Chapter 4

Criteria and options for appropriate irrigation methods

The aim of modern irrigation development must be to make the best use of water in conjunction with land and human resources, as well as with all other essential inputs (energy, machinery, fertilizers and pest control measures) so as to enhance and sustain crop production. The selection of an appropriate irrigation technology for any given combination of physical and socio-economic condition involves complex and sometimes conflicting considerations. Where water shortage is acute, the obvious overriding need is to raise the efficiency of water utilization. Where capital is short, the major requirement might be for an irrigation method with minimal de-pendence on capital investment or expensive equipment. In other cases, the deciding factor may be energy require-ments, labour availability or maintenance costs.

Since the economic considerations, along with the physical conditions and cropping patterns, are necessarily specific to each location, an irrigation system that may seem most appropriate in one country or region may not be so in another. In particular, it is a mistake to assume a priori that a modern system proved to work in an industrialized commercial economy will necessarily succeed in the context of an emerging economy.
The following sections describe and compare the various alternatives with respect to their possible applicability in developing countries, particularly in Africa. Physical factors generally involved in system selection include soils, crops, climate, topography, water quality and availability, drainage, field size and general system performance. Human factors include labour and management, training and skills. Economic factors include the costs of labour, capital and energy in relation to expectable returns. Not all of the relevant factors can be defined or weighed quantitatively in each case, so often the decision as to which system to select rests in part on subjective preferences rather than explicit analysis.

There is, altogether, no "best system" for various crops, soils and farm unit sizes. The aim should be not the "best system" but a spectrum of options that may be appropriate for the circumstances. The search for appropriate methods is necessarily guided and constrained by available knowledge as well as by local trial and error.
The first criterion for selecting and adapting one or another of the modern irrigation methods to the special needs and circumstances of developing nations in Africa is to reduce the capital costs associated with the installation of such systems. In the industrialized countries, commercial systems are designed to minimize labour requirements by substituting mechanical power and automation for human labour, and by enlarging the systems so as to achieve economy of scale. In many developing nations, the economic equation is reversed: labour is often more readily available while capital and fuel are in shorter supply. Farming operations are typically carried out by individual farmers or families who generally cannot afford major investments in machinery, especially if such machinery must be imported from distant sources. The appropriate irrigation systems for such farmers should be based, to the extent possible, on self-reliance - that is to say, on local materials and labour. The process of adaptation must also include a downscaling of the system so as to fit the size of a family holding, generally no larger than several hectares and often less than one hectare.
A wide spectrum of options exists for introducing irrigation methods consistent with the principles described. The range of possibilities includes, at one end, systems of water conveyance, distribution and application that can be fabricated entirely locally, of a sort that even small-scale subsistence farmers can adopt them and be self-sufficient in their maintenance. At the intermediate level are systems based in part on manufactured components, preferably of a type that can be fabricated by workshops or factories within each country or region. Only in the special circumstances where high-value cash crops can be produced in a well-developed market economy will systems relying entirely on imported equipment be justified.
In no case can blind acceptance be assumed of any technology or methodology designed and introduced entirely from the outside. Local trial and error (guided, to be sure, by sound basic principles) will be necessary, as systems must be proved in practice to fit the circumstances and preferences of their intended users. Local experience will evolve gradually and will take time to become local expertise. The region's own farmers should be involved from the outset and encouraged to participate and innovate. Local entrepreneurs may then develop the capability to improvise essential components and service irrigation systems.
There can be no short cut to the process of adoption and adaptation; it should not be rushed and must not be imposed from above. Rather, it should be nurtured by means of positive incentives. Extension services can provide information, demonstrations and guidance to farmers where needed, while financial institutions can offer them credit on favourable terms to invest in appropriate irrigation technology. Such technology will only be accepted if it produces adequate returns, that is to say, if its benefits clearly justify the costs. Since the benefits will depend in each case on marketing opportunities and other local factors, they cannot be predicted ahead of time by outsiders.
The HELPFUL (High-frequency, Efficient, Low-volume, Partial-area, Farm-Unit, Low-cost) irrigation methods described in this section can be divided into two broad categories: first, below-ground application methods, and second, above-ground application methods.

BELOW-GROUND APPLICATION METHODS

In this group of methods, water is applied directly to the root zone via porous or perforated receptacles that are embedded in the soil to some depth (from 10 to 50 cm), with their openings protruding above the soil surface. These receptacles, which are filled with water periodically or kept filled continuously, exude the water through their permeable walls into the surrounding soil. The moisture applied in this manner feeds the roots of the crop. When arranged in a grid, these embedded applicators produce a pattern of wetting that can be optimized with respect to the spacing and rooting habit of the crop thus irrigated.

The rate of infiltration and the distribution of moisture within the root zone also depend on the properties of the soil itself. For example, in a coarse-textured (sandy) uniform soil profile, the water would naturally tend to flow downwards, thus producing a carrot-shaped zone of wetting. On the other hand, in a fine-textured (clayey) or layered profile, more water would tend to spread laterally in the soil, thus producing an onion-shaped zone of wetting. If cylindrical porous containers are fitted to form a continuous tube that is embedded horizontally in the soil, they can constitute a line-source capable of irrigating an elongated bed. Soluble nutrients (fertilizers) can be injected into the water supply, to enhance the efficiency of fertilizer use as well as of water use by a row crop.
In principle, this type of irrigation can provide water steadily, as long as the receptacles contain water. The frequency with which they must be refilled depends on their capacity (the volume of water they can hold) as well as on the rate of water flow into the soil. The latter, in turn, depends on the permeability of the receptacle walls as well as on the rate of soil moisture extraction by the surrounding root system. If the applied water contains particulate matter (suspended sediment, either mineral or organic), or if it contains precipitable chemicals (such as calcium salts), these may eventually clog the pores of the receptacles. Algal or bacterial growth may also cause clogging. The remedy is to flush out the receptacles periodically with an acidic or fungicidal solution, and to replace them after some time (every few years).
In arid areas, where the upper zone of the soil is not leached by rains sufficiently, subsurface irrigation may cause salt accumulation at the surface, especially if the irrigation water contains an appreciable concentration of salts. Where this occurs, the topsoil must be leached each season by impounding water over the surface prior to planting time.

Porous ceramic jars

The use of soil-embedded porous jars is one of the oldest of the partial-volume, high-frequency (or continuous) irrigation methods. Although the origin and antiquity of the method cannot be established with certainty, numerous reports have attested to its use by traditional farmers throughout North Africa and the Near East (Figures 10 and 11).

FIGURE 10
The pattern of soil wetting around a single porous clay jar embedded between two crop rows

FIGURE 11
The pattern of soil wetting by a set of porous clay jars embedded between two crop rows

The method consists of placing porous clay jars (or pots) in shallow pits dug for this purpose. Soil is then packed around the necks of the jars so that their rims protrude a few centimetres above the ground surface. Water is poured into the jars either by hand or by means of a flexible hose connected to a water source. The jars used are generally made of locally available clay, so they are of no standard shape, size, wall thickness or porosity.
For best results, the jars should be fired at relatively low heat and without glazing, so they remain permeable. Trial and error experience should lead to the manufacture of jars with optimal properties of strength (to resist crushing), permeability (to exude water into the soil at a more or less steady rate), and size (to hold enough water for at least a one-day supply).
The clay jar irrigation method appears to be most suitable for fruit trees, but it can also be used for row crops. For young tree plantations, a single jar placed adjacent to each sapling should suffice initially. For example, if a single 5 litre jar wets a soil volume having an effective cross-sectional area of, say, 1 square metre, and if the rate of exudation is such as to empty the jar within one day, then the supply rate would be equivalent to 5 litres per square metre per day.
The pattern of lateral and vertical spreading of the water exuded from each jar depends upon soil texture and profile stratification. It may also depend on the shape of the jars (whether slender and long, or wide and shallow).
As each tree grows, its canopy covers a larger area and its roots tend to grow laterally and vertically to tap a larger soil volume. A mature fruit tree whose canopy covers a ground area of about 10 square metres may require roughly 30 to 50 litres per day during the dry summer period. To meet that requirement, the irrigator can place several jars in a circular pattern around the trunk of each tree. The porous jar irrigation method is flexible enough to permit adding porous jars gradually as the trees grow and the need arises for more water per day and for a larger volume of wetted soil.
The example given above is hypothetical, of course. The actual amount and rate of water application should be determined in each case in accordance with local experience. Careful observations and trials are needed to optimize the system's controllable variables.
The exposed openings of the jars may attract thirsty land animals as well as birds, and these may in turn damage the crop. For this reason, as well as to prevent clods of soil from falling into the jars and reducing their effective volumes, irrigators should cover the tops of the jars between waterings. This can be done simply by placing a stone over each jar.
The simplest but most laborious way to fill the jars with water is to do so manually, by using hand-carried buckets fitted with spouts. A more efficient way is to use a flexible hose connected to a water source. A still more labour-saving device for filling the jars is to set a narrow hose in place for the duration of the season, with perforations made over each jar. At appropriate intervals of time (daily or weekly, as the case may be) the hose can be connected to a water source so as to fill all the jars along the line simultaneously.
How long the jars last depends on several factors, including the rate of clogging by turbid water (containing suspended clay or organic matter), or by saline water. Acidity of the water as well as of the soil may also affect the durability of the jars, especially if the material from which they are made contains calcareous fragments. Careless trampling by humans or animals may crush the jars or fill them with loose earth. Simple though it is, the porous jar irrigation system must be monitored constantly if it is to be kept in continuous and satisfactory operation.

Porous and sectioned pipes

This is a variant of the porous jar method of irrigation, designed to spread water along a continuous horizontal band in the soil, rather than at discrete locations. As such, the porous pipe method is more suitable for closely spaced row crops grown in beds, such as vegetable crops. To allow water entry, the pipe is bent at one end, and the orifice is made to protrude above ground.

A good demonstration of the porous and sectioned pipe method of irrigation has been carried out by the British Institute of Hydrology in southeastern Zimbabwe, in cooperation with the Zimbabwe Ministry of Agriculture and Water Development. They use locally made clay pipes, approximately 24 cm in length and 7.5 cm in internal diameter, with a wall thickness of 2 cm. (These dimensions are arbitrary, of course.) The pipes are placed at the bottom of a shallow trench (about 25 cm deep) representing the centre-line of a 1 metre-wide bed, and are thus arranged to form a continuous horizontal tube, 3 metres long. The trench is then back-filled with earth.
To allow filling with water, an inlet is formed at one end of the pipe by tilting the first pipe section (the lower end of which was slanted during manufacture to fit the second, horizontal section). As the pipe sections are only abutted against one another but not sealed, water can leak into the soil at the joints between adjacent sections as well as through the porous walls of each section (Figures 12-14).

FIGURE 12
The pattern of soil wetting under irrigation by means of subsurface porous clay pipes: pipe sections are fitted to form parallel horizontal line sources for irrigating a row crop

FIGURE 13
The pattern of soil wetting by a horizontal porous pipe embedded between parallel crop rows

FIGURE 14
Planting a crop in rows directly above horizontal porous pipes

Experience shows that a single pipe, so arranged, can irrigate two rows of a vegetable crop, planted on each side of it. The amount of water applied is the equivalent of 6 to 8 mm per day during the growing season for a crop of rape. Okra and tomato were also grown successfully using this method of irrigation (Murata et al., 1995).

Perforated plastic sleeves

An interesting variant of the subsurface exuder method of irrigation is the use of thin plastic sheeting to form a sleeve-like casing. The chief advantage of the method is its low cost. However, the method has several distinct disadvantages that restrict the range of its applicability (Figure 15).

FIGURE 15
The pattern of wetting by a sand-filled plastic sleeve, perforated on one side and placed vertically in the root zone

Since the soft plastic material that serves for the making of a sleeve cannot retain its shape, the sleeve must be filled with sand before being placed in the soil. The sand filling reduces the capacity of the sleeve (i.e. the volume of water that it can hold) by some 50 to 60 percent. Moreover, the sand itself tends to retain a significant fraction of the moisture given it and to resist outflow. Thus the effective capacity is reduced still further.
Finally, since the plastic casing is essentially impervious (unlike the porous clay described above), it must be perforated. The need to optimize the diameter and density of the perforations introduces another variable into the system, the best solution to which must be established by trial and error. Too many perforations can weaken the plastic sheath and reduce its life span (which in any case cannot be expected to be as long as a clay jar or tube). In some cases, roots of the crop or of weeds may penetrate the perforations. As a consequence of all these factors, the ability of the sand-filled plastic sleeve to deliver water to the surrounding soil is limited, both in volume and in rate.
Notwithstanding these potential shortcomings, this method has been applied with apparent success to the growing of manioc and other crops in sandy soils in Senegal. To define its comparative usefulness better, however, the method should be tested side by side with alternative methods of irrigation. To date, this has not been done systematically.

Below-ground drip

A much more sophisticated and hence more expensive method of subsurface irrigation employs narrow plastic tubes of about 2 cm diameter. These are buried in the soil at a depth between 20 and 50 cm, deep enough so as not to interfere with normal tillage or traffic. The tubes are either porous throughout, or are fitted with regularly spaced emitters or perforations. If porous, the tubes exude water along their entire length. If fitted with emitters, they release water only at specific points. The water so released then spreads or diffuses in the soil. The pattern of wetting depends on the properties of the surrounding soil, as well as on the length of the interval between adjacent emitters and their discharge rates (Figure 16).

FIGURE 16
A line-source drip emitter with closely spaced perforations

A potential problem here is that the narrow orifices of the emitters may get clogged by roots, particles, algae or precipitating salts. Such clogging is difficult to detect as readily as when the tubes are placed over the surface in above-ground drip irrigation. Occasionally injecting an acidic or herbicidal solution into the tubes may help to clear some types of clogging, though the problem may recur periodically. Slit sections of plastic tubes may also be used to cover the emitter and thus inhibit clogging by roots without substantially reducing the discharge rate.
In underground drip irrigation, the delivery of water in the feeder tubes can be constant or intermittent. For uniformity of application, there should be some means of pressure control. If the lines are long or the land is sloping, there can be considerable differences in the hydraulic pressure and therefore in delivery rate, unless pressure-compensated emitters are used. Such emitters tend to be expensive, however.
Experience in Israel, California and elsewhere has shown that this method of subsurface irrigation is feasible in plantations of fruit trees and other perennial row crops. It may also be applicable to annual crops grown in regular beds.

ABOVE-GROUND APPLICATION METHODS

The methods described in this section are based on the steady or intermittent supply of water to a fraction of the soil surface. This is usually done by delivering the water in closed conduits (e.g. plastic tubes) to specific points, located and spaced in accordance with the configuration of the crop to be grown. At these points, the water is released on to the surface at a rate that, ideally, does not exceed the soil's infiltrability, so the water penetrates into the root zone without any of it either ponding or flowing over the surface.

Closed-conduit (piped) irrigation distribution systems are generally capable of saving water by increasing the uniformity of application and by avoiding losses of both quantity (resulting from seepage and evaporation) and quality (resulting from contamination of water in open channels). But because piped systems require pres-surization as well as costly installations, the water is saved often at the expense of increased energy consumption and capital investment. Methods are needed, therefore, that minimize those capital and energy costs.

Full-system drip

Drip irrigation is the slow localized application of water, literally drop by drop, at a point or grid of points on the soil surface. As long as the application rate is below the soil's potential intake, termed infiltrability, the soil remains unsaturated and no free water stands or runs over the surface.

Water is delivered to the drip points via a set of plastic tubes, generally weathering-resistant opaque polyethylene or PVC. Lateral lines, supplied from a field main, are laid on the surface. They are commonly 10 to 25 mm in diameter and are either perforated or fitted with special emitters. The latter are designed to drip water on to the soil at a controlled rate, ranging from 1 to 10 litres per hour per emitter.
The operating water pressure is usually in the range of 0.5 to 2.5 atmospheres. This pressure is dissipated by friction in flow through the narrow passages or orifices of the emitters, so the water emerges at atmospheric pressure in the form of drops rather than a jet or spray.
Commercial emitters are either in-line (spliced into the lateral supply tubes), or on-line (plugged on to the tubes through a hole punched into the tubing wall). Commercial emitters are precalibrated to discharge at a constant rate of 2, 4 or 8 litres per hour. The discharge rate is always affected by changes in pressure, but less so in the case of pressure-compensated emitters. The frequency and duration of each irrigation period are controlled by means of a manual valve or a programmable automatic valve assembly. Metering valves are designed to shut the flow automatically after a pre-set volume of water is applied (Figure 17).

FIGURE 17
A basic trickle irrigation system (schematic)

Water tends to spread sideways and downwards in the soil from the point where it is dripped. The fraction of the soil's total volume that is actually wetted depends on the density of the drip points (the grid), as well as on the rate of application and the internal water-spreading properties of the soil. The wetted zone, and hence the active rooting volume, is usually less than 50 per-cent of what would be the normal root zone if the entire soil were wetted uniformly.
Under frequent drip, the wetted portion of the soil is maintained in a continuously moist state, though the soil is unsaturated and therefore well aerated. This creates a uniquely favourable soil moisture regime. Drip irrigation thus offers a distinct advantage over flood irrigation and even over less-frequent sprinkle irrigation, especially for sandy soils of low moisture storage capacity and in arid climates of high evaporative demand. In contrast with sprinkle irrigation, drip is practically unaffected by wind conditions. Compared to surface irrigation, it is less affected by soil texture, topography or surface roughness.
If irrigation is applied in an amount that exceeds crop requirements, the wetted zone under each dripper becomes elongated downwards, and may eventually form a "chimney" draining the excess water beyond the reach of the roots (Figure 18).

FIGURE 18
The pattern of soil wetting under a drip emitter placed between closely spaced rows of a crop

With drip irrigation, it is possible to use somewhat brackish water (e.g. with a salt concentration of about 1 000 to 2 000 mg/litre) for the irrigation of crops such as cotton, sugar beet, tomatoes or dates that are not too sensitive to salinity. The brackish irrigation water does not come into direct contact with the foliage, which is therefore not so prone to salt-scorching as in sprinkle irrigation. Because the soil in the wetted zone is kept constantly wet, the salts are prevented from concentrating and the salinity of the soil solution in the rooting zone does not significantly exceed that of the irrigation water.
If the irrigation water is brackish, however, a fraction of the salts carried by the water tends to concentrate at the peripheries of the wetted circles, forming visible rings of salt around each drip point. In areas that receive appreciable seasonal rainfall, such salt rings are usually leached away annually.
Full-system drip irrigation can greatly reduce labour costs, but its successful operation demands constant supervision by skilled technicians with a ready supply of spare parts. It is certainly not a system that, once installed, can continue to operate trouble free by itself. Drip emitters must be inspected regularly and cleaned or replaced whenever any fail by clogging or mechanical damage.
Though the plastic tubing used in drip irrigation is weathering-resistant and flexible, it is vulnerable to kinking and cracking when bent or trampled repeatedly, as well as to puncturing by tillage implements, rodents and birds. Burying the tubes in the ground increases their longevity but makes them harder to inspect and to repair when they are damaged.
The most important aspect of drip irrigation maintenance is the prevention of clogging by suspended particles (silt), by biological organisms or their products and by chemical precipitation of salts. Algae and other biological slimes can be controlled by chlorination. Special care is needed where the irrigation water is drawn from open reservoirs that are turbid with silt and greenish with aquatic plants. Salts such as calcium carbonate can be prevented from precipitating by acidifying the water periodically.
Particles of various sorts can be removed from the irrigation water by means of screen filters, media filters (containing gravel, sand or diatomaceous earth) and centrifugal separators. Filters of one kind or another are, in fact, integral components of drip irrigation systems. Screen filters are rather delicate and require frequent inspection and servicing. Gravel and sand filters are less expensive, but tend to be bulky and to result in considerable loss of pressure. As the pores of the gravel or sand medium become clogged with retained solids or slime, pressure loss increases and flow rate diminishes, so these media require frequent back-flushing and periodic replacement.
The spacing between lateral tubes is determined by the spacing of the crop rows, as these tubes are generally laid alongside each row. In crops with closely spaced rows, it is often possible to economize in tubage requirements by using a skip-row arrangement or by placing a single lateral tube between a pair of close rows grown on a bed. This is not possible, of course, in the case of widely spaced shrub or tree crops. In principle, drip irrigation is most suited to orchard crops and to garden crops grown in rows and beds, and least suited to close-growing field crops requiring uniform wet-ting of the entire soil volume (Figure 19).

FIGURE 19
The pattern of soil wetting under drip emitters placed either side of a tree

The capital investment costs of drip irrigation systems are relatively high because large quantities of pipes, tubes, emitters and ancillary devices are necessary to control the precise delivery of water to specific sites in the field. Moreover, since standard drip-emitter orifices are narrow, expensive filtration equipment is necessary to prevent clogging. Hence drip systems tend to be more expensive, at least initially, than surface irrigation. Drip systems may prove to be economically justifiable in the long run if they can indeed prevent the waste of water and the degradation of land that is so frequent under traditional irrigation. However, to make drip irrigation more applicable to African conditions, ways must be sought to simplify the system and make it less expensive to install and operate.

Simplified drip

The highly sophisticated equipment developed to serve drip irrigation systems in the industrialized countries obscures the concept's essential simplicity. The main justification for such a capital-intensive and generally energy-intensive approach is to reduce the costs of labour. Since the relative costs involved in the promotion of irrigation for the developing countries of Africa are often the reverse of those in the industrialized countries, consideration must be given to simplifying drip irrigation systems. Efforts must be directed towards redesigning drip systems so as to facilitate installation and maintenance, while retaining the basic principles of high-frequency, high-efficiency and low-volume irrigation (Figures 20-24).

FIGURE 20
An on-line point-source emitter with a single dripper

FIGURE 21
An on-line emitter with multiple drippers

FIGURE 22
Section of an in-line emitter with capillary spiral flow path, and of an on-line (plug-in) narrow-orifice emitter

FIGURE 23
The patterns of spreading moisture under drip irrigation in sandy, loamy and clayey soils

FIGURE 24
A method of promoting the penetration of water into tight sloping ground under drip irrigation by means of a gravel-filled ring driven into the soil to a depth of several centimetres

Drip emitters need not necessarily be precision-fabricated. Instead, they can be improvised by punching holes manually in the lateral tubes. To make such perforations as uniform as possible, the use is recommended of standard round-edged cutters of the type used for leather belts. To prevent excessive outflow or blockage of the perforations, users can cover the holes with tight-fitting collars made by slitting short sections of the same tubage that is used for the laterals and slipping them over the holes. With trial and error experience, a user can make adequate emitters for a fraction of the cost of commercial emitters. Moreover, such emitters are easy to service, i.e. to clean or unclog whenever necessary. Another way to make emitters is to insert sections of microtubes into holes punched in the lateral tubes, then adjusting the microtube length to provide the desired discharge rate (Figures 25 and 26).

FIGURE 25
Making a simple drip emitter by perforating a plastic tube and covering the perforations with a sleeve cut from the same tube

FIGURE 26
Making a simple drip emitter by tightly inserting a microtube of adjustable length into a lateral hose line

Hydraulic pressure in the delivery lines need not be created by means of mechanical pumps. Elevating the reservoir just a few metres above the land to be irrigated may create a gravitational head sufficient for drip-irrigating a small area. Larger-diameter tubes and wider emitter orifices, as well as longer durations of irrigation, can compensate for the lower operating pressure. The need for precision pressure-regulators is thereby obviated, especially where the land is fairly level and the laterals are not too long or narrow.

Filtration can be accomplished by interposing a simple sand-filled container between the source of the water and the irrigation lines. Incoming (turbid) water can be introduced at the bottom of the container and made to flow upwards through the layered sand, so that the filtered water collects on top and overflows into the irrigation lines. Such a filter can be assembled locally, using either a metal or a plastic container of whatever size is found to be adequate for the flow rate and the turbidity of the water. The sand to be used should be pre-washed to remove the finer particles, and it should be rewashed or replaced at regular intervals as it gradually tends to clog.

Measurement of flow is an essential requirement of efficient water use. Where a system is not equipped with flow meters or metering valves, the flow must be monitored by recording the duration of each irrigation. The volume of discharge per unit time should be checked and rechecked periodically, as should be the uniformity (or variability) of emitters within each lateral line and of the lines within the field. This can be done by recording the time needed for the discharge to fill a vessel of known volume. The volume of water in each irrigation application should conform to the estimated irrigation requirement for the crop, given its stage of growth and weather conditions (rainfall and evapotranspiration since the previous irrigation).

Microsprayer

Microsprayers, also called mini-sprinklers or spitters, are similar in principle to drip systems in that water is applied only to a fraction of the ground surface. However, instead of dripping water from narrow-orifice emitters, microsprayer systems eject fine jets that fan out from a series of nozzles. Each nozzle can water an area of several square metres, which tends to be much larger than the individual areas wetted by single drip emitters. Microsprayers can thus help to enlarge the volume of soil available for the uptake of water and nutrients by crop roots (thereby obviating the need for multiple drippers). Enlarging the wetted volume is especially important for large trees (Figure 27).

FIGURE 27
The pattern of soil wetting by a microsprinkler

Another significant advantage of microsprayers over drip systems is that, thanks to the larger nozzle orifices and the greater rate of discharge, the hazard of clogging is reduced and the filtration requirements are not as stringent as in the case of drip irrigation. For this reason the installation costs may be somewhat lower. The pressure requirements, however, remain in the order of one to two atmospheres - lower than those for regular sprinklers but still requiring pumpage or a commanding reservoir elevation of 10 metres or more.
In other respects, microsprayer irrigation retains the potential benefits of drip irrigation: it permits high-frequency, low-volume irrigation as well as the injection of fertilizers into the water supply. Moreover, microsprayer systems can be scaled down readily to accommodate the small irrigation units prevalent in developing countries.
The disadvantages of microsprayer irrigation relative to drip irrigation must also be considered. The evaporation component of the water balance is increased because of the larger wetted area of ground, the spraying of water into the dry air and the wetting of the lower foliage of the crop. Because of the wetting of leaves, the use of brackish water and the incidence of fungal diseases can be more problematic with microsprayer irrigation than with drip.
Microsprayer systems are served by the same tubing network as drip systems. A wide variety of emitter units, generally made of durable plastic materials, is now available commercially. Such spray nozzles are harder to improvise, however, so the irrigator in this case must depend on manufactured components more than in the case of the simplified drip system described above.

Low-head bubbler

Bubbler irrigation is a partial-area, low-volume, high-frequency irrigation method based on closed-conduit delivery. It is designed to reduce investment and energy requirements by using inexpensive, thin-walled, corrugated plastic pipe of sufficient diameter that even the limited pressure available from a low-head surface reservoir might suffice. Bubbler irrigation is essentially a modification of drip irrigation, intended to make the system less dependent on industrially produced components (Figure 28).

FIGURE 28
The pattern of soil wetting by a low-head bubbler

In bubbler irrigation, no manufactured emitters of any kind are used, and the water is simply allowed to "bubble out" of open vertical tubes. This does away with the need for filtration, which is a major problem in drip irrigation. The vertical bubbler tubes (called risers or standpipes), roughly 1 to 3 cm in diameter, are connected to buried lateral irrigation tubes having a diameter of at least 10 cm. The bubblers are anchored to stakes or posts and their heights are adjusted up or down, by calculation or by trial and error, so as to deliver water at the desired rate.
Bubbler systems are particularly suited to the irrigation of widely spaced crops, like fruit trees or shrubs, in which a standpipe bubbler can be installed alongside each tree or group of shrubs. The irrigation water delivered by each bubbler is distributed uniformly by filling small level basins, surrounded by low ridges, with equal quantities of water. Such basins can be constructed by hand and may be either circular or rectangular. By the simple means described, the principles of efficient irrigation can thus be implemented.
Because of its simplicity and the absence of standardized manufactured components (such as nozzles, fittings, pressure regulators and filters), bubbler irrigation has not been promoted as a commercial product by equipment salesmen. Perhaps this is the reason why so many potential users are unaware of its advantages, including its low cost and its ease of installation and operation.
A procedure for installing and calibrating bubbler systems was described nearly 20 years ago by Rawlins (1977). Since that time, experience by the author of this publication and others has proved the system to be practical. Such systems, or variations of them, can serve as attractive options for tree crops, particularly on relatively level lands that can be converted from rain-fed farming or from traditional surface irrigation methods.

Fertilizer injection

Many soils in Africa are inherently low in fertility. Soils of the humid tropics tend to be highly leached and in places exhibit acidity, as well as aluminium or sulphate toxicity. Soils of the arid subtropics are typically coarse-textured and have low organic matter content. Such soils often require chemical amendments, manuring or fertilizing if they are to provide the higher yields needed for food security.

Conventional methods of applying fertilizers, as by broadcasting uniformly on the surface or by drilling a continuous band of fertilizer alongside the row crop, are not compatible with partial-area or partial-volume irrigation. For best results, the spatial distribution of the fertilizer in the soil should correspond to the distribution of the water.
Where water is applied only to a fraction of the soil volume, crop roots concentrate in the wetted portion of the soil. It is important, therefore, to ensure that the restricted rooting zone be endowed with the nutrients essential for crop growth. Surface application of dry fertilizer may not ensure optimal placement, especially in the case of below-ground irrigation methods. Experience has shown that fertilizer-use efficiency, as well as water-use efficiency, are enhanced when the nutrients are applied in the irrigation water.
The combined application of water and fertilizers has come to be known as fertigation. As such, it is a particular variant of the more inclusive concept of chemigation, by which different agrochemicals are introduced into the rooting zone in solution form via the irrigation system. Among the other types of chemicals similarly applied are selective herbicides to suppress weeds, fungicides to control fungal diseases and nematocides to protect crop roots against parasitic nematodes.
In closed-conduit irrigation systems, fertigation can best be accomplished by means of a fertilizer injection tank connected to the main line (Figure 29). A fertigation unit is relatively easy to assemble. It requires no specialized equipment, merely a container of appropriate volume (20 to 100 litres), preferably of corrosion-resistant material, through which the water supply is made to flow. The container should have a wide opening for pouring in and mixing the fertilizer and a watertight seal for it. For systems requiring filtration, such as drip or microsprayer, the fertilizer tank should precede the filter so that any insoluble particles originating in the tank are prevented from clogging the emitters.

FIGURE 29
A fertilizer-mixing tank for injecting soluble nutrients (fertigation) into a closed-conduit irrigation system

Of the essential plant nutrients, the one most often deficient is nitrogen, whose mineral forms (e.g. ammonium sulphate, ammonium nitrate, potassium nitrate and urea) are generally readily soluble. Applications of nitrogen often result in dramatic bursts of foliar growth and greening, especially in plants growing on leached soils of low organic matter content. However, crops given only nitrogen may soon exhibit deficiencies of the other major elements (phosphorus and potassium), as well as of several minor elements.
Potash, when required, is also available in soluble formulations, including potassium chloride, sulphate or nitrate. Fertilizers containing phosphorus may need to be acidified to make them readily soluble. In tropical soils of very low fertility, deficiencies of minor elements may call for foliar application by spraying.

Subirrigation by groundwater control

Subirrigation is the supply of water to the root zone of crops by artificially regulating the groundwater-table elevation. The method can work where the water-table is naturally high, as it frequently is along river valleys or in plains underlain by impervious strata (Figure 30).

FIGURE 30
Raising or lowering the water-table for subirrigation, by controlling the level of water in parallel ditches

Open ditches are usually dug to a depth below the water-table, and the level of the water is controlled by check dams or gates. In this manner, the ditches can serve either to drain excess water and thereby lower the water-table during wet periods, or to raise the water-table during dry periods and thereby wet the root zone from below.
The disadvantage of open ditches is that they interrupt the field and interfere with tillage, planting and harvesting. They also take a significant fraction of the land out of cultivation. An alternative is to place porous or perforated pipes (now generally consisting of corrugated plastic tubes) below the water-table, with controllable outlets. When open, the outlets serve as drains; when closed, they allow the water-table to rise. Subsoil pipes are, however, more expensive to install and more difficult to maintain, as they tend to clog with soil or with precipitated iron oxide.
Subirrigation may be used for field crops and pasture, as well as orchards. It is best suited to hydrophilic crops such as sugar cane and dates. The uniformity of irrigation depends on how level the land surface is and how uniform the soil.

Precise control of shallow groundwater is a delicate and difficult task, and it involves some serious hazards. The optimal depth of the water-table is some 30 to 60 cm below the root zone. A higher water-table tends to waterlog the soil, restrict aeration and cause capillary rise and evaporation at the surface, where salts can accumulate. On the other hand, keeping the water-table too low may deprive the crop of essential moisture. As the crop grows, its rate of moisture extraction increases and its root system extends downwards, so the water-table tends to fall, unless it is purposely maintained at a high level.
Since the water source is below the root zone, the supply to the roots occurs by capillary action. Hence the operation of the system depends on the sorption characteristics of the soil. A fine-textured (clayey) soil tends to become waterlogged and to restrict aeration. Clay soil also slows the flow of water both in subirrigation and in drainage. Such a soil requires closer spacing of the ditches or of the underground pipes. On the other hand, a coarse-textured (sandy) soil may retain too little water and tend to dry out excessively. As in other modes of irrigation, there can be no substitute for local experience in water control, based on knowledge of the specific soil conditions and crop requirements.