Pesticide extraction from soil into runoff under a rainfall simulator

Context Runoff estimation is an important aspect of pesticide environmental behaviour and is the major loss pathway to the environment. Aims To improve understanding of pesticide runoff. Methods Data from three rainfall simulator studies was used. Twelve pesticides were studied ranged from tightly sorbed (DDE, soil sorption coefficient (KD) ~15 000 L kg−1) to weakly sorbed (dimethoate, KD < 30). Key results Event runoff pesticide concentrations were closely related to soil concentrations (0–25 mm depth). The ratio of runoff to soil concentration (the runoff extraction ratio, ERO), was similar for pesticides with a wide range of sorption and across the three soils: runoff concentration (μg L−1) = 28 × soil concentration (mg kg−1). ERO decreased with time after spraying, presumably due to lower concentrations in the top few mm of soil. Conclusions This model provides improved or similar estimates of pesticide runoff than previous models. Similar ERO values between sites was probably due to similar hydrology (high rainfall intensity, surface sealing, moist subsoils) and erosion, and because the same masses of soil and water are involved in mixing. Reduction in runoff concentrations by leaching was not influential, because infiltration was small and soil sorption too high. Implications Conditions studied apply during summer storms on most cotton and grain land on clay soils in the northern grain and cotton lands in eastern Australia. The model should be applicable under these conditions.