Contents. Page. Foreword iii 1. THE TROPICAL ENVIRONMENT 1. Soils and climatic resources 1

Transcription

1

2 Foreword Soil, the most basic natural resource, is finite and fragile. It may take thousands of years to produce a few centimetres of fertile topsoil. Only 11% of the earth s land area of 13.4 X 10 9 ha is currently cultivated. With rapidly increasing population, especially in tropical regions, the arable land per caput is rapidly decreasing. Densely populated Asia, with 75% of the earth s population, has little additional land to bring under cultivation. Most of the land in Africa and Latin America that can be brought under cultivation is located within fragile and ecologically sensitive regions, for example tropical rainforest, acid savannah, drought-prone sahel, steepland of the Andes. Potentially productive agricultural land is either inaccessible, too steep, too shallow, too wet or too dry. Soil degradation is widespread, particularly in the tropics, where it is a major economic and ecological constraint. It leads to poor yields, poor standards of living, and contributes to hunger and malnutrition. The ecological consequences of soil degradation include pollution of wind and water by sediment and dust, and eutrophication and contamination of natural waters by sediment-borne pollutants. Processes of soil degradation most prevalent in the tropics include soil erosion and desertification, compaction and hard-setting, salt and water imbalance in the root zone, biological degradation including reduction in soil organic matter content, fertility depletion, and chemical degradation including leaching and acidification. Tillage, mechanical soil manipulation for seedbed preparation, affects the rate and trend of soil degradation. In addition to establishing the seed-soil contact, tillage is used to alleviate soil compaction and so improve infiltration capacity; to dispose of pathogen-infested crop residue; to incorporate fertilizer into the root zone, and to eradicate weeds. The kind of tillage and its frequency depend on the soil and its related constraints to crop production. A wide range of tillage implements has been developed to alleviate unfavourable soil conditions. Manual or animal driven tillage tools are widely used in small-scale agriculture, whereas motorized equipment is used in large-scale commercial or semi-commercial agriculture. The term mode refers to the type of tillage whether primary, secondary or tertiary based on tools for soil inversion, loosening, levelling, mixing or pulverizing. Similarly, the term means refers to the source of power and the type of machinery for tillage. Both mode and means of tillage have changed drastically since industrialization in the 19th century. Equipment has become larger and faster, so the risks of soil and environmental degradation have drastically increased. Recognition of the need to protect bare soil from soil erosion and erosion-induced degradation has led to the development of a wide range of practices involving the reduction of the intensity and frequency of tillage and the practice of leaving some crop residue or other biomass on the soil surface as a protective barrier against erosion. Reduced cultivation, an integral part of conservation tillage, has been widely adopted since the early 1950s. The widely recognized risks of soil erosion and environmental degradation in the tropics are only partly due to fragility of the soils. Their susceptibility to degradative processes can be accentuated by mechanical soil distur-

3 bance. The harshness of tropical climates (high rainfall intensity and excessive energy load of rains, and extremes of temperatures) is another important factor responsible for soil degradation. Furthermore, the loss of nutrient-rich tosoil has more marked effects on crop yield, and productivity in low-input subsistence farming than in sciencebased semi-commercial or commercial agriculture. Tillage systems in the tropics are generally designed to lower soil temperatures and conserve soil moisture, in contrast to practice in temperate regions where agriculturalists attempt to drain excess water from the root zone during early spring and so raise soil temperatures. The objectives of this Bulletin are to collate up-to-date information on soil tillage requirements for soils in the tropics; to assess the impacts of different modes and means of tillage on the soil and the environment and on crop productivity; and outline criteria for developing environmentally-friendly and economically-viable tillage techniques for sustainable use of soil and water resources. The question Is tillage necessary? is addressed to explore viable alternatives to mechanical soil disturbance. The question How much cultivation, how often, and with what equipment? is considered in relation to the short-term and long-term consequences for soil and environmental quality, with special reference to accelerated erosion, water pollution and emission of greenhouse gases. The Bulletin is written for field staff, extension agents, policy-makers, project managers and researchers interested in tillage requirements and soil surface management for sustained crop production. While the technical terminology used is simplified and explained to facilitate its use by generalists, the scientific content and references quoted address the needs of active researchers. An attempt is made to address the concern of research scientists, decision-makers and planners about the widespread problem of soil and environmental degradation. Why do soils of the tropics degrade so rapidly with intensive cultivation? How can this degradation be stopped without jeopardizing productivity? How can the productivity of degraded soils be restored? How can the labour-constraints of resource-poor farmers be addressed and their facility augmented to increase the land area under production? How can the drudgery of the farming practices in the tropics be alleviated and quality of life for the farming community be improved? How can subsistence farming be transformed into semi-commercial or commercial agriculture? These important questions require a critical and an objective approach.

4 Contents Foreword iii Page 1. THE TROPICAL ENVIRONMENT 1 Soils and climatic resources 1 Arid zone 1 Seasonally dry tropics and subtropics 2 Humid tropics and subtropics 3 Mountainous zone 3 Soil-related constraints on intensive arable land use 5 Soil degradation and intensive land use 11 Decline in soil organic matter content 11 Clay migration and clay pan formation 12 Structural deterioration and soil compaction 13 Laterization 15 Hard-setting 15 Soil erosion 16 Nutrient depletion and leaching PREDOMINANT FARMING SYSTEMS OF THE TROPICS 17 Traditional farming systems 17 Some alternative farming systems 18 Large-scale mechanized farming 18 Ranching and pasture development 18 Forestry TILLAGE SYSTEMS 21 Definitions 21 Traditional tillage 21 Plough-till 21

5 Page Conservation tillage 22 Qualifying criteria for conservation tillage 23 Factors affecting the choice of tillage system 25 Tillage methods and socio-economic factors SOIL SURFACE MANAGEMENT 27 Mulch farming 27 Beneficial effects of mulch farming 30 Erosion control 34 Soil fertility improvement 36 Cover crops and planted fallows 39 Live mulch 41 No-till systems 47 Advantages and disadvantages of no-till systems in the tropics 53 Ridge tillage 54 Tied-ridges IMPROVING ROOTING DEPTH 61 Amelioration in soil structure 61 Soil restorative cropping systems 63 Management of planted fallows 63 Killed sod as residue mulch 63 Live mulch 64 Agroforestry and alley cropping 65 Alleviating soil compaction by deep ploughing 65 Semi-arid West Africa 65 Southern Africa 66 Tropical India 67 Summary 67 Alleviating soil compaction by paraplough and subsoiling 67 Improving rooting depth by drainage 68 Managing leaching losses 72

6 Page 6. CROP RESPONSE TO TILLAGE ON VARIOUS SOILS 75 Lixisols and tropical Luvisols 75 Crop responses on Lixisols and tropical Luvisols with favourable properties 75 Crop responses on Lixisols with unfavourable properties 80 Sandy soils of low water holding capacity 101 Increasing soil moisture storage capacity 101 Decreasing soil evaporation 104 Increasing rooting depth 105 Vertisols 107 Erosion control 108 Soil compaction and cracking 114 Water conservation 114 Systems approach to management of Vertisols 119 Acid tropical soils 120 Acrisols 120 Ferralsols 125 Sodic soils 126 Reclamation of sodic soils (Solonetz) by tillage 127 Hard-setting soils CONSERVATION TILLAGE IN THE TROPICS 133 Improvements in soil structure 133 Soil and water conservation 136 Favourable soil moisture and soil temperature regimes 137 Soil chemical and nutritional properties and fertilizer response 137 Energy conservation 145 Preventing soil degradation and maintaining soil fertility TILLAGE SYSTEMS AND AGRICULTURAL SUSTAINABILITY 147 Tillage and sustainability 147

7 Page Systems approach to tillage methods 149 Agroforestry and alley cropping 150 Cover crops 154 Live mulch 157 Rotations and multiple cropping 157 Summer fallowing 158 Sustainability and agricultural productivity 160 Agricultural productivity 161 Soil properties SOIL PROPERTIES AS A GUIDE TO TILLAGE 167 Soil characterization in relation to tillage requirements 167 Soil structure assessment 169 Soil guide for tillage requirements CONCLUSIONS: RESEARCH AND DEVELOPMENT PRIORITIES 175 Past achievements 175 Knowledge gaps 176 Researchable priorities 177 Developing a soil guide 177 Reducing inputs 177 Crop stand establishment 177 Diverse farming and cropping systems 177 Agricultural sustainability 178 The holistic approach 178 Socio-economic factors 178 Soil degradation 179 Weed control measures 179 Alleviating drudgery 179 REFERENCES 181

8 List of figures Page 1. Broad categories of tropical climates according to Koppen (1936) 2 2. Major climates 4 3. Main soil patterns 6 4. Agricultural sustainability in relation to climate 8 5. Conceptual relationship between land degradation and climatic aridity 8 6. Factors and processes of soil degradation Principal farming systems of the tropics Factors affecting choice of tillage methods Objectives of seedbed preparation Strategies to enhance water infiltration Different types of mulch materials and cultural practices Mulching effects on soil moisture content under maize in southwestern Nigeria Mulching effects on cumulative infiltration Effects of surface applied mulch on percentage ground cover of a soil in semi-arid regions of Kenya Mulching effects on soil temperature in southwestern Nigeria Effects of grasses and legumes grown as cover crops on cumulative infiltration into an eroded Lixisol Effects of Mucuna cover on runoff and soil erosion Effects of Mucuna cover on water infiltration Maize grain yield for no-till and till systems of seedbed preparation in Nigeria Effect of tillage methods on grain yield of flooded rice Effects of tillage methods of soil temperature at IITA, Ibadan Change in infiltration rate with cultivation duration Schematic cross-sections of different land configurations adopted for Vertisols to facilitate water management Schematic diagram of the broadbed and furrow (BBF) system Zonal tillage concept based on three zones, e.g. traffic zone, seedling environment zone and water management zone Maize gain yield on a Lixisol for 24 consecutive crops grown with no-till and plough-till methods of seedbed preparation Effect of tillage methods on leaf water potential of maize grown on a Lixisol in Nigeria Effects of tillage methods on root system development of maize Effect of pre-sowing cultivation methods on pearl millet stand measured at ISC, Niamey Effect of pre-sowing cultivation on pearl millet grain yield at ISC, Niamey Mulching X fertilizer interaction on pearl millet yield at ISC, Niamey 87

9 Page 32. Ridging and weeding effects on pearl millet yield over three years ( ) at ISC, Niamey Labour time required for different methods of land preparation at ISC, Niamey using different tools and animal traction Yield decline with continuous cropping on a Vertisol in central India Schematic cross-section of the BBF system developed at ICRISAT Furrow or compacted zone in the BBF system is designed to decrease infiltration rate. Dimensions of a BBF system along with its infiltration rate measured at ICRISAT Comparison of the air-filled porosity in BBF and flat seedbed during the period of high soil moisture content on a Vertisol in India Effect of furrow cultivation on infiltration capacity of a Vertisol Effect of trampling wet soil on plant stand and grain yield of chickpea and safflower on a Vertisol in India Effects of tillage methods on nitrogen response of a rainy-season sorghum on a Vertisol Tillage and liming effects on root system development of an Acrisol in southeastern Nigeria Mulching and tillage effects on grain yield of upland rice in southeastern Nigeria Effects of no-till, plough-till and ploughing with motorized farm operations on water infiltration into a Brazilian Latosol Effects of long-term tillage treatments on infiltration capacity of a Brazilian Latosol Structural profile of a no-till soil Effects of no-till and plough-till systems on soil moisture reserves under maize on a Lixisol in Nigeria Soil temperature in no-till and plough-till systems in a Lixisol at Ibadan, Nigeria Effects of plant cover and tillage methods on soil temperature regime of a Ferralsol in Parana, Brazil Effects of mulch material and residue cover on soil temperature regime of a Lixisol Soil organic matter profiles of a Lixisol in western Nigeria under different land use and tillage systems Schematic diagram of crop response to nitrogen for no-till and plough-till systems Soil P profile of no-till and plough-till systems for a Lixisol at Ibadan, Nigeria Generalized rooting patterns in relation to tillage methods Effects of tillage methods on root systems of crops grown on a Lixisol in western Nigeria with no-till and plough-till systems of management Effects of tillage methods on maize root development on an Acrisol in eastern Nigeria Effects of no-till and plough-till systems on root system development in maize in soils of semi-arid region of West Africa 145

10 Page 57. Tillage effects on agricultural sustainability System approach to conservation tillage Effects of tillage methods and hedgerows of Leucaena and Gliricidia on accumulative infiltration Effects of hedgerow spacing and method of seedbed preparation on the hydrographs generated by a rainstorm of 67.3 mm Maize grain yield on a Lixisol in Western Nigeria for 17 consecutive years Maize grain yield for no-till mulch (NT mulch) and plough-till (ploughed) systems for 8 consecutive years on a sandy soil in Western Nigeria Grain yield of maize and cowpea grown in association with perennial hedges of Leucaena and Gliricidia in western Nigeria Effects of tillage method imposed on land clearing techniques on maize grain yield in Nigeria Effect of cultivation duration on (a) equilibrium infiltration rate, and (b) soil-water sorptivity of Lixisols evaluated in a 10-year watershed project in southwestern Nigeria Matrix depicting tillage needs in relation to soil properties and moisture regime Matrix depicting tillage needs in relation to soil properties and climatic aridity 174 List of tables 1. Extent of major soil groups by climatic zone 7 2. Extent of main soil groups by Continent 9 3. Production constraints of principal soils of the tropics Soil classification, land use, rooting depth, depth of compacted layer and field assessment of compaction status, Santa Cruz, Bolivia Critical soil bulk densities for optimal yield of various crops Commonly used cover crops for soil and water conservation, for improving soil fertility and for green manuring Regression equations relating mulch rate with worm activity Effects of mulch rate on soil physical properties and earthworm activity Effects of mulching practices on moisture conservation under pineapple plantation in Taiwan Effects of mulch and mechanical treatments on infiltration rate and preventing runoff in northern Nigeria Effects of mulch rate on sorptivity and transmissivity at different times after deforestation Effects of management of sugar cane residue on cane yield (t/ha) in Queensland, Australia 34

11 Page 13. Effect of mulch rate and slope steepness on runoff Effect of stubble cover from three crops on runoff from a single event (P = 23 mm) at Emerald, Queensland, Australia Effects of mulching on runoff and soil erosion in Zanzibar Mulching effects on runoff and soil loss at Tanga, Tanzania Effects of mulching on runoff and erosion for a young tea plantation in Kenya Effects of mulching and fertilizer application on relative cotton yield over a 10-year period in Zaire Crop yield response (t/ha) in 1976 under 22 different mulch treatments Effects of tillage methods and mulching on yield and yield components of yam tubers Yield response (kg/ha) of maize and cowpea to different mulch materials and mineral fertilizer in an Eastern Amazon Oxisol Effects of legume and grass covers on physical properties of a tropical Lixisol in Western Nigeria Effects of cover crops on soil moisture characteristics of a Lixisol in Western Nigeria General properties of the surface layer (0-15 cm) of cultivated degraded Lixisol in comparison with that under forest cover Grain yield (t/ha) as affected by various sod species, Gravimetric soil moisture (%) measurements (0-5 cm and 5-10 cm depth) under various cover crops and surface mulches during the dry season Grain yield (t/ha) as affected by ground cover in a live mulch crop production system Soil moisture conservation by mulching and growing cover crops in Taiwan Comparative effectiveness of different cover crops in controlling runoff and erosion in Malaysia Effect of cover by Mucuna utilis on runoff and soil erosion on 3 to 4 ha watersheds at IITA, Ibadan, Nigeria (rainfall of 29 August 1982) Effects of tillage methods on grain yields of soybeans and cowpeas grown at Ibadan, Nigeria Reduction in maize grain yield due to traffic induced soil compaction Effects of mulching on fresh tuber yield of yam Effects of tillage methods and mulching on fresh tuber yield of yam Fresh tuber yield of cassava as influenced by methods of seedbed preparation and mulching Effects of tillage systems and weed management on growth and grain yield of maize in Costa Rica Effects of tillage methods on grain yield of crops at Londrina and Rolandia, Parana, Brazil Mean yields of upland rice and groundnut as affected by different tillage systems in northern Thailand Effects of tillage methods and of mulching on yield of crops grown in Indonesia Effects of ridges on runoff in northern Nigeria Effects of ridges up and down the slope on erosion and runoff under cassava, maize and uncropped plot 56

12 42. Effects of tied-ridge system on crop yield in Tanzania Maize grain yield from a mechanized tied ridge system for three soils in Tanzania Effects of tillage methods on moisture conservation and grain yield of maize in in Tanzania Maize grain yield in a tied-ridge system in Burkina Faso Effect of tied ridges on crop yield (t/ha) in West Africa Effects of tied and untied furrows on (a) yields of sorghum and cowpeas and on (b) soil moisture conservation at Sebele, Botswana Effects of tied ridge system on soil and water conservation and on grain yield of wheat on a Luvic Calcisol in the Shaar Haneger region of Israel Effects of cover crops on infiltration characteristics with and without worm activity Effect of cover crop and method of suppression on grain yield of subsequent maize crops Mean effects of ploughing on yields of crops in the West African dry tropical area Tillage effects on crop yields for different soils of the semi-arid regions of West Africa Effects of compaction alleviation by paraplough and chiselling on maize grain yield on a tropical Lixisol Effects of tillage methods on seedling emergence, seedling fresh weight, soil temperature and soil moisture regimes in cowpeas and soybeans Effects of raised beds and spacing on grain yield of black gram (Cicer sp.) on a Vertisol in central India Effects of ridge height on grain and stover yield of sorghum on a poorly drained soil at Abet in northern Nigeria Effects of different methods of seedbed preparation on maize grain yield from a Lixisol in western Nigeria Effects of mulching and methods of seedbed preparation on grain yield of soybean and cowpea in Relative efficiency of soil surface treatments in conserving rainfall Effects of tillage methods on saturated hydraulic conductivity of a savannah soil in Nigeria Effects of tillage methods and cropping systems on soil erosion from a savannah soil in Nigeria Effects of tillage methods on soil physical properties and crop response at Samaru, Nigeria Structure effects of weed control cultivations on a lateritic soil Effects of the width of strip cultivation on maize grain yield for an Alfisol at Kongwa, Tanzania Effects of cultivation on maize/legume yields at Morogoro Effect of different tillage methods on grain yield of maize and soybean at Morogoro, Tanzania Grain yield responses (kg/ha) of pearl millet to fertilizer application and presowing cultivation, ISC, Nigeria, rainy season 85 Page

13 68. Effect of presowing tillage treatment for pearl millet and fertilizer application on hills surviving at harvest as percentage of sown hills, ISC, rainy seasons 1985 and Grain yields of pearl millet cv. CIVT, and cowpea cv. Suvita 2, in response to tillage, timing of tillage, and rotation, ISC, Niamey, Nigeria, for rainy season Effect of cultivar, sowing arrangement and method and time of pearl millet sowing on performance of pearl millet/maize intercrops, Sotuba, Mali, rainy season Effect of tillage system on profile water content to a depth of 1 m at 2 weeks after planting in soils of semi-arid regions of West Africa Effect of tillage on subsoil root weight per plant for some soils in semi-arid regions of West Africa Effect of tillage in crop yields in the West African semi-arid tropics Grain yield of wheat as affected by deep tillage on an irrigated sandy loam soil in Nigeria Effect of tillage method on grain yield of maize and dry yield of Stylosanthes (kg/ha) from a ferruginous soil in northern Nigeria Effects of tillage methods on bulk density (g/cm 3 ) and moisture storage (mm) of a ferruginous soil in northern Nigeria Effect of cropping sequence and residue management on cowpea grain yield in the Sudan savannah of Burkina Faso Effect of seedbed preparation method and residue management on grain yield of cowpea from hydromorphic soils in Nigeria Relationships between soil properties and organic matter content for soils in semi-arid Nigeria Effects of tillage methods and stubble management on infiltration rate of two soils in Queensland, Australia Effect of no-till and conventional till on the yield of maize, soybean, sorghum and groundnut on different soils in Northern Territory, Australia Effect of tillage methods on nutrient uptake in maize stover (St) and grain (G) (kg/ha, except where stated) in semi-arid tropics of Northern Australia Effect of construction of ridges on shallow sandy loam Luvisol on crop grain yield (100 kg/ha) at Hyderabad Effect of construction of ridges on physical properties of a shallow Lixisol on maize yield at Hyderabad, India Rates of water absorption by sorghum seeds as a function of soil-water and bulk density Emergence of pearl millet as influenced by various treatments under crusted soil conditions on Luvisols in Hyderabad, India Influence of sowing and cultural methods on grain yield of pearl millet on a Luvisol in Hyderabad, India Effect of crust formation on seedling emergence of different cultivars of cotton and pearl millet in India 96 Page

14 Page 89. Effect of inter-row cultivation (shallow tillage) in addition to normal tillage on runoff and soil loss from a Luvisol at ICRISAT Centre, Hyderabad, India Runoff (mm) from two surface configurations on Luvisols at ICRISAT Centre in Effect of different tillage treatments on sorghum grain yield, runoff and soil loss from a Luvisol (Luvisol), ICRISAT, Hyderabad, India Effect on grain yield (kg/ha) of shallow tillage in addition to normal tillage for weed control on a Luvisol Effect of subsoiling on root density at 89 days after emergence of maize (cv deccan Hybrid 103) on a Luvisol Cumulative infiltration equations and correlation coefficients as influenced by different tillage treatments for two Bangladesh soils Effects of shallow tillage and land shaping on runoff and soil loss from an uncropped Luvisol, ICRISAT Centre, Effect of alternative land surface configurations on crop yield, runoff and soil loss on a Luvisol in Hyderabad, India Grain yield (t/ha) of groundnut (JL 24) and pigeon pea (HY 3) on Luvisols under different land treatments, Mittimari, India, rainy seasons 1986 and Effect of trenching and hedgerow treatments on runoff and soil erosion from a Vertic Cambisol, ICRISAT Centre, mean of 1986 and Moisture retention and infiltration rates of high subsurface mechanical impedance soils in India Effect of different tillage systems on physical properties of soils in India Effects of compaction level and tillage methods on biomass yield of rice grown on a coarse textured Ultisol in southeast Nigeria Effect of compaction of highly permeable soil on crop yield at two sites in arid regions of India Grain yield in field plots of sandy loam soil in northeastern Nigeria as affected by 5 different traffic treatments Effect of compaction and clay mixing in loamy sand on grain yield at Jobner, India Effect of row zone application of farmyard manure and mulch on yield of pearl millet and cotton on an alluvial sandy loam soil at Hissar, India Tillage effects on summer maize grain yields in 1987 from a sandy soil in Punjab, India Effects of tillage methods in Faizabad, India, on profile moisture storage in 120 cm depth at harvest of maize during Effects of tillage treatments on grain yield of wheat on sandy soils of Ludhiana, India Effects of tillage methods and surface residue management on water balance components and water use efficiency of forage maize at Faizabad, India Effects of tillage, mulch and N application rates on dry-matter yields of fodder maize during 1982 and 1983 at Faizabad, India Tillage effects on root growth and yield of winter maize in relation to water retentivity of sandy soils in Punjab, India Mean crop yields from Vertisols in Arab countries in the Mediterranean region 108

15 Page 113. Effects of tillage methods and residue management on runoff and soil erosion from Vertisols at two sites in Queensland, Australia Soil-water balance of a Vertisol in Queensland, Australia Effects of different land and water management systems on runoff, peak runoff rate and soil loss in deep Vertisols at ICRISAT Centre ( ) Effect of different tillage practices and amendments on runoff and soil loss from maize plots on a Vertisol at ICRISAT Centre, Hyderabad, India, Effect of compaction on size of cracks and inter-crack structural units (ISUs) of a Vertisol at Hyderabad, India Effects of tillage methods on draught power and net energy requirements on a Vertisol in Sudan Soil-water accumulation during fallows related to rainfall amounts in a Vertisol in Queensland, Australia Soil-water balances during sunflower fallows for reduced-till (RT) (blade plough, tined cultivators) and no-till (NT) systems in a Vertisol in Queensland, Australia Effects of tillage methods during fallowing on soil moisture storage (mm, m) in a Vertisol in Queensland, Australia Yield response to no-till on two Vertisols in Israel Effects of different depths of cultivation of a Vertisol on soil properties and sorghum grain yield in Ghana Water use (mm) (0-1.8 m) from a Vertisol in Queensland, Australia, by grain sorghum crop form planting to antithesis Effects of tillage methods during fallowing on soil moisture storage (mm, m) in a Vertisol in Queensland, Australia Effect of construction of ridges on plant growth on a Vertisol at Tabalpur, India Effect of raised and sunken beds on grain yields (100 kg/ha) of different crops on a Vertisol at Jabalpur, India Effect of different tillage practices and amendments on grain yields (kg/ha) of maize and chickpea, Vertisol ICRISAT Centre (1983/84 and 1984/85), Hyderabad, India Grain yield, runoff and soil loss under improved and traditional technologies on deep Vertisols at ICRISAT Centre, Hyderabad, India, in 9 successive years Soil-water retention (%) at different suctions after 4 years of conventional tillage, chisel plough and no-till on a Ferralsol in Brazil Effects of tillage methods on physical properties of a Ferralsol in Parana, Brazil Effect of tillage on physical properties of the soil at Saavedra, Bolivia Effects of tillage methods on nutrient concentrations of the roots and leaves of soya at Saavedra, Bolivia Effect of tillage and fertilizer treatments on soya yield during a wet season from a compacted sandy soil in Bolivia Effect of treatments on grain yield of wheat on a sodic soil in Queensland, Australia Effect of treatments on plant available water stored in the soil profile at sowing and harvest, on soil water accumulation during the fallow, and on net soil water extraction by wheat crops on a Solonetz in Queensland, Australia 129

16 Page 137. Effect of tillage systems and stubble retention on water use efficiency of wheat from a Solonetz in Queensland, Australia Comparison of some morphological properties of the surface soils (a 1 horizon or top 15 cm) of the hard, friable and sandy duplex soils prone to hard setting Effects of tillage methods on wheat yield from hard-setting soils in Australia Effects of tillage methods on soil physical properties of the 0-15 cm layer of a tropical soil in Brazil Effect of tillage treatments on the percentage of water-stable aggregates of a Haplic Acrisol in northern Nigeria Effects of method and intensity of tillage on runoff and erosion from upland soils in northern Thailand Commercial energy J) used for inputs to agricultural production Issues of agricultural sustainability in fragile lands of marginal ecoregions Tillage and soil surface management effects on indices of agricultural sustainability Specific technologies for sustainable management of soil and resources for tropical ecological regions Some examples of tillage-based technological packages for sustainable management of soil and water resources on small farms (less than 5 ha) in the tropics Some examples of tillage based technological packages for sustainable management of soil and water resources on medium-sized farms (5-25 ha) in the tropics Effects of methods of seedbed preparation and of hedgerow spacing of Leucaena and Gliridia on grain yield of maize and cowpeas Effects of contour hedges of Leucaena and Gliridia on runoff, soil erosion and total nutrient loss for maize grown in the first season (April-August 1984) and cowpeas grown in the second season (September-December 1984) Dry matter yield of various grasses and legumes, Africa Effects of crop rotations on runoff and soil erosion in northern Thailand Comparative effects of depth of ploughing and of fallowing on soil moisture storage in the profile and on maize yield Yield of sorghum as influenced by previous crop and the weed control system during fallowing Effects of cultural practices on sorghum production Effects of soil surface management during fallowing on runoff and soil erosion in Senegal Total biological yield (kg/ha), water use (mm) and water use efficiency (kg/ha/mm) of the two phases of a rotation in a single year Yield, water use and water use efficiency of barley and vetch crops in two course rotations at Breda from 1982/83 to 1987/ Amount (mm) and proportion (%) of current season s rainfall stored under fallow, or barley and vetch crops at the time of maximum recharge of the soil-water profile in the seasons 1983/84 to 1987/ Recommended methods for determination of soil physical properties Plant and soil parameters to be measured for field experiments on seedbeds Desirability scores for different aggregate size ranges for some principal physical properties of seedbeds 170

17 Page 163. Soil ecoregional guide to tillage methods for upland crops in West Africa Accumulative tillage rating index and the appropriate conservation tillage system in the tropics 172 Acknowledgements The author is grateful to Dr. C.S. Ofori and colleagues at FAO for advice on preparing the publication. The cover photograph was kindly made available by RELACO.

18 Chapter 1 The tropical environment SOILS AND CLIMATIC RESOURCES The tropics cover about 40% of the earth s surface and have a wide range of climates. Water availability to plants is undoubtedly the most important factor governing the characteristics of tropical climates and, in combination with temperature, it is used to characterize them. Köppen (1936) distinguished two broad categories of tropical climates: tropical rainforest (Af, Am) and periodically dry savannah (Aw) (Figure 1). These two categories are further subdivided depending on the rainfall amount and its effectiveness on the basis of several indices (Thornthwaite, 1948; Budyko, 1974; Hargreaves, 1977; Bailey, 1979). A more detailed climatic characterization was developed by FAO (1978) in the context of defining agroecological zones for the developing world. In this system the length of the available growing period is used as a moisture availability index and defined as the period measured in days during the year when precipitation exceeds half the potential evapotranspiration plus a period required to evapotranspire an assumed 100 mm (or less if 100 mm is not available) of water from excess precipitation stored in the soil profile. The period during which the daily mean temperature is less than 6.5 C is subtracted from the length of the period during which water is available. FAO (1993) simplified and broadened the above agro-ecological zoning method described to illustrate the distribution of the major soils resources globally (Figure 2). In the context of the tropical environment only four subdivision are relevant. These are described below. Arid zone The arid climatic zone is characterized by an LGP of less than 75 days. This climate is typical for approximately million ha, which is about a quarter of the earth s land surface that is not covered by ice. Major arid areas include the Sahara, the Kalahari, the Namibian desert and the Horn of Africa; they are also widespread in the Arabian Peninsula, central Asia, the western part of the United States, the northwastern part of Brazil and along the southwestern part of Latin America, and central Australia. When the LGP is so short, almost no agricultural crops are able to complete their normal growth cycle, severely limiting yields, if any yield is obtained at all. In addition, both rainfall and the LGP are often unreliable, and may vary considerably from year to year resulting in major difficulties in planning agricultural operations. In these

19 arid climates, crop and tree productivity is marginal or virtually zero, unless groundwater or surface water is available for irrigation in the warmer climates. Some grazing possibilities exist at the wetter fringes of this zone but usually of a nomadic type. Soil characteristics are also affected by the non-availability of soil moisture for most of the year, so the weathering process proceeds very slowly in such conditions. The migration and accumulation of soluble salts, calcium carbonate and gypsum are often the only ongoing pedogenetic process, resulting in soils such as Solonchaks and Solonetz, Calcisols and Gypsisols, depending on which salts are involved. Another phenomenon typical of this climate is the constant wind erosion, resulting in shallow soils (Leptosols) and shifting sand dunes, or exposure of rock outcrops (miscellaneous lands) and a general loss from the soil of the finer particles, which can be carried by the wind over long distances, leaving behind soils that are predominantly sandy (Arenosols). Other soils are very gravelly, or show only the beginnings of or no horizon development (Cambisols, Regosols). Seasonally dry tropics and subtropics The tropics and subtropics where the dry season lasts between 90 and 285 days and the rainfall is not concentrated in winter belong to this type of climate. This zone extends over about million ha and is the prevailing climate in large areas of South and Southeast Asia, northern Australia, the major part of Africa, and South and Central America, between the deserts and the tropical rain forests.

20 Given the wide range in the length of the dry period, the natural vegetation in this climatic zone varies widely, but is generally dominated by what is commonly called savannah, a term that embraces all the mixed tree and grass types of vegetation found in the subtropics and tropics. The climate-dependent agricultural potential of this zone is largely determined by the length and the intensity of the dry season. In general the grazing potential is high, provided animal diseases can be kept under control. A wide range of agricultural crops can be grown, from millet, sorghum and cassava in the drier parts to maize and cotton in the wetter areas. The forestry potential is largely constrained by the length of the dry season, but adapted species yield reasonably well. The relatively high temperatures throughout the year, combined with the occurrence of a more or less pronounced wet season, favours the weathering of the soil parent material, liberating silica, iron and aluminium, with consequent formation of kaolinitic clays. This clay mineralogy is typical for the Ferralsols and Acrisols that occur in the wettest part of this climatic zone, and for the Lixisols that occur in what is now the drier parts of this zone, but that started their development with more pronounced weathering under wetter conditions. Where parent rocks are rich in calcium and magnesium, formation of smectites (swelling clays) occurs by the alternation of dry and wet seasons, and results in Vertisol formation. The Arenosols of this climatic zone are mostly relatively young soils in coarsely textured alluvial, lacustrine or aeolian deposits, and their development is barely influenced by the climate. Humid tropics and subtropics This climatic zone is characterized by high temperatures throughout the year and only a short to very short dry season, and is typical of some million ha, mainly located in central and coastal West Africa, the Amazon basin, Southeast Asia and the islands of the Pacific Ocean. The long growing periods combined with high temperatures are conditions that favour tropical rain forests, and also allow agricultural activities all year round, although high rainfall at harvest time and flourishing pests and diseases may be a moderate constraint. Grazing is hampered by the rather unfavourable temperatures. Soil formation is characterized by the extreme weathering of the parent material and the formation of kaolinitic (low activity) clays. Nutrients are leached out and the inherent fertility status of these soils is generally poor. Lesssoluble compounds, such as iron and aluminium oxides and hydroxides, and coarse quartz grains remain behind. These characteristics are typical of Ferralsols (extremely weathered soils) and of Acrisols (less weathered), which together cover about 57% of the land in this climatic zone. The high rainfall also favours the quite common occurrence of soils affected by groundwater (Gleysols). Extreme weathering may, on coarse textured parent material, lead to the formation of giant Podzols (considered as Arenosols) and groundwater Podzols. Mountainous zone This climatic zone is typified by a great variability in temperature and rainfall over short distances due to large and sudden differences in altitude and aspects of slopes. The temperature is lower than in the surrounding lowlands and

21

22 the topography is characterized by steep slopes. No strict climatic limits could be worked out to define this climatic zone and it contains climates which are as diverse as those in the Himalayas and in the tropical highlands, hence land use and vegetation are not uniform in this zone, though specifically adapted species and varieties of animals and plants are common (yak, llama, teff, quinoa, etc.). Sometimes, the healthy and pleasant climate and relative civil security have encouraged dense populations, often beyond the supporting capacity of these fragile lands. The lower temperatures in mountainous areas with moderate to high rainfall are often an advantage for certain agricultural crops, but the steepness of slopes increase the risk of soil degradation if no preventive measures are taken. Forestry may thrive on the slopes at lower altitudes, while grazing can be extensive above the tree limit. At the highest altitudes a climate similar to the boreal and polar zone may prevail. At medium altitudes, soil profile development is mainly hampered by the erosive influence of steep slopes. At the highest altitudes this factor is exacerbated by the prevailing boreal or polar conditions, resulting in relatively shallow soils Leptosols and Regosols with little or no profile development except for a surface horizon. Some limited profile development can take place in Cambisols. Andosols are common in mountainous areas of volcanic origin. Tabular representation of the extent of the soil groups in these four climatic zones is given in Table 1, while the global soil distribution is depicted in Figure 3. In addition to crop growth, climate is often the most critical determinant of the sustainability of agricultural systems. Stewart et al. (1990) presented a generalized view of the effects of varying temperature and moisture conditions on the difficulty of achieving sustainability (Figure 4). As temperature increases and the amount of precipitation decreases the development of sustainable cropping systems becomes increasingly more difficult (Figure 5). SOIL-RELATED CONSTRAINTS ON INTENSIVE ARABLE LAND USE Soils of the tropics have been described in relation to their agricultural potential by Obeng (1978) in West Africa, by Sanchez and Salinas (1981) in tropical America, by Dregne (1983a) in the semi-arid tropics, by ACIAR (1984) in Australia, and in a global context by FAO (1993) (Table 2). In West Africa, soils of the rainforest and semi-deciduous rainforest with rainfall of 1270 to 2000 mm are generally gravelly, acid and of medium to low inherent fertility. About 130 million hectares of forest soils in West Africa have developed concretions or ironpans (Obeng, 1978). These soils are cultivated with a wide variety of cash and food crops. The savannah soils of regions with 500 to 1200 mm rainfall are mostly gravelly with ironpans occurring at shallow depths. In the West African savannah about 113 million ha are covered by soils with ironpans (Obeng, 1978). Soils of the savannah region of tropical America and of the Amazon basin are mostly Ferralsols, Acrisols and Plinthisols (Sanchez and Salinas, 1981). These have a low ph and are generally deficient in major plant nutrients,

23

24 especially phosphorous. Soils of tropical Australia are similar to those of tropical Asia and West Africa with identical rainfall regime. Isbell (1978) reports that the most common soils in the tropics and subtropics of Australia receiving more than 500 mm of annual rainfall are Lixisols, Luvisols, Cambisols, Acrisols and Ferralsols. Most soils supporting tropical rainforests are old and intensely weathered. Clay has been eluviated from upper horizons so the texture of the surface horizons of Ferralsols and Acrisols is generally coarser than the subsoil. Heavily-textured soils occur along the floodplains of large rivers and on areas of volcanic ash. Under natural forest, soil bulk density is generally low and soils possess a favourable structure. The good soil structure is caused by the flocculation/cementation of clay by organo-mineral complexes to form stable aggregates. The stable micro-aggregates are generally within the size range of silt to sand. Soil structure breaks down easily, however, once the forest vegetation is removed. Soils in regions with a prolonged and intense dry season dry out to about 1 m depth. They are thus subjected to high temperatures when at low moisture content. Under these conditions soils acquire an extremely hard consistency and are often compact even in their natural state. With the exception of Vertisols and Cambisols, soils of the

25

26

27 semi-arid savannah are also characterized by a coarse-textured surface horizon with a low organic matter content. They tend to develop surface seals and crusts with poor water transmission properties (Mott et al., 1979). Most upland soils with predominantly low activity clays have small plant-available water reserves (generally less than 100 mm in the root zone). The effective rooting depth for seasonal crops is shallow, either from adverse soil physical properties or from nutrient toxicity and imbalance. Therefore, crops are prone to periodic drought stress throughout the growing season. The infiltration capacity of undisturbed soils with low activity clays under forest cover is usually very high, resulting from both horizontal and vertical components of water flow. The high spatial variability of the infiltration rate of the surface horizon commonly found is due to the variable activity of the soil fauna and other biotic factors (Lal, 1987b). In general, soils in the savannah and semi-arid regions have lower infiltration rates than those within forest eco-regions. Structurally stable Andosols have also high infiltration rates. Soils containing high activity clays, e.g., Vertisols, usually have low infiltration rates. With cultivation, however, the infiltration rate of most upland soils declines because of the rapid deterioration in soil structure and susceptibility of these soils to formation of surface crust and seal. General constraints to the intensive utilization of most tropical soils are summarized in Table 3 (Stewart et al., 1990). In the rainforest and semi-deciduous forest zones with soils of low ph, the major constraints are low inherent fertility and nutritional imbalance. Although soil chemical problems are overwhelming, crop growth and yields are also limited by soil physical factors once the nutritional constraints are alleviated. Principal soil physical constraints are compaction, erosion and drought. In the semi-arid savannah, the soils tend to present severe physical problems, the principal limitations being compaction, crusting and surface sealing, high soil temperatures, low water holding capacity and droughtiness. Nevertheless, nutritional constraints can limit food production, especially when soil physical problems have been overcome.

28 Fears that intensive land-use in the humid and sub-humid tropics may give rise to soil degradation have inhibited the introduction of improved farming systems. Such fears are easily exaggerated, because there are insufficient data from well designed, adequately equipped, long-term field studies. Consequently it has become customary to refer to tropical eco-regions as fragile and to the soils as easily degraded without precisely understanding the biophysical processes involved. The use of the word fragility in relation to an eco-region refers to the ability of an ecosystem to resist drastic alterations in its biophysical processes and to revert to the state that existed before human intervention. In the context of tropical forest and savanna ecosystems fragility refers to irreversible or permanent changes or disruptions in energy balance, in water balance, in cycles of C and major plant nutrients (N, P, K, Ca, Mg and S), and in major energy pathways. Few studies in the tropics have quantified the effects of alterations in nutrient cycles and energy pathways due to changes in land use and farming practices. Similarly, the terms soil degradation and desertification are often vaguely used without the benefits of quantitative data to define them in relation to the processes involved. The term soil degradation should be defined in terms of the rate of change in soil quality in relation to crop productivity. The concept of soil quality encompasses several properties that affect its life-support processes organic matter content, biotic activity of soil fauna, soil structure, water infiltrability, porosity and pore-size distribution, cation exchange capacity, ph, concentration of toxic elements and nutrient imbalance. It is therefore necessary to quantify the rate of alteration in soil properties due to change in land use and define their critical limits in relation to crop productivity. It is necessary to define the critical limits of soil organic matter content, porosity, water and nutrient status, and tolerable level of toxic elements beyond which the yield of various crops is drastically reduced. Over what time periods under varied farming systems and soil and crop management techniques do the major soils of the tropics reach these limits? What are the critical limits beyond which a soil is liable to undergo irreversible degradation? The terms desertification and degradation are used interchangeably. Dregne (1983b) defined desertification as a process of impoverishment of terrestrial ecosystems under human impact leading to (1) reduced productivity of desirable plants, (2) undesirable alterations in the biomass, and in the diversity of the micro- and macrofauna and flora, (3) accelerated soil erosion, and (4) increased hazard for human occupancy. Once again, the terms impoverishment, undesirable alterations and hazard are too vague and should be quantified by defining their critical limits with reference to the processes involved and biological productivity. SOIL DEGRADATION AND INTENSIVE LAND USE For soil degradation to be understood and counteracted it is imperative that the cause-effect relationship is known in terms of both the factors and processes involved (Figure 6). Some important soil degradation processes prevalent in the tropics are discussed below. Decline in soil organic matter content The rapid decline in the soil organic matter content of cultivated tropical soils is a direct effect of high temperatures throughout the year. Studies have shown that the rate of mineralization of organic matter in tropical soils may be four times that in temperate soils (Jenkinson and Ayanaba, 1977). Consequently, cultivated soils in the tropics may have

29 lower levels of organic matter than similar temperate soils. A decrease in organic matter content increases the risk of erosion. Water erosion also preferentially removes soil colloids, including humified organic matter. Decreasing organic matter content also increases the susceptibility to crusting, which further enhances the risk of soil erosion. Soil erosion is also increased by the reduction of soil faunal activity that occurs with decreasing soil organic matter content. Reduction of organic matter content also affects plant-available water reserves in the soil. The favourable effects of organic matter content on water retention in tropical and sub-tropical soils have been widely reported. Decline in organic matter content reduces a soil s structural stability, renders it likely to crust, increases soil erosion hazard and decreases plant-available water reserves. There are also nutritional implications, as there is a decrease in effective cation exchange capacity, acidification and reduction in plant nutrients. Clay migration and clay pan formation Removal of clay from the surface soil horizon can lead to soil degradation. The loss of clay may be caused by eluviation of the clay to subsoil horizons. Under forest the activity of soil fauna (e.g., earthworms and termites) plays an important role in returning fine particles to the soil surface. Some termites build large nests in tropical savannah and sub-humid regions, thus effecting substantial soil turnover. Reductions in soil biological activity decrease the net

30 clay translocation to the soil surface, and increase the surface runoff and sediment transport in the overland flow. Accelerated soil erosion can reduce the organic matter and clay contents of the surface soil horizons. Translocation of clay, particularly fine clay, to the subsoil or in situ formation of clay in the subsoil can form a physical barrier to roots and the movement of water. Various processes have been suggested for the formation of clay pans or argillic horizons. The most commonly-accepted is that clay, particularly fine clay, moves from the surface (A- and E-horizons) and accumulates in the subsoil (B- or argic-horizon). For this to happen alternate wetting and drying cycles are required. In many soils loss of clay from the surface cannot account for the amount of clay in the subsoil. This discrepancy leads to the hypothesis that some of the clay is formed in situ from a re-synthesis of weathering products derived from the A- horizon. Soils which have a clay content that doubles over less than 7.5 cm depth are considered degraded. Such an abrupt increase in clay would limit rooting, particularly where the soil lacks strong structure. The presence of sodium in a clay pan or argic-horizon creates an even more limiting environment for rooting, and restricts water movement because of dispersion of the clay. Structural deterioration and soil compaction A significant process leading to soil degradation is the deterioration of soil structure and the associated restriction of water and air movement through the soil. Structural degradation related to decline in soil organic matter and clay contents and reduction in biotic activity leads to crusting, compaction, reduced infiltration rate, low available waterholding capacity, increased soil detachability, accelerated runoff and soil erosion. High erosion risk can be a direct consequence of deterioration of soil structure. Soil compaction is a more severe problem than commonly believed. Soils with low activity clays have slight or negligible swell/shrink capacity. Decline in soil organic matter content, degradation of soil structure, and drying accompanied by high soil temperatures encourage consolidation and compression. As well as these factors there is the compactive effect of heavy machinery. About 90% of the soil surface can be traversed by tractor wheels during primary tillage. The smearing and compacting action of the plough sole gives rise to pore discontinuity that inhibits water movement and root development. Soil compaction is a major constraint to continuous cultivation and high yields on many tropical soils when intensively used Lixisols and Luvisols are easily compacted and attain critical levels of soil bulk density within 2 to 3 years of bringing land under cultivation. Surface soil compaction is especially severe with mechanized farm operations. The characterization of soil compaction is particularly difficult in the case of heterogeneous gravelly soils. The usual criteria, such as bulk density, total porosity, and penetrometer resistance, are not good indicators of the conditions that plant roots experience in such soils. The void ratio, the specific volume, air/water permeability, or pore-size distribution and continuity may be better indices of plant response than bulk density. The available information on these aspects for tropical soils is scanty. In Bolivia, Barber et al. (1989) observed that several Lixisols, Luvisols and Cambisols were highly compacted (Table 4). The bulk density of some cultivated soils was as high as 1.6 g/cm 3. Critical soil bulk density values for root

31 penetration and crop growth are not known for some of the major soils of the tropics. Table 5 shows, from available data, critical soil bulk densities for a range of tropical crops. Each soil type has a narrow range of optimum bulk density.

32 Laterization Laterization is a general term to describe the process of iron accumulation in soils. Other terms that have been introduced for the iron-rich laterite materials are ferricrete and plinthite. Ferricrete is used almost synonymously with laterite. Plinthite is more specific, referring to a soft, iron-rich soil material that hardens irreversibly when exposed to several cycles of wetting and drying. In general, laterization is a process which, at its extreme, involves intense weathering (resulting in a breakdown of all minerals except quartz) and intense leaching of the soil which removes all the soluble salts, much of the silica and some of the iron and aluminum. Kaolinite is the dominant clay mineral formed as a result of the process. The iron and aluminum in lateritic material has traditionally been considered to be the result of residual accumulation from the original parent material. Recent studies have, however, indicated that at least a portion of the material may have been transported vertically or laterally. Groundwater fluctuation is suggested as a mechanism for such transport, particularly in the case of plinthite. Some workers consider laterite to be a very thick B-horizon of an intensely weathered soil, for laterites and plinthites are associated with very old, stable surfaces and a wet tropical climate. The weathering zone in these landscapes is commonly many metres thick. As a result of intense weathering such soils are very low in plant nutrients. The high iron and aluminum content results in complex fertility management problems and, where the iron has hardened into ironstone, roots are restricted to the upper portion of the soil. Lateritic soils are widespread in the humid and subhumid regions of Africa, and may also occur locally in drier areas as relics of a previous climatic regime. Hard-setting Some structurally-weak soils of the tropics have a tendency to harden during dry periods and exhibit high strength. High strength and hard consistency make tillage difficult, restrict timing of cultivation, and impede or restrict root development. Soils with dense, sometimes weakly-cemented surface horizons, are identified as hard-setting. In Australia where this soil condition has been most extensively observed and studied, hard-setting is associated with deeply-weathered soils that are low in organic matter content and phosphorus. This low fertility has resulted in limited biological activity, decreased biomass production, and low organic matter content. These soils are characterized by the massive, single grain or weak pedality of the surface horizons which tend to crust. The dispersion of the clay fraction often associated with these soils is attributed to sodium, which hydrates the clay particles, diffuses the double layer and disperses the clay. The role of sodium in hard-setting is supported by the occurrence of hard-setting properties in many soils classified as Solonetz. This association is, however, not universal, because hard-setting surface horizons are also found in some podzolic soils. Plant establishment is difficult on hard-setting soils because of the low-porosity (particularly with respect to macropores), and fine laminations in the upper few centimetres. The low porosity and limited vegetation result in high runoff. In some areas under natural vegetation, more than one-third of the annual rainfall is lost to plants. Hard-

33 setting soils often have a textural contrast, with a loamy surface horizon and a high-clay content in the subsoil which adds to the problem of vegetation establishment. When cultivated, the hard-setting characteristics can be destroyed temporarily but usually reappear after a single wet season. One of the contributing factors to the formation of the low-porosity layer may be a low phosphorus status. In Australia, the application of superphosphate and establishment of grasses and legumes have reduced the incidence of crusting. Gypsum has also been found to be an effective ameliorant for some of these soils. Although hard-setting occurs under native vegetation, it is frequently associated with cultivation. Intensive cultivation or over-stocking of pastures may deplete organic matter and lead to the loss of soil structure and development of the dispersion characteristics of hard-setting. Soil erosion Soil erosion causes widespread soil degradation and desertification. Cultivated soils in the humid and sub-humid tropics are at risk of accelerated erosion. Water erosion is particularly severe in semi-arid regions. Both wind and water erosion are severe in arid regions. FAO (1984) estimates that, if soil erosion continues unchecked, between 1975 and the year 2000, about 18% of the rainfed cropland of the developing countries of the tropics will be lost and that rainfed crop productivity will fall by about 29%. Despite its obvious dangers there is a lack of quantitative data on current rates of erosion on various soils and eco-regions, their effects on crop production under different management systems, and on the acceptable or tolerable limits of soil erosion. Nutrient depletion and leaching Nutrient depletion by crop removal, leaching, runoff and erosion, and volatilization results in the loss of soil fertility unless nutrients are replaced. The leaching losses of nutrients are increased when deep-rooted perennial forest vegetation is replaced by shallow-rooted arable crops.

34 Chapter 2 Predominant farming systems of the tropics There is a wide range of farming systems in the tropics. They can be grouped into three principal categories treebased, animal-based and crop-based (Figure 7). The main farming systems are briefly described below. TRADITIONAL FARMING SYSTEMS Traditional farming systems have not changed much since the early 20th century. Shifting cultivation and bush fallow rotation are widely used in the tropics (Okigbo, 1978). These subsistence systems are diverse and based on low inputs. Several crops are grown simultaneously on the same piece of land. In some regions, especially on marginal or steep lands in densely populated regions, traditional slash and burn farming has been destructive and has caused severe soil and environmental degradation. In most of the lowland tropics with sparser population and relatively fertile soils, however, traditional systems have proved to be ecologically stable despite the minimal inputs of the shifting cultivators (Greenland, 1975). The degree of severity of soil erosion and water runoff under traditional farming systems depends on the soil, land-use intensity, relief and cultural practices adopted. Runoff and erosion are generally reduced if the fallow period is long enough to restore soil physical properties and increase the soil organic matter content. Traditionally, farmers abandon the land when crop yields are too low either because of prevalence of pests or because of deterioration in soil quality. The rate of decline in crop yield on soils under traditional farming depends on many factors soil properties, crops grown, prevalent climate and soil management practices. A study by Allan (1965) in East Africa indicates that, on soils of high inherent fertility, it took 20 years of continuous cultivation to cause severe yield decline. On soils of low inherent fertility, however, yield declined under traditional farming in 1 to 2 years. There are urgent reasons to improve productivity even on soils where yield decline does not occur rapidly. The system is ecologically stable and works as long as the farmers are willing to remain at the subsistence level.

35 SOME ALTERNATIVE FARMING SYSTEMS Cropping and farming systems that have been tried as alternatives to shifting cultivation and bush-fallow rotations include the following: Large-scale mechanized farming Mechanized farming has been tried widely in Africa, but with limited success. Major physical obstacles to intensive cropping of crops such as maize, cowpea and groundnut by motorized farm operations in tropical Africa are severe soil compaction, accelerated erosion and low soil fertility. In attempts to grow a second crop, farmers often try to speed up the harvesting procedure. Mechanized harvesting is particularly damaging if the soil is wet. Ranching and pasture development Ranching is a popular system in the humid tropics. More intensive pasture development is successful only when the initial soil fertility is high and the stocking rate is carefully controlled, otherwise severe economic and environmental problems arise. Excessive and uncontrolled grazing can cause at least as many problems of soil and environmental degradation as mechanized farming. Excessive grazing depletes the vegetation cover, changes the species composition, exposes soil to high-intensity rains, compacts the surface soil layer and decreases the infiltration rate of the surface soil layer. Forestry Tree-based farming systems are very common in the humid tropics. It is generally believed that forest plantations and tree crops can provide ecologically stable systems. The belief in the restorative effects of forestry may be an overgeneralization because, with careless management, severe soil compaction and accelerated soil erosion and land degradation occur. Plantations make heavy demands for plant nutrients on the soil but, with good management,

36 the chances of attaining ecologically compatible systems are better with trees than other crops. Tree crops protect the soil against raindrop impact and insolation by their continuous cover. Tree crops are key components of some agricultural production systems. These include: Fruit and nut orchards, plantations yielding oil, rubber, gums, tannins and drugs. Forest plantations providing timber and related forest products. Trees and shrubs maintained in natural or planted fallows to recycle nutrients for use by subsequent arable crops. Trees grown in association with food crops on farms and as plantation crops associated with animal husbandry. The Taungya system, or agrisilviculture, in which food crops are associated with forest trees for the first 2 to 3 years. Forest trees as components of integrated land development plans.

37 Chapter 3 Tillage systems DEFINITIONS The term tillage is a generic term and is used broadly. Tillage embraces all operations of seedbed preparation that optimize soil and environmental conditions for seed germination, seedling establishment and crop growth. It includes: mechanical methods based on conventional techniques of ploughing and harrowing; weed control using chemical herbicides and growth regulators; and fallowing with an aggressive cover crop that can be easily controlled for direct seeding through its residue mulch. There is a wide range of tillage systems used in the tropics, the chief of which are described below. Traditional tillage Farmers in the tropics employ several traditional methods of seedbed preparation. Traditionally, weeds and bush regrowth are slashed manually and left on the soil as mulch or are burnt in situ. The land is then hand-hoed, often superficially. Farmers also make mounds or ridges, often with manually-operated hoes or with equipment drawn by draught animals. Mounds, ridges and other forms of raised beds are widely used throughout the tropics. On poorlydrained soils in southeastern Nigeria, large mounds are constructed. These are often 3-4 m in circumference, and about 1 m high. Various crops are grown on top of the mounds, and rice is sown between them. Raised beds are specifically constructed to grow vegetable crops in swamps in eastern Cameroon. A slightly modified version of the mound system is used in Zaire. In the Mofuku system, crop residue and weeds are buried in the mound. The residue is later burnt. A similar system using ridges and burning residue is used in Ethiopia. Traditionally, farmers mulch the mounds with crop and weed residues. The practice of mixed cropping provides a continuous ground cover that protects the soil against erosion and improves soil temperature and moisture regimes. The technique of building mounds is also useful in concentrating nutrient-rich surface soil. With the native method of cultivation, mounding is clearly beneficial, especially when mixed cropping is practised with little or no fertilizer input. This practice is only slightly superior to a mulched flat seedbed on which inorganic fertilizers are applied. On some shallow soils, crops are sown in depressions rather than on mounds. The depressions help conserve water and minimize risks of drought. Plough-till This system is based on mechanical soil manipulation of an entire field, and involves mouldboard ploughing followed by one or two harrowings. Plough-till embraces primary cultivation based on ploughing or soil inversion, secondary

38 cultivation using discs, and tertiary working by cultivators and harrows. These tools are often drawn by animals or by tractors and other mechanically-powered devices. The mechanical soil disturbance involved increases the risk of erosion. Ploughing removes the vegetation cover and exposes the soil to rainfall, wind, and overland flow. The technique gives a weed-free seedbed, incorporates fertilizer and improves soil conditions. The effects of tillage methods on soil properties and on the erosion risk are hard to generalize. The effects vary depending on basic soil properties. For example, where the soil has favourable structure with a high proportion of water-stable aggregates, and is permeable, mechanical soil disturbance is likely to increase the risks of soil erosion. On the other hand, where the soil has a smooth crusted surface and compacted subsoil horizons, massive non-porous unstable structure, carefully-judged, timely mechanical tillage is likely to decrease the risks of soil erosion, at least temporarily. Conservation tillage (CT) The term conservation tillage has been used for varied tillage practices under a range of conditions (Mannering and Fenster, 1983). The vague use of the term for differing situations has created confusion and misunderstanding. The term encompasses a broad spectrum of practices ranging from no-till to intensive tillage, depending on soil conditions. Conservation tillage has been defined as any tillage sequence that reduces the loss of soil or water relative to plough-till ; often it is a form of non-inversion tillage that retains a protective layer of mulch. The key techniques used for soil and water conservation are (1) residue mulches and (2) an increase in surface roughness. Increases in surface roughness can be achieved by chisel ploughing, strip tillage, ridge-furrow systems, or tillage methods that cause soil inversion. If done at the right soil moisture content and with the right equipment, inversion tillage can produce an almost ideal rough seedbed. Adequate supplies of mulching materials are not always available. In such situations, conservation tillage techniques may include contour ridges, tied ridges, camber bed system, and broadbed and furrow systems. In the light of the above, a broad definition of conservation tillage, such as that of Wittmus et al. (1973), seems appropriate: conservation tillage includes tillage systems that create as good an environment as possible for the growing crop and that optimize the conservation of soil and water resources, consistent with sound economic practices. Some commonly used practices used under the generic term of CT are described below: No-till: When a crop is planted directly into a seedbed that has not been tilled since the previous seedbed it is called a no-till or no-tillage system. The maximum amount of crop residue is retained on the surface, and weeds are controlled by chemicals, by residue mulch, by using an aggressive cover crop, or by a combination of these methods. If a soil is disturbed at harvesting, as in the case of root crops, some workers argue the system is no longer a no-till system. According to the definition given here, however, soil disturbance at harvest is allowed in a no-till system. Minimum-till: The term minimum-till has caused the greatest confusion because the minimum cultivation required to grow a crop successfully varies from zero to a complete range of primary and secondary tillage operations

39 depending on soil properties and crops. It is commonly defined as the minimum soil manipulation necessary for crop production or meeting tillage requirements under the existing soil and climatic conditions. It often means any system that has few tillage requirements. It may also mean tillage of only part of the land e.g. strip tillage or zonal tillage. Minimum-till may also refer to a stale-bed system in which the soil is ploughed at the end of the previous crop cycle. The crop is then seeded with a minimum of seedbed preparation performed at the onset of the next rains. This is commonly recommended for soils in the semi-arid tropics in west Africa (Charreau and Nicou, 1971). Mulch tillage: A tillage system that ensures a maximum retention of crop residue on the soil surface is called mulch tillage or stubble mulch farming. The soil is prepared in such a way that plant residues or other mulching materials are specifically left on or near the surface. Mulch tillage is a broad term. It includes practices such as no-till, disk plant systems, chisel plant systems, and strip tillage systems. When a grain crop is seeded through the mulch of a chemically killed cover crop, it is called sod seeding. If the crop cover is untreated or only temporarily suppressed, the system is called live mulch. When a cover crop, usually a legume, is specifically grown within the cropping cycle to produce mulch material, the system is called planted fallow. Another variant of planted fallowing, practised in North America, is referred to as summer fallow or ecofallow. The latter is a system of fallowing in which weed growth is restricted by shallow cultivation or by using herbicides to conserve soil moisture. Crops are grown every other year or once in 3 years. This type of cropless fallow is mostly used in arid climates to conserve soil moisture. Mulch tillage is also practised within agroforestry systems. A common practice is alley cropping, where annual food crops are grown between widely spaced hedges of perennial shrubs. The hedges are planted on the contour and are regularly pruned to provide mulch. Ridge tillage: The practice of planting or seeding crops on ridges is widespread in both temperate and tropical climates. The crop row may be planted on the ridgetop, along both ridge sides, or in the furrow. Ridge tillage facilitates mixed cropping systems in which more than one crop can be grown simultaneously in the same plot of land, a common practice throughout the tropics and subtropics (Bradfield, 1970). The ridges may be made every season. Alternatively, in a semi-permanent ridge-furrow system, necessary repairs are done at the onset of a new cropping cycle. The ridges may be on the contour with graded furrows draining into a grassed waterway, or the ridges may have short cross-ties to create a series of basins to store water. The latter system with cross-ties is called the tied-ridge system. QUALIFYING CRITERIA FOR CONSERVATION TILLAGE The criteria to be met for a method of seedbed preparation to be considered as conservation tillage include: the presence of a crop residue mulch; effective conservation of soil and water; maintenance or improvement of soil structure and organic matter content;

40 maintenance of a high and economic level of productivity; minimum need for chemical amendments and pesticides; preservation of ecological stability; minimum pollution of natural waters and environments. These criteria can be satisfied by various cultural practices, including: adopting non-inversion or no-till systems; using crop rotations based on cover crops, buffer strips, and/or agro-forestry; improving infiltration capacity of soil by crop rotation with deep-rooted perennials; increasing surface detention capacity of the soil through using rough and cloddy seedbeds and ridge-furrow systems; enhancing the biological activity of the soil fauna; reducing cropping intensity to conserve soil and water resources and improve soil fertility.

41 FACTORS AFFECTING THE CHOICE OF TILLAGE SYSTEM The choice of an appropriate tillage system depends on factors listed in Figure 8. In the short-term, choice of tillage method also depends on the objectives of seedbed preparation. For the immediate crop these are to optimize soil and environmental conditions for seed germination, seedling establishment and crop growth (Figure 9). In the long-term, however, methods of seedbed preparation should aim to maintain or improve soil organic matter content and soil structure, and enhance pore stability and continuity. The methods adopted should meet both short- and long-term requirements. TILLAGE METHODS AND SOCIO-ECONOMIC FACTORS These include farm size, land tenure system, credit availability, source of power, educational status of the family, family structure, and the role of the sexes. Farm Size: Most farms in the tropics are less than 5 hectares and it is difficult for them to adopt improved tillage tools and herbicides. In some countries, as much as 80% of the population depends on agriculture. It is desirable that other kinds of jobs are developed so that pressure is taken off the farms and farm size can be increased to allow farmers to adopt modern tools and improved technology. Land Tenure Systems: Community ownership and transient rights to use land do not encourage investment to improve farm conditions. Where the land is owned by the village chief, the right to cultivate it can be withdrawn at will. Credit Availability: Credit on fair terms is rarely available for resource-poor farmers to adopt improved technology. Power Sources: It is estimated that as much as 80% of the arable land in sub-saharan Africa is cultivated manually. Draught animals are rarely used, especially in the tsetse fly-infested areas of the humid forest regions. Large farms cannot be worked effectively by hoe and machete. Education: Most farmers are illiterate, and often unaware of improved systems of soil and crop management and their availability. This is not to say that uneducated farmers cannot comprehend the usefulness of science-based technology. Family Structure: Most farm families are large. Although the causes and effects are difficult to separate, the availability of cheap labour may hinder the adoption of labour-saving technology. Role of Women: In sub-saharan Africa, food crop farming is predominantly done by women. Credit is not usually available for women farmers. The design of tools and planning of tillage systems should take their role into account.

42 Chapter 4 Soil surface management Three objectives of cultivation are to (i) optimize infiltration, (ii) enhance rooting depth, and (iii) improve soil-water management. There are various tillage systems designed to achieve these objectives. The strategies for increasing water infiltration into soil outlined in Figure 10 are based either on in-creasing the infiltration capacity, or increasing the time for rainwater to infil-trate into the soil. The other tillage practices outlined in Figure 10 are described in subsequent sections. MULCH FARMING Mulches are non-soil materials placed on the soil surface to protect it against raindrop impact and erosion, and to en-hance its fertility. Mulch farming is a key method of improving traditional agriculture. There is a wide range of mulching techniques available Figure 11). The usefulness of a particular method depends on several factors, including environmental and socio-economic aspects. Crop residue mulches are useful. They help to conserve the soil, maintain the quantity and quality of water running off the agricultural land, regulate soil temperature and moisture regimes (Figure 12), improve soil physical conditions by enhancing biological activity, and increase soil fertility by adding nutrients. Crop residues have many other uses so they are often in short supply. They can be used as fuel, to feed livestock, for fencing or roofing and various other purposes. There are four ways of ensuring adequate supplies of mulch: The residue from the previous crop can be used as mulch. For this to be successful, however, it is necessary that an appropriate rotation should be followed, so making sure that at least one crop leaves enough residue to be used as mulch. In general, cereals produce considerable amounts of a slow-to-decompose straw that can

43 be used as mulch. These include maize, sorghum, rice and millet in the tropics, and wheat, barley, and oats in subtropical and montane climates. Leguminous crops, such as soybeans and cowpeas, do not produce enough residue or it decomposes more quickly (De Vleeschauwer et al., 1978; Lal et al., 1980). Crops can be specifically grown in the rotation to produce mulch. Such crops are usually referred to as cover crops. They include kudzu (Pueraria phaseoloides), stylo (Stylosanthes guianesis), centro (Centrosema pubescens) and mucuna (Mucuna utilis). Grasses often have more beneficial effects on soil structure than legumes (Pereira et al., 1958), but some are too persistent (Lal et al., 1978). Useful grasses for the tropics include guinea grass (Panicum maximum), elephant grass (Pennisetum purpureum), molasses grass (Setaria sp.) and bahia grass (Paspalum notatum). Grasses and legumes grown for mulch are listed in Table 6. Subsequent food crops can be grown through the mulch produced by the cover crop. Some cover crops die naturally and produce a nicely mulched seedbed ready to be seeded using a no-till system. Crops are also grown through the live mulch of a cover crop that has been only partially suppressed by mowing, slashing or by using herbicide. A form of mixed cropping specifically designed to produce mulch can be adopted. This may be in the form of alternative strips planted to food and mulch-producing crops. The vegetation in the mulch strip is harvested regularly and used as mulch for the food crop. Alternatively, a narrow but contoured hedge of grass and woody shrubs can be used to control runoff and erosion and to provide a reliable source of mulch.

44

45 The mulch material may be brought in from elsewhere. For high-value horticultural crops, it may be economical to bring in natural material from some distance. Synthetic materials (polythene sheets or emulsified soil conditioners) can be used as mulch. Polythene sheets are used as mulch for pineapple production in Hawaii (Ekern, 1967), tomatoes in the United States (Geraldson et al., 1966), sisal in Tanzania (Hopkinson, 1969), and seed yam in Nigeria. Petroleum products, such as bitumen, are also used as mulch under special circumstances (Blore, 1964; De Vleeschauwer et al., 1978). Transporting large quantities of mulch for large-scale cropping can be economically prohibitive. Beneficial effects of mulch farming Mulch farming has conclusive ameliorative effects on physical, chemical and biological soil properties. Some effects are briefly described below: Improved Soil Structure: Mulching has a notable effect on the physical properties of tropical soils. Lal et al. (1980) developed the following empirical relations between mulch rate (X in t/ha) and soil properties:

46 Percentage water-stable aggregates (>0.5 mm) = X X 2 ; r = 0.98** Dispersion ratio = 26.9 exp (-0.09X); r= 0.97** Erosion ratio = 71.9 exp (-0.09X); r = 0.96** Improvement in soil structure caused by mulching is partly due to prevention of surface crusting by checking raindrop impact, and partly due to enhancement of activity of soil faunal activity, notably that of earthworms. Tables 7 and 8 show linear increases in earthworm activity with increasing application of mulch, and improvement in soil physical properties with increasing activity of earthworms. Better Water Conservation: Mulching improves water conservation in two ways. It reduces water runoff, and decreases losses due to soil evaporation. Table 9 shows that in comparison with clean cultivation the use of rice straw mulch increased the soil moisture content by 22 percent. In northern India, Prihar et al. (1979) reported significant beneficial effects of mulch in soil-water conservation. In Nigeria, Lal (1975; 1978) recorded more soilwater in the root zone of mulched than unmulched plots (Figure 12).

47 For soils in the semi-arid region of northern Nigeria, Lawes (1962) observed significantly higher infiltration rates on plots with mulch than on plots where the crust was broken mechanically. Table 10 shows that the maximum infiltration rate and the percentage of total rainfall penetrating the soil surface were highest for mulched treatments. On Lixisols in western Nigeria, Lal (1976c) observed a higher infiltration rate on mulched compared than on unmulched plots. Furthermore, the infiltration rate normally increases with increasing levels of mulch applications. Both sorptivity and transmissivity increase with increasing mulch application, as is evidenced by the data in Figure 13 and Table 11.

48 In general the infiltration rate is governed more by the proportion of surface covered by mulch than by the volume of the mulch. It is assumed, however, that the greater the quantity of mulch material, the greater the proportion of soil surface covered. Barber and Thomas (1981) show that the proportion of the ground covered increases exponentially with the weight of mulch applied (Figure 14). Table 12 shows significant effects of sugar cane residue on infiltration rate. For well-drained soils, surface mulch improved infiltration rate. The effects were, however, inconclusive for poorly drained soil (Wood, 1991). The protective effect of mulch in reducing runoff is generally related to the quantity of mulch. On permeable, well-drained soils, runoff and erosion usually decrease exponentially with increasing mulch rate. Table 13 shows that the mean annual runoff losses were 393, 81, 30, and 13 mm for mulch rates of 0, 2, 4 and 6 t/ha, respectively. Experiments conducted in Australia by Yule (1987) showed that runoff losses from sunflower, sorghum and wheat decreased with increasing stubble cover (Table 14). Mulching also reduces losses due to evaporation. Soil evaporation can be considered to take place in three stages. Evaporation is usually greatest during the first stage. Mulching drastically reduces the rate of evaporation

49 during this stage, prolongs its duration, and keeps soil moist for longer. Mulching also decreases evaporation losses by reducing the effects of insolation at the soil surface. The maximum soil temperature can be reduced by mulching by as much as 20 o C at 5 cm depth (Figure 15). EROSION CONTROL Soil erosion generally decreases with increasing rate of mulch application. For soils of low permeability the mulch rate may have no significant effect on runoff, but soil erosion may decrease with increasing mulch rate even on slowly permeable soils. Experiments conducted in East Africa indicate that residue mulch applied at 6 t/ha effectively controlled runoff and erosion even on steep slopes of up to 22 percent (Tables 15 and 16). The effectiveness of mulch in reducing runoff and erosion has also been demonstrated in Zanzibar and Tanga, Tanzania.

50 In Zanzibar, Khatibu et al. (1984) recorded more water and soil loss from bare than from mulched plots (Table 15). At Tanga, Tanzania, Ngatunga et al. (1984) evaluated the effects of 6 t/ha of straw mulch on runoff and erosion compared with unmulched treatments. Their data in Table 16 show that mulching provided an effective erosion control even on steep slopes of up to 22%. Othieno (1975) and Othieno and Laycock (1977) report notable reductions in runoff and erosion under young tea plantations when crop residue mulch was applied (Table 17). Other examples of the usefulness of mulch for erosion control are given by Roose (1977; 1980) and Shaxson (1975). The beneficial effects of mulch in controlling wind and water erosion have also been validated in India by Rama Mohan Rao (1973), and Manipura (1972); and in Chile by Pena (1981). In Thailand, application of mulch decreased water runoff from 19.0 to 2.6 mm per 30 minutes, decreased runoff rate from 63.3 percent of total rainfall to 8.6 percent, and decreased soil erosion from to 5.8 g/m 2 (Miyake, 1986). Data from Queensland, Australia, for Vertisols show that stubble mulch reduced runoff by 39 percent and soil movement by as much as 92 percent. The data showing 6-year average effects on runoff and soil loss in relation to residue management showed similar trends.

51 In Taiwan, Wang (1984) observed that soil erosion control in citrus orchards mulched with 10 t/ha of weeping love grass was as good as with level retention terraces. Mulch was more effective in erosion control than growing a complete cover of bahia grass. SOIL FERTILITY IMPROVEMENT Regular, substantial additions of crop residue enhance soil fertility and increase yields. In Zaire, Jurion and Henry (1969) observed that mulching maintained soil fertility better than the application of inorganic fertilizers (Table 18). Cotton (Gossypium hirsutum) yields on unmulched plots without fertilizer decreased to 12% of the initial yield after 10 consecutive years of cultivation. In contrast, yields on mulched plots without fertilizer were 87% of the initial yield, about eight times greater than that of the unmulched treatment. The beneficial effects of mulch on crop yields

52 have also been demonstrated for Lixisols (Lal, 1975; 1978; Maurya and Lal, 1980) and Ultisols (Okigbo, 1965, Okigbo and Lal (1982) in West Africa. The use of organic mulches as fertilizers is especially important for small landholders (1-2 ha) who neither can afford commercial fertilizers nor are sure of their effectiveness. Leguminous mulches can supply nitrogen and other nutrients needed to boost yields and are particularly useful. Okigbo and Lal (1982) compared yields of cassava, maize, cowpea, and soybeans under 22 different mulch materials. The data in Table 19 show that rice husk increased maize yields by 0.7 t/ha and cassava by 12 t/ha. On an Acrisol in southeast Nigeria Maduakor et al. (1984) observed that mulching significantly increased the yam tuber yield (Table 20). On an Acrisol in southeast Nigeria, Hulugalle and Lal (1984, unpublished data) observed that the yield of cocoyam tubers was increased from 2.8 t/ha to 5.75 t/ha by mulching. It is well known that organic farming and the use of crop residue mulches have been the basis of sustained and intensive agriculture in China and elsewhere in tropical Southeast Asia.

53

54 Sanchez and Salinas (1981) observed only slight increases in the yield of maize mulched with Melinis mintriflora compared with an unmulched control on a Haplic Ferralsol near Brasilia, Brazil. In another study at Yurimaguas, Peru, Sanchez and Salinas reported positive effects of mulching with Panicum maximum on the grain yields of many crops grown on Typic Paleudults. Mulching increased the yield of maize by 10.7%, of soybeans by 2.2%, and of groundnuts by 2.8%. The yields of upland rice and cowpea, however, were slightly lowered. These authors reported that mulching with Pueraria phaseoloides straw was more beneficial than that with Panicum, and produced 80% of the maximum yields without the addition of chemical fertilizers. Schoningh and Alkamper (1985) studied the yield responses of maize and cowpea to different mulch materials on a Ferralsol in Brazil. Their data (Table 21) show that mulches with high decomposition rates improved the yield of maize more than those resistant to decomposition. The reverse was the trend, however, with cowpea. In addition to yield, mulches positively influenced soil mesofauna, soil organic matter content and CEC. The soil moisture storage in the 0-10-cm layer was generally more in mulched than unmulched treatments, and mulching drastically decreased the soil surface temperature. COVER CROPS AND PLANTED FALLOWS When there is insufficient crop residue to provide a complete mulch, an appropriate cover crop or planted fallow can be included in the rotation. In addition to providing mulch, planted fallows are more effective in restoring soil physical and nutritional properties than long bush fallows (Tables 22 and 23). Organic matter is built up and the soil structure improved even on eroded and degraded lands by growing appropriate planted fallows for 2 to 3 years (Wilkinson, 1975; Juo and Lal, 1977; Lal et al., 1978; 1979; Wilson et al., 1982). Figure 16 shows that compared with the preseeding control, water infiltration in an eroded Lixisol was significantly improved by 2 years growth of Psophocarpus, Centrosema, Pueraria, Stizolobium and Stylosanthes. The soil organic matter content, available phosphorus and exchangeable cations were also increased (Table 24). Growth is generally better after a planted legume fallow than after grass (Lal et al., 1978). Experiments conducted at IITA showed the beneficial effects of in situ mulch from the cover crop Mucuna utilis. Maize grown without Mucuna mulch was stunted and prone to drought, suffered from high soil temperatures, and produced less grain even when 60 and 90 kg N per hectare were applied. Leguminous crops have long been grown in association with plantation crops. The most commonly used legumes for this purpose are Calapogonium muconoides, Centrosema pubescens, Indigofera spictata, Pueraria phaseoloides, Stylosanthes guianensis, Dolichos hosel and Glycine javanica. The beneficial effects of grass and leguminous crops are well established for soils in semi-arid West Africa (Pereira et al., 1954). Legumes are also grown as improved fallows. Pereira et al. (1958) confirmed that fallowing with teff grass (Eragrostis abyssinica) improved rainfall infiltration and decreased runoff and erosion. Planted fallows are increasingly being used in rotations based on seasonal food crops. Crops that have been found useful as fallows before no-till grain food crops are Centrosema pubescens, Pueraria phaseoloides, Stylosanthes guianensis, Psophocarpus palustris and Mucuna utilis. The effects of various legume covers on the following grain and root crops evaluated at IITA are shown in Table 25. In general, crop yields were better with legumes than with grasses. If grass covers can be effectively

55

56 suppressed, they give improved crop yields compared with the weed-fallow control. Another factor responsible for low yields in some cover crops is the competition for soil moisture. The data in Table 26 show that some deep rooted cover crops (e.g. Indigofera, Axonopus, Desmodium) can rapidly deplete soil moisture reserves. LIVE MULCH A fallow management system of mixed cropping in which the main crops are grown through untilled mulch is termed live mulch (Lal, 1975). A small strip is opened, with or without using herbicides, to grow a seasonal food crop. The system works if the fallow crop does not compete for light, moisture and nutrients. Wilson et al. (1982) record severe yield reductions caused by unsuppressed climbing cover crops. With a low-growing non-climbing cover crop

57

58 favourable grain yields can be obtained under suitable management (Akobundu, 1980; 1982) (Table 27). In regions with annual rainfall of less than 2000 mm and with periodic drought stress in the growing season, live mulch is likely to compete for soil moisture and suppress the yield of the food crop. Observations of soil moisture under various live mulches show that some deep-rooted mulch crops can deplete soil water from 120 cm depth (Figure 16). Table 26 com-pares soil moisture contents for 0 to 5 cm depth on a Lixisol under different live and residue mulches and bare fallow treatments. Unless properly killed by chemical or mechanical means, live mulches may become aggressive weeds and outgrow crops being grown through them. Live mulch systems may, therefore, be suited more for humid than arid or semi-arid regions. Table 28 shows no differences in moisture reserves of soils under cover crops and residue mulch in the humid climate in Taiwan. The live mulch system is more suited for perennial than seasonal crops, e.g., legume crops under oil palm, rubber or citrus plantations.

59 One disadvantage of live mulch systems is the higher incidence of diseases and insects. Light or controlled grazing can be used to minimize competition with a grain crop. A ley farming system based on the no-till technique has been deve-loped in northern Australia under which food crops are grown through lightly grazed cover crops (McCown et al., 1985). The effectiveness of a cover crop in controlling runoff and erosion varies depending on the canopy cover and biomass produced, root system, etc. Table 29 shows that Calopogonium was the most effective cover in reducing erosion in Malaysia. Calopogonium may not, however, be suitable for other soils and climates. In southwestern Nigeria, Mucuna proved to be the most effective cover crop. Effects of growing Mucuna cover on runoff and soil erosion for a Lixisol are shown in Figures 17 and 18, and Table 30. Growing grass and leguminous covers is known to improve soil structure and water infiltration rate (Pereira et al., 1954, 1958; Wilkinson, 1975). In northern Australia, Bridge et al. (1983) reported that Stylosanthes improved infiltration and macro-porosity. In Chinchina, Colombia, De Castro (quoted by Primavesi, 1982) reported that grass cover was almost as effective in reducing runoff and soil erosion as undisturbed forest cover. Figure 17 compares the water infiltration rate on two watersheds cleared in Both were used for growing maize-cowpea in an annual rotation cycle. In 1986 one of the watersheds was sown to Mucuna while the other was kept under maizecowpea rotation. The cumulative infiltration under the Mucuna cover was about double that under the maizecowpea crop.

60

61

62 NO-TILL SYSTEMS No-till systems avoid mechanical seedbed preparation and rely on mechanical slashing, herbicides and/or cover crops to kill or suppress weed growth. The crop is seeded in undisturbed soil and residue mulch is firmly anchored to the ground. Severe accelerated soil erosion and the high costs of energy associated with plough-based methods of seedbed preparation have led to a wide adoption of no-till systems for row crops, especially in temperate climates. Such systems are, however, applicable to some soils, crops and ecological regions only. No-till systems are suited for soils that are highly susceptible to erosion, are well drained and have a low water-holding capacity, are prone to surface sealing and compaction, and have limitations of supra rather than suboptimal soil temperatures during the seedling stage (Lal, 1985b; Allmaras and Dowdy, 1985; Cannell, 1985). Successful crop establishment with a no-till system also depends on soil conditions and land-use history. Under a successful no-till system: erosion is reduced to a level below the limit of tolerable soil loss; soil-water conservation significantly decreases the frequency and intensity of short (10-15 day) midseason droughts; soil structure, infiltration rate and soil organic matter contents are maintained at, or improved to, a level favourable for crop production; satisfactory seed germination and seedling emergence and acceptable weed control are achieved; acceptable economic returns and satisfactory yields are sustained over the years. The last point needs further clarification. A lower level of yields is acceptable provided that it is economic and that it is sustained without causing degradation of soils and environments. In the tropics it is desirable not to have an excess of mulch. Another factor is the susceptibility of soils in the tropics to compaction and hard-setting. Considering these two factors, a no-till system is likely to succeed better on the upland soils in the humid and sub-humid than in the semi-arid or arid tropics (Lal, 1985a). There are many examples of the successful use of no-till systems in humid and sub-humid regions with one or two long growing seasons, which permit the production of sufficient biomass to be used as mulch. The no-till system has been successful in West and Central Africa for cultivation of maize, cowpeas, soybeans, yams and cassava (Lal, 1983). Maize is grown in regions with an annual rainfall of mm. Experiments in Nigeria show that satisfactory long-term yields are obtained provided that maize is grown in rotation with legumes, weeds are adequately controlled, crop residue mulch is available, and soil compaction is avoided (Rockwood and Lal, 1974; Curfs, 1976; Wijewardene, 1981; Lal, 1982; Osuji, 1984: Lal, 1986a-c). A long-term experiment with maize at Ibadan, Nigeria, showed that where vehicular traffic is not used, satisfactory maize yields are obtainable under continuous maize crops with two crops a year for 17 consecutive years (Lal, 1982; 1984). In another experiment on similar Lixisols, Lal (1986d) reported satisfactory maize yields using no-till with residue mulch for 8 consecutive years. Lower yields were obtained for crops on ploughed or ridged seedbeds (Figure 19). In the same district, equivalent or higher yields of cowpeas and soybeans were obtained with no-till than with a plough-based system (Nangju, 1979) (Table 31). In Ghana, Leyenaar and Hunter (1976) record lower yields for ridged- than for flat-planted maize.

63 Soil compaction limits yields severely under the mechanized no-till systems, especially on soils with predominantly low-activity clays (Lal, 1985a; Kayambo et al., 1986a,b). Drastic declines in yields of maize are observed where the soil is compacted (Table 32). It is important, therefore, that soils are managed to minimize the risks of soil compaction. Soil compaction is less with manual than with mechanized land clearing methods (Lal and Cummings, 1979; Lal, 1981; Hulugalle et al., 1984), and less with manual than with motorized cultivations. Soil compaction can be alleviated by growing deeprooted legumes such as Mucuna (Hulugalle et al., 1986), pigeon pea (Cajanus cajan) (Hulugalle and Lal, 1986), or deep-rooted perennials such as Leucaena leucocephala (Juo and Lal, 1977). It can also be alleviated by mechanical means. The paraplough has proved effective in loosening a compacted Lixisol without soil inversion. This equipment is, however, energy-intensive, is too expensive for small tropical farmers, and requires a powerful tractor. Tropical root crops, such as yams and cassava, are grown in humid regions with a mean annual rainfall of mm. The soils of such areas have a ph of and have low plant-available nutrient reserves. Several experiments have shown that the best way to obtain a high tuber yield is to use mulch rather than tillage (Vine, 1981; Ohiri, 1983; Anazodo and Onwualu, 1984; Anazodo, 1986; Opara-Nadi and Lal, 1987a, b; Hulugalle and Lal, 1986).

64 Germination and emergence of yams are often faster and more complete on mulched than on unmulched soils (Hahn et al., 1979). Mulching increases tuber yield of yams through its favourable effects on soil temperature and moisture. Tables 33, 34 and 35 show the significant effects of mulch on fresh tuber yield. The tuber yield, however, can be markedly limited if the effective rooting depth is less than 50 cm. If the rooting depth is restricted either by a hardpan or by a high water table, tuber yield can be increased by using raised seedbeds. Upland rice has low germination, gives a poor stand and low grain yields when grown under no-till rather than a plough-based system (Ogunremi et al., 1986a). Poor seed-soil contact, damage to young seedlings by birds and nutrient imbalances in the soil are major factors. Transplanting of rice seedlings through the dead mulch of a cover crop, although labour-intensive, may be a satisfactory method of obtaining a good crop stand. In western Cameroon, however, Ambassa-Kiki et al. (1984) reported no differences in rice grain yield among no-till, minimum-till and plough-till methods when rice was grown on Vertisols with a high moisture retention capacity. The mean grain yields and soil bulk densities of the 6-11 cm layer were 6240 kg/ha and 1.46 g/cm 3 for minimum-till, and 5934 kg/ha and 1.56 g/cm 3 for the plough-till, respectively. Hydromorphic soils with low infiltration rates give satisfactory yields of lowland irrigated rice with an unpuddled or no-till system (Rodriguez and Lal, 1985; Lal, 1986d) (Figure 20). For rice grown on a sandy hydromorphic soil in western Nigeria, mean grain yields and soil bulk density of the 6-11 cm layer were 6240 kg/ha and 1.46 g/cm 3 for no-till, 6332 kg/ha and 1.61 g/cm 3 for minimum-till, and 5934 kg/ ha and 1.56 g/cm for the plough-till, respectively. Also, for hydromorphic soils with low infiltration rates, satisfactory yields of lowland irrigated rice have been reported with an unpuddled or no-till system (Rodriguez and Lal, 1985; Lal, 1986d) (Figure 20). No-till farming is also successful in the tropics of Central and South America. In Costa Rica, Shenk and Saunders (1981) observed that no-till and reduced-till systems produced more maize grain than ploughed treatments (Table 36). No-till plots also suffered significantly less damage by insects and pathogens. Despite the high yields, however, systems with large chemical inputs were not the most economical or efficient (Zaffaroni and Locatelli, 1980). Maize, soybeans, beans, wheat and sunflowers (Helianthus annuus) are successfully grown under no-till systems on a wide range of soils in Brazil. Yields obtained with the no-till system were often equivalent to or greater than those using the plough. In Parana, for example, Derpsch et al. (1985; 1986) reported that the mean grain yields (over 6 consecutive years) for wheat were 1597, 1418 and 1342 kg/ha, and yields for soybeans were 2314, 1856 and

65

66 1730 kg/ha for no-till, chisel ploughing and plough-till methods, respectively. Effects of tillage methods on yields, however, depended on the season, location and crop (Table 37). In 1979 wheat yield was more in no-till than in ploughed land for Londrina, the reverse was the case for Rolandia. In 1978, no-till crops outyielded those under plough-till at both sites. Soybean yield was less with plough-till than under a no-till or minimum-till system for both sites. Wheat yields were increased and soil physical properties and nutritional status improved when a no-till system was adopted in conjunction with appropriate cover crops (Kemper and Derpsch, 1981). The use of a no-till system also improved soil temperature, moisture regime and water infiltration (Sidiras et al., 1985a). In Sao Paulo, Brazil, Silveira and Kurachi (1981) observed that mechanical weeding of coffee altered soil structure. In Minas Gerais, Brazil, the grain yield of beans was 959 kg/ha with the no-till system and 712 kg/ha with plough-till. A no-till system proved effective for soil and water conservation and for the yield of pastures and grain crops on Luvisols prone to crusting and accelerated erosion in Northern Territory, Australia. McCown et al. (1980; 1985) and McCown (1984) developed a system producing maize and soybeans in association with pastures using no-till. Cultivation of sugar cane in northern and central Queensland, Australia, is widely practised under the no-till system. As well as producing equivalent yields, the no-till system has proved effective for erosion control on undulating terrain (Bureau of Sugar Expt. Station, 1984; Freebairn et al., 1986). Many crops are being grown experimentally with no-till systems in the humid regions of South and Southeast Asia. Blevins (1984) describes the potential for minimum-till in Bangladesh. Its potential is promising for many regions in India. Thamburaj (1980) reports that cassava tuber yield in Madras was satisfactory when no-till was combined with the use of residue mulch. Wijewardene (1982) reports that a range of upland crops can be grown in Sri Lanka with a no-till system to produce economic returns. Conservation tillage systems have been tried in northern Thailand, where Ratchdawong et al. (1984) report satisfactory crop yield and less runoff and erosion on no-till cropping compared with twice-ploughed plots (Table 38). Higher yields were obtained with mulch than without. Research in Indonesia has also shown the usefulness of no-till farming for erosion control and for producing satisfactory crop yields. Suwardjo et al. (1984) showed that mechanical tillage including ploughing reduces structural stability. They therefore recommended minimum-till with mulching for soils prone to erosion. The data in Table 39 show that satisfactory yields are obtained with no-till systems for maize, groundnuts, mung beans (Phaseolus aureus) and upland rice.

67

68 Lowland rice is a key crop in Southeast Asia. Puddling the soil when it is near saturation point to destroy soil structure and decrease percolation is the conventional method of seedbed preparation. The puddling system has many disadvantages including high energy and labour costs, the time needed in seedbed preparation when the turnaround time is limited for double cropping, and low yields of the following upland crops caused by poor soil structure. Sharma and De Datta (1986) evaluate the use of a no-till system for lowland rice. They report satisfactory yields in some seasons but not in others. Other workers have also reported satisfactory yields in the Philippines (Mabbayad and Buensosa, 1967; De Datta and Karim 1974; Sharma and De Datta, 1986). Advantages and disadvantages of no-till systems in the tropics In suitable circumstances in the tropics, no-till has the following advantages over plough-till systems: erosion control, moisture conservation, savings in labour and fuel, reductions in machinery cost, timeliness of operations, possibility of double cropping and lessened risks of environmental pollution. It can have some disadvantages such as high weed infestation, additional herbicide costs, creating of anaerobic conditions and possible yield reductions. No-till systems are apparently suited to tropical uplands, where the advantages outweigh the disadvantages. Effects on soil temperature At the onset of monsoons, soil temperatures are usually supraoptimal (>30 o C) for at least up to 2 months after sowing (Figure 21). Germination and crop establishment can be adversely affected by high soil temperatures. A notill system with crop residue mulching lowers the maximum soil temperature and improves germination, seedling establishment, crop growth and yields (Lal, 1982). Soil moisture Plough-till plots are generally more drought-prone than those under the no-till system. Occurrence of frequent drought stress is one of the major yield-limiting constraints on upland soils of the tropics. Hard-setting characteristics Some tropical soils with low levels of organic matter, predominantly those with low-activity clays, set hard on drying (Mullins et al., 1987). Such soils are widespread. Hard-setting inhibits seedling establishment and root growth and lowers crop yields. Hard-setting soils have extremely low infiltration rates and the surface soil is often so hard that it is difficult to sow crops without resorting to mechanical loosening, so sometimes no-till systems are not suitable. Compaction Soil compaction is a problem in soils subjected to excessive vehicular traffic. Compactability is related to the nature and amount of clay, quantity and distribution of crop residue mulch and the soil moisture content. Low levels of soil organic matter content, the absence of the effects of freezing and thawing and the predominance of soils with lowactivity clays tend to make tropical soils compactable. Lal (1984) observed that soil compaction can set in 2-3 years

69 after adopting mechanized no-till farming on a tropical Lixisol. The data in Figure 22 show a drastic decline in infiltration rate of a Lixisol used for continuous cropping to maize with motorized farm operations. Accelerated erosion The adverse effects of accelerated soil erosion (Lal, 1987a) are generally severe in the tropics because: The effective rooting depth of most residual soils is commonly limited by root-restrictive layers at shallow depth. This may be due to either adverse physical or nutri-tional properties. A high proportion of plant-avail-able nutrients and soil organic matter reserves are confined to the top few centimetres of the soil. Chemical fertilizers and other modern ameliorative techniques are expensive and not readily available. As a consequence of the above factors, the range of tolerable soil loss for most uplands is 1-2 t/ha/yr (Lal, 1990). Soil fertility As a general rule, the residual upland soils of the tropics are old, highly weathered and often excessively leached. The nutrient reserves are small and concentrated in the top few centimetres of the soil (Obeng, 1978; Sanchez and Buol, 1975; Sanchez and Logan, 1992; Lal, 1987b) and are not present in the right balance. The potential productivity of soils is, therefore, low. RIDGE TILLAGE In arid and semi-arid tropical regions crop yields are limited by the low amounts of plant-available water (Lawes, 1965; Lal, 1979). Most of the rain falls during high-intensity thunderstorms and is concentrated in a short rainy season. Because

70 of low infiltration capacity, a high proportion of rainfall is lost as runoff. To restrict runoff and give time for the rainwater to infiltrate, contour ridges are used. A graded-ridge furrow system, once installed, can last for many years provided that drainage channels are maintained and weeds are adequately controlled (Kampen et al., 1981). In the savannah regions of Nigeria and elsewhere in semi-arid tropical Africa, crops are traditionally grown on ridges, on small hillocks, or on mounds (Kowal and Stockinger, 1973). Ridge cropping is practised under a wide range of conditions. Crops are grown on ridges or hillocks on shallow soils to increase the effective rooting volume. Ridges are formed on poorly drained soils to grow upland crops; on nutrient-deficient soil to heap up the fertile topsoil; in manually-powered farming to save labour by tilling only half of the land; on steep slopes to provide drainage channels up and down the slope for disposal of surplus water; and on sloping lands to conserve water and control erosion. Ridging facilitates easy harvesting of roots and tuberous crops. Ridge-tillage has evolved as an integral part of subsistence farming throughout the tropics and is well adapted to small-scale, low-input, subsistence farming. It often breaks down on large, commercial farms. Ridges erode on structurally unstable soils, mechanical weed control is difficult and crops suffer from high soil temperatures and low soil moisture, because the ridges tend to dry rapidly and crops often lodge. If rainfall exceeds the capacity of the ridge-furrow system to check surplus water, gully erosion is often severe on sloping land. On steep land, ridges are sometimes made up and down the slope, to facilitate surface drainage and minimize the risks of landslides and mass movement. Ridges made up and down the slope may, however, cause more runoff and erosion than sowing without ridges (Kowal, 1970a, b). Table 40 shows the highest runoff and erosion occurs where ridges are made up and down the slope. Under these conditions even tie-ridges and broadbeds were less effective than planting on the flat. Experiments conducted at IITA, Ibadan, also show maximum runoff and erosion from plots with ridges up and down the slope (Table 41). On undulating slopes, however, contour ridges increase the time for water to infiltrate into the soil. The surface retention capacity is greatly increased, and both the amount and the rate of runoff are decreased by ridges constructed on the contour. Ridge-furrow systems are usually stable for gentle slopes up to 7 percent and for soils with a relatively stable structure. Severe erosion can be caused by the failure of ridge furrow systems, installed on steep slopes and on soils of unstable structure.

71

72 Tied-ridges Tie-ridging and ridging with cross ties in the furrows is an improvement over the traditional ridge-furrow system. Tieridging is similar to basin-listing as described in the semi-arid regions of southeastern USA by Unger (1984). The tie-ridge system was initially developed in semi-arid Tanzania, where it proved to be an effective water-conserving technique. Prentice (1946) compared yields of cotton, sorghum and maize on tied-ridges and flat seedbeds from 1939 to With the exception of cotton in 1942, yields were more from tie-ridges than from flat seedbeds. Yield benefits in the tie-ridge system were greatest in years of partial or modest drought (Table 42). Crop failure due to drought stress and soil degradation led to the abandonment of the Groundnut Scheme at Kongwa, in semi-arid Tanzania (Wood, 1950). Pereira et al. (1958) observed that neither broadbeds nor tied-ridges provided satisfactory erosion control, although tied-ridges prevented gully erosion. Macartney et al. (1971) at Kongwa reported no beneficial effect of tie-ridged system on maize grain yield. In Malawi there was no increase in cotton yield when planted on tied ridges (Mitchell, 1987). On a similar soil in India, Ali and Prasad (1974) report no effect of ridging on moisture conservation or on grain yield of pearl millet. This apparent contradiction regarding the beneficial effects of tie-ridge systems can be resolved in terms of soil properties. It seems to be effective in water conservation on some soils but not others. Dagg and Macartney (1968) (Table 43) show that tie-ridged land produced significantly more maize grain yield only on a red soil (Lixisol) but not on a Vertisol or an Andosol. The principal benefits of tied-ridges on soils with positive response were attributed to moisture conservation (Table 44). In Zambia, Honisch (1974) measured yields of maize, sorghum, and millet in a tied-ridge system. On a structurally unstable soil with no water-stable aggregates, > 0.5 mm tied-ridges produced the highest grain yield of maize and sorghum. On soils in the Sahel region of Burkina Faso, the tied-ridge system also conserved moisture (Table 45). In this study, the beneficial effects of the tied-ridge system were pronounced in seasons of frequent droughts during the growth cycle and on shallow soils with low infiltration capacity. Lawes (1966) observed that in northern Nigeria the highest yield of sorghum was obtained when alternate furrows were left open, and the lowest yields where all furrows were left open. The increase in grain yield ranged from 0.29 to 0.44 t/ha. Yield increases were also obtained in groundnuts by tying alternate furrows. The practice of tied ridging has also been found beneficial in semi-arid and arid regions of Burkina Faso. Rodriguez (1987) got

73

74 significant increases in yields of maize planted on tied ridges in comparison with that planted on the flat or on simple ridges. Hulugalle (1986a) observed that the grain yields of cowpea were 1529 and 2315 kg/ha in 1985, and 1981 and 2373 kg/ha in 1986 on simple and tied ridges, respectively, for similar soils and environments. Hulugalle (1986a, 1987b) observed that tied ridges increased profile water by an average of 30.5 mm and 24.6 mm per week. Consequently, the grain yield of cowpea was increased by kg/ha. The increase in profile water content is apparently due to an increase in infiltration capacity. In another study in Burkina Faso, Hulugalle et al. (1987a) observed that tied ridges in combination with rockbunds produced the maximum grain yield of sorghum. With the recommended rate of fertilizers, the sorghum grain yields for planting on tied ridges and on the flat were 620 and 260 kg/ ha with rockbunds and 490 and 170 kg/ha without rockbunds, respectively. Hulugalle (1990) demonstrated in semiarid regions of Burkina Faso that the tied ridge system was extremely useful for drought-prone shallow soils of the West Africa Sahel (Table 46). The effectiveness of ridge planting depends on duration, intensity and time of occurrence of drought stress, as well as soil properties. In the Eastern Province of Uganda, Walton (1962) observed that ridge systems had no beneficial effects on yield of an early planting on high-yielding productive soils where severe drought stress did not occur during the critical flowering stage. Ridge planting was extremely beneficial, however, on marginal lands of low-yielding potential where the frequency of drought during critical stages was high. It is apparent that ridge planting is a risk-avoidance system for small subsistence agriculture. It is effective on marginal, low-yielding soils, on structurally inert soils with low infiltration rates, and in regions with frequent drought stress during critical stages of growth. It is not an effective system for commercial agriculture or for high-yielding productive soils of good infiltration rate and favourable soil structure. Ridge planting or mechanical tillage provides a transient beneficial effect by improving infiltration. On some soils therefore flat cultivation methods with mulch in the inter-row zone and zonal tillage in the seed zone to alleviate soil compaction are preferable. The beneficial effects of tied ridges have also been demonstrated in Botswana (Min. of Agric: Dryland Farming Research Scheme, 1984). Grain Yields of sorghum and cowpeas were markedly increased when every furrow was tied (Table 47). The yield increases were apparently due to greater availability of water in the root zone under the tied ridges. The depth of penetration of wetting front after a 34-mm rainfall was 11.1, 17.3 and 24.0 cm for untilled soil, untied ridges and tied ridges, respectively. Similar results are reported from the Shair Haneger region of Israel. Morin et al. (1984) show a 44% increase in the yield of wheat sown on 1.6 m-wide beds of tied ridges (Table 48). In arid parts of Israel, Rawitz et al. (1983) observed that erosion loss from disced and ploughed plots was approximately 10 times greater than from plots managed with basin tillage or the tied ridge system. The ridge-furrow system also facilitates irrigation. The bed-furrow system is also a useful technique for improving irrigation efficiency. This technique saves water and increases irrigation efficiency. The semi-permanent bed-furrow system combined with no-till or minimum-till is an appropriate conservation tillage measure for improving water storage in the root zone. The system of irrigation using tied ridges has been successfully used in semi-arid Texas (Musick et al., 1977; Stewart et al., 1981; Stewart and Musick, 1982).

75 Irrigation often alters soil structure, probably because of salts in the water (Moreno et al., 1986). Use of mulching and no-till systems can restore soil structure and improve crop yields (Adem et al., 1984; Tisdall and Adem, 1986). Choudhary et al. (1985a) also reported an increase in maize grain yield on coarse-textured sandy soils at Ludhiana, Punjab, India, by irrigation. The maximum yield was, however, obtained by a combination of subsoiling and irrigation.

76 Chapter 5 Improving rooting depth Tillage is commonly used to improve the rooting depth of soils. Many soils in the tropics have root-restricting horizons. Root impedance may be for chemical or physical reasons. The subsoil may, for example, have excessive concentrations of Al and Mn or deficiency of P. Physically, the subsoil may have high density, high strength and a small proportion of macropores. Some soils have massive structure. Such soils are especially common in arid and semi-arid regions. Rooting depths can be improved by the techniques described below. AMELIORATION IN SOIL STRUCTURE Improvement in the soil organic matter content usually results in structural amelioration. In the tropics it is easier said than done. Fallowing, growing suitable legumes and grasses for a sufficient period, is effective in increasing soil organic matter content, improving soil structure and alleviating soil compaction. Fallowing by regeneration of natural vegetation is a slow and inefficient system and generally requires many years for complete soil recovery. Fallowing with specially planted grass or legume species or woody perennials is an efficient method. Addition of manures and organic matter may be required to establish good grass or legume cover (Pereira and Beckley, 1952). The beneficial effects of grass and leguminous crops are well recognized in the semi-arid regions of Kenya and elsewhere in East Africa. Pereira et al. (1954) and Pereira (1956) investigated various lengths of fallowing with grass and legume mixtures on the formation of water-stable aggregates, total porosity, infiltration and percolation rates, and the yield of arable crops. They show that improvements are achieved during a fallow period under grass, and that structural improvements are rapidly lost during the arable phase. Pereira and associates observed that soil improvements were more rapid with grass than legumes. From their experiments on the abandoned land of the Groundnut Scheme, Pereira et al. (1958) confirmed that fallowing with teff grass (Eragrostis abyssinica) improved rainfall infiltration and decreased runoff and erosion. However, establishment of grass can be difficult on ploughed bare soil, and is generally easier with no-till or by undersowing standing crops than by seeding after mechanical tillage (Peers, 1962). Improvements in soil structure and fertility by fallowing with grass and legume covers are also demonstrated in Uganda (Jameson and Kerkham, 1960), Nigeria (Dennison, 1959; Vine, 1953), Central Africa (Jurion and Henry, 1969) and Malaysia (Weng et al., 1979). In northern Nigeria, Wilkinson (1975) observed that infiltration rates increased during the fallow period. Wilkinson confirmed that the equilibrium infiltration rate increased with duration of the fallow period.

77 However, most of the increase in infiltration rate was lost by the end of the first cropping season. For example, an infiltration rate of 58 mm/hr at the end of the first cropping season declined to 23 mm/hr at peak rainy season and to 10 mm/hr at the end of the cropping cycle. Similar data were obtained in western Nigeria by Wilkinson and Aina (1976), where the high infiltration rate under prolonged bush fallow declined rapidly with arable cropping. Lal et al. (1979) reported significant improvement in infiltration rate by growing cover crops on an eroded and compacted Lixisol in western Nigeria. There were also differences in infiltration rate under various grass species. These improvements in water infiltration were attributed to earthworms and other faunal activity (Table 49). Juo and Lal (1977) demonstrated that deep-rooted woody perennials, Cajanus cajan and Leucaena leucocephala, were more effective in improving infiltration and macrochannels than shallow-rooted annuals. Similar positive effects of L. glauca were observed in East African soils by Pereira et al. (1954). Mehanni (1974) reports that fallowing with

78 deep-rooted plants in Australia had the same effect on structure as deep ripping. However, the choice of an appropriate fallow crop depends on many factors, including soil and climate. Economic considerations are also important, because farmers do not want to spend time growing a cover crop if circumstances permit the growth of a food crop (Vine, 1953). SOIL RESTORATIVE CROPPING SYSTEMS No-till and conservation tillage systems do not function effectively without adequate mulch. Residues from the previous crop may be insufficient especially in the arid and semi-arid tropics. Further supplies of mulch can be obtained by the methods described in the paragraphs below. Management of planted fallows Pereira et al. (1954) and Wilkinson (1975) show that the benefits of fallowing are lost by the end of first cropping when primary and secondary tillage operations are used. Several studies have shown soil structural properties are optimum when the soil is protected by vegetation and that any disturbance alters them adversely. It is, therefore, important that improvements brought about by fallowing be maintained by the tillage system which follows. Many experiments at the International Institute of Tropical Agriculture (IITA) (Lal, 1983; Wijewardene, 1982) and elsewhere in the tropics (Ogborn, 1982) demonstrate that deterioration of soil structure and decrease in infiltration capacity during cropping can be drastically reduced by managing the fallow cover using a no-till system. Cover crops and pastures are a source of producing in situ mulch for no-till farming (McCowen et al., 1980; Wilson and Akapa, 1983). Killed sod as residue mulch The destruction of pastures by herbicides or mechanical means is a good substitute for ploughing (Hood et al., 1963). In Ghana, Kannegieter (1967; 1969) successfully demonstrates the use of herbicide for growing maize through killed Pueraria sod. The yield of maize grown through an undisturbed Pueraria mulch was significantly greater than when ploughed under or used as mulch after ploughing. Wilson (1979) discussed the use of legume cover crops for no-till production of food grains and recommended the inclusion of cover crops in the rotation every 3 or 4 years. As well as soil physical properties, cover crops suppress weed growth and control nematodes. Lal et al. (1979) and Wilson et al. (1982) compare the grain yield of maize seeded through a range of suppressed grass and legume covers (Table 50). In 1976 with chemical spray, high grain yields were obtained following Psophocarpus, Stylosanthes, and Pueraria. Among the mowed treatments, the maximum yield was obtained with Stylosanthes. In 1977, the highest maize grain yield was obtained in Stylosanthes irrespective of the method of suppression. Although improvements in soil structure and infiltration rate were greater in grass than legumes, low maize grain yield with grass cover crops is attributed to difficulties in effective suppression. The beneficial effects of grass and legume mulch on crop yields are demonstrated for an Acrisol in Peru by Wade and Sanchez (1983). Their data show that incorporation of grass and legume mulch is, for some crops, more

79 beneficial than leaving it on the surface. McCowen et al. (1985) found that maize and sorghum yields on chemically suppressed legume pastures without tillage on Luvisols in Australia were superior to those on ploughed soil. Live mulch Live mulch can be used in a system of mixed cropping (Lal, 1975). A low-growing crop, preferably a legume, is grown specifically as a cover crop. The objective is to establish a fast-growing, easily controlled crop that can readily smother weeds while aggressively covering the entire ground. Legumes are used as a perennial cover throughout the cropping phase so that soil structure, infiltration rate and chemical fertility are preserved. A small strip is opened, with or without using herbicides, so a seasonal food crop can be sown. The system works if the fallow crop does not compete for light, moisture and nutrients. Wilson et al. (1982) show that mechanically mowed Stylosanthes can be effectively controlled, but unsuppressed growth of other cover crops drastically suppressed maize grain yield. The climbing legumes, Psophocarpus and Centrosema, suffocated young maize. Akobundu (1980; 1982) compared maize grain yield using three legume cover crops as live mulch with that obtained with no-till and ploughing. With effective weed control and proper suppression of live mulch, grain yields were equivalent to those for the no-till and ploughed systems. Lal et al. (1978) and Akobundu (1980) report significantly more earthworm activity under a live mulch than under a no-till system using the previous crop residue as a mulch. Although the live mulch system of planted fallow crops depends on the availability of appropriate herbicides, it has potential in the tropics (Miller and Bell, 1982). Its

80 success depends to a great extent on the choice of legume covers that can be used as live mulch with a minimal dependence on herbicides. Agroforestry and alley cropping Agroforestry is the practice of growing deep-rooted perennial woody shrubs in association with seasonal food crops. This system of mixed cropping with selected perennial shrubs and leguminous trees attempts to minimize the degradation of intensive arable use and to preserve and enhance soil structure and chemical fertility (Kang et al., 1981; Juo and Lal, 1977; Mongi and Huxley, 1979; Pereira et al., 1954; Vergara, 1982). Agroforestry can be combined with a no-till system to attain the maximum environmental benefit. In alley cropping, arable crops are grown between rows of specially planted woody shrubs or trees, which are regularly pruned during the cropping season to prevent shading, provide mulch and reduce water use. Food crops interplanted with nitrogen-fixing legume trees are generally more productive than monocultures. Either an alternate row or alternate strip system of tree arrangements can be used, and the trees are pruned at an appropriate height to promote rapid regeneration. Commonly used legume species grown in association with food crops include Sesbania, Gliricidia, Leucaena, Parkia, Ramon, Calliandra, Albizia, Acacia, Cassia, Pithecellobium, Mimosa, Prosopis and Samanea. The choice of an appropriate species depends on soil, climate, and main crops to be grown. They should have rapid growth, nitrogen-fixing capability, and a multipurpose nature (Vergara, 1982). ALLEVIATING SOIL COMPACTION BY DEEP PLOUGHING The objectives of seedbed preparation on structurally-inert coarse-textured soils containing predominantly lowactivity clays are to improve (i) total porosity and relative proportion of macropores, (ii) root growth in subsoil, (iii) water infiltration capacity, and (iv) water storage within the soil profile. The methods of seedbed preparation vary with the circumstances. Semi-arid West Africa Intensive soil management studies were conducted by IRAT in West Africa (Charreau, 1972; Charreau and Nicou, 1971; Chopart and Nicou, 1976; Nicou, 1974a, 1974b; Nicou and Chopart, 1979; Poulain and Tourte, 1970; Chopart, 1981). These studies indicate the benefits of deep tillage and soil inversion by decreasing runoff losses and conserving more water for plant growth. Nicou (1977) records significant increases in available water in the soil profile and of yield of millet and groundnuts from ploughing compared with the no-till system. In these experiments, however, the crop residue in the no-till treatment was burned. Charreau (1977) summarizes 17 years data and shows the beneficial effects of ploughing on grain yield of pearl millet, sorghum, maize, lowland rice, cotton, and groundnut (Table 51). Nicou (1979) shows from his studies at Bambey that the root weight of groundnut increased linearly with total porosity and that the pod yield increased with total root weight as shown by the following empirical equations:

81 Root weight (mg) = 148 (total porosity, %) , r = 0.91 Pod yield (kg/ha) = 1.37 (Root weight, mg) + 233, r = 0.82 Ploughing increased total porosity and the groundnut yield. Nicou (1974a), however, shows that ploughing brought about only transient improvements and that soil structure is easily degraded during the cropping phase. Nicou (1979), Chopart (1981) and Chopart et al. (1981) support the conclusion of Charreau (1977) that the structurally inert soils of the Sahel region benefit markedly from mechanical tillage. Nur and Gasim (1974) also support Charreau s conclusion that intensive tillage increases groundnut pod yield. Minimum or reduced-tillage, however, was found to be feasible for class 1 and class 2 soils in the sub-humid regions of the Côte d Ivoire (Table 52). Soils in the Boake region of Côte d Ivoire are structurally-active and respond favourably to a no-till system with crop residue mulch. Southern Africa Soils in semi-arid Botswana are mostly sandy (Arenosols). Similar to soils of the West African Sahel, they are also structurally inert, and moisture is the most important factor limiting yield. Willcocks (1979a, 1979b, 1981) shows the importance of reduction in soil bulk density by mechanical tillage and periodic subsoiling. Whiteman (1975) shows that moisture conservation by fallowing in alternate years increases crop yield. He demonstrates that chemical weed control during fallowing is more effective in soil-moisture conservation than ploughing and mechanical weed control measures. Sorghum grain yields are significantly higher after bare fallow than after cover crop, weed, fallow or maize.

82 Tropical India Studies in India support the conclusion that structurally inert soils require mechanical ameliorative measures to improve soil and water conservation and crop yield (Charreau, 1977). In Gujrat, Borole et al. (1972) report yield data from a 23-year tillage study showing that deep ploughing consistently produced higher yields of pearl millet than shallow ploughing or no ploughing. Gidnavar et al. (1972) report a low percentage of water-stable aggregates in soils receiving minimum-till treatment. Subramanian et al. (1975) report that for a red sandy loam soil near Madras deep tillage (20-45 cm) improved soil porosity, saturated hydraulic conductivity and structural stability. In the subtropical regions of northern India, ploughing also improved the infiltration in an Inceptisol and increased grain yield of pearl millet by 29% over the no-till treatment. Chiselling in the row zone was as effective as ploughing (Malik et al., 1973). Summary In most studies reported above, the beneficial effects of ploughing over untilled soil measured in terms of grain yields are impressive. The results, however, should be viewed in the light of the following: Crop residue mulch was not used in most of these studies. Wherever residue mulch was used, its advantages in soil and water conservation and improving yields were substantial in comparison with the unmulched no-till technique commonly used as a standard control. Soil physical conditions at sowing time were generally unfavourable. The results would have been different if steps had been taken to improve soil physical conditions by fallowing or by using deep-rooted cover crops prior to implementing the no-till system. Amelioration in soil structure caused by mechanical tillage was generally temporary, and the improved structure was easily degraded (Nicou, 1974a). Tillage and fertilization, although temporarily increasing crop yields, did not completely restore degraded lands unless organic matter levels were improved (Diatta, 1974). Chemical weed control during fallowing in Botswana was as effective in moisture conservation in Botswana as mechanical tillage and ridging (Whiteman, 1975). The benefits of ploughing were limited to its role as a weed control measure. The effects of ploughing entire fields were similar to those of chiselling in the row zone alone. Permanent improvements in soil structure and water transmission cannot be achieved by mechanical tillage alone. Steps must be taken to improve soil organic matter content by efficient fallowing, controlled grazing and prohibiting the use of fire in land preparation. Once soil structure has been improved, ploughing and subsoiling may not be as necessary. ALLEVIATING SOIL COMPACTION BY PARAPLOUGH AND SUBSOILING Some tropical soils have poor subsoil characteristics, and root growth into the lower horizons is checked by high soil bulk density. Such soils may also have low infiltration rates caused by surface crust. In Zambia, for example, Lenvain

83 and Panwelyn (1988) observed high soil bulk density of g/cm 3 at cm depths. Shallow, eroded and compacted soils occur extensively in West Africa (Charreau, 1970) and in the semi-arid regions of India (Vittal et al., 1983). Mechanical loosening improves crop growth on such soils. Surface soil compaction is also caused by wheeled traffic and plough-pans can form in subsurface horizons (Lal, 1985b). Such compaction decreases root growth, crop water uptake, and crop yield. In Nigeria, Kayambo and Lal (1986) and Oni and Adeoti (1986) observe that increasing tractor passes prior to sowing decrease germination and yield of maize, cowpeas, soybeans and cassava. In Southern Brazil Da Silva et al. (1981) and Klamt et al. (1986) record a decrease in the proportion of water-stable macro-aggregates and a reduction in the equilibrium infiltration rate due to repeated passage of motorized tillage equipment. Under no-till or reduced-till methods the number of passes during seedbed pre-paration is reduced. Guided trafficing is helpful tractor wheels always follow the same tracks. The use of mechanical loosening devices, such as the paraplough, is beneficial. The paraplough is a slant-legged machine that loosens the subsoil without inverting it. This can temporarily increase yield, as has been observed on compacted Lixisols in western Nigeria (Table 53). It is noted, however, that moderate compaction in the seed zone by wheel tracks may improve seed germination and seeding establishment (Sidiras and Vieira, 1984). In sub-humid and humid regions, seedling establishment on crusted soils can be improved by using strip tillage, no-till systems, or a ridge tillage system. Nangju et al. (1975) note that seedling emergence of soybeans was 1, 33, 51, and 54% with ridges, ploughed flat, strip tillage, and no-till methods of seedbed preparation (Table 54). For soils in arid and semi-arid regions, however, it is necessary to break the crust mechanically. In the West African Sahel Hoogmoed and Stroosnijder (1984) observed that beneficial effects of breaking the crust are short-lived and new crusts are quickly formed. Frequent cultivation may be needed. Attempts have been made to develop special equipment to break the crust without damaging seedlings. IMPROVING ROOTING DEPTH BY DRAINAGE Yields are often depressed on soils with slow or poor internal drainage. In Trinidad, Lindsay et al., (1983) observed that on a tropical Inceptisol with impeded drainage, maize grain yields were 5.6, 1.9 and 1.3 t/ha for plough-till, notill and minimum-till, respectively. Vertisols (poorly drained, fine-textured soils) are widespread in the tropics. Their infiltration rate is often as low as 0.2 cm/hr (ICAR, 1984). Formed predominantly from swelling clay minerals, they develop wide, deep cracks during the dry season. When dry, they have a hard consistency and are plastic and sticky when wet. The optimum soil moisture range for tillage is narrow.

84 Graded and open ridge-furrow systems help to remove surplus water from such soils. For example, cotton grown in Tanzania on a system of raised and cambered beds, 7-8 m apart, produced good yields over a number of years (Spence and Smithson, 1966). Some examples of such beds are schematically drawn in Figure 23. At ICRISAT, Hyderabad, India, a broadbed-and-furrow (BBF) system was developed to reduce runoff and erosion and to permit cropping during the monsoon season (Figure 24) (Kampen, 1982). The cultural practices for the BBF system are as follows: The BBF system is semi-permanent, and maintenance is done during the dry season. If tillage is necessary, it is done on already existing broadbeds. Traffic on wet soil is avoided, dry seeding is recommended just before the first rains. The seed is sown deep enough to avoid germination following a small amount of rainfall. The optimum sowing depth is 5-7 cm for sorghum, pigeon pea, maize and millet. Small seeds, such as Setaria, are sown at shallower depth. The rainy-season crop is harvested as soon as possible after maturity, so that the post-rainy-season crop can be established while the soil is still moist. The post-rainy-season crop is intercropped or relay cropped through the standing previous-season crop using a no-till system. Klaij (1983) has developed a tillage concept that involves the BBF and a ridge-furrow system. Klaij sees the seedbed as comprising three separate zones: a traffic zone, water management zone, and the seedling environment zone (Figure 25). He observes that the grain yield of pearl millet decreased with increase in depth or cross-sectional area of the soil disturbed. The runoff and erosion from shallow-tilled soil were less than those on deep-tilled soil.

85

86

87 In Central India, yields of black gram (Cicer arietinum) were more than doubled by the raised-bed technique (ICAR, 1984). The mean grain yields for control plots, for raised beds made along the slope, are given in Table 55. There was no difference in yield between the three spacings used on the raised beds, implying adequate surface drainage. Raised beds along the slope, however, increase the risk of soil erosion, unless the land has already been graded. A satisfactory level of soil aeration was not achieved by constructing drainage ditches in the control treatment. Poor soil drainage is common on lowlying and flat ground, and farmers overcome this by constructing mounds and ridges (Kowal and Stockinger, 1973). In northern Nigeria, Tarawali and Mohamed- Saleem (1987) observe marked increases in grain and stover yields of sorghum by increasing the ridge height to 45 cm (Table 56). Tied ridges are less useful than open ridges. Soil inversion and ploughing improve crop yields on slowly permeable soils (Lindsay et al., 1983) Flat seedbed and open ridge-furrow systems are, however, prone to accelerated soil erosion, especially at the onset of rains. The use of ridges in combination with crop residue mulch can decrease runoff losses and reduce sediment concentration in the overland flow (Loch et al., 1987). MANAGING LEACHING LOSSES Adequate leaching is necessary to avoid harmful salt accumulation in the root zone. Appropriate conservation tillage for salt-affected soils aims to improve drainage, increase leaching and improve soil porosity. Soils with a high proportion of Na + in the exchange complex commonly have a massive structure. Deep ploughing to 80 cm depths and extensive soil loosening are often necessary to reclaim sodic soils (Cockroft and Tisdall, 1978). Deep tillage and use of crop residue mulches can also help to redistribute salt within the soil profile.

88 Some acid soils require lime applications for good crop growth. Lime acts as a nutrient and neutralizes acidity. It must be incorporated into the surface horizon to neutralize acidity effectively. No-till systems preclude the mixing of lime into the soil. Repeated surface applications of lime in no-till systems may be gradually translocated to the subsoil. This means that crop growth can be adversely affected until the lime is translocated into the root zone. In coarse-textured permeable soils, surface-applied lime is more readily translocated into the root zone (Rodriguez and Lal, 1985; Maurya and Lal, 1979a,b). In the humid tropics of southeast Nigeria, Friesen et al. (1982) note rapid movement of surface-applied lime and record large leaching losses. Some soils, especially upland soils in the humid and sub-humid tropics, are likely to suffer considerable leaching of essential plant nutrients. In southwest Nigeria, Ghuman and Lal (1984) observed that leaching losses of chlorides and nitrates were large in no-till plots, because worm channels serve as preferential pathways for translocation of these chemicals in saturated or nearly-saturated flow. On a kaolinitic Acrisol in southeastern Nigeria, Arora and Juo (1982) observe that nitrate leaches readily in both the fallow and cropped plots. The recovery of applied nitrate ranged from 22 to 60% depending on the number of split applications made. Though preferential movement of water in biochannels can cause higher leaching losses in no-till systems than in ploughed soil, higher concentrations of organic matter in the no-till seedbeds retain nutrients in the surface horizon, so it is difficult to draw generalized conclusions.

89 Chapter 6 Crop response to tillage on various soils LIXISOLS AND TROPICAL LUVISOLS Lixisols are widespread in sub-humid and semi-arid regions. Those in sub-humid regions are structurally-active with more favourable physical properties than those in the semi-arid regions. Lixisols are less weathered and have better soil fertility than Acrisols and Ferralsols. Most have weak soil structure, so they are highly prone to crusting, compaction, accelerated soil erosion and other degradative processes. The tendency to soil degradation is generally greatest under intensive cropping of soil-depleting crops such as maize and rice. To help discussion Lixisols are divided into those with favourable properties and those without. The former are structurally-active, have a relatively high organic matter content and an active soil fauna. The latter are highly degraded. Crop responses on Lixisols and tropical Luvisols with favourable properties Maize, sorghum and other coarse-grain cereals Production of cereal crops is better suited to sub-humid and semi-arid than humid regions. The near neutral soil reaction of Alfisols is also favourable for cultivation of these crops. In Southwest Nigeria, Aina (1982) obtained similar or greater maize grain yields from mulch/no-till than from mulch/ridges or mounded seedbeds. In a long-term study at IITA, Lal (1982) compared maize grain yields on no-till and plough-till plots for 24 consecutive seasons (two seasons per year). Though seasonal differences due to variations in rainfall were marked, maize grain yield from no-till plots was generally greater than from ploughed seedbeds (Figure 26). Tillage-induced differences in crop growth and yield were particularly striking in seasons with a short duration drought. The data show that sustained and economic maize grain yields can be obtained with a no-till system. Another long-term experiment at IITA on sandy soils shows that comparable grain yields to those under conventional systems can be obtained without the elaborate and capital-intensive techniques of chiselling, primary and secondary tillage, ploughing at the end of the previous rainy season, or contour ridges (Table 57). The lowest yield was obtained from the most elaborate ridged treatment. The removal of crop residue mulch from no-till plots also reduced yield, while mechanized traffic caused soil compaction. Also at IITA, Couper et al. (1979) observed that on soils with severe compaction caused by mechanized harvesting of the first season crop, grain yields under a no-till system were greater than on a ploughed seedbed.

90 If initial soil conditions are not unfavourable, however, no-till systems can produce low yields. On a Lixisol near Kumasi, Ghana, for example, Ofori and Nandy (1969) observed significantly less maize yield on unploughed than ploughed seedbed. Drought Stress The high maize grain yields obtained with no-till systems on soils of low water-holding capacity are partly due to more favourable soil moisture and temperature regimes in mulched/untilled soils. Lal et al. (1978) monitored diurnal fluctuations in leaf water potential of maize grown with no-till and plough-till systems for various

91 irrigation cycles (Figure 27). The rate of decrease of leaf water potential in the days after an irrigation event was slower with no-till than with plough-till, implying that the degree and duration of soil moisture stress were both affected by tillage method. The no-till maize developed a better root system in undisturbed soil (Figure 28). Channels created by earthworms and decomposing roots provide continuous and preferential pathways for root development and water transmission (Ghuman and Lal, 1984). Opara-Nadi and Lal (1983) established an empirical relationship between the maize grain yield and available water content for the two tillage systems. Their data showed that the maize grain yields increased at the rate of 0.45 mg/ha for each 1% increase in the available water capacity in the range of 20-39%. Lal et al. (1978) also computed the water use efficiency of maize and found it to be more for notill than plough-till systems. Fertilizer Requirements The response of maize to fertilizer in a no-till system can differ from that in a plough-till system. Losses of nitrogen by ammonia volatilization from surface-applied fertilizer in a no-till system can be greater than from a plough-till seedbed where the fertilizer is incorporated (Acquaye and Cunningham, 1965). Kang et al. (1980) observe that maize grain yield from no-till unfertilized plots was significantly lower than from plough-till treatments. Differences in maize grain yield between tillage treatments are less when fertilizer is applied (Kang and Messan 1983). The higher response to fertilizer in no-till than in plough-till plots is partly due to the immobilization of available N by micro-organisms in plots that received mulch material with a wide C:N ratio. This response is less and even negative when the tillage systems have been underway for a sufficiently long period to allow the available soil N to attain a steady value.

92 For Lixisols and those less able to fix P, P placement in the soil has no advantage over broadcast application. Kang and Yunusa (1977) showed that although P movement was slow with no-till methods, P was equally effective in plough-till and no-till systems. Juo and Lal (1979) and Armon et al. (1981) also observed significant P movement up to 30 cm depth in no-till plots within 6 years. Cowpea Repeated cropping of cowpea in a no-till system can cause soil conditions to deteriorate because of inadequate mulch. Cowpea grown in rotation with a cereal, however, can be easily adapted to the system. Cowpea responds favourably to no-till provided an adequate quantity of residue from a previous crop is available. In Nigeria, it has been observed no significant differences in cowpea grain yield when planted on a freshly ploughed ridge/furrow system or on ridges formed the year before. Lal (1976b) reported cowpea yields of kg/ha with no-till compared with kg/ha on plough-till seedbeds. Nangju (1979) observed that in two of three years, cowpea grown under a no-till system outyielded that on plough-till and ridge-till treatments. He observed that percentage emergence was significantly lower on ridges (83%) and plough-till seedbeds (89%) than with strip tillage (97%) and no-till (98%) (Nangju et al., 1975). Furthermore, it took 12 days for the first emergence down on ridges, compared with 3 days in strip or no-till seedbeds. Cowpea on ridges and plough-till seedbeds suffers more from drought stress and high soil temperature than on no-till plots. Lal et al. (1978) measured the grain yield and water use efficiency of cowpea on no-till and ploughtill plots under different irrigation regimes. For irrigation every 4, 8 and 12 days, water use efficiency was significantly greater with no-till than on plough-till seedbeds. Similarly, Aina (1979b, 1982) observed less drought stress and

93 greater grain yields with a no-till mulch than on ridge or mound seedbeds. While evaluating the effects of tillage and mulching treatments on a Lixisol, Lal (1979) observed that cowpea growth on ridges was suppressed by very high soil temperature and that there was a linear decrease in the leaf area index with an increase in the maximum soil temperature. In the first growing season from April to July, the maximum temperature at 1500 hours was higher on ridges than on with flat-sown, with residue mulch, seedbeds. Lowest yield was obtained on the ridges (Table 58). In the second growing season from August to November, with frequent rains, and a mulch, yields were adversely affected more by fungal diseases than under the other treatments. Soybean Soybean is more sensitive to fluctuations in soil temperature and moisture than maize or cowpea. Soybean germination and seedling establishment are adversely affected by the high temperatures and drought stress commonly experienced during April-May in West Africa. Of all the grain crops, soybean responds most favourably to a no-till with mulch seedbed. Nangju et al. (1975) records that the emergence of soybean was respectively 0.9%, 33.4%, 50.7% and 53.9% for ridges, plough-till, strip till and no-till treatments. The days to first emergence in the same order of treatments were 12, 6, 5 and 5, respectively. Nangju (1979) obtained satisfactory soybean yields on no-till and strip tillage seedbeds, but the 1975 soybean crop sown on ridges failed completely. Table 58 shows that soybean yields were least on ridge-till treatment, due to high soil temperature and frequent drought stress. Cassava and other root crops Most tropical root crops are sensitive to soil depth and need adequate drainage. The response of root crops to tillage systems also depends on their physiology. Tuber shape and rate of development greatly influence yield response to tillage methods. The method of propagation and establishment is relevant. Most tuberous crops are vegetatively propagated. Seed-sets and root cuttings require favourable soil moisture and temperature regimes for good establishment. In general, root and tuber crops do not produce satisfactory yields on compacted or shallow soils.

94 In northern India, Verma (1973) records an increase in size and yield of cassava tubers from earthing up twice during the initial stages of development. In the humid tropics of southern India, however, Thamburaj et al. (1980) report that no-till and mulching of cassava outyielded plough-till followed by harrowing of the seedbed. In Ghana, Ofori (1973) reports significant yield increases in cassava tubers when a compacted Lixisol was ploughed compared with the traditional method. Because cassava tubers develop horizontally, Ofori observed no improvement in tuber yield by ploughing to 30 rather than 22 cm. In western Nigeria, Aina (1982) also reports lower cassava tuber yields on no-till than on plough-till and ridge-till soil. Lal (1980) obtained similar tuber yields from large-scale no-till and plough-till watersheds of 4-5 ha. Akobundu (1983) concludes that low cassava yield on no-till land was affected by severe weed competition as well as compaction. Uncontrolled weed growth caused a yield reduction of 71% in notill and 54% in plough-till treatments. Lal (1975) describes the effects of various methods of seedbed preparation and mulching on tuber yield of yam. On a shallow Lixisol, mulching of ridges and mounds significantly improved plant establishment. For early planted yams, there was an eight-fold increase in yield by mulching a flat seedbed. If adequately mulched, ridges and mounds are effective for the production of tropical root crops, particularly on shallow soils underlain with gravelly horizons. Lal (1980) also reported a satisfactory yield of sweet potato on a no-till mulched Alfisol. Crop responses on Lixisols with unfavourable properties Lixisols with unfavourable properties are widespread in the tropics, in West Africa, East Africa, Australia and India. West Africa The Nigerian savannah is dominated by ferruginous soils. Soils of the northern savanna have additions of aeolian materials. In the southern Nigeria-Benue depression the parent material is predominantly sandstone (Jones and Wild, 1975). The loess soils have a high proportion of silt and fine sand, and cap or crust under raindrop impact. Most soils are dominated by low-activity clays with CEC ranging from 1 to 10 meq/100 g and have low organic matter

95 contents (about 1%). The bulk density of uncultivated soils is usually 1.4 g/cm 3. They have weakly developed structure, so the risk of erosion, by wind and water, is high. They cover about 60% of the region. Lawes (1961; 1962; 1966) shows that crop residue mulch is extremely effective in preventing capping, in maintaining a high infiltration rate, and in preventing runoff (Table 59). Mechanically disturbed bare soil may lose 70% of the total rainfall by runoff; even with cultivation at fortnightly intervals, 50% may be lost. In his study, the mean sorghum grain yield on plots mulched with groundnut shell was 29% higher than on the unmulched control. On sandy soils in Oyo State, Nigeria, Sabel-Khoschella (1988) notes that higher hydraulic conductivity of soils under a no-till system than under plough-till (Table 60). He also shows that a no-till system with crop residue mulch effectively reduced runoff and soil erosion (Table 61). Because of the difficulties of obtaining adequate supplies of mulch in semi-arid regions, erosion control is exercised by ridging and rough ploughing. Kowal (1970a) reports that ridges made up and down the slope encourage rather than control gully erosion. Bababe (1977) confirmed Kowal s findings and observed the least runoff and soil erosion from lands with a flat seedbed. Maximum runoff and soil erosion were observed from land ridged at 1 m intervals. Even tied ridges were not effective in preventing runoff and soil erosion. Table 59 shows that runoff and soil erosion are negligible if crops are sown on flat land and mulched with 4-6 t/ha of straw. Killed weeds can be a good source of mulch, and the channels left by their decaying roots usually improve water infiltration. Dunham (1979) and Dunham and Aremu (1979) reported significantly higher infiltration rate with no-till (54 mm/hr) than with plough-till and harrowing (18 mm/hr). The percentages of water-stable aggregates in the surface horizon of no-till plots were 38%, 42% and 57% compared with 15%, 19% and 33% on plough-till plots with size ranges of > 2 mm, 1-2 mm and mm, respectively. Although grain yields of maize, sorghum and seed cotton were lower on no-till than on plough-till treatments, yields under no-till were improved by retaining the crop residue mulch and applying N and P (Dunham, 1982a). Dunham (1982b) concluded that soil physical properties of the surface horizon were better under no-till, manual tillage, and ox-driven tillage than on land cultivated using tractors (Table 62). The proportion of water-stable aggregates and the infiltration rate decreased with increased tractor tillage. The relative yield for the no-till treatment increased in the third and fourth years, and was more than that of subsoiled, plough-till and ridge-till treatments.

96 Dunham concluded that the most effective way of maintaining surface soil stability and a high infiltration rate is through minimizing soil disturbance and increasing soil organic matter content. No-till farming, if done properly, with adequate weed control and liberal use of residue mulch, is potentially useful in the savannah zone of Nigeria. In a similar region in northern Ghana, Takayi (1970) observed a slight improvement in maize yield in ridged plots over the no-till control. The difference would have been greater had weed control been adequate. Leyenaar and Hunter (1977) concluded that maize grown on ridges was adversely affected by drought, high soil temperatures, and severe lodging. Semi-arid East Africa There have been relatively few tillage experiments in East Africa, although the importance of mulching and no-till for perennial crops has long been appreciated (Fuggles-Couchman, 1934; Jameson, 1970). Pereira and Jones (1954) observed that total pore space, percolation rate, and percentage of water-stable aggregates were smaller in a clean weeded bare soil than in soil with slashed or tall weed cover (Table 63). Discing and ploughing lowered total porosity, noncapillary pore space, and percolation rate compared with hand hoeing. Pereira and Jones (1954) conclude that

97 bare soil subjected to tropical rains declines rapidly and substantially in its ability to accept and transmit rainfall. Mulching, by whatever means, should be beneficial. Northwood and Macartney (1971), studying three soils in Tanzania, compared the effects of depth and width of cultivation on maize grain yield (Table 64). On the structurally inert Alfisol at Kongwa, strip tillage (20 cm wide, 9 cm deep) was superior. An increase in width of the tilled strip beyond 20 cm was not beneficial. Macartney et al. (1971) observed a complete failure of no-till maize on a compacted Alfisol with a bulk density of 1.52 g/cm 3. The initial high infiltration rate, established by ploughing, discing and harrowing, decreased markedly over 2 or 3 cropping seasons as soil structure deteriorated and the soil surface became sealed. In Tanzania, Huxley, (1975, 1979, 1983b), highlights the merits of no-till and minimum-till in a 3 year study (Table 65). The benefits for maize yield brought about by ploughing were smaller than those obtained in other

98 experiments by mulching or N fertilizer. Khatibu and Huxley (1979) found that effects of rates of N applications on the grain yield of cowpea were not greatly improved by no-till. At Morogoro, Maseri and Jana (1979) found that grain yields of maize and soybean with mulch and no-till treatments were equivalent or better than with plough-till (Table 66). Soil moisture reserves were enhanced in mulch and no-till treatments throughout the growing season. The potential of no-till farming in East Africa has been summarized by Diheng (1979) and Huxley (1983a). They conclude that the no-till method with crop residue mulch can be used to advantage for many soils and crops. In

99 an economic appraisal Diheng (1979) reported that the production costs of maize grain were 1.25 times higher with plough-till than with notill. When evaluating the economic feasibility of a no-till system, it is important to consider whether to improve indigenous farming systems by incorporating a no-till component or to introduce new systems. The results obtained in East and Central Africa (Morel and Quantin, 1972) indicate that, in the long run, mechanized intensive cultivation results in soil deterioration, and that the yields obtained by no-till and mulch systems are very promising. West African Sahel The sandy soils of the West African Sahel are marginal for food crop production. In addition to their droughty conditions and high soil temperatures, they have hard-setting characteristics, high strength, surface sealing and crusting. Formation of surface seal and crust is a major factor responsible for low infiltration and high runoff. These soils are also prone to wind erosion and seedlings are damaged by sand blasting and high temperatures. Crop residue mulch is not readily available. It is a precious commodity, and has alter-native uses such as feed, fuel, fencing material, and for household construction. Crusting, compaction and hard-setting are major constraints to crop production. Figure 28 shows linear decline in penetration resistance with increase in sand content for soils in Texas and Niger. In contrast, penetration resistance of these soils increased linearly with increase in clay content. The higher the amount of low activity clay, the greater the penetration resistance. Experiments have been conducted by the ICRISAT Sahelian Centre (ISC) at Niamey, Niger, to develop appropriate tillage methods to overcome poor crop germination and high seedling mortality. It has been shown that rough seedbeds produced by ploughing or ridging produce better yields of millet than no-till systems (Table 67). The high yields with a rough seedbed are due to better seedling survival. Table 68 shows a significantly higher survival

100 rate in plough-till and ridge-till seedbeds compared with no-till systems. However, when used in rotation with crop residues, the no-till system also produced favourable yields (Table 69). In other experiments at ISC, crop stand and seedling establishment of pearl millet were studied in relation to pre-sowing tillage methods. Figure 29 shows that crop establish-ment and survival later in the season were better for plough-till and ridge-till than with other treatments. These tillage methods produced satisfactory stands when dust storms occurred 8 to 10 days after sowing. Grain yields also depend on tillage method. Figure 30 shows that grain yields of pearl millet were significantly affected by tillage methods, mulching and fertilizer use. The positive effect of

101 residue mulch was more pronounced in treatments without than with fertilizer (Figure 31). Application of crop residue mulch also increased soil organic matter content. Tillage methods that protected seedlings from sandblasting improved crop stand and yield. Ridging, weeding and seeding methods are all significant in seedling survival. Figure 32 shows that ridging and hand weeding are extremely beneficial to yield. Manual operations are extremely labourintensive. Figure 33 shows the major advantages of improved tools and animal traction over human-powered traditional tools. Intercropping experiments conducted in Mali showed that transplanting pearl millet at 4-leaf stage of

102

103 maize produced optimum grain yield and high Land Equivalent Ratio (LER) for both maize and millet (Table 70). These studies show that a rough seedbed and intercropping are likely to optimize crop yield on these sandy soils. Experiments by the Institut Recherche Agronomique Tropi-cale (IRAT) have also demon-strated that mechanical loosening and subsoiling are necessary for satisfactory yields on hard-setting soils in the West African Sahel in Senegal. Nicou and Chopart (1979) showed that ploughing (with residue incor-poration and with additional mulch) significantly increased profile water content (Table 71), root growth into the subsoil (Table 72), and increased yields of most crops (Table 73). Chiselling and subsoiling also increased wheat yield in one out of three years (Table 74), and yields of maize and Stylosanthes (Table 75) in northern Nigeria. Subsoiling decreased soil bulk density and markedly improved soil moisture storage (Table 76).

104 Management of crop residue is also a key factor affecting crop yields. In Burkina Faso, retaining crop residue increased yield of maize and Crotolaria (Table 77). Application of residue mulch also increased the yield of cowpea grown in dry season on hydromorphic soils (Table 78). Experiments in the semi-arid region of Nigeria (Ekwue, 1990) show that soil organic matter content is significantly correlated with soil physical properties (Table 79).

105

106 Kouyte and Wendt (1992) compare the effects of no-till and combination subsoiling, ridges and tied ridges on soil properties and grain yields of sorghum and cowpea in the Sahel. During the year of the study, ridging increased sorghum yield over no-till by 113 and 70% for grain and stover yields respectively. Cowpea grain and hay yields increased by up to 91 and 102%. Water use efficiency was similarly affected by fertilization and tillage. This supports the conclusions of Nicou (1984) that deep ploughing and ridging are effective methods of seedbed preparation for these soils. Tropical Australia A no-till system with residue mulch has given very encouraging results on Alfisols near Katherine, Northern Territory (McCowen et al., 1980a, 1980b, 1985). When weed and pasture regrowth were adequately controlled by systemic herbicide, no-till maize and sorghum generally outyielded plough-till crops. Mulch between the rows decreased the maximum soil temperature by 8-16 o C beneath the mulch and by 4-5 o C in the rows themselves. There was more soil moisture under the mulch than in ploughed bare soil. Seedling establishment of maize and sorghum was more than 40% greater with no-till. Grain yield of maize was 5.6 t/ha with ploughing and 7.1 t/ha with no-till. Stover yield of sorghum was 20% higher with no-till. Re-establishment of pasture following no-till maize or sorghum production is usually easy. Peake et al. (1983) and Fisher and Phillips (1970) showed that re-establishment of Alysicarpus vaginalis, Stylosanthes humatas and S. guianensis was excellent after a maize crop treated with the herbicide atrazine.

107 Several of these tropical soils have low infiltration rate and high losses of water due to surface runoff. Experiments using simulated rainfall show that stubble mulch significantly increased infiltration rate in a no-till treatment (Table 80). Application of residue mulch to plough-till was less effective in increasing infiltration rate. Thiagalingam et al. (1991) record greater yields of maize and soybean with no-till than with plough-till treatments (Table 81). The increase in yield was 41% for maize and 20% for soybean. Yields of sorghum and groundnut were not effected by tillage method. Table 82 shows that high yields of maize were associated with greater uptake of P and Zn in the crop, greater uptake of K and S in the crop, and greater uptake of P and Zn in the crop in maize stover in no-till compared with plough-till treatments. There were also differences in nutrient uptake in maize grains related to tillage methods (Table 82). Experiments on Lixisols and Luvisols in the drier regions of western Australia show structural stability of notill soil to be better than that of plough-till treatments (Hamblin, 1980). Structural improvements, however, take place at a slower rate than in the wet tropics. In Fiji, Chandra (1977) demonstrates the merits of no-till and minimum-till systems for sorghum and other grain crops in semi-arid regions. South-Central India Lixisols and Luvisols in South-Central India have been intensively cultivated for many centuries. Though under lowinput, subsistence farming, they are more productive than those in West Africa. One of the factors is the nature of

108 the clay minerals. Those in India have mixed mineralogy rather than low activity clays. Mixed mineralogy gives more favourable soil structure and fertility status. Tillage experiments in India have been conducted by the Indian Council of Agricultural Research (ICAR) and ICRISAT. Shallow rooting depth crusting and compaction are major factors responsible for low yields. Table 83 shows that increasing rooting depth by ridge tillage increases grain yield of sorghum, redgram and sunflower. In shallow soils of 15 cm depth, increases in yield by ridge tillage were 57% in sorghum, 45% in redgram, and 43% in sunflower. In comparatively deep soil (35 cm deep) the increase in sorghum yield by ridge tillage was about 50% (Table 83). Ridge tillage reduced soil bulk density and increased infiltration rate. Table 84 shows that ridge tillage was more effective in increasing soil infiltration capacity than incorporation of paddy husk or addition of clay to the soil. The highest crop yield was obtained when clay was mixed into the soil. Addition of clay increased plantavailable water reserves and improved crop yield.

109 Sparse crop stands are a major factor giving low yields in rainfed agriculture in the semi-arid regions of India. These arise because of poor germination and emergence caused by low soil moisture contents and high soil temperatures at sowing time, crusting and hard-setting, and seedling mortality. Water uptake by seed, necessary for germination, is directly related to soil moisture content. Improving the seed-soil contact by soil compaction can enhance take up of water by seed provided that soil-water content is above a minimum threshold. Table 85 shows that soil moisture content is the most important factor in seed water uptake. Joshi (1987) records the percentage emergence of pearl millet (Table 86). Because of the favourable soil moisture regime, seedling emergence was greatly improved by sowing in the furrow and by the application of farmyard manure (FYM) over the seed furrows. There exists a direct positive correlation between emergence and yield of pearl millet. Table 87 shows the high yields obtained in treatments with furrow sowing and application of FYM over seed furrows.

110

111 Several experiments in India show that crust strength and seedling emergence are directly related. The higher the crust strength the poorer the seedling emergence. Cotton is extremely sensitive; Table 88 shows that its emergence was as low as 2% in heavily crusted soils. In comparison, millet seedlings are more robust but their emergence can also be as low as 5% in heavily crusted soils. Some of the soil surface management techniques for moisture conservation and enhancing crop growth on these soils are described in the next section. Techniques to conserve soil moisture and improve yield in South Central India Inter-row cultivation: Inter-row cultivation is widely adopted to break surface crust, to improve crop emergence and increase water infiltration. Table 89 shows that inter-row cultivation reduced runoff and soil loss markedly, but it increased yield only slightly. Surface configuration: Surface detention capacity can be enhanced by altering the surface configuration of these soils by forming a ridge/furrow system, and changing the orientation or spacing of the ridge furrow system. Table 90 for Luvisols at Hyderabad, India, shows that runoff was slightly smaller in wide than in narrow furrows by 3 to 20%.

112 Deep ploughing and subsoiling: Luvisols in central India respond favourably to deep ploughing. Alleviation of soil compaction, increase in surface detention capacity and formation of rough/cloddy surface by deep ploughing improves the seedbed environment and increases soil-water storage. Table 91 illustrates the beneficial effects of deep ploughing. Mouldboard ploughing to 25 cm depth decreased runoff and soil losses; consequently, sorghum grain yield was increased. It is clear that the optimum depth of ploughing is 25 cm. Table 92 shows that additional shallow tillage

113 increased yield by 16% for intercrop sorghum and by 13% for sole crop of pearl millet, presumably through improvements in root system development. Table 93 shows directly that subsoiling increased root system development. As well as improving surface detention capacity and decreasing runoff and soil loss (Table 91), deep ploughing also increases infiltration capacity. Table 94 shows that cumulative infiltration increased with increasing depth of ploughing. Broad bed and furrow system (BBF): Runoff on Luvisols with low infiltration rate can be managed by the BBF system developed at ICRISAT. This system is designed to increase runoff and collect the water in a storage tank for possible irrigation use. Table 95 shows that the BBF system increased runoff by 22% from shallow tillage and 35% from a no-till system. The corresponding increases in soil loss were 84% and 147%, respectively. Accelerated soil erosion caused by increases in runoff can be controlled by engineering gentle gradients within the BBF system. Soil losses are small on a gentle gradient of 1 in 1000 (0.1%). Table 96 gives an example of the effectiveness of different systems on runoff and soil loss, and on crop yields. The most conservation-effective system for Luvisols in Central India was sowing on the flat seedbed with a gradient of 0.4%. Maximum runoff and soil erosion were observed on the BBF system with 0.4% grade.

114 Surface configuration can also be altered by contour cultivation and a simple ridge-furrow system. For resource-poor farmers in South Asia, contour cultivation and ridge-furrow systems are easier to adopt than a BBF system. Table 97 shows that crop yield with contour cultivation and trenching to conserve water was better than other systems. Trenching is a very effective means of conserving water. Trenches soak in excessive runoff and recharge the soil profile (Table 98). A combination of trenching, at appropriate spacing, and contour cultivation may be the most effective approach.

115 SANDY SOILS OF LOW WATER HOLDING CAPACITY Many tropical soils are coarse-textured, and have low soil organic matter contents. They have low water reserves, high infiltration rates and relatively high soil bulk densities (Tables 99 and 100). Crops grown on such soils are prone to drought stress. Low yields result from poor crop stands, high losses due to soil evaporation, and the interaction of drought stress and unfavourable soil temperate regime. The available water holding capacity of these soils is generally limited by their shallow depth. The management of sand soils must aim to: increase soil moisture storage capacity, decrease soil evaporation, and increase rooting depth. Increasing soil moisture storage capacity The soil moisture storage capacity of sandy soils can be increased by increasing the proportion of small pores that retain water. Sandy soils generally have more transmission (macropores) than retention pores (micropores). Some practical ways of increasing the ratio of fine to coarse pores are briefly described. Soil compaction Compacting sandy soils decreases total porosity by reducing the proportion of macropores. Onwalu and Anazodo (1989) determined the optimum range of bulk density in relation to yield. Compacting the soil to increase the bulk density from 1.05 g/cm 3 to 1.2 g/cm 3 increased crop yield. Indian experiments have also shown that appropriate soil compaction of coarse-textured soils can save 15 to 36% of irrigation water (Table 101).

116

117 The use of 200 kg and 460 kg rollers increased yields of pearl millet and wheat at two sites in India. For sandy soil at Durgapura, increases in yield of pearl millet using a 200 kg roller were 18.9% for 10 passes, 35.2% for 20 passes, and 50.8% for 30 passes (Table 102). The corresponding increase in wheat yield was 19.6, 25.9 and 25.3% for 10, 20 and 30 passes, respectively. For a loamy sand soil at Jobner the increases in pearl millet yield by a 460 kg roller were 16.5% for 4 passes, 18.7% for 8 passes, and 39.6% for 12 passes (Table 102). Results of similar experiences in northeastern Nigeria are shown in Table 103. The results were very variable and maize grain yields were not significantly different. There was, however, an increasing trend in grain yield with increasing tractor passes up to 15 in number. Subsequent passes decreased yield. In contrast, soil compaction experiments in southern Nigeria (Kayambo et al., 1986a, b, 1991) showed marked reductions in yields of maize, soybean and cowpea with 3 passes of a 2 ton roller. The limit of soil compaction, beyond which crop growth is adversely affected, is different for different soils and crops. Moderate levels of soil compaction are generally beneficial. The key is to identify the critical limits of bulk density for different soils and crops. Clay mixing A slight increase in the clay content of a sandy soil can enhance its soil-water storage capacity. A practical way to increase clay content is soil inversion, in which more clayey subsoil is brought to the surface. Limited amounts of clayey material can also be brought in from else-where and mixed with the soil. Table 84 shows that addition of 2% clay increased maize grain yield. The benefits of mixing

118 clay were greatest in shallow soil. The effects of mixing clay at 3 levels (1% and 2% compared with 0% control) on yields of pearl millet and wheat for four levels of compaction are shown in Table 104. Increasing soil organic matter content A practical means of increasing soil organic matter is the addition of compost or manure. Application of farmyard manure (FYM) is common in arid and semi-arid regions with mixed farming systems. Table 105 shows that row application of FYM increased yield of pearl millet and cotton. It is apparent from the data that crusting was an overwhelming problem in this soil. Table 106 shows that application of FYM at 30 t/ha significantly increased maize grain yield for one out of 3 soils (Soil No. 2). Decreasing soil evaporation Evaporation losses from sandy soils are usually large, especially during first and second stages. Although dust mulch is widely used to reduce evaporation, its effectiveness is questionable. Crop residue mulch reduces direct insolation to the soil surface, decreases soil temperature, reduces evaporative demand, and decreases the evaporation rate in the first stage (Figure 16). Data from Faizabad, India (Table 107) show that application of mulch increased yields under three tillage systems, the most successful being plough-till. Crop residue management is the most practical means of mulch procurement. The data in Table 108 from Ludhiana, India, show that residue removal significantly decreased wheat grain yield. The highest yield was obtained

119 in a no-till system with residue mulch. Tables 108 and 109 show the positive effects of retained residue on crop yield. Table 109 shows that water use efficiency was 5 to 13% greater with than without mulch in 1982, and 1 to 11% greater in Table 110 also shows a substantial increase in dry matter yield by mulch application. Mulch improved the dry matter yield for three tillage methods. Increasing rooting depth Deep ploughing and subsoiling are commonly used to increase rooting depth on shallow soils. These techniques are energy-intensive and require motorized equipment. Furthermore, the beneficial effects of soil loosening are transient

120 unless steps are taken to avoid recompaction. Several studies have been conducted on sandy soils to assess the beneficial effects of deep tillage on crop growth and yield. The data in Table 106 show the beneficial effect of deep ploughing. The effect of deep ploughing on maize yield was more significant (2 out of 3 soils) than that of FYM (1

121 out of 3 soils). As well as increasing water infiltration, deep ploughing increased rooting depth and enabled the crop to exploit soil moisture reserves from subsoil horizons. Other experiments conducted on sandy soils in Punjab, India, by Arora et al. (1991) showed that deep ploughing to 30 cm depth significantly increased maize grain yield and root length index in four soils (Table 111). VERTISOLS Vertisols, predominant in the semi-arid tropics, are found extensively in West Africa, India, the Caribbean and tropical Australia. Under favourable conditions, these structurally active soils exhibit self-mulching characteristics. Because of their poor trafficability when wet, they are often left fallow during the rainy season when accelerated erosion can become severe. Major constraints on intensive land-use include accelerated soil erosion during the

122 monsoon season, drought stress during the post-rainy season, and formation of deep cracks when dry. Consequently, crop yields on Vertisols using traditional systems of management are low. The data in Figure 34 for a Vertisol at Hyderabad, India, show declining trends in yield for crop species in rotations. Although there were improvements in yield when legumes were included in rotation, the overall declining trend was consistent. Low yields on Vertisols are also reported from West Asia and North Africa. The data in Table 112 indicate the range of yields obtained on Vertisols in North Africa. Given proper management, however, Vertisols have a high productive potential. Tillage methods can play a major role in reversing the degradative trends and improving yields. The following methods can be used to improve crop production on Vertisols. Erosion control Preventive measures involve the use of crop residue mulch which decreases splash by decreasing raindrop impact and by reducing the heat of wetting. Freebairn and Wockner (1982) reported that tillage systems that leave the most residue intact on the surface were the most effective in reducing soil erosion. Freebairn et al. (1986) reported the results of experiments involving the use of different types of reduce-till systems and crop residue

123 management on soil erosion. Table 113 shows that soil erosion was drastically reduced by no-till and stubble mulch. No-till and stubble mulch treatments were even more effective in erosion control at Greenwood (Freebairn et al., 1986). As well as controlling erosion, the no-till treatment reduced the peak runoff rate. The no-till system is not as effective as the stubble mulch system, especially where mulch supplies are inadequate and the soil is compacted by traffic. More water was stored in the root zone for crop use with no-till and stubble mulch farming than with the bare fallow management (Table 114). Sallaway et al. (1983) observe that runoff ranged from zero for a no-till treatment to 46.4 mm for disc ploughing treatment. Soil erosion also ranged from zero for the no-till treatment to 13.3 t/ha for bare fallow planted to sorghum. Marston (1978) observed in New South Wales, Australia, that erosion of plough-till systems generally increased with the removal of crop residue by burning or grazing and with excessive cultivation during fallowing. Erosion control measures on Vertisols are based on the principle of safe disposal of excess runoff. In this connection, considerable progress has been made by ICRISAT through the development and testing of their BBF system (Figures 35 and 36). This is in essence a drainage system. It is designed to encourage surface drainage with a minimum risk of erosion. It improves soil aeration for upland crops in the rainy season. Figure 37 shows marked increases in air-filled porosity in the top 30 cm layer of the BBF system compared with a flat seedbed. Table 115 shows that the BBF system with 0.4% grade, or the BBF system with 0.6% grade and field bunds, effectively decreased runoff amount and soil loss. Permeability of the furrow zone is affected by the tillage system. Cultivation of furrows, especially when soil is wet, can decrease the infiltration rate (Figure 38).

124 ICRISAT have shown the use-fulness of deep tillage and the use of phosphogypsum in reducing run-off and soil loss from a Vertisol. Table 116 shows that the traditional shallow tillage and no-till system produced excessive runoff and caused considerable soil loss. In contrast, 30 cm deep tillage in combination with a crop residue mulch or phosphogypsum reduced runoff amount and soil loss. While deep tillage may have improved infiltration rate, application of phosphogypsum probably improved soil structure and decreased slaking and splash.

125

126

127

128 Soil compaction and cracking Crack formation is one of the principal features of Vertisols. The width and depth of crack formation depends partly on the degree of soil compaction. Cracks are often deeper and wider on compacted than on uncompacted soil (Table 117). The degree of soil compaction partly determines the net energy requirement for tillage operations, which on these heavy soils are energy-intensive. In Sudan, Willcocks (1987) measured the energy requirements for different intensities and types of tillage operations. Table 118 indicates that energy requirements increase with increasing intensity of soil manipulation. Water conservation Vertisols suffer poor drainage, inundation and anaerobiosis during the rains and drought stress during the post-rainy season. Several soil-water conservation techniques have been tried, some of which are described below. Fallowing Fallowing during the rainy season has been advocated as a means of conserving water for intensive cultivation during the post-rainy season. Table 119 shows the effectiveness of fallowing within a wheat-sunflower rotation. Unless managed properly, fallowing can be inefficient and counter-productive. Its efficiency depends on soil surface management during the fallow phase. Yule (1987) shows that fallow efficiency can be negative (Table 120). The maximum efficiency in a 4-year study was 35%. The type of tillage, quantity of residue mulch and degree of weed infestation have significant effects on soil moisture conservation and fallowing efficiency. Thomas et al. (1990) observe that presence of residue mulch is a key factor in soil moisture storage during fallowing (Table 121). From the data available, it seems that cropping during the rainy season followed by a double cropping during the post rainy period is a possible strategy to utilize soil-water resources effectively, and so increase agronomic productivity.

129 Tillage methods A wide range of tillage methods has been used in attempts to conserve water in Vertisols. There has been an equally wide range of claims for success. Some researchers conclude that reduced or no-till systems are extremely effective, especially on soils with self-mulching properties. Others support deep tillage and subsoiling. These conflicting claims are outlined below. In subtropical Israel, Stibbe and Ariel (1970) compared sorghum grain yields under plough-till and no-till and observed that yield responses to tillage method depended on rainfall amount and distribution. Yield in 1967 with sufficient, well distributed rains was suppressed by 40% under a no-till system. Yield in 1968 with low rainfall was 10-16% lower in a plough-till than no-till system. These soils contain about 35% clay in the surface horizon. In a similar study on a Vertisol containing 64% clay, Stibbe and Ariel (1970) reported significantly lower cotton yield for no-till than plough-till in 1969, a wet season compared with 1970 (Table 122). Increased cotton yield during 1969 in the subsoiled plots was due to removal of traffic-induced pan at 30 cm depth. In 1970, however, subsoiling produced lower yields than the no-till treatment. Stibbe and Ariel conclude, that if there is no plough pan and a soil moisture deficit is expected, no-till is successful for summer crops seeded directly through winter wheat stubbles.

130 In Ghana, Mante (1979) investigated the influence of depth of ploughing on the physical properties of a Vertisol and on sorghum grain yield (Table 123). Cultivation to 7.5 cm depth gave low bulk density, a favourable soil moisture content and low power requirements. No differences in sorghum grain yield were evident. On self-mulching Vertisols, small advantages in grain yield from shallow cultivation may not be worth the extra effort.

131 Experiments on Vertisols in India also indicate that satisfactory yields are obtainable with minimum-till. Srivastava and Singh (1972) found no significant differences in seed cotton yield between mouldboard ploughing plus two harrowings, cultivation to 15 cm depth, and strip tillage in the row zone only. These conclusions were also supported by the watershed management experiments conducted by ICRISAT at Hyderabad. Extensive studies of the effect of tillage methods on water conservation have been conducted on Vertisols in Australia. Tables 124 and 125 show the effects of tillage methods and residue management on water use by sorghum in its different growth stages. Table 126 shows that the mean water use ranged from 198 mm for no-till without residue mulch to 208 mm for disc or blade cultivations with residue mulch. The use of crop residue mulch was more important than the type of equipment used for soil tillage. Several experiments have been conduced on Vertisols in central India to evaluate the effects of ridges and raised beds on soil-water conservation. Table 126 shows that plant populations, seedling growth, and grain yields of three crops were superior on ridges than on flat seedbeds. The effects were consistent on all slopes. Table 127 shows that raised beds at 6 m intervals produced the best yield. The increase in crop yield on ridges and raised beds is apparently due to improved drainage during potentially anaerobic wet periods. In contrast, sunken beds (furrows or depressions) may be more suitable for some crops during periods of drought. The data for paddy rice in Table 127 support this view because paddy yield was increased by 37% when grown in a sunken bed compared with the control. However, yield of gram grown in sunken beds was suppressed by 8% due to waterlogging. Green gram is sensitive to waterlogging but rice, a semi-aquatic plant, benefits from it. Water management is the crucial element in the cultivation of Vertisols. Excess water during rains can be managed by the BBF system or raised bed technique. Water conservation to alleviate drought requires the use of crop residue mulch and an appropriate system of seedbed preparation. Fine tuning these systems to meet local requirements is crucial to the sustainable use of fragile resources. Crop management Soil and water conservation on Vertisols can be enhanced by careful crop management. Three aspects of crop management need particular attention to obtain satisfactory yields on Vertisols. These are to ensure a satisfactory

132 crop stand, prolong the growing period, and improve soil fertility. The range of optimum soil moisture content for sowing on Vertisols is rather narrow. The seed-soil contact is often poor when the soil is dry. The soil structure is easily destroyed by trafficking by humans, animals and vehicles if the soil is too wet. It is critical that farm operations are performed when the soil moisture content is within the optimum range. Figure 39 shows the adverse effects of trampling on plant populations of chickpea and safflower. Plant stand on trampled beds was only 50 to 70% of the untrampled control. Pre-season cultivation can facilitate early sowing. Early sowing, with proper systems of seedbed preparation, can prolong the growing season and enhance the chances of a successful crop if the rains cease earlier than expected. Willcocks (1987) investigated dates of sowing on a Vertisol in south Kordovan, Sudan. The data from his 3-year study show marked reductions in cotton yield due to delayed sowing. Tillage systems also play an important role in soil fertility management. The efficiency of chemical fertilizer is influenced by tillage methods and crop residue management. Figure 40 shows the response of rainy-season sorghum

133 to nitrogen application. Deep cultivation (DC) had a favourable response because nitrogen-use-efficiency is enhanced by improvements in internal drainage. Systems approach to management of Vertisols Sustainable intensive management of Vertisols requires a system approach. Various components and subsystems must be combined to get the best management practice. A system approach is needed to develop, validate, adapt and fine-tune procedures for specific site conditions. The literature survey presented above indicates the usefulness of various techniques in conserving water, reducing soil erosion and increasing crop yields. Important among these are: (i) use of crop residue mulch, (ii) deep tillage, (iii) raised beds in form of ridges or broadbeds and furrow system, (iv) double cropping with crops both in rainy and post-rainy seasons, and (v) multiple cropping: growing more than one crop simultaneously on the same piece of land. These practices should be combined with balanced fertilizer use. Several attempts have been made at ICRISAT in India to develop soil-specific manage-ment packages. One such package is illustrated by Table 128. This shows the usefulness of crop residue incorporation at 5 t/ha with 30 cm deep primary tillage. Deep incorporation of crop residue increased yield in comparison with no-till by 43% in maize and 276% in cowpea during 1983/84 compared with 40% in maize and 268% in cowpea during 1984/85.

134 It is clear that improved management systems can double or even triple crop yields from these soils. Another example of the systems approach to the manage-ment of Vertisols is illustrated by Table 129. Kanwar and Virmani (1987) compared the productivity of improved and traditional sys-tems of management. The pro-ductivity of the former was 2 to 3 times that of the latter. It is evident that the best tillage practices must be combined with the best techniques of residue management, with improved cultivars and cropping systems, appropriate fertilizer management, and time of sowing, to enhance productivity from these soils. ACID TROPICAL SOILS Acrisols, Ferralsols and to a lesser extent Nitisols, Alisols and Plinthosols are common in the sub-humid and humid tropics, are highly weathered, strongly leached, acid, low in basic cations, and have often high concentrations of Al and Mn in the subsoil horizons. In general, Acrisols and Ferralsols have more favourable physical properties than Lixisols. Acrisols Acrisols have severe nutrient imbalance including toxicity of Al and Mn. With appropriate land use and balanced fertilizer inputs, satisfactory yields can be achieved under continuous cropping (Sanchez et al., 1982). Because Ultisols have low mineral fertility and are acid, the choice of crops to be grown is limited. Regardless of tillage method, maize does not grow well on unlimed Ultisols. Lime is not everywhere readily available and where it is, it is often necessary to incorporate it by ploughing to help neutralize sub-soil acidity. A practical alternative is to choose crops that can be grown without liming such as cowpea, banana, cocoyam, yam and cassava. Erosion, compaction, crusting, drought stress and high soil temperatures are also important limitations on many Ultisols in the humid tropics (Pla Sentis et al., 1979). There have been few experiments to help develop appropriate tillage systems for acid tropical soils. With their high chemical fertility, Alfisols are generally the first choice for arable use. In the Amazon Basin, for example,

135 Terra Rossa and flood-plain soils (Cambisols) are intensively used for food crop production, rather than nutrientdeficient Acrisols and Ferralsols. Crop residue mulch is readily available in the humid tropics, so no-till and conservation tillage systems can be suitable for some crops on acid soils in spite of their chemical limitations. Grain crops No-till and conservation tillage systems are successful for grain crops in humid regions. With adequate liming, no-till maize outyielded plough tillage on Ultisols in Liberia (Lal and Dinkins, 1979) and southeastern Nigeria (Rodriguez and Lal, 1979). In Liberia, maize grain yield was 2.8 and 1.3 t/ha with no-till and plough-till, respectively. On ploughed plots highly acid subsoil was brought to the surface giving an unfavourable physical and chemical environment for root growth. This resulted in a maize grain yield of 3.8 t/ha under a limed no-till compared with 2.6 t/ha for a limed plough-till system. Maurya and Lal (1979b) reported slightly more maize grain yield from plough-till than no-till on an Acrisol in eastern Nigeria. The superior yield was attributed to deeper root system development when lime was incorporated by ploughing (Figure 41). Hayward et al. (1980) report that it is common practice in many Central American countries to grow maize with no-till in hilly areas.

136

137 Lal and Dinkins (1979) reported upland rice yields of 2.9 and 1.9 t/ha for no-till and plough-till systems on an Ultisol in Liberia. Similarly, Maurya and Lal (1979b) reported 80% more cowpea yield from a no-till than plough-till. Crop response is greatly influenced by mulching. It not only has beneficial physi-cal effects but it supplies nutrients and stimulates the soil fauna. Ogunremi et al. (1986a) reported lower grain yields of upland rice with no-till than from ploughed treatments without a mulch (Figure 42). With mulching, however, equivalent yield was obtained in no-till and plough-till treatments. Soil compaction caused by mechanized operations may, however, eventually cause severe yield reductions in no-till upland rice. Mechanized farm operations cause severe compaction on most tropical soils, and the no-till system is not readily suited to compacted soils. Perez Escolar and Ortiz Lugo (1973) report from Puerto Rico that sugarcane yields from compacted Acrisols were improved by deep ploughing. Compacted soils are preferably restored by fallowing with deep-rooted crops rather than by deep ploughing. Root crops Tropical root crops, for example cassava and yam, are better adapted than grain crops to acid soils in humid environments. Okigbo (1979) reported no significant differences in cassava tuber yield in relation to the frequency and intensity of cultivation before planting. Similar observations were made by Maurya and Lal (1979b). Akobundu (1983), however, records significantly more cassava tuber yield on ridged and ploughed than on no-till plots: the mean tuber yield was 13.5, 13.4, and 8.3 t/ha, respectively. The lower yields on no-till plots were attributed to severe weed infestations (Akobundu, 1983). In a mulching and tillage study, Maduakor et al. (1984) show significant positive effects of mulch but no effect of tillage methods on tuber yield of yam on an Acrisol in eastern Nigeria. The mean tuber yield on a ridged

138

139 seedbed was 14.1 t/ha compared with 14.8 t/ha on a flat unridged seedbed. On comparatively deep sandy loam soils, tillage has little effect on tuber development and there are no obvious advantages of ridging over flat planting. However, mulch improves soil moisture and temperature regimes and tuber yield. Ohiri and Ezumah (1990) evaluated cassava response to tillage methods on an Acrisol at Umuahia, eastern Nigeria. Their data for do not show significant differences in fresh tuber and top yields between tillage treatments. Tuber yield was also not significantly influenced by tillage treatments in However, top yield in was significantly different among tillage treatments. The top yield in no-till plots significantly outyielded that in plough-till treatments. Ferralsols In central Brazil and the Amazon Basin Ferralsols are susceptible to water erosion and have low available waterholding capacity. Crops often suffer from drought even a few days after heavy rains. Seedbed preparation under these conditions, therefore, should aim to conserve soil and water resources, increase water use efficiency, and so sustain economic productivity. In Brazil, Sidiras et al. (1982; 1983) observe significant improvements in water-stable aggregates and soilwater retention at 0.06, 0.33 and 1 bar suctions after 4 years of continuous no-till farming (Table 130). Improvements in soil moisture retention in the no-till system were most pronounced in the top 20 cm layer. Crop residue mulch on no-till plots improved soil structure and infiltration rates, and significantly decreased runoff and soil losses on 4% slope. Compared with a plough-till system, grain yield of soybean was improved by 4% with chisel plough and by 7% with the no-till system (Sidiras et al., 1982). In a study on a Typic Haplorthox, Kemper and Derpsch (1981) reported soybean grain yields of 1434, 1505 and 1987 kg/ha on plough-till, minimum-till and no-till seedbeds, respectively. Effects of tillage methods on physical properties of Ferralsols are described by Roth et al. (1988). Table 131 shows significant differences in soil bulk density, total porosity, and on pore size distribution. The no-till plots had slightly higher bulk densities in the topsoil than minimum or plough-till plots. The topsoil in no-till plots had fewer macropores and more micropores than under plough-till. Barber et al. (1989) and Barbosa et al. (1989) show the effects of different tillage methods on Ferralsols in Bolivia. Table 4 shows that soil compaction is a serious problem in many cultivated soils. Because of predominance of low activity clays, these soils do not shrink and swell and so restore soil structure. Consequently, mechanical loosening is necessary to improve soil water movement. Table 132 shows significant decreases in soil bulk density and cone index from subsoiling and controlling traffic. Controlling traffic is an key strategy, otherwise loosened soil is easily recompacted. Table 133 shows that improvements in soil physical properties also improve plant nutrition and crop yield. Table 134 corroborates Table 133. The maximum soybean yield was obtained in the treatment where subsoiling was done every year. Tropical root crops respond differently to no-till than grain crops. In addition to the problem of weed control, additional soil volume (or root room) is required for proper development of bulky tuberous roots. Development of root tubers can be better in ploughed than untilled Oxisol. However, ploughing is not so essential for recently developed Inceptisols with favourable subsoil properties. Ezumah (1983) observed the effects of tillage methods on

140 cassava in a Ferralsol in Zaire and reported inferior plant stands for no-till compared to tilled and ridged treatments. The cassava tuber yield was 7.3, 5.9, and 4.0 t/ha for ploughed, ridged and no-till treatments in a Ferralsol, compared with 18.9, 16.6 and 16.4 t/ha, respectively, for a sandy loam Cambisol. SODIC SOILS Accumulation of excess salts in the root zone limits crop production and results in a partial or complete loss of soil productivity. The problems of soil salinity are most widespread in arid and semi-arid regions, especially under irrigation but also occurs extensively in sub-humid and humid climates particularly in the coastal regions. There are two main groups of salt-affected soils, namely, saline soils and sodic soils. Saline soils (Solonchaks) contain sufficient neutral soluble salts to adversely effect the growth of most crops. Sodic soils (Solonetz) contain measurable to appreciable quantities of salt capable of alkaline hydrolysis (e.g. Na 2 CO 3 ) but appreciable quantities of neutral soluble salt are generally absent. The soil is characterized as saline if the conductivity of saturated soil extract exceeds 4 ds/m at 25 o C and sodic if exchangeable sodic percentage (ESP) is of 15 or more and conductivity is

141 generally less than 4 ds/m at 25 o C (may be more if appreciable quantities of Na 2 CO 3, etc., are present. Climatic factors that favour formation of salt-affected soils include low rainfall, high evaporative demand, high temperatures and low humidity. Poor drainage is also a factor. Soils with poor internal drainage accumulate salt in the root zone. High concentration of sodium ions on the exchange complex deflocculates and disperses the clay, which moves into the subsoil to form a low permeable layer with massive structure. As a result, sodic soils are poorly drained, and are hard to cultivate. The salt may come from soil parent materials or groundwater, from irrigation water taken from rivers or groundwater, from sea encroachment in coastal areas; or be blown in by wind. Excessive use of fertilizers and other chemicals coupled with poor drainage may also lead to salinization. The total area of Solonetz and Solonchaks in the world is estimated at about 325 million ha (Table 2), excluding less expressed saline and sodic phases. Reclamation of sodic soils (Solonetz) by tillage Reclamation of sodic soils requires the replacement of Na + on the exchange complex by Ca ++ and leaching of Na + out of the root zone. Soil permeability and internal drainage must also be improved so the displaced Na + can be leached out of the root zone. Common mineral amendments used are gypsum, phosphogypsum, calcite and other acid-forming salts, e.g. iron and aluminium sulphates, lime-sulphur and pyrites. Table 128 illustrates the application of phosphogypsum on a Vertisol in India. Application of phosphogypsum improved soil aggregation, decreased structural slaking and increased infiltration capacity, thus reducing runoff and soil erosion. Rice culture can play a key role in reclamation of salt-affected soils, because continuous flooding helps to leach salt out of the root zone. Shallow mixing of gypsum and organic materials is very effective in reclamation of sodic soils (Gupta and Abrol, 1990). For upland crops (wheat, maize, etc.), tillage and crop residue management play a major role in reclamation of sodic soils. Key techniques are the addition of organic matter to enhance aggregation, and the application of gypsum to replace Na + by Ca ++. Deep tillage may be required if subsoil compaction and massive structure are severe. However, deep tillage should be carried out after removing and reclaiming the sodicity, otherwise it will cause complete disturbance and collapse of the soil structure. Table 135 shows that the three highest 4-year yields of wheat were for treatments with stubble retention, the two highest yields were for plots with stubble retention and gypsum application, and the highest yield was obtained on the plot with stubble retention mouldboard ploughing and gypsum application. The tillage methods had significant effects on available soil-water, soil-water accumulation, and on net soil-water extraction by wheat crops (Table 136). There was no direct correlation between average wheat yield for the 4-year period and the net average soilwater extraction. This is expected because of major differences in soil-water balance between years. Table 137 indicates water use efficiency (WUE) of wheat for different tillage and stubble management treatments. There were significant differences in WUE among tillage treatments in 2 out of 4 years. In 1984, the two highest WUEs were observed for the plough-till with blade plough and reduced-till with disc plough. In 1987, the two highest WUE s were observed for reduced-till with blade plough and reduced-till with disc plough. For 1985 and 1986, the

142 stubble removal treatment had a significantly higher WUE than stubble retention treatment. Comparison of the data in Tables 135 and 137 shows that the highest yields were not obtained for treatments with the highest WUE. With use of gypsum or amendments and the addition of organic matter or its maintenance, salt-affected soils can be adequately managed under no-till or reduced-till systems provided internal drainage is adequate and the soil structure is not too massive. HARD-SETTING SOILS Hard-setting soils are soils that set to a hard, structureless mass during drying and are thereafter difficult or impossible to cultivate until the soil profile is rewetted (Mullins et al., 1990). They occur widely in tropical Africa, Asia, Australasia and Latin America. These soils are characterized by low-activity clays and low levels of soil organic carbon. Crops grown in these soils suffer impeded root growth and drought stress. Table 138 compares the morphological characteristics of soils that set hard with those that do not. The major differences lie in soil structure and consistency. The soft and good tilth of hard-setting soils is easily lost on drying. The degree and intensity of setting increases with clay content. The extensive research on the tillage requirements for hard-setting soils in the West African Sahel done by IRAT is described in Chapter 4. The research on Luvisols in Central India is equally relevant. Useful information is available from tillage and soil surface research in Australia. The basic strategy for sustainable management of hardsetting soils is to increase the soil organic matter content, and ameliorate soil structure. These are, however, longterm goals and require frequent additions of organic materials and the adoption of appropriate land-use and cropping

143

144 systems over a long-term. In the short-term, however, it is helpful to learn to live with the hard-setting characteristic. In this connection, timing and type of tillage operations, and land-use or cropping intensity, are important considerations. For example, higher yields can be obtained with frequent and long fallowing. The latter enhances structural attributes, improves the soil s properties and increases soil-water reserves. The beneficial effects of ploughing on crop yields for soils in West Africa are well documented, but the data in Table 139 from Australian experiments do not show significant differences in wheat yield between tillage methods. One of the difficulties in objective evaluation of yield among tillage treatments is the timing of operations. Sowing on hard-setting soils is commonly delayed in plough-based systems because the soil is too wet, and with no-till or reduce-till systems because it is too dry and hard. New crops and improved cultivars should be considered for management of these soils. There may be some crops, preferably biennials and perennials with deep tap root systems, that may be more easily adapted than others. Experiments in Nigeria (Hulugalle and Lal, 1986) show that perennial pigeon pea has ameliorative effects on the structure of these hard-to-manage soils.

145

146 Chapter 7 Conservation tillage in the tropics Conservation tillage methods are used for growing crops in diverse soils and climates. There are many reasons why such systems in which crop residue is used as mulch should be preferred to plough-till systems based on soil inversion and residue incorporation. A widely used method of conservation tillage is the no-till system. Some of the benefits of no-till systems are listed below. IMPROVEMENTS IN SOIL STRUCTURE Properly implemented, no-till systems improve soil structure, increase the volume of biochannels and macropores, increase water infiltration, and decrease susceptibility to crusting. In some soils, no-till also increases the homogeneity of soil structure, and maintains relatively lower bulk density. In Brazil, Machado (1976) observed lower bulk density and more pores in the 0-15 cm layer of no-till plots than in ploughed soil (Table 140). In northern Nigeria, Ike (1986) observed more water stable aggregates in the surface (0-20 cm) layer of no-till and manually-tilled plots than in mechanically-tilled soil (Table 141). The surface (0-5 cm) layer of soil tilled by manual or mechanical methods was less dense than the no-till soil. Similar results were obtained by Aina (1979) for an Lixisol in southwest Nigeria. Adem and Tisdall (1984) record that, in southeastern Australia, the volume of water-stable aggregates increased by 48% when a cultivated red-brown earth was left untilled for 6 months. In southwest Nigeria, Lal (1982) reports significantly higher infiltration rates in no-till compared with ploughed plots. In Parana, Brazil, Sidiras and Roth (1985) also report significantly better infiltration in no-till compared with plough-till treatment. Depending on soil conditions, mechanical soil disturbance may have a beneficial effect on the structure of some soils. Figures 43 and 44 indicate the beneficial effects of no-till and conservation tillage on infiltration capacity for Ferralsols in Brazil. Improvements in soil structure by no-till are attributed to increased biotic activity, especially that of earthworms, which thrive better in no-till than in plough-till soils (Lal, 1987a; Lavelle, 1984). Use of herbicides to suppress grass decreases the activity of earthworms, which is most suppressed, however, in cultivated plots (Lal, 1982). As a result of intensified biotic activity, soil under continuous no-till develops a structural profile distinct from that of ploughed soil (Figure 45). Surface soil under no-till contains high amounts of organic matter. Under mechanized systems, soil structure can be platy. In soils with high biotic activity, surface soil may have a honeycomb-like structure. The number of visible pores is often used as an index of soil structure, and the suitability of a soil for notill culture can be assessed by the number of biochannels visible during field examination. Gowman et al. (1987)

147

148

149 suggest that a large number of visible pores (>100 ìm in diameter) is the best guide for successful application of a notill system. The infiltration rate in untilled soil can be slower than in ploughed soil because of soil compaction caused by trafficing. The use of residue cover may also in some soils decrease infiltration rate and moisture retention. In loess-derived soils in northern Nigeria, for example, Maurya (1986) observed a lower infiltration rate with crop residue mulch than without. SOIL AND WATER CONSERVATION Soil erosion is effectively controlled by no-till and other conservation tillage systems. The beneficial effects of the no-till system on soil and water conservation are extensively documented for a wide range of soils in the tropics and subtropics (Lal, 1984). In southwestern Nigeria Lal (1976a) observed that a notill system drastically reduced soil erosion on slopes of up to 15%. In Parana, Brazil, Sidiras and Roth (1985), using a rainfall simulator test, recorded greater infiltration in no-till and minimum till fields compared with ploughed land. In northern Thailand Ryan (1986) observed that water runoff and soil erosion increased with increasing frequency and intensity of mechanical tillage (Table 142). In upland rice culture, the mean annual soil erosion was 12.2 t/ha for ploughtill and 0.8 and 1.0 t/ha for no-till and minimum-till treatments, respectively. In Queensland, Australia, Mullins et al. (1984) observed that serious soil erosion losses of t/ha/yr on land with burnt trash were reduced to acceptable levels by adopting no-till or minimum till systems. Conservation tillage helps erosion control by restricting sediment sources and soil transport. The principal aims in erosion control are to: improve aggregation and encourage a large proportion of water-stable aggregates. This is done by raising soil organic matter contents, thus encouraging biotic activity. Soil detachability can also be reduced by encouraging rooting in the topsoil (Maurya and Lal, 1979a); reduce inter-rill erosion and soil splash by preventing raindrop impact on the soil surface; reduce rill erosion by decreasing runoff rate, amount and velocity by avoiding surface crusting and by encouraging stable and continuous biochannels. Crop residue mulch reduces the velocity of runoff.

150 FAVOURABLE SOIL MOISTURE AND SOIL TEMPERATURE REGIMES Through their effects on soil structure, aggregation, porosity and pore size distribution, tillage methods influence the wettability, moisture retention characteristics, water transmission, the depth of the wetting or drying front, water extraction patterns, and movement of water and solutes through the soil (Ghuman and Lal, 1984). Conservation tillage has a moderating effect on soil temperatures and moisture regimes. It prevents extreme temperatures, and regulates the rate of evaporation. The tillage method also affects soil moisture through altering root distribution. Consequently, other factors being the same, plant available water reserves in a soil managed by conservation tillage are likely to be greater than in plough-till soil. This is especially true during the first and second stages of evaporation. The large moisture reserves of no-till and conservation tillage soils have been demonstrated by data from the tropics (Lal, 1982; Ojeniyi, 1986). Generally higher soil moisture reserves in no-till soil than in plough-till soil are illustrated in Figure 46. The high soil moisture content is due to improved soil structure and decreased evaporation due to the crop residue mulch. Improvements in soil structure take a long time and are relatively insignificant in coarse-textured soils. In such soils, improved soil moisture contents are mostly due to the use of crop residue mulch (Gupta and Gupta, 1986). Soil temperature is affected by the quantity of crop residue mulch and its properties (colour, durability, reflectance, composition, contact), and by soil properties. The latter include texture, clay mineralogy, bulk density, and moisture content, which influence the soil s thermal conductivity and heat capacity. Soil temperature is also affected by the proportion of surface area covered by crop residue mulch. In southwestern Nigeria, Lal (1976a) observed that the maximum soil temperature at 5 cm depth for a no-till system was 10 to 15 o C lower than under plough-till (Figure 47). Too high a soil temperature restricts yields in some crops. In arid regions of northwestern India and in Brazil, the use of crop residue mulches lowered the maximum soil temperature by 5-10 o C regardless of the degree of mechanical tillage (Gupta and Gupta, 1986; Derpsch et al., 1985) (Figure 48). Plastic sheets used as mulch influence soil temperature in a different way to straw. Transparent plastic mulch can act as a greenhouse at the soil-atmosphere interface. In northern India, for example, Tripathi and Katiyar (1984) record that paddy straw mulch lowered the maximum soil temperature by o C and raised the minimum by about 3 o C. In contrast, a polythese asphalt emulsion raised the maximum soil temperature by 4-8 o C. Similar results were obtained for Lixisols in sub-humid western Nigeria (Harrison-Murray and Lal, 1979; Maurya and Lal, 1981). Figure 49 shows the effects of clear plastic, black plastic, ridges, bare soil and straw mulch on maximum soil temperature. The diurnal fluctuations in soil temperature were as much as 20 o C. SOIL CHEMICAL AND NUTRITIONAL PROPERTIES AND FERTILIZER RESPONSE The effect of tillage method on soil chemical properties varies with soil, climate, crop rotation, and the period the tillage system has been in operation. Soil chemistry differs because of tillage-induced alterations in soil temperature and moisture regime, biotic activity of soil fauna, and the amount of crop residue returned to the soil surface. In general, the surface layer of a soil managed with conservation tillage contains more organic matter and has higher fertility status than a similar soil managed by plough-till. The chemical properties of the subsoil, however, may be more favourable in soil managed by plough-till than with conservation tillage.

151

152

153

154 Available data for tropical soils shows that continuous conservation tillage for 5 to 10 years increases organic matter content, cation exchange capacity, and basic cations held in the topsoil (Aina, 1979a; Lal, 1986a). This is supported by the data on a Lixisol from western Nigeria (Figure 50) (Aina, 1979a). In some soils receiving chemical fertilizers, topsoil ph may be lower under no-till than plough-till, but this has not been substantiated. Lal and De Vleerchauwer (1982) reported that favourable chemical properties of the surface layer of no-till soil in Nigeria may partly be due to a large proportion of earthworm casts. Machado (1976) reported more available phosphorus and higher levels of exchangeable cations in the 0-15 cm layer of no-till soil than of ploughed soil in Brazil. Modifications in physical, chemical and nutritional properties by tillage alter the soil s response to fertilizers. Differential fertilizer response may also be due to the mode of fertilizer application. Fertilizers are usually broadcast on the surface under no-till and conservation tillage systems; they are incorporated into the plough layer in the plough-till system. Fertilizer response is influenced by root growth, soil fauna and mulch. The latter influences fertilizer response directly and indirectly. Crop residue can contribute to or immobilize available plant nutrients. Crop residues with a low C:N ratio (from leguminous plant materials) provide readily available nitrogen, while those with a high C:N ratio (cereals) may immobilize soil nitrogen through microbial activity. Indirectly, crop residues influence nutrient availability by altering temperature and moisture regimes and influencing losses by drainage and surface runoff. The effect of residue mulch in preventing nutrient losses through runoff and erosion is relevant to improved fertilizer use efficiency on steep lands (Lal, 1976a). The fertilizer response depends on the initial soil nutrient status, other soil properties, drainage conditions, the quality and quantity of crop residue mulch, and the climate. The response to applied nitrogen may follow either of the two patterns shown in Figure 51 depending on internal drainage, the initial soil conditions, and the quality and quantity of crop residue mulch.

155 Nitrogen uptake is influenced by tillage methods. Under some circumstances, additional nitrogen is needed to produce yields equi-valent to those of plough-till. In Nigeria, for example, Kang et al. (1980) observe that yields of no-till maize were less than yields of ploughed maize with no or low rates of N application, but equal with greater rates. White et al. (1985) report that the slope of the curve relating wheat yield to N rate was maximal when the resi-due of the previous sorghum crop was incorporated into the soil. Sharma (1985) observes that pro-duction of irrigated forage in northern India was restricted by low availability of N in no-till plots. The nitrogen requirements of conservation tillage can be reduced by using management systems that include leguminous crops in rotation or in association with the main crops. The nitrogen requirement under conservation tillage eventually decreases in comparison with plough-till sys-tems because of reduced erosion losses and the equilibrium attained between the amount of nitrogen immobilized and released. On soils similar to those studied by Kang et al. (1980), for example, Lal (1982) observed equivalent or better responses by maize to nitrogen on no-till compared to plough-till soil after 10 years of these treatments. Response to P in relation to tillage methods depends more on soil chemical and mineralogical composition than on physical properties. For soils with a low capacity to fix P, tillage methods have little effect on P uptake. For Lixisols in western Nigeria, Kang and Yunusa (1977) observe that broadcast and hill methods of P application were equally effective in supplying adequate P to the maize crop at P application rates exceeding 20 kg/ha. Juo and Lal (1979) observed a satisfactory rate of P movement in Lixisols (Figure 52). For soils with a high capacity to fix P and with small plant-available reserves, incorporating fertilizer makes it more readily available than when surface broadcast. The response to P is, therefore, lower with a no-till system, in which the fertilizer is broadcast, than with a plough system, in which it is incorporated into the soil. Fertilizer use efficiency is also influenced by the size and development of the root system. Tillage method influences the root-depth distribution. Root growth in relation to tillage is influenced by factors that determine pore

156 size distri-bution, stability and continuity of pores, and soil moisture and temperature regimes. In general, conservation and reduced-till systems favour root growth in the surface layer immediately beneath a residue mulch. In the immediate sub-soil, the quantity (weight and number) or the total weight of root system is less in conser-vation tillage than in ploughbased systems. Under no-till there is a much higher percent-age of roots in the surface than in subsoil horizons. Some isolated roots grow actively in the deeper horizons of the conservation tillage soil in contrast with plough-till. Deep root penetration is facilitated by worm holes and biochannels. As a consequence of the differences in root system, water extraction patterns also differ. Generalized rooting depth distribution patterns for no-till and plough-till systems are shown in Figure 53. The pattern, however,

157

158 varies with the type of conservation tillage used (minimum-till, ridge-till), land-use history, initial soil conditions, and crop and soil management practices adopted. Maurya and Lal (1979a) observe that in an uncompacted Alfisol managed with manual farm operations, there was higher root density in the surface horizon of a no-till than a plough-till soil (Figure 54). A few roots penetrated deeper than 1 m through biochannels in the undisturbed soil of the no-till plot. In another study, Maurya and Lal (1979c) obtained slightly more maize grain yield from plough-till than no-till. The superior yield was attributed to deeper root system development in limed plough-till (Figure 55). Similar results have been reported for Nigeria by Osuji (1984). In Cameroon Ambassa-Kiki et al. (1984) observed that root distribution was restricted in the no-till soil. Restricted root growth in no-till soil was also observed in the West African Sahel by Chopart (1984). In easilycompacted, hard-setting sandy soils, plough-till had greater root weight between 35 and 80 days after seeding. ENERGY CONSERVATION Through its emphasis on reducing inputs, conservation tillage saves significant energy costs without jeopardizing productivity. Mechanical tillage and chemicals such as fertilizers and pesticides are energy intensive inputs. Conservation tillage saves energy by reducing the frequency and intensity of cultivations and decreasing fertilizer and irrigation needs. Currently, subsistence farmers in the tropics use little energy compared with North America and Europe (Table 143). A good reason for using conservation tillage is energy-saving.

159 PREVENTING SOIL DEGRADATION AND MAINTAINING SOIL FERTILITY Agriculture should aim to prevent soil degradation, maintaining soil s productive potential, and reducing environmental pollution. Non-point source pollution is a major environmental hazard of modern agriculture. Where the risk of soil degradation and desertification are severe, the question of crop yields is of secondary importance. In such circumstances, the important question is not whether conservation tillage works but how to make it work.

160 Chapter 8 Tillage systems and agricultural sustainability Agricultural sustainability implies increasing per caput productivity to meet present needs without jeopardizing future potential. The strategy is to use natural resources to enhance progressively the productive capacity and life support processes of an ecosystem without adverse effects on the environment. With this perspective, an important issue relevant to tropical ecosystems concerns achieving sustainability on the basis of minimum inputs. Input is a relative term. At present, tropical agriculture is mostly based on little or no commercial input. In contrast, agriculture in North America and Western Europe is heavily dependent on them. Ecologically, no-input agriculture can be as harmful to the economy and tropical environment as are excessive inputs and intensive agriculture. Whilst farmers in North America must take steps to reduce inputs, subsistence farmers in the tropics can achieve substantial yield improvements by marginal increases in added inputs. Sustainable agriculture aims to optimize the use of energy-related inputs. High crop yields are, however, possible if non-industrial inputs are increased. These include improved cultivars, new crops, efficient cropping systems, improved tools, increased fertilizer-use efficiency and systems of integrated pest management. Such inputs accord with the principles of good farming and land stewardship. Good farming, by this definition, is that which is ecologically, economically and culturally compatible. TILLAGE AND SUSTAINABILITY Tillage plays a major role in agricultural sustainability through its effects on soil processes, soil properties, and crop growth (Figure 57). Tillage may enhance or curtail these processes depending on initial conditions and the type of tillage tools used. Tillage affects agricultural sustainability through its long-term effects on the life-support processes of the soil, including soil structure, soil organic matter content, rate and capacity for supplying water and nutrients to crops, soil productivity and economic profitability. An important effect of tillage on sustainability is through its impact on the environment, for example on soil degradation, and pollution of surface and groundwaters. While conservation tillage can reduce soil erosion, it may increase risks of water pollution through increased use of pesticides, fertilizer and other agricultural chemicals. In contrast, plough-till systems may enhance risks of soil erosion, increase rates of mineralization of soil organic matter, and accentuate emission of greenhouse active gases from soil-related processes. There is now a greater need to attain agricultural sustainability than ever, especially in the fragile ecosystems and marginal lands of the topics and subtropics. These are found mainly on the semi-arid, arid and humid tropics, and

161

162 on steepland in all regions (Table 143). The productivity and land carrying capacity of such areas are low, but demographic pressures and demands on the limited resources are high. As a consequence, resources are used to the limit, and the risks of soil and environmental degradation are great. Issues of agricultural sustainability in these areas, especially those relevant to soil tillage, are shown in Table 144. High risks of soil degradation (due to erosion, compaction, deterioration of soil structure, and decline in soil organic matter content), low soil productivity, and low carrying capacity of land are widespread problems. The risks of soil and environmental degradation are accentuated by the resource-based and low-input agriculture widely practised. Adoption of proper tillage systems and techniques of soil management can help attain agricultural sustainability by reversing the degradative trends and restoring the productive capacity of the soils. Soil tillage and conservationeffective technologies influence sustainability at different levels (Table 145). Soil tillage influences sustainability at crop level through its effect on agronomic yield, at cropping system level by influencing productivity, and at farming system level by enhancing profitability. Specific tillage-based technologies for sustainable management of soil and water resources are outlined in Table 146. These are based on the principles of conserving soil and water resources, preventing or minimizing degradation of soil and environments, restoring degraded lands and reducing dependence on off-farm purchased inputs while enhancing productivity and increasing profitability. The overall objective is to increase yields, generate income and transform subsistence farming into commercial agriculture. Soil tillage is a basic part of agricultural production technology. In addition to preparing the desired seedbed, tillage is needed to manage crop residue, eradicate weeds and reduce pest incidence, mix fertilizer with the soil, improve aeration, alleviate compaction, and optimize soil temperature and moisture regimes. The exact nature of appropriate tillage operations depends on soil and crops (Lal, 1985b). Examples of suitable tillage methods for agricultural sustainability on small and medium sized farms in the tropics are listed in Tables 147 and 148. Specific components or sub-systems of tillage-based technologies in relation to sustainability are described in the section below. SYSTEMS APPROACH TO TILLAGE METHODS The effectiveness of conservation tillage on soil and water conservation and resource management is greatly enhanced by adopting a systems approach. There are strong interactions between various aspects of agricultural technology. A holistic approach views tillage as an integral part of a whole system. Conservation tillage requires a

163 special set of cultural practices that may be different than those needed for a plough-till system. Some crops and varieties are more suited to conservation tillage than others. The rate, time, mode and type of application of fertilizers and other amendments are likely to be different, as are measures for pest control. Conservation tillage also requires different seeding equipment and farm machinery to manage the uneven and trashy soil surface. Some crop rotations and farming systems are better suited to conservation tillage than others. Mulching is a key technique in conservation tillage, so cultural practices that ensure the availability of large quantities of residue mulch are most compatible. Vehicular traffic must be so used to reduce the risks of soil compaction. At harvest time method, time and type of equipment may affect soil compaction. Conservation tillage, therefore, is a package of cultural practices that are specifically developed and adopted to conserve soil and water resources, sustain high and satisfactory returns, minimize degradation of soil and environment, and preserve the soil resource. The interrelationship between conservation tillage and supportive cultural practices is shown in Figure 58. Some of the cultural practices specifically developed to enhance the effectiveness of conservation tillage are described below. Agroforestry and alley cropping Agroforestry is a technique of growing food crop annuals in association with woody perennials to optimize the use of natural resources, minimize the need for inputs from nonrenewable resources and reduce the risks of environmental

164 degradation. A range of woody plants and numerous agricultural crops are grown in one or another form of agroforestry in the tropics (King, 1968). The most common tree species used are Nauclea diderrichii, Lovoa trichilioides, Khaya ivorensis, Acacia, Erythrina, and Tectona grandis; woody perennials include Leucaena, Gliricidia, Flemingia and Caliandra. Alley cropping is a form of agroforestry in which annual food crops are grown between adjacent hedgerows of leguminous shrubs and woody perennials (Kang et al., 1981; 1985). The woody perennials are regularly pruned to limit shading and to provide nitrogen-rich mulch for the food crop annuals. Satisfactory crop yields are obtained provided that compatible species are chosen, and that the plant available reserves of soil-water are sufficient to meet the evapotranspiration needs of both species. The system is best suited to humid and sub-humid regions in which precipitation exceeds evapotranspiration during the growing season. At present, the system is labour-intensive and is suited more for resource poor farmers than for large-scale commercial farming. Maize-Leucaena alley cropping can be economic if hired labour is available at low cost (Guy et al., 1984).

165

166 Field experiments in the sub-humid tropics show that, when properly esta-blished, maize grown in association with contour hedges of Leucaena leucocephala and Gliricidia sepium produces satis-factory yields. Table 149 shows that despite the reduction in cropped area, maize grain yield with alley cropping was equivalent to that of no-till treatments. The yield of cowpeas, however, was sharply reduced by alley cropping. The reduction was due to poor stand esta-blishment and reduced germination. An allelopathic effect is a likely reason for the poor germination. In semi-arid and arid climates, suppressed growth and reduced yields of food crop annuals are caused by competition for soil moisture (Singh and Van Den Beldt, 1986; Nair, 1984). Contour hedges decrease runoff velocity and thus sediment transport. Sediments trapped by the contour hedges facilitate the formation of terraces. Experiments on relatively steep land in the Philippines show that contour hedges of Leucaena reduce sediment transport by several orders of magnitude (Loch, 1985; Pacardo and Montecillo, 1983). The effectiveness of contour hedges in trapping sediments has also been demonstrated in Indonesia (Sukman et al., 1985). Closely spaced narrow strips of shrubs or woody perennials are likely to be more effective in soil and water conservation than widely spaced single row hedges. There is, however, a need to determine the optimum spacing for erosion control and for satisfactory growth and yield of food crop annuals. The optimum spacing depends on slope, soil type and its susceptibility to erosion, rainfall, crop species, and the soil and crop management system. Loch (1985) observed that erosion control by contour hedges depends on the sediment carrying capacity of the water runoff. Hedges trap sediments as long as the sediment transport capacity of the overland flow is not attained. Table 150 shows that contour hedges of Leucaena 2 m apart are more effective in reducing runoff and erosion than hedges 4 m apart. In comparison with plough-till, hedges of Leucaena and Gliricidia also reduced losses of cations and plant nutrients. Figure 59 shows that growing contour hedges of perennial shrubs markedly increased accumulative infiltration, which was 83, 82, 70, 55, 54 and 40 cm per 2 hour period for Gliricidia at 4 m spacings, Leucaena at 2 m spacings, Leucaena at 4 m spacings, ploughed soil, Gliricidia at 2 m spacings, and no-till treatments, respectively. There were also notable differences in runoff hydrographs between methods of seedbed preparation and hedgerow spacing treatments (Figure 60). Despite their established advantages, more specific local research is needed to develop alley cropping and agroforestry systems for particular soils, crops and ecological environments. Research is needed on the choice of appropriate crop and tree species, on suitable spacing, management of trees, and soil and crop management practices for the food crops. There is a need to determine the proportion of land allocated to trees so that ecological stability is maintained enabling intensive use of the remaining land as arable. The nutrient recycling effects of growing deep-rooted perennials need to be evaluated. Aspects of alley cropping tillage need fur-ther research. What

167 are the most appropriate tillage methods to maximize the nutrient value of the prunings? How can tillage tools be used to minimize competition for moisture from the trees and woody perennials? Which methods of seedbed preparation are best suited for grain crops and also optimize stand establishment of trees and woody perennials? How can possible allelopathic effects be minimized? In addition to these questions, there are also soil and crop-specific issues that need to be addressed. Cover crops Diversifying a cropping system often increases its stability and reduces the incidence of diseases and pests. Growing grass or leguminous cover crops at frequent intervals (once every 2-3 years in the tropics) is necessary for a suc-

168 cessful system. Cover crops have many advantages for conservation tillage systems. They restore fertility, control weeds, avoid repeated traffic, conserve rainwater and reduce energy costs. They also help to control pests, improve soil physical properties including soil tilth, and reduce soil erosion. Cover crops have long been used in the tropics for soil and water conservation in plantation crops. The importance of cover crops for the management of some uplands in Ghana, West Africa, was demonstrated by Kannegieter (1967a). Some grasses and legumes can produce large quantities of biomass even under average conditions (Table 151). The biomass produced is useful as forage, mulch, and for domestic uses. Kannegieter (1967b; 1969) developed a technique combining short-term fallowing with Pueraria and zero cultivation to reduce the fertilizer nitrogen needs of the following maize crop and to counteract soil erosion. As well as augmenting soil fertility, cover crops also improve soil structure and increase macroporosity. In northern Nigeria, Wilkinson (1975) observed significant benefits of grass fallow rotations on the infiltration of water into the savannah soil. The infiltration rate increased with increasing length of the fallow period. Similar observations are described for soils of western Nigeria (Lal et al., 1979; Wilson et al., 1982). Mucuna utilis is now widely recommended as a cover crop in western Nigeria. The benefits of cover crops on soil structure and tilth have also

169 been demonstrated for East Africa (Pereira et al., 1954; 1958; Pereira, 1956; Wallis, 1960; Peers, 1962; Stephens, 1967). Cover crops have proved useful for erosion control in soils of tropical America (Kemper and Derpsch, 1981). In Parana, Brazil, Sidiras et al. (1985b) observed that cover crops (such as Avena strigosa, Raphanus sativus, and Lupinus albus) grown during winter have beneficial effects on the yield of the following summer crops of beans, soybeans and maize and control erosion. The choice of a cover crop for different soils and ecological regions depends on several considerations, including: easy and economical establishment, including availability of seed; quick ground cover and growth rate during the off season; N-fixing rather than N-consuming properties; a deep root system and consumptive water use; its feed value for livestock; its role as a host for pests and cover for wildlife; canopy height; ability to suppress weeds; growth duration (i.e., permanent or annual); shade tolerance, and ease of management for growing a food crop with conservation tillage. There remains considerable scope for selecting appropriate species and cultivars for cover crops.

170 Live mulch It can be expensive and energy-intensive to suppress a cover crop so the concept of live mulch or a green seedbed was developed in the early 1940s (Spivack, 1942, 1984). A live mulch system is a form of mixed cropping. A fast growing perennial legume is established to smother weeds and a seasonal grain crop is grown through it in such a way that growth and yield of the food crop is not unduly depressed. A small strip is opened, with or without herbicides. The system works if the live mulch is low-growing and does not compete for light, moisture or nutrients. Live mulch systems are most likely to succeed in humid and sub-humid regions with little or no water deficiency. The concept has been tried in West Africa with modest success (Voelkner, 1979; Ogborn, 1980; Akobundu, 1980; Wilson et al., 1982). Akobundu (1980a) reported satisfactory yields of maize using a live mulch system of Arachis, Centrosema and Psophocarpus. Yields of food crops can be severely depressed, due to allelepathic effects, smothering and competition for moisture during periods of drought. Tillage systems should be developed to manage live mulch so that competition with the crop is minimized, particularly in the row zone. Rotations and multiple cropping Crop rotations are an integral component of successful conservation tillage systems. Their benefits are widely recognized. Ideal rotations involve one or more of the following: sequential cropping of cereals followed by legumes; following shallow-rooted crops by deep-rooted crops; fertility-depleting by fertility-conserving crops; soil-degrading by soil-regenerating crops, and crops demanding heavy inputs by those that can survive on low inputs. Mixed and multiple cropping are usual in the tropics. The most common rotations in West Africa are maize-cowpea, milletcowpea, sorghum-cowpea, and sorghum-yam as sequential crops, and cassava with maize, cassava with cowpea, and maize with yams as mixed crops (Okigbo, 1978). For tropical Lixisols, Lal (1976a, b) observed that maize-cowpea and maize-soybean rotations were compatible with no-till and conservation tillage systems. Aina et al. (1979) reported significant reductions in water runoff and soil erosion from mixed maize-cassava than either maize or cassava monocultures. Mixed cropping has been shown to maximize water use efficiency in Australia (Hulugalle and Willatt, 1985) and Nigeria (Hulugalle and Lal, 1986). Intensive monocropping encourages soil erosion (Table 152). Fallowing, with spontaneous or planted cover crops, improves the soil s physical condition and reduces the risks of soil erosion. Tillage methods for mixed cropping systems, particularly those that limit the use of herbicides, can differ from monocultures.

171 Summer fallowing Bare fallow is commonly practised in arid and semi-arid regions to improve soil-water reserves for the succeeding crop. All plant growth is killed by cultivation or chemicals and production for one season is forfeited. There is often at least partial compensation by increased production next season (Haas et al., 1974). It is difficult to trace the origin of this practice, because it is used throughout the world in regions with low and erratic rainfall and marginal soil conditions. On a Luvisol in Western Australia, Hamblin (1984) compared soil properties and crop performance in plots cropped every year with those cropped every other year. Soil properties and yield deteriorated progressively in the continuously-cropped treatments but not in those fallowed. In arid regions of western India, fallowing conserves more water in the soil profile than ploughing (Table 153). In Botswana, Whiteman (1975) reported significant improvements in the grain yield of sorghum by fallowing. Tables 154 and 155 show significant differences in grain yield

172 where sorghum was seeded in fallow compared with land previously growing a cover crop, maize or weeds. The beneficial effects of fallowing were greater in dry years than in those with normal rainfall, and greater where weeds were controlled than on plots left weed infested. Weed control during fallowing improves water conservation (Min. of Agric., Botswana, 1977). Experiments in semi-arid central Tanzania showed that weed-free bare fallow conserved more water and increased yields of groundnuts, compared with continuously cropped land (Pereira et al., 1958), but severe soil erosion occurred on unprotected bare ground. Protection of the fallow by sowing shallow rooting teff grass (Eragrostis abyssinica) provided efficient soil conservation, controlled weeds, and enabled sufficient subsoil-water to be stored to produce a satisfactory crop of groundnuts. Similar observations were made in Senegal, West Africa, by Charreau (1970) (Table 156). Experiments conducted by ICARDA in West Asia and northern Africa (Tables 157, 158, 159) show the benefits of planted fallows on biological yield and on water use efficiency. Table 160 shows substantial increases in yield by incorporation of vetch in crop rotation. Water use efficiency was also substantially increased (Tables 157 and 158). The maximum depth of soil wetting and total recharge were substantially improved (Table 159). The inclusion of vetch in rotation also improved nitrogen status of the soil and crop yield. Considering the needs to provide residue mulch at 2-4 t/ha and to enhance the nitrogen status of the soil, it is logical that a cover crop be sown during the fallow. It must, however, be managed to conserve soil, water and nutrient reserves; tillage can play a key role in this.

173 SUSTAINABILITY AND AGRICULTURAL PRODUCTIVITY To assess and quantify sustainability some kind of an index is needed. It must have a time dimension because sustainability can be defined as an increasing trend in production overtime per unit consumption of the non-renewable or the limiting resource, or per unit degradation of soil and environmental characteristics (Lal, 1991a). An index of sustainability based on this definition was proposed by Lal (1991): I s = f(o i, O d, O m ) t where O i is the output per unit input that maximizes the per caput productivity or profit, O d is the output per unit decline in the most limiting or non-renewable resource, O m is the minimum assured output, t is the time. Although

174 soil-specific, the index based on the parameters outlined above needs to be developed. Relevant examples of trends in productivity and soil quality for different tillage systems are described below. Long-term experiments of this kind are needed to establish similar trends for other soils, crops and ecoregions. Agricultural productivity Figure 61 shows maize grain yield (t/ha/yr for 2 crops a year) for two tillage methods. Yield declined in both. The rate of decline was 136 kg/ha/yr in no-till compared with 170 kg/ha/yr for the plough-till system. The declines could have been reversed or minimized if a legume or other crop had been growth in rotation. Figure 62 shows again that maize grain yield was maintained at a higher level with mulch-based no-till compared with plough-till. Nonetheless, yield declined in both. The rate of decline, however, was 0.59 t/ha/yr for no-till compared with 0.71 t/ha/yr for the plough-till treatment.

175 Yield trends in a plough-based alley cropping system are shown in Figure 64. Yields of maize and cowpea declined over the 6-year period even where prunings from woody perennials and supplementary chemical fertilizers were applied. The rate of decline was 340 kg/ha/yr for maize and 96 kg/ha/yr for cowpea. Methods of land clearing interact strongly with tillage systems to affect agricultural productivity. Figure 64 from a watershed experiment conducted at IITA (Lal, 1992) shows that maize grain yield declined with time under all treatments, irrespective of the management system. The sharp decline in the shearblade treatment from 6th year

176 onward is probably due to severe soil compaction. The yields were generally greater in those no-till systems with Mucuna included once in the 3 year rotation. Experimental data from the tropics that show increasing productivity with time are very scanty. Maybe this is difficult to achieve. It is exactly what is needed, however, if per caput productivity is to be increased or enhanced from shrinking resources. Reversing the declining trends in productivity on tropical soils could be the greatest challenge facing agriculturists during the 21st century. Soil properties To evaluate the relationship between agricultural productivity and soil properties, it is necessary to establish what changes in soil properties take place under different tillage and land-use systems. Reliable data are needed from well-designed long-term experiments to establish these. Lal (1992) shows that soil organic carbon content declined in parallel with maize grain yields in the IITA watershed experiment. The decrease in soil organic carbon, and increase in soil compaction with continuous intensive cultivation observed in these experiments reduced infiltration rate and soil-water sorptivity (Figure 65). Measurement of infiltration characteristics is a useful way to quantify the impact of management systems on structural properties.

177 Structural degradation of Vertisols in Australia under continuous cultivation was studied by Cook et al. (1992). Continuous cultivation increased the soil bulk density, and decreased the size and strength of structural aggregates. Organic carbon contents also decreased with cultivation, and dispersibility was increased, probably because of an increase in total and exchangeable Na. Increased dispersibility increased the risk of soil splash and erosion. These kinds of data from Vertisols in Australia are exactly those needed for other tropical soils. Regrettably, only few long-term experiments are maintained in the tropics to provide such information. Work on benchmark soils is needed to assess the sustainability of various soil management systems.

178

179

180 Chapter 9 Soil properties as a guide to tillage The choice of the most appropriate type of tillage depends on many factors. Physical factors include soil properties, rainfall regime, climate, drainage conditions, rooting depth, soil compaction and erosion hazards, and cropping systems. Socio-economic factors include farm size, availability of inputs, and marketing and credit facilities. The socioeconomic factors are often the most important. Impressive progress has been made in measuring soil nutrient status using laboratory and field tests and in recommending optimum fertilizer applications. Soil scientists have developed reliable tests for soil acidity, aluminum toxicity and nutrient deficiencies, and for prescribing corrective measures. There are, however, no reliable routine tests to determine the type of tillage needed for sustainable use of soil and water resources. Tillage operations are energy-intensive, a major cost, and have long-term effects on soil productivity and environmental quality. Their study therefore deserves high priority. Attempts have been made to develop and standardize soil evaluation techniques in relation to tillage needs. The most important factors to be considered are drainage, erosion, rooting depth, soil temperature regime and compactability. SOIL CHARACTERIZATION IN RELATION TO TILLAGE REQUIREMENTS Research managers and planners need to recognize the role of soil physical characteristics in sustaining the productivity of tropical soils. There has been no standard methodology to evaluate soil physical properties. This has limited objective, comparative evaluation of the available data. The methodology adopted should be uniform, so that results obtained from one soil or region can be compared with others. Progress has been made in standardizing methodology for evaluation of soil physical properties (ASA, 1986; Smith and Mullins, 1991), but most methods developed are for soils of temperate regions. These are not applicable for some tropical soils because of different soil characteristics, including hard-setting, presence of plinthite, high sesquioxide contents, presence of concretionary materials, cracking and swell-shrink properties. Table 160 suggests possible standard methods of soil physical analyses for use in relation to tillage methods. Relevant plant and soil parameters, that should be monitored during assessment of tillage effects, are listed in Table 161. Important among soil characteristics are crusting, soil roughness, cracking patterns and aggregate strength. These are not routinely measured properties.

181

182 SOIL STRUCTURE ASSESSMENT Although soil structure has been extensively studied, it is still the least understood of physical characteristics affecting crop production, sustainability, environmental quality and tillage requirements. Soil structure is so complex that it is difficult to study directly. It is always evaluated by indirect measurement of related parameters. Soil structure has been defined as the size, shape, arrangement and packing of soil particles into identifiable units. This definition seems to ignore the key aspect of soil structure the voids or pores. Soil pores are essential for the transmission, diffusion and retention of water, gases and plant nutrients. They also provide pathways for root growth and development. Soil structure should take voids into account so it can be defined as the size, shape and arrangement of solids and voids, and forces that affect those characteristics. Pores facilitate plant growth; they are most useful when they are continuous and stable. Soil structure may therefore be defined as the size, shape, arrangement of soil particles and continuity of pores and voids; their capacity to retain and transmit fluids and organic and inorganic substances, and ability to support vigorous root growth and development. This is a working definition of soil structure in relation to crop production. There is a need to distinguish pores from voids. Functionally, they are identical as long as they have similar size, shape, tortuosity and continuity. A major difference perhaps lies in their mode of origin. In the present context, pores are channels created by the activity of soil flora and fauna. Voids, on the other hand, are created by nonbiotic factors, for example, cracks formed by swelling and shrinking, wetting and drying or freezing and thawing, or inter-aggregate pores between domains or micro-aggregates formed by inter-particle and inter-domain forces of cohesion and adhesion. There is no single method of measurement that satisfies all the aims of structural assessment. More often than not, assessment of soil structure is subjective, qualitative, visual and arbitrary, indirect, fragmentary and piecemeal. On the whole the procedure is confusing and frustrating. There are several methods, and the suitability or desirability of a method depends on the objectives. To pedologists, assessment of soil structure involves qualitative description of shape, size and packing arrangement of structural units observed in the field. To civil engineers