Comparison of two modeling frameworks for unifying the toxicant-affected percentage of aquatic species, individuals, and time into a single metric.

To address time-variable exposure to toxicants, this work compares simple and complex approaches to unifying the affected percentage of aquatic species, individuals, and time into a single metric. The simple approach uses only information on the probability distribution of exposure concentrations, a species sensitivity distribution (SSD) of chronic values, and the distribution of tolerance within species. The complex approach involves time-series simulation with a kinetics-based toxicity model coupled with a population model for each species in the SSD. Unlike the simple approach, this takes into account the exposure duration needed to elicit toxicity, differing sensitivities of life stages within a species, differing effects on survival versus reproduction, and species differences in their model population's response to press disturbance and recovery time from pulse disturbance. The probability distribution approach indicated that, for SSD assemblages challenged with moderately variable toxicant concentrations exceeding the aquatic life criterion a few percent of the time, most of the predicted aggregate effect is usually experienced by the most sensitive 10% of species and individuals. The kinetics-population simulation approach indicated that, for time-variable exposure (but not for constant exposure), the severity of the population effect depended on the type of effect and life stage affected. The results from both approaches suggest that moderately time-variable exposure is best viewed as a fluctuating press disturbance, not as a sporadic pulse disturbance. Integr Environ Assess Manag 2022;18:1364-1374. © 2021 SETAC.