**2.4 Results**

236 Soil Erosion Studies

Fig. 5. Procedure for characterizing the agricultural practices in Calvados

structural stability carried out on aggregates (Le Bissonnais and Le Souder, 1995).

During two years, we carried out nearly 8000 soil boreholes, which represent one borehole per 40 ha of agricultural land. Pedogenesis, soil thickness, coarse fragments, texture and hydromorphy have been registered. Data and Progressive knowledge of the soil landscape allowed us to produce a first soil map on which we selected 150 representative soil boreholes. Physical (granulometry, structural instability of surface horizon), hydrologic (AWC taking into account per cent of coarse fragments) and chemical properties have been determined. Structural instability has been evaluated starting from the INRA test of

Consequently, the soil features do not come from the application of the pedotransfer rules. This analytical step has been led us to finalize a global soil map on a 1 : 50 000 scale, to suggest at the same scale a map of soil structural instability, to propose another map in connection with spatial distribution of available water content of the soil, and, finally, combining the latter with interpolated rainfall data, to define the yearly positive hydrological balance. Classes of structural instability of the soil aggregates come from a Combination of potential sensitivity of areas to erosion and rainfall erosivity in a "normal" climatic context leads to evaluation and mapping of mean erosion hazard (Fig. 6a). This document highlights of the existence of all levels of hazard. Level 0 shows parcels promoted with permanent grassland or orchards and/or with a slope lower than 1%. These parcels represent 1600 km² of agricultural surface (42 %). They are localized by the form of coherent spatial units in the north-west and center-east of the territory.

Fig. 6. Soil erosion hazard in Calvados at parcel scale (A) for a normal climatic contex, (B) for a rainy year

Levels 1 and 2 of soil erosion hazard cover more than 500 km² of agricultural surface (13 %). It is located in majority in the central part of the Calvados along a North/South axis. Level 3 is

SCALES: An Original Model to Diagnose Soil Erosion Hazard

and Assess the Impact of Climate Change on Its Evolution 239

Fig. 7. Aggregation of the soil erosion hazard data: (A) at the administrative scale

parcel scale to the administrative scale due to aggregation procedure

**3.1.1 Plant covering** 

periods, in Normandy.

(municipality); (B) at the hydrologic scale; (C) significant deterioration of results from the

The plant covering on cultivated areas can be estimated by the foliar surface of plants. In French regions with temperate climate, it is very difficult to get reliable data about monthly rates of plant covering and their intra-annual evolution. The data that we show (Fig. 8) have been sent by the Technical Institute for Plants specialized in the agronomic research (ARVALIS – Institut Technique du Végétal), a Calvados-based and nationally recognized organism. These data concern the principal plants cultivated annually or in intercrop

Fig. 8 enables to show on a monthly scale three types of evolution for the plant covering by plants in crop areas. A first type gathers the plants characterized by a poor rate of plant covering at the end of the first month of growth and then by a very quick and totally covering rate at the end of the second month. This type concerns maize silage and ray grass. The following type refers to rapeseed and mustard. The rate of plant covering is very

the most represented since it covers 1400 km² (37%) of agricultural surface. It is located southwest of the department and locally in the East. The fourth level is the only one representing important hazard since the level 5 is absent. The affected parcels cover a surface of 100 km² (2,6%). They are essentially gathered at the extreme south-west of Calvados.

The assessment of the soil erosion hazard for a rainy year (2001, + 15% compared to the "normal" 1970-2000) has been achieved in order to study the potential impact of the rainfall conditions supporting high erosivity (Fig. 6b). Results are particularly interesting because they show that agricultural surfaces affected by a level 4 (strong hazard) extended more than 800 km² compared to a year with "normal" climatic conditions. The most affected areas are South-West, the central North/South axis and secondarly the eastern part of the territory. This brings us to considerate that Calvados is a department presenting a strong predisposition for the genesis of soil erosion by water, erosion which express itself as soon as climatic conditions induce strong erosivity.

Finer representation with zoom effect of levels of hazard at the scale of parcel unit reveals the occurrence of a mosaic of colors expressing very frequent spatial disjunction at local scale about erosion hazard (Fig. 7C). The rapid and brutal spatial variations of physic properties of the area associated with the interpenetration of grassland and crops parcels contribute to the strong heterogeneity of the results at local scale. The precision of input data of the model SCALES allows to obtain this type of conclusion. It also comes to the idea that the management of this issue assumes in priority an approach at the scale of a parcel or a group of parcels.

The aggregation of soil erosion hazard data at administrative and hydrologic scales (Fig. 7A-B) shows a significant loss of information when the basic scale is given up (Fig. 7C). SCALES model loses quickly its interest but can become a communication tool about the question of soil erosion.
