The levels and distribution of diquat, paraquat, and difenzoquat were determined by solid phase extraction (SPE) and high-performance liquid chromatography (HPLC) in water samples from irrigation channels, rivers, and lagoons taken during 1 year from three different marsh areas of the Valencian community. These areas are representative of the typical Mediterranean coastal ecosystems. All three compounds were detected. Diquat was found most frequently at all the sampling sites. Although the spatial distribution of diquat and paraquat showed a maximum concentration near the fields where they were originally applied, their location fluctuated due to irregular large spill and/or loading. The herbicide concentration tended to be highest during the summer (June, July, and August) because these are the months with the least rainfall and highest evaporation rates, when weeds grow best and pesticides are needed more often. The average concentration found for diquat was 0.09 &mgr;g/L, with a maximum of 3.10 &mgr;g/L. The average concentration for paraquat was 0.01 &mgr;g/L, with a maximum of 3.95 &mgr;g/L. Samples that were below the method detection limit are included in the mean calculation as zero. Difenzoquat was only detected in one sample at a concentration of 1.75 &mgr;g/L.

No abstract

No abstract

No abstract

Key WordsColumn liquid chromatography Solid phase extraction Graphitized carbon black Diquat, paraquat and difenzoquat SummarySelective on-line solid phase extraction (SPE) and liquid chromatography determination (HPLC) of diquat, paraquat and difenzoquat from environmental water samples has been accomplished with Graphitized Carbon Black (GCB) as both extraction and analytical columns. The method involved passing of 50 mL of water through a cartridge filled with Carbograph. In the elution step, the herbicides were transferred from the cartridge to the analytical column (Hypercarb) by mean of a gradient of pH 3 aqueous solution of tetramethylammonium hydroxide (TMAOH) and ammonium sulphate and methanol. Hypercarb columns were found to give a low probability of false positives for bypiridylium herbicides and are very selective for polar compounds. Recovery was better than 80 %. The breakthrough volume was studied with distilled water spiked with the herbicides at various concentration levels (from 0.1 to 20 ktg L-I). The limits of quantification of the method were lower than 0.1 ktg L -1. The total analytical method was applied to surface waters from Torreblanca Nature Park (CasteU6, Spain).