Investigating micropollutant partitioning in five environmental and biological matrices collected in replicate artificial streams

Abstract

The presence of micropollutants (substances occurring in sub-ng/L concentrations) is a growing concern due to their potential risks to both the ecosystem and human health. Although they can be introduced into the environment via point (wastewater treatment plants) and non-point sources (urban and agriculture runoff), a lack of comprehensive regulation has overlooked their potential persistence, mobility, and adverse impacts on aquatic ecosystems. The main objective of this thesis was to investigate the partitioning of a diverse group of micropollutants, including pharmaceuticals and personal care products in the following environmental compartments: (1) water, (2) sediments, (3) invertebrates, (4) biofilm, and (5) fish. Then, the impact of more advanced levels of wastewater treatment (i.e., ultrafiltration, ozonation, reverse osmosis) on the occurrence and partitioning of these compounds was evaluated. Sampling campaigns were carried out at the Advancing Canadian Water Assets (ACWA) in Calgary, Canada as this facility is equipped with 12 naturalized artificial streams (320 m long) that receive 95% Bow River water and 5% effluent v/v from an operational municipal (Pine Creek) wastewater treatment plant (WWTP) and two pilot WWTPs (reverse osmosis and ozonation). This thesis first focused on improving sample preparation methods in these complex environmental matrices so defensible analytical data (via liquid chromatography, triple quadrupole mass spectrometry) on trace concentrations can be obtained. After evaluating different sampling preparation techniques for the solid matrices, the QuEChERS method (Quick, Easy, Cheap, Rugged, and Safe) for sediment, biofilm, invertebrate (Gammaridae spp), and fish (longnose dace [Rhinichthys cataractae] and spoonhead sculpin [Cottus ricei]) tissues were found to be an appropriate sample extraction method with analytical recoveries from 70% to 120% for most of the compounds analyzed. Overall, the compounds that were frequently detected at high concentrations in all the matrices include analgesics (diclofenac), antibiotics (sulfamethoxazole), antiepileptics (carbamazepine), and antidepressants (venlafaxine). Furthermore, 18 of the 22 compounds were detected in the water matrix, and <10 compounds were detected in the solid matrices (sediment, biofilm, fish, gammarids). High concentrations were observed in the water matrix for diclofenac, venlafaxine, O-desmethylvenlafaxine (venlafaxine metabolite), and carbamazepine at 162 ± 3 ng/L, 381 ± 28 ng/L, 149 ± 3 ng/L, and 45 ± 1 ng/L, respectively. Concentrations in the streams as well as the seasonal trends observed were linked to the Bow River conditions given that it represents a large portion of the stream volume. The Bow River near the ACWA facility has already accumulated micropollutants as a result of WWTP discharges from two Calgary WWTPs that service ~75% of the population. It was also clear that the streams receiving effluent from the Pine Creek WWTP had higher levels of micropollutants compared to effluents that underwent ultrafiltration, reverse osmosis, and ozonation. Finally, an increase in effluent contribution (5% to 15%) was also reflected in the streams and was more detectable in solid matrices. The concentrations ranged from below the limits of quantification (2000 L/kg in fish, gammarids, and biofilm). For sediment, the solid-water distribution coefficient (Kd) for fluoxetine and triclosan have the highest values (>4000 L/kg), indicating that these compounds tend to sorb more into the solids as they were non-detected or present at low concentrations in the streams. Overall, the results suggest that certain micropollutants partition in the solid matrices more and monitoring of the water alone can underestimate the overall pollution levels and potential risks.