Variability in the toxicity of oil sands process-affected water to rainbow trout ( Oncorhynchus mykiss )

Temporal variability in acute toxicity to rainbow trout was assessed in four oil sands process-affected water (OSPW) samples collected from a single active settling basin over an annual operating cycle. Baseline toxicity tests were conducted on each sample to identify initial levels of toxicity, followed by an investigation of cause(s). The level of toxicity was relatively consistent, varying up to 2-fold between samples (i.e., LC50s ranged from 35 to 71%). There was no indication that metals, total dissolved solids or total suspended solids contributed to toxicity, but toxicity was consistently removed by treatment with non-polar (i.e., C8) solid phase extraction columns and recovered in methanol elutions, indicating the organic fraction was responsible for observed effects. High resolution chemical analyses and toxicity tests conducted on sequential elutions of C8 columns across an increasing methanol gradient demonstrated toxicity was primarily related to the classical naphthenic acids (NAs) when adjusted for potency based on the carbon numbers present (P < 0.0001), suggesting a non-polar narcosis mode of toxicity. Importantly, concentration data for total naphthenic acid fraction compounds (NAFCs) and total classical NAs were not significant indicators of toxicity across the fractions tested, highlighting the importance of the carbon number profile. Good agreement between toxic units for NAs estimated from the fractions relative to the level of toxicity observed in the untreated samples suggested that all significant sources of toxicity were accounted for. Overall, results suggested general consistency in the level and cause of toxicity across the time-frame evaluated, and demonstrate a comprehensive approach for similar investigations that could be applied to other sources and endpoints to support discussions of treatability and environmental risk.