Solute transfer in agricultural catchments: The interest and limits of mixing models

This paper deals with the applicability of the endmember mixing analysis (EMMA) to agricultural catchments. The basic assumptions of the method are discussed in the case of semi-natural and rural ecosystems. EMMA was applied to two storm events in a farmland catchment in Brittany (France) using four endmembers and three tracers. The results were validated using additional tracers. The sensitivity of the method to uncertainty in the endmember composition was analysed using a Monte-Carlo procedure. The results were satisfactory although the method is very sensitive when the endmember composition is close to the streamwater chemistry. It is concluded that mixing analysis is a valuable tool for studying water pathways in agricultural areas, provided it is applied over short periods and used with other hydrological approaches.