MODELING TOXICITY DUE TO INTERMITTENT EXPOSURE OF RAINBOW TROUT AND COMMON SHINERS TO MONOCHLORAMINE

摘要

We evaluated the ability of three mathematical models to predict toxicity to common shiners and rainbow trout during intermittent (pulsed) exposures to monochloramine, based on data from continuous-exposure toxicity tests. If a power term for the exposure-water concentration was included in the models, a concentration x time (Cxt) model and the Mancini uptake-depuration model predicted pulse LC50s to within ±50% of the observed pulse LC50s, for the first four pulses in toxicity tests with 2-h pulse/22-h recovery cycles. Beyond the fourth pulse cycle, though, the pulse LC50s predicted using the Cxt model appeared to diverge considerably from the trend of the experimental pulse LC50s, partly because this model does not predict an incipient lethal level (C_(ILL)) for either continuous or intermittent exposures. The Mancini model predicted the C_(ILL) moderately well in the common shiner intermittent-exposure test but not in the rainbow trout intermittent-exposure test. The Breck three-dimensional damage-repair model did not predict pulse LC50 or C_(ILL) values as well as did the other two models, probably because not enough partial-mortality data were available to parameterize the model adequately. Although the underlying processes appear to be more complex than what these simple models assume, the models may still be adequate for use in regulating a few pulse discharges of monochloramine.