|
THEMATIC PAPER |
The purpose of this paper is to review irrigation scheduling methods using information given in the different papers presented to this FAO workshop. It will not only be a simple report of these papers, but it will contain part of the personal experience and opinion of the authors. It will not present methods and techniques in detail but will focus on the main subject of Theme 1, 'Applicability and limitations'.
The objectives of irrigation management are well stated in Huygen et al. (1995): 'Maximize net return... minimize irrigation costs, maximize yield, optimally distribute a limited water supply, minimize groundwater pollution...'. To reach these goals, it is necessary to schedule irrigation accordingly, in other words, to decide 'which fields to irrigate, when and how much' (Hess, 1996), keeping in mind that overirrigation can have negative effects on quantitative and qualitative yield (Deumier et al., 1996). Obviously, there are interactions with on-farm irrigation systems, but these will be discussed under Theme 2 (see L.S. Pereira report).
Timing and depth criteria for irrigation scheduling (Huygen et al., 1995) can be established by using several approaches based on soil water measurements, soil water balance estimates and plant stress indicators, in combination with simple rules or very sophisticated models. Some of the methods will enable 'transfer to farmers' while other will be considered as research tools or, even more, 'gadgets' (see Hill and Allen, 1996). This transfer will be made possible by following different procedures or strategies, depending on the farmers' social and technical development. These methods can range from very simple calendars (Hill and Allen, 1996; Varlev et al., 1996) to intermediate and high technology methods requiring not only some technical assistance, but also the contribution of farmers themselves through field measurements (de Jager and Kennedy, 1996; Specty and Isbérie, 1996).
The following sections are devoted to, first, discussions on methods of measurements or estimates and, second, presentation of methods of disseminating scheduling advice.
SOIL WATER MEASUREMENTS
Soil water affects plant growth directly through its controlling effect on plant water status. There are two ways to assess the availability of soil water for plant growth: by measuring the soil water content, and by measuring how strongly that water is retained in the soil (soil water potential).
Soil water content
For irrigation purposes, it is generally expressed as a fraction of the available water. This fraction is given by the ratio of available water content over available water capacity which is defined by water contents at field capacity and at wilting point.
Whatever the technique used to determine this measurement, one has always to deal with the problem of spatial variability, even when using tensiometers to control the wetted volume under furrow irrigation and drip irrigation (Peymorte and Chol, 1992).
Gravimetric measurements are mentioned only by Specty and Isbérie (1996) reporting the case of a local climatological association which takes 15 000 samples per year. Besides the sampling, which makes this technique time consuming and cumbersome, the main problem for farmers' use is the accuracy and reproducibility of the mass determinations and drying process. Phene et al. (1989) give a simple way to do it by using a set of ordinary scales and a household microwave oven.
Neutron probes use the property of scattering and slowing down neutrons by the hydrogen nuclei of the water molecules. Its main advantages are that non-destructive and direct measurements can be performed without disturbing the soil. This allows one to follow water content changes with time. This equipment can be considered affordable since it can be used at several locations in one field and for several fields, once access tubes are installed and conveniently distributed in the field. The main limitation relates to safety rules which have to be followed to operate, transport and store the probe. This is the main reason why there is a tendency to replace this equipment by the recently developed Time Domain Reflectrometry apparatus (TDR).
The TDR (Topp et al., 1980) measures the propagation of an electromagnetic wave through the soil. The characteristics of this propagation depend on soil water content through the dielectric properties of the soil. There is a good agreement between the TDR and neutron probe measurements reported by Dalton and Van Genuchten (1986). The main limitations seem to be due to gaps and cracks which may arise during installation of the rods or as a result of shrinking of the soil during drying (Topp, 1987).
Because TDR was beyond the range of equipment affordable by farmers, Malano et al. (1996) decided to use the Aquaflex, which is another soil moisture monitoring (SMM) device that measures the dielectric constant of the soil surrounding a five-metre flexible ribbon. They concluded that this SMM proved to be highly reliable.
Two other SMM devices are mentioned in Malano and Wood (1996): the Microlink, which uses thermal dissipation, and the Enviroscan, which uses a capacitive technique. While the latter was considered to perform satisfactorily, the former was declared ineffective for irrigation scheduling but they did not give any details relative to the problems encountered. This remains open to discussion keeping in mind that Phene et al. (1989) reported good field evaluations of a commercial sensor based upon thermal dissipation.
Soil water potential
Water potential measuring devices give direct information about soil water status. The more precise techniques are based on thermocouple psychrometry, but unfortunately their use is not only far beyond any farmer's reach but also beyond current agronomic field experiments.
Gypsum blocks exhibit a wide range relationship between their electric conductivity and soil water potential. However, their use is limited because temperature and salinity effects make them not very reliable.
Based upon the direct measurement of soil water pressure, through a simple hydrostatic system consisting of a porous ceramic cup connected to a manometer, tensiometers measure in situ and in real time soil matric potential down to -0.8 bars. This allows an empirical use of this device to schedule irrigation not only by advisers but also by well-equipped farmers, especially for drip irrigation and even for furrow irrigation (Specty and Isbérie, 1996). The use of electric or electronic tensiometers able to measure down to -2 bars has increased recently because they are cheaper and can be logged (Malano et al., 1996; Specty and Isbérie, 1996). However, they need correction for temperature and exhibit some problems under waterlogging conditions. To account for variability, when using tensiometers and electronic tensiometers, some sampling rules must be followed. For instance, Humicro 2000 (Deumier et al., 1996) is connected to eight sensors in the soil.
Besides A and B types of measurements, Guillaume (1995) describes a device which can be used in heavily clayed soils: 'Theresa' (Cabidoche and Ozier-Lafontaine, 1995) measures soil displacement between two levels by means of coaxial pipes. The system can be run both to start and to stop irrigation but there is a lack of information for farmers' use concerning the operative conditions to implement this technique (depth, sampling).
All soil sensors can be used alone or included in a real-time irrigation scheduling system together with evapotranspiration (ET) calculations and rain forecasts to predict the water balance in the following days (Malano et al., 1996).
SOIL WATER BALANCE
The aim of soil water balance is to predict the water content in the rooted soil by means of a water conservation equation:
D (AWC
× Root depth) = Balance of entering + outgoing water fluxes
where AWC is the available water content.
The soil water balance approach has been used in a lot of papers, either explicitly or included in simulation models (e.g., Plauborg et al., 1996; Mailhol et al., 1995; Huygen et al, 1995; Przybyla, 1996; Thoradeniya, 1995; Hess, 1996, etc.). In some of them, assumptions related to some terms of the balance equation are adopted to simplify the model and the main sources of error are given (e.g., Hess, 1996). It would be worth considering the different terms in the equation separately and analysing the problems related to each of them.
D (AWC × Root depth)1
1
It is very important to know the available water content at the date of sowing. If winter rains are sufficient, its value will be set equal to AWCap. Otherwise, it is necessary to measure or to estimate this initial value to run the water balance during the growing season.
At farm level, soil textural class, crop type and variety, and planting dates can be available. It is then necessary to estimate the root depth and available water capacity (e.g., Hess 1996). Concerning the root depth (RD), Mailhol et al. (1995), give the expression RD = RD0 + a t, where t is time and a an empirical coefficient. In no case do these authors give information to relate the maximum root depth to soil type. This question of root depth seems very important for transfer purposes. From farmer interviews, Hill and Allen (1996) found that the root depths estimated by farmers were, on the average, three times larger than the values obtained from field observations. Farmers should be provided with simple rules to determine root depth as is the case with the 'code of practice' in the United Kingdom (BSI, 1988) where a figure gives minimum and maximum values for different crop groups (for example, 1 to 2 metres for winter grains).
Concerning the available water capacity, Specty and Isbérie (1996) note that it is one of the major bottlenecks in the use of 'Irritel' advice, especially under stony soil conditions. Once again, simple rules giving the AWCap per metre depth for main soil types will be helpful to farmers (see BSI, 1988, or Maucorps et al., 1988).
Another problem is the depletion point (p) defining the readily AWCap, i.e., the portion of this available water capacity that can be depleted by the crop without affecting growth. Doorenbos and Pruitt (1977) give some values for several crop groups and different levels of evaporative demand, but in many cases field experiments are far from confirming these values (e.g., Katerji et al., 1987, for tomatoes grown on sandy soils). This is important for two reasons:
· the relative evapotranspiration is governed by the ratio AWC/(p × AWCap);
· consequently, for deficit irrigation, water balance will depend on p.
Entering and outgoing water fluxes
At the soil surface, there are: Rainfall (+ R), Irrigation (+ I), Runoff (± R.O.) and Actual evapotranspiration (- ETa).
Below root depth, there are Capillarity Rises (+ C.R.) and Drainage (- D).
Hess (1996) gives assumptions such as 'deep water table1... no capillarity rise' and 'rainfall and or irrigation assumed effective... no runoff. Even where not explicit, these assumptions are common to many papers!
1
The water conservation equation giving soil water balance can be used for operational purposes if there is no high water table. When this is not the case, it is necessary to add either soil measurements (tensiometer, for instance) or plant water stress measurements (see next section).
Deep percolation is generally considered as the term to be minimized by avoiding overirrigation.
Rainfall makes the difference not only between types of irrigation scheduling but also between levels of technology. Under temperate climate conditions, rainfall is 'typically erratic, unreliable and spatially variable' during the period of irrigation (Hess, 1996), while under arid climates, one can assume no water input from rainfall during the irrigation season and provide farmers with simple rules of thumb for adjusting irrigation dates when rain occurs (Hill and Allen, 1996). Under temperate (and possibly semi-arid) climate conditions, the rainfall regime leads to the use of high (or intermediate) technology with a weather station on farm or a least a rain gauge. Hess (1996) points out that the 'main source of error is in the measurement (or estimation) of input data and in particular of irrigation and precipitation amounts'. They are due either to inadequate installation of rain gauges or to human error in reading or entering the data.
Last, but not least, comes actual evapotranspiration (ETa). It could be a chapter per se (e.g., Itier, 1996). When running a water balance, there are two ways to obtain this major term: either from measurements or from estimates.
Measurements of actual evapotranspiration are not mentioned in any paper. Despite the efforts made during the last thirty years to promote simple systems, generally based on energy balance, it does not seem that, up to now, any of them can be considered simple and cheap enough for farmers' use. The Bowen ratio is still limited by difficulties in humidity gradient measurements, even with recent improvements which avoid the use of psychrometers (Cellier and Olioso, 1992). Systems based on the aerodynamic method (Itier and Riou, 1985) use the same type of reliable instruments as those included in automatic weather stations but the need to change levels of measurements with crop growth makes it very difficult for farmers to use. The direct measurement by means of lysimeters (Phene et al., 1989) still remains a way to calibrate empirical crop coefficients.
Estimates of actual evapotranspiration are mentioned in several contributions. They are obtained by the now very classical equation (Przybyla, 1996; Hess, 1996):
ETa = kc * ETref
kc = kcb * ks + ke
where
ETref is reference evapotranspiration (Doorenbos and Pruitt, 1977),
kc is the crop factor,
kcb is the basal crop factor, i.e., corresponding to a crop grown under no water shortage,
ks is a soil water availability factor (0-1), also called stress coefficient, and
ke is the soil water evaporation coefficient.
ETref is reported by Specty and Isbérie (1996) as the basic information which is of most interest to farmers.
The use of Class A Pan is not reported in the proposals. However, due to its simplicity of use, it is still widely used all over the world, especially in African irrigation schemes. Depending on the nature of surroundings, evaporimeter readings must be corrected to avoid overestimation resulting from high oasis effects.
Comparison of ETref estimation methods was not among the objectives of this workshop. Where data are available, there is now a world wide agreement to use the modified Penman (Doorenbos and Pruitt, 1977) or Penman-Monteith equation (Allen et al., 1994) equation, or any simplification of them (as Brochet-Gerbier in Specty and Isbérie, 1996). There is still debate concerning the choice between Penman and Penman Monteith (Thoradeniya, 1995 and Mastrorilli et al., 1995) mainly due to the use of a constant value for surface resistance in P-M equation. Whatever the choice, these is no reason to adopt different time scales when both equations are given (Specty and Isbérie, 1996). Although Penman-Monteith is seen by most researchers as the best method of estimating ETref when sufficient data are available, there is still no consensus on the most appropriate method when data are limited. The publication of new FAO guidelines is unlikely to be the last word on estimating ETref with limited data. Hence, further research is required on approaches to estimating ETref with limited data that are reliable, robust and widely applicable.
Concerning spatial extrapolation, little information is given except to use Piche evaporimeter readings to extend the aerodynamic term in both formulae (Deumier et al., 1996) or to have a weather station common to a number of farms in the case of intermediate level of technology (de Jager and Kennedy, 1996).
Crop coefficients, kc: The use of kc curves giving variations of that coefficient along the growing season has been generalized (Burman et al., 1983). The case of partial covering by the crop is considered in Plauborg et al. (1996) and Hess (1996) to separate soil water evaporation and crop transpiration. The ratio of maximum transpiration to ETref is consistent with the results given by Riou et al. (1994) concerning the relationship between the maximum transpiration of vines or trees (in vineyards or orchards) and Penman equation estimates. This ratio corresponds to the percentage of soil covering by the crowns. It could be adopted by advisers for transfer to users.
Concerning maximum kc, Hess (1996) reports values of 1.1 or 1.2 depending on crop type. It is well known that values higher than 1.2 are sometimes used. It should be pointed out that these values were probably obtained from field experiments performed under advective conditions (very small plots) and that they are not valid for fields larger than one hectare. Finally, following Thoradeniya (1995), it is important to underline that crop coefficients should be used in conjunction with the equation with which they were obtained. Up to now, it has generally been with the FAO 24 modified Penman equation (Doorenbos and Pruitt, 1977), so that they may need to be corrected when using P-M equation.
Soil water availability factor, ks: When regulated deficit irrigation is planned (Varlev et al., 1996), relative transpiration has to be related to relative available water (Plauborg et al., 1996). We are faced once again with the limitation mentioned earlier (see AWC measurement) with the depletion point varying with crop groups and evaporative demand (ETref) (Doorenbos et al., 1978). Bailey and Spackman (1996) propose linking ks to soil moisture tension using an empirical relationship to fit the data of Denmead and Shaw (1962). Besides the fact that these results were obtained on corn growing over shallow containers, how could a farmer running a simple water balance estimate the soil moisture tension?. To our knowledge, this could only be approximated by pre-dawn leaf water potential (see next section) with the limitations and difficulties corresponding to this measurement1.
1
In a paper presented after the workshop, Raine and Shannon (1996) give a simple way to estimate threshold values of ks by linking relative growth observations to water level in evaporation mini-pan.
Among all these problems, it is of interest to note that the water balance calculation is self-correcting relative to ks: an overestimation of fluxes will lead to an underestimation of remaining water and then will reduce losses during the next time step.
A last point concerning kopt, (Specty and Isbérie, 1996), i.e., the optimum value of ks to maximize quality and quantity of yield under no water restriction: it is a good idea but, are there tables of kopt, available to provide farmers with?
Irrigation scheduling procedures and time scale: There are two possibilities (Deumier et al., 1996): real-time irrigation scheduling systems (high value crops) and strategy in the choice of crops (agricultural crops).
As regards real time irrigation systems, ETref for the forthcoming days can be directly estimated from meteorological service forecasts (Malano et al., 1996; Cabelguenne et al., 1996; Specty and Isbérie, 1996). In other cases historical data can be used for full season irrigation scheduling based on long-term average data (de Jager and Kennedy, 1996; Specty and Isbérie, 1996). Both possibilities are given by Hess (1996). The first one seems to be more adapted to temperate climate with rapidly changing weather conditions. The time step is seven days in Plauborg et al. (1996) and five days in Cabelguenne et al. (1996), who recommend resetting data day after day due to the unreliability of five-day forecasts (see also Table 1 in Malano et al., 1996).
The problem of rainfall arises when using calendars based on long-term data. Calendars can be determined assuming either no rain or average rainfall conditions (Hill and Allen, 1996) or by applying different percentages of probable rainfall during the crop growing season (Thoradeniya, 1995). The latter makes calculations more complex and the author thinks that it is necessary to evaluate the consequences in terms of transfer before application.
USE OF PLANT WATER STRESS CRITERIA
Instead of measuring or estimating, by means of a water balance, the amount of water in the rooted soil, it is possible to get messages from the plant itself indicating that the time has come to irrigate. Here the question is not 'how much', but 'when' This message can either come from individual plants, then it will be necessary to have a correct sampling, or from the canopy as a whole.
Plant messages
Trunk or branch diameter change: It is simply mentioned in Deumier et al. (1996). The 'Pepista' system (Huguet et al., 1992) gives interesting information, mainly on trees, when it is possible to analyse the following two factors: mean trend in diameter growth and diurnal changes. Sensors have proven their reliability after several years of measurement. They are not difficult to install on branches and are connected to a logging system. As for other plant measurements, it is necessary to sample several trees. The main problem encountered is that, sometimes, the same response is obtained with an excess and a lack of water. Furthermore, small diurnal changes are also observed in the case of high water stress (as a result of important stomatal closure). For mild water stress conditions, diurnal changes depend on species and varieties, so it could be useful to have groups of crops exhibiting the same type of response before any possible transfer to farmers.
Leaf water potential: The main interest of this measurement lies in the possibility of linking values of pre-dawn leaf water potential to relative evapotranspiration. Itier et al. (1992) showed that it is possible for C3 plants such as wheat, tomato, alfalfa and soybean, but experiments performed since that time show that it is not the case for C4 plants such as corn and sorghum (Katerji and Itier, 1995). This technique presents other limitations such as the sampling needs which are too difficult for farmers and cannot be done automatically. Furthermore, safety rules imposed on pressure chambers make them expensive (around $US5 000). Measurements limited to pre-dawn values will allow chambers with safety vent at 10 bars, hence probably cheaper.
Sap Flow: Sap flow measurements (Valancogne and Nasr, 1989) are sometimes included among the plant water stress criteria for irrigation scheduling. In fact, one can obtain relative ET values by measuring sap flow along the trunk and comparing trees under water shortage to well irrigated trees. Two techniques are available: 1) the sap flux density technique (e.g., Cohen et al., 1981); and 2) the mass flux technique (e.g. Sakuratani, 1981). The former is limited by the need to determine the cross-sectional area of the water conducting tissue. The latter is restricted to estimation of small trees' transpiration. Both techniques need tree sampling and, if badly used, can lead to necrosis of the trunk. Their possible future use at the farm level is linked to the possibility of logging, as for diameter change measurements.
Canopy measurements
Temperature: Surface temperature measurements performed by means of I.R. radiometers are mentioned in Mailhol et al. (1995). They use the stress degree day (SDD) (Idso et al., 1981) to calibrate a model. The problems related to the use of this index are: 1) it is not a relative index, which means that threshold values need to be adjusted to each crop; and 2) it does not work correctly when small rainfall occurs during the drying cycle, because the question of knowing whether it must be reset or not is not solved.
These problems do not occur with the crop water stress index (CWSI) (Jackson et al., 1981) but its use is limited by the problems related to the baselines definition (Idso, 1982). Despite efforts to simplify calculations (Itier et al., 1993), and the continually decreasing price of I.R. radiometers, its use seems to be limited to industrial farms or advisers.
Furthermore, both SDD and CWSI can be used only if weather conditions are not rapidly changing (wind and radiation) and only for fully developed crops (in order to avoid soil surface temperature influence on measurements).
Radiation: There are some possibilities of using the ratio of intercepted radiation in the near infrared over visible ranges to anticipate water shortage. This is the aim of some tools such as Piquhelios (Baldy, 1973). However, to use it, it is necessary to know the evolution of the optimum ratio.
USE OF MODELS
Two types of models are presented:
· Models based on soil water balance, where irrigation starts when a threshold value of water content in the soil is reached (Pilote and Pilotman in Mailhol et al., 1995; Cavazza et al., 1996; Przybyla, 1995; Putu in de Jager and Kennedy, 1996; Hess, 1996; Bipode in Specty and Isbérie, 1995; Irriguide in Bailey and Spackman, 1996).· Mechanistics models, in which growth and yield are simulated and depend upon water status of the plants at different stages. (MODIS in Chang et al., 1995; Markvand in Plauborg et al., 1996; Epicphase in Cabelguenne et al., 1996; Lunardi et al., 1995; Hydra in Huygen et al., 1995; EPIC in Santos et al., 1995; Tarjuelo et al., 1995; Varlev et al., 1996). Obviously, this last type of model includes a water balance submodel generally similar to those mentioned above.
As mentioned earlier, keywords are: economic viability, environmental soundness, water conservation (Chang et al., 1995) so that users are facing a multiple criteria decision problem.
Models can be used either for strategic or tactical purposes. In the first case, the question is: how large an area to irrigate, which crops to plant and how to distribute the available water supply along the season (Huygen et al., 1995). For agricultural crops, this is the major problem while real time irrigation scheduling is important for high value crops (Deumier et al., 1996).
Models, per se, seem to be difficult to run for farmers. Irrigation simulators originated from these models could be easier to use. Deumier et al. (1996) present one software (LORA), which makes it possible for farmers to find the optimal crop planning on irrigated farms, and two irrigation simulators (MIRA and IRMA), designed for testing of management decisions. For this workshop it would be interesting to point out that decision rules result from collaboration between farmers and researchers (see L.S. Pereira report). For real-time irrigation scheduling, Plauborg et al. (1996) present the Markvand, a model being used by 200 farmers in Denmark, while Bailey and Spackman (1996) indicated that their package is currently being used by over 300 farms in the United Kingdom. Specty and Isbérie (1996) present some examples of irrigation warning in France and the results of an inquiry among users. This leads to the more general question: which method should be used for transfer of information?
TRANSFER OF INFORMATION
This point has seldom been treated in the papers. Nevertheless, some proposals have been made. For instance, regarding the transfer to farmers, de Jager and Kennedy (1996) define three levels of technology.
(a) High technology: this corresponds to weather stations on farm with additional expert advice. In Specty and Isbérie (1996), this corresponds to industrial farms.(b) Intermediate technology: this corresponds to weather stations shared by several farmers, located centrally to a number of farms; the farmers receive personal assistance from an adviser. In Specty and Isbérie (1996) this corresponds to the use of Irritel.
The cases of BIPODE (Specty and Isbérie, 1996), MARKVAND (Plauborg et al., 1996) and IRRIGUIDE (Bailey and Spackman, 1996) are between (a) and S(b) because national meteorological services provide ETref and rainfall occurrence forecasts while farmers provide effective rainfall and site specific soil and agronomical data.
(c) Minimum technology: fixed irrigations of given amounts are applied at regular intervals. This corresponds also to the simple calendars proposed by Hill and Allen (1996), or to the fixed amounts of required irrigation proposed by Varlev et al. (1996) for different periods of corn growth. This minimum technology corresponds also in Specty and Isbérie (1996) to the case of advice limited to dates of first and last irrigations.
The irrigation scheduling options are linked to the level of technology. In de Jager and Kennedy (1996) a) and b) are demand driven while (c) is supply driven. Scheduling criteria refer to two categories: timing and depth criteria. Huygen et al. (1995) give five timing criteria which, for simplicity, can be reduced to three:
(a) Allowable daily stress, or relative ET = a (with 0.5 < a < 1), in the case of deficit irrigation (obtained by means of soil water balance or from plant water stress indicator).(b) Readily AWC consumed or some fixed percentage of AWC consumed (obtained by means of soil water balance).
(c) Critical pressure head or moisture content at sensor depth (obtained by means of a SMM device).
The depth criteria are then either 'back to field capacity' or 'fixed depth'. The former can be used when coupled to the use of soil water balance while the latter is more common. In de Jager and Kennedy (1996) Table 1, one can see that both the use of a soil water balance and that of the depth criteria consisting of refilling soil water capacity belong to high and intermediate technology levels.
In the case of the minimum technology level, fixed intervals or simple calendars, as those presented by Hill and Allen (1996), appear to be highly preferable to irrigation at random or 'when one's neighbour irrigates' as often occurs, even in developed countries.
CONCLUSIONS AND RECOMMENDATIONS
The objective of this workshop was to analyse what really works and also what is really workable under minimum technology conditions. Two types of opinions arose:
For a first group of participants, farmers are information saturated and too tired at night to comply with the requirements of real-time scheduling. It is therefore necessary to develop simple calendars. These calendars will replace units of depth by units of time, which are more understandable by farmers.
For a second group of participants, the time has come to transfer existing knowledge while improving methods and techniques.
The debate is still open. However, for all the users of methods and techniques, whether farmers or advisers, there is a need for: 1) research concerning both knowledge and development of devices, and 2) improvement of transfer schemes
Research needs
Soil water measurements:
· Safe, simple, cheap and automatic devices for soil moisture measurement must be developed.· There is also a need to develop remote techniques able to perform spatially integrated measurements to overcome the problems related to spatial variability.
Soil water balance:
· Improving the reliability of rainfall forecast.· Developing methods aimed at mapping ETref in order to interpolate network data.
· Developing studies devoted to quantifying local advection effects on water consumption of irrigated fields surrounded by arid zones.
· Developing thematic mapping enabling farmers to obtain easy estimates of the available water capacity per metre depth of soil.
· Improving and revising tables giving allowable depletion levels for different crop groups and evaporative demand.
· Developing tables giving root system growth velocity for different crop groups and soil types.
· Improving definitions relative to critical periods for water (i.e., differentiating skipping from drought sensitivity).
· Establishing generalized tables giving optimum crop coefficient kopt, (fruit trees, pea, sunflower, etc.).
· Developing studies to analyse the influence of planted roll banks on the water balance of irrigated schemes.
Plant water stress criteria:
· Developing research on qualitative and quantitative plant water stress indices to enable farmers to have simple and cheap warnings concerning the water status of their crops.· Defining groups of crops having similar responses to water stress in trunk or branch diameter growth and diurnal changes.
· Developing studies to discriminate changes in diameter due to excess water from those resulting from lack of water.
· Designing safe, cheap and possibly automated devices for pre-dawn leaf water potential measurements.
Use of models:
· Developing methodologies to recalibrate models by using real-time information.· Developing water balance models taking into account water fluxes which are generally neglected (capillarity rises, seepage and runoff).
· Developing irrigation simulators taking into account hydraulic or socio-economic network constraints.
· Developing easy to use irrigation simulators originated from models, devoted either to strategic purposes (choice of crops and irrigated plannings) or to real-time irrigation scheduling.
Transfer needs
· Writing simple booklets devoted to guidelines for irrigation scheduling. The guidelines should be for high, medium and low levels of technology transfer to meet the needs of low-input and high-input farmers. This has to be done both by researchers and extensionists. The national committees of ICID could help people from different institutes or organizations to meet as did the French committee for updating the RNED-HA1 booklet. In some developing countries, they should be in the language used by the farmers (Hill and Allen, 1996).
1RNED-HA: Réseau National d'Etude du Drainage et de l'Hydraulique Agricole.
· Spreading simple ways of weighing and drying soil samples.
· Training farmers to use soil water potential devices to identify high water tables.
· Spreading simple ways to estimate effective rain for its use in the soil water balance equation.
· Spreading simple rules for adjusting irrigation dates or amounts in case of rainfall occurrence.
· Spreading simple ways of correcting class A pan evaporation data with wind velocity, air humidity and nature of surroundings.
· Ensuring when spreading new techniques for estimating ETref (e.g., as P-M equation) that users are provided with appropriate crop coefficients.
· Training extensionists in the use of software or irrigation simulators originated from models.
The authors are grateful to panels member J. de Jager and R. Allen, in participants to the workshop discussion and to reviewers for helpful comments on the manuscript.
REFERENCES
Allen, R.G., Smith, M., Perrier, A. and Pereira, L.S. 1994. An update for the definition of reference evapotranspiration. ICID Bulletin 43 (2): 1-34.
Bailey, R.J. and Spackman, E. 1996. Irriguide: an irrigation scheduling system to take account of variable rainfall, soil texture and cropping pattern. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Baldy, Ch. 1973. Contribution à l'étude de l'extinction de l'énergie d'origine solaire dans des peuplements de blé tendre (Triticum aestivumL. EM. Thiell). Annales agronomiques 24: 507-532.
B.S.I. Standards. 1988. Draft code of practice for the planning designing and installation of irrigation schemes in the United Kingdom. Part 2. Irrigation water requirements. B.S.I. Standards, Technical Committee; AGE/30, Irrigation and drainage equipment, no 88/73803 DC.
Burman, R.D., Cuenca, R.H. and Weiss, A. 1983. Techniques for estimating irrigation water requirements. In: Advances in Irrigation. Vol. 2. D. Hillel (ed.). Academic Press, New York. pp. 336-394.
Cabelguenne, M., Puech, J., Debaehe, P., Bosc, N. and Hilaire, A. 1996. Tactical irrigation management using real time EPIC-phase model and weather forecast: experiment on maize. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Cabidoche, Y.M., Ozier-Lafontaine, H. and Theresa, I. 1995. Matric water content measurements through thickness variations in vertisols. Agric. Water Manag. (In press).
Cavazza, L., Patruno, A., Rossi, P., De Seneen, M.O. and Gammino, M. 1996. Evaluation of an irrigation scheduling programme for providing advice to farmers. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Cellier, P. and Olioso, A. 1993. A simple system for automated long-term Bowen ratio measurement. Agric. For. Meteorol. 66: 81-92.
Chang, Y., Vanclooster, M. and Feyen, J. 1995. A multiple objective dynamic programming approach to schedule irrigation for sustainable irrigated agriculture. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome.
Cohen, Y., Fuchs, M. and Green, G.C. 1981. Improvement of heat pulse method for determining sap flow in trees. Plant Cell Environ. 4: 391-397.
Dalton, F.N. and Van Genuchten, M.T. 1986. The time-domain reflectometry method for measuring soil water content and salinity. Geoderma 38: 237-250.
De Jager, J.M. and Kennedy, J.A. 1996. Weather-based irrigation scheduling for various farms (commercial and small-scale). In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Denmead, O.T. and Shaw, R.H. 1962. Availability of soil water to plants as affected by soil moisture content and meteorological conditions. Agron. J. 54: 385-390.
Deumier, J.M., Leroy, P. and Peyremorte, P. 1996. Tools for improving management of irrigated agricultural crop systems. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Doorenbos, J. and Pruitt, W.O. 1977. Guidelines for predicting crop water requirements. FAO Irrigation and Drainage Paper 24. FAO, Rome.
Doorenbos, J., Kassam, A.H., Bentvelder, C. and Uittenboogaard, G. 1978. Yield response to water. FAO Irrigation and Drainage Paper 33. FAO, Rome.
Guillaume, P. 1995. Pilotage d'un système d'irrigation sur vertisol au moyen du suivi des variations d'épaisseur de couches de sol. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome, Italy.
Hess, A. 1996. A microcomputer scheduling program for supplementary irrigation. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Hill, R.W. and Allen, R.G. 1996. Simple irrigation calendars: a foundation for water management. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Huguet, J.G., Pelloux, G. and Lorendeau, J.Y. 1992. PEPISTA: Programmateur d'irrigation par les microvariations de diamètre de la plante. In: Proceedings of the Workshop on Real Time Sensing and Control of Automated Irrigation Systems. ICID, 16th European Regional Conference, Budapest (Hungary), 26 June 1992.
Huygen, J., Van den Broek, B.J. and Kabat, P. 1995. Hydra Model Trigger, a soil water balance and crop growth simulation system for irrigation water management purposes. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome, Italy.
Idso, S.B. 1982. Non-water stressed baselines: a key to measuring and interpreting plant water stress. Agric. Meteorol. 27: 59-72.
Idso, S.B., Reginato, R.J., Reicosky, D.C. and Hatfield, J.L. 1981. Determining soil induced plant water potential depressions in alfalfa by means of infrared thermometry. Agron. J. 73: 826-830.
Itier, B. and Riou, C. 1985. Une possibilité de suivi de l'état hydrique des cultures en zone semi-aride à partir de mesures micrométéorologiques simples. In: Les besoins en eau des cultures. Crop water requirements, Perrier A. and Riou C. (eds.). ICID International Conference, Paris (FRA), INRA, Paris, 831-840.
Itier, B., Flura, D. and Belabess, K. 1993. An alternative way for C.W.S.I. calculation to improve relative evapotranspiration estimates: results of an experiment over soybean. In: Irrigation of horticultural crops. Lopez-Galvez J. (ed.). Int. Symp., ISHS, Almeria (SPA), 23-27/11/1992, Acta Hortic., 335, 333-340.
Itier, B., Flura, D., Bellabes, K., Kosuth, P., Rana, G. and Figueiredo, L. 1992. Relations between relative evapotranspiration and predawn leaf water potential in soybean grown in several locations. Irrig. Sci. 13: 109-114.
Itier, B. 1996. Measurement and estimation of evaporation. In: Sustainability of Irrigated Agriculture. L.S. Pereira, R.A. Feddes, J.R. Gilley and B. Lesaffre (eds). NATO Workshop, Vimeiro (Port.), 1994 NATO ASI Series, Kluwer, Dordrecht, 171-192 pp.
Jackson, R.D., Idso, S.B., Reginato, R.J. and Pinter, P.J. 1981. Canopy temperature as a crop water stress indicator. Water Resources Res. 17: 1133-1138.
Katerji, N. and Itier, B. 1995. Utilisation d'indicateurs de l'état hydrique des plantes pour la conduite de l'irrigation. E.E.C. Contract no. 8001 - CT 91-0108 - Final report.
Katerji, N., Itier, B. and Ferreira, I. 1987. Etude de quelques critères indicateurs de l'état hydrique d'une culture de tomate en région semi-aride. Agronomie. 8 (5): 425-433.
Lunardi, L., Muni, M., Francaviglia, R. and Costantini, A. 1995. A simulation model for irrigation scheduling. Applicative examples on soybean in two environments of the Po plain. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome, Italy.
Mailhol, J.C., Revol, P. and Ruelle, P. 1995. Outils opérationnels pour déceler l'apparition du stress hydrique en grande culture: comparaisons sur l'exemple du sorgho. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome, Italy.
Malano, H.M., Turral, H.N. and Wood, M.L. 1996. Surface irrigation management in real time in southeastern Australia: irrigation scheduling and field application. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Mastrorilli, M., Katerji, N., Venezian Scarascia, M.E., Losavio, N., Rana, G. and Colucci, R. 1995. Reference evapotranspiration in Mediterranean region: from theory to applied procedures. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome.
Maucorps, J., Solau, J.L. et al. 1988. Carte des sols du Département de l'Aisne, Saint-Quentin, no 7/8, éch. 1/25000. Chambre d'Agriculture de l'Aisne Ed., Laon (FRA).
Peyremorte, P. and Chol, P. 1992. Pilotage automatique optimisé des irrigations localisées. In: Proceedings of the Workshop on Real Time Sensing and Control of Automated Irrigation Systems, ICID, 16th European Regional Conference, Budapest (Hungary), 26 June 1992.
Phene, C.J., Allee, C.P. and Pierro, J. 1989. Measurement of soil matric potential and real time irrigation scheduling. Agric. Water Management 16 (3): 173-185.
Phene, C.J., McCormick, R.L., Davis, K.R. and Pierro, J. 1989. A lysimeter system for precise evapotranspiration measurement and irrigation control. Trans. ASAE. 32 (2): 477-484.
Plauborg, F., Andersen, M.N., Heidmann, T. and Olesen, J.E. 1996. MARKVAND: An irrigation scheduling system for use under limited irrigation capacity in a temperate humid climate. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Przybyla, C. 1996. Irrigation scheduling in large-scale sprinkler irrigation in the Wielkopolska region (Poland). In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Raine, S.R. and Shannon, E.L. 1996. Increasing the irrigation efficiency of Burdekin Cave growers through participatory action learning. Presented at Irrigation Associations of Australia Conference, 14-16 May 1996, Adelaide, Australia (in press).
Riou, C., Pieri, P. and Le Clech, B. 1994. Consommation d'eau de la vigne en conditions hydriques non limitantes. Formulation simplifiée de la transpiration. Vitis 33: 109-115.
Sakuratani, T. 1981. A heat balance method for measuring water flow in the stem of intact plant. J. Agric. Meteorol. 37: 9-17.
Santos, F.L., Serralheiro, R.P., Santos, A.M. and Bica, M.A. 1995. Maize tactical irrigation management in the Southwest of Portugal using real time Epic model. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome, Italy.
Specty, R. and Isbérie, C. 1996. How farmers use information obtained from irrigation advisory services in France. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
Tarjuelo, J.M., De Juan, J.A., Valiente, M. and Garcia, P. 1995. Model for optimum irrigation water management. Application to an extensive horticulture farm. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome, Italy.
Thoradeniya, W.B.M. 1995. Optimization of irrigation water releases for rice growing in the wet season using probable rainfall. Paper submitted to ICID/FAO Workshop, Sept. 1995, Rome. Irrigation Scheduling: From Theory to Practice. FAO, Rome, Italy.
Topp, G.C., Davis, J.L. and Annan, A.P. 1980. Electromagnetic determination of soil water content: measurements in coaxial transmission lines. Water Resources Res. 16: 574-582.
Valancogne, C. and Nasr, Z. 1989. Measuring sap flow in the stem of small trees by a heat balance method. Hortsci. 24 (2): 383-385.
Varlev, I., Dimitrov, P. and Popova, Z. 1996. Irrigation scheduling for conjunctive use of rainfall and irrigation based on yield-water relationships. In: Irrigation Scheduling: From Theory to Practice, Proceedings ICID/FAO Workshop, Sept. 1995, Rome. Water Report No. 8, FAO, Rome.
