Catchment sourcing urban pesticide pollution using constructed wetlands in Melbourne, Australia

A survey of 111 urban constructed stormwater wetlands (median watershed area= 86.8 ha) was conducted to identify the major pesticides present and to determine their major catchment sources (residential, industrial, commercial, sporting ovals) and associations with catchment imperviousness. Melbourne, Australia, has separate stormwater and sewerage systems and these wetlands are designed to treat urban stormwater. To maximise the pesticides that could be detected, three types of passive samplers (POCIS, Chemcatcher (R) SDB-XC and Chemcatcher (R) C18) were deployed, along with collection of fine sediments. A total of 231 pesticides were screened using these methods. Pesticides that were detected in >5 % of wetlands were checked to determine their registered use in urban areas using an Australian government database (PubCris). Twenty-five pesticides were detected in >5 % of wetlands: 4 pesticides were associated with non-urban land uses (agriculture and forests), another 4 pesticides had no known registered use in urban areas and 17 were associated with urban areas. The pesticides associated with urban areas were the herbicides simazine, diuron, metolachlor, bromacil, propyzamide and paclobutrazol, the fungicides tebuconazole, propiconazole, metalaxyl, trifloxystrobin, iprodione and carbendazim and the insecticides fipronil, bifenthrin, chlorantraniliprole, thiamethoxam and permethrin. Atrazine was also detected in 59 % of wetlands but has not been registered for urban uses in Australia since 2010. It's presence in Melbourne may be due to legacy issues or aerial transportation from rural areas where it's still widely used in crop cultivation. Generally, the major urban catchment source of pesticides is from residential areas (particularly fipronil and simazine), most likely in wood preservatives, paints and fromweed or insect control. Many of these widely used pesticides were correlated with increased catchment imperviousness. Some pesticides (bromacil and imidacloprid) were correlated with commercial premises and chlorantraniliprole was correlated with the presence of sporting ovals in the catchment. No pesticides were specifically correlated with industrial areas. The use of passive samplers and fine sediments, in conjunction with detailed land use mapping of stormwater wetland catchments is very effective and efficient in monitoring and sourcing pesticide contamination in urban environments.