Yield response factors of field crops to deficit irrigation

P. Moutonnet,
International Atomic Energy Agency,
Joint FAO/IAEA Division,
Vienna, Austria

Summary

Water is essential for crop production, and any shortage has an impact on final yields. Therefore, farmers have a tendency to over-irrigate, an approach that runs counter to the conservation of scarce resources. At present, owing to the global expansion of irrigated areas and the limited availability of irrigation water, there is a need to optimize WUE in order to maximize crop yields under frequently occurring situations of deficit irrigation. When water deficit occurs during a specific crop development period, the yield response can vary depending on crop sensitivity at that growth stage. Therefore, timing the water deficit appropriately is a tool for scheduling irrigation where a limited supply of water is available. A standard formulation relates four parameters (Y_a, Y_m, ET_a and ET_m) to a fifth: k_y, the yield response factor, which relates relative yield decrease to relative evapo-transpiration deficit. Two series of k_y values obtained from FAO data sets and from an International Atomic Energy Agency (IAEA) coordinated research project (CRP) showed a wide range of variation for this parameter 0.20 < k_y < 1.15 (FAO), and 0.08 < k_y < 1.75 (IAEA). The two data sets, whilst showing the same trends, gave neither identical average values for k_y nor similar ranges of variation.

Water is a finite resource for which there is increasing competition among agricultural, industrial and domestic sectors. According to Kemp (1996), in Mediterranean countries, "The World Bank argues that the allocation of water to agriculture, which accounts for about 90 percent of regional water use, no longer makes economic sense... In Morocco, for example, it is estimated that the value added by a cubic meter of water in irrigated agriculture is a mere 15 cents; used in industry it is a striking $25. In Jordan, which uses highly efficient drip irrigation for over half of its irrigated agriculture, the equivalent figures are 30 cents for agriculture and $15 for industry." Therefore, there is an urgent need to maximize crop yields under conditions of limited water supply. Kang et al. (2000) have shown that regulated deficit irrigation at certain periods during maize growth saved water while maintaining yield.

The upper limit for yield is set by soil fertility, climatic conditions and management practices. Where all of these are optimal throughout the growing season, yield reaches the maximum value as does evapotranspiration. Any significant decrease in soil water storage has an impact on water availability for a crop and, subsequently, on actual yield and actual evapotranspiration. A standard formulation (Vaux and Pruitt, 1983) relates these four parameters to a fifth: the yield response factor, which links relative yield decrease to relative evapotranspiration deficit, as follows:

    (1)

where:Y_a=actual yield (kg/ha)
Y_m=maximum yield (kg/ha)
ET_a=actual evotranspiration (mm)
Et_m=maximum evapotranspiration (mm)
k_y=yield response factor

Calculations of Y_m, ET_m and ET_a are well documented (FAO, 1977; and FAO, 1998) and the literature has provided values for k_y (FAO, 1979). From these four parameters, it is possible to calculate Y_a where the available water supply does not meet the full moisture requirements of the crop. Where water deficit occurs during a specific growth stage, the yield response will depend on crop sensitivity during that period. Therefore, the timing of the deficit is a tool for scheduling the use of a limited water supply and in setting priorities among several irrigated crops. As an example, the World Meteorological Organization recommended the utilization of Computer-Aided Learning (CAL) software in meteorology (Bell, 1994), especially the French educational computer program BILHY (Bilan Hydrique). BILHY is useful for training extension workers and meteorologists in the subject of soil moisture; the user has to decide, according to pedological, agricultural and meteorological parameters, whether or not to irrigate.

FAO has facilitated the calculation of crop water requirements and irrigation planning through a series of technical papers (FAO, 1992; and FAO, 1993). Nevertheless, the process is still difficult and requires several data sets. Another approach is based on field experiments on crops exposed to deficit irrigation, with soil moisture status monitored using the soil moisture neutron probe (SMNP) and sets of tensiometers (Vachaud et al., 1978). The SMNP is useful for assessing the soil hydraulic conductivity versus the soil water content throughout the internal drainage process, as described by Hillel et al. (1972) and Libardi et al. (1980). The monitoring of soil water content profiles (with the SMNP) and gradients of hydraulic heads below the rootzone (with tensiometers) allows the periodical calculation of water balance and water flows, and hence access to ET_a and subsequently to the yield response factors k_y.

Mannocchi and Mecarelli (1994) showed that, using Equation (1), it was possible to model relationships between crop yield and water applied. These relationships acted as a constraint in a mathematical programming framework, with the aim of optimizing (in economic terms) the application of available irrigation water, taking into account the possibility of varying the cropping pattern. An optimal solution was possible only on an annual basis; there was an attempt to define a method for determining a single, constant and optimal solution.

The objective of this study was to compare two series of yield response factors k_y obtained separately by FAO, through the literature or calculations, and by an IAEA coordinated research project (CRP) under monitored field conditions.

Materials and methods

Implementation of the CRP "Nuclear techniques to assess irrigation schedules for field crops" involved a network of eleven developing countries from 1990 to 1995; results were published in a technical document (IAEA, 1996), with a later synthesis of this research in book form (Kirda et al., 1999). The measurements of crop yield responses to deficit irrigation related to two sets of conditions:

• A reduced amount of irrigation water and a deficit imposed throughout the season. The code for this treatment was Tr.₀₀₀₀. An SMNP took weekly measurements of the soil water content profile; the irrigation scheduling was such as to maintain the soil water storage at 50 - 70 percent of its capacity.
• Water stress was imposed during specific growth stages of the crop under consideration. In general, four physiological growth stages for each crop are sufficient to describe their sensitivity to water stress: (a) initial (planting to 10 percent ground cover); (b) crop development (10 percent ground cover to effective full cover and initiation of flowering); (c) mid-season (effective soil cover to onset of maturity); (d) late season (onset of maturity to harvest). The deficit irrigation was applied only during one specific growth stage to be assessed for water sensitivity. The codes for these treatments were Tr.₀₁₁₁, Tr.₁₀₁₁, Tr.₁₁₀₁ and Tr.₁₁₁₀ (i.e.₀ and ₁ correspond to the stages during which irrigation water was or was not restricted, respectively). During the period of restricted irrigation, the threshold of 50 - 70 percent was enforced.

Periodical SMNP profiles and tensiometer readings were used both for monitoring soil water storage and for calculating ET_a throughout the successive crop growth periods.

Results

Table 1 collates values of k_y from FAO publications for 11 crops or crop yields. Table 2 collates values of k_y obtained by research-contract holders for the IAEA CRP for ten crops in nine countries. Some crops were grown in more than one country; e.g. cotton was cultivated in three countries, and some values were not calculated/obtained for certain crop growth periods.

Table 1
FAO yield response factors

Table 2
CPR yield response factors

FAO vs. CRP comparisons were possible for 21 pairs of k_y values (Table 3). Some crops (cotton, wheat and bean) may be over-represented as they were grown in more than in one country. The t-Test gave a significant difference between the k_y pairs at the 1-percent level of probability (and at the 2- percent level with the two-tail distribution).

Table 3
Twenty-one pairs of response factors values obtained
from the CRP and FAO publications

The average k_y value was higher (+38 percent) for the CRP series than for the FAO series. Therefore, in-field experiments indicated a higher impact of deficit irrigation practices on crop yield than previously expected.

Figure 1 shows the same data; the correlation was weak (Pearson correlation = 0.43) and the range of variation of k_y was wider for the CRP data than for the FAO data (slope = 0.65). Therefore, although there was a significant relationship between the two sets of data, the in-field values obtained by the CRP research group were greater than published FAO data, particularly for small reductions in yield through deficit irrigation (intercept = 0.44). At the highest levels of yield reduction, in-field CRP data were also higher than FAO data. It would be worthwhile completing these data sets by further calculations and with well managed field experiments on different soils and in different geographical areas.

Figure 1
Comparative assessment of response factors, FAO vs. CRP

Conclusions

Crop production depends mainly on soil water status throughout the growing season. A high level of soil water availability usually ensures an optimal yield with maximum ET_a with potential losses of water and N fertilizer through leaching. Any restriction in the supply of irrigation water is likely to induce a decrease in crop yield. However, the impact of deficit irrigation on crop yield can be insignificant where the water stress is applied to the crop during specific growth stages that are less sensitive to moisture deficiency. The two series of yield response factors, k_y, showed wide ranges of variation of this parameter: 0.20 < k_y < 1.15 (FAO), and 0.08 < k_y < 1.75 (CRP). The two data sets, whilst showing the same trends, gave neither identical average values for k_y nor similar ranges of variation.

Therefore, it will be necessary to extend these data sets to other crops and cultivars, and to other soils and weather conditions, to achieve mathematical optimization of deficit irrigation systems.

References

Bell, I. 1994. Computer-aided learning in meteorology. WMO Bulletin 43: 318-324.

FAO. 1977. Crop water requirement. Irrigation and Drainage Paper No. 24. Rome.

FAO. 1979. Yield response to water. Irrigation and Drainage Paper No. 33. Rome.

FAO. 1992. CROPWAT. A computer program for irrigation planning and management. Irrigation and Drainage Paper No. 46. Rome.

FAO. 1993. CLIMWAT for CROPWAT. A climatic database for irrigation planning and management. Irrigation and Drainage Paper No. 49. Rome.

FAO. 1998. Crop evapotranspiration; guidelines for computing crop water requirements. Irrigation and Drainage Paper No. 56. Rome.

Hillel, D., Krentos, V. & Stilianov Y. 1972. Procedure and test of an internal drainage method for measuring soil hydraulic characteristics in situ. Soil Science, 114: 395-400.

IAEA. 1996. Nuclear techniques to assess irrigation schedules for field crops. IAEA-TECDOC-888. Vienna.

Kang, S., Shi, W. & Zhang, J. 2000. An improved water-use efficiency for maize grown under regulated deficit irrigation. Field Crops Research, 67: 207-214.

Kemp, P. 1996. New war of words over scarce water. Middle East Economic Digest 49: 2-7.

Kirda, C., Moutonnet, P., Hera, C. & Nielsen, D.R. (eds.) 1999. Crop yield response to deficit irrigation. Dordrecht, The Netherlands, Kluwer Academic Publishers.

Libardi, P.L., Reichardt, K., Nielsen, D.R. & Biggar, J.W. 1980. Simple field methods for estimating soil hydraulic conductivity. Soil Science Society of America Journal, 44: 3-7.

Mannocchi, F. & Mecarelli, P. 1994. Optimization analysis of deficit irrigation systems. Journal of Irrigation and Drainage Engineering, 120: 484-502.

Vachaud, G., Dancette, C., Sonko, M. & Thony J.L. 1978. Méthodes de caractérisation hydrodynamique in situ d'un sol non saturé: application à deux types de sol du Sénégal en vue de la détermination des termes du bilan hydrique. Annales Agronomiques, 29: 1-36.

Vaux, H.J. & Pruitt W.O. 1983. Crop-water production functions. In: D. Hillel, ed. Advances in irrigation, Volume 2, p. 61-93. New York, United States of America, Academic Press.