Experimental drainage ditches used as mesocosms were treated with linuron three times, at 4‐week intervals, at nominal concentrations of 0, 0.5, 5, 15, and 50 μg/L. After each treatment, the mesocosms were kept static for 1 week. They were then flushed with noncontaminated surface water, with a residence time of 5 d, until the next treatment. Structural and functional measurements were made. A companion paper discusses the structural aspects. Dissolved oxygen concentrations and pH were measured continuously at 15‐min intervals. Both variables declined during the static periods, and a lowest‐observed‐effect concentration (LOEC) of 5 μg/L was noted for both response variables. All mesocosms recovered during the flushing periods. Diurnal oxygen curves were used to calculate ecosystem gross primary production (GPP) and ecosystem oxygen consumption. Ecosystem oxygen consumption was not affected by the linuron treatments, but GPP had an EC50 of 21 μg/L during the static period after the first treatment. The EC50 for GPP increased to 36 μg/L after the second treatment and to 80 μg/L after the third treatment. This increase in EC50 is interpreted as an adaptation of the ecosystem, the mechanism of which could not be identified. The LOEC was 15 μg/L for all treatments and did not change over time. Between treatments, GPP recovered to the level of the controls. The LOEC for the long‐term GPP was 50 μg/L.