An indicator approach for describing the spatial variability of artificial stream networks with regard to herbicide pollution in cultivated watersheds

Abstract : Artificial stream networks play a major role in water and pollutant transfer within cultivated watersheds. Since the characteristics of these artificial stream network (e.g. width, vegetation cover, presence of sediments, etc.) are highly variable in space and in time, describing their variability appears as a prerequisite for assessing environmental risk of pollution at watershed scale. In this perspective an indicator approach is presented which includes three steps: (a) collecting properties of reach (i.e. elementary stream) along an artificial stream networks by means of a field and GIS-based approach, (b) deriving from these properties a set of quantitative indicators of hydrological and chemical processes involved in herbicide transfer, (c) classifying reaches in terms of their role in herbicide transfers on the basis of these indicators. The application of this method to three different watersheds located in France - Roujan watershed (South), La Morcille (East-Centre) and Le Cétrais (North-West) - confirms that artificial stream networks of cultivated watersheds are highly spatially variable in regard to pollutant transfer. This spatial variability includes both variations between the watersheds and variations within the watersheds. Furthermore, the study reveals that the stream network properties can also be variable in time.