Radiotracer studies on waterborne copper uptake, distribution, and toxicity in rainbow trout and yellow perch: a comparative analysis

Rainbow trout (Oncorhynchus mykiss) are often used to estimate important biotic ligand model (BLM) parameters, such as metal-binding affinity (log K) and capacity (Bmax). However, rainbow trout do not typically occupy metal-contaminated environments, whereas yellow perch (Perca flavescens) are ubiquitous throughout most of North America. This study demonstrates that dynamic processes that regulate Cu uptake at the gill differ between rainbow trout and yellow perch. Rainbow trout were more sensitive to acute aqueous Cu than yellow perch, and toxicity was exacerbated in soft water relative to similar exposures in hard water. Whole body Na loss rate could account for acute Cu toxicity in both species, as opposed to new Cu uptake rate that was not as predictive. Time course experiments using radiolabelled Cu (64Cu) revealed that branchial Cu uptake was rather variable within the first 12 h of exposure, and appeared to be a function of Cu concentration, water hardness, and fish species. After 12 h, new branchial Cu concentrations stabilized in both species, suggesting that metal exposures used to estimate BLM parameters should be increased in duration from 3 h to 12+ h. In rainbow trout, 71% of the new Cu bound to the gill was exchangeable (i.e., able to either enter the fish or be released back to the water), as opposed to only 48% in yellow perch. This suggests that at equal exposure concentrations, proportionally more branchial Cu can be taken up by rainbow trout than yellow perch, which can then go on to confer toxicity. These qualitative differences in branchial Cu handling between the two species emphasize the need to develop BLM parameters for each species of interest, rather than the current practice of extrapolating BLM results derived from rainbow trout (or other laboratory-reared species) to other species. Data reported here indicate that a one-size-fits-all approach to predictive modeling, mostly based on rainbow trout studies, may not suffice for making pred