Multicompartment Examination of Micropollutant Partitioning in Replicate Artificial Streams Highlights the Limitations of Assessing Water Matrices Alone

Abstract

While numerous assessments of micropollutant exposure primarily focus on monitoring the water column, a growing body of research indicates that differences in micropollutant partitioning in other compartments require additional consideration for risk evaluation. This study investigated the partitioning of antibiotics, antiepileptics, antibacterials, and antidepressants and their metabolites in water, sediment, macroinvertebrates (gammarids), biofilm, and fish (spoonhead sculpin and longnose dace) found or exposed in replicate naturalized streams (Calgary, Alberta, Canada). All target micropollutants were detected in the water and sediment, and >5 substances were detected in the biotic matrices at concentrations between the limit of quantitation and 244 ± 16 ng/g_dw. Triclosan and triclocarban (antibacterials) were frequently detected in sediments, but very rarely in the water column. The solid–water partitioning (K_d) and organic carbon–water partitioning coefficients (K_oc) indicate that fluoxetine, norfluoxetine, and triclosan have a stronger affinity for sediments and/or organic matter (log K_d > 2.7, log K_oc > 1.5). More specifically, fluoxetine was found to be up to 10× higher in sediments, biofilm, and gammarids than other substances, whereas its concentration in the water column was very low or nondetectable. Finally, bottom-dwelling fish (spoonhead sculpin) were also found to have higher concentrations of fluoxetine and its metabolite than longnose dace.