Six of these indicators are simple and highly descriptive. They are given rather quickly from data related to agricultural practices. They are: a certain number of crops (NB), of treatments (PHY), a mass of product (MAS), or a period of flowering (FLO) or of a soil cover (CVA, CVT). The five other indicators are composite and more complex and are expressed in the terms of balance or ratio.

These indicators are based mainly on a reading of practices without taking the soils or the local weather conditions into account. In fact, only two indicators take the environment into account: the indicators IRR and MO.

Nine indicators give a raw value which cannot be directly interpreted in this form, without any reference to an objective or to a recommended value.

Only two indicators refer to a local objective (IRR and H/P). In these cases, the aim is reached with a value of 1 for the ratio IRR and a value of 0 for the deviation H/P/.

The values gave us the possibility to compare several rotations for a given indicator, and therefore to classify performance in a particular field. As a matter of fact, it is enough to evaluate the difference between two rotations, and then to see what is the best rotation in terms of durability. But with a lack of reference, they do not give us the possibility to estimate an isolated rotation, particularly for the nine indicators mentioned above: for example, what would be the performance of a rotation with an energizing efficiency of 4 ?

That is why, and also to have a general view and make the performance of a given rotation clearer in the different fields, we chose to transform these values into marks varying between 1 and 9 within a scale of increasing durability.

To compare and classify rotations, these marks allowed us to establish means on all, or some of, these indicators and to give a global definition of each rotation.

On a practical level, the value became a mark between 1 and 9, by a simple classification, according to the relative place of the value between two references (graph 1).

Except IRR and H/P, the marks of indicators depend on chosen values from V1 to V9. They correspond to extreme values encountered in the variation of studied rotations for all 7 different areas in 2001. Therefore it is more here a relative classification than a real estimation.

Example of rotations in a region:

This evaluation was carried out to compare rotations in the framework of an agricultural area. Each of the indicators was estimated within a region on the level of the rotation. For each studied rotation, we described the succession of crops, the management of each crop from the preceding crop to its harvest, as well as the production results obtained. Generally, we described only one crop system by rotation: we reduced the diversity of crop managements for each crop to only one type-management, except for winter wheat for which we modified the crop management and the results following two preceding crops (wheat/wheat, wheat following another crop).

The presented region is “Champagne berrichonne” in the middle of France, where oilseed crops are associated to winter cereals in “dry” systems (without irrigation).

We present here a system with a sample of five different rotations. The table below indicates the marks given by rotation for each indicator (scale from 1 to 9).

A first analysis shows that certain indicators mark clearly the difference between these five rotations. For example: the number of phytosanitary treatments, the period of melliferous flowering, the soil covers and the nitrogen balance. On the other hand, these rotations obtain similar marks when the energizing efficiency and the humic balance are concerned.

The comparison of rotations indicates that the average mark is generally better with the length and diversity of the rotation. The shortest rotations tend to be particularly handicapped by their nitrogen balance, the mass of phytosanitary products and of course the criterion of the number of crops. The long and diversified rotations tend to offer certain advantages, concerning essentially the number of phytosanitary treatments, the nitrogen balance, the period of flowering and the balanced proportion between winter and spring crops.

An analysis of the part played by oilseeds in the rotations

A two-by-two comparison of these rotations allows us to understand better how the replacement of a crop by another crop modifies the performance of each of these indicators. The detailed study of the introduction of oilseed crops in cereal rotations gave us the possibility to identify a certain number of interesting points, but also of tricky questions.

In the field of the utilization of resources, the energizing efficiency of the agricultural production is equivalent for the rotations with winter crops with a nitrogen fertilization higher than 150 units. It improved a little with the introduction of a crop like sunflower, for which the supplies of nitrogenous fertilizers are reduced to less than 100 units.

In the field of the quality of waters (item: phytosanitary products), the introduction of sunflower in a cereal rotation gave us the possibility to reduce the number of phytosanitary treatments as well as the mass of applied active matter. Considering the present behaviour of present varieties towards their tolerance to diseases, sunflower is a crop which requires little in matter of fungicides and insecticides. Hoeing being often applied to sunflower, the use of herbicides is then reduced. But the introduction of rapeseed penalizes cereal rotations from the point of view of these indicators ; in fact, even if crop diversification allows to reduce the use of pesticides on cereals, it does not compensate for additional supplies given to rapeseed. It can explained by the use of rather old herbicides used in high quantities, several applications of insecticides, great applications of anti-slug products, rather systematic applications of phytosanitary products and little developed systematic control.

In the field of the quality of waters (item: nitrogen), the introduction of sunflower allowed us to reduce the nitrogen balance of the rotation thanks to low nitrogen needs of this crop. But the short cycle period of this crop leads to an uncovered soil in the autumn between the harvest of the preceding crop and sunflower sowing. Rapeseed and its volunteer plants following harvest cover the soil over two successive autumns, play the part of nitrate trap and also give the possibility to valorize farm-fertilizers applied in the summer. But they lead to nitrogen balances in the rotation which are less favourable than sunflower. The use of intermediate crops in these rotations could improve the soil cover in autumn without modifying the nitrogen balance.

In the field of the quality of the air and of the soils, the introduction of rapeseed in a cereal rotation seems to improve the humic balance of the soil. The nitrogen storage in the soil would be reduced in rotations with sunflower. Considering the similar evolution of the indicator CVT according to the rotations, erosion risks seem to be low in rotations based on winter crops, more particularly in rotations with rapeseed. Sowing without ploughing being a characteristic of this crop most of the time, erosion risks are also decreased. But these risks are higher with the introduction of a summer crop like sunflower. As a matter of fact, this crop is penalized by its short cycle and also by its wide spacing sowing.

In the field of the biodiversity, landscapes bearing winter cereals and winter rapeseed offfer the advantage to give a crop cover which will last over the year for a long period, but which is totally absent at the end of the summer. The introduction of crops like sunflower gives a certain balance between summer and winter crops, as well as plots which are covered in the summertime. Rapeseed and sunflower are an important resource for bees, while contributing to the beekeeping production (indicator FLO).

In the field of the landscapes, long and diversifed rotations lead to varied rotation systems where the crops are as many variables in the landscape. Oilseeds contribute to give a nice yellow color at the time of flowering, rapeseed early in the spring, followed by sunflower early in the summer for the example of France.

Discussion

To sum up this study, 11 indicators were used. Is this number enough to deal with the main aspects of durable development ? Is it necessary considering the priority items for the concerned agricultural region ? These were inevitable questions. Thus, to start with, the number of indicators was more reduced, but with all the problems concerning the quality of waters and nitrogen, we had to have another indicator. Now for the setting up in each of the regions, the 11 indicators were not kept for the final classification of rotations. In spite of the problems linked to its utilization [8], we used the average to compare the rotations globally. But the priority was given to the 5 indicators which corresponded to the main stakes in a given region, and we carried out a weighted average to stress out the most important indicators for the durable development in the region.

We also thought of using more complex indicators, which would have been more difficult to reckon but more representative of the durability stakes. For example, to the nitrate losses in a crop system, it is clear that an indicator including the soil cover and nitrogen balance would be more performing from this point of view. But the absence of indicator for the soil cover in autumn would be difficult for other items, such as erosion or the run-off water pollution with phytosanitary products.

In order to conform to initial tender specifications, the indicators had to be easy to compute and meaningful for farmers. That is why we were led to take environmental variables (except for IRR and MO) rather rarely, and never the properties of active matter for the indicators MAS and PHY. We have then to deal with a probably reduced readability of consequences to be expected from practical changes on performance of these systems, like the scientific validity. Choosing indicators doesn't avoid compromises [9].

These indicators are often simple describers of practices, and only nine of them are not based on a reference value, which would be meaningful and valid from a scientific point of view. Therefore, we cannot evaluate an isolated rotation but can only make comparative evaluations. Lastly, these comparative evaluations offer a lack linked to their relativity: they can lead to a certain perfectionism if the rotation under study is already highly performing, and contrarily, they can lead to a certain kind of easy going.

These limits raised no major problem to define the rotational measure since the question dealt with the comparison between long and diversified rotations in comparison with monocultures with very short rotations. In order to widen the possible utilization of these indicators to estimate rotation performance, it is now desirable to improve some of them in order to help users to take their decisions.

RÉFÉRENCES

1. Pouzet A. Systèmes de grandes cultures et environnement. Situation et perspectives d'évolution. CETIOM 2001 ; 7 p ; + annexes.

2. Reau R, Wagner D, Pilorgé E. Première évaluation environnementale des systèmes de cultures avec oléagineux. Oleoscope, CETIOM, Paris 2002 : pp 24-26.

3. Girardin P, Bockstaller C. Les indicateurs agro-écologiques, outils pour évaluer des systèmes de culture. OCL, 4, n^o6, 1997 : pp 419-426.

4. Halberg N. Indicators of resource use and environmental impact for use in a decision aid for Danish livestock farmers. Agriculture, Ecosystems and Environment 1999 ; 76 : pp 17-30.

5. Lavoux T, Trocherie F, 2001. Les indicateurs agro-environnementaux en agriculture. AGPM, Commissqion technique du maïs, Paris, 15 mars 2001. 12 p.

6. Viaux Ph. Une troisième voie en grande culture. Ed. Agridécisions, Paris, 1999 : 211 p.

7. Vilain L. La méthode IDEA, indicateurs de durabilité des exploitations agriciles. Guide d'utilisation. Educagri éditions, Ministère de l'agriculture, Bergerie nationale de Rambouillet 2000 : 100 p.

8. Bockstaller C, Girardin P. Some methological issues in the construction of environmental indicators. Europena Society of Agronomy, VII congress, Cordoue. Book of proceedings, 2002 ; pp 551-552.

9. Girardin P, Bockstaller C. Indicators : tools to evaluate the environmental impacts for farming systems. Journal of sustainable agriculture 1999 ; 13(4) : pp 5-21.

10. Maurizi B, Verrel J-L. Des indicateurs pour des actions de maîtrise des pollutions agricoles. Ingénieries 2002 ; 30 : pp 3-14.