Effects of several variables on whole effluent toxicity test performance and interpretation

Protocol changes and options contained within U.S. Environmental Protection Agency whole effluent toxicity tests represent variables that have the potential to affect bioassay performance and interpretation of results. Variables evaluated in this study include: the change in allowable age in the Pimephales promelas acute bioassay from up to 90 d to a maximum of 14 d, age‐specific acute responses of P. promelas among the allowable ages of 1 to 14 d, change in the chronic growth endpoint definition from final mass to biomass, differences between hemacytometer and fluorometer measurements in the Selenastrum capricornutum protocol, and options for statistical interpretation of species sensitivity in multiple test/species screening bioassays. Clear age‐related sensitivity and precision differences were observed in acute responses of P. promelas. Results obtained using the younger age classes were typically more variable in studies of both 1‐ to 14‐d‐old and 14‐ to 90‐d‐old P. promelas. In the experiments on 1‐to 14‐d‐old organisms, larvae at 1 d of age were significantly less sensitive. In the tests on 14‐ to 90‐d‐old organisms, the 14‐d‐old organisms were significantly less sensitive. The change in endpoint definition in the P. promelas chronic bioassay resulted in an apparent increase in toxic response in the inhibition concentration (ICp) value for each bioassay, evaluated by the biomass method, with no general improvement in statistical intertest precision estimates and no predictable impact on the no‐observed‐effect concentration endpoint. Fluorometric scoring in the Selenastrum bioassay was significantly more precise and better capable of estimating counts than hemacytometer measurements. Discrepancies associated with commonly used statistical endpoints used to determine the most sensitive species were identified, and potential solutions were proposed.