~~~~~~~~~~~61220 36582 2531~~~~~~..... . ..... . USAt00- i0(0-00 -00i---00 000;-189915 ; 46700t; 002-- 00000 0 041 .640000000 .. ........ Source: FA0 1990. Box. 5~... .4 . Pcae ....enei .an ...ects Sapldscription of mehnzto ' fro ooc n Tunisi ;";-'include'''"-'''--"jSX the 'follow''ing:- SSaal-farm (1# d a) pac.Ia00e baled0 onaimalrapoweTnonst * . fed~umsize (greter than 0 ha ) fam packagesinclude tactors of40'50 HP presumabl drils and00 trullers Thris amutst US$21to,000 wit abie and |0 U$5 0 exludngth N:~. aler. Weed ct eq4ipment.....in.....T Larg... r-farm. (greater. ..a 2(. . u pk.e may. . n.d rcor n te eupeta ...... ...o e lsaco iieh retr(4m trct b...er..... to. .ot ....... d.... th....... tra.t. an....... . qipen.sow ude ..~ meiu pa..epute...h..adbae amounts........ .... ... .n.983.rices The cobn cots...20 &rnrce: Authors.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. ... . 77 Machinery Specification in Project Documents Except in projects where specialized equipment is required for land clearing, destoning, or land- leveling and where relatively small numbers of machines are involved, the specifications in appraisal reports tend to be poorly defined. For example, the size (HP) of the tractors is only indicated in one or two cases. Cultivating and tillage equipment is often described vaguely or ambiguously--particularly where minimum and reduced tillage (MART) techniques are implied through statements such as "to facilitate early and better plowing and seeding, soil moisture infiltration, and sowing through mechanization of pre-planting tillage practices. In most cases tillage implements are not defined in any identifiable way, but where they are defined, specifications suggest equipment already obsolete or with demonstrated adverse effects such as creating hard pans just below their operational depth, for example by rototillers. Deep ploughing was advocated in a few cases, although it raises costs, and experimental evidence does not support arguments that it increases moisture conservation (ICARDA 1991b). With the exception of one or two elements of the Xinjiang Project in China such as the laser-guided self-propelled scrapers for land leveling, most of the agricultural equipment described in the Bank projects is conservative. Even the latest projects do not reflect recent trends or are too unspecific in their references to innovative practices such as MART. Major Gaps in Bank Provision for Farm Mechanization There are significant gaps and deficiencies in Bank provision for mechanization in the projects reviewed. The following are prominent. Seeding and planting equipment. Only four projects refer directly to planting equipment. Several other projects refer to the provision of tractors and ancillary but unspecified equipment. Even where direct reference is made to seeding equipment, its nature and purpose is not clear, nor is it stated whether it is also capable of applying fertilizer at the same time as the seed. While there are situations where farms are so small or fields so steep that the use of a seed drill is impracticable, efficient seed or seed/fertilizer drills should be a preferred part of equipment packages based on tractors. Small five-row drills designed for animal draft power have also been developed, for example in Morocco, and could be considered for smaller farms, provided costs are acceptable. The machine developed in Morocco costs US$750 (manufactured locally). Including undepreciated costs, this ranges from US$28 each year for farms of 5 ha to US$7/ha annually for farms of 20 ha, assuming an eight-year working life for the machine (Bansal and El Gharras 1993). The importance of efficient modem seeding equipment relates both to its versatility, and to the need to economize on seed and fertilizer by placing it accurately both at optimum lateral spacing between rows, longitudinal spacing within the row, and vertical spacing in the soil to take advantage of soil moisture and prevent exposure to sun and wind. Empirical comparisons of drilled versus broadcast seed in semi-arid environments show a clear advantage for the former. Seed rates, especially for wheat and barley, specified in projects are often excessively high, and could be reduced by more precise placement. In Morocco seed rates for wheat are around 180 kg/ha when broadcast, but only 120 kg/ha is needed when drilled. Precision drilling eliminates the need for the extra pass with a disc harrow to cover the seed, which is necessary when it is broadcast, and also facilitates subsequent inter-row cultivation and weeding. Average yields in Morocco from on-farm wheat experiments were higher with a seed drill than with the farmers' methods of broadcasting seed followed by harrowing, and net returns per hectare were also higher (annex table A29). 78 The calculated break-even point for generating enough revenue to cover loan payments is 2.4 ha sown per annum at the yields achieved. Fertilizer application techniques. Fertilizer applicators in Bank projects are sometimes described as "distributors" and sometimes as "spreaders". It is rarely possible to identify whether the fertilizer is to be applied as part of the seeding operation (for example via an attachment to the seed drill), through some separate operation with a relatively precise placement machine, or via a spinner or some other surface applicator. No mention is made in any project of anhydrous ammonia applied in liquid form as an alternative to the use of powdered or granular forms of nitrogen fertilizer. This is now relatively common practice in North America, often via contractors. Crop spraying requirements. Apart from references to knapsack sprayers for small farms in a couple of credit projects the subject of crop spraying is very poorly handled, both in terms of the objectives, the nature of the equipment, and the materials to be applied. The importance of having reliable and versatile spraying equipment has increased with the expansion of MART techniques, especially no-till. Along with efficient means of placing seed and fertilizer through residues, weed control as a substitute for tillage is the most critical component of conservation tillage. Effective control of perennial weeds, especially certain grasses with Glyphosate (Roundup) or another total herbicide, prior to or immediately after seeding, is crucial to the success of the technology. Subsequent control of broad-leaved weeds during the growing season of the crop through the use of 2-4D, MCPA or other selective herbicides is also important. In Westem Europe and North America, crop sprayers, like seeders, have become extremely sophisticated in recent years, with electronic sensors linked to each nozzle on the spraybar to provide instant warning of any blockage or other malfunction, and even a weed sensing capacity. Very small quantities of active ingredient can be applied with great precision, and with minimum risk of drift on to sensitive crops. Such machines are only appropriate to larger farms, and to situations where servicing is reliable and efficient. Less complex, relatively cheap, low volume sprayers have been available for many years, however, with variable spraybars suited to different farm sizes and crops. These are not difficult to calibrate or maintain and can be mounted on tractors equipped with power takeoffs from 35 UP upwards. This type of sprayer is quite well suited to custom operations where average farm size is small, as it can be operated in a restricted space when tractor mounted. The main constraint in such circumstances is likely to be the proximity of herbicide-sensitive crops to the area being sprayed. As an alternative to the relatively cheap but laborious knapsacks, portable hand-carried C02-powered sprayers are in use in Asian countries, although only knapsacks are referred to in WANA project appraisals. Tractors. A World Bank study on agricultural mechanization (Binswanger and Donovan 1987) casts doubt on the rationale for supplying tractors as a means of achieving yield gains. In WANA, however, there is a strong body of evidence from on-farm as well as experiment station trials that increasing use of tractors for tillage does lead to yield increases by facilitating early sowing, as well as increasing the area cultivated. (Oram 1956, Cleaver 1982, ICARDA 1991b and 1992a, Bansal and El Gharras 1993). 79 Mechanization and WANA Bank projects in WANA base their reasons for investing in mechanization primarily on those benefits that Binswanger and Donovan (1987) either discount (timely cultivation, better soil management, and increasing crop intensities) or view as potentially socially undesirable (mechanical harvesting, saving labor). Are those projects thus wrongly oriented, or are there mnitigating circumstances peculiar to WANA that justify mechanization? Timely cultivation and better soil management. The critical period in rainfed farms in WANA is the period immediately prior to and after the beginning of the single rainy season. There is abundant and incontrovertible evidence tllat the longer seeding is delayed after the onset of the winter rains, the lower are the yields of the key cereals, and also of grain and forage legumes. To the extent possible farmers like to prepare land in advance of the first rains, but this is difficult with animal draft, because the soil tends to be hard and the animals are often undernourished at that time of year. With hand labor it is virtually out of the question. The WANA region has the most variable rainfall of any developing region, and it is not uncommon for a crop to have to be resown if there is a serious drought after the first seeding. Time is even more vital under those circumstances. Cropping intensity. With respect to cropping intensity, the benefit of mechanization in WANA is related to the substitution of annual fallow by a cool-season crop (usually a grain legume, oilseed, or forage crop). This raises land use intensity from around 60 percent to 100 percent in any one year. Some of these crops are fall-sown and some are sown in the early spring, but the impact is to increase the total area that has to be cultivated, sown, weeded, harvested, and transported every year. Control-intensive operations. According to Binswanger and Donovan (1987), operations such as seeding and applying fertilizers and pesticides can better be performed by machine. It is clear from numerous experiments in WANA countries, that this leads to greater efficiency in the use of the input and often to yield improvement. Box 5.5. Mechanization Policy in WANA The five main reasons advanced for World Bank support to on-farm mechanization in the projects reviewed are: * More timely and better cultivation, facilitate water conservation, good seedbed preparation and early planting, to take advantage of the full rainy season Reduced tillage and residue management to reduce runoff, improve infiltration, conserve organic matter, and control weeds Elimination of fallow and increasing crop intensities * - Reduction of harvest losses * Saving labor in the face of rising wages. Source- Authors. 80 Conservation tillage. In WANA and in northern China, this technique has been shown by experimental station and on-farm trials to increase yields through improving soil moisture storage and controlling weeds. It also combats soil erosion (Chekli 1991, Kacemi 1992, and Siming and others 1994). To be adopted by farmers, appropriate tillage implements, seed drills, and sprayers must be provided, possibly through Bank projects. Combine harvesting. The argument for combine harvesting seems more open to question, because it is capital-intensive and because of the risk of labor displacement. Nevertheless, up to 20 percent of grain from cereals and more from food legumes is lost because of shattering, damage from pests, weather, and encroaching livestock when harvesting is delayed by lack of labor. Considerable losses of leaf from forage legumes are also reported because of delays in harvesting. Because of their short straw, especially in dry years, however, using combines to harvest legumes is difficult and often damaging (Oram and Belaid 1990). Because of these a conflicting views, a review of the technical altematives and their impacts on yield, and of the socio-economic issues--including the availability and opportunity costs of labor ought to be undertaken to guide Bank policy on financing mechanical harvesting equipment. If casual labor is costly or hard to get, why have some project appraisals not included the costs of family labor in their rates of return calculations? Also, it may be important to look at off-farm employment of farm families. If a genuine case can be made for farm mechanization not reducing employment opportunities this would be a significant argument in its favor. Even so, it is difficult to understand the rationale of the Fifth Agricultural Credit project in Morocco for recommending a combine on a 200 hectare "fully-mechanized" rainfed cereal/pulse-based farm in a 300 mm rainfall zone of Morocco that has nine employees and 80 hectares of fallow, while retaining the same labor force afterwards and not intensifying the cropping pattern. In a similar rainfall area of Canada or Australia a 1,000 ha farm with a comparable cropping system and a cattle fattening enterprise would probably only have two to three employees. Bank Supportfor Investment in Machinery Contracting An important question is whether farmers wishing to undertake custom operations for their neighbors should be encouraged to do so and could then get financing for those requirements, especially if their own farm operations did not conform to any particular farm model predicted in an appraisal report. In all three Maghreb countries of North Africa and in Syria there is a strong element of mechanized contracting, especially for land preparation and harvesting, but increasingly also for baling straw because of its high current value for animal feed. The Fifth Moroccan Agricultural Credit project is ambivalent on the contracting issue, since it argues on the one hand that large farmers should purchase combines and balers to reduce their dependence on contracting, but then asserts that they can help to amortize the cost of those machines by undertaking custom operations. Contractual machinery services run by private operators have played a valuable role since World War II in mechanizing key operations in North America, Western Europe, and Mediterranean Europe. They have declined significantly in importance since the 1970s in North America and Western Europe, while remaining important in Spain and Italy, and increasing in North Africa, Syria, and Turkey. Farm size is obviously an important factor determining the optimum level and type of owner or user of mechanization, and in this context it is interesting that Egypt, a country of predominantly small irrigated farms, now counts one tractor to 49 ha of cultivated land, and Turkey counts one tractor to 41 hectares, about the samne as in the United States (annex table A30). Only three WANA countries have at least one combine to under 400 ha of cereals: Cyprus, Algeria and Tunisia. With the exceptions of Cyprus and Algeria, the most heavily mechanized countries in terms of tractors do not have the most combines in 81 relation to their cropped area. On small farms under 10 hectares, are most of the private holdings in WANA countries (and in China and South Asia), farmers use contract operators primarily for land preparation and rely more on family labor and casual hired help for haymaking and harvesting, with animal draft power being used for cultivation and transportation. Full-time contractors, as opposed to local landowners undertaking custom operations for neighbors, seem to be most successful in countries with a climatic gradient that enables them to operate over a long season, for example in Spain, where they can commence combining winter cereals in Andalusia in June and work their way north as the harvesting season commences progressively later. This situation can apply with altitude as well as latitude. In order to be more efficient and cost-effective, and in order to avoid excessive stress on machinery and operators, contractors try and avoid working on small farms with scattered plots, especially on hilly land. Given that in WANA and in other semi-arid sub-tropical and tropical regions, such holdings represent the majority in numerical terms (if not in terms of the area cultivated), Bank missions face difficult decisions in recommending investments in farm mechanization. On the whole (Tunisia and Morocco being an exception), attempts to bridge the gap between private ownership of on-farm machinery and private commercial contractors by the provision of government machinery services or machinery cooperatives have not succeeded when it comes to meeting the needs of small- to medium-sized farms in the 15 ha to 40 ha bracket. Public mechanization schemes have some of the worst records of any government enterprises in developing countries. Under these circumstances, credit should be made available for rental purposes, subject to supervision of contractors' charges, the levels of training of their machinery operators and their maintenance facilities. Environmental Effects of Mechanization The Bank projects reviewed pay little attention to the potentially adverse effects of inappropriate mechanization. The major problems of inappropriate mechanization in WANA are tractor cultivation up and down slopes rather than on the contour (leading to massive water erosion in hilly areas and related downstream externalities through siltation of dams and irrigation canals), and cultivating fallows base after crops have been removed and stubble grazed down by sheep, leading to wind erosion. These problems can be dealt with if proper provision is made for surveys, related to soil type and slope, so that the provision of credit can be tied to the introduction of appropriate cultivation techniques such as conservation tillage, contouring, or terracing. Provision for surveys (unless already available) and for monitoring and evaluation of these trends should be built in to the projects. In hardly any projects are these particular needs provided for at present. In some countries inheritance laws are a reason for long, narrow, fields running up and down slopes. Unless farmers on the hillsides can be persuaded to act collectively, this can be a major impedimnent to soil conservation. Optimum Balance Between Mechanization and Animal Draft The reports reviewed and the survey data on farm size show that animal power is still used quite extensively in the WANA region by small farmers, particularly for light cultivation and for transportation. With snall mechanized equipment unavailable or too expensive, there is little choice for small farmers in WANA between animal traction or the purchase of a tractor, other than combining rental services with the use of their own animals. Thus, in determining how best to support mechanization in future Bank projects, the crucial requirement is to strike an optimum balance between owner-operation of farm machinery and contractual 82 operations, and, in relation to ownership, between mechanically powered and animal draft equipment. This requires a careful assessment of the suitability and cost of animal-drawn equipment, especially that required for control operations, since in Morocco, India, and Ethiopia experience shows that there are both design and cost problems in developing toolbars and seed/fertilizer placement machines that are efficient, light enough for animal draft, yet robust and within the purchasing power of small farmers. Such implements are usually of limited interest to commercial manufacturers of larger machines, while universities and public agricultural research institutions generally have a poor record in machinery design. The International Agricultural Research Centers have not done much better. Hence there is an R & D gap that applies both to the design and manufacture of certain equipment and to the mechanization of some important crops such as food legumes, com and fodders. In addition, virtually all such equipment is imported, and experience with local manufacture has been frustrating (annex table A3 1). The right balance of equipment and ownership will clearly vary with farm size and layout- including topography, soils, farming systems, and the proportion of land irrigated. It will also vary with labor availability and wage-rates and with the availability, cost, and servicing facilities for the appropriate machines. In reviewing the pros and cons of animal traction the opportunity costs of animal feed must be taken into account. Subsidies or other financial distortions (including cheap fuel) that favor certain types of machinery should be included in the calculations. Future Demand For Farm Machinery The data in annex table A30 reveal wide differences in current levels of mechanization in the region, particularly with respect to combine-harvesters. The areas theoretically served by one tractor, and even more by one combine are still very large in many WANA countries. The extent to which these figures translate into a genuine need for further mechanization can only be adequately assessed at the national level. The gaps are presumably being filled by human labor, animal draft power, and stationary threshers, since the FAO figures for tractors and combines include rental equipment. Farmers' enthusiasm for mechanization should not be taken for granted; for example, the actual purchases of tractors and combines were well below 50 percent of the targets foreseen in the Tunisian Seventh Plan, because of a combination of drought (affecting financial means), rising prices, and import tariffs for protection of domestic manufacturing. Efforts to attract private farm machinery contractors evoked little response. In Morocco the growth of tractors has leveled off recently even though current numbers are estimated to be only about 50 percent of potential "need." In Turkey, one Bank report states that ancillary equipment is needed more than additional tractors, and the recent data from Morocco and Tunisia support this conclusion. Moreover an increasing proportion of new tractor purchases are to replace unserviceable or obsolete machines, but this is not revealed by the FAO data. Many of the Bank's projects reviewed seem to take a rather casual attitude to mechanization, with inadequate assessment of demand and very little discussion of the technical, social, or environmental aspects of what is to be financed. In Turkey, for example, it is stated in one project that shortage of ancillary equipment is the major issue, yet instead of concentrating on filling this gap, the project provides for a relatively limited number of additional tractors, which represent only about half of the annual replacement rate. The Bank mechanization study states "to consider accelerating mechanization as an end in itself is to confuse means with ends" (Binswanger and Donovan 1987). New proposals for financing mechanization in Bank projects should therefore be evaluated with greater rigor by appraisal missions. In assessing future needs, especially under small farm and/or hilly, erosion prone conditions, the Bank should proceed circumspectly in support of mechanization 83 Considerationsfor WANA Experience in WANA suggests that other things being equal, initial priority should be accorded to support of power operations (especially conservation tillage, threshing, pumping, haymaking, transportation) where current levels of mechanization are low, and to control operations where mechanization in the form of tractors or combines has already made significant progress (for example, an arbitrary figure of 100 ha/tractor or 600 ha/combine harvester). As the area worked per hectare for power operations declines, especially in rainfed areas, it may be necessary for owner-operators to seek opportunities for performiing rental services to other farmers, possibly competing with full-time contractors. While they may be able to underprice the contractors, their efficiency and equipment is not necessarily better. Where tractor densities are high, priority may be required for annual replacements rather than for new tractors to increase their numbers still further, and also to the provision of quality ancillary equipment for precision operations such as minimum tillage, seed/fertilizer placement, weed and pest control, baling, and harvesting. Turkey may be a case in point. Rental operations have proved valuable to smaller farmers in WANA for the performance of tasks that require power beyond the capacity of draft animals or family labor and where timeliness is important. These mainly involve pre-planting tillage and 'cover crop' operations and harvesting/threshing, although contract baling is also reported to be increasing in Morocco. It is not clear why contractors seem unwilling or unable to undertake seed/fertilizer application or spraying operations, and efforts should be made to encourage them to move in this direction. Efforts should also be made to stimulate the development of efficient and reasonably priced seed/fertilizer drills, planters, sprayers and walk-behind conveyor reapers suited to small farm and animal draft power, as well as the design of simple tools for operations such as seed hulling, chemical seed treatment, seed cleaning, weeding and inter-row cultivation. The role of various fiscal incentives to machinery manufacturers and farmers also requires careful consideration in Bank dealings with national governments. Subsidized imports and loans to farmers, tax- breaks, etc., have been an important and socially costly incentive to mechanization in several countries, especially with respect to combine harvesters, and should be removed in Bank-supported projects. The Moroccan government has subsidized the purchase of tractors and combines, but Bansal and El Gharras (1993) question the value of continuing this type of support since it is now contributing little to the further growth of mechanization, but is maintaining high prices of machinery and spare parts that (particularly for control operations) is acting as a deterrent to their adoption. They also claim that this acts as a deterrent to local manufacturers who have to pay high interest rates on capital, whereas importers--often with a much larger global market, can get the same levels of subsidy. Providing tariff or non-tariff protection to local manufacturers is another thomy issue. In Tunisia farmers were discouraged from investing in farm machinery by its high costs and poor quality because of protection of the farm machinery industry. Imports have therefore been liberalized. In Morocco the situation is somewhat similar. Subsidies for tractors and other machines maintain high prices both of imported and local equipment, large domestic manufacturers are unwilling to venture outside the production of tractors, disc harrows, and trailers, imported drills and other implements for control operations are expensive and not suited to animal draft, and the smaller entrepreneurs have not had a good record technically or financially in producing and marketing innovative locally designed implements. Although in Morocco the national research institute seems to be making good progress in designing such equipment, the issue is complicated in many other countries by the inability of public sector research institutions or universities to develop machines acceptable to small farmers and their animals, and the 84 reluctance of local machinery manufacturers or importers to do so themselves. Thus there is an almost insoluble dilemma. Chapter 6 ANIMAL PRODUCTION The improvement of ruminant livestock production is an important part of virtually all World Bank projects reviewed, whether or not they were livestock-only projects. The importance of livestock in agricultural development also transcends regional boundaries of the agro-ecological situations covered in this study; animals figure largely in the projects reviewed in South America, in WANA, and in China, despite the different political, social, and structural conditions (farm size in particular) among the countries covered by the various projects. In WANA, livestock are of primary importance on small family farms. Glenn (1988) reports that in Morocco farms of between 1 and 7 ha carry four times as many animals per hectare s farms over 20 ha. In Tunisia, farms under 5 ha have a stocking rate 8 times as high as those of 50 ha-100 ha, while in Egypt farms of 0.5 ha to 1 ha support six times as many animals per unit area as those of 30 ha -50 ha. Animal production represented 36 percent of agricultural GDP in WANA in the mid-1980s (39 percent in North Africa and 35 percent in West Asia), and in several of the drier countries its share exceeded 45 percent. Thus livestock are of great social and economic significance in this region, and their importance increases with declining rainfall. This is not generally reflected in the weight accorded to the livestock sector in government policies nor in the allocation of resources to research and extension, however (Oram 1988). Rising demand for red meat because of rapid growth in both population and income has driven up livestock prices in WANA, and in the absence of increased productivity in beef and small ruminants, sheep numbers have risen from 127 million in 1961-65 to 186 million in 1988-90, and cattle numbers increased from 40 million to 56 million, although goats declined from 74 million to 71 million over that period (Annex 5.1, 5.2, and 5.3). 9 The main gains have been in poultry meat, which has increased to 35 percent of the total supply in the eight countries surveyed, while the share of sheep and goat meat has declined from 42 percent to 33 percent, and that of beef and buffalo meat from 42 percent to 30 percent. The chicken population has almost doubled since the period 1979-81 (Cunningham 1992). Animals are also an important component of family farms in the cold, low-rainfall areas of Northern China. The number of small ruminants has also risen significantly since the period 1961-65 in China, with sheep increasing from 64 million to 109 million (an increase of 70 percent) and goats from 53 million to 89 million (68 percent). Cattle increased more slowly from 61 million to 75 million (23 percent). While regional differences within China can not be identified, a large share of the increase in small runinants is probably in the colder and drier areas of northern China, where pastures are a major component of the land. 9 The cattle population is dominated by Turkey, where it has declined by 30 percent between 1961-65 and 1988-90, and the Sudan where it has increased by 55 percent. These two countries comprise roughly 60 percent of the total cattle in WANA. 86 These increases in animnal numbers have Figure 6 1 North Africa Changes in Cattle, Sheep been pareled in amanuntrs e and not been paralleled In Latm Arican countnes 8000 . Pb *atialot;196.4 141, Lnau1 90 with similar climates to WANA, for example in X 60000 - USheeji ... iArgentina, Chile, and Mexico. Overall, cattle s 20000 j-,w ...Go.......t... .........represent a more important share of total animal ; units there than in WANA, although their 1961-65 1971-81 1980-90 numbers have declined in Argentina and Chile from 1979-81 to 1989-91 and remained more or Figure 6.2 West Asia: Changes in Cattle, Sheep and less stable in Mexico. Sheep populations fell over GoatPbpdatioi; 1%1-65,1971-81, and 1980-90 the same period in Argentina and Mexico, and C-- ate rose by only 2 percent in Chile. Goat numbers z 1 -- --; --;--*s .................. increased slightly in Mexico and Argentina and X-.............. ........ .... ... ........ stagnated in Chile. The reasons for the slower growth of ruminant livestock numbers in those 0 196165 1971-81 19090 countnes, especially with respect to meat, probably lies in the already high levels of consumption, which although still rising, are expanding more slowly than in WANA and China. Also, production depends heavily on export markets, especially in Argentina. Low international beef prices and restrictions placed on unprocessed meat from Latin American countries by European and North American importing countries have depressed the meat industry there. Individual livestock enterprises tend to be much larger than in WANA or China, and hence are more influenced by market fluctuations. This applies to mutton as well as beef, and to wool--where the collapse of world prices in the late 1980s has drastically reduced incomes of farmers rearing sheep primarily for wool. With different structural and social situations in these three geographical regions, the problems facing ruminant livestock production are different as well. In the more land-abundant, upper-middle income Latin American countries the emphasis tends to be on cattle, farms are relatively large, pastures rather than forage crops are the basic source of feed, and veterinary problems--particularly those such as foot and mouth disease which affect export potential--are very important. Degradation due to overgrazing is not as pervasive a problem as in WANA or in China, because pastures are privately rather than communally owned. Whether capital intensive measures to raise productivity will be adopted under current market conditions is uncertain In WANA, the needs are different. A continuing increase in numbers of small ruminants under existing systems of management is widely held to be a recipe for environmental and possibly economic disaster (Cooper and Bailey 1990). Despite continuing strong demand for the preferred mutton, growth rates of animal population and absolute numbers seem to have peaked in several countries. Therefore, if escalating meat imports are to be avoided, productivity per head must be increased. This applies to cattle, sheep, and goats; but particularly to sheep, since the ecology of natural grazing land in much of the region is better suited to small ruminants than to cattle. The problems facing the WANA countries appear particularly difficult to resolve, for five main reasons. The productivity of sheep and goats in WANA is higher than that in any other developing or developed region, except North America and Western Europe--which exceeds WANA for milk but not for meat (annex tables A35 and A36). This complicates the target of increasing productivity to meet demand. 87 * Ecological conditions in much of the main livestock rearing areas in WANA are severe; in particular, rainfall is extremely variable both between years and within seasons. * Measures to take the pressure off the natural ranges through introducing annual or perennial forage or pasture species into arable rotations with cereals have been unsuccessful, despite years of research and considerable inputs of expatriate expertise from analogous ecological regions in Australia, South Africa, and the United States. * Genetic improvement programs for sheep and goats have had a limited effect, largely because of constraints imposed by inadequate feeding and management, but also in some cases by the failure of the breeding technologies. Table 6.1. Productivity of Regional Sheep Populations (1981) TOTAL READ CARCASS MILK YIELD REGION NUMBER SLAUGHTER YIELD (kg) (millions) (percent total) (g) North America 13 47 11.8 - Western Europe 92 62 9.4 27.9 USSR 142 37 5.8 0.7 North Africa MiddleEast) 225 38 5.9 15.2 Africa (Central & 125 28 3.4 2.4 Southern) Note: Carcass and milk yields are per head in regional herd. Source. Livestock International. Population and production statistics for 1981 are summarized from FAO 1982. Table 6.2. Productivity of Regional Goat Populations (1981) TOTAL HEAD) CARCASS MK REGION NUMBER SLAUGHTRE YIELD YIELD (milions) (% total) (Ig) (kg) North Amenca 1.4 -- -- -- Western Europe 10.1 80 7.6 145.0 USSR 5.9 46 6.9 67.6 North Af&ica (Middle 112.2 39 5.2 22.2 East)__ _ _ _ Africa (Central & 123.3 32 3.7 6.6 Southern) Note: Carcass and milk yields are per head in regional herd. Source: Livestock International. Population and production statistics for 1981 are summarized from FAO 1982. 88 Four main avenues have been followed by the Bank and other donors in assisting WANA countries to improve animal production and management. These include (a) the improvement of nutritional status (forage and water supplies, use of by-products and treatment of straw), (b) upgrading the genetic potential of the animals, (c) control of animal diseases, and (d) the provision of adequate housing and shelter for animals and implements and of storage for feed supplies. A related area of attention in some projects is the development or supply of improved animal draft equipment. Off-farm measures include the development of milk-collection schemes, milk and cheese processing plants, slaughterhouses, cooperative broiler production units, and day-old-chick-breeding centers, wool grading and processing facilities, and feed processing and mixing plants. These can play an important part in providing incentives to farmers and flock owners but will not be discussed in detail in this review, which concentrates on production technologies for farms and for ranges. In China, the needs with respect to small ruminants appear again similar to those of WANA. At the national level the situation is less critical, however, since unlike WANA, China has vast populations of ducks and pigs. Also, productivity of cattle and buffaloes in China has risen significantly over the past decade. Similarly, social and political problems appear to be more difficult to resolve in WANA than in China (where the influence of state and provincial government in resource use and management is much more pervasive), or in Latin America (where private ownership is dominant and farms are relatively large). These problems are due partly to the dispersed and fragmented nature and low capitalization of family farms in WANA countries, and partly to the dismantling by governments of traditional tribal arrangements over grazing rights without a viable management alternative for the common property resources being put in their place. IMPROVING NUTRITION Shortage of feed in the right quantity and quality is generally accepted as the most formidable technical constraint on livestock development in the WANA region (Qureshi 1991, Carter 1978, ICARDA 1988). Hence, the Bank has given priority to improving animal nutrition in many projects in the WANA region as well as in Northern China; and, in combination with veterinary measures, in South America . Despite the difficulties which have affected the achievement of this goal, the choice is correct both in order to increase animal productivity, and to counter escalating imports of feedstuffs. A recent paper by Nordblom and Shomo (ICARDA 1994) underlines the importance of giving high priority to efforts to increase feed production for livestock in WANA, as the region becomes increasingly dependent on concentrate feeds, not only for poultry, dairy, and feedlot production, but also in barley grain for small ruminants. In most countries there is also an increasing reliance on crop residues, including cotton seed cakes and meal, sugar beet and cane residues, wheat and barley straw, rice and maize straw from irrigated lands, pulse crop residues, and to a lesser extent forage crops. While the relative degree of dependence on the three main sources of feed (natural ranges, forage crops and crop residue; and grains and concentrates) varies from country to country, the general trend is a progressive decline of the contribution from the natural grazings in pastures and forests. Although the contribution of crops and sown pastures is expected to increase in absolute terms to the year 2000 and beyond, the growth of demand with rising populations and incomes is likely to outrun domestic production in eight of the nine countries studied by the year 2000, compared to only six in 1989 (annex table A37) The aggregate gross feed deficit for those countries would be around 100 million tons, more than double the level in 1989. By 2025, only one country would not be in deficit, and the gross imports required to meet demand at the constant assumed level of 0.2 Livestock Units (LU) per capita 89 would be around 225 million tons.10 These figures, although fraught with the usual difficulties of measuring feed derived from pastures and forests, are probably based on somewhat better data than earlier projections by FAO Nd IFPRI are substantially higher than other estimates for the year 2000, and 2025.11 The ICARDA feed estimate for the latter year is more than double the entire year 2000 net cereal deficits for all developing countries for all end uses (food, feed, seed and waste) projected in a number of recent studies (Oram 1993). In this context it should also be noted that China has recently increased its estimate of feed needs for the year 2000 to 99 million tons of cereals; highlighting the need for continuing priority Bank assistance to the nutritional needs of livestock in that country also. Increasing Domestic Feed Production With a critical shortage of new land to bring into agricultural use, the key approach is to find ways of using existing resources more productively. Major changes in the entire farming system to increase feed supply have already been discussed in chapter 3. The simpler "component options" to increase feed supplies from domestic resources include: * Improving natural ranges by reseeding (usually with legumes), and/or by the application of phosphoric fertilizer, sometimes supported by chemical weed control to eliminate noxious, or unpalatable species or alternatively planting or seeding adapted species or shrubs (in either case necessitating grazing management ) * Improving the methods of conservation and utilization of forage crops and of by-products ( such as straw, oilseed residues, beet pulp, and animal excrement), including measures such as urea or ammonium treatment of straw to improve its digestibility and palatability * Supplementary feeding of concentrates, either to prevent catastrophic losses of animals as a result of drought, or to provide extra nutrients at critical periods of the production cycle (such as during breeding, pregnancy, and fattening) - "Stratifying" ruminant production across ecozones to take advantage of ecological complements, especially to reduce reliance on natural grazings during certain periods of scarcity. Concerning the last point, traditionally whole flocks moved around under nomadic or transhumant grazing systems with the animals finished on the range for direct sale. Given the heavy pressures of animals on a diminishing area of natural pastures, alternative outlets have emerged with the range producing the immature stock, and sedentary systems for fattening. These fattening systems can range from fattening one or two animals at the household level for home consumption to larger fattening enterprises on arable farms (using cereals, legumes, and byproducts) to commercial feedlot operations that may be cooperatively or privately owned. ° Conversion factors used for a 500 kg LU are sheep, .09; goats, .08; cattle, .70; buffalo, 1.00; camels, 1.10; horses and mules, .80; asses, .60 (Kosilla 1988). " ACSAD/AOAD (1985) assessed grazing offtakes from pastures and forests in ten WANA countries. These values show Saudi Arabia at 20 kg/ha; Jordan, 79; Egypt, 90; Libya, 100; Iraq, 110; Syria, 200; Algeria, 300; Morocco, 500; and Tunisia, 600 kg/ha. 90 These options have been tested in development assistance programs and in Bank-funded projects in WANA during the last forty years. Despite these efforts, the feed situation in most countries of the region is becoming increasingly critical. Improving Natural Pasture Natural, peimanent pasture occupies the largest share of WANA's agricultural land, except in Tunisia, Cyprus, Lebanon, and Turkey. The area remained roughly constant or has increased slightly between 1974 and 1989 in all countries except Algeria, where it declined sharply by 15 percent, and Syria and Turkey, where it has declined by 14 percent and 17 percent respectively. In those countries, the amount of land characterized as "other" rose by 21 and 10 percent respectively in the same period. Where the share of pasture has decreased over time it has sometimes been accompanied by an increase in the forested area, and in one country (Algeria), by an expansion of cropped area and "other land" as well, but generally the overall change in other components of agricultural area has been smaller than the decrease in the area of pasture. In China, the very large area described as pasture (319 million hectares) has remained unchanged. Pasture area has risen by 13 percent in Chile, apparently at the expense of "other land", while it has declined by 1 percent in Argentina (where arable area and "other land" have both risen almost 4 percent). in Mexico, where no change is recorded in pasture area, arable area has risen 4 percent and 'other land' by almost 19 percent, primarily at the expense of forests, which have declined 17 percent over the same period. Broadly, two classes of "pasture" can be distinguished in the region: Meadows in lowlands or upland valleys. This is the smallest pasture class, which may be privately owned, but is generally communally grazed by village-based flocks. Efforts to improve these grazings have focused on developing water supply, phosphate application, and occasionally by reseeding in a few Bank- funded projects for example, in Erzerum, Turkey. While these pastures are often situated in slightly more productive environments than steppe or mountain grazings, and may be used to provide hay as well as grazing, their communal nature and often limited extent makes their improvement a difficult task, unless agreement can be reached with their users to control stocking rates and movements of animals. Upland Pastures. The area of these grazings is very considerable in West Asia (in Turkey alone they probably represent about 70 percent of the national total of approximately 20 million ha) but relatively small in North Africa. They are principally under communal use, although in areas of West Asia where transhumance is important (that is, where flocks move from lower to higher altitudes in late spring to take advantage of upland pastures, and return to the lowlands in fall and winter ) there are well-recognized tribal rights to upland pastures with traditional routes of access. In evaluating the potential for improvement of these grazings, it is important to distinguish mountain/valley inter-relationships as in Eastern Turkey, and Morocco from the situation in the high cold steppes of Central Asia (Northern Turkey, Kazakhstan, Mongolia, Tibet, Northern China) which suffer both from cold and drought and which are a high altitude version of the semi-arid lowland marginal steppes which mostly lie between the cultivable land and the deserts in Syria, Iraq, Iran, Jordan, Egypt, Libya, Algeria, and Morocco. 91 Bank Experiences Considerable Bank resources have been directed toward improving pastures in upland watersheds of Morocco, Algeria, Tunisia, Jordan, Turkey, and Iran. This is a continuing priority in several of those countries. The measures adopted vary from project to project with the ecology of the project area, the steepness of the slopes, and the altitude. On moderate slopes at intermediate altitudes where cropping is possible, several projects have encouraged the introduction of contour banks or terraces, which may be either managed on a modified ley farming system alternating cereals with annual forage legumes or temporary pastures; or sown to permanent grass-legume pastures, or planted with shrubs. In some cases fruit trees, olives, or almonds are planted on the banquettes, together with permanent pastures. Out of twenty-seven projects reviewed with livestock components, eleven involved some activities aimed at improving the quality of natural forest and range grazings at higher altitudes through ground and/or aerial reseeding or shrub planting; fertilizer application, weed control, and better management. The latter generally included resting and regeneration of selected areas, rotational grazing, and regulation of stocking rates, all supported by investments in fencing, road improvement, and range water supplies (sometimes also in the improvement of village water supplies). Despite these investments, it often proved difficult to persuade the users to cooperate whole-heartedly in the programs proposed, (although some projects failed to consult them adequately in the planning stage), so some form of discipline or coercion had to be introduced or the project staff attempted unsuccessfully to manage the pastures themselves (or both). The results of these projects were quite mixed. The impact of the projects on raising the productivity of forest and rangelands also varied, although with no particular pattern. In some cases the forest grazings responded well, in others the results were extremely disappointing. The large range improvement program in the Moroccan Middle Atlas project can be rated as successful though costly, while in northwestern Tunisia the expected results were not achieved and appear to have been conceptually unrealistic. The following experiences came to light in summarizing these project. * Although conceptually attractive, it is by no means assured that increasing the area of forage crops on arable farms at lower altitudes will reduce the pressure of animals (probably owned by different people) on upland rangelands (grazed communally with mixed flocks). In some cases (for example transhumance, where winter and summer pastures are at different altitudes)measures to improve one resource must be closely integrated with the capacity of the other to avoid imbalances in nutrient availability and excessive pressures on the neglected area. * Fodder production on terraced land at higher altitudes or within strip-cropping contour farming systems seems better suited to perennial than annual species. The costs are lower than those of annual forage crops in rotations, and there is less likelihood of soil disturbance or damage to structures since cultivation operations are greatly reduced, and the perennial forages can either be grazed or cut for hay. Costs of routine management once the anchor crops are established are mainly confined to weed and rodent control, plus periodic phosphate application (not necessarily every year); and maintenance of banquettes and drains. Nevertheless, farmer acceptance of perennial pastures on their own land has been mnixed, being somewhat better in Eastern Turkey, (where sainfoin is grown quite extensively in the colder areas) than in Morocco (where attempts to introduce contour strips of pasture or shrubs on hillside farms had to be abandoned). Plantings of forage shrubs on eroded collective lands were reasonably successful. 92 * While local people respect shrubs or improved pastures on private land, this does not necessarily apply to common property resources. Thus it is important to involve local communities in planning and management. If upland or steppe grazings and shrubs can be protected from grazing for appropriate periods and soil degeneration has not gone too far, the regrowth of desirable species can be astonishing--especially where aridity is not excessive."2 Given the existing free-for-all exploitation of common grazing and fuelwood resources, such progress is virtually impossible without the users' cooperation. "hile the Middle Atlas Project (box 6.1) was quite successful, similar success did not attend the Northwest Rural Development Project in Tunisia, which adopted a more top-down regulation of upland pastures; including taking over rangeland improvement and then renting the improved lands back to their "owners". Not surprisingly, this approach did not capture the hearts and minds of the local people, who were often treated as mere laborers in the various project works. The project staff did not have the competence needed to manage the pasture properly. * The provision of social services through project funds, including rural roads, water supplies, schools, clinics, milk collection and processing plants, has provided incentives to local populations in some areas to adopt technical resource management or other measures which meet the project goals. * The production targets appear to have been over-optimistic. For example, there were targets of a 50 percent increase in sheep production and 100 percent in goat production in the Loukkos project, and 54 percent for cattle and 42 percent for sheep in Erzerum. The proposed 80-90 percent vegetation cover throughout the year in the Anatolia Watershed Project seems unrealistic. The projected wool yields were 6.5 kg for a finewool sheep in China, 25 percent above top Merino yields in Australia. Over-estimation of forage supply from the project was a main reason for failure of the livestock component in North West Tunisia Rural Development project. Lowland Steppe and Marginal Rainfall Rangelands Range and livestock production systems are an important part of the agricultural system of the dry areas of the WANA region (those with less than 250 mm annual rainfall), which occupy about 100 million ha or 30 percent of the total land area and accommodate most of the region's flock of 186 million sheep and 71 million goats. These areas are not well suited to cattle, which are mainly kept in irrigated or high- rainfall areas or on upland pastures. Most of the small ruminants spend some time during the year on these marginal rangelands, but the amount of sustenance they obtain from is generally less than half of their total annual feed intake. The balance is being provided by grazing arable fallows and stubbles, orchards, wastelands, weeds, and crop residues, supplemented increasingly by feed grains (usually barley), by products such as oilcakes and by beet pulp, straw, and dry forage. The flock owners may own arable land themselves or have a variety of arrangements with land owners in arable fanning areas and move to farmland for several months once the natural pastures dry up in the summer. Despite the conceptual 12 Tadros (1992) cites dry weight forage yields of 850 kg/ha to 1,380 kg/ha at protected range reserve sites in Jordan compared with about 100 kg/ha average on open range and rainfed barley yields of between 250 kg/ha and 800 kg/ha. 93 attractiveness of a closely stratified system of rearing young stock on the steppe when it is at its peak, and selling them to farmers in better-rainfall ecozones or to feed lots, this organized integration has yet to be achieved. Consequently, the steppe remains heavily stocked and young animals do not move systematically on to a higher plane of nutrition off the range once its productivity declines. The situation has been aggravated in the last 25 years by the nationalization of the tribal rangelands in several countries of the region without putting an effective system of grazing control in place of the traditional checks and balances on animal numbers and movements agreed among users. In the absence of any firm control the steppe has become an open access area, and in recent years its users have sequestered pieces of the steppe on which they may remain for long periods even after range productivity falls off, supported by water tankers and supplementary feed. A related trend is the use of tractors to cultivate steppeland in the less arid areas for barley production as supplementary feed. Entrepreneurs from urban areas have also moved in, with purchased flocks supervised by hired shepherds who have no stake in the long-term sustainability of the resource. Box 6.L Community Range Development In The Middle Atlas Some projects have been more successful than others in developing community ranges. One example of relative success is the Middle Atlas project in Morocco, which observed that other ongoing activities with similar goals in that area were being hampered by failing to involve local people. The project set:out to develop a participatory approach to range and forest management. In addition to a range improvement program involving fencing and water supply of 50,000 ha, fertilization of 1,300 ha and reseediig of 2,000 ha of collective pastures; forest range improvement and management with grazing control, fertilization, and shrub planting. Grazing regulation, grazing charges, and enforcement by range police, includirg collective "sanctions" vere instituted with agreement by the users. The program was supported by a special extension effort and by training. Despite the disciplinary measures for collective-range management, the proposals seem to have been well accepted and adopted quite successfully by'the people, who cooperated in enforcement. The project measures aimed at increasing productivity on private fairs fell well below expectations, however. Forest range management efforts were also disappointing.' Only a third of the target area was covered Source: Authors. As a result, the natural grazings (especially the low-rainfall steppelands) have come under severe pressure, with stocking rates declining from 2.8 Tropical Livestock Units (TLU) in 1961 to 1.7 TLU in 1988 (Glenn 1988). In 1969, Le Houerou estimated that rangelands in Tunisia were overstocked 24-45 percent, depending on the region, although the extent to which they were degraded and the degree of reversibility of an damage is uncertain. De Haan (1990) states that qualitative assessments from World Bank reports indicate that 50-80 percent of the rangelands in the Maghreb, Yemen, Eastern Turkey, and Pakistan are degraded and substantially underproductive. Information from other sources (FAO 1987, ICARDA 1992 for Turkey and Syria; Tadros 1992 for Jordan) supports these assessments. Despite this large and apparently convincing body of evidence concerning overstocking of the range, some skepticism is being expressed both about its validity and the impact on degradation of the resource. Past assessments of optimal stocking levels were mainly based on the concept of carrying capacity, and the validity of using carrying capacity as a means of regulating range use is being increasingly questioned (de Haan 1993). This concept 94 * Depends strongly on the management objective * Varies highly inter-annually, depending on rainfall (and is therefore of low predictive value) * Is difficult to estimate accurately, since it tends to be arbitrarily defined and does not account adequately for landscape heterogeneity, animal mobility, and different grazing behavior among species * Has a low practical value since no case has been documented where pastoralists are willing to voluntarily reduce stock numbers to conform to some estimated carrying capacity. Doubt is also being cast on the evidence concerning major degradation of arid range lands, especially allegedly irreversible degradation. An altered spectrum of vegetation does not necessarily mean that a pasture is degraded. The idea that the higher stocking rates now being recorded in many semi-arid regions must inevitably result in higher levels of degradation is also being challenged. Examples to support such statements come from South Africa, where a range reported to be 300 percent overstocked in 1948 is now just as productive, more densely stocked, and with a higher rate of annual offtake. Also, in Botswana where a drastic reduction in the stocking rate made only a minor difference in erosion and soil loss, but significantly reduced the gross margin per hectare. This leads to the conclusion that the level of grazing pressure in one year has little effect on biomass production and soil characteristics of the next year (Behnke, Scoones, and Kerven 1994). These concepts involve important conceptual and practical issues, of which perhaps the most difficult is the question of irreversibility. They are mainly based on Sub-Saharan Africa and these new concepts of range management need further research under the conditions of WANA because (a) they are based on a summer rainfall moisture regime, where cattle dominate rather than sheep, (b) many studies on grazing pressure are not based on long-term trials or do not include a variety of essential data, and (c) some countries do indicate declines in the numbers of small ruminants and in range biomass production. In the meantime, an extremely dynamic and fluid situation has developed in WANA with respect to sheep and goat production in general, and the management and use of the range grazings in particular. Technology exists to improve the productivity of the range. Its chances of success, however, appear small if the policy, institutional, and legislative framework favorable to rational management of the resource is not in place, and the users of the range cannot be convinced of the benefits. Such a framework should support not only the sound management of the natural grazings, but also of the structural and institutional links between the low rainfall and upland range areas and the arable farms or feedlots in areas of higher agricultural potential (where range-reared animals could be finished to economic carcass weights). This requires a holistic view of the small ruminant production system and how it fits into the agricultural sector and the economy as a whole. Recent evidence from four countries where range livestock are important shows that each has different approaches and attitudes to range management and property rights. These range from nationalization of the range and detribalization in Syria (with no effective alternative management system), to nationalization with explicit acknowledgment of tribal and cooperative rights in Jordan, to encouragement of privatization through land registration for example in North Africa. Often more than one govemment agency is involved, and sometimes their objectives seem contradictory, one promoting communal or cooperative action by range users, another privatization. Different situations exist within countries with respect to invasion of the grazings. In southem Tunisia olive cultivation is tihe main 95 incentive in Central Tunisia it is barley. In several countries, a main incentive for cultivating steppe range- lands is to stake a de factor claim to the land. In two countries (Morocco and Tunisia) commissions have been established to review the situation and while in Jordan a study of the problem is under way with IFAD support. New legislation is contemplated. There are, however, weak links in communication between research staff and administrators (and even within governments) that impede the development of solutions. Two factors present a paradox. Solutions must be found to the property rights problems before the technical measures discussed above can succeed. At the same time, flock owners are unlikely to accept discipline in their use of the resource if they see no gains in its productivity from the use of technology. Government policies have given contradictory signals with respect to ruminant livestock production, and especially concerning the exploitation of the rangelands. In addition to the problems caused by nationalization of the range, other actions which appear to have aggravated the difficulties of its rational use include: * High levels of protection of locally-produced meat prices that have made sheep and goat production attractive * High support prices for domestic cereals that encourage cultivation of marginal lands * Feed concentrate subsidies that discourage investments in range improvement and forage production on farms * Fodder banks and free feed distribution during droughts that are disincentives to stocking rate adjustments and that impede range recuperation after the drought * Subsidies to fuel and tractors, that enable flocks to be moved around opportunistically to exploit good grazings regardless of their need for recuperation, or to park for long periods on a piece of the steppe supported by water tankers and supplementary feed'3 * Policies for resettlement of nomads that have concentrated animals on grazings near to population centers, with especially damaging consequences on the quality of the vegetation * Intensification of production in higher rainfall areas, and subsidized combine-harvesting of cereals that have reduced the availability of fallow, stubble, and orchard grazings after the range productivity declines in summer (thus maintaining pressure on range grazings) Two other major points aggravate the difficulties of rational use of rangeland. The first of these is inconsistent or laissez-faire policies towards small ruminants generally, and grazing management in particular. Government attempts to impose regulations on range users have included prohibition of cultivation of steppe lands for barley below specified rainfall limits, herd registration for grazing permits, head taxes on animals in excess of a registered number, and various "carrot and stick" measures such as provision of water points on the range, access to supplementary feed in time of drought, or access to 13 A 1989 survey in Jordan shows that 52 percent of 85 sheep farmers interviewed owned a tractor, 71 percent owned a truck or a pickup, 19 percent owned a water tank, and only 18 percent had no wheels. Of those farmers, only 21 percent were fully nomadic, 23 percent had a home base but no land, 7 percent had land but no home, and 48 percent had both land and a home base (Tadros 1989). 96 government fodder reserves and shrub plantations; in return for acceptance of grazing discipline. On the whole these measures do not seem to have made a large contribution to improving range management or productivity, and in some cases--provision of water points--may have exacerbated overgrazing. They have rarely formed part of a consistent national plan for ruminant livestock production, and have not provided sufficient incentives to offset the negative effects of the wider government price and subsidy policies referred to above which mostly benefit a more influential urban political constituency. Finally, low-rainfall ecozones are a low priority in national research and extension programs in most WANA countries and there is a shortage of well trained specialists (partly because of inadequate University curricula). It should be noted here that until very recently this has also been the case with international agricultural research priorities--both in WANA (ICARDA) and in Sub-Saharan Africa (ILCA). Only one regional center in WANA--ACSAD--has consistently been dedicated to research on problems of range improvement, although the USAID-supported Aridoculture Program with INRA in Morocco and the Arid Zone Institute of Medicine in Tunisia are also developing innovative research on low-rainfall agriculture. This perception of low priority to problems of agricultural development in marginal areas, especially steppe grazings, clearly affects Bank policies in WANA. Of the numerous projects reviewed here, only two--Syria First Livestock development, and Jordan First Agricultural Credit--include any significant investment in steppe management, and both of those are old projects, whose long-term impact is difficult to track. Bank Experience The Bank-funded Syrian First Livestock Development project did pursue some innovative ideas- including the establishment of sheep breeding cooperatives, emergency feed reserves and a national feed revolving fund. There are several lessons from that project. * Range surveys with an agreed methodology are needed early in a project. The survey for this project was undertaken by ACSAD, but was not completed during the project's life- span, and ACSAD apparently failed to devise and implement any rapid large scale appraisal technique to provide infornation on range condition.'4 * The administration of the proposed large-scale emergency feed reserve (EFR) proved difficult, and its costs high and hard to predict because of great inter-annual weather variability and hence unstable demand for the feed. It was concluded that the purchase of emergency feed should be left to the flock owners, with the state's role being confined to monitor the feed adequacy and availability. The audit report also questions the wisdom of the government encouraging routine reliance on concentrates of up to 30-40 percent of breeding flock requirements, when the range grazing potential is apparently under-utilized. * The availability of feed at subsidized interest rates to sheep breeding and range improvement cooperatives, emergency feed reserves, and a successful animal health programn led to a doubling of the number of sheep. The project never mentions goats. 14 Recently ACSAD has begun to remedy this gap by developing a sophisticated remote sensing and imaging capability, linked to GIS technology, with assistance from Germany. 97 Because there was no emergency feed reserve, however, the increase proved unstable and numbers declined dramatically during the next drought. * Fattening cooperatives did not increase in numbers and used very little funds, partly because of being charged higher interest rates than breeding cooperatives, but also because of self-financing from the high financial returns to fattening. * The National Feed Policy Committee and National Feed Revolving Fund provided crucial institutional support, which contributed greatly to the project's success. * The PAR/PCR report provides little detail about the nature and operating procedures of the cooperatives, although their success was crucial to the project. The whole system of rearing, finishing, marketing, and pricing small ruminants, including the range, the low rainfall barley sheep farms, the arable farmns in higher rainfall ecozones, and the fattening cooperatives should be looked at critically and in-depth. * Although the audit notes that the range potential is greatly underutilized, the project did little to develop that potential and the SAR does not suggest any specific technical measures to introduce. Lack of monitoring and evaluation of range conditions and stocking rates is noted as a serious gap in information with respect to its future management. Uncertainty as to the true build-up of numbers during the project, the extent of losses due to drought in the final year, and inadequate data on costs of sheep production impeded assessment of the ERR. OPTIONS FOR TECHNOLOGY IN MARGINAL STEPPE AREAS The amount of "best bet" technology for range improvement in arid zones is limited. The constraints imposed by the extremely low and unreliable rainfall in steppe areas preclude capital intensive investments in fencing and reseeding, and past efforts of this nature in WANA have not been very successful or sustainable. Simple, relatively low-cost techniques are required, the results of which can be readily demonstrated to range users in order to gain their confidence and cooperation in applying them on open rangelands through appropriate communal grazing agreements. : The first of these are environmentally passive techniques for rangeland restoration based on development of small block seeding techniques that strategically utilize natural wind and water forces to distribute seeds to revegetate degraded areas accompanied by deferred grazing. In less severely degraded rangelands resting alone to allow regrowth of palatable species can give good results as recent work in areas with less than 200 mm annual rainfall in Jordan shows (Tadros 1992). Second, phosphate application to pastures accompanied by deferred grazing for a season to allow seed biomass to increase before returning to continuous grazing appears to be a promising and relatively low-cost technology, especially for legume rich swards. The effect of P205 application is cumulative over years, resulting in more and better quality grazing from January through April, with increased total biomass. In trials at ICARDA (1988) 25 kg/ha P205 gave almost as good yields as 60 kg/ha (1,440 kg/ha versus 1,580 kg/ha), compared with 1,050 kg/ha without phosphate. Micronutrient application has given 98 large responses in some countries, for example molybdenum in Australia, but has not yet been reported in WANA countries. ; Results fromcoparsons of a wside range of annual forage legumet species a*rosssevea alt anua raifal; l C varyig rom. 1S'50mmto: 330 mm, both in pure fstads andin m:a' on wit barle :show average :yields for grazing ranging from 1.71 tons of dry; matter per hectare to:E 2.63 to.os per hectar.e, with total ymatter yields: 'of the; bet -species 'such iasi J narbonesis, V elslia, ind .. .....! at........ sites 0e xceeding lthose fof barley.0 However,0 trials in Syria'and Jordanl show 'lltha conlsiderable 'differec exist both between and within species in- palatability, digestibility tand 0proteincontent as0 well smot0 chaatrsticsthat can improve quality ad yield; such as leafiness and leaf reetion.'t Amn ceels a straw is superior to most zother species, -but-barley straw is generally superior to wheat -straw: Howver work at ICARDA 1(1988) fshows thfat ithere is Dimportant variability tamong: barley icultivars, wththe local ecotype 'being among' thef most palatable. 000:Similar intra and i nter-specifi c Cdifferences "exist amn foag legulmes, pea -straw gbeing dinferior ito lentil, vetch :or chickling. Literally ;hundreds lofC genotyspes- of VicA,l Lthyrus, 2and Medicago species tare :being evaluated: for their agronomiic characteristicsa yield, and quality: -Laithyrus sahiva (common ebickling) iappears to -beD particularly Spromising 'in 'terms0 of y3;ield Cand quality6 0S~~~~~~~ ~~~~ ~~~~~~~ .;00 . .. .. .. ..Ii; .. .. .. .S ... .. . . .. ..:- .! ; .. ....... , i : f : E i ft f i. f0iiS- : -, xSeveral speces Eexhibit Vconsiderable flexibility in performance over a wide sraifall grdint in particulare rr. E ti - -. Ee : E E Ei0 ie W Eg. . . . ..E . . . . . . . . . . . ..-.;gE v, nacrbonensis EE . EE- XSource+: A uthors.::020:7S:- 0 :i :-: V;i ; !: i ;i0:: ::0002: it:;:l- t: A third technique is establishment of perennial shrubs in steppe or upland rangelands, especially where steep slopes make regular seeding of pasture species difficult or in eroded areas. Numerous species hlave been identified, including several of the genus A triplex (saltbush), A. halimus, A. semibaccata, A. nuammularia, A. canescens (Four wing saltbush), Acacia cyanophylla and Acacia tortilis; Artemisia hterba-alba; Pistacio altantica; Opuntia inermis; and Salsola vermiculata. Most of these are indigenous, or were until virtually eliminated by overgrazing; some are salt tolerant (Atriplex halimus) and several are palatable and of high nutritive value, especially Salsola. Other drought tolerant species not native to the region include several acacias, jojoba, and Prosopis species, but the range of local ecotypes covers a wide spectrum of conditions within WANA. '5 Even so care needs to be exercised in their choice: Atrpipex nZummulania does well in Morocco, but not in Syria which is more loose-prone. Direct seeding in late fall just before the onset of the rains, rather than planting is currently the preferred establishment technique. Seedlings often suffer considerably from root disturbance and damage in transportation to planting sites (especially if not planted promptly), and in semi-arid areas c 250 mm, normally have to be watered at planting and for up to two years thereafter. Seedling production and transplanting is therefore more expensive than seeding, and it is argued that after a normal to good rainfall year when the seeds germinate well and get off to a good start, the young plants need little watering thereafter. Several species will tolerate brackish water and sewage effluent from urban areas. In order to facilitate direct seeding a combined cultivator/seeder has been developed and tested by the USAID-Jordanian National Agricultural Development Project with satisfactory establishment rates. Whether the plants are grown in stu from th Grazing trials of pasture species (mainly shrubs) in several old-established range reserves in Jordan show very large differences in initial biomass (ranging fom 2,000 kg to 950 kg to 140 kg/ha at different sites, none of .hich have more than 200 mm rainfall. 99 seeding, or transplanted, they require protection from grazing for several years until they are well established. When this can be assured there is generally regeneration of natural pasture grassland legume species as well. Animals need to be introduced gradually to the shrubs since there are some examples of them rejecting the unfamiliar forage.'6 A problem with Atriplex species is that animals fed on them require to drink a lot of water because they have a high salt content. This is why many range workers prefer Salsela, which does not create thirst and is preferred by stock. Opuntia can actually store water, but is less nutritious. Fourth, probably one of the fastest ways of increasing meat production from small ruminants in low-rainfall areas would be to encourage the expansion of fattening operations. In Jordan, for example, range-produced lambs are slaughtered traditionally at between 60 days and 90 days old and at liveweights between 18 kg and 22 kg in good years or 16 kg and 20 kg in dry years. It is estimated that if the approximately 295,000 lambs weaned each year could be fattened to 40 kg, this would produce another 2,100 tons of finished red meat. An additional 300,000 to 400,000 head of live sheep or lambs are imported annually, and fattening them would provide a further 5,000 tons. A similar opportunity exists in nearly all WANA countries, and rising prices of meat with favorable meat/feed price ratios make this a profitable venture, as is the case in Syria However, while cottage type smallholder fattening has developed well over the last decade, industrial sheep fattening has never really taken off, because of problems associated with the development of feedlots including: * Availability of feedstuffs and reliability of supply, including roughage to balance increasing levels of higher-energy feed in the diet. Urea or ammonia treated straw, sugar industry and cotton byproducts, and poultry wastes are possible sources of feed for fattening operations but caution needs to be exercised with poultry manure to avoid copper poisoning. * Identification of suitable sources of feeder sheep; requiring an effective livestock marketing system. * Lack of experience of fatteners in the management and feeding of range reared lambs and sheep, and the need for training so that cost-effective diets using local resources can be established and losses due to metabolic disorders, parasites, or other causes of poor performance minimized. * Inadequate government policies, especially with respect to prices of meat and feedstuffs, credit, and inputs such as urea and pharmaceuticals, which can adversely affect returns to fattening. * Inadequate quality differentials for specially fattened sheep. Experience shows that feedlots are only profitable, if the kg. price of fattened lamb is at least 10 percent higher than the price of non-feedlot lamb. 16 Trials in Jordan suggest that cutting mature Atriplex nummalaria at 85 percent of height killed about 40 percent of the plants, while cutting too high at 15 percent left much of the growth out of reach of the animals. 50 percent of plant height appears optimum, and yields in a 200 mm rainfall zone were about 950 kg/ha. 100 Inadequate research on factors affecting fattening, and lack of extension staff with relevant training and experience. Ammonium or Urea Treatment of Straw The technique of treating straw with ammonium or urea is proposed in a few Bank-funded projects, including Morocco (Oulmes-Romani), and China (Gansu), but except in China it is not part of standard farm practice. In China, it is quite widely used. Treatment of straw with urea or anhydrous ammonia can increase its digestibility. Urea application alone raised daily liveweight gains of cattle in trials with farmers in Northern China by 150 grams. When the treated straw was supplemented by cotton-seed cake (CSC) as an additional source of energy the daily gains rose to 650 grams. Comparison of 3 percent anhydrous ammonia with 5 percent urea in solution for straw treatment showed the former to be marginally better (8 percent superiority in animal performance). However the key factor appears to be the supplementation, since raising the quantity of CSC from 1.5 to 2.5 kg/day in the same trial increased daily liveweight gain by 41 percent and 42 percent respectively for the groups receiving anhydrous ammonia and urea (annex table A39) Results are also affected by management (selection of good quality cattle and ad libitum feeding of straw). On-farm data show that average daily gains can vary from 390 grams to 630 grams using a standard 2 kg CSC supplement. Straw dry matter at 3.5 percent of animal bodyweight is a guide to feeding, equivalent to 4 kg dry straw per 100 kg liveweight and a somewhat higher allowance of wet urea treated straw. A cost-efficient supplementation level is 1.5 to 3 kg CSC per day. The return on investment over a 300-day fattening period is around 14 percent. The urea method seems effective under moderate frost levels (0-+ 5°C minimum winter temperature). The lower temperature limit for its use is not yet clear. The straw can be treated with urea at the household level by soaking it in a liquid urea solution in plastic bags. This is widely practiced; mobile plants to mass-treat straw with ammonia have been developed, but apparently with less success than with on-farm urea treatment. Liquid ammonia treatments require 10 tons of straw or more to be treated for cost effectiveness, thus is more suited to cooperatives than to individual small farmers. The implications of the Chinese national target for the year 2000 are considerable, since treatment of 35 million tons of straw is estimated to require 1.5 million tons of urea, supplemented by up to 8 million tons of CSC. The World Bank projects output of CSC by 2000 in China at 4.9 million tons and urea is already in short supply, being the key fertilizer for many crops. If these supply problems drive up prices. This could lead to selective use of the mixture for dairy and intensive beef cattle fattening rather than for all ruminants, and/or modified levels of feeding. Since CSC and straw are already important sources of animal feed in several WANA countries the ammoniation of straw and its supplementation by CSC certainly seems worth pursuing there. Awassi ewes fed chopped barley straw ad-lib with concentrate supplement from late pregnancy through lactation to the dry period lost weight even at the high concentrate level because the energy density of the diets was limiting. (ICARDA 1988) Studies in Jordan show the advantage of supplementing barley pasture or natural grazings with concentrate feed to lactating ewes, compared to barley or natural pasture alone. These trials suggest the introduction of urea treatment of straw combined with concentrates in future Bank- funded projects, at least on a trial and demonstration basis, seems worthwhile. However this is a sophisticated technique with considerable danger of poisoning of individual animals. For example, in Egypt, where this practice was introduced three years ago after successful trials with sheep, a number of adopters discontinued, apparently as a result of sickness and even deaths of animals. An alternative procedure being tested is feeding straw together with urea/molasses blocks which provide additional 101 protein uptake and better utilization of straw (ICARDA, 1994b). If this proves successful local industries could develop. In this context it is noteworthy that the FAO/UNDP project in China states: While the work of the extension service is facilitating the adoption of technology, the understanding of the relevant aspects of rumen physiology concerning feeding and the supplementation of crop residues is inadequate for proper scientific backstopping of the extension service and the farmers. This is one area requiring follow-up. Other follow-up research needed include the identification of alternative protein supplements to CSC; and possibilities of supplements containing highly digestible fiber to stimulate numbers of cellulolytic rumen bacteria; as well as developing treatments suited to sheep and goats.Because these treatments are relatively labor intensive, and require the use of supplementary ingredients that have cost and supply implications, it is important that any proposal to introduce them to WANA countries through Bank-funded projects is preceded by a careful analysis of the availability, costs, and empirical evidence of suitable levels of application and expected benefits from the different options. Policy and Socio-Economic Measures The above technological measures are unlikely to succeed in improving the productivity of rangelands or in developing a more sustainable form of land use, unless the policies, which encourage uncontrolled resource exploitation in low rainfall ecozones change. The objective must be to provide incentives to cooperation among range users so that the grazings can be stocked and managed for their mutual benefit over the long term; and to encourage more systematic and organized offtake of young stock and culls from the range and low rainfall arable farms, to fattening areas, so that pressures can be reduced on the ecologically fragile zones. Measures that might help to facilitate this include: * Phasing out government subsidies on feed grains and concentrates to encourage on-farm fodder production and the adoption of range improvement measures * Developing flexible land tenure mechanisms, building on customary institutions and systems of grazing regulation and grazing rights (such arrangements should ensure mobility and access to key grazing resources and water points to all registered users; and would have to be worked out cooperatively among them, with the government acting as facilitator) * Improving marketing structures to cope more effectively with fluctuations in feed supply, and support to systematic stratification of rearing and fattening * Establishing an effective drought contingency program, including early warning, development of reserve areas, and possibly drought insurance schemes (other measures such as food-for-work schemes might be incorporated--for example to establish strategic plantings of shrubs, stone clearance, soil conservation and water harvesting structures for range improvement). 102 An important issue in this context is what institutional framework is best to allow government to contribute most effectively, without stifling the initiative of range users and local communities to develop sustainable and mutually beneficial approaches to resource management. (Scoones 1994). Bank Experiences The Syrian Livestock Development Project established a National Feed Policy Committee (NFPC) and a project coordinating unit in the Ministry of Agriculture. The head of the Project Coordinating Unit acted as Executive Secretary to the NFPC; together with a National Revolving Fund and an Emergency Feed Reserve (EFR) to provide feed and credit to purchase it in drought years. The aim was to stabilize the national flock and to prevent catastrophic losses and distress sales in drought emergencies. These mechanisms proved very valuable in coordinating efforts of the numerous government agencies involved, and contributed significantly to stabilizing flock numbers at a record high level. They did not prevent increasing resource mining in the steppe and neighboring marginal barley areas, and did not achieve mutually acceptable arrangements among range users for its sustainable long-terms management. In fact, even without the proposed EFR (which proved too difficult to manage), the effect of the various feed support measures probably led to increased overstocking of the range and higher dependence on supplementary feed. Furthermore, it clearly shows that stabilizing national flock numbers through top- down measures is unsuccessful. Such matters are more likely to be understood and solutions to their effective resolution worked out by consensus and cooperation among the users themselves, with the government acting as convener, mentor, and facilitator, rather than using coercion or legislation. MAINTENANCE OF ANIMAL HEALTH The crucial link between animal production, nutrition, and health, especially in areas of severe climatic stress has already been indicated in the preceding section. Diseases affecting ruminants in the regions include the following. * Foot and mouth disease, especially in cattle, has a direct impact on intensive dairy production and greatly limits exports of beef from Latin America to markets in Western Europe, North America, Japan, Australia, and New Zealand; but to a lesser extent also affects sheep and goats. * Anthrax, Rinderpest, Brucellosis, Mastitis, Johne's disease, and sheep-pox, which is a major cause of mortality in WANA (especially in drought years when sheep resistance is low). Entero-toxemia, is a second widespread cause of losses, both in lambs and adults. * Acidosis and other metabolic disorders appear frequently, often resulting from sudden changes in the diet, especially when animals are introduced too rapidly to a higher plane of nutrition with unbalanced proportions of roughage and concentrates.. Sheep tend to be particularly susceptible to metabolic dysfunction during late pregnancy, including toxemia and hypocalcaemia (calcium deficiency). 103 In some cases where sheep are moved to a more intensive system of husbandry they are also more confined, in conditions which are less hygienic, leading to increased parasite infestation and other stresses. The major factors affecting flock productivity of sheep are low lambing percentages and high mortality of new-born lambs, including ones that manage to survive the critical period immediately after birth (Bahhady 1987). Mortality in Syria averaged 15 percent, ranging from almost zero to 39 per cent. Birth weights of lambs were strongly influenced by levels of feeding during gestation, as were liveweight gains throughout the suckling period. Major diseases which cause heavy losses of lambs and kids Syria after birth include pneumonia, colibacillosis, enteritis, septicemia, and (after weaning) coccidiosis in kids. In Egypt, enteritis accounts for 40 percent of all lamb deaths and pneumonia for 27 percent. It is obvious from research from different countries in the region, and from China and South America, that theprevalence of the various diseases varies quite considerably from country to country and that some are quite country specific, particularly with respect to nutritional problems at susceptible periods in the life cycles of ewes and of lambs. In China, for example, the joint Chinese-Australian team working on sheep and wool improvement in Xinjiang and Gansu found widespread mineral deficiencies in pastures, particularly of sodium. but also of selenium and copper. Providing a macro-mineral supplement or treating each deficiency individually increased weaning rates, growth, and wool production. At another site in Inner Mongolia, few signs of mineral deficiency were observed, however, and supplementation yielded no measurable benefits. Further studies are in progress. A related project looked at worm burdens in sheep and found that although the numbers were lower than those normally present in more temperate environments, their impact in this harsh environment was regular and severe, causing liveweight losses of 15 percent losses in wool production of between 5 and 15 percent. Apparently, most farms in China do not treat for worms. This underlines the need both for strong national research capabilities and veterinary services capable of providing feedback to researchers concerning such problems, as well as of monitoring the effects of curative measures. Animal Health Provisions in Bank-Funded Projects Animal health is a major concern in a few Bank-funded projects in WANA or China. These projects include Syria First Livestock, Argentina Credit III, Morocco Middle Atlas Agricultural Development, Morocco Oulmes-Romani Agricultural Development, Morocco Fez Karia-Tissa Agricultural Development, China Hebei Agricultural Development, and China Xinjiang Agricultural Development. In some projects with an important livestock component, it was assumed that improved nutrition or improved breeding (or both) will largely eliminate losses of productivity from disease. Experience in WANA shows that this is not the case--the combined action of better feeding and improved veterinary care are synergistic--while also the introduction of exotic genes through breeding actually increases susceptibility to disease. Apart from better nutrition and direct measures of preventing diseases through vaccination or control through dipping, drenching, or other medication, a number of Ban- funded projects included support for improving housing, shelter, and water supplies. In areas of severe climatic stress due to heat, drought, extreme cold, these measures may help to reduce vulnerability to disease. They may also raise productive efficiency by enabling animals to use their available feed for production of meat, milk, or wool rather than to cope with climatic stress. In northwest China, for example, sheep spend most of the winter in sheds. In some cases, provision of separate sheds for pregnant or lactating animals (or for segregation 104 during breeding) may increase herd efficiency, and may also reduce diseases resulting from overcrowding in unhygienic housing. Where specific references to veterinary services are made in Bank-funded projects the three main concerns appear to be availability of appropriate vaccines or drugs, how to reach farmers and herders, particularly in more remote areas, and how to recover the costs of providing veterinary services and treatments. These problems are not entirely independent. For example, treatment costs may be higher where drugs have to be imported, or are in short supply, and services are likely to be more costly where "customers" are scattered over large areas with poor communications. Availability of Vaccines and Drugs Vaccines are produced in most countries of the region. Syria has expanded sheep pox vaccine output to supply Jordan, Lebanon, and now develops vaccines against enterotoxemia in sheep, rinderpest in cattle, and Newcastle disease. Tunisia, Egypt, Cyprus, Turkey, and Morocco have vaccine production facilities. Turkey produces vaccines for brucellosis, and local FMD strains. Thus lack of availability of vaccines against key diseases does not appear to be a major constraint in WANA. Provisions for vaccines are referred to only vaguely in the two Chinese projects. Non-technical problems can cause problems within a country. For example, foreign exchange restrictions, changes in government financial provisions to agriculture, or high rates of inflation, have led to shortages of veterinary medicines as well as cutbacks in veterinary services and/or support to mobile clinics and extension staff mobility, in some countries e.g. Turkey and Morocco. Bureaucratic procedures for purchasing imported pharmaceuticals by centralized government agencies without adequate provision for technical advice from veterinarians have also been reported as a problem. Reaching Farmers and Herders One of the main obstacles to the adoption of sound animal health practices is the difficulty of actually reaching the producers, in more remote areas. Not only is it more costly, time-consuming, arduous, and sometimes even more dangerous, than contacting livestock owners in lowland areas, but it may also be less rewarding since people who dwell in those areas are often more conservative and less able or willing to innovate.'7 Consequently animal health services aimed at servicing farmers and shepherds in such areas tend to be cut back most severely in periods of budgetary stress. There have been several attempts to overcome this problem, including the establishment of nucleus veterinary service centers in appropriate rural population centers, usually where other government services exist (often in towns) with veterinary advisors operating from there in cars or on motor-cycles. Aside from often being cost effective, the development of mobile veterinary service units is another means of bridging the gap between the regional and provincial headquarters of the veterinary service and the field, the concept of mobile clinics is attractive. They can be equipped to provide a wider range of diagnostic and preventive services to farmers, writh appropriate staff, than is normally available from one locally-based veterinary officer. The main problem is sustaining such mobile services after project '7 As the Second Livestock Project in Turkey shows, local people may also be right in rejecting novel ideas. In this case they refused to abandon closed indoor fattening of tethered bulls in favor of an outdoor system. They not only gave good reasons for this resistance,but considerably exceeded the appraisal estimates of 600 g/day gain for cattle and 150 g/day for sheep: achieving on average 750-800 g/day and 175-200 g/day respectively with comparable initial bodyweight. 105 completion. In Syria, the units were meant to operate principally in areas without fixed veterinary centers, and the mobiles were distributed proportionately to the sheep population. This made it possible to reach almost 100 percent of the estimated 1983 flock of 8.5-9.0 million, compared to a coverage of only 6.7 percent of the national flock of 6.0-6.5 million prior to the project. Tentative calculations by the FAO cooperative program based on a flock of 9 million indicate an incremental benefit on US$15 million from control of sheep pox alone, not counting other benefits from parasite and other disease control campaigns conducted by the mobile health units. A third means of coping with budget difficulties is strengthening and unifying all crop and livestock services into one national extension system. Theoretically, the latter has many advantages; in practice it has not worked out well for livestock, particularly with respect to back-stopping disease prevention and artificial insemination (Al) campaigns, where locally-based extension workers ought to be the vital link between the livestock technicians and the flock owners but normally are not, because of other duties. Furthermore, village-level extension workers are rarely adequately trained in animal management or health. This is a vital gap identified by several project reports from WANA countries, as well as by the Australian project in northwestern China (Lehane 1991). A second (or parallel) step could be to encourage producer-owned livestock cooperatives or other organizations and assist them in providing such services to their members. The cooperatives of other organizations should be backed by government technical advice, credit, access to appropriate drugs and equipment and market intelligence. With government support, such organizations have played a major role in other regions. Establishing a network of private animal health auxiliaries ("paravets"), who can move with the herders is a fourth way to reach farmers and herders. This systems is quite common in Bank-funded projects in Sub-Saharan Africa, but has not yet been tried in Bank-funded projects in the WANA region. Paying for Animal Health Services To provide adequate services to producers with small land holdings can be a heavy burden on government budgets, which may become increasingly unpopular politically in an urbanized society. Consequently, in most cases where Bank-funded projects are specific concerning animal health provisions there is pressure on the government to charge users of the services, so as to recover at least a proportion of the costs. Generally cost recovery has been more successful than anticipated by the project authorities for those interventions felt critical by the herders. In Syria a revolving fund (RFVM) was to institute graduated charges after a three-year free demonstration period for drugs and vaccines, up to full cost after five years, just at the end of the project. Unfortunately, severe drought conditions in the final two years resulted in the charged being deferred indefinitely by the government. There is no record of whether or when those deferred charges were paid, or what has happened to the RFVM, which was still not operative in 1986, except that government instructed the Ministry concerned to institute a study of the costs of treatment by mobile units and to provide suggestions for methods of collection from flock owners. Although it is stated in the PPAR of the Syrian Livestock project that flock owners are now convinced of the value of vaccination, and willing to pay for such services in future, this did not apparently extend to the dipping program. Seven large dips were constructed, but these appear to have been used only to a limited extent, partly because of design and quality deficits in construction. Because the results are less easy to demonstrate than those of vaccination, the demand for their services was low. It seems that an inadequate analysis of the nature and profitability of the intervention preceded the decision regarding cost recovery (Umali and others 1992). 106 New Technologies in Veterinary Medicine "Best bet" technologies specifically for animal health in low rainfall ecozones, will in the immediate future concentrate on vaccination against major endemic diseases and parasites (which can be very effective, particularly when combined with improved nutrition), and improvements in management techniques to reduce losses of pregnant ewes, lambs and kids. In parasite control, increased attention will be given to genetic resistance. Especially in sheep, the existence of significant genetic variation in the resistance to Strongylosis has been demonstrated (ILCA, 1994). The identification of genetic markers that characterize particular resistant genotypes, should greatly improve selection and accelerate genetic progress. In traditional production systems the use of lay staff (paravets, herders) in the provision of veterinary services will increase. This will increase the demand for easy diagnostic tools and the development of "cow-side" probes and tests, which will greatly facilitate the move towards paraprofessional staff. In vaccine development, the attention will be on cost reduction. This will be through the increased use of multivalent and thermo-stabile vaccines. Recent developments in foot and mouth vaccine confirm these trends. ANIMAL BREEDING, GENETICS, AND BIOTECHNOLOGY Progress in raising the productivity of ruminant livestock in Western Europe, North America, and Oceania has depended considerably on genetic improvement. In some cases (especially in the United States), this progress rests on new, genetically stable breeds of cattle and sheep. Animal breeding has also played an important role in the "Southern Cone" countries of South America, especially in Argentina and Uruguay. Consequently, Bank-funded projects have tended to allocate high priority to genetic improvement in developing regions, especially with respect to cattle. In some projects, the focus has been primarily on importing exotic breeds, in others on cross-breeding of selected local genotypes; often, both approaches have been followed. In order to accelerate progress, efforts have been made in a number of countries to expand or to introduce artificial insemination techniques, and more recently to explore the potential of embryo transfers. Thirteen of the World Bank projects surveyed explicitly provide for genetic improvement of dairy cattle, five for beef cattle, fifteen for sheep breeding, four for goats, and three for horses."8 In six other projects, references are too imprecise to identify the type of animal involved, the breeds, or the breeding objectives. Only seven out of the overall total of 27 projects with livestock components actually included provisions for support of Al services. In others, however (for example Argentina), Al services may already have been in place. The breeds most commonly imported were Holstein and Brown Swiss in the case of cattle. Merinos (aLrnost without exception) for sheep; Murcia and Angora goats, and Haflinger horses. In several projects it was also proposed to upgrade local stock through selection and crossbreeding, although it is not altogether clear whether the program involves only indigenous bloodstock, or is mainly oriented towards crossbreeding of local breeds with introduced exotics. 1s This does not mean that 39 separate projects had breeding components, since in several instances a project included more than one type of animal. The Turkey Fifth Livestock Improvement Project, for example, provides for genetic improvement of dairy cattle, sheep goats, horses, and poultry. 107 While considerable funding has clearly been spent in support of efforts to improve national flocks and herds through breeding, the returns to this investment are difficult to quantify. This is partly because of problems of attribution and separation of the benefits of breeding, feeding, and disease control in the total package of livestock improvement measures, and partly because of the fact that evidence of the impact of the dissemination of improved genetic material among the national herd or even within the animal population of the project area is often hard to find. The main difficulty is that very often the number of imported animals is quite small relative to the total number of locally bred stock in the project area (which may already include exotics and their crosses with local animals), so that when the exotics introduced under Bank auspices are distributed, their impact on this much larger pool is inevitably limited. In Turkey, for example, the dairy cattle actually imported under the four major Bank livestock projects between 1978 and 1984 numbered approximately 18,300, representing only 1.7 percent of the total recorded number of purebred dairy cattle in 1988 (1.05 million) and .04 percent of the total population of 4.62 mnillion pure and cross bred dairy cattle in Turkey (annex table A40). In the case of the Fifth project, the project distributed 1,224 imported purebreds among 17 provinces with a total population of 285,000 exotic purebreds--0.4 percent. The PPAR argues that, when compared with the number of purebred cattle existing and reproducing in Turkey the cattle imports under the Bank-funded projects have not been significant, and the need for imports of live cattle can be questioned in view of the existing number of purebreds and the expansion of the Al program in this area. The supervised credit program could have been carried out without cattle imports. Eventually, over 60 percent of the dairy sub-loans in the two projects had no cattle imports, without a significant negative impact on the milk output by sub-borrowers. The appraisal mission may thus have erred in its judgment on the need for imported cattle, may have been put under pressure by the national government, or have been influenced by a need to have a high foreign exchange element in the project to justify a targeted Bank loan size. In addition, in the case of the Fourth project, the extent to which intensive milk production should have been encouraged in its target production environment of Eastern Anatolia is questionable. Certainly, the plan to introduce 12,500 unadapted, imported dairy animals can hardly be justified under the production, management, service, and infrastructural support circumstances of the Eastern regions. Formulation of the Fourth project within a logical national livestock development strategy would very likely have led to this component being smaller in scale for local dairy purebreds and to an expansion of crossbred enterprises. The foregoing are the most detailed and severe criticisms of breeding practices in Bank-funded projects in any reports reviewed. Similar, if less well-documented comments, are made in PPARS of projects in Morocco and Tunisia, particularly with respect to cattle. The same general tendency, however, to introduce imported breeds in insignificant numbers is apparent with respect to sheep, goats, and horses: for example the Fifth Turkish Livestock project planned to import 130 mares and 30 stallions of the Haflinger Breed to a country with some 700,000 horses and a rapidly declining horse population. Neither the rationale for this, nor for the choice of breed is explicit. The choice of breed is not clearly stated for imported exotics in several other projects. Generally, the number of recipients of exotic stock is also small. In Turkey, the Fourth and Fifth Livestock projects together covered 47 provinces having a total population of around 413,000 exotic dairy cattle and another 1.5 million local dairy cattle. The two projects combined targeted only about 3,150 farmners to receive exotic cattle, and a majority of those were located in remote Eastem Anatolia. This was as part of a scheme to establish 2,000 specialized dairy farms there with on imported, high-yielding cows, an idea condemned as unsound by the PPAR in light of conditions in that region. In the case of Merino sheep, the number of loans to farmers in the Fifth Turkish Livestock Project was only 628, although the loan covered 32 provinces with around one million Merinos and 30 million sheep in all. Some 59,000 animals were distributed. The Angora goat component involved 554 loans, with 108 5,300 does distributed in provinces with a population of some 1.7 million Angoras and a total of 7 million goats altogether. A related set of questions not adequately dealt with in most PPAR's or PCR's concerns the recipients of these imported animals, their selection, and the performance of their cattle, sheep, or goats. In some projects, for example the Northwestem Tunisia Rural Development Project, the farmers will be eligible to receive bulls when their herds have reached a minimum of fifteen cattle and they can feed and house the bull properly, insure it, use it on neighbors' herds ,and register its activities. This implies rigorous monitoring of the conditions under which the bulls are kept, and their performance. There are indeed serious obstacles to success for exotic cattle imports. Exotic animals and their progeny have often proved ill-adapted to the climate, low plane of nutrition, poor management, and lack of health care on farms. Low levels of literacy, inadequate capitalization, poor communication, lack of good supporting infrastructure (veterinary, extension, Al, credit), and difficulties of marketing produce, severely constrain the ability of farmers or herders to provide better conditions for high-potential stock. If such animals go to government stud farms for breeding they may be better fed and managed, but unless those farms are linked to an efficient performance recording and Al system the end result in terms of their impact on upgrading the national herd is not likely to be much better than that of animals handed out to individual private farms. The risks of failure are highest with dairy cows imported to establish purebred milking herds, such as attempted in Eastem Anatolia. Exotic dairy animals are larger and need more feed than local cows, yet are unsuited to the natural pastures in most of the region. Thus they usually have to be kept in confinement, with minimum or zero grazing, and may require irrigated forage to support them, supplemented by imported concentrates. The opportunity cost of dairy production under these conditions is questionable except in a highly protected market. The chances of replicating results widely within most countries are low. Yet importation of dairy cattle continues to play a large role in Bank-funded projects in WANA. Both Glenn (1988) and Cunningham (1992) are critical of this approach. Noting that most modern dairy improvement in WANA has been based on imported Holstein-Freshen cattle, Cunningham states, Investment in local improvement programs for these stock is probably not justified, particularly where Al gives continued inexpensive access to breeding programs elsewhere... However where smallholder dairy is being encouraged, which implies a substantial dependence on home-produced forage, these imported genotypes are seldom the best choice." Nevertheless, he is somewhat skeptical of repeating in WANA the potential gains which have been achieved in India from crossing unimproved native breeds with imported European breeds (Bos Taurus) because local breeds in WANA are also Bos taurus, so the gains from heterosis are liable to be smaller (page 72). Three major institutional obstacles impede progress in exploiting the potential for breed improvement through cross-breeding and selection in WANA, and with sheep in northwestern China. The first is the general absence of any organized system of performance recording the importation and distribution of improved ruminant livestock to private farmers or to government stud farns or breeding cooperatives trying to upgrade local stock. Without adequate time-related records of feed consumption, management, growth, reproduction, health, yields, sales, prices, and costs, the selection of "improved" local animals by tracking their perfornance and that of their offspring is extremely difficult. Some attempts have been made to induce recipients of imported or improved stock to keep records. The Turkish Second Livestock Project (in the context of feed lots) developed a system based on five different cards showing feeds, feeding regimes, health controls, a project evaluation card, and a feedlot visit card showing details of 109 visits by the technicians. The attempt to persuade the fatteners to keep records failed because of their lack of cooperation and a lack of thorough supervision by the technicians. The second factor impeding breed improvement through cross-breeding is the inadequacy of artificial insemination services in most countries, and the resistance of traditional stock owners to the concept of Al. A related problem is the practical difficulty of applying Al to sheep under the extensive systems of husbandry that are dominant in WANA and in Northern China, where telephones and roads are often poor. This is not simply a matter of getting both access to the scattered flocks and agreement from their owners to treatment (as is the case for vaccination programs), but of getting to the flocks at the right time in the sheep's ovulation cycle, and if necessary repeating the insemination. In WANA there are probably 80 million breeding ewes. This puts great demand on the efficiency and mobility of the service, and in the absence of any charges for the work, most governments do not provide it with sufficient resources. Finally, the inadequacy of the national extension services with respect to animal production is a limiting factor. Genetic Potential of Local Breeds in Semi-arid Regions An important policy issue with respect to the emphasis on importation of new blood versus reliance on selection and improvement of local breeds is the quality of the local gene pool. Cattle In WANA, the genetic potential of the local cattle is poor with respect to milk production, although local animals are more tolerant to poor feeding and management practices and diseases. For meat, local breeds are capable of responding well to a higher plane of nutrition and gaining weight fast up to a point where they put on too much fat, even though their carcass weight at slaughter is low. This makes a case for importing more pure bred bulls or semen for cross-breeding with selected local animals, but generally not for importing exotic breeds to form pure bred dairy herds. Importation of semen is cheaper than importing livestock, although its success might be affected by poor performance of the local AI service, and the resistance reported by farmers to its use, could make it a high risk venture. Consequently several Bank-funded projects have chosen to use natural breeding until Al services have been strengthened and farmer education concerning AI has had a chance to make an impression. Sheep. While there may be a lack of suitable genetic material for improving dairy cattle, this is not the case with sheep. Mason (1969) characterized almost 360 breeds or local races of sheep including 46 from WANA. In addition there are analogs (various types of Merino, Barbary, Karakul) from neighboring countries such those in the former Soviet Union, southern Europe, and from similar climatic zones in more distant countries (Australia, Argentina, the United States, Sub-Saharan Africa. Other breeds were added later. In sharp contrast, however, only eight breeds are listed from China (all from Mongolia), and only five from Latin America (of which one is a Merino type). The other four are hairless tropical breeds. The numerous breeds listed for WANA countries provide a wide range of choices of size; meat, milk, and wool production, and conformation. A majority, but by no means all, have a fat tail or rump. None has fine wool, however, presumably because no Merino is identified as coming from the region, even though numerous local Merino breeds are listed in areas that abut on WANA. Mongolia and northwest China, surrounded by ex-USSR countries, are only shown as having one fine-wool , non fat- tail sheep, but it is not described as a Merino. Imported breeds include Merinos, Baritone du Cheer, Lie de France, and Suffolk sheep. The latter, next to the Merinos, seems to be among the most widely used exotic breeds for crossing purposes. 110 Goats. Goats receive much less attention than other ruminants in Bank-funded projects, although in two cases the importation of Murcia goats from France is proposed. What is remarkable about this project is that although one project was operational in 1981 (Morocco, Middle Atlas), and the other in 1990 (Northwest Tunisia Rural Development), the animals could not be imported in either case because of an outbreak of disease in their country of residence. On reflection, perhaps this was a blessing in disguise! Mason (1969), classified 74 goat breeds of which 22 originate in the Middle East or North Africa- 30 percent of the total and outnumbered only by the 24 breeds of West European provenance. Several of these breeds are prolific millers and suited to tethering, for example. Maltese goats from Tunisia, Nubbin goats from Egypt, and the Damascus goats from Syria. Several of the breeds he lists have similar names to some of his sheep breeds, probably indicating that they come from the same area. The main characteristics include: * Relatively short gestation and lactation periods (circa 150 and 60 days respectively) * The possibility of obtaining three lambings in two years * High levels of profligacy (multiple births) in some breeds * Considerable heterogeneity among breeds which offers the possibility of substantial hybrid vigor from cross-breeding or from the creation of new synthetic breeds through interbreeding. These characteristics not only serve to shorten the period required to achieve a response to selective breeding, but also provide the flexibility to increase flock numbers after a drought or to respond rapidly to market. All Merino Fl crossbred ewes had a higher lamb rate than their respective local breeds; this is attributed to heterosis and the better maternal capacity of the local breeds. This also suggests that a mating system to produce FIs might be a better strategy than attempting to develop a synthetic for the following reasons. * Merino crosses with local sheep in Egypt produced leaner carcasses with better conformation ,while Suffolk crosses with locals produced heavier carcasses with more fat, more prime cuts and better conformation. * With few exceptions, exotic breeds and their crosses had a heavier fleece. Crosses with Merinos produced finer fibers, less kemp, and a finer secondary/primary follicle ratio, although with a higher variability of fiber thickness than local breeds. The latter generally had longer wool staple and fiber, with crosses being intermediate. * The local breeds have better adaptive mechanisms for living and grazing under a warm climate than imported breeds. Studies in Tunisia show that changes in management, feeding of ewes, and dates of lambing, of local sheep breeds can significantly influence liveweight of ewes, proflicacy, milk yields, lamb growth rates, estrous, and fertility. Work in Morocco shows large differences in conception rates according to the date of mating, with June matings considerably superior to those of December. An attempt to use hormonal treatment to reduce the interval from lambing to first estrous and achieve three lambings in two years gave a reasonable conception rate, but unsatisfactory levels of multiple birth mortality due to individual variability to PMSG (Lahlou-Kassi 1987). ill * More analysis is needed of genotype-environment interaction in WANA. An overall conclusion from a review of past and ongoing sheep and goat breeding work in WANA is that the potential of local ger-otypes has probably been under-estimated. This may be because field observation of animals kept under poor management on small farms does not provide a fair assessment. For example results from the Cape Serrat project in Tunisia show that, although the local type was not as productive as the Boer goat, it outstripped the European dairy goats in weaning performance, in biological and in economic terms, because of its greater fecundity, shorter kidding interval, and lower mortality. In addition, the local goat was found to be more salt tolerant than the imported genotypes, being able to maintain weight while drinking sea water at 3.5 percent salt concentration. (Steinbach 1987). Breeding Strategy TheWorld Bank technical paper "Sheep and Goats in Developing Countries" (De Boer and others 1983) identifies a number of criteria to be considered when undertaking a genetic improvement program. These include the production environment and projected market requirements, the relative technical and economic advantages of attempting to change the environment or changing the genotype and the characterization of the local animal population for the traits which determine productivity and efficiency. A suggested list of important factors that need to be considered is shown in table 6.3. Four strategies for genetic improvement are listed in the technical paper. These are: * Selection, basically involving progressive breeding and selection within an established local type so as to enhance favorable characteristics of the population such as proflicacy or disease resistance * Crossbreeding, where two or more established genetic populations are intermated to capitalize on hybrid vigor and to combine complementary traits * Upgrading, which is introduction of a new breed by breeding the improved male purebreds to the local breed and to successive generation of topcrossed stock * Introduction of improved breeds for direct production rather than primarily for crossbreeding or upgrading of local populations The latter strategy, introducing improved breeds, has generally had limited success when followed in Bank or other projects, as prevailing feeding and management conditions are frequently too poor for the imported animals. Except in very special situations therefore, it is better to introduce proven breeds from other developing countries rather than temperate genotype. Embryo transfer may be valuable in this respect. In general, the rationale behind the choice of imported sheep and goat breeds in Bank-funded projects is not very well articulated, nor is the breeding strategy clearly defined. In particular there is relatively little attempt to characterize the local animal population along the lines suggested above. -i ........... ... .. ...... .... ... .. ... . ... .. .. 9 1 00 > ......... .. .... ... ...... ....... (JQ .. ......... ...... ...... :X X ...... ...... . ... ... ...... ... ......... ... 0 . ....... . . ...... .. . ... ...... .. .. . ... . .. ...... ... ... . ...... .... . .... ... ............ . ..... ::X. ....... .. . ...... ............ ... . ... .. ...... . . . ...... .... .. ....... .. .....:.. ..... ........ ..... ..... ............ ..:.t -... .... . ...... ....... . ...... ::: ... . . . .... ........ . ..... ....=. ........ ..... . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . ..... ...... . ..... .... ... . ........... . . . . .. . . . . . . . . . ... . . . .. . .. . . . . . . . . . . .... ..... ...... .. ... ..... ..... e4 ....... .... . .... ..... .... .. ........ .... .. ..... .N-4 ... . .. ... ...... ... . ... ..... ...... ... ......... . . ...... ....... ....... .... .... . ............. ............. ..... .. ....... rA . .. ... .... x .. ... . ................. . .... .......... ... .... . . . ....... ......... .. ..... .. ..... ... .... . .... ... ............ .... ... . .. ...... . . .... . ........ ....... ... .... .. ... . . .. .... ... ....... ..... ........... ...... .. . .. . . . . . ............. .... .. .. .... ..... .. ...... ........ .... ........ ....... ...... .......... ............. .. . . tA mm .:.  .: : : : : ::::... : :...... ........ ... . ..... ... ...... x ..... .... . .... ......... .. ................. ...... .... ...... ...... ... . .......... . . .......... . . .. . ........ C D  ::.: .. . . .. ... : ::: ..... . ... . .... . ... ............. .. .. ... . . . . . . .. . . . . . . . . . . . ... . . .. . . . . . . . . . . . . . . . . . . . . .......... ......... .... ............ .... . . .. . .. .. ........ ....: ....... ..... .. ....... ... :: ... :-: : .. ... ... . ..... ....... .. . ...... ... .. .. .. . ..... ................ ..... . .. ............ ...... ..... ..... . ............. . ... ..... . ....... . t.4 X ... ................ .. ... .......... r 0 r-L ....... .. . ... ... ..... .. .. .. . . .... .. . ... ... ......... .... M. 4r.) ....... .... . . .. .. ..... o o 0 tz. ... . ... ::] O..:. . : ?0. El C.. x.. ...... ..... .... .... O" Ea 4rA 17, ... W.. 0:.,n x. j. .. . r4- 0. -4 rA ci : g ... ... ... . ..... . Z . 0 0 ...... . ......... ... ...... . . ..... CD 0 Cb 46: CD X.. co .......0 .... .. CD Im P) CD 0 ... ... ...... P) FD ... ..... 0 . ..... .. . ... . .... ... ... .. ....... .... ....... . . .....0 V. ... 25. . .... ...... -x:. . .. . .... .... CD .... . .'oz:: ... V..... 1* . . ... ... ..... . .. . ...... CD ... .. ...... ... ...... 0 . ....... ..... .CV . rA N cl . .. . ....... tz CD ....... F rA =r. :. E. -t Cj t3 OQ. P, QQ P I Pl CD 0 O = (W tCD CD EL ... X.. .... ... ...... CD: .... ..... . ... . . . . . . . .. . . . . . . . . .. .. . . . . . . ... . . . . . . . . . . ... .... .. ........ . :: ::..9.... .. . . ... ... . . . . . . ........ . .. ... . . . . .. . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . = I .. ......... ...... O .. r"A .............. . ... .. . ...... . ....... ............. CD (JQ W KP ....... . .. .... . .... .. .... ..... .... ... ..: : ........ ............... . ..... ........ ...................... ..... . .... &; :: ...... ..m: .. .. .............. .. .... .. ..... . . . . .. . . . . . . .. . . . . . . ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . r L ... .. . . .. ............ ..... :.:x : .... ...... ....... .... . ........... .......... .. ........ . ............. . . ........ ............... .... ...... ..... ....... ... . ...... ...... . .... .... .. ..... ......... . .. .. . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . .. .. . . .......... ... ......... 0 QQ - ::x .... .. . .. ... .... Ot I ... . - . .... ...................a ... ....... .. c y - a ....... . ...... ......... ...... ..... . ......... . ... . ...... . ......... .. . . ........ . . . . ..... ... ....... .. .: ........ ... ............ r D . . .... ........ ....... .. .... ... .. X . ...... ... .... ........ .......... ..... ........ ..... ........ ........ .. .. . ..... ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......... ..... . .... ........0. .. ... .. .... .... .... ........... . .............. m.: O" - : :. - :: . . ... . ..... ...... ..... .. ....... ....... .............. C D ........... .... ............. ....... ... O Q .. . ... .... C D 5  ............... . ... ........ . .... .... ..... . ... ...... .. ........ .... .... ....... C ) C D ............. . .... .. . ...... .......... ........... . ....... . . .......... ....... ..... ............ ...... .. ... ....... . .. ... ......... . . ..... ... ...... .-..... .... ... ... ... ... ....... ... ... .. . ......... . .............. . . . ........... .. . ...... ....... ......a .. .......... .. . .  :: : : ... . . . ......... I.. . .O .: .... . . . . ....... ....... ... . ................. ...... . ........... ...... ... . ....... .. ... .... ..... ........ ........ . ........ .... ...... ..... 11 .... . .. ... .... . .............. ... ...... . .. ...... ........ ........... ....... .... . . . . . . . .. . . . : . . . . . . . .. ..... . . .. . . . . ........ ... .. . .....m.. ........ ...... ........ .. .......... ... ..... ........ .......... Irc ........ .-W .. .............. .... ... .......... .... .... ... ......... ... . .. con ... ... .... . ... .. ..... ..... ..... ......... .. ....... CD co) E; ::... p, . ............. ....... ............. ...... ..... ........ . . .... . . . . ....... .. .. .. .r ... . . .. ..... . .. .......... .. . ...... ........... . .. . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . ... . . . . . . . . . . . . . . . . . . . . . ... .. . . . . .. .. . . . . ........... . ..... .. . ... ........ ... . ..... .. ..... ..... . .......... ..... ...... ... .. . . ....... .. .... . ... .... ..... .. . . .. . ...... ..........I ........... ......... ........... ......... .... ...... ..... . .. E'r CD CD . ................. . .... ..... .... . .... .. C J C D ......... ..... . ...... ... . .... ..... .... .... ....... ............. ..... . .. .. ...... . ................ . . ............ . .. ... .. .... ... .. ...... ... . . . .. ........ .... ... ... .. . ..... . ............ .......... ... .. ..... . ...... .... .. . ...... ... .... ..... ..... .. .. ... . . .. ....... . ..... ....... .... . ... . ..... ....... .... ....... ......... ... . . ..... . .... .. .. ..... .. . . .. ..... .. .... ... ... . .. ... ... . ..... .. .. . ...... . .. C) :::: . . . . . ........ . . . . . . ...... . ........ .... ...... .... ... .. .. ... ............ ......... .. .... ..... .. . ... ... .::X: . ..... ... ... . m. m... O Q :: ..... .. .  .... m : : ...... ...... ... ........ .......I...... ...... ... ... . .. ..... ........ .. ... ....... ..... . . ..... .. .. ... ..... ....... .. ...... .... ...... .......... .... . ..... ..... .... .... ..... .... . .. m : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .. ..... . . .. ...... .. ............. ... .. ... ... C 5 C'D 113 Artiflcial Insemination Artificial insemination has undoubtedly played a key role in genetic improvement in the developed countries, especially in upgrading national dairy herds. It has had mixed and more limited success in developing countries. The need to give priority to establishing a national Al program has been questioned by several Bank-funded projects and other authoritative sources in WANA countries, including Egypt, Tunisia, and Turkey. In Turkey, the PPAR of the Fourth and Fifth Livestock Project states that the higher levels of production that could be achieved with an Al program would be in any case limited by feeding or other management practices. Thus, the priority which should be given to establishing an AI service in Bank-funded projects requires very careful thought. The issue is not whether AI is a useful technology, but under what circumstances it is likely to be an appropriate one. The factors which need to be considered in reaching a decision include: The gap between the current and production potential per animal. When physical factors would allow a higher level of production, Al is more easily justified than when production is mainly constrained by the environment. * The nature, composition, and size of the national herd. Dairy cattle merit higher priority than sheep and goats, where genetic potential is already high. * The reproductive physiology of the target livestock population. Experience in Egypt and South Asia indicates that conception rates and the impact of Al on performance of buffalo are both lower than is the case with cattle. Greater difficulties may also attend attempts to improve sheep through AI than with cattle or goats. Close herding or tethering of small ruminants may increase the chances of successful Al. * The location and degree of dispersion of the animal population. The task of insemination and the rate of success is likely to be higher and the costs of service lower where animals are relatively concentrated and easily accessible. The attitude of the local people towards AI. Several projects report skepticism or resistance of farmers or flock owners either to the principle of Al or to its value. The IFAD project appraisal report for the Egypt Livestock Intensification Project covering 5,411 km2 of densely populated middle Egypt (1,220 people/km2) where there are 139 clinics (4 fixed points and 14 mobile job lines--a series of villages along main roads to which veterinarians travel at set times), states that the project failed to establish a viable system, even in an area were it should theoretically have been successful. * The budgetary and government incentive systems which may affect the efficiency and cost- effectiveness of the AI service. If a government is unwilling to provide adequate financial support to an Al service, especially in the critical period between its inception and the achievement of demonstrable benefits among the livestock owners, AI should not be promoted. The issue of cost recovery is raised in several project reports, and there is evidence from Syria, Egypt, Morocco, and Turkey that dairy farmers are willing to pay a modest fee. While initially Al may be provided free of charge, after a relatively short period partial cost recovery (at least) needs to be introduced. Experience in Morocco, with 114 a decentralization of the service to dairy cooperatives and a payment of the cooperatives' operating costs, has been favorable. Cost recovery improved the administrative efficiency and had a direct positive result on the conception rate. The availability of adequately trained staff to administer the program and deliver Al to herd owners. Inseminators should be specifically trained both to undertake their work and to answer questions of users or potential users. The structure of the AI system. In tracing the progress of modem animal selection programs since its inception in the 1930s, Cunningham (1992) contrasts the evolution of the AI-cum-breeding services in the United States and Europe. The former were dominated by a few large commercial breeding cooperatives and cooperatives, and the latter by producer cooperatives. Thus, he sees investment in livestock improvement in two broadly separate contexts: as a company investment with all the attendant market risks, and taxes, or as a public service-type investment, with artificial insemination usually carried out on behalf of a producers' cooperative. The same investment might be appraised quite differently in these different circumstances, with lower discount rates being appropriate to the public service-type of approach. * The adequacy of monitoring and evaluation of the Al service as a whole and its results, including both the performance of the selected bulls which provide the basic semen, and their progeny; and the financial and economic returns to the investment in the service. In general, as has been noted in other sections of this report, this is a crucial area of weakness in Bank and government services. * The demand/supply/price structure for livestock products. The benefits of farmers' investment in Al can be greatly affected by the nature of consumer demand on local and international markets for more or different quality animal products and the impact of changes in demand and supply of those products, (domestic and imported), and government price policies towards them. Distance from markets, market structure, and infrastructure also play a role in determining what producers receive for their products. The interplay of these factors largely determines demand for Al and the financial and economic rates of return to expenditure on Al services as a whole. Potential of Biotechnology in Animal Breeding The initial promise of biotechnology was over-estimated and oversold, and there have been important changes in the nature of demand for agricultural products since 1980 as well as in the economics of production. Further important changes are likely to follow the successful outcome to the current Uruguay round of GATT negotiations, which will affect the goals and returns to investmnent in biotechnology. Cunningham (1992) points to the paradox that while the major challenges lie in the developing countries, biotechnology has so far proved less profitable in those countries than in the developed world. Biotechnology advances with important implications for animal breeding and genetic improvement include 115 embryo technology, gene manipulation and transfer, and hormone and immunological treatments (including new diagnostic tests). Embryo Transfer Over the last twenty years, the techniques for recovering, storing, and implanting embryos have been perfected. It is now possible to superovulate cows and to recover nonsurgically on average four and up to thirty embryos at a time. Freezing in liquid nitrogen and thawing are now routine operations, with relatively small effects on viability. Implantation with an insemination gun is also now a nonsurgical routine. It has been advocated, in Bank projects because importing embryos is usually cheaper than importing live animals. There are several benefits of increasing the reproductive rate of selected cows via embryo transfer. First, the contribution of genetically outstanding cows to the breeding program can be increased. Specifically designed breeding schemes with multiple ovulations (MOET) to take advantage of the increased intensity of female selection, combined with increased generation turnover, can double the rate of genetic change over that achievable in a conventional selection program. In this respect, open nucleus breeding schemes, which rely heavily on selection at village level, supplement by breeding of elite animals on large (state or private) farms have generated high rates of genetic improvement, especially in sheep and goats. Second, the twinning rate can be increased by either implanting two embryos, or by following a normal insemination with the transfer of a single embryo. Similarly, the birth weight of young stock can be increased by implanting embryos with specialized genotypes (for example, pure beef breeds) of higher value into cows of lower potential (for example, undeveloped local breeds). This is, however, only advisable if the feeding, management, and veterinary care are sufficient. Finally, embryo transfer allows the rapid expansion of rare genetic stocks and new breeds. The following new and related developments could provide additional benefits. * Sexing of embryos (now possible) or semen (not yet possible, but being actively researched) could increase selection intensities further, and could permit greater specialization in beef and milk production functions of a dual purpose population. * In vitro fertilization, now showing great promise in research studies, has enormous potential for increasing the benefits achievable in breeding programs. Its main impact will be to reduce the cost of embryos, and therefore to make embryo transfer techniques economically feasible on a wide scale, and in circumstances where the cost of present techniques cannot be justified. * Embryo splitting to produce identical individuals has been possible for several years and has considerable application in improving the efficiency of research studies. * The production of multiple copies of an embryo via embryo cloning (nuclear transplanting) is now possible. The main limitation is the supply of host oocytes. However, the current expectation is that repeated cloning will be a commercial reality within a few years. 116 For each of these existing or potential reproductive technologies, the uptake will depend on the tradeoff between costs and benefits. The cost is measured as the cost per cow bred and the benefit can most generally be measured as the bonus in genetic improvement that is achieved over the baseline technology of natural breeding and selection. This will include the benefits of sex determination. Hormone and Immunology Technology The most dramatic product of biotechnology currently available is recombinant growth hormone, or somatotropin (BST). When administered regularly to dairy cows during lactation, it increases milk output by 15 to 30 percent, and also increases the efficiency of milk production. No problems of undesirable residues in milk are apparent, and this technique is coming into general use in the years ahead. The main technical problem still to be resolved is an efficient delivery system, since current use requires regular injections. There are, however, important socio-economic side effects, since it has been shown in practice that BST treatment yields highest returns on well-managed high technology dairy farms, and that in technologically advanced countries its widespread use on such farms may drive less efficient producers out of business. Vaccines and probes are two aspects of immunology that are benefitting from biotechnology. The opportunity to develop new vaccines via genetic engineering could help to overcome not just against viral but bacterial diseases as well as a variety of parasitic agents. In addition, tailor-made vaccines for greater effectiveness and less risk may be possible. This area holds considerable potential for increasing control over endemic diseases like foot and mouth disease, rinderpest (Mariner and others 1994), theileriosis and possibly trypanosomiasis. Because of the difficulty of operating conventional improvement schemes in developing countries, the use of DNA probes to diagnose diseases ("cow side diagnostics') and to genetically identify superior individuals for particular traits, like resistance to a specific disease, could become important (Beckman 1988, ILRAD 1994). Future Perspectives The immediate impact of biotechnology on animal production in arid and semi-arid zones will be limited. Cunningham (1992) states: Gauging the future pace and utility of biotechnology in both the developed and developing world is difficult. In the past ten years interesting technical developments have been achieved more rapidly than was generally predicted: successful gene transfer in animnals, in vitro fertilization, cloning of embryos, and sexing of semen. With the exception of diagnostics, each development seems to uncover new layers of complexity in application, and as a result pushes the prospect of practical application further into the future. A good example is in the case of embryo manipulation, which took a decade of effort to overcome the initial technical problems and to achieve a certain level of success in nonsurgical transfers. Research has now reached a plateau at a level where its cost inhibits widespread uptake. The emphasis has therefore shifted to in vitro fertilization as a cheaper way to produce embryos. Beyond that the prospect of cloning, with its higher genetic benefits but lower costs, beckons (p. 94). 117 Bearing this in mind, much of the promise of animal biotechnology is equally applicable in the developed and developing world. However, the experience is that, as with many aspects of advanced technology, the capacity of small-scale producers in poor countries to use it is often limited. Conservation ofAnimal Genetic Resources Unlike plants, no global center for conservation of animal genetic resources currently exists. There are also no specialized international centers for genetic improvement of specific livestock with identified responsibilities for maintaining germplasm collections as is the case with crops. The livestock research center of the CGIAR, located in Eastern Africa (ILRI), has not yet taken a major role in genetic conservation. There us realcause for concern, for two main reasons: * As mentioned previously, many of the local breeds of domestic livestock of sheep and goats in developing countries, have not been adequately characterized. * There is a possibility that the advent of new techniques and the wider use of Al will force out some of the valuable characteristics of local breed and/or result in their eventual disappearance; this has already happened with some of the old landraces of sheep, cattle, horses, and poultry in Europe. It is imperative to manage and conserve these unique genetic traits for future use in livestock improvement, but this implies evaluating the indigenous breeds under their prevailing environmental and management situations, and disentangling genetic from environmental or non-genetic factors. Modeling may be useful in this respect; the Texas A & M Systems Group has developed a dynamic interactive sheep production model that simulates responses of specified genotypes to different nutritional and management practices, for example responses of sheep to fluctuating rainfall patterns on arid rangelands in Northern Kenya. It has been suggested that in order to identify the relevant genetic parameters (such as heritability, repeatability, genetic correlation, and non-additive effects) genotype reservoirs should be established in the different ecological zones of a region (Steinbach 1987). Autochthonous genotypes to be evaluated would be selected following a survey of the available breeds with actual evaluation carried out in the field under variable husbandry conditions as well as on stations with standard conditions of adequate management to investigate the production potential. Those situations might be focal points for producing performance tested, disease free, seed stock or embryos for wider use. Agreed international standards of record-keeping would be needed. Such a program would probably require long-term support, perhaps under the aegis of an international organization such as FAO and/or an appropriate IARC. In this context a recent report from the U.S. National Research Council (NRC 1993b) is timely. It focuses on the management and conservation of the genetic resources of domesticated animals that have potential economic value. Strategies for conservation, management, and application of appropriate technologies are presented. 118 Conclusions Genetic improvement of ruminants may be a best-practice technology if it enables animals to perforn more efficiently within the constraints imposed by their environment. Genetic improvement designed to increase yields of meat, milk, or wool is not advisable if it depends on importation of purebreds or results in an animal population requiring a higher plane of nutrition and management, and health care than can realistically be foreseen within a decade. Insufficient attention has been paid to characterizing and evaluating the potential of local genotypes, especially with respect to sheep and goats, There is a tendency in Bank-funded projects to rely on importation of Merino sheep without clear specification of the reasons for this, and to ignore the attributes of local breeds and the potential for selective breeding and/or crossing among them. Artificial insemination is a valuable tool, but one which should be used with discretion. It can play an important role in accelerating breed improvement in dairy cattle if other factors are not severely limiting their performance. It is not appropriate to large-scale genetic improvement of sheep or goats both because of the high cost and practical difficulties of implementing such a program and the environmental constraints. Importation of purebred animals and their distribution to small numbers of farmers within large animal populations, and without workable means of monitoring their future use or recording their performance is wasteful (if not futile) and should be rigidly controlled. Instead, priority should be given to improving the nutrition, management, and health of livestock; and to building the institutional capacity all of which could make eventual large-scale genetic improvement programs under government auspices more fruitful. Such measures include strengthening support to research and characterization of local breeds and Landraces (including DNA mapping); developing effective monitoring, evaluation, and recording procedures to facilitate breeding efforts both at government stations and through on-farm research establishing a self-financing Al service for cattle where sufficient demand exists, and improving extension advice to herders with strong backing from well-qualified specialists. Biotechnology holds great promise for the future with respect to livestock improvement, perhaps most particularly in the use of embryo technology, and in immunization and vaccine development. At present, however, it is mainly a research tool; embryo transfer (although a practical possibility) is too expensive for use in large-scale breeding programs (for example as a substitute for Al); and BST is proving to be most valuable in herds with high standards of management. Neither technique is therefore appropriate to situations where inadequate management, nutrition, and health services are major constraints to increasing ruminant productivity. 119 Table 6.4. Possible Biotechnical Solutions to Current Problems in Developing Country Livestock Production :::PROBLEM POSSIBLE BIOTECH SCALE OF PROBABLE. M |-_______',___ SOLTON ' ECONOIC -PAC CO A,imal poultry, fish New vaccines' Laxge.' r' -, sew e,wdiagnostics , ' Moderate S'ort orquality f forages Microbial treatment of . - : .. forages Modere. , -' Me::'-' : Modifications of ' Moderatef:: rumen mcroflora Large: ,on:g ~~~~~~~~~~~~~. . ---. . . -. . . . . Genetic improvement of forages and their ._ __ ._ ._._ .._ .symbionts . M oderate. .......... .... Difficulty of Selection in nucleus implementing . herds, using enbr-o transfer, sexing Large, Mediu Selection progranms Use of RFLP markers to assist selection Moderate: Med i ~~~~~~~~~~~~~~~~~~~~~~~. . . . . . . . . . . . ......... . : D,cultt of ' , : ,'-Use of In vitro naintaiiing dairy cattle fertilization, embryo performance ater F1 transfer, and sexing to cross give continuous F is- Large:: ong. Cost and Use of embryo transfer environmental challenge to import embryos to imported cattle instead Limited Srt. Need for increased Use of rBST and rPST efficiency in intensive- in dairy and pig systems. production. Larg Shortg RFLP = restriction fragment length polymorphism, or direct DNA typing of individuals. rBST, rPST = recombinant bovine or porcine somatotropin (genetically engineered growth hormone). Timeframe :Short= now or <5 years Medium =5 to 10 years Long= over 10 years. Source: Cunningham 1992 p. 97. References Abdel Monem, M., M.J. Ryan, and A. Azzaoui. 1990. "Economics of N on Dryland Wheat in Morocco: Soil, Rotation, and Season Effects." Rachis: Barley and Wheat Newsletter. Vol. 9, No. 1 (January). Aleppo, Syria: ICARDA. ACSAD/AOAD (Arab Center for the Studies of Arid Zones and Dry Lands / Arab Organization for Agricultural Development). 1985. Evaluation of Present Status and Potential Development of Animal Feed Resources in Arab Countries. Khartoum, Sudan: AOAD. AMIMA (Association Marocaine d'Industrie de Materiels Agricoles). 1993. Estimates of Sales of Agricultural Equipment (unplublished data). Available from: 23 Bd. Mohammed Abdou, Casablanca, Morocco: AMIMA. Bansal, R.K., and 0. El Gharras. 1993. "Progress and Prospects of Agricultural Mechanization in Morocco." Paper prepared for the World Bank. Settat, Morocco: CRRA (Centre Regional de la Recherche Agronomique Settat). Mimeograph. Bahhady, F. 1987. "Sheep Husbandry Practices in Syria and the Possibilities of Improvement." In A.W. Qureshi and H.A. Fitzhugh, eds., Small Ruminants in the Near East. Vol. 1, Selected Papers. Rome, Italy: FAO. Barbero, ., and Queezel. 1981. pp. 69-74. Beckman, J.S. 1988. "Genomic Genetics: An Essential Link between Conventional and Transgenic Breeding." Proceedings of FAO Expert Consultation, Havana, September. Behnke Jr., R.H., I. Scoones, and C. Kerven (eds.). 1993. Range Ecology at Disequilibrium: New Models of Natural Variability and Pastoral Adaptation in African Savannas. Published by the Overseas Development Institute in association with the International Institute for Environment and Development. Nottingham, U.K.: Russell Press Ltd. Belaid, A., and M.L. Morris. 1991. Wheat and Barley Production in Rainfed Marginal Environments of West Asia and North Africa: Problems and Prospects. CIMMYT Economics Working Paper 91/02. Mexico,D.F.: CIMMYT. Benbelkacem, A. 1988. "Current Status of Crop Production in High-Altitude Areas of Algeria." In J.P. Srivastava, M.C. Saxena, S. Varma, and M. Tahir, eds., Winter Cereals and Food Legumes in Mountainous Areas, Vol. No. ICARDA-136 En. Aleppo, Syria. Binswanger, H.P., and G. Donovan. 1987. Agricultural Mechanization: Issues and Options. World Bank Policy Study. Washington, D.C. Boers, T.M., and J. Ben-Asher. 1982. "A Review of Rainwater Harvesting." Agricultural Water Management, Vol. 5. 122 Bohn, A., and D. Byerlee. 1993. "The Wheat Breeding Industry in Developing Countries: An Analysis of Investments and Impacts." Part 1 of 1992-93 CMLM7YT World Wheat Facts and Trends: The Wheat Breeding Industry in Developing Countries: An Analysis of Investments and Impacts. Singapore: CIMMYT. Boutonnet, J-P. 1989. "La Speculation Ovine en Algerie: Un Produit de La Cerealiculture." Serie Notes et Documents, No. 90. Paris, France: INRA (Institut National de la Recherche Agronomique). Bruins, H.J. 1986. "Desert Environement and Agriculture in the Central Negev and Kadesh-Barnea during Hisotical Times." Ph.D. thesis, Agricultural University Wageningen. Nijkerk, The Netherlands: Midbur Foundation. Buddenhagen, I. 1990. "A Global and Regional Overview of the Role of Legumes in Farming Systems." In A.E. Osman, M.H. Ibrahim, and M.A. Jones, eds., Developments in Plant and Soil Sciences: The Role of Legumes in the Farming Systems of the Mediterranean Areas. Dordrecht, Netherlands: Kluwer Academic Publishers. Byerlee, D., and D. Winkelmann. 1980. Accelerated Wheat Production in Semi-arid Developing Regions: Economic and Policy Issues. CIMMYT Economics Program Working Paper 80/2. Mexico, D.F.: CIMMYT. Cameron, D., and P.A. Oram. 1994. "Minimum and Reduced Tillage: Its Use in North America and Western Europe and its Potential Application in Eastern Europe, Russia and Central Asia." Report prepared for the Europe and Central Asia, Middle East and North African Technical Department and the EMTAG (Europe, Middle East, and North Africa Technical Department - Agricultural Division). Washington, D.C.: World Bank. Mimeograph. Cameron, S., K. White, and E. Dubbelde. 1991. "Sunflower." In R.S. Jessop and R.L. Wright, eds., New Crops: Agronomy and Potential of Alternative Crop Species. Melbourne, Australia: Inkata Press. Carr, S.J. 1989. Technology for Small-Scale Farmers in Sub-Saharan Africa: Experience with Food Crop Production in Five Major Ecological Zones. World Bank Technical Paper No. 109. Washington, D.C. Carter, E.D. 1978. "A Review of the Existing and Potential Role of Legumes in Farming Systems of the Near East and North African Region." Report to ICARDA. Glen Osmond, South Australia: Department of Agronomy, Waite Agricultural Research Institute, University of Adelaide. Mimeograph. Chekli, H. 1991. "Elements du Choix des Sequences d'Installation de la Culture du Ble dans la Region de Meknes." These de Doctorat Es-Sciencees Agronomiques. Rabat, Morocco: Institut Agronomique et Veterinaire Hassan II. 123 Christiansen, S., and B. Boulanouar. 1993. "Progress of Ley Fanming in Morocco." In S. Christiansen, L. Materon, M. Falcinelli, and P. Cocks, eds., Introducing Ley Farming to the Mediterranean Basin. Proceedings of an International Workshop, Perugia, Italy, 26-30 June 1989. Aleppo, Syria: ICARDA. CIMMYT. 1991. 1990-91 CIMMYT World Wheat Facts and Trends: Wheat and Barley Production in Rainfed Marginal Environments of the Developing World. Mexico, D.F. CIMMYT. 1993. 1992-93 CJdMMYT World Wheat Facts and Trends: The Wheat Breeding Industry in Developing Countries: An Analysis of Investments and Impacts. Singapore: CIMMYT. CIRAD (Centre International en Recherche Agronomique pour le Developpment). 1993. "How Research Can Accelerate the Generation and Dissemination of New Technologies Needed by Farmers in the Tropics." Seminar presented by Didier Picard, at the World Bank, 19 October 1993. Cleaver, K.M. 1982. The Agricultural Development Experience of Algeria, Morocco, and Tunisia: A Comparison of Strategies for Growth. World Bank Staff Working Paper No. 552. Washington, D.C. Cocks, P.S. 1988. "Increasing Feed Resources for Small Ruminants in the Mediterranean Basin." In E.F. Thomson an F.S. Thomson, eds., Increasing Small Ruminant Productivity in Semi-Arid Areas. [this ref. is not complete] CTIC (Conservation Technology Information Center). 1993. 1993 National Crop Residue Management Survey. West Lafayette, Indiana: NACD's Conservation Technology Information Center. Cooper, P.J.M. 1984. "The Effect of Fertilizer on Water Use and Water Use Efficiency of Barley in Dryland Regions." SAR-MAAR-SD/ICARDA (Syrian Arab Republic - Ministry of Agriculture and Agrarian Reform - Soils Directorate / International Center for Agricultural Research in the Dry Areas). Aleppo, Syria. Cooper, P.J.M. 1986. "The Implications of Crop Rotation, Year, and Residual Effects on Fertilizer Responses of Barley." Annual Report 1985. Aleppo, Syria: ICARDA. Cooper, P.J.M., and E. Bailey. 1990. "Livestock in Mediterranean Farming Systems, a Traditional Buffer Against Uncertainty: Now a Threat to the Agricultural Resource Base." Paper presented at the World Bank Symposium, "Risk in Agriculture," Washington, D.C., 9-10 January 1990. Cramer, H.H. 1967. Plant Protection and World Crop Production. Pflanzenschutz, Nachrichten Bayet, 20: 1- 524. Cunningham, E.P. 1992. Selected Issues in Livestock Industry Development. EDI (Economic Development Institute) Technical Materials. Washington, D.C.: World Bank. 124 Dahan, R., and M. Solh. 1988. "Mechanized Lentil Harvesting." In K. Starks, M. El-Gharous, A. Tanji, S. Lhaloui, R. Riddle, and D. Mulitze, eds., Programme Aridoculture: Annual Research Report 1986-87. Settat, Maroc: INRA-MIAC (Institut National de la Recherche Agronomique / MidAmerica Intemational Agricultural Consortium) at the Centre Regional de la Recherche Agronomique de la Chaouia Abda et Doukkala. De Boer, A.J., H.A. Fitzhugh, R.H. Bermsten, W. Getz, and D.W. Robinson. 1983. Sheep and Goats in Developing Countries: Their Present and Potential Role. Winrock International Livestock Research and Training Center. World Bank Technical Paper. Washington, D.C. Durutan, N. 1992. "Sustainability of Fallow Reduction in Turkey." Paper prepared for the EDI/WB/ICARDA/AOAD (Economic Development Institute / World Bank / Intemational Center for Agricultural Research in the Dry Areas / Arab Organization for Agricultural Development) seminar on "Natural Resources and Environmental Management in the Dry Areas," Aleppo, Syria, 16-27 February 1992. Mimeograph. EDI/WB/ICARDA/AOAD (Economic Development Institute / World Bank / International Center for Agricultural Research in the Dry Areas / Arab Organization for Agricultuml Development). 1992. Seminar on 'Natural Resources and Environmental Management in the Dry Areas," Aleppo, Syria, 16- 27 February 1992. Mimeograph. El-Amami, S. 1977. "Traditional Technologies and Development of the African Environments, Utilization of Runoff Waters: The 'Meskats' and Other Techniques in Tunisia." African Environment, No. 3, Vol. 3/4. Erkan, O., A. Mazid, and K. Somel. 1990. Barley Production Economics in Syria and Turkey: A Comparative Study. Report No. ICARDA-169 En/150/Aug. Aleppo, Syria. FAO (Food and Agriculture Organization of the United Nations). 1978. Report on the Agro-Ecological Zones Project, Vol. 2: Results for Southwest Asia. World Soils Resources Report-48/2. Rome, Italy. _ 1981. 1980 FAO Production Yearbook, Vol. 35. Rome, Italy. 1982. 1981 FAO Production Yearbook, Vol. 36. Rome, Italy. 1989. Agrostat.PC: Computerized Information Series: Land Use. Rome, Italy. 1989-91. Agrostat.PC: ComputerizedInformationSeries: Production. Rome,Italy. _ 1991. 1990Production Yearbook Vol. 45. Rome, Italy. _ 1992. Agrostat.PC: Computerized Information Series: Production. Rome, Italy. 125 Furtado, J.I. dos R., ed. 1994. Sustainable Agricultural Development in the Dry Areas of West Asia and North Africa. Report based on the Semninar on "Natural Resources and Environmental Management in the Dry Areas of West Asia and North Africa," Aleppo, Syria, 16-27 February 1992. Washington, D.C.: Economic Development Institute of the World Bank. Glenn, J.C. 1988. Livestock Production in North Africa and the Middle East: Problems and Perspectives, World Bank Discussion Paper No. 39. Washington, D.C. Guler, M., and M. Karaca. 1988. "Agroclimatological Criteria for Determining the Boundaries of Fallow Practice." In J.P. Srivastava, M.C. Saxena, S. Varma, and M. Tahir, eds., Winter Cereals and Food Legumes in Mountainous Areas, Vol. No. ICARDA-136 En. Aleppo, Syria. de Haan, C. 1990. "The Possible Role of the World Bank in Livestock Lending in West Asia and North Africa." World Bank Internal discussion paper, prepared for EMTAG's (Europe, Middle East, and North Africa Technical Department -- Agricultural Division) sector work on Agricultural Diversification in West Asia and North Africa. Washington, D.C. Mimeograph. de Haan, C. 1993. "The Delivery of Livestock Services to Smallholders." In P.W. Daniels, M. Wilson, eds., Livestock Services for Smallholders: A Critical Evaluation of the Delivery Health and Production Services to the Small-Scale Farmer in the Developing World. Proceedings of "Livestock Services for Smallholders" Seminar, Yogyakarta, 15-21 November 1992. Indonesia: SMT Grafika Desa Putera. Haddad, A., M. Pala, and A.Djaloul. 1989. "Algeria: Summary Results of a Survey of Food Legume Production Practices in the Wilaya of Sidi Bel Abbes." Annual Report for 1988: Farm Resource ManagementProgram. Aleppo, Syria: ICARDA. Hadjichristodoulu, A. 1988. "The Use of Hordeum spontaneum to Breed Barley for Grain and Self- Regenerating Pasture." Rachis: Barley and Wheat Newsletter. Vol. 7, No. 1 & 2 (January & July). Aleppo, Syria: ICARDA. Halse, N.J. 1993. "Australian Attempts to Introduce the Ley Farming System to West Asia and North Africa." In S. Christiansen, L. Materon, M. Falcinelli, and P. Cocks, eds., Introducing Ley Farming to the Mediterranean Basin. Proceedings of an International Workshop, Perugia, Italy, 26-30 June 1989. Aleppo, Syria: ICARDA. Harris, H. 1989. "Crop Water Use and Water Use Efficiency in Contrasting Barley Rotations." Farm Resource Management Program Annual Reportfor 1988. Aleppo, Syria: ICARDA. Herzenni, A. 1988. "Farmer Evaluation of Winter Chick-Peas (Chaouia)." In K. Starks, M. El-Gharous, A. Tanji, S. Lhaloui, R. Riddle, and D. Mulitze, eds., Programme Aridoculture: Annual Research Report 1986-87. Settat, Maroc: INRA-MIAC (Institut National de la Recherche Agronomique / MidAmerica International Agricultural Consortium) at the Centre Regional de la Recherche Agronomique de la Chaouia Abda et Doukkala. 126 Le Houerou, H. 1969. "La vegetation de la Tunisie Steppique," Vol. 42(5). Tunis, Tunisia: Ann. Institut National Recherche Agronomique de Tunisie. ICDF (nternational Conference on Dryland Fanning). 1988. In P.W. Unger, W.R Jordan, T.V. Sneed, and RW. Jensen, eds., Challenges in Dryland Agriculture: A Global Perspective. Proceedings of meeting, Amarillo/Bushland, Texas, 15-19 August 1988. College Station, Texas: Texas Agricu}tural Experiment Station. ICARDA (International Center for Agricultural Research in the Dry Areas). 1988. Annual Report for 1987: Pasture, Forage and Livestock Program. Aleppo, Syria. 1989. Annual Reportfor 1988: Farm Resource Management Program. Aleppo, Syria. ICARDA (international Center for Agricultural Research in the Dry Areas). 1991a. Annual Report for 1990: Cereal Improvement Program. Aleppo, Syria. 199 lb. Annual Reportfor 1990: Farm Resource Management Program. Aleppo, Syria. _____ 1991c. AnnualReportforl990: FoodLegumeImprovementProgram. Aleppo, Syria. 1992a. Annual Report for 1991: Farm Resource Management Program. Aleppo, Syria. 1992b. Annual Reportfor 1991: Food Legume Improvement Program. Aleppo, Syria. 1992c. Annual Reportfor 1990/91: Pasture, Forage and Livestock Program. Aleppo, Syria. 1992d. ICARDA Annual Report 1991. Aleppo, Syria. 1993. In S. Chrisiansen, L. Materon, M. Falcinelli, and P. Cocks, eds., Introducing Ley Farming to the Mediterranean Basin. Proceedings of an International Workshop Perugia, Italy, 26-30 June 1989. 1994a. Annual Reportfor 1993: Farm Resource Management Program. Aleppo, Syria. 1994b.; Annual Report for 1993: Farm Resource Management Program. Aleppo, Syria. ELCA (International Livestock Centre for Africa). 1994. Improving Livestock Production in Africa: Evolution of ILCA 's Programme 1974-94. Addis Ababa, Ethiopia. ILRAD (International Laboratory for Research on Animal Diseases). 1994. ILRAD 1993: Annual Scientific Report of the International Laboratory for Research on Animal Diseases. Nairobi, Kenya. Jiapeng, H. 1988. "Dryland Agriculture and Food Legume Production in High-Elevation Areas of China." In J.P. Srivastava, M.C. Saxena, S. Varmna, and M. Tahir, eds., Winter Cereals and Food Legumes inMountainousAreas, Vol. No. ICARDA-136 En. Aleppo, Syria. 127 Jordan/ICARDA (International Center for Agricultural Research in the Dry Areas) "Results of a Survey on Characterization of Barley/Livestock Farms in Jordan." Third Coordination Meeting, Amman, Jordan, 23-25 September 1991. Kacemi, M., G.A. Peterson, and R.M'Rabet. 1994. "Water Conservation, Wheat-Crop Rotations and Conservation Tillage Systems in a Turbulent Moroccan Semi-arid Agriculture." In M. Karrou, F. Nassif, M. El Bouhssini, and M. El Gharous, eds., proceedings on Acquis et Perspectives de la Recherche Agronomique dans les Zones Arides et Semi-Arides du Maroc, Rabat, 24-27 May 1994. Settat, Morocco: Institut National de la Recherche Agronomique, le Centre Aridoculture. Kassam, A.H. 1988. "Some Agroclimatic Characteristics of High-Elevation Areas in North Africa, West Asia, and Southeast Asia." In J.P. Srivastava, M.C. Saxena, S. Varma, and M. Tahir, eds., Winter Cereals and Food Legumes in Mountainous Areas, Vol. No. ICARDA-136 En. Aleppo, Syria. Khaldi, N. 1992. The Emergence of Barley in the Middle East and North Africa: The Case for Tunisia. Report prepared for the Arab Fund for Economic and Social Development, the United Nations Development Programme, and the Arab Agronomists Union. Washington, D.C.: IFPRI. Koppen, W., and H. Geiger. 1936. Handbook of Climatology. Berlin, Germany: Gerruder Borntrager. Kutsch, H. 1983. "Currently used Techniques in Rainfed Water Concentrating Culture: The Example of the Anti-Atlas." Applied Geography and Development, Vol. 21. Eahlou Kassi, A. 1987. "Improving Efficiency of Intensified Small Ruminant Production in Morocco through Improved Reproductive Performance." In A.W. Qureshi and H.A. Fitzhugh, eds., Small Ruminants in the Near East. Vol 1, Selected Papers. Rome, Italy: FAO. Lehane, B. 1991. "China and Wool: A Giant Constrained." Partners in Research for Development. Canberra: ACIAR (Australian Centre for International Agricultural Research). Mann, C.K. 1980. "An Economic Analysis of the Benefits of Chemical Weed Control on Wheat in Anatolia." Mimeograph. Mariner, J.C., D.M.O. Akabwai, J. Toyang, N. Zoyem, and A. Ngangyou. 1994. "Comnmunity-Based Vaccination with Thermostable Vero Cell-Adapted Rinderpest Vaccine (Thermovax)." In G.J. Rowlands, M.N. Kyule, and B.D. Perry, eds., The Kenya Veterinarian (Special Issue): Proceedings of the 7th International Symposium on Veterinary Epidemiology and Economics, Nairobi, Kenya, 15-19 August 1994. Vol. 18, No. 2. Nairobi, Kenya: ILRAD (International Laboratory for Research on Animal Diseases). Martin, A. 1988. "Tritordeum: the First Ten Years." Rachis: Barley and Wheat Newsletter, Vol. 7. No. 1 & 2. (January & July). Aleppo, Syria: ICARDA. Mason, I.L. 1969. A World Dictionary of Livestock Breeds, Types, and Varieties. Famham Royal, England: Commonwealth Agriculture Bureau. 128 Mazid, A., and M. Hallajian. 1984. Crop-Livestockinteractions: Informationfrom a Barley Survey in Syria. Research Report No. 10. Prepared for the "First ICARDA Regional Fanning Systems Research Workshop," Aleppo, Syria, May 24-26, 1983. McCown, RL., A.L. Cogle, A.P. Ockwell, and T.G. Reeves. 1987. 'Nitrogen Supply to Cereals in Legume Ley Systems Under Pressure." In J.R. Wilson, ed., Advances in Nitrogen Cycling in Agricultural Ecosystems, Symposium Proceedings, Brisbane, Australia, 11-15 May 1987. Wallingford, U.K.: C.A.B. International. McCown, RL. 1993. "Being Realistic about Ley Farming." The 27th International Grassland Congress, Rockhampton, Australia, 19-21 February, 1993. (in press page 4 of mimeograph). Morris, M.L., A. Belaid, and D. Byerlee. 1991. "Wheat and Barley Production in Rainfed Marginal Environments of the Developing World." Part 1 of 1990-91 CIvIYT World Wheat Facts and Trends: Wheat and Barley Production in Rainfed Marginal Envronments of the Developing World. Mexico, D.F.: CIMMYT. NCARTr (National Centre for Agricultural Research and Technology Transfer). 1991. "Characterization of Barley/Livestock Farms in Jordan." Report prepared for the Jordan/ICARDA (international Center for Agricultural Research in the Dry Areas) Third Coordination Meeting, Amman, Jordan, 23-25 September 1991. Nordblom, T.L., D.J. Pannell, S. Christiansen, N. Nersoyan, and F. Bahhady. 1994. "From Weed to Wealth? Prospects for Medic Pastures in the Mediterranean Fanning System of North-West Syria." Agricultural Economics. Vol. 11 (September 1994). Nordblom, T.L., and F. Shomo. 1994. "Food and Feed Prospects to 2020 in the West Asia / North Africa Region." Paper prepared for the IFPRI workshop on "Sustainable Agricultural Growth in the Major Ecoregions of the Developing World: Prospects to 2020," held at Airlie House, Warrenton, Virginia, November 7-9, 1994. (NRC) US National Research Council. 1993a. Soil and Water Quality: An Agenda for Agriculture. Washington, D.C.: National Academy Press. (NRC) US National Research Council, Board on Agriculture. 1993b. Managing Global Genetic Resources: Livestock. Washington, D.C.: National Academy Press. Oram, P.A. 1956. Pastures and Fodder Crops in Rotations in Mediterranean Agriculture. Agricultural Development Paper No. 57. Rome, Italy: FAO. Oram, P.A. 1979. "Crop Production Systems in the Arid and Semi-Arid Warm Temperate and Mediterranean Zones." In D.W. Thome and M.D.J. Thomne, eds., Soil, Water and Crop Production. Westport, Connecticut: AVI Publishing Company, Inc. 129 Oram, P.A. 1988. "Agricultural Production and Food Deficits in West Asia and North Africa: Future Prospects and the Role of the High-Elevation Areas." In J.P. Srivastava, M.C. Saxena, S. Varma, and M. Tahir, eds., Winter Cereals and Food Legumes in Mountainous Areas, Vol. No. ICARDA-136 En. Aleppo, Syria. Oram, P.A., and A. Belaid. 1990. Legumes in Farming Systems. Aleppo, Syria: ICARDA. Oram, P.A. 1993. Global Perspective on Population, Resources, and Agricultural Production. Keynote Paper presented at the "Seventh Australian Agronomy Conference," Adelaide, South Australia, September 19-24. Oram, P.A., and A. Belaid, with S. Bhattarai, B. Hojati, and E. Jacinto. 1994. The Current and Potential Role of Barley in Moroccan Agriculture: Results of a Survey of 1062 Farms, and the Implications for Agricultural Policy. Report prepared for the United Nations Development Program, Regional Bureau for Arab States, and the Arab Fund for Economic and Social Development. Washington, D.C.: IFPRI. Ouassou, A., and H. El Baghati. 1988. "Agriculture in the Mountainous Regions of Morocco." In J.P. Srivastava, M.C. Saxena, S. Varma, and M. Tahir, eds., Winter Cereals and Food Legumes in Mountainous Areas, Vol. No. ICARDA-136 En. Aleppo, Syria. [c 4 p 45 ?? change in text] Pacey, A., and A. Cullis. 1986. Rainwater Harvesting: The Collection of Rainfall and Runoff in Rural Areas. London, U.K.: IT Pubication. Pala, M., D. Tully, A. Rassam, A. Mazid, and P. Cooper. 1987. "Fertilizer and Herbicide Effects on Farmers' Wheat Production in Northwest Syria." Paper presented at the "Farming Systems Research Symposium," University of Arkansas, 18-21 October 1987. Pala, M. 1991. "Sustainability of Fallow Replacement." EDI/ICARDA (Economic Development Institute / Intemational Center for Agricultural Research in the Dry Areas) meeting, Aleppo, Syria, 1992. Primov, G., A. Herzenni, H. Ben Aouda, and L. Croy. 1988. "Producer Evaluation of Juanillo 97 Triticale (Transect Study)." In K. Starks, M. El-Gharous, A. Tanji, S. Lhaloui, R. Riddle, and D. Mulitze, eds., Programme Aridoculture: Annual Research Report 1986-87. Settat, Maroc: INRA-MIAC (Institut National de la Recherche Agronomique / MidAmerica International Agricultural Consortium) at the Centre Regional de la Recherche Agronomique de la Chaouia Abda et Doukkala. Qureshi, A.W. 1987. In A.W. Qureshi and H.A. Fitzhugh, eds., Small Ruminants in the Near East. Vol 1, Selected Papers. Rome, Italy: FAO. Rafsnider, G.T., RL. Zimdahl, and M. Hamnmida. 1990. "Costs Associated with Weed Management in Cereals and Food Legumes in Safi Province, Morocco." Plant Protection Bulletin, Vol. 38, No. 4. Rome, Italy: FAO. 130 Regehr, D.L. 1988. "Effects of Hessian Fly and Weed Stress on 'Nesma' Wheat." In K. Starks, M. El- Gharous, A. Tanji, S. Lhaloui, R. Riddle, and D. Mulitze, eds., Programme Aridoculture: Annual Research Report 1986-87. Settat, Maroc: INRA-MIAC (Institut National de la Recherche Agronomique / MidAmerica International Agricultural Consortium) at the Centre Regional de la Recherche Agronomique de la Chaouia Abda et Doukkala. Reij, C., P. Mulder, and L. Begemann. 1988. Water Harvesting for Plant Production. World Bank Technical Paper No. 91. Washington, D.C. Ruthenberg, H. 1980. Farming Systems in the Tropics, 3rd edition. Oxford, England: Clarendon Press. Saade, M.E. 1993. "Triticale Production and Utilization by Tunisian Farmers: Preliminary Results of a Farm Survey (1992)." Paper presented at the Rockefeller Foundation Social Science Fellows Meeting, CIMMYT, Mexico, November 9-13, 1992. Tunis, Tunisia: Rockefeller Foundation. Saade, M.E., A. Amri, H. El Baghati, F. Nassif, A. Benyassine, and K. Benahnia. 1994. "Constraints to the Adoption of New Barley Varieties in Morocco." Paper prepared for L'Institut National de la Recherche Agronomique (INRA), La Direction de la Production Vdgtale du Ministere de l'Agriculture et de la Reforme Agraire (DPV-MARA), and ICARDA. Aleppo, Syria: ICARDA. Sarna, J.S. 1986. Cereal Feed Use in the Third World: Past Trends and Projections to 2000. Research Report No. 57. Washington, D.C.: IFPRI (International Food Policy Research institute). Scoones, I., ed. 1994. Living with Uncertainty: New Directions in Pastoral Development in Africa. London, U.K.: IT Publications. Siming, H., S. Juntung, and Y. Chunfeng. 1988. "Research on Stubble Mulch Tillage on Rainfed Land." In P.W. Unger, W.R. Jordan, T.V. Sneed, and R.W. Jensen, eds., Challenges in Dryland Agriculture: A Global Perspective. Proceedings of the International Conference on Dryland Farming, Amarillo/Bushland, Texas, 15-19 August 1988. College Station, Texas: Texas Agricultural Experiment Station. Somel, K., A. Mazid, and M. Hallajian. 1984. "Survey of Barley Producers, Northern Syria (1981/82)." Descriptive Statistics, Vol. III. Research Report No. 12-III, Aleppo, Syria: Farming Systems Program, ICARDA. Somel, K. 1991. Tradeoffs Between Yield Levels and Yield Variability an Example of Fertilizer Use on Rainfed Barley. Seminar on "Natural Resource Management in Dry Areas," Aleppo, Syria, 16-27 February 1992. Somel, K. 1992. "Agriculture Diversification in the Middle East and North Africa." Paper prepared for the EDI/WB/ICARDA/AOAD, seminar on "Natural Resources and Environmental Management in the Dry Areas," at Aleppo, Syria, 16-27 February 1992. 131 Springborg, R. 1990. "Human Constraints in Extending the Use of Forage Legumes in Mediterranean Areas." In A.E. Osman, M.H. Ibrahim and M.A. Jones, eds., Developments in Plant and Soil Sciences: The Role of Legumes in the Farming Systems of the Mediterranean Areas. Dordrecht, Netherlands: Kluwer Academic Publishers. Srivastava, J.P., P.M. Tarnboli, J.C. English, R. Lal, and B.A. Stewart. 1993. Conserving Soil Moisture and Fertility in the Warm Seasonally Dry Tropics. World Bank Technical Paper No. 221. Washington, D.C. Steinbach, J. 1987. "The Possibilities and Development Requirements for Improving Small Ruminant Production in the Near East." In A.W. Qureshi and H.A. Fitzhugh, eds., Small Ruminants in the Near East. Vol. 1, Selected Papers. Rome, Italy: FAO. Stewart, B.A. 1988. "Research on the Conjunctive Use of Rainfall and Irrigation." In P.W. Unger, W.R. Jordan, T.V. Sneed, and R.W. Jensen, eds., Challenges in Dryland Agriculture: A Global Perspective. Proceedings of the International Conference on Dryland Farming, Amarillo/Bushland, Texas, 15-19 August 1988. College Station, Texas: Texas Agricultural Experiment Station. Stone, B., C. Greer, T. Zhong, C. Friedman, M. McFadden, and M. Snyder. 1985. "Agro-Ecological Zones for Wheat Production in China: A Compendium of Basic Resource Materials." Washington, D.C.: IFPRI (International Food Policy Research Institute) Tadros, K.I. 1989. Questionnaire conducted to "Characterize Sheep Production in Jordan" [unpublished]. Cited in Tadros, 1992. Tadros, K.I. 1992. "Current Situation and Future Potential of Dry Rangelands in Jordan." Aleppo, Syria: ICARDA. Mimeograph. Tahir, M., and P.M. Hayes. 1988. "A Synthesis of Winter Cereal Research and Production Status in High-Elevation Areas of West Asia and North Africa." In J.P. Srivastava, M.C. Saxena, S. Varma, and M. Tahir, eds., Winter Cereals and Food Legumes in Mountainous Areas, Vol. No. ICARDA-136 En. Aleppo, Syria. Thomson, E.F., F. Bahhady, and T.L. Nordblom. 1982. The Role of Livestock in the Farming System. FSRP (Farming Systems Research Program) Research Report (December). Aleppo, Syria: ICARDA. Troll, C., and K.H. Paffen. 1965. "Seasonal Climates of the Earth." In H.E. Landsberg, and others, eds., WorldMaps of Climatology, 2nd edition. New York: Springer-Verlag. Tully, D. 1986. "Rainfed Farnning Systems of the Near East Region." Discussion Paper No. 18. Prepared for the workshop on "Soil, Water and Crop/Livestock Management Systems for Rainfed Agriculture in the Near East Region," Amman, Jordan, 18-23 January 1986. Aleppo, Syria: ICARDA. 132 USDA/SCS (United States Department of Agriculture / Soil Conservation Service). 1992. "Executive Summnary 1991." Soil and Water Conservation Research Needs. Washington, D.C.: SCS. * UNDP/FAO (United Nations Development Programme I UN Food and Agriculture Organization). Study of Beef Production Based on the Use of Crop Residues, China (table 5-8) [not complete] Wickens, G.E., and C.L.P. White. 1978. "Land Use in the Southern Margins of the Sahara." In B.H. Walker, ed., Management of Semi-Arid Ecosystems, Amsterdam, The Netherlands: Elseviers Scientific Publications. World Bank. 1986. World Development Report. Washington, D.C.: World Bank. World Bank. 1994a. "Integrated Pest Management (IPM): An Environmentally Sustainable Approach to Crop Protection." Agriculture Technology Notes, No. 2, (February). Washington, D.C.: World Bank, The Agricultural Technology and Services Division (AGRTN). World Bank. 1994b. "Participatory Rural Assessment: Techniques to Improve Communication with the Rural Poor." Agriculture Technology Notes, No. 6, (October). Washington, D.C.: World Bank, The Agricultural Technology and Services Division (AGRTN). World Bank. 1994c. "Integrated Soil Management for the Tropics: A Profitable and Environmentally Sound Approach to Land Husbandry." Agriculture Technology Notes, No. 7, (November). Washington, D.C.: World Bank, The Agricultural Technology and Services Division (AGRTN). Umali, D.L., G. Feder, and C. de Haan. 1992. "The Balance between Public and Private Sector Activities in the Delivery of Livestock Services." World Bank Discussion Papers, No. 163. Washington, D.C. Zimdahl, R.L., G.T. Rafsnider, M. Boughlala, and A. Laamari. 1992. "Costs Associated with Weed Management in Cereals and Food Legumes in the Chaouia Region of Settat Province, Morocco." Weed Technology, Vol. 6, Issue 1 (January-March), reprint. Annex DA TA The following tables present data on many of the issues discussed in this volume. Although much of the data will be familiar to professionals in the field, the authors regret that it was not possible to locate the original sources of all of the data compiled in the course of this study and presented in this annex. Table Al. Agricultural Technology Assessment: Projects Reviewed by Project Type and Country COUNTRY ALGERIA ARGENTINA CHILE CHINA JORDAN MIOROCCO SYRIA TUNISIA TURKEY PROJECr TYPE . . Ausicurw I t credit Ist credit 2nd credit Ist credit Ag. SA 2nd credit 2nd credit Credit APIS AP7 AP0 AP63 PAs9 PA8S45 AP83 PA8I PA88 2nd credit National credit PA76 PA90 2nd credit PA74 APS9 4th credit 3rd credit AP8S 2nd credit PC92 APS9 PA79 AP87 3rd credit At Ew. Review AP76 1991A S.A. 4th credit Loan AF79;PCS4 RepS6 5th credit Acusural Hebei AiL Dev. Mid-Allas S. Retion Northwest Developnmt AP90 APSI;PC91 PC90 APS5 Xinmiiag Oulmes-Romani AP87 AP82 Shainli Fez-Karia Kissa APS9 AP7S _ Rural Gafusu Loukkos Northwes Erzcrum Dcvvloprmt Provincial PC90;APS0 APSI PA92 API? Meknes AR. Dev. PC9 1 Corum-Cankir _________ __________ AP75;PC90 PAS8;PC86 Livestock BalcaI Uvestock ISt LiVesO& 2nd Livestock AP67 AP76;PAS6 PASI PAS I 4th Livestock PA90;AP7S Sili Livestock ,___ _ _ __ _ APSO;PA90 Reaar and R&E (Piloti Irtitutional RhE Extemion R&E RhE Extanion AP90 AP91 AP90 AP85 AP90 APR4 R&E 11 AP90 Research _ _ _ _ ~~~~~~~~~AP92 Fruit =, Lveok diiri 2nd F&V Vega"ble FruiL and EXPOW AP81 Vegetable Marketing PA90 PC83 PA90 . _. COUNTRY AL(ERIA ARGENTINA CHILE CHINA JORDAN MOROCCO SYRIA TUNISIA TURKEY Waler Pilot Proiet N.W. Mountain E_Analoli xa- -w;--t 1:,ks Revtew~~~~~~~~~~~~~~~~~~~~~~~9 Meeting S'9 Land ISt ad 2nd Lower Recl naioa aait prieds EuphLtcs (see credi above) Draina . AP AppraiI rqtst (vith yar of report) PC Projed unpletion repmt (with yr of report)993 PA Perfonmance audit report (with year of repoet) SA Sector adjusmeit loan R&E Research and extension Note: Abbreviated projct titles are ulined. Table A2. Land Suitabe for Growing Rainfed We inWestAsia(thousandacres) __ MAJOR COOL SUBTROPICS COOL TEMPERATE COOL TROPICS TOTAL WEST ASIA CLIMATE (Whnter RaIfI) VS/S MS NS TOTAL VS/S MS NS TOTAL VSIS MS NS TOTAL VS/S MS NS TOTAL 330-364 davs _______10 253 263 o10 253 263 270-329 days 299 51 1,663 2,013 _ 299 51 1,663 2,013 210.269 days 2,284 14 7,225 9,523 _ 2,284 14 7,225 9.523 150-209da 5,852 966 17,688 24,506 558 25 1,054 1,637 6,410 991 18,742 26,143 120-149 days 2.610 2,084 12,532 17,226 . 153 4,151 4,304 2,763 2,084 16,683 21,530 90-119days 303 5,302 29,812 35,417 2 181 551 734 78 6,542 6,626 383 5,483 36,911 42,777 75-99 days I 1 3,114 3,125 4__ 11321 1,385 5 1 4,495 4,510 1-74 days _ 618 23,161 23,779 _ _ _ 618 23,161 23,779 75-S9 (Or) 199 5,616 5,815 6 2,746 2,752 205 8,362 8,567 0 292,075 292,075 = _ 2.983 2,983 295,058 295,058 TOTAL- 12.176 2,427 393,139 413,742 564 206 2,986 3,756 237 16,428 16.665 12,977 2,633 412,553 434,163 VS/S = Very suitable or suitabe for growing wheat MS - MargialWly suitable for growing wheat NS - Not sutable for growing wheat _ Note: Zones are classified as "very suitable' fo whea production where the yield ia more than 80 pe t of the maximum attainable at high input levels, 'suitable' where the yield is from 40-80 percent of the maximum, an 'marginally suitable' where it is 20-49 parcat of the manmn atinable. l than 20 peroat ig "no suitable.' Anodter 243,854,000 ha in West Asia are unsuitable for wheat in three other major clinatic zones: warm tropical lowban1 (130,187 ha unuitable, oold subropics (100,165 Ha unuitabley and cold temperate 13,502 ha unsuitable). Source: FAO 1978. Table A3. Statistics on Land Use for West Asia and North Africa, 1973-80 Year 1973 1974 1975 1976 1977 1978 1979 1980 Land area 1000 ha 1543401 1543373 1543365 1543363 1543357 1543356 1543354 1543350 Arable and perm. crops 1000 ha 106041 105675 104935 105064 104495 104209 106117 103445 Percent of land area 6.87 6.85 6.80 6.81 6.77 6.75 6.88 6.70 Arable land 1000 ha 98717 98230 97439 97321 96555 96215 97944 95146 Percent of land area 6.40 6.36 6.31 6.31 6.26 6.23 6.35 6.16 Perm crops 1000 ha 7324 7445 7496 7743 7940 7994 8173 8299 Percent of land area 0.47 0.48 0.49 0.50 0.51 0.52 0.53 0.54 Perm pastures 1000 ha 384702 384922 384482 384739 385084 385315 385420 386552 Percent of land area 24.93 24.94 24.91 24.93 24.95 24.97 24.97 25.05 Forest and woodland 1000 ha 114389 114190 113905 113549 113169 112809 112717 112394 Percent of land area 24.93 24.94 24.91 24.93 24.95 24.97 24.97 25.05 Other land 1000 ha 938269 938586 940042 940011 940609 941023 939100 940959 Percent of land area 60.79 60.81 60.91 60.91 60.95 60.97 60.85 60.97 Irrig. Agr. 1000 ha 19940 19776 19729 19759 19594 19647 19403 19189 Percent of arable 20.20 20.13 20.25 20.30 20.29 20.42 19.81 20.17 Souirce-. Agrostat Data Tape, FAO. Table A4. Statistics on Land Use for West Asia and North Africa, 1981-1989 Year 1981 1982 1983 1984 1985 1986 1987 1988 1989 Land area 1000 ha 1543350 1543350 1543350 1543351 1543351 1543351 1543351 1543351 1543351 Arable and perm 1000 ha 104348 103683 102831 104020 104240 104411 105271 105195 105609 Percent of land area 6.76 6.72 6.66 6.74 6.75 6.77 6.82 6.82 6.84 Arable land 1000 ha 96094 954% 94536 95704 95809 95911 96661 96544 96892 Percent of land area 6.23 6.19 6.13 6.20 6.21 6.21 6.26 6.26 6.28 Perm crops 1000 ha 8254 8187 8295 8316 8431 8500 8610 8651 8717 Percent of land area 053 0.53 0.54 0.54 0.55 0.55 0.56 0.56 0.56 Perm pastures 1000 ha 381906 381875 382165 381967 381292 380752 380626 380480 320225 Percent of land area 24.75 24.74 24.76 24.75 24.71 24.67 24.66 24.65 24.64 Forest and 1000 ha 112142 111891 111629 111339 111082 111123 110861 1105% 110491 woodland 24.75 24.74 24.76 24.75 24.71 24.67 24.66 24.65 24.64 Percent of land area Other land 1000 ha 944954 945911 946725 946025 946737 947065 946593 946810 9470266 Percentof land area 61.23 61.29 61.34 61.30 61.34 61.36 61.33 61.35 61.36 o Lrrig. Agr. 1000 ha 19701 20108 20274 20437 20587 20725 20764 21635 21725 Percent of arable 20.50 21.06 21.45 21.35 21.49 21.61 21.48 22.41 22.42 Source. Agrostat Data Tape, FAO. 139 Table A5. Total and High-Altitude Production and Area of Wheat in Selected WANA Countries COUNTRY AREA PRODUCTION YIELD (T/ha) (000 ha) (000 MI) TOTAL HAA TOTAL HAA OVERALL HAA Algeria 1,735 1,040 1,446 520 .83 .50 Morocco 2,226 500 3,809 275 1.71 .55 Afghanistan 2,100 1,260 2,500 1,071 1.19 .85 Iran 6,200 4,500 6,400 2,835 1.03 .63 Iraq 700 320 800 230 1.14 .72 Turkey 10,000 7,500 19,000 12,750 1.90 1.70 Total 22,961 15,120 33,954 17,785 1.48 1.18 Percentage of (66%) (52%) (79%) Total WANA I I HAA = High altitude areas (greater than 1,000 meters) Source: Adapted from Tahir and Hayes 1988, p. 230. Table A6. Total and High-Altitude Production and Area of Barley in Selected WANA Countries COUNTRY AREA PRODUCTION YIELD (T/ha) (000 ha) (000 MT) TOTAL HAA TOTAL HAA OVERALL HAA Algeria 1,200 650 1,100 423 .92 .65 Morocco 2,445 500 3,563 350 1.46 .70 Afghanistan 300 250 300 300 1.00 .80 Iran 1,850 1,350 1,650 986 1.89 .73 Iraq 850 250 700 225 .82 .90 Turkey 3,500 2,350 7,000 4,230 2.00 1.80 Total 10,145 5,350 14,313 6,414 1.41 1.20 Percentage of (53%) (45%) (85%) Total for WANA HAA = High altitude areas (greater than 1,000 meters) Source: Adapted from Tahir and Hayes 1988 p. 230. 140 Table A7. Annual Yields of Wheat and Barley in Six Provinces in Turkey 1946-50, 1966, 1977, and 1987 (metric tons per hectare) 1946-50 1966 1977 1987 PROVINCE Wheat Barley Wheat Barley Wheat Barley Wheat Barley Adana 1.1 1.0 1.4 1.8 2.5 1.9 4.0 2.5 Edirne 0.8 0.9 1.3 1.3 2.6 2.6 3.2 3.2 Corum 0.9 1.0 1.4 1.7 1.4 1.8 1.9 2.3 Konya 0.8 0.9 0.9 1.0 1.7 2.4 1.9 2.4 Erzurum 0.9 1.1 0.9 1.3 0.9 1.0 1.1 Kars 0.9 1.1 1.0 1.2 1.0 1.1 1.1 1.5 Note: Adana and Edirne are low elevation coastal areas with mild winters. Corum and Konya are middle elevation areas in Central Anatolia. Erzurum and Kars are high elevation areas in Eastem Turkey. Source: Govermment of Turkey, 1954, 1968, 1979, and 1989. Table A8. Wheat Area, Yield, and Production in Rainfed Marginal Mega-Environment in Selected Countries in the WANA Region | BREAD WHEAT | DURUM WHEAT COUNTRY Area Yield Production Area