Use of the biotic ligand model to predict pulse-exposure toxicity of copper to fathead minnows ( Pimephales promelas)

Abstract

Concentrations of cationic metals (e.g., Ag, Cd, Cu, Ni, Pb, Zn) and other water quality parameters (e.g., pH, alkalinity, hardness, dissolved organic carbon (DOC) concentration) often cycle daily in surface waters, and the toxicity of the metals to aquatic organisms is altered by variations in those water quality parameters. Consequently, a method is needed to predict the LC50s (median lethal concentrations) of dissolved metals in temporally varying water quality. In this study, we combined the biotic ligand model (BLM), which predicts toxicity of cationic metals across a wide range of water quality conditions, with a one-compartment uptake-depuration (OCUD) model, which predicts toxicity of a chemical at any exposure time in either continuous or time-variable exposures, to test whether we could accurately predict pulse-exposure toxicity of Cu to fathead minnow (FHM; Pimephales promelas) larvae. First, we conducted continuous-exposure toxicity tests to calculate 1- to 96-h Cu LC50s for the FHM larvae. Then we re-parameterized the default Cu BLM for FHM until the corresponding predicted Cu LA50s (medial lethal accumulations at the biotic ligand) collapsed together into a narrow band and also fit the generalized pattern of an OCUD model [i.e., a steeply sloping plot of ln(LA50) versus ln(time) at short exposure times, followed by a gradual approach to an incipient lethal level at longer exposure times]. Next, in 72-h tests, we exposed FHM larvae to 2- or 8-h square-wave pulses of elevated Cu concentration followed by recovery in uncontaminated water for the remaining 22 or 16 h in each of three consecutive 24-h pulse-and-recovery cycles, at pH 6 or 7 in water containing either 0.5 or 2 mEq/L hardness and 0 or 20 mg DOC/L. Using the combined BLM-OCUD model developed from continuous-exposure data, we then predicted the Cu LA50s in the pulse-exposure tests and compared those LA50s to the observed pulse-exposure Cu LA50s. Although predicted pulse-exposure LA50s were within ∼4× of the observed pulse-exposure LA50s, delayed deaths during the recovery phases of the exposures precluded more accurate predictions of pulse-exposure Cu LA50s and, as a consequence, of pulse-exposure dissolved Cu LC50s. We conclude that one global OCUD equation linked to a re-parameterized Cu BLM for FHM can be used to predict the acute toxicity of continuous and pulse exposures of Cu to FHM larvae across a range of water quality conditions; but to improve the accuracy of those predictions, a mechanism must be developed to account for delayed deaths.