R. Specty, Météo-France, Direction Interégionale Nord-Est, Parc d'Innovation, Illkirch, France C. Isbérie, CEMAGREF, Groupement d'Aix-en-Provence, Division HIAX, Le Tholonet, France
Firstly, this paper discusses the characteristics of the information provided by irrigation advisory services in France, in particular that based on water balance estimates. The classification takes into account the degree of elaboration of the message that is delivered to the farmers. Irrigation scheduling tools using soil water measurements, which are sometimes associated with the delivery of information, are also described. Secondly, the results of a survey among customers of the videotex service 'Irritel' are studied. The reply rate is important for this kind of survey (20%), and enables the farmers', and agricultural advisory services', reactions to the information provided to be analysed. Proposals are then presented for improving the service.
To schedule irrigation, it is possible to refer to the soil, to the plant or to the atmosphere. Information supplied by advisory services generally relates to the atmosphere. The information varies in its degree of detail and how directive it is, ranging from potential evapotranspiration (reference evapotranspiration), PET (ETref), to advice about irrigation dates and amounts. It is distributed by various means: letters, weekly agricultural papers, answerphone, fax, Minitel (videotex), radio.
SOME ASPECTS OF INFORMATION OBTAINED FROM IRRIGATION ADVISORY SERVICES IN FRANCE
Providing information about the water balance elements
Many services, particularly Météo-France, provide information on water balance elements, including the PET (ETref) of the moment, leaving it to the farmers to do the calculations and draw the graphs. The daily PET (Penman-Monteith) is delivered on answering machines and, generally, the Penman PET (ETref) in weekly bulletins or newspapers. When this information is given in a weekly newspaper (for example, the Meteorological Department Centre [CDM] of Colmar), but also on Minitel (for example, the Agro-Meteorological and Economic Regional Information Centre [CIRAME]), it may be more condensed (greenhouse PET) and include other elements of the water balance: pluviometry, crop coefficient (kc) or optimized crop coefficient (kopt). Agronomic advice is often added. Moreover, it would be more exact to say that the water balance data exist in order to complete the agricultural advice (for example: 'Agro-Conseil' from the Chamber of Agriculture in the Aisne department).
In the Drôme department, the weekly bulletin of the Chamber of Agriculture, called 'Zoom 26' is made up of four editions, one for each kind of crop. It is doubled with a videotex edition (Télézoom).
The Météo-France service is often the provider of basic meteorological data, particularly PET (ETref), that is sold or supplied free, or delivered via the 3615 METEO answering machine (or even 3615 Irritel, but with more difficulty). In the department of Morbihan, the CECAB cooperative receives these meteorological data by fax.
The National Institute of Agronomical Research (INRA), with its Cimel station, provides a software calculation of Penman PET that is a little particular, since the sunshine duration is calculated from a theoretical global radiation.
Forecasts of PET (ETref), and even pluviometry, are sometimes distributed. This is the case with Irritel for the Valunion Cooperative (Côte d'Or, Saône-et-Loire, Jura). This is also the case in Calvados, and in Hérault for the PET ('Agri-Conseil' bulletin of the Hérault Climatological Association which contains other mainly climatological information).
Sometimes the water deficit (rains less maximum evapotranspiration [ETm]) of various crops, particularly maize, is indicated, for example each week, and for various areas, by the Development Company for the 'Côteaux de Gascogne'.
Farmers generally like to know what is happening elsewhere. Even if they often do not use the PET (ETref) directly (except for drip irrigation), they are interested in information on the water balance data at different meteorological stations, possibly on a map. They then locate the places that are similar to their own.
How to use the information
In the Haut-Rhin, Météo-France and the departmental Chamber of Agriculture used to give the farmers cardboard forms for copying out the weekly PET (ETref), the crop coefficient, kc and the pluviometry of the nearest station given by a bulletin (or if available, the data of their own pluviometers), and for making calculations and graphs. The system still exists, or rather it has reappeared, in some departments. But more generally only PET (ETref) or ETm normals are taken into account. An example is the 'Équilhydre' abacus distributed by the Société du Canal de Provence (and other services). Weekly adjustment advice is published in the newspaper 'L'Agriculteur Provençal' ('The Farmer of the Provence'). Note: Press articles generally give instructions for using the water balance elements.
Providing the water balance itself
In order to provide water balances, it is necessary to receive data from the farmers: by mail, for the data that change only once a year (crops, soil water reserve, places), by phone or Minitel for the vegetative stages, irrigation dates and amounts. Under these conditions, the adviser is often tempted to be directive ('Bipode'1). Only 'Irritel'2 delivers field water balance without being directive (however, with two conditions: irrigation allowed only when the first soil water reservoir is empty, and maximum irrigation rate): the farmers themselves note how much water they apply on the fields.
1 'Bipode': before the campaign, an agricultural adviser notes down the farmers' 'field files' and enters them in a computer. During the campaign, after phone discussions, the adviser updates the water balances and proposes an irrigation programme (ETm normal and no-rain hypothesis).
2 'Irritel': interactive videotex service based on field calculation of the double-reservoir water balance, supplied every day from the data provided by the farmer and by Météo-France (Specty and Martini, 1993; Bonnemort et al., 1995).
Water balances may be calculated not only by advisory services, but also by the farmers themselves: instead of collating the agricultural data at a central level, meteorological data are provided to the farmer using micro-computer software. Such software packages are used by the agricultural advisers themselves (departmental Chambers of Agriculture, cooperatives, the Agricultural Technical Study Centre [CETA], or private companies) to write bulletins, or directly by the farmer. This is the case with 'Météopro' of the Agricultural Technical Coordination Association (ACTA), used by the Chamber of Agriculture of Loire, 'Bipode' of the CIRAD in Montpellier, used by MAISAGRI in Castelsarrasin, 'Métir' of Météo-France Ouest used by COOPAGRI in Bretagne. The 'Mira' software ('Bilaneaumètre') of Société du Canal de Provence will soon be distributed by ISAGRI.
Some software programs are still used only by their authors, particularly in the case of research tools: 'BYM'1 of the INRA in Avignon, 'Irricane III' of the CIRAD in Reunion Island, 'Pilote' of Research for Engineering in Agriculture and Environment (CEMAGREF) in Montpellier. Many microcomputer software packages are still only used as simulating tools. This may explain why many people ask only for PET (ETref).
1 'BYM': this decade water balance can be adapted to all crop models. It separates evaporation and transpiration, takes into account the evolution of the root system and generates reduction functions that also interact with growth and development functions.
Providing irrigation dates and amounts
In this case, it is necessary either to know the water balance of each field (Irritel), or to simulate irrigation scheduling on standard fields. This was done by the 'Genie Rural' service in Colmar in 1964. In various places, on three kinds of soil and for various crops, these amounts were supplied by a written bulletin for each crop stage. But farmers did not always apply these water amounts, and the second piece of advice was often already wrong.
Such very directive 'advice' can reappear: the Chamber of Agriculture of Eure-et-Loir mails postcards (free) to the farmers with information obtained from a water balance simulation.
Finally, we could imagine a situation where, from Irritel (with a printer linked to Minitel or a computer), an agricultural adviser sends (for example, by fax) farmers irrigation advice, according to their own water balances. Unfortunately, this does not happen: often, the advisers use Irritel with data obtained from standard field simulations.
At present, many services providing irrigation dates and amounts sometimes seem to be directive, but are only giving an example, and, moreover, trying to provide other information obtained from the field. So, besides 'Équilhydre' (already mentioned), the Chamber of Agriculture of Lot-et-Garonne supplies, through a bulletin, PET (ETref) from values provided firstly by the INRA and then readjusted in accordance with their own Piche evaporation measures. It also gives advice about irrigation dates and amounts for two crops (maize and soybean). The Chamber of Agriculture of Isère takes PET (ETref) from Irritel and puts it on its answering machine. Moreover, it helps farmers with information obtained from tensiometric measurements.
Some services only provide advice about beginning and ending the irrigations. They help to determine the crop stage and to check, generally with tensiometric measurements, that there is still enough water in the soil to begin the irrigations at the intended date, and that there is as little water as possible left in the soil at the end of the irrigation season (generally recommended strategy). Then, between the beginning and the ending of irrigations, the farmer systematically irrigates with the provided amount, only interrupting the cycle in the event of considerable rain.
Tools related to the soil
In some cases, soil water controls may be used as a basis for or a complement to irrigation warnings. They enable one to follow, on a certain number of fields, the evolution, either of the soil water content, or of the potential energy linking water and soil, which is called 'tension' or 'matric potential' and indicates the availability of this water for the crop.
Gravimetry: soil sampling with auger and oven-drying. The soil moisture is measured at different depths. This method is a reference, but it is heavy or too localized, laborious and slow. Nevertheless, the Climatological Association of Moyenne-Garonne uses it to take about 15 000 soil samples a year.
Neutronometry: counting of the neutrons thermalized by inelastic shocks on the nuclei of hydrogen atoms using a 'neutron probe'. The volumic water content is determined on the whole soil profile. The equipment is expensive, calibration is delicate, and its use is subject to constraints (radioprotection). However, it can provide rapid and sequential measures at the same place without disturbing the soil. The probe is mainly used by researchers. It was once also used by commercial services in France. The CIRAME uses these measurements as an indicative tool.
Tensiometry (Peyremorte and Isbérie, 1986): Two kinds of devices are widely distributed on the French market: 'standard' tensiometers (porous cup linked to a depression manometer by a tube filled with water); and watermark sensors (electrical measure block buried in the ground and removable meter). Tensiometers measure the soil matric potential (but only below 100 kPa). A little delicate to implement, exacting to pick out (automation is desirable), nevertheless they are a good tool for controlling the irrigations (particularly drip). Nowadays they are more used by agricultural advisers than by farmers, unless the latter are well trained. The watermark sensor measurements need correction in relation to soil temperature. They are easier to implement than tensiometers. Nevertheless, the lack of means for controlling if they are working well means increasing the number of repetitions.
Conclusions about these tools
The control of the soil water status, in order to improve the irrigation scheduling, can only be individual. Moreover it is often localized. Therefore, one must ensure that the information obtained from such measures is representative, temporally and/or spatially (kind of soil, of exposure). Here again, agricultural advisers make more use of it than do farmers. Nevertheless, the complete automatization of irrigation using these tools is possible (one farm in the Haut-Rhin department).
RESULTS OF A SURVEY AMONG THE USERS OF AN IRRIGATION ADVISORY SERVICE (IRRITEL)
In addition to the 'Bulletin of the irrigating farmer', the Irritel videotex service also gives other technical information concerning: the use of drip irrigation, the irrigation of golf courses, gardens, irrigation courses, meteorological forecasts. It is widely distributed in France and is used in many departments. The farmers, or their agricultural advisers, send the data corresponding to each field to the computer service. Then the meteorological data are added to calculate the field water balance. This is supplied to the farmers, and allows them to plan their irrigations. In 1993 a sample survey concerning mainly the 'Bulletin of the irrigating farmer' was conducted among the farmers using this service.
The farmers' opinions (very good, good, poor, useless) were requested through squares to tick off. The questions turned on the use of the product, the data provided, the benefits gained from it. Some more general questions were also asked ('How did you hear about Irritel?', 'Do you otherwise use the Minitel?', etc.). Then in a 'Comments and general observations' section they could freely express themselves. Out of 2000 questionnaires sent, approximately 400 replies were received, 300 of which had a free comment.
Answers to the directive questions
How did you hear about Irritel?: A majority of farmers (50%) said that they heard about the service from an agricultural adviser. However, sometimes this agricultural adviser boycotted the product or misappropriated it as will be seen later. Information by word of mouth had almost no effect (6%), the directory even less (5%). The specialized press seemed quietly efficient (19%), but the indication 'another medium' (not specified) prevailed (23%).
The use of the product: The product is generally considered available, easy to use, logical in its progression and well displayed (good: 60%, very good: 12 to 29%). The 29% very good concerns the availability, the 20% poor applies to the ease of use. Further contrasting opinions emerge in the comments.
Data supply: 'Good' also prevails for the characteristics of the data supplied, quantity, accuracy, precision, actualization, access speed and display form (from 41 to 67%). However, the 'very good' no longer exceeds 'poor', except for the quantity of information supplied (very good: 16%, poor: 7%) and the display form (very good: 13%, poor: 11%). As regards access speed, the 'nil' appears (5%). In the comments, criticisms are more virulent and praises are more flattering.
The benefit gained: Also here, the 'good' prevails as regards general interest (good: 55%, poor: 19%) and gains obtained (good: 36%, poor: 29%). There are as many 'nil' (3%) as very good (5%). It may be supposed that this is due to an a priori judgement on Minitel, not based on objective criteria. Contrary to that concerning data supply and product use, the free comments here are more discreet than the answers to the questions.
Analysis of the comments and general observations
The comments and general observations are very eclectic, ranging from emphatic praise: 'excellent service', 'thanks to the inventor', 'essential tools for irrigation scheduling', 'very practical and very reliable', to ironic rejection: 'totally useless, let's come down to earth'. The main themes are summarized below.
Irritel 'misappropriation': A great number of Irritel users only look for the evapotranspiration and sometimes do not find it, or with difficulty (it is necessary to 'introduce fields', possibly imaginary), and they complain about this. So, Irritel is often misappropriated to verify other calculations, made by hand or with other software. The agricultural advisers often do this, 'opening' control fields and then finding that Irritel 'is not adapted to a collective use'.
Comparison with other references: Even some identified, so regular, users of Irritel only rely on their own experience: 'we have been irrigating for 34 years'; or they use other methods, for example, soil water control (once even gravimetric measurements of soil water content with soil sampling and drying oven), and particularly tensiometers (watermark). Then they compare their results to Irritel ones, and sometimes talk about overestimation, sometimes about underestimation, more often about similarity. In this last case, they abandon one method for the other one. For example, when they begin to use Irritel, they give up calculating from the normal PET(ETref): 'by comparison with the PET curve that we used for several years, Irritel is very accurate'.
The competition felt by other services: Agricultural advisers sometimes view Irritel as serious competition for bulletins that are already implemented and, therefore, they boycott it:
· 'It is a great pity that the information and the way to use it are not well known and disseminated by our professional advisers';
· 'An exchange between other users, advisers, meteorologists, in order to take a bearing each year, would be necessary. I could never get it in my department';
· 'Hard to promote';
· 'Not enough agricultural advice';
· 'It would be better to introduce it more'.
Irritel complexity: As it needs interactivity, Irritel is found complicated: 'very good product but not easy to use', 'it is taxing to insert these data'. Though the majority of farmers who answer have not met any difficulties ('Irritel easy to use'), some others point out difficulties in accessing the service (loss of codes or mistakes), in data insertion, or in modification of the latter. Often the irrigation histogram is found easy to understand, but sometimes not the graphic representation (particularly both soil reservoirs).
Slowness: Though some people find the product 'reliable and quick to use', this is not generally the case; many people complain about its slowness in recording (linked to the Minitel medium, slow in comparison with a computer, and not to Irritel itself), and data input, particularly the first time and at the beginning of the year: 'the implementation is heavy', as is the slowness of development ('access should be shorter'), that is nevertheless found 'pedagogical'. These last points are valid reasons for modifications, although one person captivated by Irritel has pointed out: 'On the whole, the product is good to very good, do not touch anything'. Another user comments: 'I would prefer an easier and more rapid bulletin'. The short cuts are unknown and a 'more direct coded access' is requested.
Training: It is obvious that, allowing for exceptions, an initial, indeed yearly, training is essential and requested: 'even briefly, from two to three hours', 'you have to follow a training course to understand'. If Irritel is 'difficult for a farmer who is not assisted by an agricultural adviser', it is generally only 'at first sight', for it has an undoubted pedagogical interest.
Irritel's pedagogical interest: 'Irritel was one of our major products used for training in the use of farm 'telematics' (in Professional Training and Agricultural Promotions Centre [CFPPA])'. The 'quality of the information supplied' is also 'useful for research', but perhaps too costly for the farmers who want to spend less money.
Cost of use: Numerous farmers point out the excessive cost of using Irritel: 'comes up to my expectations, but to be used in moderation'. Then Irritel is used as little as possible.
Under-equipment in irrigation material: Another reason (sometimes brought to the fore by Irritel) for not using Irritel during the year is the lack of irrigation material. Then some farmers only use Irritel in order to begin: 'good tool for beginning the irrigations if the total available soil water holding capacity is well known'. So, Irritel may sometimes be more used during rainy years, when irrigation is often switched off (once the soil is filled again).
Total available soil water holding capacity/easily available soil water holding capacity: Nevertheless, the estimation of the total available soil water holding capacity largely acts as a 'brake upon Irritel use'. It is particularly hard to determine for chalky soils and stony soils.
Availability of the resource: Irritel is 'well adapted for farmers who have water at will, but not for rotational irrigation', 'the management of the resources is a priority'.
Actualization: Updating is sometimes considered late, 'in particular at week end', 'the last available dates are generally those from D-3, indeed D-4, can't we do better with an automatic system?'
Other suggestions concerning this service: Many improvement ideas have been expressed, showing that the farmers are interested in the product and its improvement, even if some suggestions are not realistic and/or not feasible. We have tried to group them, though some of them relate to several categories.
· Considerations of a practical nature:
· To couple together Minitel and fax.
· Automatic disactivation and reactivation of the calculation fields.
· Definitive suppression of the identified fields in the computer service made possible for the user himself.
· To propose an Irritel operating on micro-computers.
· Data collection:
· To increase the density of meteorological stations.
· Nature of the supplied elements:
· To provide data on huidity.
· To propose a 'good' meteorological forecast.
· To bring back the bulletin on vegetable protection.
· Data processing:
· To take the objective of productivity into account.
· To recalculate the vegetative stages with the temperatures sums.
· To develop the aspects concerning the vegetable gardens.
· To take rain efficiency into account.
· Data display:
· To do a daily histogram and another one covering several years.
· Information storage:
· To keep information on past irrigation campaigns ('so much lost information!').
Irritel impact: Farmers remain discreet about their productivity, but as 'the supplied information seems to fit with reality', Irritel is 'well adapted to the soil' and 'can help to save water'. It enables 'a better calculated irrigation, less waste'. Sometimes Irritel modifies habits: 'watering is no longer done on the basis of time but on water height'.
SYNTHESIS: A FEW PATHS FOR POSSIBLE AND DESIRABLE EVOLUTION IN PROVIDING IRRIGATION INFORMATION
Irrigation information is not used very much...
The percentage of farmers who irrigate and who actually use information from advisory services seems, for various reasons, very low. The farmers who accept an help are often people who have never irrigated before (so Irritel is more successful in the departments of the north of France). The others sometimes put forward their experience, which can be misleading: for example, on wheat, in Alsace, too much water is (or was) applied during the 'dry' years (and too late) and... not enough water during 'wet' years. In the areas where irrigation is traditional, the implementation of an irrigation advisory service will perhaps provide an opportunity to point out erroneous practices. (In Irritel, the capacity to store data and compare the present year with the previous ones is an advantage).
... but this 'might' change
Until now the price of water (without pressure) has been a minor factor, and sometimes of no importance at all, in the (rather high) cost of irrigation. It might change during shortage periods, when the bans on irrigation are double-edged. Groundwater pollution must also be taken into account, even if following advisory service information is sometimes an alibi. So the information should not be intended for farmers only, but also for the 'public services', potential partners for a more general service of water management.
The proposed warnings are varied...
This profusion is a blessing and the various approachs should be managed in order to complement each other. Nevertheless, it does not always happen in practice. So, as seen above, after a period of comparison, users of tensiometers will not use Irritel anymore and vice versa (example of Haut-Rhin). It is then preferable to propose tools that already integrate various approachs.
... and are renewed continually (Isbérie, 1977)
Some kinds of warnings disappear and sometimes reappear in order, it seems, to maintain farmers' interest. Some programmes have been totally abandoned ('Irrigest', 'Irridoc', 'Irri 80').
The possibility for an information service to evolve and to adapt to changing needs must be safeguarded.
Quite often, irrigation is essential for farmers. However, they remain subjected to constraints, such as the 'water rotation' and the material limits. In this case, they cannot choose the irrigation frequency. On the other hand, with some kinds of equipment, they will be able to vary the amount.
Furthermore, there are many other necessities to manage on a farm, and this is not easy for the farmer. Perhaps it would be possible to identify them and to take them into account in the advice concerning the dates and amounts of irrigation for the fields of each farmer, considering the irrigation that is really done. However, at present this is often beyond the tasks and/or abilities of a single service.
The rational management of water in agriculture (and elsewhere) concerns everyone; the adjustment of irrigation equipment and its optimal use should become a collective undertaking.
Bonnemort, C., Bouthier, A, Deumier, J.-M. and Specty, R. 1995. Gérer l'eau avec Irritel: intérêts et limites. (How to manage irrigation with Irritel: advantages and limits). Perspect. Agric. no. 203 (in press).
Isbérie, C. 1977. Avertissement - irrigation et automatisme à la parcelle. (Irrigation warning and automation on the field). Bull. Tech. Inf. pp. 317-318; 127-139.
Peyremorte, P. and Isbérie, C. 1986. Fiches techniques (Technical cards): (1) Les tensiomètres permettent de contrôler l'évolution de l'eau dans le sol, 4 p.; (Tensiometers enable the evolution of soil water status to be controlled); (2) Les tensiomètres permettent de piloter l'irrigation par aspersion, 8 p., (Tensiometers enable sprinkler irrigation scheduling); (3) Les tensiomètres permettent de piloter l'irrigation goutte-à-goutte, 8 p., (Tensiometers enable drip irrigation scheduling). Dossier 'Maîtrisez votre irrigation' (File 'Command your irrigation'), RNEDHA, Montp.
Specty, R. and Martini, F. 1993. Bien gérer l'eau a la parcelle grâce à Irritel (Field water managing with Irritel). First Eur. Conf. Applic. Meteorol., Oxf. 1.