Acute toxicity of copper to three species of pacific salmon in water with low hardness and low dissolved organic carbon

Abstract

Despite a history of investigation, the toxicity of copper (Cu) to fishes remains difficult to predict due to the complex influential effects of water chemistry. Water hardness and dissolved organic carbon (DOC) concentrations can vary significantly within a watershed and also attenuate the toxicity of Cu to fishes. To account for location-specific water chemistry and predict Cu toxicity to aquatic organisms the United States Environmental Protection Agency (USEPA) endorses use of the biotic ligand model (BLM). Though the BLM has proven useful in many instances, it has performed inaccurately for waters low in hardness and DOC; this has raised questions regarding the model's applicability in certain regions. One such region is Alaska's Bristol Bay watershed, where tributaries low in hardness and DOC support an abundance of Pacific salmon (Oncorhynchus spp.) life. The Bristol Bay watershed also contains one of the largest Cu deposits on earth. Here, to determine empirical lethal Cu concentrations for water conditions relevant to the Bristol Bay watershed, and to assess the accuracy of the BLM in such waters, juvenile sockeye salmon (O. nerka), Chinook salmon (O. tshawytscha), and coho salmon (O. kisutch) were exposed to Cu in low-hardness (5.6-13.7 mg/L) and low-DOC (0.4-2.7 mg/L) water during 96 h flow-through bioassays. Juveniles were used to assess toxicity at a known Cu-sensitive life stage. The experimentally determined acute median lethal concentrations (LC50s) were 35.2 µg Cu/L (95% confidence interval [CI]: 32.3, 38.1) for sockeye salmon, 23.9 µg Cu/L (95% CI: 20.3, 27.4) for Chinook salmon, and 6.3 µg Cu/L (95% CI: 5.6, 7.0) for coho salmon. The BLM consistently under-predicted LC50s; predictions were 62.6 µg Cu/L for sockeye salmon, 35.4 µg Cu/L for Chinook salmon, and 15.1 µg Cu/L for coho salmon. These discrepancies demonstrate that the BLM is inaccurate for waters with low hardness and DOC and that currently assessed levels of risk of Cu to salmon are incorrect. These findings reveal a need for further calibration of the BLM for use in areas like the Bristol Bay watershed and provide information necessary to accurately assess Cu toxicity to three important species of Pacific salmon.