Discussion papers

1. Causes and consequences of soil moisture scarcity

Low and erratic rainfall is a major constraint on rainfed agriculture, particularly in seasonally dry and semi-arid tropics and subtropics (henceforward, dryland regions or "dryland"). In these areas, soil moisture for plant growth is insufficient for the fullest expression of the production potential of plants over time. Thus, soil moisture deserves to be treated as a valuable resource. It has been estimated that low levels of and uncertainty in rainfall (in combination with few choices for crop and livestock production, poor yields and continuing natural resource degradation) limit agricultural production in about three-quarters of the world's cultivated lands.

Drylands may have low crop yields not only because rainfall is irregular or insufficient, but because up to 40 percent of the rainfall may disappear as runoff. This poor utilization of rainfall is partly the result of natural phenomena (relief, slope, and rainfall intensity), but it is also caused by inadequate land management practices (burning of crop residues, excessive tillage, eliminating hedges, etc.) that destroy soil structure, reduce organic matter levels, eliminate beneficial soil fauna, and do not favour water infiltration. However, that water "lost" as runoff for one farmer is not lost for other water users downstream as it is used for recharging groundwater and river flows.

A key challenge is how to manage limited rainfall so that avoidable surface runoff (representing lost potential soil moisture and groundwater) does not occur. Crops depend not only on precipitation but also on the ability of the soil to absorb and store water. Inadequate agricultural practices that reduce this ability may add their negative effects to the natural main causes of soil moisture scarcity, and then add water stress in crops.

Once rainfall reaches the land surface, it can infiltrate into the soil, run off over the surface as overland flow, or accumulate on plant leaves or in puddles from where it evaporates back to the atmosphere. For optimal capture, infiltration, storage and use of soil moisture, three physical capacities of the soil are important:

• the capacity to allow water to enter, referred to as infiltration;
• the capacity to allow water to move readily through the profile, referred to as permeability;

• the capacity to store the acquired moisture in the rootzone and release it to plant roots, referred to as soil waterholding capacity.

Where any of these characteristics does not function well, potential soil moisture is lost as runoff. An increase in sloping terrain will further aggravate sealing and crusting of topsoil layers, the presence of compacted soil layers and plough pans.

Impact by high-energy raindrops on bare soil causes breakdown of soil aggregates. These soil particles block the surface pores, resulting in a slowing or halting of infiltration through the soil surface. This results in runoff over the surface when the rate of rainfall arrival exceeds the rate of infiltration. This sealing occurs because the original pore spaces in the top 1 mm of the surface have collapsed to very much smaller voids. These become saturated with water almost immediately, through which the incident rainwater can move only much more slowly than before, where it can move at all. Thus, the soil surface partitions rainfall between infiltration and runoff. A further partitioning occurs through dense leafy plant canopies by their interception of rainfall. Interception can divert up to 30 percent of rainfall from the soil surface. The greater is the proportion of a rainfall event that is lost as runoff, the smaller is the proportion of that rainfall that can become soil moisture and groundwater.

Plate 1
The components of soil structure. Soil minerals with organic matter form soil structure units, called "peds". Micropores inside the peds and macropores between the peds carry air and water and facilitate root penetration

USER MANUAL, 1992

Several important soil factors affect soil water movement and management. They include: soil texture, soil structure or aggregation, organic matter, ground cover, and soil tillage methods. Texture refers to the proportions of sand, silt and clay present in a given soil. For example, a sandy loam has much more sand and much less clay than does a clay loam. A loam soil is a more balanced blend of sand, silt and clay. Most soils are some type of loam. Texture is an innate characteristic of the soil type. Unlike aggregation, organic matter, and ground cover, texture cannot be changed through agronomic practice. However, by knowing the innate texture of the soil, the farmer can select and adjust practices that optimize soil moisture management.

Soil aggregation or soil structure (Plate 1) refers to how the sand, silt and clay come together to form larger granules. Good aggregation is apparent in a crumbly soil with water-stable granules that do not disintegrate easily. Well-aggregated soil has greater water entry at the surface, better aeration, and more waterholding capacity than poorly aggregated soil. A stable system of soil pores allows easy exchange of air and water. Plant roots occupy a larger volume of well-aggregated soil; better rooting increases the depth and area plants can reach for water and nutrients.

The texture and aggregation of a soil determine air and water circulation, erosion resistance, looseness, ease of tillage, and root penetration. However, while texture is an innate property of the native soil that cannot change with agricultural activities, the soil aggregation could be improved or destroyed readily through the choice and timing of farm practices.

Some practices that destroy or degrade soil aggregates are:

• tillage in general;
• tilling when the soil is too wet or too dry;
• using anhydrous ammonia, which accelerates the decomposition of organic matter;
• excessive N fertilization;

• excessive sodium (Na) buildup from salty irrigation water or Na-containing fertilizers.

Aggregation is closely associated with biological activity and the level of organic matter in the soil. Organic matter acts as the "glue" to hold the framework of soil particles and pores together, and can build a stronger internal and superficial structure in the soil profile to a condition allowing easy entry of water and its storage in plant-available form. Organic matter in the form of mulch and leaf litter can also be a significant protection against surface sealing by raindrops.

Another effect of organic matter (and microbial activity) is an increase in the macrofauna population (especially earthworms). One of the consequences of an increased earthworm population is the formation of channels and pores. Shallow-dwelling earthworms create numerous channels throughout the topsoil, which increases overall porosity (Plate 2).

Plate 2
Binary images of: (a) conventional and (b) zero-till soil profiles (depth 24 cm) showing the dramatic increase in earthworm macropores with zero till. In images, black is pore, and white is solid. Plough pans occur in conventional plot as friable topsoil and worm holes to depth occur in the zero-till plot

MCGARRY, BRIDGE AND RADFORD, 2000

The large vertical channels created by the deep-burrowing earthworms and the channels left in the soil by decayed plant roots increase water infiltration under very intense rainfall conditions and percolation to deeper soil layers.

The waterholding capacity of a soil in a particular place depends on: the depth of the soil, the volume of pore spaces, and the proportion of the voids that retain water against the pull of gravity. The smaller the particles, the greater will be the surface tension effects in holding the water in place. Water in soil pores larger than 0.05 mm in diameter drain down in response to gravity. In a sandy soil, there is usually a relatively large total volume of pore space among the large mineral particles. However, most of the pores are so large that rainwater drains through most of them and relatively little is retained within the profile. In clayey soils, the opposite can be expected. There may be a large proportion of the pore spaces so small that the water cannot drain out and can only be removed by plant roots and/or by slow evaporation into any air-filled spaces within the soil. A good comparison is with a sponge that is completely dry. When soaked in water, the sponge will absorb water and all spaces will be filled with water. The phase is called "saturation". All available pores of the sponge are filled with water. When taken from the bucket of water, part of the water will drip out in response to gravity. When the sponge stops dripping, it contains an amount that relates to what in soils is termed "field capacity" or the amount of water that can be held against gravity. Part of the water that will not drain automatically can be removed by wringing the sponge (to be compared with PAW). When all water is removed by wringing, the sponge still feels wet. This phase compares with the "permanent wilting point" in soil. This point is reached when plants are no longer capable of removing water for daily use, and irreversible wilting occurs. The little remaining soil water is bound to the soil particles and aggregates so tightly that root extraction by the plant cannot overcome the force by which this water is bound. The amount of water that can be held between "field capacity" and "permanent wilting point" is called "plant-available water" (PAW). The remainder of water after completely wringing the water out of the sponge is retained in the finest pores of the sponge and can be removed by placing the sponge in the sun to dry. Similar water extraction beyond permanent wilting point can be achieved in soil by spreading it out in the sun or drying it in an oven.

Where the pores are too small, as when the soil has become compacted, or pulverized or otherwise collapsed, or they were never there in the first place, much of the soil moisture may be held at tensions that make it less than 100-percent plant available.

The larger the volume of exploration by roots, the higher the moisture reserves to which the plants have access and the larger the "buffer" to allow plants to survive rainless periods. Where soils are very shallow or physically degraded, the ability of plants to withstand rainless periods will be reduced. Some crops are able to deepen their rooting system in search of more water. However, as shown in Plate 3, not all crops have the same ability. This is one of the reasons why rotation of crops from different species is important.

Plate 3
Where no physical or chemical obstacles are present, different crops do not have the same potential for exploring deep soil water (and nutrient) resources. For example, because of its extended root system, pearl millet is particularly adapted to severe environmental conditions

CHOPART, 1999

Compacted soils do not provide adequate space for the storage or movement of air and water. Soil animals and root growth are also restricted. Most importantly, large, continuous soil pores are lost or are reduced in size, leading to poor water infiltration, slow drainage and reduced aeration for healthy root growth and nutrient uptake for maximum crop yield.

Water is also able to move through soils as water vapour, the most important example of this being the loss of water vapour by evaporation from soil surfaces. This occurs when the concentration of water vapour in the soil close to the surface is higher than that in the atmosphere immediately above. Water vapour will then move from the soil into the atmosphere in an attempt to equalize concentrations. The drier and hotter the atmosphere compared with the surface soil, the greater will be the rate of evaporation from the soil, provided sufficient water can be supplied to the surface by capillary movement from below (Box 1). Fine-textured soils have an abundance of small pores. Thus, more capillary movement of water to the surface will generally occur in fine-textured than in coarse-textured soils.

One of the most important ways to reduce water evaporation from the soil surface is to keep the soil covered with a mulch or cover crop. Surface evaporation can be up to five times less under surface mulch compared with bare soil. Because less water is lost to evaporation, more water will be available for plants. Another way to reduce evaporation is to allow water to penetrate as deep into the profile as feasible.

Understanding the process of cracking, and how water moves into cracks, is an important issue for managing soils that have substantial cracking. In soils that contain montmorillonitic clay, cracks often develop to depths of 30-60 cm or even deeper. If precipitation occurs while these cracks are open, some of the water will move quickly to the bottom of the cracks. However, unless the crack is open to the surface, water will not move into the crack regardless of how much precipitation is received.

For example, if a soil that is cracked to 30 cm is tilled to a depth of 10 cm, then any precipitation that occurs will be retained largely in the upper 10 cm unless the amount of precipitation is very large, and even then it will move downward only by gravitational and capillary forces. On the other hand, if the land is not prepared, some of the water will move quickly to the bottom of the cracks where it will be much less. Organic material falls into the cracks, along with granular soil material from the surface. When the soil is wetted and swelling occurs, these materials are incorporated into the soil at depth. This explains why these soils do not have distinct horizons in their profiles.

In dryland regions, soil dries quickly and it dries to the depth that it has been tilled. Consequently, all PAW is lost. For example, a silty clay loam might have a volumetric water content of 40 percent at field capacity and 20 percent at permanent wilting point. Thus, the top 15 cm of the soil would hold 6 cm of water when the soil was at field capacity, and would still contain 3 cm of water at permanent wilting point. However, if the soil is tilled and air dried, then essentially all of the 3 cm of soil water still contained in the soil after the plant could no longer extract any water from the soil would be lost. When precipitation occurs, this 3 cm of water must be made up before the water content is sufficiently high enough for plants to begin extracting water. This factor is also the reason why "droughts" are so devastating in dryland regions. Prolonged droughts result in plants using all the water that they can extract, but then additional water is lost by evaporation. The longer the drought, the deeper the soil dries out below the permanent wilting point. When it begins to rain, all the water lost from below the permanent wilting point must be replenished. This is why long drought periods are so difficult to recover from, and it usually takes a prolonged wet period to really remedy a long dry period. Some drying processes of soils may not be reversible, e.g. Andosols that may lose their initial physical properties completely.

The perceived water scarcity may, besides low or erratic rainfall, be caused by choices made by the farmer, e.g. of a crop or variety sensitive to water stress, to inadequate management of available water from rainfall, or to increasing potential evapotranspiration as a result of local or global modification in atmosphere circulation. Practices such as clear-felling of trees, grazing on very steep slopes and the compacting effects of farm machinery always result in excess water runoff and erosion. Replacing forest vegetation with grassland or annual crops may increase deep drainage and so provide higher base flows in streams and rivers, and modify the global and local water balance. Changes in soil management can also affect the quantity of deep drainage replenishing groundwater. The introduction of poor management practices that increase the proportion of rainfall lost as runoff will reduce base flows and increase peak flows, i.e. the incidence of flooding. Flood flows in streams and rivers, which rise quickly after heavy rainfall, derive mostly from rapid overland flow of water. Flood flows are often muddy with eroded materials. Clear water stream-flow originates from rainwater, which has infiltrated the soil and percolated through pores of a range of sizes at different slower speeds that act as a filter not only for solids but also for solutes and pollutants.

Soils with suboptimal moisture conditions are less productive, and substantially reduce yields and profits. This is particularly true in arid and semi-arid regions where optimal soil structure conditions are required to maximize both water entry into the soil and water storage during fallows. In dry seasons, root growth must also be optimal and fine structure is required to enhance root proliferation to fully tap soil water reserves in order to enhance grain or fibre production through to harvest.

Source: Maraux and Rapidel (1990).

The effects of drought on crop production are aggravated because farmers do not have reliable methods of predicting occasional droughts (owing to seasonal or interannual variations in rainfall) or appropriate means to cope with long-term droughts. Rainfall predictions can help farmers to anticipate their actions (Box 2). Short-term weather forecasts (1-5 days) or seasonal predictions (1-3 months) based on statistical tools may be available. The former are often reliable, but not very useful for agriculture. The latter would be very useful, but are generally less reliable.

Inappropriate practices in a particular context can dramatically affect water resources management and soil moisture availability. No short-term action can improve the rainfall patterns and, therefore, major efforts should focus on improving agricultural practices and land use to make better use of the rainfall. This may include: improved rainwater capture and infiltration; the reduction of water losses through evaporation, runoff and deep percolation; water-use-efficiency improvement by choice of water-efficient crops or varieties and spatial and temporal arrangements (intercropping and rotations); and reduction of transpiration through weed control or windbreaks.

Water shortages often affect whether or not there is a response to fertilizers, and how much fertilizer should be applied. This is particularly common with N fertilizer, where the optimal response is frequently higher in good seasons than in poor seasons. This creates difficulties in that, as it is not possible to predict the distribution and amount of rainfall reliably, farmers cannot know how much fertilizer to apply, eventually resulting in low fertilizer efficiency.

One of the challenges a farmer faces in making the soil more drought resistant is how to allow more of the incidental rainfall into the soil in such situations so as to diminish the speed of development of water-stress in the plants, as well as the duration of each episode and the frequency of moisture-stress periods (short term).

A second challenge is to retain, develop or restore a condition of porous soil architecture capable of retaining and releasing a high proportion of infiltrated water to plant roots (long term).

REFERENCES

Chopart, J.L. 1999. Relations entre état physique du sol, systèmes racinaires et fonctionnement hydrique du peuplement végétal: outils d'analyse in situ et exemples d'étude en milieu tropical a risque climatique élevé. Grenoble, France, Université Joseph Fourier. (doctoral thesis)

Maraux, F. & Rapidel, B. 1990. La simulación del Balance Hídrico. Aplicación para la determinación de fechas de siembra. Collections CATIE. 30 pp.

McGarry, D., Bridge, B.J. & Radford B.J. 2000. Contrasting soil physical properties after zero and traditional tillage of an alluvial soil in the semi-arid tropics. Soil Till. Res., 53: 105-115.

USER manual. 1992. Understanding soil ecosystem relationships. Script and 2 videos. Australia, Department of Primary Industries.

2. Creating drought-resistant soil -technologies and impacts of improved soil moisture management at field level

The sustainable use of natural resources, especially in the case of water and soil, depends greatly on control of and respect for natural processes and cycles.

When rain lands on the soil surface, a fraction will infiltrate into the soil to replenish the soil water or flow through to recharge the groundwater. Another fraction may runoff as overland flow, and the remaining fraction will evaporate directly back into the atmosphere from unprotected soil surfaces and from plant leaves.

The above-mentioned processes do not occur at the same moment, but some are instantaneous (runoff), taking place during a rainfall event, while others are continuous (evaporation and transpiration).

In order to create a drought-resistant soil, it is necessary to understand the main factors that affect soil moisture. This paper examines practices that threaten or that protect and improve this characteristic, the impacts (economic, environmental, etc.) of their implementation (locally and downstream), and the requirements of the different improvements in terms of labour, investment, support, etc.

With severe drought an all-too-common occurrence, some farmers turn to irrigation for a solution. However, irrigation may not be feasible or even desirable. In this case, there are management options that can increase the ability of the soil to store water for plant use. Soil can be managed in ways that reduce the need for supplemental watering and increase the sustainability of the farm. Any worthwhile strategy for drought management optimizes the following factors:

• capture of a high percentage of rainfall (infiltration);
• maximum storage of water in the soil for later use (waterholding capacity);

• efficient recovery of stored water (plant rooting).

In places where conventional agriculture is characterized by the burning of residues and intensive tillage for seedbed preparation and weed control, it has contributed to land degradation and contamination of surface waters. Most important, residue burning and soil tillage result in the loss of valuable soil moisture, while compaction of soil layers through the use of machinery and equipment prohibit the entrance and percolation of rainwater.

The burning of crop residues and natural vegetation in the field is a common practice. Residues are usually burned to help control insects or diseases or to make fieldwork easier in the following season. Burning destroys the litter layer and, thus, diminishes the amount of organic matter returned to the soil. Conservation of a cover will lead to more soil moisture being stored in the profile.

Tillage leads to pulverization of soil, and small particles can easily be washed away by runoff during rain showers. These usually very fine clay or organic particles can easily block micropores at the soil surface. Thus, they form a very thin film-like layer, also referred to as surface sealing. This continuous impermeable layer on the surface prevents rainwater from infiltrating and facilitates runoff.

The less the soil is covered with vegetation, mulches, crop residues, etc., the more it is exposed to the impact of raindrops. When a raindrop hits bare soil, the energy of the velocity detaches individual soil particles from soil clods. These particles can clog surface pores and form many thin rather impermeable layers of sediment at the surface, referred to as surface crusts. They can range in thickness from a few millimetres to more than 1 cm, and they are usually made up of sandy or silty particles. These surface crusts hinder the passage of rainwater into the profile. The breakdown of soil aggregates into smaller particles depends on the stability of the aggregates, which largely depends on the organic matter content.

Although they are meant to generate the opposite, the use of machinery and implements and even the trampling of animals can destroy or reduce greatly the sizes of soil pores. Compacted soil does not provide adequate space for the storage or movement of soil air and water. Soil animals and root growth are also restricted. Most importantly, large continuous soil pores are lost or are reduced in size, leading to reduced water infiltration rates, slower drainage and reduced aeration for healthy root growth and nutrient uptake for maximum crop yield.

Drainage of water beyond the rootzone of a crop may reach the groundwater and help to maintain the level of water in wells, streams and rivers. However, the water "lost" by drainage could have been used for crop production. Deep drainage occurs where rainfall exceeds the amount of water that is needed to bring the rootzone to field capacity. Sometimes, deep cracks in clay soils at the surface during the dry season can cause deep drainage. In addition, drainage beyond the rootzone can be favoured by biopores. These are continuous pores of a diameter of 0.5 mm and wider formed by earthworms, ants and termites, and they extend from the soil surface to the subsoil. The amount of water lost through deep drainage is higher in coarse-textured soils than fine-textured soils.

The capacity of soil to retain and release water depends on a broad range of factors, such as soil texture, soil depth, soil architecture (physical structure including pores), organic matter content, and biological activity. However, it can be improved through appropriate soil management.

Practices that increase soil moisture content can be categorized in three groups: (i) those for increasing water infiltration; (ii) those for managing soil evaporation; and (iii) those for increasing the storage capacity of the soil.

Water infiltration

Approaches to increase water infiltration can be grouped into three categories:

• protection of the soil surface porosity from the direct impact of rainfall;
• improvement of soil structure through biological processes;

• detention of runoff by means of physical or structural barriers.

Whether or not the soil is covered with some form of evenly distributed mulch of crop residues and/or cover crops determines whether the infiltration capacity of the soil surface will be protected from, or damaged by, the impact of high-energy rainfall. It affects how much water can penetrate, and to what depth, before the rainstorm stops. Runoff can still occur when rainfall intensity is higher than the infiltration rate of the soil, or when the pore spaces in the soil are already filled with water because the soil is shallow, the waterholding capacity is low, or its subsoil is only slowly permeable. The physical contact between a protective cover and the soil surface will slow runoff and allow more time for infiltration.

Soil covers can be distinguished according to their management, and they include mulches (vegetative and non-vegetative), crop residues, cover crops, and natural vegetation. The impact of soil cover as a protective layer over the soil is instantaneous and effective during the rainfall event.

Biological processes to improve soil structure generally seek to increase the organic matter in the soil. Organic material added to the soil will increase biological activity. Micro-organisms use it as food. The waste products produced by microorganisms become SOM. Organic matter plays an important role in the formation and stabilization of soil aggregates, resulting in greater resistance to disintegration. Organic matter loosens the soil, which increases the amount of pore space. This has several important effects that are continuous and last for a long period. Although their activity is temporary and will be substituted annually, sticky substances on the skin of earthworms and those produced by fungi and bacteria help bind particles together.

As a result of these biological processes, the density of the soil decreases (it becomes less compacted) and physical soil properties improve. This means that the sand, silt and clay particles in the soil stick together forming stable aggregates or crumbs.

The use of vegetative mulches, crop residues, cover crops and natural vegetation are examples of practices that increase the organic matter content of a soil. In this sense, the effect of soil cover is much longer.

Physical practices to increase rainwater infiltration are based on holding the water on the surface as long as possible. Runoff may be detained, or at least slowed down, by using physical or vegetative structures aligned across the slope parallel to the contour. Trash lines, live barriers and stone walls will slow down runoff, while earth bunds and fanya-juu terraces will detain the runoff.

Evaporation

Approaches to manage soil evaporation can be grouped into two categories:

• modifying soil cover;

• modifying the microclimate.

In most cases, the use of a mulch is the most effective way to minimize soil water evaporation. There are three reasons for this:

• Mulch protects the soil from insolation and reduces the soil temperature and the temperature just above the soil surface, thus increasing the water vapour concentration.

• As a result of this phenomenon, having mulch on the soil surface breaks the capillarity movement of water to the surface where it is evaporated.

• Maintaining mulch on the surface minimizes tillage that dries the soil.

Soil cover can be applied in vegetative and non-vegetative forms. Examples of non-vegetative covers or mulches are the use of stones, plastic mulch, and dust mulching. In dryland regions, soil dries quickly, and it dries to the depth that it has been tilled.

Modification of the microclimate is usually achieved by reducing the wind speed across the surface, and providing controlled shade to crops and soil surface. Windbreaks are single, double or triple rows of trees and shrubs, but also sugar cane and tall erect grasses that protect areas from wind. Wind speed is reduced and, in addition to a decrease in physical damage to the crop, crop transpiration rate and soil evaporation are reduced.

Shade to crops and soil can be provided by trees and soil cover, but also artificially by using nets, cloths, plastic sheets, and plant-derived products, such as branches, palm leaves, and grass clippings. In tropical areas, shade is often used to protect seedlings during the first few weeks of their development. Shade also reduces soil temperature.

Soil temperature not only influences the absorption of water and nutrients by plants, seed germination and root development, but it also affects microbial activity and the crusting and hardening of the soil. Roots absorb more water when soil temperature increases up to a certain maximum, which depends on the crop. High temperatures restrict water absorption. Soil temperature can be reduced and adjusted through the use of soil covers, either in the form of mulch covers or as cover crops and natural vegetation. All practices that generate shade to the crop or soil can also be used.

Water storage capacity of soil

Approaches to increase the water storage capacity of the soil can be grouped into three categories:

• increase organic matter content;
• increase in effective soil depth;

• use of fallow.

The addition of organic matter to the soil will usually increase the waterholding capacity of the soil. This is based on the fact that addition of organic matter increases the number of micropores and macropores in the soil, either by "gluing" soil particles together or by creating favourable living conditions for soil organisms. Soil water is held by adhesive and cohesive forces within the soil (pores), and an increase in pore space will lead to an increase in waterholding capacity of the soil.

As a consequence, less irrigation water is needed to irrigate the same crop, as shown in Table 1 for the Brazilian Cerrados.

All practices that increase the effective soil depth will result in an increase of in-soil water storage. The effective soil depth may be limited by compacted layers, hardpans or plough pans. By removing these layers, plant roots and soil biota can explore a larger volume of soil and so create favourable conditions for water storage. It is advisable to remove such layers before implementing practices to improve the soil moisture content. This is usually done by using subsoiling equipment adjusted in such a way that the point of the subsoiler is brought right under the compacted layer, which is typically at soil depths of 25-60 cm. Disadvantages include:

• very high power requirements;
• the soil must be at crumbling point or drier up to the depth of intervention;
• the point can break easily when used incorrectly, and worn points reduce the effect;

• when not used correctly, water infiltration rates are not improved much.

A study to evaluate the resilience of agro-ecosystems was conducted in 1999 in Guatemala, Honduras and Nicaragua (World Neighbors, 2000). This showed that 3-15 percent more water was stored in the soil under more ecologically sound practices (Table 2).

Subsoiling to break compactions should not be considered a periodic work but an absolute exception. After subsoiling, measures are to be undertaken to stabilize the loosened structure. Very great care has to be taken not to recompact the subsoiled fields. Driving with heavy machinery on a freshly loosened soil or applying intensive tillage with disc harrows might destroy the effect of the subsoiling and produce even more serious compactions than before subsoiling. In view of the high power requirement of subsoiling, biological ripping with plants developing strong and deep roots might be the cheaper option.

The use of planting pits is another way of increasing the effective soil depth. In West Africa, in the dry season, farmers dig out pits 15 cm deep and 40 cm in diameter on degraded plots every 80 cm, tossing the earth downhill. The dry desert Harmattan wind blows various organic residues into them. These are attacked quickly by termites that dig tunnels through the crusted surface, allowing the first rains to soak down deep, out of danger of direct evaporation.

TABLE 1
Economy of irrigation water through soil cover, Brazilian Cerrados

Source: J. Pereira (personal communication, 2001).

TABLE 2
Average soil depth at which moisture starts

Source: World Neighbors (2000).

Two weeks before the onset of the rains, farmers spread one or two handfuls of dry dung (1-2.5 tonnes/ha) in the bottom of the pits and cover it with earth in order to prevent runoff from carrying away dry organic matter on its surface. Millet is sown into the pits at the onset of the rainy season. The first rains run copiously over the surface crust of the degraded land. The microbasins capture enough of this runoff to soak a pocket of soil up to 1 m deep. The crop seeds germinate together, break up the slaking crust and send roots down deep to where they find stores of both water and nutrients recycled by the termites.

A common practice in semi-arid and arid areas is to increase in-soil water through the practice of a clean weeded fallow in the first year. The water stored in the soil in the fallow period adds stability to crop yields the following year. It can be used for seed germination and initial crop growth. While optimizing the potential for soil moisture storage, this fallow system has a number of negative aspects. Mycorrhizal problems frequently occur because of the absence of growing plants for such an extended period. Ground cover is almost negligible, contributing to water runoff and erosion problems. Depending on soil texture (clay content) and management, deep cracks may increase soil evaporation and reduce to zero the advantages of storing water in soil.

On the other hand, quite satisfactory yields have been obtained in drylands from crops grown after a 6-month fallow from cereals (sorghum and millet) or mung beans in years of above average rainfall. This is because:

• The fallow allows for the retention of high levels of cereal stubble on the soil surface until planting, and even beyond, to maximize rainfall entry and storage in the soil.

• Standing stubble reduces runoff and soil movement on both sloping and floodplain areas.

• Retention of cereal stubble on the soil surface after planting protects vulnerable young seedlings from sandblasting.

• Timely control of weeds and volunteer cereals will preserve soil moisture.

• Delaying cultivation in the early part of the fallow will help retain post-harvest soil cracking patterns, which aids soil profile wetting from the bottom up. Less soil disturbance slows moisture loss from the surface layers and reduces weed germination.

Rainwater harvesting - Runoff farming

There are many types of water-harvesting technologies. In general, water harvesting is characterized by a runoff-producing area (catchment) and a runoff-receiving area (cropped area or storage structure). A distinction can be made between runoff-harvesting systems that use water from the ground surfaces, and floodwater harvesting that involves the diversion of floodwater from watercourses. Water spreading involves the distribution of floodwater over the land surface for crops or pasture. Runoff-farming systems require proper planning prior to construction in order to ensure acceptance by the community and individuals. They also require regular maintenance.

Rainwater harvesting is appropriate in semi-arid and arid areas where droughts are common and irrigation is not feasible. In dry seasons, yields can increase by as much as 300 percent compared with yields without runoff harvesting. In the wet season, yields may be reduced because only a part of the land is cropped or because of waterlogging.

Rainwater harvesting includes different practices. With sheet-flow runoff harvesting, runoff is collected from gently sloping surfaces. Sheet-flow runoff is usually collected from a larger catchment/collection area and is concentrated in a smaller cropping area. The ratio of catchment to cropping area generally ranges from 1:1 to 1:3. Bare catchment areas yield most runoff, but work is needed to maintain the land in this condition. They can also be left under natural vegetation and may sometimes be sown to short-season crops, but their efficiency in terms of water collection will be less than under bare soils. Diversion ditches may be necessary upslope of the area used for runoff harvesting in order to prevent excessive damage by runoff.

Concentrated runoff is collected from narrow channels such as footpaths, cattle tracks, temporary streams, residential areas and roads. Floodwater harvesting and water spreading is the diversion of floodwater from watercourses for spreading water either over land that is to be cultivated or for storage in deep farm ponds. Rooftop harvesting is the direct harvesting of rainwater from roofs.

At farm level, farmers and land users may combine various technologies and systems, and it is very difficult to classify them. For example, in Kenya, runoff from the road is directed into canals, spread into large retention ditches with bananas, into basins with crops, and into water pans.

Conclusion

In order to minimize the impact of drought, soil needs to capture the rainwater that falls on it, store as much of that water as possible for future plant use, and allow plant roots to penetrate and proliferate. These conditions can be achieved through management of organic matter, which can increase water storage by 120 m3/ha for each 1 percent of organic matter. Organic matter also increases the ability of the soil to take in water during rainfall events, ensuring that more water will be stored. Ground cover also increases the water infiltration rate while lowering soil water evaporation. When all these factors are taken together, the severity of drought and the need for irrigation are reduced greatly (Sullivan, 2002).

High aggregation, abundant surface crop residues, and a biologically active soil are keys to drought-proofing a soil. All these qualities are advanced by reduced-tillage systems. In short, maintaining high residues and adding organic matter while minimizing or eliminating tillage promotes maximum water conservation.

References

Sullivan, P. 2002. Drought-resistant soil. Appropriate technology transfer for rural areas. USA, NCAT.

World Neighbors. 2000. Lessons from the field. Reasons for resiliency: toward a sustainable recovery after hurricane Mitch. Honduras. 32 pp.

3. Environmental consequences of drought-resistant soil and improved soil moisture management

Depending on its type and condition, soil can be an excellent temporary storage medium for water. In most cases, maintaining a high infiltration rate is desirable for a healthy environment. Proper crop and soil management of the soil can help maximize infiltration and capture as much water as allowed by a specific soil type.

However, soils that transmit water freely throughout the entire profile or into tile lines need proper management of chemical inputs in order to ensure the protection of groundwater and surface-water resources.

Soils that have reduced infiltration have an increase in the overall amount of runoff water. This excess water can contribute to local and regional flooding of streams and rivers or result in accelerated soil erosion of fields and streambanks.

Soils that have reduced infiltration can become saturated at the surface during rainfall. Saturation decreases soil strength, increases detachment of particles, and enhances the erosion potential. In some areas that have a steep slope, surface material lying on top of a compacted layer may move in a mass, sliding down the slope because of saturated soil conditions.

Where water infiltration is restricted or blocked by inadequate practices, water does not enter the soil and it either ponds on the surface or runs off the land. Thus, less water is stored in the soil profile for use by plants. Runoff can carry soil particles and surface-applied fertilizers and pesticides off the field. These materials can end up in streams and lakes or in other places where they are not wanted.

Decreases in infiltration or increases in saturation above a compacted layer can also cause nutrient deficiencies in crops. Either condition can result in anaerobic conditions that reduce biological activity and fertilizer-use efficiencies.

Despite the above-mentioned considerations, farmers and the agriculture sector are often criticized because of the negative environmental impact of their activities. This impact is caused in part by supposed inadequate management of the soil, which has led to environmental problems. These include: imbalances in the water cycle; inefficient rainwater use; pollution of land and water resources (with possible impacts on human health); soil degradation; loss of biodiversity; and increased emissions of CO₂, nitrous oxide (N₂O), etc. from the soil to the atmosphere.

For several reasons, it is difficult to arrive at universally valid statements about the impacts of improved soil moisture management on land and water resources. The impacts depend on natural and socio-economic factors. Natural factors include climate, topography and soil characteristics (texture, structure and depth). Socio-economic factors include economic ability and awareness of the farmers, management practices, and the development of infrastructure, e.g. roads and markets. Furthermore, the impacts of agricultural land use may be difficult to distinguish from natural or other human impacts.

Topic 2 examined some practices and their direct effects on local water-balance modifications. This paper considers their upscaling in order to highlight the cumulative direct and indirect effects of such practices.

With regard to natural resources, both land and water, one can distinguish impacts on quantity and quality of these resources.

Impacts on water quantity can be divided into impacts on surface-water resources and groundwater resources, of which the former can be divided into: (i) impacts on the overall water availability or the mean annual runoff; and (ii) impacts on the seasonal distribution of water availability, in particular, peak and dry-season flows are of importance.

The impact of improved soil moisture management is a function of many variables. The most important of these are: the water regime of the plant cover in terms of evapotranspiration; the ability of the soil to receive water (infiltration capacity); and the ability of the plant cover to intercept moisture. A change from a diverse soil cover to bare soil will usually increase the mean surface runoff. Conversely, a change of land cover from low to high will lead to a decrease in surface runoff.

Peak flows may decrease as a result of a change in land use if the infiltration capacity of the soil is increased. Conversely, peak flows can increase as a result of soil compaction or erosion, or if drainage capacity is increased.

The effect of land-use changes on dry-season flow depends on competing processes, most notably changes in evapotranspiration and infiltration capacity. The net impact is likely to be highly site specific. In contrast, dry-season flows from deforested land may decrease if the soil infiltration capacity is reduced, e.g. through soil compaction.

Infiltrated rainwater that has not been retained by the soil pores of the soil body and that has not been returned to the atmosphere in evapotranspiration by plants will drain down to the impermeable layer or saturated zone of the water table or groundwater. The base flow of streams and rivers derives from the groundwater table. Thus, soil conditions affect not only the availability of water to plants but also groundwater recharge.

The groundwater recharge may increase or decrease as a result of changing land-use practices. The major driving forces are the evapotranspiration of the vegetative cover and the infiltration capacity of the soil. Groundwater recharge is often linked to dry-season flows as groundwater contributes much of the river discharge in the dry season. Recharge may also increase as a consequence of an increased infiltration rate.

The water table may fall as a result of decreased soil infiltration, e.g. through non-conservation farming techniques and compaction. Where the infiltration capacity is reduced substantially, this can lead to water shortages in dry seasons, even in regions where water is usually abundant.

Although most consequences of a higher infiltration rate seem to be positive, if there is less runoff, then less water will reach rivers, and other users (irrigated areas, and industrial and domestic users) could be negatively affected.

Land-use practices can have important impacts on water quality. In turn, these may have negative or positive effects on downstream uses of water. Impacts include: changes in sediment load and concentrations of nutrients, salts, metals and agrochemicals; the influx of pathogens; and a change in the temperature regime.

Sediment can act as both a physical and a chemical pollutant. Physical pollution characteristics of sediment include turbidity (limited penetration of sunlight) and sedimentation (loss of downstream reservoir capacity, wear on hydroelectric turbines, loss of coral reefs, and loss of spawning grounds for certain fish). Chemical pollution of sediment includes adsorbed metals and phosphorous (P), as well as hydrophobic organic chemicals, leading to higher eutrophication of surface waters.

Deforestation can lead to temporary high nitrate (NO₃) concentrations in water as the result of decomposition of plant material and reduced nutrient uptake by the vegetation.

Continuous soil cover reduces N leaching while fallow periods and soil disturbance increase leaching. Ploughing can increase nitrate concentrations in surface water and in groundwater as oxygenation of the soil causes nitrification.

The application of pesticides generally poses a danger to surface water and groundwater resources as pesticide compounds are designed to be toxic and persistent. Pesticide leaching into groundwater depends on the persistence and mobility of the chemicals as well as the soil structure. As many pesticides are transported in association with suspended matter, all practices aimed at reducing runoff will reduce the chance of pollution.

Especially in arid areas, where subsurface drainage water always has higher salt concentrations, irrigation and drainage activities may lead to increased salinity of surface water and groundwater as a consequence of evaporation and the leaching of salts from soils. Evaporation will be reduced with a soil cover.

Concerning land properties, impacts of improved soil management practices can be divided in quantitative and qualitative impacts. The former include reduced soil erosion, mass movement and increased fertilizer efficiency, while the latter include increased biodiversity, reduced soil temperature and salinity problems, and reduced landscape deformation.

Soil erosion and land degradation occur where water fails to infiltrate into the soil and starts to flow over the soil surface. Practices that reduce the impact of raindrops on the soil surface and maintain soil pores intact will reduce soil loss through erosion and improve water infiltration.

Mass movement of soil, e.g. dust storms and landslides, mainly occurs where the living soil cover has been removed, either for agricultural purposes or construction activities, and where natural factors have full play. Mass movement is often perceived as a natural phenomenon. However, human activities often influence these processes. In the long run, land-management practices that allow for a soil cover will reduce the loss of soil through mass movement.

Generally, runoff water not only transports soil, but also organic matter, seeds and fertilizers. The result is a very inefficient farming system. Conversely, a reduction in runoff and an increase in water infiltration and holding capacity of the soil will maximize fertilizer efficiency. However, in some cases (e.g. humid areas), increased infiltration can lead to excessive amounts of water percolation, and, consequently, more N losses by leaching.

As mentioned, improved soil management practices affect the diversity of species, the soil temperature and salinity levels, and landscape deformation.

In general, biodiversity tends to increase where moisture-saving practices are applied. This is not only because higher moisture and lower soil temperature levels generate living conditions for more and diverse species. Compared with monocrop cultures with conventional tillage, the increased production of foliage in a system with cover crops and reduced or zero tillage leaves a protective blanket of leaves, stems and stalks from the previous crops on the surface. In this way, organic matter can build up on the soil surface. This creates favourable conditions for the activity and the population development of micro-organisms. In this way, a complete foodweb of species is built or can be re-built. Agricultural production systems in which residues are left on the soil surface, e.g. direct seeding and the use of cover crops, stimulate the development and activity of soil micro-organisms and create habitats for micro- and macro-organisms.

Soil temperatures that are too high are a major constraint on crop production in many soils and ecoregions of the tropics. High temperatures adversely affect not only seedling establishment and crop growth but also growth and development of the micro-organism population. Mulch of crop residues or cover crops regulates soil temperature. The soil cover reflects a large part of solar energy back into the atmosphere, thereby reducing the temperature of the soil surface. This results in a lower maximum soil temperature in mulched compared with unmulched soil, and reduced fluctuations.

Practices aimed at improving soil moisture levels are usually practices to reduce salinity problems. Salinity occurs when evaporation from the soil is higher than infiltration or when soils are very compacted and water fails to infiltrate. Mulches and cover crops reduce evaporation levels and at the same time protect the surface soil from the impact of raindrops that may seal and obstruct the surface pores. The increased humus level in the surface soil is important because of its negative electrical charge, which can retain cations (these in turn being exchanged with hydrogen). Reduced- or zero-tillage systems improve compaction and internal drainage of the soil.

Landscape deformation is caused mainly by severe soil erosion. In the long run, practices aimed at increasing rainwater infiltration and reducing runoff will automatically lead to conservation of the landscape. The conditioning of the soil is the first step towards the development of an erosion- and drought-resistant landscape. After crops are sown, one of the most effective methods of water conservation is weed control. Weeds can consume more water than all of the crops combined. An effective, long-term method of controlling weeds is to put a layer of mulch (5-10 cm) over the soil surface. This limits the need for hand weeding and using herbicides.

Determining the direction of prevailing winds and planting or erecting windscreens can decrease water consumption drastically. Wind draws a lot of moisture from soil and plant tissue. Reducing air movement over plants will reduce moisture loss. Most windbreaks modify air movement for a distance of about twice their height. More than one windbreak may be needed for a large planted area.

Based on the possible direct impacts of improved soil moisture management, a number of far-reaching indirect impacts on society can be generated. These can be summarized as follows:

• reduction in silting in reservoirs, lakes and watercourses proportional to 70 - 90 percent less erosion;

• consequent reduction in the pollution and eutrophication of surface waters by agricultural chemicals carried in erosion runoff;

• substantial reduction in treatment costs of water drawn from surface sources;

• reduced wear on hydroelectric turbines from the passage of cleaner water;

• considerable reductions in maintenance costs of rural roads;

• flooding risks are reduced by 30-60 percent through greater rainfall infiltration and delay to overland flows by surface residues;

• by the same token, aquifer recharge is enhanced, improving groundwater reserves and dry-season flows in springs and streams;

• provision of off-season feed (crop and weed seeds not incorporated) and shelter, lower soil temperatures and reduced water pollution levels increase populations of terrestrial, soil and aquatic fauna;

• high-yielding, prosperous and sustainable agriculture ensures lower food costs and improved food security for the population as a whole;

• reduced land preparation activities result in reductions in diesel fuel of 50-70 percent or more, and proportional reductions in emissions of greenhouse gases (GHGs);

• promoting high-yielding sustainable agriculture and increasing pasture carrying capacity (through rotation with annual crops) takes pressure off the demand for agricultural frontier expansion by deforestation.

The interlinking in space and time of all these modifications may modify many aspects of the overall functioning of the watershed (e.g. recharge of the water table and maintenance of the hydrological regime; land use; landscapes; GHG production and carbon sequestration; organic and mineral balances; and biodiversity).

4a. Adequate tools and technologies to support soil moisture management

About 95 percent of all agricultural land and 83 percent of cropland depends on precipitation as the sole source of water (Wood, Sebastian and Scherr, 2000). The interaction of precipitation and evapotranspiration as mediated by crop and soil properties determines the availability of water for plant growth. For the 17 percent of the agricultural land that is irrigated, water too is often insufficient. Wood, Sebastian and Scherr (2000) reported that of the 9 000-12 500 km³ of water estimated to be available globally for use each year, 3 500-3 700 km³ was being extracted in 1995. Of that total, about 70 percent was extracted for irrigation. According to the World Bank (2000), the share of extracted water used for agriculture ranges from 87 percent in the low-income countries, to 74 percent in middle-income countries, and to 30 percent in high-income countries.

The growing population of the world coupled with an even higher growth demand for food and fibre because of increasing living standards makes increasing the efficient use of water one of the highest priorities facing humanity. Because agriculture uses such a high percentage of the available water, agricultural practices must become more water efficient to meet these challenges. One of the important factors required for increasing water-use efficiency by plants is to know how much water is available in the soil at key times in the season. This is highly important for both rainfed and irrigated cropping systems. For the former, it is important to efforts to manage risk and make more efficient use of inputs such as fertilizers and pesticides, while for the latter it is important to efforts to use irrigation water and other inputs more efficiently.

The objective of this paper is to review some of the tools and technologies for measuring soil moisture. The emphasis will be on simple, low-cost methods that can be used by small farmers in developing countries.

Soil moisture basics

It is important to have a common understanding of some soil moisture basics before discussing methods for measuring soil moisture. The amount of water that any particular soil holds depends on several factors but most important among them are soil texture, soil structure, and the amount of SOM. Water is held in the soil at widely different potentials. Some water is tightly adsorbed around soil particles and organic matter, some is held in small pores, and some in larger pores. When a soil is completely saturated, the water potential is 0, indicating that there is free water.

With time, some of the water from a saturated soil will drain to the underlying layers of soil or materials. Gravity can exert about -33 kPa of potential. This is sufficient to drain the water from large pores in the soil. As a simple guideline, about one-half of the water that a soil contains when it is saturated completely will drain to underlying layers because of gravitational forces. This may take from one to three or more days depending on the soil. The amount of water remaining in the soil after gravitational flow has ceased is considered the field-capacity water content. This is somewhat difficult to measure in the field. This is because by the time gravitational flow has ceased, there may have been considerable evaporation of water from the soil surface, and so a portion of the soil may be below the field-capacity content. When such measurements are made in the field, the surface soil should be covered with plastic or other material to prevent or minimize soil water evaporation during the time that gravitational water movement is occurring. Field-capacity values are commonly approximated using a laboratory method that applies pressure to a saturated soil sample and then determining the water content after equilibrium has been reached. As gravitational flow will not reduce the soil water content below the field-capacity content, a soil that is shielded from evaporation and that has no plant growing on it will remain at this water content. A second simple guideline is that plants can utilize about one-half of the water that is present in the soil at the field-capacity water content. As the soil becomes drier, it becomes increasingly difficult for plants to utilize the water. At some point, the plants wilt and do not recover even overnight, and the soil moisture content at this stage is called the permanent wilting point. The water potential at this stage is about -1 500 kPa. This value can also be estimated under laboratory conditions by applying pressure and determining the moisture content after equilibrium has been reached. The difference between the field-capacity moisture content and the permanent-wilting-point moisture content is considered the plant-available water content. Table 1 presents representative values for various soil texture classes.

TABLE 1
Field capacity, permanent wilting point, and plant-available water content of various soil texture classes

Source: adapted from USDa (1955).

The amount of plant-available soil water that can be retained in silt loam is more than two times that retained in medium sand. Silt loam contains about 50 percent more PAW than does a clay soil. This is not because silt loam holds more water against gravity but because clay holds more water at the permanent wilting point. Although one of the guidelines presented above suggested that about one-half of the field-capacity water content can be extracted by plants, the actual amount varies considerably. The values in Table 3 show that 30 cm of medium sand soil holds only 3 cm of water and that 2.3 cm, 77 percent, can be used by growing plants. In contrast, 30 cm of clay soil contains 10 cm of water at field capacity, but only 3.5 cm, 35 percent, can be utilized by plants. A silt-loam soil holds the largest amount of plant available soil water.

A very important fact is that there are large amounts of moisture remaining in soil after plants have utilized all that they can extract. For example, 30 cm of loam soil contains 3 cm of water at the permanent wilting point. Although plants cannot extract this water, most of it can be lost by evaporation. This is particularly true where tillage is performed. Soil will become air dry to the depth of tillage in a fairly short time. This water must then be replaced before additional water can be saved that plants can utilize. This factor becomes increasingly important in dryland areas where water is the most limiting factor in crop production. It is also a very important factor in drought areas. This is because after plants have utilized all the water they can extract, additional water can be lost by evaporation and so the soil dries below the permanent wilting point percentage. Moreover, the longer the drought conditions persist, the drier the soil becomes and the deeper the soil dries below the permanent wilting percentage. When the rains finally return, the water that was lost below the wilting-point content must be replaced before there is water available for growing plants. This is why a common saying in dryland areas is: "One rain will not break a drought." The longer the drought, the deeper and drier the soil profile becomes, and the more the rain needed to refill the profile.

Another important point is that water in the soil profile does not move downward in a significant manner unless the water content is above the field-capacity content. Then, the water in excess of the field-capacity level will drain to the underlying layer. Although this is somewhat oversimplified, it is helpful to understand how most water moves in the soil profile. In humid areas, the soil profile seldom dries below the permanent wilting percentage and even this level is rare except at the upper layers. In contrast, soils in dryland areas can become essentially air dry to considerable depths. In areas where there is a monsoon period, the entire soil profile may be quickly replenished, while in temperate climates, months or even years may be required to fully recharge a dry soil profile.

Depth of soil moisture measurement

The depth to which soil moisture should be measured depends primarily on two factors: depth of soil profile; and the crops growing in the soil. A soil profile is a vertical section of the soil through all its horizons and extending into the C horizon. The C horizon is little affected by pedogenic processes. A more practical definition for producers is the depth of soil from which crops can extract water and nutrients. In general, soil profiles in arid regions are shallower and less fertile than those in more favourable areas. The effective depth of a soil profile is determined by depth of rooting. Chemical or physical constraints can prevent rooting below specific depths in some soils. There are also major differences in the rooting depths of various crops. Table 2 lists the approximate rooting depths of some major crops. Corn, cotton, sorghum and wheat are among the most widely grown crops in rainfed areas, and these crops normally have rooting depths of 90-120 cm.

Perhaps even more important is to understand that crops do not extract water from the lower depths as quickly and efficiently as from the upper layers. As a general guideline, plants take 40 percent of the water they use from the top 25-percent layer of the rootzone, 30 percent from the second 25-percent layer, 20 percent from the third 25-percent layer, and 10 percent from the fourth 25-percent layer of the rootzone. Thus, if a wheat crop has a rooting depth of 120 cm, about 70 percent of the water used by the crop will be extracted from the upper 60 cm of the soil. These values will vary among years and among climates. A higher percentage of the PAW present in the lower portions of the rootzone in dryland areas will usually be extracted than in more humid areas because the crops are produced under more water-stressed conditions. These guidelines also show the importance of choosing crops with relatively deep rooting patterns for rainfed areas because water will probably become limited if plants do not have access to water stored in the deeper portions of the soil profile.

TABLE 2
Approximate rooting depth of major crops

Source: Adapted from TWDB (2004).

Under rainfed conditions, it is perhaps more important to measure soil moisture deeper than under irrigated conditions. This is because the amount of soil moisture stored in the soil profile at the time of seeding is extremely important. This is particularly true for dryland regions where the precipitation during the growing season is considerably less than necessary to produce a good crop. The stored soil moisture can supplement the growing-season precipitation and result in good crops where they would otherwise fail. In such areas, it is extremely important to know how much PAW is stored at time of seeding. This is because this information can be used in combination with probabilities of receiving various amounts of precipitation during the growing season in order to make estimates of seasonal evapotranspiration amounts. Many studies have shown a close relationship between the amount of seasonal evapotranspiration and yield of various crops. Therefore, knowing the amount of plant-available soil moisture is critical for managing risk and utilizing resources efficiently.

Methods for measuring soil moisture

There are many methods for measuring soil moisture in soil. They range from simple to complex, from cheap to expensive, and from low to high in terms of labour requirements. As the focus of this conference is on measuring soil moisture at farm level by small-scale producers, only methods that are inexpensive and simple to use are presented.

Gravimetric method

The gravimetric method is simple and reliable and is the standard against which all other methods are compared. Its chief drawback is that it is somewhat labour-intensive. Another disadvantage is that it requires more than 24 hours to yield a result. The method involves taking a sample of soil, placing it in a container of known weight, weighing the container with the moist soil, drying the sample in an oven set at 105 °C for 24 hours, and then weighing the container with the sample again. The difference in weights before and after drying equals the weight of the water. The difference in weights of the container containing the dried soil and the container weight equals the weight of the dry soil. The mass wetness content of the soil is then calculated by dividing the weight of the water by the weight of the dry soil, and then converting it to a percentage by multiplying the result by 100. For example, a can that weighs 30 g is filled with moist soil, and the weight of the can plus the moist soil is 145 g. After drying the soil in the oven for 24 hours, the weight of the can plus the dry soil is 125 g. Therefore, the weight of water is 20 g (145 - 125 g) and the weight of the dry soil is 95 g (125 - 30 g). The mass wetness percentage is: (20/95) x 100 = 21.05 percent.

Under field conditions, soil moisture samples should be determined for each 30-60-cm layer of the rooting depth in order to obtain a representative picture of the soil moisture status. As major crops in rainfed areas usually extract water to a depth of at least 120 cm, then about 5 or 6 samples are required for each site measured, and several sites are usually suggested to characterize a field. Therefore, this method requires considerable time, but the results are accurate and the equipment required is minimal and inexpensive. Any type of oven can be used as long as the temperature can be maintained at about 105 °C.

The percentage mass wetness is the water content based on the weight of the dry soil. It is usually more desirable to know the volume of water in the soil. This is because it can then be compared with other water sources such as irrigation water or precipitation. For example, it is much more desirable to know that 30 cm of silt-loam soil at field capacity contains 8.7 cm of water than it is to know that it contains 19.8 percent water on a dry weight basis (Table 1). The dry bulk density must be known in order to convert mass wetness percentage to volumetric water percentage. Dry bulk density is the mass of the dry soil divided by the per-unit bulk volume; often expressed in units of grams per cubic centimetre. Soil bulk densities range from about 1.1 g/cm³ for freshly tilled soil to about 1.7 g/cm³ for coarse sands. Most soils are in the range of 1.45-1.55 g/cm³.

The volumetric water percentage is equal to the mass wetness percentage multiplied by the dry bulk density. It is easy to convert the mass wetness moisture percentage to the volumetric water content. In Table 3, the mass wetness percentages were multiplied 1.48 to obtain volumetric water percentages. Although it is strongly recommended that dry bulk densities be determined at the time mass wetness percentages are determined, an assumed value of 1.5 g/cm³ is sometimes used as an approximation. However, the best method is to take a soil sample of known volume using a device that collects uncompacted samples without mixing material from other strata. A commonly used device is shown in Plate 1.

Plate 1
A commonly used device for taking a relatively uncompacted soil sample for determining dry bulk density and mass wetness percentage values needed to determine volumetric water percentage

Feel and appearance method

The feel and appearance method is based on the fact that the feel and appearance of soil vary with texture and moisture conditions. With experience, soil moisture conditions can be estimated to an accuracy of about 5 percent (USDA, 1998).

Soil moisture is typically sampled in 3-cm increments to the root depth of the crop at three or more sites per field. It is best to vary the number of sample sites and depths according to crop, field size, soil texture, and soil stratification. For each sample, the method involves the following steps:

1. Obtain a soil sample at the selected depth using a probe, auger or shovel.

2. Squeeze the soil sample firmly in the hand several times to form an irregularly shaped "ball".

3. Squeeze the soil sample out of the hand between thumb and forefinger to form a ribbon.

4. Observe soil texture, ability to ribbon, firmness and surface roughness of ball, water glistening, loose soil particles, soil/water staining on fingers, and soil colour.

(Note: a very weak ball will disintegrate with one bounce of the hand. A weak ball disintegrates with two or three bounces.)

5. Compare observations with photographs and/or charts to estimate the percentage of water available and the millimetres depleted below field capacity.

Figure 1 presents an example for loam soils. Pictures for other textures are available in USDA (1998).

Leopold (2003) concluded that using feel and appearance is a simple, low-cost method that may be used effectively by land managers to determine a close approximation of the amount of plant-available soil moisture in the rootzone. It was primarily developed for use in irrigated agriculture in order to:

• determine when irrigation is needed;
• estimate the available water in the rootzone prior to irrigation;
• estimate the amount of irrigation water to apply;

• determine the depth of penetration of irrigation water.

Source: From USDa (1998).

In rainfed agriculture, the most important uses of the method are to determine how much plant-available soil water is in the rootzone prior to seeding a crop and to determine soil moisture status during the crop-growing season. The method is based on estimating the percentage of plant-available water capacity left in the soil. The plant-available water capacity for the soil must be known or it can be estimated by information in Table 3. For example, Table 3 shows that a silt-loam soil contains 5.2 cm of PAW when it is wetted to field capacity. If the manager takes a soil sample and follows the method outlined above and compares the results with the pictures in Figure 1 and concludes that there is 60 percent of the plant-available soil moisture capacity remaining, then there are 3.1 cm of available water remaining in that 30-cm layer. Taking additional samples down the soil profile to the bottom of the rootzone will allow an estimate to be made of the total PAW in the rootzone at any time.

In irrigated agriculture, knowing the amount of PAW allows the farm operator to know how much irrigation water needs to be added to bring the soil moisture level to the plant-available moisture capacity. This concept works well because the soil moisture content of an irrigated soil seldom drops below the permanent wilting point. Using this method in dryland soils is still valid for estimating the amount of PAW present, but soils in dryland areas can dry to levels significantly below the permanent wilting point. In these situations, water in excess of the plant-available moisture capacity for a soil layer may be required to wet the soil layer to the field-capacity level.

Soil probe method

Soil probes are simple in design, low cost, easy to use, and they yield reasonably good estimates of the amount of plant-available soil moisture in the rootzone. The principle is that a probe can be pushed easily into a moist soil but cannot be pushed into a relatively dry soil. Robinson (2003) stated that most probes are made of a 9.5-mm diameter high-tensile steel rod with a handle (Plate 2), and the tip of the probe is flared to about 12.5 cm and can be either pointed or rounded (Plate 3). The purpose of the flared tip is to reduce the friction between the soil and the probe when inserting and removing the probe from the soil.

In the soil moisture basics section, it was stated in simplistic terms that a soil layer containing more water than required to wet the soil to field capacity will drain by gravity to the underlying layer. Thus, water will continue to move deeper into the soil profile until all the layers above the wetting front have reached the field-capacity level. Once the water movement by gravitational forces has stopped, there is little or no additional downward movement of water. Therefore, the depth that the probe can be pushed into the soil profile represents the depth that the soil is wetted to the field-capacity level. However, in reality, there is soil evaporation occurring at the soil surface. Therefore, the top portion of the soil may very well be below the field-capacity value even though the soil probe can be pushed through the layer. Moreover, if probing is done during the growing season, the soil moisture level may be significantly below the field-capacity level for the upper part of the rooting depth, so making it difficult to push the probe until moister soil is met. Therefore, some adjustments, particularly in the upper layers, must be applied while using the method. Again, determining the amount of plant-available soil moisture in the profile depends on knowing the texture of the soil at the different soil depths. If the rooting depth is 90 cm, and the texture is fine sandy loam, then the data in Table 3 indicate that each 30-cm layer contains 5.2 cm of plant-available soil water, so there is about 15.6 cm of PAW in the rootzone. If the top 30 cm is lower than the field-capacity value, then an adjustment should be made and the feel and appearance method described above can easily be used to make this adjustment. If the soil texture changes with depth of profile, then the soil moisture capacity value must be changed for each depth. For example, if the top 30-cm layer of the 90-cm rootzone is a sandy loam, underlain by 30 cm of loam, and then by 30 cm of clay, there would be about 12 cm of available soil moisture in the rootzone based on the values listed in Table 3. It would be better to have actual data for the soils being probed, but where no data are available, reasonable estimates can be made using the values in Table 1.

Plate 2
A common soil probe used in the Great Plains, the United States of America, to estimate the amount of plant-available soil moisture in the rootzone. The probe is 150 cm long

Plate 3
Flared tip of the soil probe for reducing friction between the probe and the soil

The soil probe method for estimating soil moisture has limitations and these should be clearly recognized and considered. The method should not be considered an accurate method for quantifying the amount of available soil moisture in the rootzone, but it is an excellent way of assessing the moisture status of soils. The method is widely used in the Southern Great Plains of the United States of America by producers and consultants. One of the limitations of the method in that area is that the soil surface layer dries by evaporation to the extent that the probe cannot be pushed into the profile. In these situations, it is necessary to dig through the dry layer with a shovel and then use the probe for assessing the moisture status of the lower layers. Some researchers are also critical of estimating PAW based on soil texture because organic matter levels and soil structure have large effects on PAW and these can vary considerably between soils of similar textures. Crop- and tillage-management systems can alter PAW on adjacent plots of the same soil. Therefore, while this method should not be considered a precise measurement of plant-available moisture in the rootzone, it does provide quick and valuable information.

The use of the method in the Great Plains of the United States of America has had its greatest impact for estimating the amount of plant-available moisture stored in the rootzone prior to seeding a crop. The potential evapotranspiration during the growing season in this area far exceeds seasonal precipitation, and satisfactory grain yields can only be achieved in most years when there is a significant amount of stored soil moisture in the rootzone at time of seeding to supplement the growing-season precipitation. Using a soil probe to quickly estimate the amount of stored soil moisture prior to seeding, coupled with using probabilities of receiving differing amounts of seasonal precipitation, is an excellent procedure for risk management in dryland areas. Many producers will not seed a crop unless there is a satisfactory amount of soil moisture stored in the rootzone.

Stewart and Koohafkan (2004) suggested as a general guideline that, for each additional cubic metre of water used for evapotranspiration in dryland regions, 1.7 kg of maize grain can be produced, 1.5 kg of grain sorghum, and 1.3 kg of wheat. Therefore, an additional 25 mm of seasonal evapotranspiration could potentially increase the yield of maize, grain sorghum, and wheat by 425, 375 and 325 kg/ha, respectively. FAO (1996) reported that the 1988-1990 average yields of maize, sorghum and wheat in developing countries in semi-arid regions were 1 130, 650 and 1 100 kg/ ha, respectively. Therefore, storing and using an additional 25 mm of precipitation could potentially raise average yields of maize, sorghum and wheat in those areas by 38, 58 and 30 percent, respectively. These estimates show the importance of assessing the amount of stored soil moisture in dryland areas at time of seeding, and the soil probe is a quick and satisfactory method. CA principles (FAO, 1994), particularly keeping residues on the soil surface, have been effective in increasing the amount of soil moisture storage in the rootzone at seeding time. However, this does not apply to dryland areas where monsoon precipitation events exceed the capacity of the soil profile to retain the moisture. In these areas, the rootzone is usually fully charged at time of seeding, but CA principles may still have dramatic effects on retention and use of precipitation occurring during the growing season.

Other methods

There are several other methods of measuring soil moisture. They including: time domain reflectometers, capacitance probes, neutron probes, gypsum-porous blocks/ electrical resistance, calcium-carbide gas pressure meter, and tensiometers. These are not reviewed in this paper because they are not easily and cheaply adapted to field conditions, particularly in developing countries. However, BLM (2003), Ley et al. (2004) and NRCS (2004) provide information about these methods.

Summary

Soil moisture measurement is critical for managing water resources in an efficient manner. This applies to both irrigated and rainfed cropping systems. Water is increasingly becoming the most limiting resource needed to meet the food and fibre demands of a growing and more affluent population. The feel and appearance method and the soil probe method for measuring plant-available soil moisture are quick, low cost, and suitable for many situations. Although they are not as accurate as some other methods, the information they provide can be extremely useful. The first decisions that must be made in choosing a method are to determine what the information is going to be used for and how accurate the information must be before its usefulness is muted. Neither the feel and appearance method nor the soil probe method would be an acceptable method for conducting a research water-balance research study. However, both of them can be very helpful for producers and consultants making decisions about when and what crops to grow and for assessing the probabilities of producing economic crop yields.

References

Bureau Land Management (BLM). 2003. Soil moisture measurement methodology. Washington, DC, Bureau Land Management, U.S. Department of Interior (also available at http://www.blm.gov).

FAO. 1996. Prospects to 2010: agricultural resources and yields in developing countries. In: Volume 1, Technical Background Documents 1-5. World Food Summit, pp. 26-36. Rome.

FAO. 2004. Intensifying crop production with conservation agriculture. Rome (also available at http://www.fao.org).

Leopold, R.P. 2003. Soil moisture measurement by feel and appearance. In B.A. Stewart & T.A. Howell, eds. Encyclopedia of water science, pp. 847-851. New York, USA, Marcel Dekker Inc.

Ley, T.W., Stevens, R.G., Topilec, R.R. & Neibling W.H. 2004. Soil water monitoring and measurement. Washington-Oregon-Idaho, USA, Pacific Northwest Publications (also available at http://cru.cahe.wsu.edu).

Natural Resources Conservation Service (NRCS). 2004. Irrigation water management. National Engineering Handbook Part 652. Washington, DC, NRCS, U.S. Department of Agriculture (also available at http://www.wcc.nrcs.usda.gov).

Robinson, C.A. 2003. Soil water storage measurement by soil probes. In B.A. Stewart & T.A. Howell, eds. Encyclopedia of water science, pp. 908-910. New York, USA, Marcel Dekker Inc.

Stewart, B.A. & Koohafkan, P. 2004. Dryland agriculture: long neglected but of worldwide importance. In: Challenges and strategies for dryland agriculture. Special publication. Madison, USA, Crop Science Society of America.

Texas Water Development Board (TWDB). 2004. Agricultural water conservation practices (available at http://www.twdb.state.tx.us).

United States Department of Agriculture (USDA). 1955. Water. Yearbook of agriculture. Washington, DC, USDA.

United States Department of Agriculture (USDA). 1998. Estimating soil moisture by feel and appearance. Program Aid No. 1619. Washington, DC, USDA NRCS (also available at ftp://ftp-fc.sc.egov.usda.gov).

Wood, S., Sebastian, K. & Scherr, S.J. 2000. Pilot analysis of global ecosystems: agroecosystems. Washington, DC, International Food Policy Institute and World Resources Institute (also available at http://www.wri.org).

World Bank. 2000. World development indicators. Washington, DC.

4b. Using models for optimizing soil moisture management strategies

Farmers and extension services are interested in or are actively experimenting with new crop options. These include CA methods, direct seeding, minimum-tillage methods, and crop-residue management. In order to develop cropping systems that are well suited to the environmental conditions, the agronomic performance of the envisaged systems needs to be understood. Given the large number of options and the desire to develop viable cropping systems as rapidly as possible, it has been suggested that these options could be investigated by using crop modelling as virtual field research experimentation. The modelling effort investigates the environmental conditions (water, light, nutrients, etc.) required by various crops to develop, grow, and produce economical yields. This information is matched with long-term weather and environmental data in order to assess crop suitability to specific agronomic zones and to aid the development of complementary crop sequences. The modelling effort requires field experiments in order to parameterize the model for multiple crops, and to validate model predictions through comparisons with actual field data. Subsequently, model and field experience can be combined to design field experiments for evaluating the most promising crop sequences, sowing dates, sowing densities, crop-residue management, etc.

The search for the best combinations in varieties and sowing dates, or changes in crop management towards different cropping systems (less intensive systems, CA systems, dry-farming systems, etc.), or the simple prospective studies that farmers need to conduct before deciding a long-term investment require that the possible performances of these systems be forecast for various climate conditions in space and time.

The experiments generally conducted with this aim often give a partial response. This is because the ranges of soils types, climates, and crop managements covered are limited compared with the agricultural conditions in which these systems are able to be used, and because the time required to obtain a response is too long in relation to the rapid change in varieties and all the cropping-system components used and modified by farmers.

It is possible to use new forecasting tools to accelerate this step. This paper presents the use of crop models to analyse and predict the production and environmental effects of different cropping systems in a large range of environmental conditions. It describes an actual story, concerning CA, in which modelling options have been taken for agronomic research, following several years of practices in field experimentation either in experimental agronomic centres or at farm level.

Specific considerations on modelling CA cropping systems

In response to the negative impact of soil degradation processes under "conventional" cropping systems in tropical areas, direct-seeding mulch-based cropping systems without tillage practices and with a protective cover of crop residues are being developed in wide areas of the tropics. These cropping systems are based on four essential farming practices: (i) not tilling the soil; (ii) maintaining a mulch of crop residues at all times; (iii) direct seeding into crop residues; and (iv) using suitable crop successions. A key principle of these cropping systems is the retention of crop residues on the soil surface in order to reduce surface water runoff and erosion. Under semi-arid conditions, surface plant residues are also supposed to play a role for conservation of soil water through reduced soil evaporation. In addition, the use of crop residues as mulch moderates temperature fluctuations of the topsoil layer and enhances the activity of soil fauna, which increases the release of nutrients, improves water infiltration, and facilitates root development. Finally, a mulch of crop residues may also contribute to the control of weeds, by smothering them or through allelopathic effects.

Combining all these effects, the water conservation effect in soil moisture may increase crop yields in tropical environments where drought stress is occurring, and reduce production risks related to climate variability and change.

Step 1: understanding basic biophysical mechanisms

The following three effects were studied in detail: (i) reduction in evaporation; (ii) increase in infiltration rate; and (iii) reduction in runoff.

These effects were studied separately, through experimental and theoretical aspects.

Soil evaporation

Evaporation was studied using a mechanistic model dealing with water and heat transfers in the soil-atmosphere system that had been adapted to mulched soils. Experimental data were collected on four uncropped evaporation plots. The plots measured 3 x 3 m² and were covered with 0, 1.5, 4.5 and 15 tonnes/ha of residues, which represented a mulch-covering rate of 0, 0.3, 0.7 and 0.95. Measurements concerned soil moisture and temperature in the layer 0-30 cm.

A relationship between the evaporation and the mulch-covering rate was derived from the results, and simplified in a linear way with a simple evaporation reducing factor of the mulch that was related to the mulch physical properties (porosity and thickness).

Infiltration

The soil hydraulic properties obtained by the Beer-Kan method, which consists in a combination of theory and measurements of infiltration rates in beer cans with standardized protocols. Infiltration trials were made at the soil surface removing the 0-2 cm soil layer and at 50 cm. The mulch effects on the soil moisture at saturation were negligible as similar values were found for the different situations. However, large differences were observed for water conductivity at saturation (from 1 to 10). The difference was strongly attenuated at 2 cm, and not significant at 50 cm. This behaviour is explained by the formation of a crust, which reduced infiltration dramatically.

Runoff

Runoff on mulched soil was simulated by a physically based model, which accounted for rain interception by the crop and the mulch, water friction on the mulch, pathway tortuosity, and channelling in rills. Four experimental runoff plots (RPs) were used to collect the data necessary to run the model. During the experimental campaign, 22 runoff events were recorded and used for modelling. The results showed a strong mulch effect when comparing RP1.5 with RP0. The small amount of residue of RP1.5 cut down runoff dramatically. Mulch increased friction and pathway tortuosity, which slowed the flow down and improved infiltration rate. A comparison of RP4.5P with RP1.5P showed that mulch density also had an important impact on runoff.

These three experimentations, combined with theoretical approaches, give some consistent information on how the system works. It also legitimates further integration into global functioning models. However, the derived modules need several parameters in order to be used. Figures 1 and 2 show how crop residues may modify runoff conditions. Physical equations may be used for describing these considerations.

Step 2: integrating the results in a global functioning model

The focus in this paper is on using simulation models for predicting and evaluating crop growth and yield under a wide range of alternative management scenarios. In specific cases such as CA systems, the very important effect of practices and residues management on water availability for crops necessitates the accurate predicting of the effects of surface crop residue on soil moisture (and the environment) as they have a strong influence on the final crop yield.

STICS is a generic crop model that is able to simulate the development, growth and yield of various crops (e.g. wheat, maize, soybean, tomato, lettuce, banana, and sugar cane). It links crop growth to soil water and C and N dynamics. Hence, it is suitable for predicting the effects of management practices such as fertilizer application, sowing date and irrigation on crop yield.

As a means of taking into account complex interactions between management practices and soil and weather conditions, STICS was then selected for considering simultaneously the major effects of direct sowing with mulch, and for providing a detailed water balance of the soil-mulch-plant continuum, including soil moisture prediction at any moment of the crop cycle. This upgrade was based partially on simplifying the results described above.

The following concept considers mulch as being a partial (discontinuous) screen between atmosphere and soil for radiation and for rainfall or irrigation water. The mulch also modifies the roughness of the soil-atmosphere interface, and thus the conditions for partitioning rainfall between infiltration and runoff. The concept considers that:

• Mulch is degraded during the crop cycle - a function of time.
• Mulch intercepts rainfall water, like a sponge, and varies according to the type of residue.
• Mulch evaporates according to the energy intercepted by the mulch (radiation and wind).

• Mulch modifies roughness conditions in the soil-atmosphere interface. This is taken into account in the model: if rainfall is less than a calibrated threshold, then there is no runoff; if rainfall is higher than the threshold, then runoff is equal to a fraction of the rainfall.

All these options are much simpler than the ones established in the model described above. However, they are homogeneous with the general complexity of the other crop and soil parameters required to feed the model.

Step 3: building a simulation framework

In this example, data from two experimental sites with maize were used for model testing and application. One site was located at La Tinaja, Mexico (1 200 m altitude), the other at Planaltina, Brazil (1 100 m altitude). The sites represent contrasting conditions of tropical environments: the La Tinaja site has a semi-arid tropical steppe climate, whereas the climate of Planaltina is humid tropical of the savannah type. Mean annual rainfall at La Tinaja is 525 mm with 80-90 percent of the rain occurring between June and September. However, dry spells of ten days or longer are common during the rainy season. The mean annual reference evapotranspiration at La Tinaja is about 1 710 mm and the mean temperature during the growing season is about 25 °C. At Planaltina, the mean annual rainfall is 1 400 mm, with a dry season from May to September. Reference evapotranspiration is fairly constant throughout the year with a mean annual total of 1 480 mm. The mean maximum and minimum temperatures during the growing season are 17 and 27 °C, respectively. The parameters used in the simulation were extracted from previous field experimentations conducted since 1995. For climate conditions, long-term data (30 years) on daily radiation, temperature, rainfall and reference Penman evapotranspiration were used in the simulations.

In order to quantify the impact of crop-residue mulching on soil moisture and maize-yield likelihood for a range of management and soil conditions at both sites, simulations using long-term weather records yielded sequences for a typical local maize cultivar for three sowing dates (early, medium and late) and two soil depths (low and high plant-available soil water storage capacity), and examined the interactions with three levels of surface residue (no mulch, low and high quantity). This resulted in a simulation matrix of 36 combinations.

Step 4: analysing results

At La Tinaja, surface-residue management had a large impact on yield likelihood. A crop-residue mulch of 1 tonne/ha resulted in a considerable increase in the yield values compared with the simulations with no mulch. The absolute increase in median grain yield with 1 tonne of residues per hectare varied between about 1 000 kg of grain per hectare (late sowing and low soil moisture storage capacity) and 1 600 kg/ha (late sowing and high soil moisture storage capacity). The effect of an extra 2 tonnes/ha of surface residue was much smaller. Averaged over all combinations, the median grain yield increased by about 500 kg/ha when passing from 1 to 3 tonnes/ha of mulch surface residue. Simulations also showed that delayed sowing (from 20 June to 30 July) decreased yields for all levels of surface residue.

At Planaltina, the highest median yield and lowest variability were predicted for an 18 October sowing with 6 tonnes/ha of surface residue. However, the impact of surface-residue management on maize yield likelihood was much smaller than at La Tinaja. For example, for early sowing, with retention of 6 tonnes/ha of surface residue, increases in median values were less than 500 kg of grain per hectare, and almost zero on the 75-percentile yield values. This indicates that crop-residue mulching at Planaltina is most effective in years of low yield potential (low water supply).

Surface residue management under the climate conditions at Planaltina also has a major impact on the risk of water drainage. As an example, the model predicted for an 18 October sowing date an increase in median water drainage of about 150 mm (from about 470 mm to 620 mm) with the retention of 6 tonnes/ha surface residue. While drainage exceeded 570 mm in 25 percent of years for the zero mulch treatment, this level becomes 700 mm for 6 tonnes/ha of surface residue.

Discussion

The approach presented in this paper has produced very interesting results concerning soil moisture management through CA systems in two different tropical environments. These results, and the strategies that may be deduced from these results and later adopted by farmers, are quite different. In dry areas of Mexico, it is a good tool for managing limited crop residues. It is possible to define thresholds for using part of crop residues for feeding cattle, and another part for limiting soil evaporation and stimulating rainwater infiltration within CA systems. In Brazil, it is a good tool for managing sowing dates, and it may also be useful when considering N fertilization strategies.

Because of soil diversity and climate interannual variability, many supplementary years of field experimentation would have been necessary to obtain these results without the support of a model. Moreover, the results would probably not have been so precise, and, hence, nor would the deduced optimized strategies.

However, these results may be considered at various levels, for example:

• A model is a simplification of the reality. What is the legitimacy of the simplifications?

• Does the model take into account the really important phenomena? What about other phenomena, such as biological activity in soils, and nutrient bioavailability?

• The models are not always used in the place where they have been calibrated. Are the sets of parameters still appropriate in a new situation?

• How can models convince farmers?

A lot of questions may be addressed; some of them are generic and may concern all agronomic models. Others need more specific answers within the framework of soil moisture management. However, the conclusion of this paper is that models are tools (field experimentation, physical or biological indicators, etc.) that may help researchers, extensionists, farmers and politicians in taking decisions. If they are used in the appropriate way, they may be very powerful.

5. Conditions for adoption of drought-proofing practices by farmers

Agricultural practices that consider and work with elements of agro-ecosystems (soil, plant nutrients, water, vegetation and biota) in isolation from one another and from the ecosystem of which they are part have a common bottleneck: site specificity. Such approaches cannot guarantee an adequate upscaling under great heterogeneity of socio-economic and agro-ecological conditions. Instead, approaches based on agro-ecological principles, which have universal applicability, are much more effective for drought-proofing soils, land and watersheds.

It is being recognized that minimal mechanical soil disturbance is an essential component of a resource-saving, practicable concept for crop production that strives to achieve acceptable profits and high and sustained production levels while conserving and enhancing the environment.

In the past, the activity of soil and water conservation has been advocated as a necessary starting point for raising crop yields. Conventionally, soil erosion has been perceived as a major cause of land degradation and the main reason for declining yields in tropical regions. Another important factor for declining yields is the fact that ever more marginal (degraded) lands are being brought into production. However, for the moment, the use of fertilizers, improved varieties, pesticides, etc. may be masking the seriousness of the situation.

Experience has shown that none of the recommended physical and institutional anti-erosion methods has been widely adopted by smallholder farmers in tropical regions. Together with the understanding that water erosion is not a cause but a consequence of poor rainwater infiltration and consequent runoff, this indicates the need for a switch from "stopping erosion" to assisting farmers to achieve a higher, more conservation-effective, and more stable production.

Similarly, especially in the arid and semi-arid tropics, it is important to emphasize with farmers the management of rainwater as a productive resource more than merely as a means of saving soil. While favouring agricultural production, achieving better infiltration and in-soil storage of rainwater also reduces soil and water movement. In this regard, in order to enhance water availability and retain soil productivity, it is important to consider those practices that affect rainfall catchment before considering those that aim to control runoff. They are complementary in a sequence and not competing alternatives.

Who is the customer?

Anyone with a product or practice to promote will benefit from understanding the customer. The same applies to conservationists seeking to convince farmers to adopt drought-proofing practices. Any practice or system to be recommended to farmers should be feasible and acceptable in the prevailing socio-economic conditions. For example, soil crop-residue management should be economically viable if it is to be recommended where livestock now uses a major part of crop residues.

Currently, enabling farmers to improve their land use provides a more effective response than efforts to combat erosion alone. It specifically recognizes farmers' desire to raise yields and incomes as they stabilize or reverse resource depletion. It also provides opportunities for governments to harmonize certain national objectives (better management of natural resources and development of sustainable agriculture) with major objectives of farm families (secure livelihoods). However, by seeking both improvements in land and increased production, this approach requires many adjustments in common thinking.

Usually, farmers' decisions about change are influenced strongly by their assessment of the risks attached to an innovation, including its possible side-effects. Their seemingly conservative attitudes may in fact be an essential caution in weighing the possible benefits and hazards that could follow change. Where new drought-proofing practices have met farmer requirements for risk aversion, create no major conflicts and have an assured beneficial effect, adoption has been shown to be very rapid. In this sense, farmers have the ability to make development sustainable - to ensure that it meets the needs of the present without compromising the ability of future generations to meet their own needs.

Farmers have latent skills and enthusiasm that are tapped when they are involved in doing things that concern and interest them. Experiences with resource-poor small farmers show that in cases where drought-proofing practices can increase their cash incomes, they are keen to adapt and adopt such techniques even if this may lead ultimately to a complete change in their farming system. Nevertheless, farmers often do not have contact with people and information that can help them develop appropriate solutions to the problems that they really face (which are not always those in fashion at policy and technical levels).

Adaptation processes in agriculture involve technical and organizational innovations that are both collective and individual. In considering and adapting to technical change, farmers and all concerned stakeholders have to make decisions and act.

Conditions for adoption of drought-proofing practices by farmers

The challenge is for advisers to develop the sense of partnership with farmers, participating with them in defining and solving problems rather than only expecting them to participate in implementing projects prepared from outside.

The success of improving land husbandry in a catchment depends not just on the motivations, skills and knowledge of individual farmers, but also on action taken by groups, communities and regions as a whole. Simple extension of the message, even coupled with demonstrations, will usually not suffice. Increased attention to community-based action through local institutions and users groups will also be required.

The development of common-interest groups serves to provide encouragement and mutual support to members as they make the changeover. These groups are usually very effective in farmer-to-farmer spread of the beneficial ideas and practical technologies. In addition, they can develop into significant local pressure-groups for improvements in the policy and institutional environment in order to obtain political and legal support for their own initiatives.

Creating drought-resistant soils starts from a thorough understanding of the current land-use situation of which only the farmers themselves have an intimate knowledge. This is why they have to be the architects of change from the outset (design) and supported during the change process, and even further in order to make sure that the new situation is sustainable. The principal focus must be on the farmers rather than on the land alone, as they are the ones who make the ultimate decisions on land use and management.

For better results in the adoption process, the priorities and context of the farmers and farmer families have to be considered while addressing the problems and designing the strategies. Solutions should help reduce risks inherent in markets and weather fluctuations, and intensify output per unit area in ways that are conservation-effective. This approach is more realistic than expecting farmers to adopt standardized recommendations.

When given the opportunity, farmers have been innovative in adapting technologies to their own conditions, often having a significant reverse impact on research and extension institutions in the process. Empowerment of farmers to solve their own problems is achieved not only through training but also by introducing ideas and information from which farmers can make their own choices. In addition, it is widely accepted that the experiences of early adopters of new technologies are an important source of information for other people in making their adoption decisions. Thus, communication between adopters and others helps to promote adoption and diffusion through the information externalities or services of early adopters. However, if people are strategic, the communication opportunities can hinder diffusion where agents intentionally delay in order to learn from others.

Advice and technical support need to be provided in order to assist these groups in putting their ideas and plans into action in ways that can be self-sustaining. Emphasis should be on helping rural people to develop and implement their own solutions and plans rather than imposing programmes from outside. In order to favour such synergy, the advisers should also make sure that the tools necessary for farmer experimentation are physically and economically within reach.

It will be important to be able to push for improvements in moisture capture and management on the basis of benefits that farmers really appreciate, not only in terms of yields but also in terms of yield stability, drought minimization, income improvement, etc. This has been one of the problems with soil and water conservation as a discipline. It has often been "sold" on the basis of it "obviously being a good thing", "will raise your yields", etc., but in fact failed to ignite farmers' interest because of not producing the goods that farmers had been told to expect. In some situations, downstream considerations such as better regularity of water supply may provide prime motivation for action.

From the viewpoint of farmers too, the matter of the relations between various soil moisture situations and subsequent yields, and between these and net benefits from marketing the crops, will be of great concern. Therefore, even if a technical solution is found, it is necessary to consider its economic, social and cultural implications at the stage of encouraging farmers to adapt it before adopting it, in order to make sure that it produces economic advantages in terms of higher yield, lower unit costs, or other benefits that they desire.

This requires:

• engaging in the change process by considering formal (rules and public policies) and informal social processes (professional networks and collective actions);

• designing methods to carry out multidisciplinary research in partnership with the farmers to test and adapt techniques to local contexts;

• building a common conceptual framework embracing adaptive management, technical change, learning and coordination.

Decision-makers will need to develop and maintain policies that facilitate the initiatives and respond to the needs and desires of local groups and land users, especially smallholders. The land users should be provided with inter alia relevant information on the benefits to be gained from good practices at farm, village and environmental levels. Contradictory policies, lack of technical support, inadequate infrastructure and market access, and inadequate communication channels between the different stakeholders would undermine the success of good initiatives.

The creation of an adequate policy framework is a necessary step for fostering the design and implementation of improved soil moisture management practices and, more generally, adequate land-management systems that lead to economic benefits for the farmer as well as environmental benefits for the community.

The first step might be to identify constraints and resolve conflicts, especially regarding the use of land, water and other natural resources. This requires proactive research with stakeholders in identifying opportunities including strengthening capacities for diagnosis, planning and management as well as participatory monitoring and impact assessment.

Once started, the process will consist of several different phases that are all characterized by different training needs. The "motivation" stage is to bring together the leaders (municipal), technicians and rural farmers through a series of contacts, visits, meetings and field excursions with the objective to exchange experiences, with the idea that external experiences help in developing appropriate participatory methods. The next stage consists of training and updating technicians of the public and private sectors and especially farmers. After initiating work, extensionists together with regional leaders participate in courses, seminars, and meetings dealing principally with soil management. The objective is to keep them up-to-date and aware of recent technical and scientific advances in sustainable rural development, leadership, group dynamics, farmer organizations, environmental education, and agro-ecological production in its widest sense. The training of farmers is a basic activity in the change process both in specialized training centres and in their own communities. Other training and motivation activities may include municipal and community seminars that focus on topics like soil moisture management, erosion control, advantages of maintaining the soil covered for as long as possible, use of green manure and cover crops, and direct-sowing methods. At the same meetings, health-related issues and nutrition can be discussed.

The private sector should form part of the rural extension component through cooperatives and agro-industries. As they provide technical assistance to their members, one can think of agreements made with their class organizations and members to guarantee quality of the information provided.

In the absence of adequate government technical or credit support, non-profit, non-commercial and non-political entities such as farmer clubs at municipal level and farmer field schools may provide farmer self-help. They usually have regular monthly meetings on topics of interest and are highly effective both in spreading the technology and in raising the level of farmer achievement. Their activities include: exchange of experiences, clinics for beginners, tests of new practices, farmer-driven on-farm research, feedback to research, field days and excursions, panel debates, technical lectures, short courses, and disclosure in the mass media.

Facilitation of the change process should have the following features:

• Attention is directed as a priority to farms that develop family agriculture, ones that manage natural resources inadequately and ones that have a low grade of association.

• Technical support to farmers does not have just a preventive nature; it is part of a rural development strategy.

• The result of the work is expressed in economic, social and environmental indicators, such as increases in production and productivity, the improvement of sanitary conditions, and progress achieved in recuperation of natural resources

• Supervision is continuous and lasting so that all impacts of the programme can be evaluated.

• The norm is collective attention provided to the community with the perspective of working together to solve common problems and the discovery and analysis of the potential of the area.

• The work is characterized by the integration of extension services, research, promotion and assistance in the fields of health (nutrition) and education.

• There are political and administrative opportunities to establish vertical articulation of government bodies linked to agriculture with decentralization of decision-making.

• Development of collective awareness of the potential of the people.

• The concept of sustainable development is the basis of the activities.

As an example, and assuming that implementing CA practices is the right option for optimizing soil moisture management, the following section proposes a set of steps for implementing a strategy for stimulating adoption by farmers.

Steps for implementing drought-proofing practices

Step 1: awareness, information and training

Awareness

In order to create awareness effectively, it is necessary to:

• indicate how CA can reduce or resolve the current problems farmers and the agriculture sector are facing;

• inform parties of the environmental and socio-economic benefits of sustainable natural resources management provided by CA;

• address resistance to change issues from local to international levels;

• interact with a broad spectrum of stakeholders involved in agricultural development, including farmers and agricultural workers, civil society organizations, indigenous peoples, public and private organizations, consumers and policy-makers.

Information

It is crucial to guarantee access to more specific and technical information by interested people, associations and institutions. Capacity to inform farmers is essential in this bottom-up approach. The main issues relating to CA that would be of interest to farmers are:

• potential benefits (mainly short term);
• potential risks (mainly short term);

• how CA can be implemented in specific situations (it is particularly important to clarify the operations needed where major effects of soil degradation are present).

Organizations involved in this action should be coordinated in order to provide up-to-date and consistent information.

Training

The aim of training is not to spread the system over the whole territory but to establish "lighthouses" where farmers can learn CA principles and concepts and how to implement the system in their local ecological and economic conditions. If they consider CA effective and beneficial, the adoption and spread process will be farmer-driven.

Farmers are the main actors. Their expertise and creativity is essential for adapting CA principles to their own conditions and to new challenges.

Step 2: equipment and input supply, technical support, farmers organizations

Equipment and input supply

Access to CA equipment and inputs is indispensable for its adoption. Partnerships between farmers organizations and the private sector (local craftspeople/artisans and companies) can facilitate:

• purchase of inputs;
• acquisition or hire of equipment;

• modification of conventional tools.

Technical support

The provision of technical support serves to overcome limitations that farmers face, mainly caused by the lack of experience with drought-proofing systems. Capacity building of technical personnel of local public institutions is necessary for ensuring quality and appropriate support.

Extension workers and researchers can undertake deeper studies, long-term and expensive experiments, thereby creating a solid knowledge base. They can also transfer useful outputs between different areas or to a larger scale.

Farmers organizations

Farmers organizations ensure the effective implementation and spread of drought-proofing practices as they understand and defend the needs and interests of farmers. Such organizations can:

• stimulate farmers to analyse and improve their systems and activities;
• provide useful information and support individuals on different issues related to drought-resistant soils;
• facilitate knowledge sharing with other farmers, farmers associations and institutions;
• help to concentrate efforts in resolving common problems;
• strengthen and secure equal communication and collaboration with research and technical institutions;
• facilitate communication with civil society and within their communities;

• reinforce their influence on policy-making.

These organizations serve to strengthen the role of farmers at community, national and international levels. Support for the creation of new farmers organizations, or for existing ones, and for linking such organizations will facilitate consistent agricultural development.

Step 3: financial support, adequate policy framework

Financial support

Adequate funding programmes are needed to support farmers, particularly resource-poor ones, during the 3-5-year transition phase. This support is crucial in areas where there is major soil degradation because the farmer has to undertake preparatory and expensive operations. However, it is important to ensure that existing incentives and subsidies do not hinder the implementation of CA.

Non-governmental organizations (NGOs) can facilitate smallholder adoption of CA through technical support to farmers with low assistance from extension services, facilitating machinery access and encouraging the creation and strengthening of farmers organizations.

A possible option for facilitating CA adoption could be to provide incentives to farmers for environmental services. This will:

• stimulate farmers to undertake sustainable land management;

• promote awareness and understanding of the socio-economic and environmental impacts of sustainable resource management.

Adequate policy framework

Decision-makers must enable an appropriate environment in order to meet the requirements and facilitate the initiatives and needs of local groups and land users, especially undervalued smallholders. Policy support should:

• establish effective communication channels with rural communities, NGOs and the private sector;
• address land tenure and related social conflicts (with a special focus on gender issues);
• adjust legislation to the new goals and needs;
• ensure the development of infrastructure to facilitate farmer access to markets;

• institute policies for fair access to markets in order to increase economic security for farmers.

The agriculture sector must be re-oriented to include local environmental, economic and social development and not only to support industrial and export agriculture that is easily translated into national economy terms.

In conclusion, when correctly implemented and under adequate circumstances, CA (based on drought-proofing practices) can constitute a valid tool for improving rural livelihoods and fostering bottom-up sustainable rural development.

Certification process for drought-proofing practices

Agriculture needs some subsidies in order to survive these difficult economic times. It may be possible to give something back to taxpayers for their generosity, in this case better air and cleaner water, and an improved soil on which their food is produced. Certification processes for drought-proofing practices for decision-making at farm level have to be increasingly recognized by governments and civil society, including food industries, as the essential prerequisite to food safety from farm to fork. The consumer is still not willing to pay to preserve natural resources (soil and water) in spite of all the extolling of environmental values.

The option of supporting sustainable agriculture through segregated markets for sustainable products is interesting. However, a limitation is that some markets may remain limited. Therefore, one proposal is for practitioners to engage in dialogues with the food industry and a broad range of stakeholders in specific food-supply chains in order to better include sustainability and environmental concerns along with food safety and quality concerns, leading to a special certification.

The certification process already embraces actions, technologies and systems that are accepted as most effective for optimal management of soil and water, and for crop and livestock production. The details of a CA certification protocol in a given production environment cannot be generalized and prescribed from a central information source. They but must be adapted locally, taking into consideration local conditions and market requirements.

The hope is that some ideas provided during this electronic conference will eventually lead to an international certification system for drought-proofing practices.