Critical source areas for herbicides can change location depending on rain events

• We quantify herbicide loss rates from individual fields within a catchment.
• The spatial pattern of herbicide losses to the stream changed between events.
• Characteristics of rain events and land management influenced the spatial patterns.
• Implications for using the critical source area concept to mitigate diffuse pollution.

During rain events, herbicides can be transported from their point of application to surface waters, where they may harm aquatic organisms. Since the spatial variability of herbicide losses to streams can be large, the identification of critical source areas could help to target mitigation measures efficiently to those locations where they reduce herbicide pollution the most. We performed a controlled herbicide application on wheat (isoproturon) and corn fields (atrazine, S-metolachlor and sulcotrione) in an agricultural head-water catchment (about 1 km2) in the Swiss Plateau to investigate the spatial variability of herbicide losses. We performed spatially distributed discharge measurements and high-resolution water sampling during rain events after herbicide application to determine herbicide loads and loss rates for individual fields. The dry weather conditions after the wheat herbicide application resulted in a very low isoproturon loss rate (0.005% of the applied amount). In contrast, the corn herbicide application was followed by several rain events with varying intensities and magnitudes causing loss rates of 0.26, 0.16 and 0.26% for atrazine, S-metolachlor and sulcotrione, respectively. The spatial differences in loss rates between fields were about a factor of three in most events. However, the spatial loss pattern varied between events implying that being a critical source area is not a temporally stable field property. No correlations were observed with several field characteristics (connectivity of the fields to the stream, the tendency for topsoil saturation, field-specific herbicide dissipation rates and sorption affinities). However, the data suggest that critical source areas may depend on the type of rain event because infiltration and saturation-excess runoff affects different parts of the catchment.