Optimal Irrigation: Considerations for Semiarid Regions
Irrigation is required for high crop productivity when precipitation is inadequate to meet the crop evapotranspiration (ET) demand. The soil provides a storage capacity for water from which the crop withdraws water. The overall quantitative storage capacity depends on both the type of soil and the rooting depth of the crop plants. Soils whose texture and structure result in large pore sizes have lower storage capacity than soils with smaller pore sizes. More deeply rooted crops have a higher storage capacity as compared to crops whose roots are shallower. In principle, the amount of irrigation applied should provide for recharge of the soil storage capacity between irrigation events. The time to irrigate is before the soil has become too dry to allow good crop production. The combination of a soil and crop rooting system that have a low storage capacity dictates the use of smaller-quantity, more frequent irrigations, whereas larger storage capacity of soil and crop rooting systems allow less frequent though higher quantity irrigations.
The salinity of the irrigation water affects irrigation management. Plants transpire pure water, causing the soil solution to become concentrated with salts as transpiration proceeds. Because of this effect, irrigation amounts must not only recharge the storage capacity, but additional water may be necessary to leach excessive salts from the root zone.
Irrigation uniformity adds complexity to irrigation management. A uniform irrigation is one in which the same amount of water infiltrates the soil at all points in the field. Most frequently, the amount of water that infiltrates into soil is variable for different parts of the field. Nonuniform irrigation creates a dilemma. If irrigation is programmed to restore the storage capacity in the parts of the field that receive the most water, the other parts of the field will be underirrigated,
causing yield reduction in those underirrigated areas. Conversely, if irrigation is programmed to recharge the storage capacity for those zones that have the lowest infiltration rate, the other parts of the field will be excessively irrigated leading to unrecoverable water lost to deep percolation. Uniformity of irrigation, therefore, is one of the most critical factors affecting irrigation management.
Irrigation systems can be broadly categorized as being pressurized or nonpressurized. Pressurized systems are those in which water is delivered through a pipe, under pressure, and discharged by one of a variety of different outlet designs, including sprinkler heads and drip emitters. Nonpressurized systems are those in which water is delivered in an open channel and allowed to flow across the field. Various configurations for flow across the field, such as furrows or basin borders, are possible. The advantage of pressurized systems is that the quantity of water applied to the field can be precisely controlled by valves. Furthermore, if the system is properly designed, all the water infiltrates into the soil with no runoff from the field.
Although the amount and timing of water delivery to a nonpressurized system can be controlled, the amount of water that actually infiltrates the soil depends on the soil properties, among other conditions. Although the irrigator has some design control features for nonpressurized systems, such as length of furrow, rate of water discharge, and time (duration) of water application, the quantitative control is limited. Runoff from the field is usually an unavoidable condition for nonpressurized systems. Water must be maintained on the soil surface at the lower end of the field to allow adequate infiltration. During that period of time water is flowing off of the field.
Properly designed and maintained pressurized systems can deliver water very uniformly across the field. However, if water is emitted into the air, such as through sprinkler systems, the wind currents can greatly affect the distribution of water. Therefore, although sprinkler systems can theoretically be designed to be very uniform, wind can cause a very nonuniform distribution. Pressurized systems that emit water without spraying into the air, such as drip systems, allow for very uniform irrigation.
Nonpressurized irrigation systems have two sources of nonuniformity. First, water is on the soil for a longer period of time at the top end of the field as compared to the lower end of the field. This provides the opportunity for more water to infiltrate at the top end as compared to the lower end of the field. The term for this kind of nonuniformity is opportunity time nonuniformity. Variability of soil infiltration rate across the field due to textural or structural changes also leads to nonuniform infiltration across the field. The total nonuniformity, therefore, is a combination of the opportunity time and the soil variability.
Accurate measurement of irrigation uniformity is important in developing
the optimal management scheme for a given irrigation system. Unfortunately, measurement of uniformity is complex. Uniformity of sprinkler systems is measured by distributing containers throughout a collection area and measuring the amount of water collected in each container. The data are then statistically analyzed for variability. The numerical result depends on the size of the container. Using larger containers will result in a higher uniformity value than using smaller containers under the same conditions. Thus, the value is already recognized as being somewhat subjective based on the measuring technique. Even a drip system is very nonuniform if the measurement is made on a very small scale. The amount of soil water is high at the location adjacent to where water is released from the emitter, and the amount of soil water between the emitters is very low. Nevertheless, depending on the type of plant, the plant root system can integrate differences in soil water content in different parts of the root zone and even things out. This factor raises an additional point concerning uniformity. The plant root system can accommodate uneven water distribution and can extract more water where the soil water content is high, as long as the scale of the uneven soil water distribution is proportional to the scale of the plant root system. A tree with a large root system can accommodate considerable nonuniformity of water application under the canopy. A shallow-rooted vegetable crop would be more highly impacted by the same distribution.
The nonuniformity for surface systems is determined by measuring the rate of advance of water down the furrow, and then inserting these numbers into equations developed to compute the nonuniformity associated with opportunity time. These measurements do not include the nonuniformity associated with soil variability, which can be considerable. Therefore, the numerical values for furrow systems are overestimates of the true uniformity.
As long as the measurement procedure for a given irrigation system is consistently used, the comparative uniformity of different fields for that particular system can be determined. In other words, the uniformity from one sprinkler system can be compared to a different sprinkler system. However, it is not appropriate to compare a uniformity coefficient that has been measured for a furrow system to a sprinkler system.
With all of the factors considered, a drip system has better uniformity than a sprinkler system, which is better than a nonpressurized system. However, the costs for these systems are in the reverse order. The improved performance from a drip system compared to a pressurized system may not justify the costs for the drip system. Furthermore, pressurized systems require an energy supply that may not always be present in the field.
PLANT RESPONSE TO WATER
A major portion of the sun’s energy striking leaf surfaces is dissipated by evaporation from plant surfaces (transpiration). As long as the plant’s roots can
extract water from the soil and transfer it to the leaf surfaces at a rate equal to the transpiration rate, the plant can grow at its potentially highest rate. If the rate of transfer of water from the soil through plant tissue to the leaf surface is less than the plant’s loss of water by transpiration, the plant responds by closing stomata to reduce the water loss. Carbon dioxide (CO2) used for plant photosynthesis enters the plant through these same stomata. Therefore, closure of the stomata to reduce water loss also reduces CO2 intake and, therefore, reduces the rate of photosynthesis. Thus, the plant has a dual mechanism for protecting itself under a limited water supply. It reduces transpiration, thus attempting to maintain turgidity. At the same time, it reduces photosynthesis, which would otherwise increase the plant surface area through growth, leading to greater interception of energy and more transpiration in a continuing cycle.
Numerous studies have reported that total dry matter production in plants is linearly related to evaporation (ET). This observed result is consistent with the mechanisms that the plant uses to protect itself against inadequate water. However, the marketable product of some crops is not linearly related to total dry matter production. Therefore, the relationship between production of the marketable product and ET must be established independently for individual crop species.
WATER USE EFFICIENCY AND WATER CONSERVATION: DEFINITIONS
Water use efficiency and water conservation are commonly used terms in irrigated agriculture. However, confusion can arise because these terms can have multiple definitions. Water use efficiency is calculated as the ratio of two terms. Yet, different measurements may be used for the two terms in the ratio, resulting in different numbers, and yet all are referred to as water use efficiency. For example, water use efficiency can be defined as the ratio of beneficial water use to applied water. However, beneficial use sometimes is represented by ET, and at other times it is represented by ET plus the amount of water required for leaching salts from the root zone. Some individuals include all of the water delivered to the field as applied water, whereas others might only consider the infiltrated water that would be available for crop use. Obviously, different numbers result for any combination of these terms. Furthermore, the computations can be made on different land area sizes. For example, the ratio can be calculated for a field, the total farm, or the total basin. Different numbers result depending on which is selected. Possibly the biggest problem, however, is the common belief that a higher efficiency number is always better than a lower number. As will be discussed later in this paper, this is not usually the case.
Water conservation likewise is subject to different definitions. One definition is to use less water. This can be accomplished by various means, each with a specific consequence. A farmer can use less water by not growing a crop. Or, a
farmer can grow a crop and apply a small amount of water resulting in very low crop production. Farmers can also grow a crop for high production and either eliminate runoff or capture runoff and use it as part of the irrigation supply.
The main point is that in using these terms, the definition must be clearly specified and the consequences of the action properly evaluated. A further complication is that all of the definitions are based on water quantity without reference to water quality, and water quality cannot be ignored in a management scheme.
CROP-WATER PRODUCTION FUNCTIONS
Agricultural production is a business operation, and irrigation management must be evaluated in context of the business. The goal of any business is to maximize profits. Maximizing profits can include sustaining the business through a period when profits are not possible and generating growth in anticipation of future profits. Therefore, one definition of optimal irrigation management is that management which maximizes profits.
Water is an input to the production system. The functional relationship between crop yield that is marketed and the amount of water applied must be known for the economic analysis. Only water that infiltrates into the soil has an opportunity to contribute to crop production. The water running off the field cannot contribute to crop production, so the crop-water production function can only be based on the amount of infiltrated water. The runoff water has economic implications that must be accounted for separately.
The uniformity of irrigation significantly affects the crop-water production function. The relationship between cotton lint yield and the amount of infiltrated water for various irrigation uniformities is presented in Figure 1 for climatic conditions of the San Joaquin Valley of California. The uniformity is characterized by the Christiansen uniformity coefficient (CUC), where a value of 100 represents perfectly uniform irrigation and decreasing values of CUC represent increasing nonuniformity. Also depicted in Figure 1 is the amount of water that would percolate below the root zone for the various irrigation uniformities and infiltrated water amounts.
Considering the uniform irrigation first, crop yields increase with increasing amount of infiltrated water until a maximum yield is achieved and additional applied water does not contribute to more production. When the irrigation is nonuniform, more average water must be infiltrated for the field to get the highest yields. For a given amount of infiltrated water, the yield decreases as the irrigation uniformity decreases.
Under uniform irrigation, no deep percolation occurs until enough water has been applied to reach maximum yield. In other words all of the water applied is used by the crop. After maximum yield has been achieved, any additional applied water goes directly to deep percolation. In contrast, for nonuniform irrigation
some deep percolation occurs before a maximum yield can be achieved. This is a consequence of some parts of the field having more water than necessary for maximum crop yield and other parts of the field having less water than required. In all cases, increasing water application increases the amount of deep percolation. For nonuniform irrigation there is a tradeoff between irrigating for high crop yield and low deep water percolation.
The salinity of the irrigation water is another factor that affects the crop-water production function. The salts in the irrigation water become concentrated through evapotranspiration and, therefore, some water to leach the excess salts
from the root zone must be applied. The amount of water to be applied, however, depends on the salinity level of the irrigation water and the crop sensitivity to salinity.
Letey et al. (1985) developed a model to compute relative crop yield in relationship to the seasonal applied water for crops with different tolerance to salinity. The crop-water production functions are presented in Figure 2 for corn, which is a salt sensitive crop, and for cotton lint, which is a salt tolerant crop. The applied water (AW) assumes that all of the water infiltrates, and Ep is the pan evaporation. The ratio AW/ Ep may be used to facilitate comparisons between different climatic zones with different Ep values.
Note as the salinity of the irrigation water increases, more water must be applied to get the same yield. Irrigation water salinity may get to a level where maximum yield cannot be achieved regardless of the amount of water applied. Larger differences in yield for a given amount of applied water, or larger differences in applied water for a given yield, occur for the salt sensitive corn than for the salt tolerant cotton. Indeed for cotton, irrigation water salinities up to 4 deciSiemens per meter (dS/m) require relatively small amounts of additional water to achieve the maximum yield.
All of the curves depicted in Figure 2 assume that the irrigation is uniform. Nonuniform irrigation would modify the results in a manner as depicted in Figure 1. In other words, increasing nonuniformity would cause each of the curves on Figure 2 to be lowered for a given water application.
ECONOMIC IRRIGATION EFFICIENCY
Economic irrigation efficiency will be defined at the farm level as the irrigation management that maximizes profit. In a broader context, economic irrigation efficiency could be defined as irrigation management that maximizes net social benefits. The difference between the two definitions is the result of externalities. An externality arises when some of the costs or benefits of irrigation agriculture accrue to society as a whole and the costs (as reflected in market prices) are not borne by the farmers or the consumers of their products. Externalities can be positive or negative. An example of a positive externality occurs when water purchased by a farmer runs off his farm and serves some beneficial societal use. However, if the water is polluted, it can impose a cost to society and create a negative externality.
The crop-water production function as depicted in Figures 1 and 2 can be converted to benefit curves by multiplying the yield and the market price for the crop. Such curves are depicted in Figure 3 for a hypothetical case representing two irrigation uniformities. The total benefit (TB) in $ ha–1 for a given infiltrated water (IW) is higher for the more uniform (TB1) than the less uniform irrigation (TB2). The total cost of water (TC) is also depicted in Figure 3 for two water prices, where the price of water for case 1 (TC1) is greater than for case 2 (TC2).
The highest profit is achieved where the difference between TB and TC is the greatest. These points are identified by arrows in Figure 3. Some general conclusions can be derived from the information depicted in Figure 3. The economically optimal (profit maximizing) level of IW depends on the shape of the crop-water production function and the price of water. Improving the uniformity of irrigation results in a decrease in the value of IW to achieve economic efficiency. Also, raising the price of water lowers the value of IW to achieve economic efficiency. However, note that raising the price of water has a greater effect on decreasing the economically efficient IW value under the nonuniform irrigation system as compared to the more uniform irrigation system. Indeed, raising the price of water had relatively little effect on changing the economically efficient level of IW for the most uniform system.
A shift in technology or management to achieve more uniform irrigation usually imposes a cost. The increased cost must be offset by the increased benefits associated with improved irrigation uniformity to justify the investment. This factor must be evaluated on a case-by-case basis.
The economically optimal irrigation under nonuniform irrigation results in much more deep percolation than for the more uniform irrigation. Unless the
benefits or costs associated with deep percolation are borne by the farmer, they become externalities. With externalities, the economically optimal management from the perspective of the farmer may not be economically optimal from a social perspective. Because externalities can be either positive or negative, and the magnitude varies based on individual situations, it is not possible to make general statements concerning the comparative economic efficiencies of the farmer vs. society in general.
Letey et al. (1990) did an economic analysis of irrigation systems for cotton production in the western San Joaquin Valley of California, where high water tables require the installation of a drainage system. In this case the drainage waters are highly concentrated with salts and selenium, and there is a cost associated with their disposal. If the cost for drainage water disposal was not borne by the farmers, the cheaper nonuniform irrigations were determined to be most profitable for the farmer (Figure 4). However, if there was a constraint on the amount of drainage water generated, or a cost for disposal imposed on the farmer, investment in more expensive uniform irrigation systems was justified.
The average benefit (AB) as a function of IW is also depicted in Figure 3 for the two irrigation uniformities. AB is calculated as TB divided by IW and has the units of $ (ha cm)–1. AB is the average dollar return per hectare-centimeter
(ha cm) of IW and is not affected by the cost of water. Note that the maximum AB value occurs at a lower IW value than the economically optimal quantity. Even though the average return decreases, the cost for additional irrigation water is exceeded by additional returns associated with an incremental increase in water. The economically optimal input is at the point where the marginal benefit equals the marginal cost. The marginal benefit is defined as the benefit associated with the next incremental increase in input, and likewise the marginal cost is the increase in cost associated with the next incremental input. The marginal benefit is the slope of the TB curve, and the marginal cost is the slope of the TC curve depicted in Figure 3. Note in Figure 3 that the slope of the TB curve equals the slope of the TC curve for the optimal irrigation.
The shape of the AB curves in Figure 3 is identical to the shape of the ratio of yield (Y) to infiltrated water (IW). This ratio (Y/IW) is one definition of water use efficiency. Note that the IW that achieves maximum water use efficiency by that definition is not the IW that is economically efficient. Irrigating to achieve the maximum commonly defined water use results in a significant reduction in yield. Clearly, maximizing water use efficiency by this traditional definition can lead to results that are not the most economically beneficial.
A shift in production function from less uniform to more uniform irrigation does result in a higher water use efficiency for a given value of IW. Therefore, increasing the numerical value of water use efficiency by a change in management that entails a change in production function is positive. However, it is not necessarily economical. It is not obvious that the shift in production function from the less uniform to more uniform irrigation is economically efficient. The main conclusion is that generalizations cannot be made and each situation has to be thoroughly evaluated from production and economic considerations.
Irrigation scheduling refers to the time duration and quantity of an irrigation. Although the crop water production functions depicted in the Figures 1-2 provide the scientific and economic basis for optimizing irrigation, farmers do not necessarily have such complete detailed information available to guide their irrigation management. Nevertheless, the general principles still apply. Since the purpose of irrigation is to replace the water lost from the storage zone between irrigations, knowing the amount of ET that has occurred since the last irrigation is important. Alternatively, the farmer could monitor the soil water content as a function of time to determine when the soil is sufficiently dry to warrant recharge. Therefore, a method of monitoring either ET or the change in soil water content is required for irrigation scheduling.
Climatic conditions drive ET. Therefore, monitoring the potential ET by an evaporation pan, or determining potential ET from other climatological data, is required. In California, several weather stations have been established through-
out the state to form a California Irrigation Management Information System (CIMIS). Farmers with computer systems can get daily information from the weather station located nearest to their farms.
The climatological data identify the potential ET that can occur. Crop ET is not always equal to potential ET. For example, during the early part of the season for row crops the plant is small and the crop ET is much less than the potential ET. As the crop grows and its canopy cover increases, the crop ET approaches potential ET. Thus, to estimate the crop ET the potential ET is multiplied by a crop coefficient (Kc), which must be empirically determined for each crop as a function of time. Studies have established the crop coefficients for several crops in California. Results from these studies can be used to guide irrigation management. Nevertheless, the study results are not absolute, and the farmer must use judgment and make observations in the field to assure that his irrigation is appropriate.
Monitoring the soil water content as a function of time requires instrumentation. The neutron probe can be used to measure the water content in a soil profile, but this method is labor intensive. It requires the installation of neutron probe access tubes and then measurements on some predetermined schedule. Other instruments, such as tensiometers, can be installed at various depths and require reading on a timely basis. Some of the instruments have electrical signals that can be connected to a recorder for continuous monitoring with minimal labor input. Soil moisture monitoring requires capital investment and then some level of operational expense.
Water reuse implies that the irrigation water had previously been used. It may have been used in an urban setting and discharged as wastewater from a sewage treatment plant, or it may have been previously used for irrigation and discharged as drainage waters. The important factor in determining its suitability for irrigation is the chemical and biological composition of the water and not the fact that it had previously been used. Agricultural drainage waters and sewage waters generally have increased salinity levels that must be considered in irrigation. However, the basic principles presented earlier for using waters of different salinity apply under these conditions.
Sewage waters may have health implications. Regulations on reuse of water for protecting health will be specified by public health agencies and dictate when and how these waters may be used for irrigation. However, they do not affect the basic irrigation principles that have been presented. An irrigator simply needs to consider the chemical composition of the waters, along with the other basic irrigation principles, in guiding his use of reclaimed water for irrigation.
Many factors contribute to the selection of the irrigation system and management. The scientific and economic principles are well established and can be used to select the optimal set of options. However, the optimal solution is case-specific and cannot be specified in a general manner.
Letey, J., A. Dinar, and K.C. Knapp. 1985. Crop-water production function model for saline irrigation waters. Soil Sci. Soc. Am. J. 49:1005-1009.
Letey, J., A. Dinar, C. Woodring, and J.D. Oster. 1990. An economic analysis of irrigation systems. Irrig. Sci. 11:37-43.